EPA/340/1-91/012
States Air and Radiation EPA340/1-91-012
imental Protection (EN-341W) September 1991
EPA Reference Methods 24 and 24A
Compilation of Procedures/References
Printed on Recycled Paper
-------
EPA-340/1-91-012
EPA REFERENCE METHODS 24 AND 24A
COMPILATION OF
PROCEDURES/REFERENCES
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Stationary Source Compliance Division
Washington, DC 20460
September 1991
-------
CONTENTS
Section Page
1. INTRODUCTION 1
2. EPA REFERENCE METHODS 24 AND 24A, 40 CFR 60,
APPENDIX A 3
3. ASTM PROCEDURES APPLICABLE TO RM 24/24A 11
4. CITATIONS OF RM 24/24A IN 40 CFR 60 195
5. BIBLIOGRAPHY 199
Page i
-------
SECTION 1
INTRODUCTION
Documented procedures for use with coating and ink sample analysis by
Reference Methods 24 or 24A (RM 24/24A) as found in the 7-1-90 edition of 40 CFR
60, Appendix A are presented in this compilation report. In addition, procedures to
be used during collection of samples are provided for reference purposes along with
guidance information on related technical issues.
Section 2 of this report contains EPA Reference Methods 24 and 24A from 40
CFR 60, Appendix A, Section 3 contains a compilation of ASTM Procedures
pertaining to the application of sampling and analytical methods relevant to the use
of RM 24/24A, Section 4 contains a list of citations of RM 24/24A in the Standards
of Performance for New Stationary Sources (NSPS), and Section 5 contains a biblio-
graphy.
EPA is currently in the process of amending RM 24 to make it applicable to
multi-component coatings and exempt solvents, therefore, procedures to be used
for these special cases are also presented in this report.
Page 1
-------
Page 2
-------
SECTION 2
REFERENCE METHODS 24 AND 24A, 40 CFR 60, APPENDIX A
(7-1-90 Edition)
REFERENCE METHOD 24-DETERMINATION OF VOLATILE MATTER CONTENT,
WATER CONTENT, DENSITY, VOLUME SOLIDS, AND WEIGHT SOLIDS OF SURFACE
COATING
1. Applicability and Principle
1.1 Applicability. This method applies to the determination of volatile matter
content, water content, density, volume solids, and weight solids of paint,
varnish, lacquer, or related surface coatings.
1.2 Principle. Standard methods are used to determine the volatile matter
content, water content, density, volume solids, and weight solids of paint,
varnish, lacquer, or related surface coatings.
2. Applicable Standard Methods
Use the apparatus, reagents, and procedures specified in the standard
methods below:
2.1 ASTM D1475-60 (Reapproved 1980), Standard Test Method for Density
of Paint, Varnish, Lacquer, and Related Products (incorporated by
reference-see §60.17).
2.2 ASTM D2369-81, Standard Test Method for Volatile Content of Coatings
(incorporated by reference-see §60.17).
2.3 ASTM D3792-79, Standard Test Method for Water Content of Water-
Reducible Paints by Direct Injection into a Gas Chromatograph
(incorporated by reference-see §60.17).
2.4 ASTM D4017-81, Standard Test Method for Water in Paints and Paint
Materials by the Karl Fischer Titration Method (incorporated by reference-
-see §60.17).
3. Procedure
3.1 Volatile Matter Content. Use the procedure in ASTM D2369-81
(incorporated by reference-see §60.17) to determine the volatile matter
Pages
-------
content (may include water) of the coating. Record the following
information:
W1 = Weight of dish and sample before heating, g.
W2 = Weight of dish and sample after heating, g.
W3 = Sample weight, g.
Run analyses in pairs (duplicate sets) for each coating until the criterion
in Section 4.3 is met. Calculate the weight fraction of the volatile matter
(Wv) for each analysis as follows:
. 24-1
Record the arithmetic average
3.2 Water Content. For waterborne (water reducible) coatings only, deter-
mine the weight fraction of water (WJ using either "Standard Content
Method Test for Water of Water-Reducible Paints by Direct Injection into
a Gas Chromatograph" or "Standard Test Method for Water in Paint and
Paint Materials by Karl Fischer Method." (These two methods are
incorporated by reference - see §60.17.) A waterborne coating is any
coating which contains more than 5 percent water by weight in its volatile
fraction. Run duplicate sets of determinations until the criterion in Sec-
tion 4.3 is met.
Record the arithmetic average (WJ.
3.3 Coating Density. Determine the density (Dc kg/liter) of the surface
coating using the procedure in ASTM D1475-60 (Reapproved 1980)
(incorporated by reference - see §60.17). Run duplicate sets of
determinations for each coating until the criterion in Section 4.3 is met.
Record the arithmetic average (DJ.
3.4 Solids Content. Determine the volume fraction (Vs) solids of the coating
by calculation using the manufacturer's formulation.
Page 4
-------
4. Data Validation Procedures
4.1
4.2
Summary. The variety of coatings that may be subject to analysis makes
it necessary to verify the ability of the analyst and the analytical proce-
dures to obtain reproducible results for the coatings tested. This is done
by running duplicate analyses on each sample tested and comparing
results with the within-laboratory precision statements for each parame-
ter. Because of the inherent increased imprecision in the determination
of the VOC content of waterborne coatings as the weight percent water
increases, measured parameters for waterborne coatings are modified by
the appropriate confidence limits based on between-laboratory precision
statements.
Analytical Precision Statements. The within-laboratory and between-
laboratory precision statements are given below:
Volatile Matter Content,
wv
Water Content, Ww
Density, Dc
Within-laboratory
1.5%W7
2.9% W7
0.001 kg/liter
Between-laboratory
4.7 % W7
7.5% W7
0.002 kg/liter
4.3
4.4
Sample Analysis Criteria. For Wv and Ww, run duplicate analyses until the
difference between the two values in a set is less than or equal to the
within-laboratory precision statement for that parameter. For Dc run
duplicate analyses until each value in a set deviates from the mean of the
set by no more than the within-laboratory precision statement. If after
several attempts it is concluded that the ASTM procedures cannot be
used for the specific coating with the established within-laboratory
precision, the Administrator will assume responsibility for providing the
necessary procedures for revising the method or precision statements
upon written request to: Director, Emission Standards and Engineering
Division, (MD-13) Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC 27711.
Confidence Limit Calculations for Waterborne Coatings. Based on the
between-laboratory precision statements, calculate confidence limits for
waterborne coatings as follows:
To calculate the lower confidence limit, subtract the appropriate between-
laboratory precision value from the measured mean value for that
Page 5
-------
parameter. To calculate the upper confidence limit, add the appropriate
between-laboratory precision value to the measured mean value for that
parameter. For Wv and De> use the lower confidence limits, and for Ww>
use the upper confidence limit. Because Vg is calculated, there is no
adjustment for the parameter.
5. Calculations
5.1 Nonaquepus Volatile Matter.
5.1.1 Solvent-borne Coatings.
W=WV Eq. 24-2
Where:
W0 = Weight fraction nonaqueous volatile matter, g/g.
5.1.2 Waterborne Coatings.
W0-WV-WW Eq. 24-3
5.2 Weight Fraction Solids
W,^-WV Eq. 24-4
Where:
W8 = Weight solids, g/g.
Pages
-------
REFERENCE METHOD 24A-DETERMINATION OF VOLATILE MATTER CONTENT
AND DENSITY OF PRINTING INKS AND RELATED COATINGS
1. Applicability and Principle
1.1 Applicability. This method applies to the determination of the volatile
organic compound (VOC) content and density of solvent-borne (solvent
reducible) printing inks or related coatings.
1.2 Principle. Separate procedures are used to determine the VOC weight
fraction and density of the coating and the density of the solvent in the
coating. The VOC weight fraction is determined by measuring the weight
loss of a known sample quantity which has been heated for a specified
length of time at a specified temperature. The density of both the coat-
ing and solvent are measured by a standard procedure. From this
information the VOC volume fraction is calculated.
2. Procedure
2.1 Weight Fraction VOC.
2.1.1 Apparatus.
2.1.1.1 Weighing Dishes. Aluminum foil, 58 mm in
diameter by 18 mm high, with a flat bottom. There
must be at least three weighing dishes per sample.
2.1.1.2 Disposable Syringe. 5ml.
2.1.1.3 Analytical Balance. To measure to within 0.1 mg.
2.1.1.4 Oven. Vacuum oven capable of maintaining a
temperature of 120 ±2°C and an absolute
pressure of 510 ±51 mm Hg for 4 hours. Alterna-
tively, a forced draft oven capable of maintaining a
temperature of 120 ±2°C for 24 hours.
2.1.2 Analysis. Shake or mix the sample thoroughly to assure that
all the solids are completely suspended. Label and weigh to
the nearest 0.1 mg a weighing dish and record this weight
Page?
-------
Remove a sample of the coating using a 5-ml syringe without
a needle. Weigh the syringe and sample to the nearest 0.1
mg and record this weight (McY1). Transfer 1 to 3 g of the
sample to the tared weighing dish. Reweigh the syringe and
sample to the nearest 0.1 mg and record this weight (MCY2).
Heat the weighing dish and sample in a vacuum oven at an
absolute pressure of 510 ±51 mm Hg and a temperature of
120 ±2°C for 4 hours. Alternatively, heat the weighing dish
and sample in a forced draft oven at a temperature of 120
±2°C for 24 hours. After the weighing dish has cooled,
reweigh it to the nearest 0.1 mg and record the weight (MJ.
Repeat this procedure for a total of three determinations for
each sample.
2.2 Coating Density. Determine the density of the ink or related coating
according to the procedure outlined in ASTM D1475-60 (Reapproved
1980), (incorporated by reference - see §60.17).
2.3 Solvent Density. Determine the density of the solvent according to the
procedure outlined in ASTM D1475-60 (Reapproved 1980). Make a total
of three determinations for each coating. Report the density D0 as the
arithmetic average of the three determinations.
3. Calculations
3.1 Weight Fraction VOC. Calculate the weight fraction volatile organic
content W0 using the following equation:
Report the weight fraction VOC W0 as the arithmetic average of the three
determinations.
3.2 Volume Fraction VOC. Calculate the volume fraction volatile organic
content V0 using the following equation:
PageS
-------
4. Bibliography
4.1 Standard Test Method for Density of Paint, Varnish, Lacquer, and Related
Products. ASTM Designation D1475-60 (Reapproved 1980).
4.2 Teleconversation. Wright, Chuck, Inmont Corporation with Reich, R. A.,
Radian Corporation. September 25, 1979. Gravure Ink Analysis.
4.3 Teleconversation. Oppenheimer, Robert, Gravure Research Institute with
Burt, Rick, Radian Corporation, November 5, 1979. Gravure Ink Analysis.
Page 9
-------
SECTION 3
ASTM PROCEDURES APPLICABLE TO RM 24/24A
In this section, all ASTM procedures that are relevant to the application of EPA
Methods 24 and 24A are compiled and assembled into two groups: primary ASTM
procedures and secondary ASTM procedures. Primary ASTM procedures are those
which are directly referenced in the above EPA methods. All other relevant ASTM
procedures are compiled into the secondary ASTM procedures group.
3.1 Primary Procedures
• D1475-60 (Reapproved 1980) Standard Test Method for Density of Paint,
Varnish, Lacquer, and Related Products.
• D2369-81 Standard Test Method for Volatile Content of Coatings.
• D3792-79 Standard Test Method for Water Content of Water Reducible
Paints by Direct Injection into a Gas Chromatograph.
• D4017-81 Standard Test Method for Water in Paints and Paint Materials
by Karl Fischer Method.
3.2 Secondary Procedures
The ASTM procedures referenced in ASTM D2369-81, D3792-79, and D4017-
81 are shown below. No ASTM procedures are referenced in ASTM D1475-
60. In compiling them, the most recent edition of the method available is
included when no edition year is cited in the reference.
• D362-84 Standard Specification for Industrial Grade Toluene.
• D1193-77 Standard Specification for Reagent Water (Reapproved 1983).
• D1364-87 Standard Test Method for Water in Volatile Solvents (Fischer
Reagent Titration Method).
• D3728-88 Standard Specification for 2-Ethoxyethyl Acetate (99% Grade).
• D3925-81 Standard Practice for Sampling Liquid Paints and Related
Pigmented Coatings (Reapproved 1985).
Page 11
-------
• D3980-88 Standard Practice for Interlaboratory Testing of Paint and
Related Materials.
• D4057-88 Standard Practice for Manual Sampling of Petroleum and
Petroleum Products.
• E145-68 Standard Specification for Gravity Convection and Forced-
Ventilation Ovens (Reapproved 1987).
• E180-85 Standard Practice for Determining the Precision of ASTM Meth-
ods of Analysis and Testing of Industrial Chemicals.
• E203-75 Standard Test Method for Water Using Karl Fischer Reagent
(Reapproved 1986).
3.3 ASTM Methods Referenced in Secondary Procedures
ASTM procedures referenced in secondary ASTM references or otherwise
containing information relevant to the technical application of EPA Methods 24
and 24A procedures, including sampling and quality assurance/quality control
are shown below.
• E300-86 Standard Practice for Sampling Industrial Chemicals.
• E380-89 Standard Practice for Use of the International System of Units
(SI) the Modernized Metric System).
• E691-87 Standard Practice for Conducting an Interlaboratory Study to
Determine the Precision of a Test Method.
• E1267-88 Standard Guide for ASTM Standard Specification Quality
Statements.
3.4 ASTM Standard D3960
ASTM issued standard D3960 to serve as a guide for the selection of proper
ASTM Method for determining VOC and other components. This standard
incorporated ASTM standards D1475, D2369, D3792, and D4017, which were
also incorporated into RM24, and additional sampling and testing procedures.
The latest available editions of these standards are listed below:
• D3960-89 Standard Practice for Determining Volatile Organic Compound
(VOC) Content of Paints and Related Coatings.
Page 12
-------
ASTM procedures and methods referenced by ASTM D3960-89 but not
included in the preceding ASTM secondary reference compilation are:
• D1475-85 Standard Test Method for Density of Paint, Varnish, Lacquer,
and Related Products.
• D2369-87 Standard Test Method for Volatile Content of Coatings.
• 02697-86 Standard Test Method for Volume Nonvolatile Matter in Clear
or Pigmented Coatings.
• D2832-83 Standard Guide for Determining Volatile and Nonvolatile
Content of Paint and Related Coatings.
• D3792-86 Standard Test Method for Water Content of Water-Reducible
Paints by Direct Injection Into a Gas Chromatograph.
• D4017-88 Standard Test Method for Water in Paints and Paint Materials
by Karl Fischer Method.
• D4457-85 Standard Test Method for Determination of Dichloromethane
and 1,1,1-Trichloroethane in Paints and Coatings by Direct Injection Into
a Gas Chromatograph.
Page 13
-------
PRIMARY ASTM PROCEDURES
Page 15
-------
(1
Designation: 0 1475 - 6O (Reapproved 1980)'1
Standard Test Method For
DENSITY OF PAINT, VARNISH, LACQUER, AND RELATED
PRODUCTS'
TTui standard u iuucd under the fucii Uaitnauon 0 1475: the number immniuuiv followtnt the desttnauon indicate the
year of original adoption or. in me cu* 01 revision, me vear ot lau revmon. A numocr in parcntneici indicates me year of lau
rcapprovaL
TJiii mmW kai tarn afpnvra far tut dv arrnrrrj of ikt Otfarmm of Dtftnsi re rmlact Mtttiod 41X4.1 of Ftdtni Tta
Mtlimd .Standard ,Vo. 141A ana for lunnf m lit* DoO Inatx a/ Spttuicaiioniana Siandanu.
" Non—Ediiorui chaneq w«re maue tnrouinout in Octoeer i*80
1. Scope
l.l This method covers the measurement of
density of paints, varnishes, lacquers, and com-
ponents thereof, other than pigments, when in
fluid form. It is particularly applicable where
the fluid has too high a viscosity or where a
component is too volatile for a specific gravity
balance determination.
Nort 1—The method provides for the maximum
accuracy required for hidin« power determinations.
U u equally well suited for work in which leu accu-
racy is required, by ignoring the directions for recal-
ibntion and consideration ot" temperature differen-
tials, by using the container as a -weight-per-galloa"
cap.
2. Definition
2-1 density—the mass (weight in vacuo) of a
unit volume of the liquid at any given temper-
Mure. In this method, it is expressed as the
w«ght in grams per cubic millilitre. or as the
*«ght. in pounds avoirdupois, of one U. S.
gallon measure of the liquid at the specified
temperature in the absence of other tempera-
ture specification. 23"C is assumed.
1 Summary of Method
3.1 The accurately known absolute density
of distilled water at various temperatures (Ti-
w* I) u used to calibrate the volume of a
r. The weight of the paint liquid con-
*« M«ne container at a standard tern*
perature (25«Q or at an agreed-upon temper-
u then determined, and density of the
calculated in terms of grams 'per mil-
lilitre. or pounds per gallon at the specified
temperature.
4. Apparatus
4.1 Pycnometer—Any type, or weight-per-
gailon cup. having a capacity of from 20 to 100
mL may be used, provided that it may be filled
readily with a viscous liquid, adjusted to exact
volume, and covered to exclude loss of volatile
matter.
4.2 Thermometers, graduated in O.I°C such
as are supplied with glass pyconometers.
4.3 Coiutant'Tempenuure Bath, held at 25
±0.1*C is desirable.
4.4 Laboratory Analytical Balance.
Non 2—The usual weight-per-gallon cup and
similar specialized pycnometen may have filled
weights which exceed'the capacity of'the usual lab-
oratory analytical balance. In such cases, use of a
hanging pan. tnpie-beam balance, with scales grad-
uated to 0.01 g has been found to provide results the
mean of which was consistent with the overall pre-
cision and accuracy of the method.
4J Desiccator and Desiccated Balance, or a
room of reasonably constant temperature and
humidity are desirable.
5. Calibration of Pycnometer or Cup
5.1 Determine the volume of the container
at the specified temperature by employing the
following steps:
1 Thu mciaod is under the jurodictum of ASTM Com-
ounce 0*1 on Paint and ReUtta Coaunu ind Mstenau.
Current edition aeoroved Scot. 19. I960- Onguuiiv uuicd
I9J7.^epUee O 1475 - 57 T.
Page 17'
-------
5.1.1 Clean and dry the container and bring
it to constant weight. Chromic acid (see 5.1.1.1)
cleaner and nonresiduai solvents may be used
with glass containers, and solvents with metal
containers. For maximum accuracy, rinsing.
drying, and weighing must be continued until
the difference between two successive weigh-
ings does not exceed 0.001 % of the weight of
the container. Fingerprints on the container
will change the weight and must be avoided.
Record the weight. Af, in grams.
5.1.1.1 Chromic acid cleaning solution is
corrosive to skin, eyes and mucous membranes
and can cause severe burns. Avoid contact with
eyes, skin or clothing. In making dilute solu-
tion, always add acid to water with care. In
case of contact, flush skin with water, using a
shower if exposure is severe. Flush eyes for 12
minutes with copious amounts of water. Im-
mediately call a physician. Remove clothing
immediately and wash before reuse. Chromic
acid cleaning solution is a strong oxidizer.
Avoid contact with organic or reducing sub-
stances as a fire could results. See Supplier's
Material Safety Data Sheet for further infor-
mation.
5.1.2 Fill the container with freshly boiled
distilled water at a temperature somewhat be-
low that specified. Cap the container. leaving
the overflow onfice open. Immediately remove
excess overflowed water or water held in de-
pressions by wiping dry with absorbent mate-
rial. Avoid occluding air bubbles in the con-
tainer.
5.1.3 Bring the container and contents to
specified temperature. Use the constant-tem-
perature bath or room if necessary. This will
cause further slight flow of water from the
overflow onfice due to the expansion of the
water with the rise of the temperature.
5.1.4 Remove the excess overflow by wiping
carefully with absorbent material, avoiding
wkxing of water out of orifice, and immedi-
ately cap the overflow tube when such has
been provided. Dry the outside of the con-
tainer, if necessary, by wiping with absorbent
material Do not remove overflow which occurs
subsequent to the first wiping after attainment
of the desired.temperature (Note 3). Immedi-
ately weigh the filled container to the nearest
(LOO I ?e of its weight (Note 4). Record this
weight, tf, in grams.
0 1475
NOTE 3 — Handling the container with bare hands
will increase toe temperature and cause more over*
now from the overflow onilce. and will also leave
fingcrpnnis: hence, handling only with tongs and
with hands protected by dean, dry. absorbent mate-
rial is recommended.
NOTE 4— Immediate and rapid weighing of the
filled container is recommended here to minimize
loss of weitm due to evaporation of the water throuca
orifices, and from overflow subsequent to the first
wiping after attainment of temperature where this
overflow is not retained wuhin a capped enclosure.
5.U Calculate the container volume as fol-
lows:
» - iff -
where:
» • volume of container. mL.
.V - weight of container and water, g (5.1.4),
M - weight of dry container, g (5.1.1). and
a - absolute density of water at specified tem-
perature. g/'mL (see Table 1).
5.1.6 Obtain the mean of at least three de-
terminations of v to provide the value of V
required in 6.2.
6. Procedure
6.1 Repeat the steps in Section 5. substitut-
ing the sample for the distilled water and a
suitable nonresiduai solvent for the acetone or
alcohol (see 5.12 and Note 5). Record the
weight of the filled container. W, and the
weight of the empty container, w, in grams.
NOTE 5— Trapping of paint liquids in ground
glass or metal joints is likely to result in high values
of density which appear to increase with the viscosity
and density of the material: such errors shouid be
minimized 'by firm seaung of the joints.
6.2 Calculate the density in grams per mil-
lilitre as follows:
where:
Dm m density. g/mL.
6.3 Calculate the density in pounds per gal-
Ion as follows: •
where:
D - density. Ib/gaL
AT-8J455(Note6),and
V - volume of container. mL (see 5.1.6)-
NOTE 6— Tne factor K. S J435. is calculated fn*»
voiume-wctght relationship as follows:
Page 18
-------
((SI)) D 1475
«*x Ib/ga! at 25'Q; state the mean, the range.
453J9243 U the coavenion fa=«, r *'' ^^ the Procedurc for maximum ac-
pouad*. waven,on factor for grin* ,„ curacy, a single determination by one person
in one laboratory should not differ from the
7. Report mean of determinations by that person by more
7.1 In reporting the densitv. — ... ._. $?. .*?*" il^ «««V^ ^"S^ «»•
TABLE i Abwiur* o«n.ir. 0^ w,
OcsC
15
16
17
IS
19
20
21
22
23
24
21
26
27
21
25)
30
0.999127
0.998971
0.99S772
0.99S62J
0.99M»
0.99123 1
0.995020
0.997791
0.997566
0.997124
0.997072
0.996111
0.996540
0.99AMO
U«QW
DL995972
0.9956*4
orfortddino**
ft u mtfuty of ill*
Page 19
-------
Designation: 0 2369 - 81
Standard Test Method for
VOLATILE CONTENT OF COATINGS'
I. Scope
1.1 This method describes two procedures
for the determination of the weight percent
volatile content of solvent reducible and water
reducible coatings. Test specimens are heated
at 110*C ± S'C for 60 mm. or optionally fur
20 min. Although the technique used is the
same, residence times in the oven differ. The
two procedures are designated as follows:
I.I.I Procedure /(—Volatile Content of
Coatings Determined for 20 mm at 110*C *
5-C.
1.1.2 Procedure R (Preferred)— Volatile
Content of Coatings Determined for 60 mm at
|IOV*5V
1.1.2.1 Choice of and preference for 60 min
at 1IO*C *• S'C as a general purpose method
is based on the precision data presented in these
methods that was obtained on both solvent
reducible and water reducible coatings (see Sec*
lion 7). These coatings (single package, heat
cured) are commonly applied in factories to
automobiles, metal containers, flat (coil) metal
and large appliances and many other metal
pans. Procedure B is presumed applicable, sub-
ject to further precision studies, to most kinds
of paints and related coatings intended for
either ambient or baking film formation, except
where substantial amounts of volatiles may be
consumed or produced in chemical reactions
during Him formation. If an oven residence
(ime of 20 min at I IO*C ± 5*C is used the
analyst must recognize that poorer precision
was obtained using Procedure A (see Section
7).
Ncm I—Tesnn§ at 110'C x 5*C for 20 mm was
utili/.ed for the ntahlishmem of the anginal method
i / .V^m/iriunMTtfiM/ Siuuiardi.
in 196). Precision data are not available and OUT •*
have been property generated at the time. The •*
paints tested then were all solvent reducible, ft**
conditions. 20 mm at IIO'C £ 5*C. are no bMf*
. saiislactory for the determination of volatile coot*
of many coalings currently hetn; listed in 19X0. **
icr reducible and solvent reduable coaunp •*"
tested in the development of the present method *
I IO*C x i*C for 60 mm and 20 rain for «»**
precision data have been generated.
1.2 This method does not cover muiti-P*^'
age coatings wherein one or more pans may."
ambient conditions, contain liquid coreacttf0
that are volatile until a chemical reaction W
occurred with another component of the awto*
package coating.
IJ This method may not be applicable »
all types of coatings such as printing inks. J**
other procedures may be substituted with o>r
tuai agreement of the producer and user. S«
Note 5.
2. Applicable Documents
II A STM Standards:
O343 Specification for 2-Ethoxyethyl Acf
D 362 Specification for Industrial Grade Td*
uene*
O 1193 Specification for Reagent Water'
E 145 Specification for Gravity Conveetio»
Tttae meinou* an unUcr the (urndKUoa uf .«i*
v. ommiiic* 0* I on Pami ami Related Coaimts ana Mann*
Mil are the Oirevt mr.wwh.liiv.^SuhcimimHict 001.21*
(.MfmealAnalyrauT Pami anil Paint Matcnati.
Current edition approved June i6. 19*1
lemher |r^- Pln* y- :J- :1:
Page 20
-------
tad Forced-Ventilation Ovens'
E |$0 Recommended Practice for Develop-
iag precision Data on ASTM Methods tor
Analysis and Testing of Industrial Chem-
icais'
; Apparatus
j.l forced Draft Oven, Type 1IA or Type
|U is specified in Specification E 145.
jj Synnge. 5 mL. capable of dispensing the
ogling under test at sufficient rate that the
—omen can be dissolved in the solvent (see
r^
5.&
3J Wei$hin% or Dropping Buttle.
j4 Tea Tube, with new cork stopper.
JJ Aluminum Foil Dish. 58 mm in diameter
to IS ntm high with a .-.mooch bottom Mirt'uce.
Recondition the dishes for 30 mm in an oven
tf ||0*C ± 5"C and store in a desiccator prior
iJtttfcau
4.1 Purity of Reagents— Reagent grade
^coicals shall be used in all tests. Unless
otherwise indicated. it is intended that all re-
items shall conform to the specifications of the
Committee on Analytical Reagents of the
American Chemical Society, where such spec-
ifications are available.* Other grades may be
psed. provided it is ascertained that the reagent
jj of sufficiently high purity to permit its use
lessening the accuracy of the determt-
42 Purity of Wow— Unless otherwise indi-
cated. references to water shall be understood
to mean Type II reagent grade water conform-
io| to Specification D 1 193.
4J Toluene, technical grade. Specification
DM2.
4.4 2-Ethoxvethyl Acetate, technical grade.
Specification D 343.
Hun 2— The solvent* anil samples used in these
mtih-** may. under som* conditions. be hazardous.
infer u> the manufacturers Material Safety Oau
5lcci (or specific handling and safety precautions.
jsfe laboratory handling procedures and all appli-
cable US. Occapaiioaai Safety and Health Act reg-
ginjoat art to b€ followed in the handling of samples
104 solvents.
5. Procedure
j.l Mix the sample, preferably on a mei-han-
,cal shaker or roller, until homogeneous. If air
02369
bubbles become entrapped. Mr by hand until
the air has been removed.
5.2 Using an appropriate weighing container
(4.2. 4.3. or 4.4. with the synnge preferred for
highest precision), weigh to I mg. by difference.
a specimen of 0.30 * 0.10 g lor coatings be-
lieved to have a volatile content less than 40
weight % or a specimen of 0.20 ± 0.10 g for
coatings believed to have a volatile content
greater than 40 weight **. into a tared alumi-
num foil dish (4.5) into which has been added
3 ± 1 mL of suitable solvent (3.1. 3.2 or 3.3).
Add the specimen dropwise, shaking (swirling)
the dish to disperse the specimen completely in
the solvent. If the material forms a lump that
cannot be dispersed, discard the specimen and
prepare a new one. Similarly prepare a dupli-
cate.
NOTF 3—If the specimen cannot be dispersed in
the solvents listed (3.1. 3.2 or 3.3) a compatible
solvent may be substituted provided ii is no leu
volatile than 2-«thoxyetIiyl acetate (3.3).
5.3 Procedure A—Heat the aluminum foil
dishes containing the dispersed specimens in
the forced draft uven (4.1) fur 2U nun at I IO*C
±5'C.
5.3.1 Caution—Provide adequate ventila-
tion, consistent with accepted laboratory prac-
tice, to prevent solvent vapors from accumulat-
ing to a dangerous level
5.4 Procedure B—Heal the aluminum'foil
dishes containing the dispersed specimens in
the forced draft oven (4.1) for 60 min at 110*C
± 5*C. Caution: See Section 5.3.1.
5 J Remove the dishes from the oven, place
immediately in a desiccator, cool to ambient
temperature and weigh to I mg.
NOTC i—If unusual decomposition or degrada-
tkta of the specimen occurs during heaung. the actual
time and temperature used to cure the mating in
practice may be substituted for the time and temper-
.aiure specified in this method, subject to mutual
agreement ol' producer and user.
6. Calculation!
6.1 Calculate the percent volatile matter in
• Aumttl AM* tfASTM Amriw*. Ptni 39 ami 41.
' Alum* AM* ffASTM Sun^uf^i. ton 30
* "Xntcm OwnuoU. American Chemical !!>««« Spec-
iftaiNMU.'Am. ChcmnaJSoiu Wellington. U C' FIKMI|-
ICHIOIU on ih« icsun; of rngcau not lined by ifte Amcncaa
Chemical Society, xc -Reijem Chemicab and >umJanta.~
by Juttpfl RoMiil D. Van Nowrand Co.. Inc. New York. N
Y. and the 'UniiciiSuici "
Page 21
-------
the liquid coating as follows:
Volatile matter. % - 100 - {« W. - W,)/S> x I00|
where:
W< - weight of dish.
W-i <• weight of dish plus specimen after heat-
ing, and
S •• weight of specimen.
6.2 The percent of nonvolatile matter in the
coating may be calculated by difference as
follows:
Nonvolatile mauer — 100 — volatile matter
7. Precision
7.1 Procedure A (20 rain at 110«C ± 50Q:
7.1.1 The precision estimates are based on
an interiaboratory study7 in which one operator
in each of 15 laboratories analyzed in duplicate
oa two different days seven samples of water-
based paints and eight samples of solvent-based
paints containing between 35 % and 72 % vol-
atile material The paints were commercially
supplied. The results were analyzed statistically
in accordance with Recommended Practice
E180. and the within-laboratories coefficient
of variation was found to be I.! % relative at
193 degrees of freedom and the between-labo-
ntorics coefficient of variation was 2J % rela-
tive at 178 degrees of freedom. Based on these
coefficients the following criteria should be
used for judging the acceptability of results at
the 95ft confidence level.
7.1.1.1 Repeatability—T<*o results each the
mean of duplicate determinations, obtained by
the same operator on different days, should be
considered suspect if they differ by more than
relative.
MncOT Sociftr for Ttaaif tn4 Mmunti* <•*«»»
I nut mmr turn »umtfii*^m ilutamm J.p< I*—/1.*.
02369
7.1.1.2 Reproducthtlity—Two results*' tx*
the mean of duplicate determinations, obtained
by operators in different laboratories should be
considered suspect if they dirfer by more ihtf
7.1% relative.
7.2 Procedure B (60 min at I 10"C ± 5*O:
7.2.1 The precision estimated for tests at tf
min at I IO*C ± 5*C are based on an inieriab-
oratory study7 in which one operator in each of
15 laboratories analyzed in duplicate oa t**
different days seven samples of water-bated
paints and eight samples of solvent-bawd
paints containing between 35 % and 72 ^ »«*•
atile material The paints were commercially
supplied. The results were analyzed statistically
in accordance with Recommended Practice
£ ISO. and the within-laboratories coefficient
of variation was found to be 0J % relative at
213 degrees of freedom and the berweenrlabo-
ratones coefficient of variation was 1.7 "a reii-
live at 198 degrees of freedom. Based on these
coefficients, the following criteria should be
used for judging the acceptability of resuiu at
the 95 % confidence level
7.2.1.1 Repeatability—Two results, each in*
mean of duplicate determinations, obtained by
the same operator on different days should be
considered suspect if they differ by more than
1.5 % relative,
7.11.2 ReproducibHiiy—T»o results, eat*
the mean of duplicate determinations, obtained •
by operators in different laboratories should be
considered suspect if they differ by more than
4.7 % relative.
'VicymiiH tfau an tvnUblc on tout from ASTM
lanm. 191* KM* Sc niMdpfcu. Pa, I9IOX *•"
JUOOt. IOW.
•• fiaMtaf^t. /«/« JUn St. rrtrtrtftm. f*. 1910). Mtar
Page 22
-------
DMifnation: O 3792 - 79"
Suniv-
Standard Test Method for
N,™ W*™fWHDUCIBLE HA.NTS BY
INJECTION INTO A GAS CHROMATOGRAPH'
..
a- . i i w 4>T« UN IIUMMT tOUMrfl
ac. anh« an of rr»i»ri«. ta« yof of Un irnamu A B»
11 In iltiipiima iiMtinm itir
«•—• iMictut U» j «*r of iut
!• Scope
I.I This method is for the determination of
otal w«ercome« of water-reducible
' h ' * for
vHe
acrylic. acrylic). t a* not yet been evaluated
for other w«er.reducible paints but is believed
-------
&auo« « available.* Other
.
available.* Other md«
used, provided it a first asceniiaed that the
«*!** « or suxTkieatly high punty to permit
5.2 f^iy of Jf^-Unjea othen>iae ^
caiei refereact to water shall be understood to
meaa Type n reageat water coa/brmiax to
Speafiatuaa O 1 193. ^^
C!"— HeUum of 99.9*5 * or
nitrs}8«» "My also
5.4
(DMF) (Anhydnus)
such
or 23^,1 capacity
3.6 ^^(^/rv,tn«,)L.S
5.7 Sipiwa Sampl, Vialt. lO^nl aptatv
wth flu««artoa.»aeed septa are prelerrcT *
6. S»ferr Prccaatkra
6.1 DLitttkylformainiiie is harmful if inhaled
or absorbed throusa ,kia, Uj« onlv with ade-
quate- veatilatioa. Avoid conua'wnh »kio,
vtn. aad clothing, b ,3* of contact, unmedi-
ait y flush , km or cy« w,u» ?|ea,y of water for
at least 13 mm. If eVCT are ^^ al, a
phyMcian. Remove and wash elothmt before
feux.
03792
6J2. 2-Propanol is flammable. Its vapor is
harmful especially to the eyes. Use with ade-
quate ventilation. Avoid contact with skin and
eyes. In case of contact with liquid, wash skin
with soap and water, for eye contact with high
concentration of vapor or liquid. Hush eyes for
15 min and see a physician immediately. Re-
move contaminated clothing immediately.
7. Calibration
7.1 Before each calibration and series of de-
terminations (or daily) condition the column ax
200* C for 1 b with carrier gas flow.
7.2 Determination o/JUloar* Resparut Foe*
ton^— Anhydrous 2-propaaol is used as an in-
leraal standard. The response factor to water
relative to the standard is determined by means
of the following procedure. See Fig. 1 for a
typical chromaiognm. It is good practice to
determine (he response factor daily or with
•g»«»fo senes of dftrrnirT''^'*
7.2.1 Weigh about 02 g o( water and &2 j
of 2-propanol to the nearest 0.1 mg into a
septum sample viaL If it has b^p determined
that a correction for the water con teat .is net*
essary, weigh 2 ffll of dimethyl formamide into
the viaL If the dunethyUbrmamide is anhy-
drous, simply add 2 ml of it as weighing is not
Inject a I /tl aliquot of the above solu-
tion onto the column and record the caromai-
ogram. Tbe retentiea order and approximate
retention times after the air peak are (/) water,
about 0.7 min: (2) Z-propanoU about IS min.
aad (J) DMF. about 7 min.
7.2J The preferred procedure to obtain the
water content of the OMF is by Karl Fischer
utration (Note I). IT this **** been done, calcu-
late the response factor for water by means of
the following equation:
._
where:
Jt * response factor.
W. •• weight of 2-propanol
* M,U - weight of water added.
* -**»»!»< Chcmiiau. Amman Oirmi^ji Santtj Sort-
ificaiuia.- AmtncM anmai Scam. WuHmjion, 0. C.
For farIOII.M.. on IM mut) of mf cni» net laud •* in«
CJwanoi Swiat. m -tnftai Chonicatt *nd
* M J.^ftA R.M. 0 V« H.Mramt C.t. Inc.
> or*, s V 4«4 uu -Caned Sum
Page 24
-------
W. • weight of dimethyl fcrmamide.
/
-------
relative at 34 decrees of freedom and the be-
tweea laboratory coefficient of variation w«
16 It relative at 30 degrees of freedom. Baud
on these coefficients, the followiC| criteria
ihould be used for judging the acceptability of
resuiu at the 95 * confidence level:
10.1.1 Atpmabiluy—TwQ results, each the
mean of duplicate determinations, obtained by
D3792
the same operator on different days, should be
considered suspect if they differ by more than
19 * relative.
10.1.2 Reprodueihiltty—Two results, each
ihc m*""! of duplicate eif*f\ »»"«»im»^_ obtained
by operaton in different laboratories, should
be considered suspect if they differ by more
than 7 J % relative.
Cebuna
CururOu
MO*C A/Wr ite Lpn^Moi ktf Marad tkc
Utt MHfmim u 170^ uaui OMP ckui
KMI tta umpmum u I40*C for MkM^oi
Page 26
-------
W
D 3792
i.SZ 4-
nc.:
, <»<««nrirM4HWftri
•JSiE""'"1"* V.
*=^^^-^is-^sgs»iT=s=^stfM53
I •H7«aiM MM*. PM ^^^ —^ w rtr<< ,-./ -^ -//W—._,
•frniiftltrASTU *~***Dint*n.
Page 27
-------
Designation: 0 4O17 > 81
Standard Test Method for
WATER IN PAINTS AND PAINT MATERIALS BY KARL
FISCHER METHOD1
"««*«* lh« A*** JoitiuiHm D 4017: the number imnoiiaieiy followmj IRC dcuenaiion indicates ihc
nuiM or. in inc cue o< rcvum*. the year uf Uu rmuun. A numner in nefemhcw indicates the ycirufUu
1.1 This method is applicable to all paints
0& paint materials, including resins, mono-
ocfs, and solvents, with the exception of aide-
tydci and certain active metals, metal oxides
tad metal hydroxides. While the evaluation
ttf limited to pigmented products containing
,000011 of water in the 30 to 70 % range, there
0 lessen to believe that higher and lower con-
oatrations can be determined by this method.
j, AffUcaUc Documents
11 ASTM Standard*
D U93 Specification for Reagent Water1
E ISO Recommended Practice For Devel-
oping Precision Data on ASTM Methods
for Analysis and Testing of Industrial
£203 Test for Water Using Karl Fischer
Reagent
22 Othtr Doountnr.
Archer. E. E. and Jeeter. H. W.. Anoint, VoL
ft1945. p. 357. ^
3. Summery of Method
3.1 The material is dissolved in pyridine. or
igodter appropriate solvent, and titrated di-
KdTf with standardized Kari Fischer Reagent.
0 an electrometnc end point. The sluggish
Ksction with water in pyridine is accelerated
•iih a chemical catalyst. 1-ethylpiperidine.
U Pyridine is used as a solvent to minimize
gicrference problems caused by ketones. It is
ibo used because the more commen solvent.
oethanoi. will not dissolve many common
(tains and because methanol reacts with some
loins to produce water.
4. Apparatus
4. 1 Kari Fucker Apparaaa. manual or auto-
fnatir. encompassed by the description in
Method E 203. Apparatus should be equipped
with a 15-mL bureu Class A. or equivalent.
4.2 Syringe. 100-nL capacity, with needle.
4J Syringes, 1-mL and 10-mL capacity,
without needle, but equipped with caps.
5.1 Purity of Rtegenu — Reagent grade
chemicals shall be used in all tests. Unless
otherwise indicated, it is intended that all re-
agents shall conform to the specifications of the
Committee on Analytical Reagents of the
American Chemical Society, where such spec-
Utcations are available.4 Other grades may be
used, provided it is ascertained that the reagent
is of sufficiently high purity to permit its use
without lessening the accuracy of the determi*
5.2 Purity of Wettr— Unless otherwise indi-
cated, references to water shall be understood
to mean Type II reagent grade water conform*
1 This Betted » rade* tn* jwodicuoo trf .ASTM Coot-
(BtaeeO-l on Paint ud Related Coutafiud Matcnalsaod
Hi**dimifc»poiiiiMiiyorSvbcoaMBweOOUieoChcB-
ical Aaaiyta of Paim aod Paint MMenata.
Cumm cd>ue« approved JUM 2k l«ll. Publtthcd Sep.
HMfeer IVHI.
' .«MMW *M* e/ .<5TM tandw*. Ptni M. 21. H. M.
29.3l.40.aod41.
«M* ffASTM Siv***. Pan 30.
4 "Kmacai Cbcinxaii. Amcncu Cbcaical Society Spec-
Uicaumu.' Am. Chemical Soc. Wukugien. O. C. For >u;-
gaiMMaeo ih« icuwg a( mccaa noc lined by the American
Chcoucai Sooety. ice 'ttciieni Ckemieati and Siandvdi."
by JeMpa Houm. O. Van Neurand Co. Inc. New fort N.
T. aad to* -Uniied Sum Phtnucepeu."
Page 28
-------
D4O17
ing to Specification D 1193.
5.3 Karl Fischer Reagent*
5A Pyndine. reagent grade."
53 l-Ethyipipendine.'
6. Safety Precautions
6.1 Karl Fischer reagent contains four toxic
compounds, namely iodine, sulfur dioxide, pyr-
idine. and methanol or glycoi ether. The re-
agent should be prepared and dispensed in a
hood. Care must be exercised 10 avoid inhala-
tion or skin contact. Following accidental con-
tact or spillage, wash with large quantities of
water.
6.2 Pyridine and methanol solvents should
be treated with the same care as Karl Fischer
reagent.
6J 1-ethyipiperidine is of unknown toxicity
and. therefore, should be handled with the
same care as the above materials.
6.4 Many paint materials are highly flam-
mable and should be transferred in a well-
ventilated area free from sources of ignition.
7. Procedure
7.1 Standardisation of Karl fucker Reagent:
7.1.1 Add enough fresh pyndine to cover the
electrode tip. plus I mi, of l-«thylpiperidine
catalyst per 20 mL of pyndine. Catalyst per-
forms best at a concentration of about 5 "*• of
the volume present.
7.12 Fill the 100-y.L syringe to about half
full with distilled water and weigh to the nearest
0.1 mg.
7.1 J Pretitrate the pyndine to the endpoint
indicated by the equipment manufacturer, by
adding just enough Karl Fischer Reagent I
(hereafter referred to as KFR) to cause the end
point to hold for at least 30 s.
7.1 J.I The use of the catalyst greatly in-
creases the reaction rate between water and
Karl Fischer reagent. To obtain reliable result*.
increase the electrode sensitivity and reduce
titration rate to a minimum. Most instruments
have controls for these functions. Consult in-
structional manual for information on these
controls.
7.1.4 Empty the contents of the syringe into
ihe utrator vessel. Immediately replace the
stopper of (he sample port and titrate with
KFR lo the cmipotnt as described in 7.1.3.
7 1.5 Repeat standardization until replicate
values of F agree within I *&. Determine ^
mean of at least two such determinations. O"?
out calculations retaining at least one e»w
decimal figure beyond that of the »cqui««
data. Round off figures after final calculation
7.1.6 Calculation:
F-JfP
where:
F - KFR litre.
J - water added. g. and
P - KFR used. mL.
The value for F should be recorded to the fotf
significant digits and should be the mean <*•
least two determinations. Typical values are*
the range of 0.004000 to 0.006000 g/mL.
7.2 Analysis of Samples With Men V+
0.5", Water.
7.11 The titrauon vessel should already cot-
tain preutrated pyndine and catalyst. « 4*
jcnbed in steps 7. 1 . 1 and 7. 1 .3 in the stand**
ization procedure. Best results are obtain^
with fresh solvent, that is. contain no previous!?
titrated specimen in the vessel
7.12 With a 1-mL or 10-mL syringe. <*«•
the amount of material indicated in Table I-
7.11 1 Remove the syringe from the s»mp*
pull the plunger out a little further, wipe "*
excess material olT the syringe, ami place » of
on the syringe tip. Weigh the filled synnje*
the nearest O.I mg.
/.2J Remove the cap. and empty the synnf
contents into the pretitrated pyndine *e»
Pull the plunger out and replace the cap. TiW
ihe specimen with KFR to the endpoi"1 «•
scribed in 7.1 J.
7.14 Reweigh the emptied syringe.
culate the specimen weight by difference.
7.2J Calculation:
* water - (milliliires KFR used
X F x
7.3 Analysis of Materials With Lest I**
0.3 "o Water.
7 J.I For 0.1 to OJ •&. follow procedure *
7.2 (1-g specimen), except substitute a !••*
raicroburet for the 23-mL buret in (he ""
Fischer apparatus.
7.3.2 For less than 0. 1 5>. use a 1-mL mic*
Co.
ct
-------
D4O17
tad increase specimen size as much as
(ceded. up to 10 g. It should be possible to
ecasure moisture levels down to I ppm
(0.0001 ft) by (his approach (see Note).
Hon— Specimen with leu than 0. 1 4 wucr may
jftwfCfpccuJ handling techniques to prevent pickup
rf aiaoipherie moisture. Toe precision of this test
•0 4*tennuud with specuncns containing higher
i Maiateaance
tl Cleanup— Clean the titration vessel by
rinsing with fresh pyridine. Do not use metha-
goi or other solvents.
12 0/7»?«w— Check frequently to be sure
that til drying lubes are in good condition and
rightly connected. Replace dessicam when in-
dicator color changes through half of the tube.
SJ Electrode Performance— \t electrode re-
sponse is sluggish or otherwise off standard.
take the following steps, in turn, to correct the
problem. Test the electrode with a titration
lAer each step, to determine if the next step is
icouired.
til Wipe the electrode tip with a clean
piper towel
H2 Wash the electrode by dipping in con*
ceatrated hydrochloric acid for at least I min.
Rinse first with distilled water, then with meth-
aooL
UJ Follow manufacturer's instructions on
rescuing endpoint meter.
8.3.4 Replace power source. See manual for
replacement procedure.
8.3.2 Replace the electrode.
9. Precision
9.1 The precision estimates are based on an
interiaboratory study in which one operator in
each of seven different laboratories analyzed in
duplicate, on two different days, seven samples
of water-based paints of various types contain-
ing between 25 to 75 % water. The results were
analyzed statistically in accordance with Rec-
ommended Practice E180. The withtn-fabora-
lories' coefficient of variation was found to be
1.7% relative at 98 degrees of freedom, and the
between-laboratories' coefficient of variation
was 5.3 % relative, at 42 degrees of freedom.
Based on these coefficients, the following cri-
teria should be used for judging the acceptabil-
ity of results at the 95 % confidence level.
9.1.1 Repeatability— Two results, each the
mean of duplicate determinations, obtained by
the same operator on different days should be
considered suspect if they differ by more than
4.7 % relative.
9.1.2 Reproducibility— Two results, each the
mean of duplicate determinations, obtained by
operators in different laboratories should be
considered suspect if they differ by more than
13.0% relative.
TAIU I S*
0-J-I.O
I 3
3 10
10-»
30-70
>70
Spoon W«n fee
I
1
2 S
i:
0*1.0
0.1-0.4
0.1
iraM Volvm M
J Mf/niL mm.
5 in
10 a
10 so
2013
15 23
3)
TV.
r *iW MM tf rmrmfd »wrr/!»» tttn
.
i« ittr 4STM Comamtr M Sui*u*i. /Vf» Jbrr St
w»r«r0rf cwm*ratiMtfi • mettmg •/***
fCftftw tf f9t
'A loini.,
Page 30
-------
SECONDARY ASTM PROCEDURES
Page 31
-------
Designation; D 362 - 84
Standard Specification for
Industrial Grade Toluene1'2
This stanoard is issued under the fixed designation D 362: the numeer immediately following the designation mil""- the vear of
onpnaladoplion or. in the cue 01 revwon. the year at last revtaon. A numcer in parentneses uaicaies ttt year ot Ust respprovai. A
superscript epnlon n» indicatesan editorial cfaaate since tne last revision or reapprovai.
"^ ** "" *V agmae* mthe 0*1*"mm of Defense and for listmtmtneDoD Index of Speaftcaaons
ffor this stanaard Has voted us withdrawal. In the absence of substantial reasons that u should be continued.
Thecomn
ihe Society wiil appro** withdrawal from puoitcaaon in Julv 1990.
1. Scope
1.1 This specification covers industrial grade toluene.
2. Referenced Documents
2.1 ASTMStandards:
D847 Test Method for Acidity of Benzene. Toluene.
Xylenes. Solvent Naphthas, and Similar Industrial Aro-
matic Hydrocarbons'5
D848 Test Method for Acid Wash Color of Industrial
Aromatic Hydrocarbons3
D849 Test Method for Copper Corrosion of Industrial
Aromatic Hydrocarbons3
D 850 Test Method for Distillation of Industrial Aromatic
Hydrocarbons and Related Materials3
D853 Test Method for Hydrogen Sulfide and Sulfur
Dioxide Content (Qualitative) of Industrial Aromatic
Hydrocarbons3
D94I Test Method for Density and Relative Density
(Specific Gravity) of Liquids by Lipkin Bicapiilary
Pycnometer4
D1078 Test Method for Distillation Range of Volatile
Organic T iqnirfg3
D 1133 Test Method for Kauri-Butanol Value of Hydro-
carbon Solvents3
D1209 Test Method for Color of dear Liquids (Platinum-
Cobalt Scale)3
D1296 Test Method for Odor of Volatile Solvents and
Diluents3
D1298 Test Method for Density, Relative Density (Spe-
cific Gravity), or API Gravity of Crude Petroleum and
Liquid Petroleum Products by Hydrometer Method*
D1364 Test Method for Water in Volatile Solvents
(Fischer Reagent Titration Method)3
D 1613 Test Method for Acidity in Volatile Solvents and
Chemical Intermediates Used in Paint. Varnish. Lac-
quer, and Related Products3
This specification ts under the jurisdiction of ASTM Committee D-16 on
Aromatic Hydrocaroons and Relates ChfimnH and is the direct remonsbiiitv of
Subcommittee DI6.0A on flTX Cyctoaeaane. and Their Dew-ttrveT^^
Current edition March 30.19S4. Published June 198.4. Orieaaily published u
D 362 - 33 T. Ust previous edition D 362-80. •»"—y p™u»eo as
: This matenal was formerly known as -industrial pure toluol"
1 Annum Book of ASTM Stanaaras. Vol 06.03.
4 Annuat Book ot'ASTM Stanaaras. Vol 01.01.
D 1616 Test Method for Copper Corrosion by Mineral
Spirits5
D 3437 Practice for Sampling and Handling Liquid Cyclic
Products3
D3505 Test Method for Density or Relative Density of
Pure Liquid Chemicals3
E 1. Specification for ASTM Thermometers4
2.2 Federal Specmcaiion:
PPP-C-2020 Packaging of Chemicals. Liquid. Dry, and
Paste7
3. Properties
3.1 Industrial grade toluene shall conform to the following
requirements:
AP
Specific fnv)iy
dear liquid free of i
I078TOSJ to 23.6
-------
4. Test Methods
4.1 The material shall be sampled and the properties
enumerated in this specification shall be determined in
accordance with the following ASTM methods:
4.1.1 Sampling—Practice D 3437.
4.1.2 Appearance—Visual inspection.
4.1.3 Specific Gravity—Any convenient method that is
accurate to the third decimal place.
NOTE 1—See Test Methods O 941. D 1298. and O 350S.
4.1.4 Co/or—Test Method D 1209.
4.1.5 Distillation—Tea. Method D 1078. or Method
D850. using ASTM Solvents Distillation Thermometer
having a range from 98 to 152"C and conforming to the
requirements for Thermometer 41C as prescribed in Specifi-
cation E 1.
4.1.6 Odor—Test Method D 1296.
4.1.7 Water—Place 10 mi- of the sample in a dry, loosely
0362
stoppered test tube < 16 by 150 mm; and place in a water bath
at 20*C (68*F). On viewing the tube transversely at the end of
3 min. no turbidity should be observed and no free water
should appear on the bottom of the tube.
NOTE 2—If a more precise test for moisture is needed the purchaser
and the seller should agree on specification limits using the Karl Fischer
quantitative procedure described in Test Method D 1364.
4.1.8 Aciditv—Test Method D 1613 or Method D 847.
4.1.9 Acid Wash Color—Test Method D 848.
4.1.10 Sulfur Compounds—Test Method D 853.
4.1.11 Corrosion—Test Method D849 or Test Method
D1616.
4.1.12 Solvent Power—Test Method D 1133.
5. Packaging and Labeling for UJS. Government Procure-
ments
5.1 United States Government procurements shall be
packaged and labeled in accordance with the applicable
paragraphs of Fed. Spec. PPP-C-2020.
ngnat
n mconntcoott
. Uttn or mn atrava urn nontw «oviwo mm mmnmmnitf«» ramuy * in? wen
mr*™MUwtm»frmtri3oan3mtettnwiMiniTut™ttnrwMnv*<^fvwtMY»™*tt
H»ASrMHMooMmn. Your eemmtna w# rtcxv* emtul eamataaon at« mMBng o/ tf» i
..____ — ^. . -T.iV01' "^ •«•"* " vw rt« «i« your eemnwn n*v» nor nemM • ter MMI? you sne«« maw your
i*no«wwin«/»STMCam™it».c«stano»o*. J9»fl Rm St. PMMIWPM /* 19101
34
-------
(jhlM Designation: 0 1193 - 77 (Reapproved 1983)*1
NO. /910
Standard Specification for
Reagent Water1
This standard is mued under the fixed designation D 1193: the numocr immeaiatety following the designation indicates toe vear 01'
original adopooa or. in tne case of revision, the year ot last revision. A number in parcntneMS indicates tne year of last reapprovai. A
superscript epnion it) inrtiratn an eoiioruu ciuuge since the last revision or reapprovai.
This specification Has been opprovea for use ev agencies at the Department of Defense and for listing in the DoD Index oi'Specuicauons
and Stanaaras.
" Nan—The saietv hazards caveat was added editonailv m January 1989.
1. Scope
1.1 This specification covers requirements for water suit-
able for use in methods of chemical analysis and physical
testing. Four grades are specified:
1.1.1 Type I Reagent Water.
1.1.2 Type II Reagent Water.
1.1.3 Type III Reagent Water, and
1.1.4 Type IV Reagent Water.
Total miner, max. me/litre
Electrical conductivity, max. umno/cm at 298 K.
(23*0
Bectncai resmivity, mm. MQ-on at 298 K.
(2TO
pHat298K(2TQ
Minimum color retention time of potassium
Type!
0.1
0.06
16.67
60
net detectable
TypeU
0.1
1.0
1.0
60
noc
Type HI
1.0
1.0
1.0
6.1 to 7.3
10
10 ui/litre
Type IV
10
5.0
0.2
5.0108.0
10
Maximum TTluhte silica
MioobtoiogicBi dasBncanon
* The measurement of nH in Type I and 11 M^BI w.t>r* « m......,!^ 1mj hm t-n-n »t.mm..«4 tw^ thr p^Ane. «aee etocaodes used in tto tea coBtimmate the
' When bacterial levels need to be controlled, reagent grade types should be further <-tf««t«i«»f u follows:
Type A
0/ml
Mixiinuin toui bacteni count
Typefl
10/ml
TypeC
100/ml
1 .2 The method of preparation of the various grades of
reagent water determines the limits of impurities and shall be
as follows:
1.2.1 Type I grade of reagent water shall be prepared by
the distillation of feed water having a maximum conductivity
of 20 umfaos/cm at 298 1C (22*Q followed by polishing with
a mixed bed of ion exchange materials and a 0.2-um
membrane filter.
NOTE— The "mho" is used in this standard as the unit of electrical
conductance until such time as the equivalent SI unit, the Siemens
(symbol SX is accepted by the committee having jurisdiction.
Type II grade of reagent water shall be prepared by
distillation using a still designed to produce a distillate
having a conductivity of less than 1.0 jimho/cm at 298 K
(25*O. Ion exchange, distillation, or reverse osmosis may be
required as an initial treatment prior to distillation if the
punty cannot be attained by single distillation.
1.2.3 Type III grade of reagent water shall be prepared by
distillation, ion exchange, reverse osmosis, or a combinauon
1 This speculation is under the )urodiction of ASTM Committee D-19 on
Water and is the rewonabiiirr of Subcommittee D 19.02 on General Speculation!
and Technical Resources.
Current edition approved Jan. 28. 1977. Published March 1977. Orignuuy
issued as D 1193 -51 T. Las previous edition O 1193-74.
thereof, followed by polishing with a 0.45 urn membrane
filter.
1.2.4 Type IV grade reagent water may be prepared by
distillation, ion exchange, reverse osmosis, electrodialysis. or
a combination thereof.
1J The choice of one of the various grades may oe
ri^gnatpH by the method or by the investigator.
1 J.1 Type I grade of reagent water shall be used where
maximum accuracy and precision is indicated, providea
dissolved organic matter is not a possible interference.
1.32 Type II grade of reagent water shall be used for most
analytical procedures and all procedures requiring water low
in organics.
1JJ Type IH grade of reagent water shall be recom-
mended for general laboratory testing. .
1.3.4 Type IV grade of reagent water shall be used in
procedures requiring large amounts of water of lower punty,
parricuiariy for the makeup of synthetic test solutions, nnse
water, and wash water.
1.4 This standard may involve hazardous materials, oper-
ations, and equipment, nis standard does not purport to
address all of the safety problems associated with its use. It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine tne
applicability of regulatory limitations prior to use.
Page 35
-------
01193
2. Referenced Documents
11 ASTM Standards
D 839 Test Methods for Silica in Water
Dll2S -TeSt ^°^ for ^^noi Conductivity and
Resistivity of Water
01129 Definitions of Terms Relating to Water
D1293 Test Methods for pH of Water
D1888 Test Methods for Partxcuiate and Dissolved Mat-
ter. Solids, or Residue in Water
F 60 Methods for Detection and Estimation of Microbio-
. logical Contaminants in Water Used for Processing
Electron and Microelectronic Devices3
3. Significance «*"* Use
3.1 The method of preparing Type I reagent water adds
organic contaminants to the water by contact with the
ion-exchange matrriah It should be noted also that the
method may or may not remove non-ionized dissolved gases
3.1.1 The dissolved or paniculate organic contamination
would normally range from 1 to 10 ug/L. The concentranon
of nomonized dissolved gases may exceed 10 mg/litre
32 The quality of the effluent water depends upon the
type, age, and method of regeneration of the ion exchange
mairriak Likewise, the flow rate through the ion exchange
resin bed will change the conductivity of the effluent water.
The instructions of the manufacmrer of the resins or the
resin cartridge bed should be followed.
33 The use of the membrane filter in the preparation of
Type I and Type HI water may add a small amount of
organic components to the water first produced.
3J.I Some membrane niters contain as much as 8
mass % of soluble components resulting from the manufac-
turing process.
If the contamination of the water by the organic
component is of significance to the test. Type II water should
be used or the membrane should be rinsed by discarding the
fust 10 mL of water produced per square centimetre of filter
area, or until a test for the organic components is negative.
3.4 Type II grade of reagent water should be sterile and
pyrogen free as produced and generally may be "vrt when-
ever freedom from organic or biological contaminants is
desirable. However, the method of storage and handling of
the water may itself result in contamination.
3.4.1 To obtain sterile water, any of the preceding listed
types of reagent water may be produced bottled, and heated
to 394 K. (lirQ for 20 min. This procedure is most easily
carried out by autodaving at 103 kPa (15 pa) for 20 min.
3.4.2 Type II water should be pyrogen-free. but must be
tested in conformance with the requirements of the current
edition of U.S.P. if proof is needed.
3.5 Types I. II. and HI reagent water should be protected
from atmospheric contamination and from solution of
container and tubing: ' '
3.5.1 Extreme care must be exercised in handling samples
when making an analysis. Sample containers and tubing
should be made of TFE-fluorocarbon. titanium, tantalum.
block tin. quartz. 18-8 stainless steel, polyethylene, or other
] Annual Book ofASTMSunaordt. Vol 11.01.
' Annual Book oi'ASTMSuuaatds. VoU 11.01 ana 11.01
material proven to be suificientiy resistant to chemical attack
so as not to cause contamination in the intended use.
3.6 Because atmospheric gasr* and impurities rapidly
reconiaminate exposed water, in-line electrodes should be
employed for determining the electrical conductivity pi
reagent water Types I. IL and III. The measurement of pH in
Type 1 and II reagent waters is meaningless and has been
efoTMpared from the procedure, since electrodes used in the
test contaminate the water.
3.7 Since freedom from biological contaminants may be
important in the test procedure using any of the reagent
waters specified, a classification of bacterial levels is included
and should be specified if of significance to the test being
penonned.
4. Definitions
4.1 The terms -paniculate matter." "dissolved matter."
"total matter." and others related to water constituents
determined in these methods, are defined in accordance with
Definitions D 11 29 as follows:
4.1.1 paniculate matter— tiax, noniiquid matter, exclusive
of gases, which is heterogeneously dispersed in water.
4.1.2 dissolved matter— (hat matter, exclusive of gases.
which is dispersed in water to give a single homogeneous
phase.
4.U total matter— the sum of the panicuiate and dis-
solved matter. _
42 For definition of other terms used in this specifica-
tion refer to Definitions D 1 129. For an explanation of the
metric system including units, symbols, and conversion
factors see Standard E 380.
5. Reagents
5.1 Paray of Reagents— Reagent grade chemicals shall be
used in all tests. Unless otherwise indicated, it is intended
that all reagents snail conform to the specifications of the
Committee on Analytical Reagents of the American Chemi-
cal Society, where such specifications are available. Other
grades may be used, provided it is first ascertained that the
reagent is of sufficiently high purity to permit its use without
the accuracy of the determination.
_ g__
52 Purity of Water— Unless indicated otherwise, refer-
ences to water shall be understood to mean Type I or Type II
reagent water as denned in this specification.
5 J Potassium Permanganate Solution (OJ16 &'«™—
Dissolve 0.316 g of potassium permanganate (KMnOJ in
water and dilute to 1/L in a volumetric flask.
5.4 Sulfuric. Acid dp gr /.«;— Concentrated aad
6. Requirements
6.1 Reagent water shall conform to the requirements
specified in Section 1.
62 Additional requirements concerning specific contami-
nants or methods of preparation may be included in tnis
ChenwalJ
«» Sot. WMfaafioit. OC For warned on the
the Amman Cbemnt Soaety. xc -Re«e« CbemM*
UMM.*Am.Oiem-
Ibv
bv
Phamucopea.'
Page 36
-------
D 1193
specification by mutual agreement between the parties con-
cerned.
7. Test Methods
7.1 Total Matter—Method B of Test Methods D 1888.
7.2 Electrical Conductivity—Test Methods D 1125.
7.3 pH—Test Methods D 1293.
7.4 Consumption of Potassium Permanganate—Deter-
mine the consumption of potassium permanganate by
adding 0.20 mi. of KMnO4 solution to a mixture of 500 mi.
of the reagent water and 1 mL of H:SO4 in a stoppered bottle
of chemically resistant EI!"« Consider the reagent water as
having passed the test if the permanganate color does not
disappear completely after standing for the indicated period
of time at room temperature. This test should be run against
a blank using water known to be free of organic substances.
7.5 Silica—See Test Method D 859.
7.6 Clinical Reagent [Voter—The total count of viable
bacterial organisms shall be determined by Methods F60.
Th» Amtnem SOCMV tor Testing tna Mmnta omt no oeamon nsotcmg tn* vwotv ot m ptuot nona isstma m eomteoon
wrtftwirnm nwMiiMu n mo ntnatra. Uan a ms sunatra trm •jrorasw KMSM oat atmiiaauon at tM *
ilnetn*Ma. MMfntegmna orwnnonwn. Yourcomrnnan^neatnnvmrwmanatnasanatraortorvjoaianutttnaim
•^yy* l»^o
-------
Standard Test Method for
Water in Volatile Solvents (Rscher Reagent TOration Method)'
This staadini H M»i»4 .....< ,.-<- . .
r '"«•*••'» fe"°*«»f *• tewaanon mourn tne vear o«
n™
«-'S38S==3±£
m$*fa*\ editomiiv in Decetnoer
1. Scope
- •
2. Referenced Document
2.1 ASTM Standard:
DI500 Test Method for ASTM Color of Petmi«,«
Products (ASTM Color Scale)2 Petroleum
3. Descriptions of Terms Specific to This Standard
3.1 instrumental end
3L2 color end powr-During the titration, the
edinoo approved Nw. 27. 1917.
.364 . 5 J T.
wi ru:
' °n*IBUy
4. Summary of Test Method
4.1 This test method is based essentially upon the reduc-
tion of iodine by sulfur dioxide in the presence of water. This
reaction can be used quantitatively only when pyridine and
an alcohol are present to react with the sulfur trioxide and
hydriodic acid produced according to the following reac-
tions:
H,0 + I, T SO, -r 3CfH.N — 2C.H.N-HI + C«H5N-SO, •
C,H,N-SO, + ROH —< ,_„,
4.2 To determine water, Fischer reagent (a solution of
iodine, pyridine. and sulfur dioxide, in the molar ratio of
1 -HO+3) dissolved in anhydrous 2-methoxyethanoi is added
to a solution of the sample in anhydrous pyridine-ethyiene
glycoi (144) until all water present has been consumed. This
is evidenced by the persistence of the orange-red end-point
colon or alternatively by an indication on a gt'lvanornrw or
similar current-indicating device which records the depolar-
ization of a pair of noble-metal electrodes. The reagent is
standardized by titration of water.
5. Sipntfio^n^ atflj Use
5.1 Volatile solvents are used in a variety of chemical
processes which may be affected by water. Therefore, this test
method provides a test procedure for assessing compliance
with a specification.
6. Apparatus
6.1 Titration Vessel—For color end point titrations. use a
100 or 250-mL volumetric flask which need not be cali-
brated: a 250-mL flask fitted with interchangeable electrodes
(rig. 1) may also be used for the instrumental end point and
aparncularly suitable for titrations at ice temperatures. For
permanently mounted assemblies, the vessel should have a
capacity about equal to that of a 300-mL tall-form beaker
and be provided with a tight-fitting closure to protect the
sample and reagent from atmospheric moisture, a stirrer. and .
a means of adding sample and reagents and removing spent
reaction mixture. It is desirable to have a means for cooling
* 1°?^°° vessd to ice temperature.
6.2 Instrument Electrodes—Pktinum with a surface
equivalent to two No. 26 wires 5 mm long. The wires should
be 3 to 8 mm apart and so inserted in the vessel that 25 mL
of liquid will cover them.
1 Such flasks are made by Raakin CU
Road. Manmex. CA.
i fllowint Co- J920 Franklin Canyon
Page 38
-------
D1364
CONNECTOR
SLECTROOE
" LEADS
250 mt
VOLUMETRIC
FLASK
(PYREX1
INSULATED
LEADS ~
T ELECTRODES
3 TO 4 mm. APART
Note—All dimensions n manning.
FKSL 1 Tttratton RMK A«m«molv
6.3 Instrumen: Depolarization Indicator, having an in-
ternal resistance of less than 5000 fi and consisting of a
means of impressing and showing a voltage of 20 to fo mV
aoos we electrodes and capable of indicating a current How
of 10 to 20 uA by means of a galvanometer or radiotunme
eye circuit.4 B
6.4 Buna Assembly for Fischer reagent, consisting of a 25
or 50-mL bum connected by means of glass (not rubber)
connectors to a source of reagent: several types of automatic
dispensing burets5 may be used. Since the reagent loses
strength when exposed to moist air. all vents must be
protected against atmospheric moisture by adequate dryine
tubes containing indicating calcium suifate drying agent. All
stopcocks and joints should be lubricated with a lubricant
not parnculariy reactive6 with the reagent.
6.5 Weighing Bottle, of the Lunge or Grethen Type or
equivalent. ^ ur
7. Reagents
V • fta^y °fRea^ents—^gtteaa. grade chemicals shall be
used in all tests. Unless otherwise indicated, it is intended
that all reagents shall conform to the specifications of the
Committee on Analytical Reagents of the American Chem
ical Society, where such specifications are available7 Other
grades may be used, provided it is first ascertained that the
" Mvpesimiiar w we ft«aon Scientific Co. 'Aquatraior.- or Fisher Sciennfic
Co. -Fhher r,(nm«er.- ,5 smable for the measurement of the inswimeTena
poinu
> A type similar to Cauloi No. J-821 ot Snenuiic GUu Apparatus Co
Bloomiield. NJ. or Catalog No. 750 or Eck and Kietos. New York. NYhas h-,'
specuicalWdesigned for tius purpose ana presents me minimum contact 01 reagent
with stopcock lubneant. «=>»em
- • Suitable lubricants are Apiezon N (James G. Biddle and Co Philadehihu
PAk High Vacuum SUicone Crease lOow Corning Co- Midland. Mlk SiscoJOO
(Swedish iron and Sled Co. New York. NY\ ""taueowo
Soc. wSnloC Fors^ons on ^^^l
American Chenucai Sooert. see -Rea*e« Chemtcaa an
Rosm. D. Van Nostnna Co. Inc. New York. NY.
Pharmacopeia.
reagent is of sufficiently high purity to permit its use without
lessening the accuracy of the determination.
7.1.1 Fischer Reagent t equivalent to 6 mg of H;O per
mL)*—For each liter of solution, dissolve 133 ± I g of I; m
425 ± 2 mL of anhydrous (less than 0.1 % H;O) pyndine in
a dry glass-stoppered bottle. Add 425 ± 2 mL of anhydrous
(less than 0.1 % H2O) 2-methoxycthanol. Cool to below 4 C
in an ice bath and add gaseous SO:. dried by bubbling
through concentrated H:SO« (sp gr 1.84); determine the
amount of SO: added by mcasunng the change in weight of
the SO; cylinder (102 ± 1 g) or the increase in volume (70 ±
1 mL) of"the reagent mixture. Alternatively, add 70 mL of
freshly drawn liquid SO, in small increments.
7.1.2 Solvent Mixture—Mix. 1 volume of anhydrous (less
than 0.1 % H3O) pyndine with 4 volumes of anhydrous (less
than 0.1 ?S HJO) ethyiene glycol.
NOTE 3—Pyndine. ethyiene glycol. and 2-raethoxyetfaanoi. each
containing iess than 0.1 % water, are available and should be used.
NOTE 4—If adequately dry reagents cannot be procured, they can
each be dried by distillation through a multiple-plate column, discarding
the first 5 % of material distilling ovemead and using the 9S %
rcmainmz. Drying may also be accomplished by the addition of t
^olume or benzene to 19 volumes ot' the pyndine^lycol. or of the
pynoine emvtene giycoi raonomethyi ether mixture, followed by disul-
iation: me lira 5 "a distilled is discarord ana the residual 95 % is useo.
8. Standardization of Reagent
8.1 Standardize the Fischer reagent each day used by
either the color or instrument end point (Section 2) method
using the same procedure as used for titrating the sample.
8.1.1 Add to each flask 25 mL of ethyiene glvcoi-pyridine
mixture, and titrate this as described in Section 9. Add to the
solvent thus titrated, in place of the sample. 0.15 to 0.18 g of
water from a weighing pipec weighed to the nearest 0.1 rag-
Complete titration with Fischer reagent as described in
Section 9.
8.1.2 Calculate the equivalency factor. W. of the reagent
in terms of water content per miiiilitre by means of the
following equation:
WmAjV
where:
W ** equivalency factor, mg of water per mL.
A ** weight of water used in the standardizauon. mg, and
V ** volume of Fischer reagent required. mL.
9. Hazards
9.1 This test method includes the use of a U.S. Occupa-
tional Safety and Health Administranon (OSHA)-designated
hazardous chemicals, iodine, pyridine. sulfur dioxide, and
2-methoxyethanol. For specific hazard information and
guidance, consult the supplier's Material Safety Data Sheet.
10. Sampling
10.1 Lacquer Solvents—It is essential to avoid changes in
the water content of the material during sampling opera-
tions. Errors from this source are particularly significant in
the analysis of materials having low water content, when
analyzing materials that absorb water readily (for example.
absolute alcohols), limit as much as possible contact with air
in transferring the specimen into the titrauon vessel. Avoid
intermediate sample containers, if possible.
1 Karl Fischer Reagent is available from ranous laboratory suppuen. fvndine.
frecadaptauons ne aviuable and mav be usea if precision can be cnaolishfrt One
lueh mature B HYORANAL a indemsn of Riedette Haen AC. distributed by
Fisher Soemntc/AUied. 711 Forbes Ave. Pittsburet. PA 13219.
Page 39
-------
4SD» D 1264
* roccoi
11.1 Introduce 10 to 25 mi. of the anhv*
mixture fpyridine-givcoL I-M* ?„,« ?•. anfaVdro«» solvent
making sure/if anTrSmrat end Sim an^f*"011 vcsscl-
that the eieetrodes are covered 3* 1S
pomt is to be
end point (see 4.1), or the coo
Fischer reagent. If the color end B
titrate one flask to the orange-red end DOSr ? 5e,.otaerved-
to match the first. Set as^the fiS £ L ^ second
standard for titrating tfaeipecimen^ a comPans°n
when the specimen is
absorbed from the air.
humidity. Again, titrate the
the same instrumental or coloi
ployed. Record the amount $ '
water in the specimen.
NOTE 5f a) Kama— In among
care
1S-
" ot
r reaeat to
NOTE 7— In tmanng with
12. CalcnJadon
C0ntent of
as
wnere:
fof titrating by the
^-equivalency factor for Fischer
nullffitre of reagent, mg, and
•^ « specimen weight vied, g.
13. Report
13.1 For water concentrations below 0.5 «. repon ail
TABLE 1 R«
ISJZM
W
MMMT Soaonwi
Tatan
2Jto 10
0.5 to 2.5
BatowO.S
0.18 9 at H,O
lOim.
25 mt
0.15 to
WOMt
SMW is tor 0 J to i5
0 °-001 ^' Two determinations which agree withio
. absolute, are acceptable for averaging (95 % conn-
aence level).
14. Precision and Bias9
14.1 On the basis of an interiaboratory study in which one
operator made duplicate determinations in "each of eight
aitterent laboratories, on two days, on samples of acetone
containing 0.118 and 0.406 % water and of methyl ethvl
ketone containing 0.050 and 0.176 % water, the following
criteria should be used for judging at the 95 % confidence
level me acceptability of results on samples containing less
than u.o % water
1 4. 1 . 1 Repeatability— Two results, each the mean of du-
plicate determinauons. obtained by the same operator
n n,uJdflrbe consjde«a suspect if they differ by more than
0.015 ~o absolute.
14.1.2 Reproducibiluv—Two results. <^rh the ™?*n of
duplicate determinauons. obtained by operator in different
laboratories should be considered suspect if they «ftflfcr by
more than 0.027 7o absolute.
^ In another interiaboratory study, one operator in
each of seven different laboratories, on two different days,
made duplicate determinations on five randomly coded
samples of hexyl acetate containing 0.015, 0.034, 0.05Z
u.u/r, and 0.098 % water. One of the seven laboratories used
toreedifferent equipment or procedural variations, or both.
•"^bymaking **"« in effect, a nin? interiaboratory study.
The statistically designed study covered a variety of equip-
ment. analytical methods, and reagents »«pg the Karl
Fischer chemistry. ^^
14.2.1 Repeaubility—Two results, each the mean of du-
plicate determinations, obtained by the same analyst should
beconsidered suspect if they differ by more than 0.030
Reprodudbiiity—Two results, rarfa of the mean of
duplicate determinations, obtained by analysts in different
laboratories should be considered suspect if they differ by
more than 0.060 vT % absolute.
14.3 Bias has not been determined for this test method.
Refer to Notes 5, 6, and 7 for factors that might influence
test bias.
{nm
Reqim Rfc
'*'*'~
or tort
«f« among at trtt nmonuo*
/raw your
Page 40
-------
Designation: 0 3728 - 88
Standard Specification for
2-Ethoxyetnyi Acetate (99 % Grade)1-2
This standard is issued under the fixed designation O 3728: the nuraoer immediately tbilowinj the destination indicates me year 01"
onanat adoption or. m the case 01 revision, the year 01 last revision. A number in parcmaeaes mdicaia tne year 01 last reapprovai. A
supencnpt epaion it) indicates an editorial chance nncc the lass revision or reapprovai.
1. Scope
'1.1 This specification coven the properties of 99 % grade
2-ethoxyethyi acetate.
1.2 This material may be suitable for use in uretfaane
coatings, providing that the water content is acceptable.
1.3 For specific hazard information and guidance, see the
supplier's Material Safety Data Sheet for materials listed in
this specification.
2. Referenced Documents
2.1 ASTM Standards:
D268 Methods of Sampling and Testing Volatile Solvents
and Chemical Intermediates for Use m Paint and
Related Coatings and Material'3
D1078 Test Method for Distillation Range of Volatile
Organic Liquids3
D1209 Test Method for Color of Clear Liquids (Platinum-
Cobalt Scale)4
D1296 Test Method for Odor of Volatile Solvents and
Diluents3
D1364 Test Method for Water in Volatile Solvents
(Fischer Reagent Titration Method)3
D1613 Test Method for Acidity in Volatile Solvents and
Chemical Intermediates Used in Paint. Varnish. Lac-
quer and Related Products3
D3545 Test Method for Alcohol Content and Purity of
Acetate Esters by Gas Chromatography3
E 1 Specification for ASTM Thermometers2
E 300 Practice for Sampling industrial Chemicals3
22 US. Federal Specification:
PPP-C-2020 Chemicals. Liquid. Dry, and Paste: Packaging
3. Properties
3.1 2-Etnoxyethyl acetate (99 % grade) shall conform to
' This specmeanon is under the jtmsdtction of ASTM ConumneeD-l on Paint
and Related Coannts and Materials and u the direct responsibility oi Suborn-
mnee DOIJS on Solvent. Hatocnm. and Chemical Intermediaies.
Current edition approval March 25. 1988. Published May 1988 Oritmaily
published as O 3728 - 79. Las previous edition O 3728 - 84.
: Also known as cthytene fjycoi monoeuivt ether acetate (EGMEA)
1 Annual Book of ASTM Stonoants. Vol 06.03.
' Annual Book of ASTM Stonaardi. Vols 06.01 and 06.03.
' Annual Book of ASTM Stanaards. Vols 03.03 and 14.03!
* Naval Publications and Forms Center. iSOl Tabor Ave_ Philadelphia. PA
19120.
the following requirements:
Punty wettnt %. nun
Alcohol (as .2-ethoxv etnanol)
. max
Apparent Tpmnr irtvitv.
20/20-C
2J/2TC
Distillation range. 760 mm rig
Bdow UO'C
Above i60"C
Acidity as acetic acid.
weifnt %. max
Water, weifnt S. max
Color Pt-Co icaie. max
Odor
99.0
0.3
0.973 to 0.976
0.969 to 0.972
none
nooc
0.02. equivalent to 0.19 mi oi KOH
per (ran oi ti'iiiMf
0.10
13
nonresiduai
4. Sampling
4.1 The material shall be sampled in accordance with
Practice E 300.
5. Test Methods
5.1 The properties enumerated in this specification shall
be determined in accordance with the following ASTM test
methods:
5.1.1 Purity and Alcohol Content—Test Method D 3545.
5.1.2 Apparent Specific Gravity—D&crnaac the apparent
specific gravity by any convenient method that is accurate to
the third -Wma' place, the temperature of both speamen
and water being 20 or 25'C. See Specific Gravity section of
Methods D 268. tft_e .
5.1.3 Distillation Aange-Tcst Method D 1078. using an
ASTM Solvents Distillation Thermometer 102C having a
range from 123 to ITTC and conforming to the require-
ments in Specincation E 1.
5.1.4 Acidity—Tat Method D 1613.
5.1 J Water—Test Method D 1364.
5.1.6 Color—Test Method D 1209.
5.1.7 Odor—Test Method D 1296.
Non-Due to the low votatility of 2-ethoxyethyi «cewe. conader-
able time (more than 30 min> may be required to detenmne restairn
odor adequately.
6. Packaging and Package Marking
6.1 Package size shall be agreed upon between the pur-
chaser and the supplier. .
62 Packaging shall conform to applicable earner rules
and regulations or when specified shall conform to red.
Spec. PPP-C-2020.
Page 41
-------
D3728
n»*iMncMtecM«vMr temp*
Page 42
-------
Designation: 0 3925 - 81 (Reapproved 1985)*'
Standard Practice for
Sampling Liquid Paints and Related Pigmented Coatings1
This sundnd is issued under the fixed 4
ontmaiuoooonor.
toe vtv of
WM eonomily etaaya in Septemoer 1985
*1. Scope
l.l This practice describes methods of taking representa-
tive samples of fluid paint or pigmented coating products
from containers of any type.
1.2 The sampling of dry powder paints, clear coatings.
mixed solvents, and nonpigmented materials of any type is
not covered in this procedure.
1.3 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the satiety problems associated with its use. It is
the responsibility of whoever uses this standard to consult ana
establish appropriate safety and health practices and deter-
mine the applicability of regulatory limitations prior to use.
2. Referenced Document
2.1 ASTM Standard:
D1475 Test Method for Density of Paint. Varnish. Lac-
quer, 3"d Related Products*
3. Description of Term Specific to This Standard
3.1 batch—(he quantity of liquid paint or coating pro-
duced in the final mixing operation after all production
processes are coropiott, For example, when a number of
pigment dispersions are reduced with additional vehicle
together in a large tank, the resulting final im>tuzt.is.jone
batch."- •'•"•''..:::-~.''%^"'! '"..."-' •- ' ' • '••' •'""•
4. S|«fn{fl5*«itM» and Use
4.1 Samples are taken from batches, lots, and shipments
of paint in order to dcifi'"*"* their uniformity and compli-
ance with specification requirements. It is very important
that these samples be of convenient and economical size and
that they be representative of the batch of paint at the time it
was filled into shipping or storage containers.
4.2 The time and effort necessary to ensure that the
sample is representative of the original material will be
repaid in reduction of laboratory work and elimination of
possible rejections of acceptable material.
5.. Sampling Considerations
5.1 The use of common sense and good judgment is
* This practice is under the junsdienon of ASTM Committee D-l on Mm and
&*i*tr*t Coamtjand Mnenis and a the data respassiadur of Subeonumnee
001.20 on Sampum. Smaoo. etc.
Cuneni edition approved July 31.1981. Published September 1981. OritjnaUv
published as D 3925 - 80. Las prevwn edmm D 3925 - 80.
1 Annuat Book of ASTM StanOana. Vol 0641.
important even in the apparently simplfc task of taking
samples.
5.2 Use care to assure that all container! agitating equip-
ment, and sampling apparatus are clean and that they can in
no way contaminate the sample being taken. CTght contam-
ination of the paint sample may lead to false test results.
5.3 The sample container should be dry and not cooler
than the temperature of the area in which the «""p>e is to be
taken.
5.4 Because pigmented coatings are dispersions and not
solutions, finely divided pigment particles disturb in the
coating vehicle may settle upon standing. Consequently,
thorough atrd careful agitation before
-------
put ail containers from the same batch together. From each
batch select at random 1 % of. but not more than five
containers, using the next larger whole number if a fraction
results. For example, if there are 275 containers in a batch.
select three for test.
6.2.1.2 After selection of the filled, unopened containers.
thoroughly agitate or stir the contents by the best means
available. Acceptable methods of mixing are mechanical
shaking or stirring or hand stirring with a paddle, followed by
-boxing," that is pouring back and forth between the original
and a clean empty container. Mechanical shakers are desir-
able for most materials since there is thorough agitation in a
closed container. To prevent evaporation, agitate in a closed
container lacquers and other coatings containing a highly
volatile solvent.
6.2.1.3 Before mechanical shaking, open the container
and check to be sure that the pigment has not caked on the
bottom of the container. If this condition exists, stir manu-
ally to break up the hard settling and then put the containers
on the mechanical shaker. Agitate paints having a weight per
gallon of 11 Ib (1.3 g/mL) or less on the shaker for 10 mm
and those with a weight per gallon of more than 11 Ib for 10
mm.
6.2.1.4 Once the contents have been thoroughly agitated.
pour half of the material into an empty container and take a
1 qt (1 U (or smaller if taken from a container of 1 qt or less)
sample from each half. Determine the weight per gallon on
each sample in accordance with Test Method D 1475. The
two determinations should not differ more than 0.5 %. If the
results differ by more than this the paint is not thoroughly
mixed. Return the material to the original container, stir
again, and repeat the test.
6.12 Containers Larger than 5 CaL
6.12.1 30 and 55-Gal Drums—From each batch select at
random 5 % of but not more than three containers. Drums
may be stirred satisfactorily by several means. With open-
head types, mechanical or manual stirring may be ysfd,
Some drums contain their own agitators: drum shakers or
rollers may also be used. After thorough agitation, take
03925
samples from the top and bottom of the drum and compare
weights per gallon as described in 6.2.1.4.
6.222 250 to 500-Gal Containers (Tote Tanks}—Select
at random 25 % of all containers for test. Take samples from
top and bottom of the container and compare weight per
gallon determinations as described in 6.2.1.4.
622.3 Tank Wagons and Tank Cars—Sample each com-
partment of the wagon or car. Pigmented paints and coaungs
packaged in large containers are generally formulated to be
essentially nonsettling. Therefore, take samples from the top,
middle, and bottom of the container and make weight-per-
gailon determinations before any vigorous stirring is done. If
the resulting tests tall within the limits described earlier, no
further agitation is necessary. Samples may be obtained with
a Bacon-bomb sampler or a "thief" apparatus.
6.3 Pigmented Coatings Containing Water:
6.3.1 Handle pigmented coatings containing water (latex
paints, etc.) in a slightly different manner from solvent-
thinned coaungs. Water-thinned paints, if stirred too vigor-
ously, have a tendency to incorporate air bubbles, which
sometimes result in changing the physical properties of the
paint.
6.3.2 With the above consideration, take the samples in
accordance with the same general procedure outlined in 6.2
for paints containing organic solvents. If it has been neces-
sary to shake, stir, or agitate a water-thinned paint vigor-
ously, deaerate the samples before the weight-per-gallon tests
are run.
6.4 Sampling from Tanks at the Factory—Mix the mate-
rial in the tank thoroughly before completely filling two 1-qt
(1-L) containers. If the containers are to be filled from a
valve on the bottom or side of the tank, mix the material
draw off at least 5 gal (20 L) through the valve and mum to
the tank before taking the sample.
6 J Sampling During Filling of Containers at the Fac-
tory—After the material is thoroughly mixed in the tank and
filling of containers has commenced, take a i-qt (1-L) sample
when about 25 gal (100 L) have been filled and another when
about 25 gal remain to be filled.
•^^^
Page 44
-------
Designation: 0 3980 - 38
Standard Practice for
Interiaboratory Testing of Paint and Related Materials1
under the toed donation D 3980: the numoer rmmemateiv ibUowm. we deacnation indicate* ite year of
«»» . ^"CaSC°' revtston-«ne vear of las ttvwon. A numoerm twenuesainoicaies we vearor last naoprovai. A
epsioa .„ mdicata an eaitanal change since tne last tevmon or reappwvai.
1.. Scope
1.1 This practice covers the interiaboratory testing of
paint and related materials. The information presented is
intended to assist task groups in the effective planning of
round-robin test programs.
1.2 The various parts appear in the following order
Pin A—Definitions
Pin fl—Planning
Part C—Statistical Analvns
Pan O—Analysis Presentation
Sections
3
•i to II
11 to 16
17 «>:o
1.3 This practice, in spue of its lenzm. is not intended to
cover ail aspects of statistical design and analysis For
example, the analysis of variance used" in the practice as the
basis for estimating the precision is abridged in comparison
to that normally used for establishing the significance of
experimental factors. The publications listed in the appendix
should be consulted for runner information or explanations
F™nVCnienCe- comParaoie Practices developed bv other
ASTM technical committees are also shown.
2. Referenced Documents
2.1 ASTM Standards:
D968 Test Methods for Abrasion Resistance of Organic
Coatings by Falling Abrasive2
D 3793 Test Method for Low-Temperature Coalescence of
Latex Paint Fflms-
E 180 Practice for Determining the Precision of ASTM
Methods for Analysis and Testing of Industrial
Chemicals^
PART A—TERMINOLOGY
3. Definitions
3.1 The statistical terms used in this practice are defined,
as nearly as possible, in every day language. For convenience
the terms are listed in order of use in this practice.
3.2 experimental design—ihe complete plan for con-
ducting an investigation or an interlaboratorv studv of
materials, processes, or test methods. It is conveniently set
forth m the form of a single comprehensive table showine
number of laboratories, materials to be tested, test methods!
number of operators, testing conditions, time of testing, and
' This practice is under the nimdienon or ASTM Commnw O-l on Paint and
u - — — - -» -«— s
2 Annual Book 01 ASTM Stonaam. Vol 06.01.
1 Annual Book of ASTM Stanaarat. Vol
degree of replication and repetition. The experimental design
should provide the basis for an efficient statistical analysis of
the results (see Pan 3. Planning;.
3.3 replication—the execution of multiple chemical or
physical determinations on the same specimen or physical
measurements on the same panel at the same time, as closely
as possible. Two replicates are called duplicates, three are
called triplicates, etc. The individual results are not suitable
for esumaung precision but are used only to ensure by
comoanson with similar measurements that there is not a
gross error among them (see 15.1.3.3).
3.4 repetition—The execution of multiple physical mea-
surements on the same panel or chemical determinations at
different times, or physical measurements on- different
panels, areas of large structures, or specimens of a liquid
sample within a short time interval, to establish the precision
of a method. In most cases the measurements are the mean
of replicates as defined in 3.3.
3.5 population—the totality of observations on or deter-
minations of a certain property or component obtained by
the same procedure: theoretically, an infinite collection of
measurements on a given item of interest but practically, a
large number of measurements of the item. Thus, when
statistical procedures are used to determine whether or not
two (or more) materials or test methods differ significantly
with respect to some measurable property, in effect, the
determination is whether these materials belong to the same
or to different populations.
3.6 sample—A randomly selected subset of a population
intended to be representative of it so as to enable obtaining
an estimate of the property or composition of the entire
population. The reliability of such an estimate can be
expressed in terms of confidence limits (see 3.19). Through
tests of significance (see 321}, it is possible to state, with a
specified degree of confidence, whether two or more samples
are drawn from the same or from different populations.
3.7 average—a typical numerical value that attempts to
summarize or reflect the location of a group of observations
by a single number. While it is a measure of central
tendency, it does not provide information on the variability
of the individual observations. The following are different
types of averages: arithmetic mean, weighted mean, algebraic
mean, geometric mean, harmonic mean, median, and mode.
3.7.1 mean (arithmetic)—the value obtained by dividing
the sum of a ser of observations or results by their number.
This value is an estimate of the mean of the parent
population. Although the arithmetic mean is affected by
extreme values and therefore may not be typical, it is
amenable to statistical treatment and is the most commonly
used average.
Page 45
-------
, Jt'I'2 medlan—\t\e c««™ observation in an ordered set
that contains an odd number of observations or theinean of
the two central observations in an ordered set w2 ST
number or observations, The median is not distorted
extreme values so that where the observation? "
1S a raorc
3.8 expected vdtc^-the mean of an infinite series of
independent determinations on the same item obtained bv
the same procedure: also thought of as the true value that
would be obtained if all extraneous variations were euS
nattd. In practice, it can be closely approached bv a finite
series ot tests, the number of tests needed being dependent
on the extent of variation (scatter) of the observed vaiues.
19 accepted reference vo/ue-a value that serves as an
f re"n" for measured values. It is derived as a
3Sed °n Sdentific PrindPi« or an assigned
OMnte0°nnaieXPenmentai W°rk by (a) COmPeKn< «no5
or international organization! s).
NoTE-When the accept* reference value is me theoretical value u
is somenmes reienea to as the "true" value. ««™w vaiue. it
3.10 accuracy- the degree of agreement of indimuai or
mean measurements with the expected or accepted reference
3.1 1 error—the deviation of a measured vaiue or grouo of
measured vaiues trom the expected vaiue or accepted refer-
cnce value.
3.1 1.1 random error— the chance variation that occurs in
all experimental measurements despite the closest nossible
control of all factors. It is characSzed by SfaSSS
occurrence of both poanve and negative deviations from the
mean or the expected value, The algebraic mean of Se
deviations is zero m a large series of measurements.
3.1 12 bias— a systematic as opposed to a random error
that contributes to the difference between
n
mean of the population and the accepted refereacem
true vaiue.
3.12 dispersion— the variability (scatter) of the observed
values, usually measured about some central vaiue sucfaas
the mean.
3.13 range— the difference between the lowest and highest
values m a set of observations or results. The rangeiH
simple but useful indicator of the variability of test results.
3.14 oudier— an extreme value far enough from other
results m a series to be suspected of not belonging to the
particular population under consideration. Statistical criteria
are available for judging whether a given outlier shouldbe
included in the analysis of results or discarded (see Sewi™
13). Also called "wild," "rogue." "maverick."
3.15 variance—*, measure of the dispersion of a series of
results around their mean. The variance of the narem
population is estimated by summing the squares of the
individual deviations from the mean and dividing by one less
than the number of results.
DISCUSSION— Since the variance of a set of results and
the estimated variance of the parent population are not equal
because of different divisors (n for the set. n - I for the
population), the same symbol should not be used for both
3.16 standard deviation— A measure of the dispersion of a
series of results around their mean, defined as the positive
square root of the variance.
Page 46
03980
DISCUSSION—The advantages of the standard deviation
are that it is in the same units as the original results and is of
the same order of magnitude as the deviations from the
mean. The standard deviation is the basis for most state-
ments of precision and may be obtained from an analysis of
variance of results of an interiaboratory test program.
3.16.1 pooled value—the weighted mean obtained by
combining in accordance with statistical rules two or more
separate vaiues shown to be members of the same (homoge-
neous) population. Variances, if shown to be homogeneous
by appropriate statistical criteria, may be combined by
weighting each variance in accordance with its degrees of
freedom, summing them and dividing by the sum of the
degrees of freedom. Standard deviations must not be arith-
metically averaged: they must be squared to convert them to
variances, combined, and the square root taken to obtain the
pooled vaiue.
3.16.2 coefficient of variation—a measure of relative pre-
cision calculated as the ratio of the standard deviation to the
mean of a series of values, expressed in percent
3.17 degrees of freedom—in a set or subset ofobserva-
tions, the number of values minus the number of constraints.
In general, there is only one constraint—the number of
values. For example, a set of n observations can be arranged
into g groups. The group degrees of freedom is then g — 1
because n is already determined. Similarly the total df is
n - 1 (see Table 8).
3.18 probability—Ac chance of occurrence of an event
expressed in terms of a relative frequency, a fraction, or a
percent. For example, the probability that a tossed coin will
land head up is one in two. or 0.5, or 50 %.
3.19 confidence limits—the limits on either side of the
mean value of a group of observations that will, in a stated
fraction or percent of the cases, include the expected vaiue.
Thus the 95 % confidence limits are the values between
which the population mean will be situated in 95 out of 100
cases.
3.20 confidence level—the probability level (usually with
reference to a statistical table) with which the significance of
differences between measurements is asy^v* Thus, a differ-
ence that is significant at the 95 % level (0.05 probability
level, sometimes called significance level) would represent a
real difference 95 times in 100: however. 5 times in 100 a
difference this large might arise by chance even with identical
material because of experimental error.
3.21 significant difference—a difference between two
values, means, or variances that is shown by tests of
significance to be a real difference at the stated level of
confidence.
321.1 Student's i test—a statistical test for assessing the
significance of a deviation from the mean or of the difference
between two means. The /-value is based on the ratio of the
observed deviation, or difference, to the standard deviation
and is compared with tabulated /-values that indicate the
frequency with which a difference of this magnitude should
occur by chance in samples having the appropriate degrees of
freedom.
3.21.2 variance ratio fF) test—a statistical test for as-
sessing the significance of the difference between two or
more variances. The F test (named in honor of R. A. Fisherj
is based on the ratio of the larger variance to the smaller
-------
popujauon- In a complete analysis of vana
. facwrs and
results in
3.22 conaam
means and
323
eR results
amons
which the
324 aww o/vanance-a systematic statistical oroce
dure for detenmnmg the sources and the nanS* 01 £e
aion present m ^measurement process and for assess^ he
1 °
called least
., as the repeatability and reproducibih^f^ £
"^00 of a '« niethod ex
in terms of the agreement
measurements may influence the repeatability
and a value
when
no
exact
ility that there is a conetation (see Table
328 wwwv/rv*-theabiKtyofatest
PARTS— PLANNING
•*• Scope
4.1 This pan covers simplified statistical
*
5. Purpose
5.1 The purpose of an interiaboratory evaluation is to
determine any or all of the following: «H«wn is w
03980
5.1.1 The sensitivity of a test method for discriminating
between materials that are known to differ in the property to
be measured.
5.12 The variability of results obtained in different labo-
ratories by different personnel on the same t>pe of equip-
ment using a prescribed test method.
5.1.3 The consistency of this variability tor duTerrnt
materials, and
5.1.4 The comparative merits of two or more alternative
test methods.
6. Problem Formulation
6.1 The objectives of the task group should be clearly
formulated before interiaboratory evaluations are initiated.
Possible objectives for the evaluation of a test method are to
determine the following:
6.1.1 Whether test results correlate with other test
methods in common use for measuring a property,
6.1.2 Whether the test method actually measures the
property intended and supplies adequate.sensuivity in dis-
crimination of materials that are known to differ in perform-
ance.
6.1.3 The limits and fields within which the test method
has value.
6.1.4 The precision of the test method, and
6.1.5 If certain defects of the test method can and should
be corrected.
7. Preliminary Test
7.1 Survey known sources of information related to the
test method to establish how results are affected by variations
in operating conditions, atmospheric conditions, differences
between operators, etc. Select what appears to be the
optimum procedure.
72 Draft instructions for the test method and. without
comment, observe a laboratory T^rriT""*" perform a test
according to these instructions. Revise any pans of the draft
causing difficulty.
7.3 Make a comparative study with other test methods for
measuring the property by using specimens with a wide
range of values of the property under test (and possibly with
wide ranges in other properties).
8. Preparation for Task Group
8.1 Prepare a clear statement of the type of information
the task group expects to obtain from the interiaboratory
evaluation.
82 Based on the study made in one laboratory, prepare a
proposed master plan for the interiaboratory evaluation. Ask
ill members of the task group and other competent authori-
ties (including apparatus manufacturers) for comments on
and criticism of the proposal. Discuss the plan in an open
meeting.
8J Select the materials to be used in the interiaboratory
evaluations so as to:
8.3.1 Cover the applicable range of the property or
component to be measured, and
8.32 Represent ail classes of materials to which the test
method will be applied.
Page 47
-------
D3980
9. Pilot Evaluation
9.1 If the test method is new or the procedure for an old
test methoa is greatly altered, a pilot studv bv one laboratory
involving a lew materials (two or three, mav reveal senouslv
misleading directions in the written procedure.
10. First Interlaboratory Evaluation
10.1 For the first interiaboratory evaluation use at least
three materials to cover the expected range in propenv
values and include all of the laboratories that will participate
in the main mterlaboratory test. This evaluation will train
the participants in the test method. clarirV the procedure
eliminate laboratories that cannot comply with the proce^
dure because of nonstandard conditions or equipment and.
together with the main study, give some idea of the time
constancy ot laboratory results.
11. Main Interiaboratory Evaluation
11.1 Variables-The major variables that can be included
m an interiaboratory evaluation of a test method arc-
materials, laboratories, apparatus, operators, and time ce-
scnoed as lollows:
11.1.1 Materials—A minimum of six materials aitTenne
in the property or component to be measured and covenne
the useful scales of the apparatus are needed to evaluate the
sensitivity of a test method.
11.1.2 Laboratories—lo. the evaluation of a test method.
an absolute minimum of three laboratories, or locales.
shoukl be used but more are generally required, as shown in
11.1.3 Apparatus—If different types of apparatus (or dif-
ferent procedures) are to be included in the study an
absolute minimum of three is required for each type. At least
five are needed to obtain a reasonable estimate of reproduc-
ibuity.
11.1.4 O^rawrr-Obviousiy, at least one operator must
be used at each laboratory or locale. If operator variability is
to be ascertained, two or more operators at each locale must
be used.
11.1.5 Time— Repeatability (intralaboratory precision) is
usually ascertained by having repeats made at different times
or, for physical measurements, on different panels. If consid-
ered desirable, a completely unbiased estimate of inna-
laboratory error can be obtained by use of'blind' repetition.
In this procedure two samples of at least one. and preferably
more, of the test materials are distributed as separate
materials. If all the materials are duplicated, the total
amount of work can be kept the same by not repeating the
test at another time. However, elimination of the time
variable might alert cooperators to the use of blind repetition
which, m any event, is not readily applicable when all the test
materials differ markedly in properties.
11.2 Experimental Design—The design selected for the
interiaboratory study should be based on the example given
in Fig. 1. There is an inverse relation between the numbers of
laboratories and materials required to obtain a reasonable
estimate of precision, while the required numbers of repli-
cates and repeats are related to the variabaity of the test
method! s) being investigated.
11.2.1 For analytical test methods, the precision of which
is usually very good, two repeats each in duplicate are often
sufficient. However, owing to the much more variable nature
of measurements of the physical or resistance properties of
coatings, the repeats should be increased, unless it has been
decided to use a test method mainly for ranking a series of
coatings (Note i). For most test methods of this type, three
repeats should be used but even when the precision is quite
poor the suggested maximum is four. If the intralaboratory
variability is high the number of replications should also be
increased.
NOTE 1 —Examples of test methods m which ranking is preferred are
Ten Methods O 968 and D 3793.
11.2.2 In order to obtain sufficient interiaboratory degrees
of freedom (see Table 6) so that the estimated interiaboratory
precision will not appear poorer than it really is. the number
of laboratories must increase as the number of test materials
decreases. If the six materials recommended in 11.1.1 are
used, then three laboratories are required if one test method
only is under study. The following is the minimum number
of laboratories and materials that should be used in a study
the associated interiaboratory degrees of freedom:
N'umoer of
Materials
^
3
4
5
6
7
Uboratonn
6
5
4
3
3
3
Oetfca of Freeoom id!)
2T6-U-IO
3(3-1) -»2
4(4-1) -12
6(3-1) -12
7(3-0-1*
11.3 Instructions—Use the master plan agreed upon by
the task group after careful discussion. This plan should
include instructions on the following:
11 J.I Care of round-robin specimens and what to do in
case of loss of specimens or results (missing results can be
ignored only if a sufficient number of participating laborato-
ries and materials are included).
11.3.2 Adjustment and calibration of the test apparatus.
11.3.3 Order of testing the specimens.
11.3.4 Recording results on the test form.
11.3.5 Detailed test procedure, to include:
11.3.5.1 Scope.
11 J.5.2 Test method.
1IJJ.3 Other instructions relevant to use of test method
or operation, as for example, replication and standardiza-
tion.
11J.5.4 Dates for performance of tests.
11J.5.5 Instructions about personnel.
1IJ.5.6 Instructions on compilation, calculation, and
reporting,
11.3.5.7 Standard report form for results and conditions.
11.3.5.8 Instructions on return of reports and materials
(including address), and
11.3.5.9 Closing date.
11.4 Allocation of Specimens—Specimens may be pre-
pared in several locales, but should be allocated and distrib-
uted from one place. Prepare from each material enough
specimens to provide the required test material for the
participating laboratories and a sufficient number of addi-
tional specimens for replacement of lost or spoiled speci-
mens. Label each specimen by means of a code symbol and
idemify the specimens on a separate key sheet for future
reference. Completely randomize the specimens of a panic-
Page 48
-------
Fip«nm»ni»i Ocnqn for EviHiMon
Slnfl* MMftoe Utina On* Typ* el Appwnn
1 Eumpto at Dtugn for InMrtabomory Study
ular material before dividing them into groups to be distrib-
uted among the laboratories. Where necessary the same
specimens may be sent in turn to each participating labora-
lory.
Now 2—Test puds ire frequently prepared in one laboratory to
«£** the variety only of .neuuimems ob^nedTSeS
laboratories. Consequently, if pud preparation has a siuificameflect
on ten results, the reroltu* precision wiU be better than where Dane*
are independently prepared by each laboratory. I£ because of difficulty
in obtaining a sufficient quantity, toe some ten panels are circulated to
all pamciputs. the test method may. depending upon the variability of
panel preparation and the sensnvity of the ten method, also annar
more precise than if different panels were sent to each paruapam.
11.4.1 Effect of Aging—If the specimens are such that
their properties may change noticeably in a few days or
*eeks. coordinate the tests among the laboratories so that
each laboratory performs the test on specimens of the same
age.
11.5 Report Form—Supply each laboratory with report
forms hke Fig. 2 to ensure that all results and peranent
information are reported in a uniform manner. In addition
to space for measurement results, the form should provide
space for such information as: relative humidity, tempera-
ture, instrument type, deviations from the specified proce-
dure, unusual observations, and constructive comments, as
required.
PART C—STATISTICAL ANALYSIS OF INTERLABORATORY
TEST RESULTS
12. Scope
12.1 Appropriate statistical methods are described for
computing the correlation, precision, and sensitivity of a test
procedure from interlaboratory test results.
12.1.1 Outliers—For each material, ranges are computed
between replicates run at each time by all laboratories.
between means for each time for each laboratory, and
between laboratory means of all repeats (usually two or
three). Results from the different materials exhibiting similar
variability or having similar mean values can be grouped
and the ranges for each material calculated. For each type of
range, the results are examined for suspected outliers and the
Page 49
-------
0 3980
port pomt f
ASTW fiubcormtiitm
Intwiabontofy Test on
T««k Orouo
Data received
Memoa of aoDacaaon (if reoureo).
F3m micxness uf requreoL
Rim tnioawss metnoa (if reaureal
Test temoerature M^M^.^^.^. Relative numoiiy,
Eouoment oescnotionc
Name
Convfwitst
. Model No.
Name.
.Ufa
* tf more man on* test imtnoa Deng stuoed.
* tf more man on* operator per taoorwory or locate.
RQ. 2 Report Form
est described in 13.3 applied, as illustrated in 15.1.3.4 and
15.1 3 5 to determine whether any values can be rejected.
l2.L2Correlation-.lfiL test procedure is being evaluated
for its ability to provide results that correlate with the known
property values of man-rials, a correlation coefficient is
computed and tested for significance.
12.1.3 Precision—The precision of a test procedure is
determined by computing intialaboratory and interlabora-
tory variances from the test results obtained for each
material. If the variances for the materials are homogeneous.
they are pooled and an overall imraiaboratoryltandard
deviation and an overall interiaboratory standard deviation
are calculated and used in the precision calculations. If the
variances are mhomogeneous. means of providing homoee-
neity (such as use of the coefficient of variation or other
transformations of the results) are given.
12.1.4 Sensitivity—A sensitivity criterion (figure of merit)
is computed and used as a measure of the ability of a test
procedure to distinguish between materials differing in the
property being measured.
13. Test for Outliers
13.1 An outlying observation, or outlier, is one that
appears to deviate markedly from other members of the set
Page 50
in which it occurs. This outlier may be a single value, a
mean, or a range.
13.2 When the experimenter is clearly aware that a gross
deviation from prescribed experimental procedure has taken
place, the resultant observation should be discarded, whether
it agrees or not with the rest of the results. If a reliable
collection procedure is available the observation may be
corrected and retained.
13.3 In many cases, evidence for deviation from the
prescribed procedure consists of the discordant value itself.
In such cases a statistical test is applied to determine if the
doubtful value should be discarded. A simple technique for
this purpose is the rejection quotient procedure. It consists of
(/) ranking the values (single observations, means, or ranges)
in order of their magnitude: (2) taking the difference
between the extreme value and the nearest value to it and
(J) dividing that difference by the appropriate range of the
values. The resulting quotient Q, is compared with the
established rejection quotient (RQ) for the number of items
in the series of values. If Q exceeds the RQ, the extreme value
is suspect and may be discarded. Table 1 gives the equations
for computing Q and the RQ values for three significance
levels as a function of the number of items in the test results.
It is recommended that outliers be discarded on the basis of
the 0.01 significance level for intialaboratory results and the
-------
TABLE 1 Tabl* for Tsaiutg Eztnm* Values
Nort—<> 0 «nMeoi in* reiKBon ouoMnt. tn« susocct vaiut may M
Nmoar n of „ Sigwheanw Levw4
' VMJM
3
4
5
6
7
8
9
10
11
12
13
14
15
18
17
18
19
20
21
22
23
24
25
26
27
28
29
30
0.10
0.941
0.765
0 m '*"*' or *• ~ *—• 0.642
X. -X, X,- X, 0 .560
0.507
0.544
0303
0 » ** ~ *' or *• ~ **-' 0.470
X_, - X, X.- X, 0.445
0.423
0.563
Q m Xs — x. or *« " x— » o .539
*.-» - *i *„ - X, OJ518
0300
0.483
0.469
0.457
0.446
0.435
0.426
0.418
0.410
0.402
0.396
0.389
0.383
0.378
0.373
0.05
0.970
0.829
0.710
0.628
0.569
0.608
0.564
0.530
0.502
0.479
0.611
0.566
0.565
0.546
0.529
0.514
0.501
0.489
0.478
0.468
0.459
0.451
0.443
0.436
0.429
0.423
0.417
0.412
3.01
0.994
0.926
0.821
0.740
0.640
0.717
0.672
0.635
0.605
0.579
0.697
0.670
0.647
0.627
0.610
O.S94
0.580
0.567
0.555
0544
OS3S
3.526
0.5J7
0.510
0.502
0.495
0.489
0.463
. * Risk or prooattty ot r»)*ctrg a vaM axnme mut.
&05 significance level for interiaboratory results. When
results from several materials with similar property levels are
being analyzed, all the results must be included in the
calculation of Q. Also, the test should be applied only once to
a set of laboratory results. Although two or more values can
be rejected at the same time, the remaining results should
not again be tested for outliers.
13.4 Following is an example of the use of the rejection
quotient procedure for evaluating extreme values:
13.4.1 In the evaluation of the precision of a test method.
a sample was tested within a single laboratory twice on *acfa
of two days by each of two operators for a total of eight
determinations. Differences between replicates for both op-
erators on both days are ranked as follows:
KM*
Diflerence
I 2
0.7 3.4
3
3.9
4 J
4.6 3.2
6
6.0
7 I
6.7 7.0
replicates (the smallest
One of the differences between
difference) appears to be suspect.
From Table 1. for eight items:
Q - (.V, - *,)/(*_, - X,) - (X, - V,)/wr7 - A',)
- (3.4 - 0.7)/(6.7 - 0.7) - 2.7/6.0 - 0.450
To reject one of the eight differences. Q must exceed 0.544 at
the 0.10 significance level (90 % confidence level), 0.608 at
the 0.05 significance level (95 % confidence level), and 0.717
at the 0.01 significance level (99 % confidence level). Since a
significance level of 0.01 should be used for intralaboratory
test results, it is concluded that the difference of 0.7 is not
rejectable and must be included in the calculation of test
precision.
14. Correlation
14.1 To be useful, a test method must provide values for
materials that either relate directly to the known property or
component values for these materials or that relate to the
values from another test method that is related to the known
values. The degree of association between the values of a test
method under study and the values of a standard test method
can be determined by computing the correlation coefficient.
14.1.1 The correlation coefficient, r, may be defined by
the equation:
[lur-Wicr-r~)2P I
where*
X = value obtained for a material by standard test method.
Y = corresponding value obtained for a material by new
test method.
T = mean value obtained for a material by standard test
method.
? = mean value obtained for a material by new test
method, and
n = number of values obtained with each test method.
14.1.2 The reliability of the correlation coefficient je-
pentis on the number of materials tested as well as the degree
of linear relation between the two test methods. It should be
pointed out that a calculated r value represents the relation
of the two test methods only over the range of values
obtained. Thus, if the materials selected for evaluating a test
method differ only slightly in their property or component
levels, the calculated r may not correspond to the correlation
that would be obtained if materials differing widely in these
values had been tested.
14.1.3 The correlation coefficient, r, is dimennonless and
its values range from 1.0, perfect direct relation, to -1.0.
which indicates perfect inverse relation. An r of 0 shows no
relation between the two test methods. Whether a value ofr
between 0 and 1 is significantly different from 0 can be
determined from Table 2 that gives the critical values at
several probability levds as a function of the degrees ot
freedom, n - 2. where n is the number of materials tested.
14.1.4 The correlation coefficient squared, r. caUed the
coefficient of determination) provides additional informa-
tion about the degree of correlation between the two test
methods. It is a useful concept in the sense that r- **P*i°
the fraction of the variation of the dependent variable r mat
may be ascribed to the effect of the independent vanawe Ji.
To claim good correlation exists between two test methods, a
value of at least 0.9 for r is required, since r~ = OJ1 indicates
only 81 % of the variation in r can be ascribed to Ji. 11 a
value of 0.7 or less is obtained for r. the degree of association
is considered to be inadequate for assuming thatthej two
procedures are measuring the same property. It should DC
pointed out that a high value for r does not guarantee that
values of r can be predicted precisely from values of*
14.1.5 An example of the computation of the correlauon
coefficient to determine the degree of associauon_between
values of two test procedures is shown in Table 3. This single
laboratory test consisted of measuring the drying time ot lu
Page 51
-------
D3980
TABLE 2 Critical Values tor Correlation Coefficient r
a*
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0.10
0.988
0.900
0.805
0.729
0.669
0.821
0382
0.549
0321
0.497
0.476
0.457
0.441
0.426
0.412
0.400
0.389
0.378
0.369
0.360
Proomoatv Levet
0.05
0.997
0.950
0.878
0.811
0.754
0.707
0.666
0.632
0.602
0376
0353
0332
0314
0.497
0.482
0.468
0.455
0.444
0.433
0.423
0.01
1.00
0590
0.959
0.917
0.874
0.834
0.798
0.765
0.735
0.708
0.683
0.661
0.641
0.623
0.605
' 0390
0.575
0.561
0.549
0.537
of tTMOom (numoar ot ours minus 21.
samples by two test procedures. The drying time values
shown represent the mean of two replicate determinations.
Since the computed r » 0.853 and r2 « 0.728. it is concluded
that the degree of association is adequate to assume that both
procedures are measuring drying time because r exceeds
TABLE 3 Samp* of Calculating Correlation Coefficient
Si
Samp*
• 2
3
4
5
6
7
8
9
10
v_
mpie
2
3
4
5
6
7
8
9
10
X
73
8
9
83
79.0
KY-tixx
Ti
f
y
73
7
9
8
63
8
9
9
83
793
SY/n)
Dryi
lunoand
'.-••lin
^wnnosr
7
73
9
83
xr
48
5625
56
81
64
39
64
72
81
7225
634.5
ig rime, n
Thu
Bl
X»
49
5625
64
81
64
36
64
64
81
7225
6313
no mo
Janee
7
73
7
9
8
63
8
9
9
83
y«
49
5625
49
81
64
4225
64
81
81
7225
639.75
n jL n
634.5 - '79 x 793/101
' [(631.5 - 624.1X639.75 - 632J3251]n
6343 - 628.05 L 6.4S
' ((7.4X7.72511"* " 57.165"
.-2^1-0853
7361
•0.728
0.765. the critical value at the 99 % confidence level for 10
samples (8 degrees of freedom) in Table 2.
14.2 There may be instances where no actual instru-
mental values are available for the materials tested, only their
relative ranking being known. In these cases, a rank correla-
tion coefficient, r,', can be computed to express the degree of
relationship between the ranking of the A' values of one test
method and the Y values of the other test method.
14.2.1 In computing r,', the results from the two proce-
dures are each arranged in order of performance and each
material assigned a rank. In the cases where two or more
materials have the same rank, each is assigned the value
corresponding to the mean of the ranks that would be
assigned to them if the rankings were sequential By taking
the differences. RD. between the two sets of rankings for the
materials, r/ is calculated from Spearman's equation as
follows:
/V = l - [6KXDf/H(H2 - 1)] - I - {tfZCJUnVOi* - «)]
where:
n = number of materials and
RD = difference between the rankings of a material.
A test to determine if r,' is significantly different from 0 can
be made conveniently with Table 4 in which KRD)1 values
at two probability levels are given as a function of the
number of pairs.
14.22 An example of the procedure for computing a rank
correlation coefficient is shown in Table 5. In this example,
the performance of 10 products subjected to two test
procedures is represented by a relative ranking of their
performance. It is seen from the significance table o(2(RD)
as a function of number of pairs that the value of 1 1 obtained
in this example is highly significant (99 %) since 2(RD)2 is
less than 39 for the probability level of 0.01. It is concluded
that the degree of association between the two test proce-
dures is good.
15. Precision
15.1 The precision of a test method is expressed in two
terms, repeatability and reproducibility. Paragraphs 15.1.1
and 15.12 provide the complete mathematical formulas for
calculating precision, but a shortened procedure for analysis
of results from a balanced d^gn is given in 15.1.3. Where
values have been discarded as a result of application of the
test for outliers in 13.1, the mean of all the values for the
respective determination can be used to replace them to
retain a balanced design. It should be recognized that the
analysis used in this practice is abridged because interactions
between factors are disregarded. Consequently, some infor-
mation that could be obtained from a complete analysis is
sacrificed, as in Practice E 180.
15.1.1 Repeatability (lntraboratory)—Ttst repeatability is
determined from the estimated intralaboratory variances
computed from the repeat determinations made within each
laboratory on each material.
15.1.1.1 For this practice, the equation defining the esti-
mated variance. jw2, (Note 3) within a single laboratory is as
follows:
s
*>
Page 52
-------
TABLE 4 Critical VaUiM et Z(HO)' for tn« Ran* Correlation M«tftod
n
S
6
7
8
9
10
11
12
13
14
IS
16
17
18
•' 19
20
21
22
ProeaoBtv Lav*
0.05
0-40
4-46
12-100
22-146
40-300
61-269
88-352
121-451
163-565
213-C97
272-948
342-1 018
423-1 209
515-1 423
621-1 659
740-1 920
873-2 207
1 022-2 520
0.01
0-70
4-108
10-158
24-216
39-291
58-382
84 488
115-613
154-756
201-919
257-1 103
322-1 310
398-1 540
484-1 796
583-2 077
695-2 385
820-2 722
n
23
24
25
28
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Preeaoff
0.05
1 187-2 861
1 370-3 230
1 570-3 360
1 789-4061
2 028-* 524
2 287-5 021
2 569-5 551
2 873-6 117
3 199-« 721
3 550-7 362
3 926-8 042
4 328-4 762
4 757-9 523
5 213-10 327
5 698-11 174
6 213-12 065
6 758-13 002
7 334-13 986
rvLavat
0.01
960-3 088
1 115-3 485
1 287-3 913
1 475-1 375
.1 681-4 871
1 906-5 402
2 149-5 971
2 414-« 576
2 700-7 220
3 008-7 904
3 338-8 630
3 693-9 397
4 073-10 207
4 476-11 064
4 908-11 964
5 366-12 912
S 853-13 907
6 367-14 953
Us* ol rjo/»—*i tr» oosarvM tot*. 2floy. (or tn« narrow of ramao an
taMar VMM. vt» common a uonhmit lor tn« racawa orooaaaty. Hign v«i
9TTBSI
10 or was man tne nwtr uouar vim*, or eouai to or grtatar mm DM ngnar
one to rwgaova conwawni. tow vaiun to powrv* uaiMUona.
TABLE 5 Examoie ot Rank Correlation Calculation
Rant
nes SuaacMO to
Natural Exposure
CycttCon
1
2
3
4
5
6
7
8
9
10
B
C
A
D
EandF
H
G
1
J
A ana 8
0
C
E
FarxJH
Q
J
I
Pantt Du^fiawn
XMauai axooaw ran*
r-Cyeac axooture ran*
RankoftimaiflOl
Rank oftaranea aouaraa iflOr1
A
3
14
14
223
S
1
14
-04
025
C
2
4
-2
4
0
4
3
1
1
E
54
S
04
025
F
54
64
-1
1
Q
8
8
0
0
H
7
64
04
025
1
9
10
-1
1
J
10
9
1
1
TOM
55
55
0
11
Rank tot* - - 55
'.'-«- f«KRO)*/(na - rt|| - 1 - {(6 x 11VOOOO - 10)J - 1 - (66/990) - I - 0.057 - 0.933
where:
- value obtained in a single determination for
a material by a laboratory,
- total number of repeat determination^ made
on that material by a laboratory,
- mean value for the nK determinations.
« sum of squares (ss). and
nx - I «• degrees of freedom, df.
For ease of computation, the equation is arranged as follows:
.
<«*-«)
NOTE 3—The subscript w. to the variance r. as wetl as Uter to the
standard deviation, s. and the coaficieat of vanauon. v. derives from the
previously used term "within-iaboniory".
13.1.1.2 When the results obtained in the interiaboratory
test are from a balanced design and there are no "lining
values, compute the sums of squares and variances for nrh
material by the analysis of variance technique, described in
15.1 J. If the results are not balanced, the computations are
made on the repeat determinations performed within each
laboratory on each material
15.1.1.3 If. by inspeoion. the inualaboratory variances of
all laboratories for a material appear to be homogeneous
(approximately the same) they can be pooled to give a single
intraiaboratory variance for that material.
15.1.1.4 If the intraiaboratory variances for a mawnal do
not appear to be homogeneous, a statistical test, such as the
Cochran4 or Battles' tests, should be used to determine
whether the variances can be pooled.
15.1.1.5 Pool the variances for all the materials (unless
they appear to be nonfaomogeneous) to give a single mtra-
4Coehon. W. G. and COL G. M- Espmmtfuai Designs, iota Wifcy and
SOUL New York. NY. 1937.
1 Youden. W. J. Sutuuctt Utikods for Chtmitu, John Wilev »«J SOBS, New
YoitNY. 1931.
Page 53
-------
D3980
laboratory variance for all materials. Obtain the intra-
laboratory standard deviation, s^ by talcing the square root
of the variance and report.
15.1.1.6 Compute the maximum allowable difference
(MAD) by multiplying jw by the q factor (Table 6) for the
appropriate number of replicates in each laboratory and the
degrees of freedom (nm materials) x (nL laboratories) x (time
-I). In case test results are missing, compute the nL (t - 1)
degrees of freedom for each material and total them.
15.1.1.7 Examine the results from each laboratory for
each material to establish whether all results are within the
MAD for mtralaboratory determinations. If this range is
exceeded and outliers were not previously rejected in accord-
ance with 13.3. discard the discordant results and recalculate
the variance, standard deviation, and MAD. Compare with
the original values to establish whether the rejection is
justified: for example, a marked reduction in MAD with the
eiiminauon of only one set of results is evidence that they
differ significantly from the other results and their retention
would adversely affect the precision of the test method.
15.1.1.8 If the mtralaboratory variances for the materials
do not appear homogeneous but appear to be dependent on
the level of the mean value of the materials, homogeneity
may be achieved generally by convening the standard
deviations to coefficients of variation:
vw- \OOsjX
15.1.1.9 If the coefficients of variation are homogeneous.
pool them in accordance with the equation, as follows:
j"(nt - I) vi2 + * nm, this condenses to (Z v/rtm)'/j
where:
IVj2 = sum of squared coefficients, and
nm = total number of materials
Calculate the *pa«™itH allowable difference in percent
relative by multiplying the coefficient of variation by q for
the number of repeats and degrees of freedom.
15.1.1.10 If the coefficients of variation do not appear to
be homogeneous, a transformation of the test results (such as
to loearithms- aw «««• nt- «••««—i root) ' " '
15.1.1.11 If homogeneous variances or coefficients of
variation cannot be obtained from the intralaboratorv re
suits, calculate the precision for appropriate levels of the
material value.
15.1.2 Reproducibility (Imeriaboniory^T^ reproduc-
ibihty is determined from the estimated interlaboratorV
variance which is the variance of the mean values obtained
by the laboratories isL2) plus the intralafaoratorv variant
Thus, mteriaboratory variance is as follows: "
V - <*w:/n*) + SL-
where nR = number of repeats in each laboratory.
NOTE 4 — t ne suoscnm. b. derives from tne previously usea term
"between-labontones".
1 5. 1 .2. 1 The equation defining s w: and the procedures for
its computation are discussed in 15.1,1.1 and 15.1.1.2.
15.1.2.2^ The equation defining the variance of laboratory
means. SL~, is as follows:
where:
•7 » mean obtained for a material by a labora-
tory,
= grand mean of the values obtained for a
material by all laboratories.
« number of laboratories.
m sum of squares, ss. and
= degrees of freedom, df.
To facilitate calculations, the equation is convened to
X
«4 .
(X - X)2
n - 1
•
15.1.2.2 When the results obtained in the interiaboratory
test are from a balanced design and there are no missing
values, compute the sum of squares and variance for each
material by the analysis of variance technique, described in
15.1.3. Calculate the interiaboratory variance using the
equation given in 15.12.
15.1.2.4 Pool the interiaboratory variances for all mate-
rials (unless they appear nonhomogeneous) to give a single
interiaboratory variant for the test. Calculate the interiabo-
ratory standard deviation. s& and report.
15.1.2.5 Compute the maximum allowable difference by
multiplying sb by the q factor (Table 6) for the appropriate
number of laboratories and the degrees of freedom of (nL -
1 ) times the number
.
15.1.2.6 Examine the results from all laboratories to
establish that the range of laboratory means does not exceed
the MAD. If the range is exceeded and laboratory outliers
were not previously rejected, discard the discordant results
and recalculate the MAD as in 15.1.1.7.
15.1.2,7 If the interiaboratory variances for the materials
do not appear to be homogeneous, but appear to vary with
the level of the mean* of the material*, convert the standard
deviations to coefficients of variation, and if homogeneous.
pool them as shown in 15.1.1.8 and 15.1.1.9. Calnilarr the
MAD in percent by multiplying the coefficient of variation
by q for the number of laboratories and degrees of freedom.
15.1.2.8 If the coefficients of variation do not appear
homogeneous, a transformation of the test results (such as to
logarithms, arc sine, or square root) may provide homoge-
neous variances. If neither variances nor coefficients of
variation are homogeneous for the interiaboratory results.
calculate the reproducibility for each level of test value of the
materials as in 15.1.1.11
1 5. 1 .3 Analysis of Variance of Balanced Results:
1 5. 1 .3. 1 Where a design recommended in Pan B has been
used and balanced results are available (that is, no missing
values from a balanced design or missing values replaced by
the appropriate mean value), the analysis of variance tech-
nique should be used to compute the intralaboratory (within)
Page 54
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
24
30
40
60
120
18.0
6.09
4.30
3.93
3.61
3.46
3.34
358
350
3.15
3.11
3.08
3.06
3.03
3.01
3.00
2.98
2.97
2.96
2.95
2.92
2.89
2M
2J3
2JO
2.77
^ — ^-«_
26.7
858
5.88
5.00
4.54
4.34
4.16
4.04
3.95
3.88
3.82
3.77
3.73
3.70
3.67
3.65
3.62
3.61
3.59
3.58
3.53
3.48
3.44
3.40
3J8
3.32
•• 01 mm BUM
32J
9.80
6.83
5.76
5.18
4.90
468
4.53
4.42
4.33
4.26
4.20
4.15
4.11
4.08
405
402
4.00
3.98
3.96
3.90
3.84
3.79
3.74
3.69
3.63
ftfQ (Mwoon or i
MRCZM Rang
5
375
10.89
7.31
6J1
3.64
5.31
5.06
489
4.76
466
458
4.31
4.46
4.41
4.37
4.34
4.31
4.28
4.26
424
417
4 11
404
3.98
3.92
3.86
BOMftainf n» v>n
6
405
11.73
8.04
6.73
5.99
5.63
5.35
5.17
5.02
491
4.82
4.75
469
4.64
459
456
4.52
449
447
445
•»37
430
«23
4.16
4.10
403
7
43.1
12.43
8.47
7.06
658
5.89
5.59
5.40
554
5.12
5.03
495
4.88
4.83
4.78
4.74
4.70
467
4.64
462
434
446
439
4.31
4.24
8
43.4
13.03
8.85
7.35
6.52
6.12
5.80
5.60
5.43
5.30
550
5.12
5.05
4.99
4.94
4.90
486
483
4.79
477
468
460
4.52
4.44
4.36
9
47.3
13.54
9.18
7.60
6.74
6.32
5.99
5.77
3.60
5.46
5.33
557
5.19
5.13
5.08
5.03
4.99
4.96
4.92
490
481
472
4.63
4.55
4.47
10
49.1
13.99
9.46
7.83
6.93
6.49
6.15
5.92
5.74
5.60
5.49
5.40
5.32
355
550
5.15
5.11
5.07
5.04
5.01
4.92
4.83
4.74
4.65
4.56
and interiaboratory (between) variances.* In the analysis, it
must be ascertained mat the intralaboratory variances and
mean values are consistent among the laboratories. Interiab-
oratory values are not necessarily consistent, depending
upon the precision of the test method(s). Lack of consistency
may arise from a variety of causes, such as materials or
equipment. As noted in 3.3, the individual replicates are
used only for determining gross errors in test results.
15.1.32 Following is an example of the application of the
analysis of variance to interiaboratory test results. In this
round robin, one laboratory coated, baked, and distributed
the panels to five co-operators. Farh co-operator received
three panels and was asked to determine the Knoop hardness
at three different times within a given period to eliminate
aging effects. The results submitted arc given in Table 7.
15.1.3.3 The first step (whether using the analysis of
variance or not) is to examine each replicate value for
possible errors. This is most easily done by calculating the
range in results for each time and comparing the ranges at
similar levels in results. In the example given here, in which
the range appears to increase with increasing hardness, the
results for each enamel must be compared separately, but in
cases where the values for all materials are at about the same
level all results must be considered together. The calculated
ranges in the example are presented at the bottom of Table 7.
*^^^Z^^^^J^tt^ "£* o^^lSTthM^P^u^that
from fcTM Information Semes inc. 1S29 Chaine Court. Glouccoer Omara * a I0ta* °* SU anaiyscs- It IS necessary to compute lirst U1C
KIC 2W6, Canada. Equivalent propams nuv be available from otter aources. variances for each material because the variances may differ
significantly for different types of materials or for different
15.1.3.4 Next, apply the test for outliers. 13.3, or the
control chan technique (D3) of Practice E 180. to see if a
suspect value can be rejected. When more than two replica-
tions are made at a time and only one result is divergent, that
result can be discarded and the mean calculated from the
remaining values, thus retaining the balanced design of the
interiaboratory study. When only duplicate determinations
are made it is usually necessary to reject both results. In the
example. Replicate b of Laboratory V on Enamel A at Time
3 has caused the range to be larger than acceptable, possibly
the result of inverting figures when taking the readings or
preparing the report The mean for Time 3 is therefore
^dilated from the two remaining values which drange8 the
subtotal for Laboratory V to 3.9. the total and mean of
Enamel A to 18.7 and 1247, and the Laboratory V total to
140.2.
15.1.3.5 The ranges between the mean results at different
times and the laboratory means (or totals) are also examined
for discordant results to see if any should be discarded before
conducting the analysis of variance. In the example none are
rejectable: although the mean (and total) for Enamel E
obtained by Laboratory IV appears suspect, the calculated
quotient does not exceed the R.Q. of 0.710 for a significance
level of 0.05 and five values, as shown:
Page 55
-------
D 3980
Uoofnorv
enamw> 'cm
Mean
TABLE 7 Knoop Hwttmw of Baked Enamets
(a) Summary at Test Results
IV
Laboratory Total
Gram Total
2.61 2.61
2-61 2.5 I 2.6!
2.9! 2.41 2.61
4.5 i 461 4.01
4.01 441 431
4.1 : 451 431
1 ' 1.3 1
21 1.61
3! 1.3!
2! 3.7' 1.41
1.3:
1.5.
1 T'
1.5.
1.6 i
1.41
1.51
1.51
1.31
1.3!
1.61
441 1.41
1.2:
1.51
1.2:
1.3:
: 1.3;
! 2.1 i
: 1.1 1
: 1.5 1 42! 19.0 i
1 287
3.41 3.01 3.1 !
3.2! 2.8!
3.1 I 3.2 I 3.4 I
2.9 I 3.6 I 3.1 1
H
il 791
II
II
II
!l 12.9!
i
• 199!
321
3.31
5.71
6.01
5.71
5.81
8.5!
8.31
8.11
8.3 ,
3.41
32!
62 i
6.01
6.41
6.21
8.7'
8.2!
8.3!
8.4-
2.81
2.9 1 9.41
5.61
5.71
6.1 1
5.81 17.81
8.71
9.3!
9.01
9.0 i 25.71
3.31
3.11
4.71
4.91
5.1 i
49)
8.0!
8.41
821
82!
3.41
3.41
4.61
5.1 1
5.0!
491
7.71
7.1 :
7.41
741
3.41
3.3 ! 9.81
5.7!
5.11
5.4(
5.41 1521
7.7 i
7.81
821
7.9 1 23.5!
43.7 ' 2.913
75.8 1 5.053
1102 : 7.347
16.6 i 16.5: 15.9!
16.0: 17.0. 15.4.
16.3; 16.9 16.1 •
11.9! 11.7' 11.3!
11.8! 11.51 11.01
1231 11.0! 10.71
9.6l 8.9' 28.61 16.3! 16.81 15.8 I 46.9112.0111.4111.01 34.41 168.9 I 11.260
15.4 , 13.8 I 13.9 i
15.2 : 13.2 : 14 6 I
14.7 13.5i 14.1 :
17.4116.1: 17.31
1821162: 16.71
17.81 16.6 i 17.01
1&6 I 202 I 19.1 !
18.01 19.6 I 18.71
18.3 I 19.9 I
115 Jl
157.31
1144.41
669.0 I
For15 vatuesO-•
0.4
2
4
5
4
1
OJ
1
4
5
4
2
0.6
1
4
6
6
0.7
1
3
4
OJ 0.
1
2 1
> 1
.
.0
1
levels of the property measured. The analysis of variance of
test results is facilitated by the construction of an analysis of
variance table as shown in Table 8. «*uy»» ui
15.1.3.7 After calculating the means for the replicates,
rejecting outhers where necessary, prepare a tabte of Se
means, which are considered to be the individual results, for
each material as shown in Table 9 for Enamel A.
15.1.3.8 Perform the computations outlined in Table 8 to
obtain the total sum of squares and the laboratories and the
mtralaboratory sums of squares. For each material from the
sums of squares and the degrees of freedom, compute mean
squares and variances for intralaboratorv and for laboratory
means as outlined in Table 8 and shown by the example 2
NOTE 5-lnthisKcampie.mplicaicmosuiwnenawremadesothat
^equals 3 but m cases where dupJicates are sufficient (see
laboratory totai sums of squares and the laboratory mean
obtained usrng two as the divisor. Tie number of
f^02 " §i * °-75'Ra « a°1 *9«*c«no» - 0-6*7.
15.1.3.9 Calculate the intraiaboratory and interiaboratory
standard deviations for »ach material as in Table 10.
Determine the maTir^^m allowable differences for each and
compare with the ranges in the results as shown in Table 9. If
required, discard discordant results and recalculate.
15.1.3.10 If the standard deviations for the different
materials are homogeneous, pool the variances and calculate
the overall repeatability and reproducibility, as described in
15.1.1.6. 15.1.1.7, 15.12.5, and 15.1.2.6.
15.1.3.11 If the results are not homogeneous, convert the
standard deviations to coefficients of variation, pool them
and calculate the repeatability, as described in 15.1.1.8.
15.1.1.9, and 15.12.7 and shown in Annex Al.
15.1.3.12 Prepare a table showing the standard deviations.
coefficients of variation (if required) and the precision as
given in Table 11.
16. Sensitivity
16.1 If a reasonably linear relationship (r > 0.7) has been
Page 56
-------
TABLE » Ar«ly«i« of Variance) Table
Sourca of Variance)
Sum of Souares IMI
at Fre
om lOF)
Mean Souare Exoocieo Mean Square"
TOM
Eot - ZX* ~ -
-------
D3980
Enamel
A
B
C
D
E
f
5X
18.7
43.7
75.8
1102
168.9
25t.4
Tota*— 668.7
Enamel
A
a
c
0
E
F
S3
0.0467
02267
0.620
1.3133
3.3133
6.1267
Enamel
A
B
C
0
E
F
at M
4
4
4
4
4
4
Totat— 24
•* See Annex A1 for oscuswon of si
TABLE 10 Summary tor All Enamels'4
128.75
388.20
821.38
2011.15
4267.24
01
"To"
10
10
10
10
10
Tota>—60
Meen Square
0.0677
0.3027
1 1343
2.616
263057
11.9123
staosncat analysis.
0.063
0.280
1 072
2.485
26.174
11.30
(IX^/n
23.3127
127.3127
383.0427
809.6027
1901.814
4213.464
Net Total ss
0.3173
1.4373
5.1573
11.7773
109.336
53.776
ZL'T/na Net laooratorv ss
233833
1283233
38738
820.0667
2007.8367
4261.1133
02707
1.2107
45373
10.464
106.0227
47.6493
l/ilfUoDoraiorv
Mean Square -
s,*
0.00467
0.0227
0.062
0.1313
0.3315
0.6127
*w
0.068
0.1505
0249
0.3625
0.5755
0.783
Mean
1247
2.913
5.053
7.347
1126
16.76
»»*
5.48
5.17
4.93
4.93
5.11
467
intenaooratorv
S,*
0021
00933
3.3574
03282
3.7248
3.7667
s,1/^
0.0016
0.0076
0.0207
0.0438
0.1104
02042
*«,' «,
0.0226 0.15
0.1009 0.32
0.3781 0.615
0.872 0.93
8.8352 2.97
3.971 1.99
V8.X
12.05
10.90
12.17
12.71
26.4
11.89
TABLE 11 Summary of Intartaboratorv T««t hv M«K^
SCfA
where, if SK is greater than 1.0. Test Method A is superior to
B, and if SR is less than 1.0. Test Method B is superior
16.5 Obviously, a sensitivity ratio should not be used to
compare test methods unless it has been established that one
of these test methods provides instrumental values that
correlate well with known property or component values of
materials. For example, if Test Method A has been found to
give instrumental values (observed or transformed) that
correlate reasonably well (r > 0.7) with known propenv
values, then a plot can be made of Test Method B values
versus Test Method A values for the materials tested.
16.5.1 If the plot of points shows a curvilinear relation-
ship, then the values for Test Method B should be trans-
formed to provide a linear relationship. When this has been
done, the slope of the straight line relating the points is
A/4/AJ3 and
SR - siope/u .,/*«)
16.5.2 In some instances, the SR for two test methods will
not be constant over the range of materials tested because of
Page 58
changes in the test errors with a change in parameter level. In
such cases, an SR should be calculated for each of the
parameter value levels. A plot of SR values versus parameter
values is useful in comparing the sensitivities of the two test
methods.
PART D—ANALYSIS PRESENTATION
17. Scope
17.1 This pan describes the essential requirements for the
preparation of a comprehensive report on an ASTM cooper-
ative test program, for the guidance of working groups and
committees or for publication, and of the precision state-
ment to be used in test methods.
18. General Report Requirements
18.1 The essential requirements for a report on a cooper-
ative test program are as follows:
18.1.1 The presentation in one document of details per-
taining to the program, including a complete description of
the experiment with a clear statement of purpose, pro-
cedure, instructions, and list of participants.
18.1.2 The presentation of all the original results reported
by the participants, together with their comments.
-------
TABLE 12 Summary ot Results Irom Intertaboratory Test on .
. by Method.
Maun* i
MatenaiZ
etc. to Mjtanii M
Laboratory
NunKMr of
Datarmna- Mean Range
toons
Numoer ot Numoar ot
Oetarmma. Mean Range Oatarmna- Mi
un Range
It
ate. to
L
Column means
18.1.3 An integrated compilation of the results into tables
. and in the form of graphs or charts that may be studied to
obtain a clear view of the experiment as a whole.
18.1.4 The presentation, in the most concise form, of the
conclusions drawn from an analysis of the results.
18.1.5 The presentation of a summary, conclusions, or a
list of recommendations.
18.2 The report should provide maximum ease of refer-
ence to the details of the experiment, of observing results in
juxtaposition, and the essential conclusions. The report
should be available to all members of the task group well m
advance of meetings so that there is sufficient time for study
and development of new ideas for discussion at meetings.
19. Procedures for Preparing Summaries
.19.1 For each material or each test method, prepare a
table that compiles in an orderly block arrangement the
results from all of the laboratories (including operators and
time). A format based on the one given in Fig. 1, but modi-
fied to include means and ranges, can be used. Using the
design shown in Table 12. prepare a table for each test
method giving the laboratory mean values and ranges for the
different materials. Prepare a table for each test method
listing for each material the means, the intra- and interiabo-
ratory standard deviations, the degrees of freedom, and. if
applicable, the coefficients of variation. The format can be
based on pans of Table 10 or on Table 11. It should be
evident from this tabulation whether the standard deviation
vanes with the magnitude of the property being measured
and. consequently, whether the coefficient of variation is
required.
19.2 An appropriate block of results may be compiled
into a frequency distribution table that can be prepared in
several different ways as follows:
19.2.1 Tabulate the values in increasing order of magni-
tude and express the frequency with which each value occurs
as a fraction or percent calculated by dividing the number of
times the value occurs by the total number of values in a
group. When a large number of values is being treated, they
may be divided into classes of equal range (for example, ten
classes) and the frequency of each class calculated.
19.2.2 Calculate the deviation of each value from the
mean value of its group noting the sign and tabulate these in
order of magnitude. Calculate the frequency of each devia-
tion as a fraction or percent of all the values treated. When a
large number of values is being treated, the deviations may
be divided into classes of equal range.
19.2.3 Frequency distributions may be expressed in a
cumulative way by arranging the values in order of magni-
tude and summing at each stage of this arrangement the total
occurrence of all values up to and including the selected
stage.
19.2.4 Graphical presentations are very useful and concise
allowing rapid inspection and judgment of a collection of
related results. They may be prepared from all the results
from an experiment or an appropriate segment to obtain a
••iew ot'the nature of the distribution. If it is revealed that an
adeauate normal distribution exists, relatively simple math-
ematical treatments may be applied. Frequency distributions
may be presented in graphical form, as histograms whereby
-alues or classes covering a range of values are plotted
against their frequency, or as cumulative frequency graphs
prepared by plotting at each point the cumulative frequency
of ail the values up to that point against the value at that
point. One of the best ways of displaying results is the dot
diagram.' Figure 3 shows the relation of dry-to-touch time to
the day on which a test was made and to the type of drier
used, and afco shows the overall distribution of the results. It
can readily be seen that there is definitely a difference
between days but none between drier types, as compared to
the overall distribution.
19.3 The final summary should be a description of results
from the cooperative program, expressed in terms of repeat-
ability and reproducibility stated with supporting informa-
tion, that is. degrees of freedom.
19.4 Prepare a draft report, with the test results and their
analysis, for deposit in ASTM Headquarters files.
20. Applicability and Precision Statements for Test
Methods
20.1 The major functions of the interiaboratory test of a
measurement procedure are the determination of:
20.1.1 The applicability of the procedure for measuring a
property of a material or groups of materials, and
20.1.2 The precision with which repeated measurements
according to the test method can be expected to be made.
202 Applicability implies both the correlation of a mea-
surable property with an unmeasurable property of a mate-
rial and the sensitivity with which differing levels ot that
property are discriminated by the test method. Correlation
may be evaluated by means of a rank order correlation
coefficient. Sensitivity may be evaluated by means of the
sensitivity criterion (ratio of the range in property levels of a
material "to the standard deviation). As applicability, or
* Box. G, Hunter. W.. sod Hunter. S_ Statutes tor Expenmtiutn. John Wiley
ind Son*. New York. NY. 1978. p. 221.
Page 59
-------
03980
80
75
M
•
1 70
I
» £K
JC 65
U
1 60
i
k 55
Q
50
45
1
1
• ••
•
•
.
••
•
• •
••
.
•
*••*
•
• *
•
•
•
*
•
•
•
1
• •
•
•
• •
•
•
• •
•
•
••
*•
2 3 4 5 A 8 C 0 E
Doy of Test *
7 Type of Oner
•
• •••
• *• •
*•• »•
• •• •
• •• •• •
•
Overall
Oistnoution
PIG. 3 Illustration of Dot Diagram
validity, of a test method depends on the magnitude or the
values of correlation and sensitivity, as well as on the
number of materials used in the imerlaboratorv test, mciu-
fo°Uow° a Stattment °f amicability is recommended as
Applicability of Test Method-Based on measurements of
materials having known qualitative ratings that range from
(good to poor, for example) the rank correlation
20.3 Precision is useful only if the applicability function
J «?1LtettJn?10d " found to ** favorable. Precision £
defined as the degree of agreement among repeated indepen*
dent measurements of the same property. StatememTof
precision may be given in terms of the standard deviation.
but this form is not very useftiL In addition, when values
obtained vary directly with the property levels of the
materials tested, the coefficient of variation "must be used to
express the general precision. Statements of precision can
also be expressed in terms of the range in values witSn
which the correct value can be expected to lie a specified
percent of the time. The most common precision *£££
range that should not be exceeded for a stated number of
results at a given confidence level, assuming that the
frequency distribution is normal and that a random samole
was used in the interiaboratory test. Because only a sample of
the total population of measurements is obtained in a test.
statements of precision should indicate the number of
degrees of freedom and the number of observations that were
used m the interiaboratory test to obtain the indicated
V31UCS.
20.3.1 As can be seen from Tables 11 and 12. the
measurements made in an interiaboratory test permit a
number of parameter variations. Precision indexes can be
devised with respect to differences within and among labora-
tories. materials, operators, and repeat measurements and
with respect to the interactions among combinations of these
parameters. It is evident from this discussion that to provide
all of these precision indexes would make this section
unnecessarily long and impair its usefulness. The decision on
how many of the possible indexes are to be included in a test
method is left to the discretion of those who formulate and
use the test method. In this practice, however, the precision
statement is based on the two indexes — repeatability and
reproducibility as defined in 326:
Precision of Test Method— la an interiaboratory study of this test
method in which ___ operators in __ laboratories T*rtfd ^^_
coatings with a broad range nf (property) levels (analyzed __
materials containing ?. the intralaboratory standard deviation
(coefficient of variation) was found to be _ units (percent) with
- degrees of freedom (df) and the interiaboratory standard
deviation (coefficient of variation) __ units (percent) with ___ df
after rejecting - resuitsfs) from _ laboratory (ies) for one time
because the range between replicates (repeats) differed significantly
from all other ranges for material ___ or all results from one
laboratory for tna^fpaj ___ bprauy the T"99^ differed significantly
from all other means. Pared on thnr standard deviations (coeffi-
cients) the following criteria should be used for judging, at the 95 %
confidence level, the acceptability of results:
Repeatability— ___ results. M"TI the mean of — lepticatcs (if
applicable), obtained by the same operator should be considered
suspect if they differ by more than _ units (% relative).
Reproducibility— Two results, each the mean of _ replicates.
obtained by operators in different laboratories should be considered
suspect if they differ by more than _ units (95 relative).
NOTE 6 — Where the results obtained are in percent instead of some
unit add "absolute" after X ?a in the repeatability and reproduability
statements to distinguish from cam where use of the coefficients of
vanauon results in percent relative precision limits.
NOTE 7— Users of ASTM test methods should be aware that the
precision obtained from an interiaboratory study is for measurements
made on the same batch of »fta material. Consequently, normal
batch-to-batch variability, which is governed by m*'"tf'mT"tt quality
control is not included in the precision values established for a test
method.
Page 60
-------
ANNEX
(Mandatory Information)
Al. Discussion of Table 10
A 1.1 Two points are evident: 11) neither the imraiab-
oratory nor interiaboratory variances can be pooled because
the standard deviations increase with higher hardness values:
(2) the intralaboratory coefficients of variation are homoge-
neous. but there is one discordant value (Enamel E) in the
interiaboratory coefficients.
; A1.2 The tests referred to in 15.1.1.4 can be applied to
confirm that the interiaboratory coefficients cannot be
pooled or. as shown below, the coefficients are pooled to
determine if any maximum allowable difference is then
acceded. Pooling the coefficients of variation in accordance
with the formula in 15.1.1.8 yields the following:
153-j873V
o /
Pooled vw - /
\ ,
For three determinations v
Pooled vb
-0 — f.055 x 3.40 - 17.19
- (235.3061)" - 15.34 %.
For five laboratories vb-f - I5J4 x 4.17 - 63.97 %.
Therefore the acceptable range for Enamel E is 0.172 x
1126 - 1.94 for intralaboratory results and 0.64 x 1 1.26 »
12 for interiaboratory results. The actual maxima obtained
with Enamel E in the round robin are for the intraiaboratory
range 1.3 for Laboratory n, and for the interiaboratory range
7.4 between Laboratories I and IV. Thus, while the results
from Laboratory IV for Enamel E could not be rejected in
15.1.3.5 on the basis of Enamel E alone, they can be on the
basis of all six enamels.
A1.3 Instead of rejecting the results from all laboratories
for one materiaL it is preferable to discard only those of the
divergent laboratory. When this procedure is followed for
Enamel E the revised values are Uf - 120 0 n - 12, X -
10.00. Z.r2 - 1213J8. (IX? fn - 1200, net total ss - 13.58,
ZIr2/nR - 1210.7667, net laboratory ss - 10.7667,
intralaboratory ss - 2.8133, df - 8, sj =» 0.35 17, j. - 0.593,
vw " 5.93 %, interiaboratory df — 3, mean square « 3 5889*
2 = 3.2372. JL2- 1.0791, Jw2/nR = 0.1172. v- U963!
1.09. vb- 10.9455.
A 1.4 The imeriaboratory coefficients of variation are now
all homogeneous so they can be pooled with total degrees of
freedom of 58 for intralaboratory and 23 for interiaboratory
coefficients.
m /776.44035r
: 4) J . \ 29
-(26.7738)" * 5.174 %: vw-
-------
D3980
APPENDIX
(1) ASTM Manual for Conducting an Interiabonaory Studv or" a
Tea .Method. ASTM STP 335. ASTM. 1963.
(2) ASTM Manual on Quaiuy Control of Materials. ASTM STP 1!.
ASTM. 1951.
• (3) ASTM Practice O 1749. for Interiaboratory Evaluation of Test
Methods Used with Paper and Paper Products. Annual Book of
ASTM Standards, Voi 15.09. -"
(4) ASTM Practice D 2904. for interiaboratory Testing of Textile
Materials. Annual Book of ASTM Standards, Voi 07.01.
(5) ASTM Practice E 173. for Conducting imeriaboratory Studies of
Methods for Chemical Analysis of Metals. Annual Book of
AST^f Standards, Voi 03.05.
(6) ASTM Practice £691. for Conducting an Interiaboratory Test
Program to Determine the Precision of Test Methods. Annual
Book of ASTM Standards. Voi 14.02.
(7) Bennet. C. A., aad Franklin. N. L. Statistical Analvsis in
Chemistry and the Chemical Industry, John Wiley ana Sons.
New York. NY. 1954.
(8) Brownies. K. A.. Industrial Experimentation. Chemical Pub-
lishing Co- 1947.
(9) Davies. O. I™ Design and Analysts o/Industrial Experiments.
Hafner. 1954.
(Nonmandatory Information)
(10) Flnkner. M. D.. 'The Reliability of Collaborative Tests for
AOAC." Journal. Association 01 Official Agricultural Chemists.
Voi 40. 1957. p. SSI
(11) Freeman.H. A... Industrial Statistics. John Wiley and Sons. New
York. NY. 1942.
(12) Kempthorne. O.. Design and Analysis of Experiments, John
Wiley and Sons. New York. NY. 1952.
(13) Mandei. J« and Lashof. T. W., "The interiaboratory Evaluation
of Testing Methods." ASTM Bulletin. No. 239. July 1959. p. 53.
(14) Mandei. J.. ed. Interiaboratory Testing Techniques. American
Society for Qualitv Control. 1978.
(15) Mandei J.. "The Measuring Process." Technometncs. Voi 1.
No. 3. 1959. p. 251.
(16) Mann. H. B.. Anaivas and Design of Experiments, Dover. 1949.
(17) Natreila. M. C.. "Experimental Statistics." NBS Handbook 91.
1963.
(18) Quenouille. M. H.. The Design and Analvsis of Experiments.
Hafner. 1953.
(19) Youden. W. J.. "Graphical Diagnosis of Interiaboratory Test
Results." Industrial Quaiuy Control. Voi XV, No. 11. 1959, p.
28.
lin
--- -mat ni
a -- — ^-mrnfmrtmronnritTinnf ..... 11 inmrinii
Page 62
-------
dLJM Designation: 0 4057 - 88*1
//ft) Designation: MPMS (Chapter 8.1)
An American National Stanoaro
Standard Practice for
Manual Sampling of Petroleum and Petroleum Products1
This standard is issued under the fixed designation D 4057; the number immediately following the designation indicates the year of
onfina! adoption or. in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epstlon d) indicates an editorial change since the last revision or reapproval.
This lest method has been approved by the sponsoring commuters and accepted by the Cooperating Societies in accordance mth
established proceaures. This metnod was issued as a joint ASTM-AP1 standard in 1981.
Annex Al is tauter revision and will be contained in subsequent revisions to the standard.
«' NOTE—An editorial correction was made in 8.10.3J (2). 10.3.1.2. and 10.3.1.7 in July 1989.
I. Scope
1.1 This practice covers procedures for obtaining repre-
sentative samples of stocks or shipments of uniform petro-
leum products, except electrical insulating oils and fluid
power hydraulic fluids. This practice also does not cover
butane, propane, gas liquids at or above 26 psi (179 kPa)
Reid vapor pressure (Rvp), and other petroleum products
that are gases at atmospheric temperature and pressure.
NOTE I—The procedures described in this practice may also find
application in sampling most noncorrostve liquid industrial chemicals.
provided that all safety precautions specific to these chemicals can be
strictly followed.
NOTE 2—The procedure for sampling liquefied petroleum i*
-------
D4057
total volume agree within the precision of the laboratory
tests.
3.1.2 "official" sample—a sample taken at the point of
custody transfer and used for the custody transfer laboratory
determination.
3.1.3 representative sample—z sample representing a
small portion of its total volume of material (for example,
tanks, ships, compartments, containers, and pipeline ten-
ders) obtained with a precision equal to or better than the
precision of the laboratory method by which this sample is to
be analyzed.
3.J.4 all-levels sample—a sample obtained by submerging
a stoppered beaker or bottle to a point as near as possible to
the draw-ofl" level, then opening the sampler and raising it at
a rate such that it is about three-fourths full (maximum
85 95) as it emerges from the liquid. An all-levels sample is
not necessarily a representative sample because the tank
volume may not be proportional to the depth and because
the operator may not be able to raise the sampler at the
variable rate required for proportionate filling. The rate of
filling is proportional to the square root of the depth of
immersion.
NOTE 3—The tube sampling procedure. 8.5.3. may be used to obtain
an all-levels sample from a barrel or drum.
3.1.5 running sample—* sample obtained by lowering an
unstoppered beaker or bottle from the top of the oil to the
level of the bottom of the outlet connection or swing line.
and returning it to the top of the oil at a uniform rate of
speed such that the beaker or bottle is about three-fourths
full when withdrawn from the oil. A running sample is not
necessarily a representative sample because the tank volume
may not be proportional to the depth and because the
operator may not be able to raise the sampler at the variable
rate required for proportionate filling. The rate of filling is
proportional to the square root of the depth of immersion.
3.1.6 spot sample—a. sample taken at a specific location in
tank or from a pipe at a specific time during a pumping
operation.
3.J.7 top sample—a spot sample obtained 6 in. (152 mm)
HATCH
*«"
x— j- TOP SAMPLE
* UPPER SAMPLE
•* . »— MIDDLE SAMPLE
-i * LOWER SAMPLE
* CLEARANCE SAMPLE
UPPER THIRD
MIDDLE THIRD
LOWER THIRD
helow the top sunacc of the liquid (Fie. 1).
3.1.8 upper sample—a spot sampie taken at the mid-point
of the upper third of the tank contents (Fig. I).
3.1.9 middle sample—a spot sample obtained from the
middle of the tank contents (a point halfway between the
upper and lower sample points) (Fig. 1). (See 9.3.2.)
3.1.10 lower sample—a spot sample obtained at the
mid-point of the lower third of the tank contents (Fig. 1).
3.1.11 clearance sample—a spot sample taken 4 in. (102
mm) below the level of the tank outlet.
3.1.12 bottom sample—a sample obtained from the mate-
rial on the bottom surface of the tank or container at its
lowest point.
3.1.13 A bottom water sample is a spot sampie of free
water taken from beneath the petroleum contained in a ship
or barge compartment or storage tank.
3.1.14 drain sample—a sample obtained from the water
draw-off valve. Occasionally, a drain sample may be the
same as a bottom sample, as in the case of a tank car.
3.1.15 nutlet sample—a spot sample taken at the level of
the bottom of the tank outlet (either fixed or swing pipe) but
not higher than one meter above the bottom of the tank (see
Fig. 1)."
3.1.16 automatic sampler—a sampler used to retrieve a
representative sample from the liquid flowing in a pipe. The
automatic sampler generally consists of a probe, an ex-
tracting mechanism, an associated controller, and a sample
receiver.
3.1.17 single-tank composite sample—a blend of the
upper, middle, and lower samples. For a tank of uniform
cross section, such as an upright cylindrical tank, the blend
consists of equal pans of the three samples. For a horizontal
cylindrical tank, the blend consists of the three samples in
the proportions shown in Table 1.
3.1.18 multiple tank composite sample (ships, barges,
etc.)—a mixture of individual samples from the several
'compartments each of which contains the same grade of
petroleum material. The mixture is blended in proportion to
the volume of material in each compartment.
3.1.19 composite spot sample—a blend of spot samples
mixed volumctricaily proportional for testing. Some tests
may also be made on the spot samples before blending and
the results averaged. Spot samples from crude oil tanks arc
collected as follows:
3.1.19.1 liiree-way—Qn tanks larger than lOOO-barrcl ca-
pacity that contain in excess of 15 ft (4.6 m) of oil.
equal-volume samples should be taken at the upper, middle.
TABLE 1 Sampling Instructions tor Horizontal Cylindrical Tanks
Sampling Level. * ol
Qvmntr Above Bottom
Composite Samtw.
Pioportionate Parts
^ BOTTOM SAMPLE
NOTE—The outlet sample location snown applies only to tanks witn side
outlets it aoes not aopty wnen me outlet comes from me floor of me tank or turns
Down rnto a sump.
RG. 1 Sampling Depths
Uooer/MKiae/Lower
100
90
80
70
60
50
40
30
20
10
80 50
75 SO
70 SO
50
50
40
20
20
20
20
20
20
20
15
10
5
Upper/MKttte/Lowef
3
3
2
4
4
5
6
5
4
3
3
3
4
5
6
10
10
10
10
Page 64
-------
and lower or outlet connection of the merchantable oil in
the order named. On tanks of 1000-barrel capacity and less
. this method may also be used.
3.1.19.2 two-way—On tanks larger than 1000-barrel ca-
pacity that contain in excess of 10 ft (3.0 m) and up to 15 ft
(4.6 m) of oil, equal-volume samples should be taken at the
upper and lower, or outlet, connection of the merchantable
oil. in the order named. On tanks of 1000-barrel capacity
and under, this method may also be used.
3.1.20 middle spot sample—On tanks larger than 1000-
barrel capacity containing 10 ft (3.0 m) or less of crude oil.
one spot sample should be taken as near the center of the
vertical column of oil as possible and outlet connection.
3.1.21 dipper sample—a sample obtained by placing a
dipper or other collecting vessel into the path of a free-
flowing stream so as to collect a definite volume from the full
cross section of the stream at regular time intervals for a
constant rate of flow, or at time intervals varied in propor-
tion to the rate of flow.
3.1.22 tube or thief sample—a sample obtained with a
sampling tube or special thief, either as a core sample or spot
sample from a specified point in the tank or container.
3.1.23 tank-side sample—a. spot sample taken from a
sample connection in the side of a tank.
3.1.24 surface sample—a. spot sample skimmed from the
surface of a liquid in a tank.
3.1.25 flow proportional sample—* sample taken from a
pipe during the entire transfer. The rate of sampling is
proportional to the flow of the liquid in the pipe at anv
instant.
3.1.26 entrained water—water suspended in the oil. En-
trained water includes emulsions but does not include
dissolved water.
3.127 free water—water that exists as a separate phase
and typically lies beneath the oil.
3.128 emulsion—an oil/water mixture that does not
readily separate.
3.1.29 sample receiver (receptacle)—container in which
all sample bites are collected during sampler operation. A
receiver may be fixed or portable.
3.1.30 borings sample—z sample obtained by collecting
the chips made by boring holes with a ship auger from top to
bottom of the material contained in a barrel case bat or
cake. '
3.1.31 grab sample—a sample obtained by collecting
loose solids in equal quantities from each pan or package of
a shipment and in sufficient amount to be characteristic of
all sizes and components.
3.1.32 grease sample—* sample obtained by scooping or
dipping a quantity of soft or semiliquid material, such as
grease, from a package in such a manner that the material on
the scoop or dipper is representative of the material in the
package.
4. Summary of Practice
4.1 A basic sampling method is available: tank sampling.
which is covered in this practice.
4.1.1 If the tank contents are not homogeneous from top
to bottom of the tank or if the conditions in 4.1.2 are not
met. automatic sampling is recommended.
4.1.2 Tank samples will be representative if the tank
D4057
contents are homogeneous from top to bottom. This is rarely
the case in actual practice. However, tank samples are
acceptable if all of the following conditions prevail:
4.1.2.1 The tank contains a heavy component (such as
water) that clearly separates from the main component: and
4.1.2.2 The tank is equipped with either a swing suction
or a weir on the outlet that prevents any shipment of the
heavy component: and
4.1.2.3 The tank samples are taken so that none of the
heavy component is included.
4.1.3 See Section 8 for additional precautions and instruc-
tions.
5. Significance and Use
5.1 Samples of petroleum and petroleum products are
examined by various methods of test for the determination
of physical and chemical characteristics. The test results are
often used for custody transfer and pricing determinations. It
is accordingly necessary that the samples be representative of
the petroleum or petroleum products in question. The
precautions required to ensure the representative character
of the samples are numerous and depend upon the type of
material being sampled, the tank, carrier, container or line
from which the sample is being obtained, the type and
cleanliness of the sample container, and the sampling proce-
dure that is to be used. A summary of the-sampling
procedures and their application is presented in Table 2.
Each procedure is suitable for sampling a number of specific
materials under definite storage, transportation, or container
conditions. The basic principle of each procedure is to obtain
a sample or a composite of several samples in such manner
and from such locations in the tank or other container that
the sample or composite will be truly representative of the
petroleum or petroleum product.
6. Apparatus
6.1 Sample Containers may be clear or brown glass
bottles, or cans. The clear bottle is advantageous because it
may be examined visually for cleanliness, and also allows
visual inspection of the sample for free water or solid
impurities. The brown glass bottle affords some protection
from light. The only cans permissible are those with the
seams soldered on the exterior surfaces with a flux of rosin in
a suitable solvent. Such a flux is easily removed with
gasoline, whereas many others are very difficult to remove.
Minute traces of flux may contaminate the sample so that
results obtained on tests for dielectric strength, resistance to
oxidation, and sludge formation may be erroneous.
6.1.1 Plastic Bottles made of suitable unpigmented linear
polyethylene may be used for the handling and storage of gas
oil, diesel oil, fuel oil, and lubricating oil. They should not be
used for gasoline, aviation jet fuel, kerosine. crude oil, white
spirit, medicinal white oil. and special boiling point products
unless testing indicates there is no problem with solubility.
contamination, or loss of light ends.
NOTE 4—In no circumstances shall nonlinear (conventional) poly-
ethylene containers be used to store samples of liquid hydrocarbons.
This is to avoid sample contamination or sample bottle failure. Used
engine-oil samples that may have been subjected to fuel dilution should
not be stored in plastic containers.
NOTE 5—Plastic bottles have an advantage. They will not shatter like
glass or corrode like metal containers. They are generally used oniv once
Page 65
-------
TABLE
bauds of more man 16 os (110 kPa) ana not more
man 26 ps (179 kPa) Rvp
bauds of more man 16 psi (110 KPa) ana not more
than 26 ps (179 kPa) Rvp
Liquids of more than 2 psi (13.8 kPa) ana not more
than 16ps(110kPa)Rvp
bauds of more man 2 ps (13.8 kPa) ana not more
than 16 psi(110 kPal Rvp
bauds of 16 ps (110 kPa) or less Rvp
bauds of 2 ps (13.8 kPat or toss Rvp
bauds of 2 psi (13.8 kPa) or less Rvp
bauds of 2 ps (13.8 kPa) or less Rvp
bauds of 2 ps (13.8 kPa) or toss Rvp
Bottom or thief samptmg of kquds of 2 ps (13.8
kPa) Rvp or toss
Liquids and semoquds of 2 psi (13.8 kPa) Rvp or
less
Crude petroleum
B industrial aromstc hydrocaiuuiis
Waxes, solid bitumens, and other soft sodds
PotfOfcBurn cok0i lumpy sohds
Greases, soft waxes, asphalts
Asphatoc materials
Emutsifod aspfiatts
Bottom water a
men
2 Summary of Sampling Procedures and Applicability
Type of Container
storage tanics. snip ana barge tanks, tank cars.
tank trucks
storage tanks with taps
storage tanks, ship and barge tanks, tank ears.
tank trucks
storage tanks twin taps
pipes or knes
storage tanks, ship and barge
storage tanks with taps
free or open-ascnarge streams
drums, barms, and cans
tank ears, storage tanks
free or open-discnarge streams. Open tanks or
kettles with open needs: tank ears and tank
trucks: drums
storage tanks, stop ana barge tanks, tank ears.
tank trucks, and pipelines
storage tanks, ship and barge tanks
barrels, eases, bags, cakes
freight ears, conveyors, bags, barrets, boxes
Kettles OJUKS. eans. tubes
storage tanks, tank cars, tries, packages
storage tanks, tank cars. hnes. packages
ship and barge tanks, storage tanks
Proceoure
precoowd bone sampung
cooler tap sampling
bottle sampkng
tap sampbng
in-line sampling
botBesampmg
lapsampmg
dipper sampmg
tube samptng
truef safnpttng
dipper sampbng
automatic samotng; thief sampbng:
bottle sampfeng; tap sampbng
bottle sampeng
bomg sampbng
grab sarnp&ng
grease sampeng
and then discarded so that recleamng and recovery procedures are not
required.
6.2 Container Closure—Cork or glass stoppers, or screw
caps of plastic or metal, may be used for glass bottles- screw
caps only shall be used for cans to provide a vaportight
closure seal. Corks must be of good quality, clean, and free of
holes and loose bits of cork. Never use rubber stoppers
Contact of the sample with the cork may be prevented bv
wrapping tin or aluminum foil around the cork before
forcing it into the bottle. Glass stoppers must be a perfect fit.
Screw caps must be protected by a disk faced with tin or
aluminum foil, or other material that will not affect petro-
leum or petroleum products.
6.3 Cleaning Procedure—All sample containers must be
absolutely clean and free of water, dirt, lint, washing com-
pounds, naphtha, or other solvents, soldering fluxes or acids.
corrosion, rust, or oil. Before using a container, rinse it with
Stoddard solvent or other naphtha of similar volatility (It
may be necessary to use sludge solvents to remove all traces
of sediment and sludge from containers previously used)
Then wash the container with strong soap solution rinse it
thoroughly with tap water, and finally with distilled water
Dry either by passing a current of clean, warm air through
the container or by placing it in a hot dust-free cabinet at
40'C (104T) or higher. When dry, stopper or cap the
container immediately. In the ordinary field sampling of
crude petroleum, washing with soap and rinsing with water
may be eliminated.
6.4 Sampling Cage—This shall be a metal or plastic
holder or cage, suitably constructed to hold the appropriate
container. The combined apparatus shall be of such a weight
as to sink readily in the material to be sampled and
provision shall be made to fill the container at any desired
level (see Fig. 2a). Bottles of special dimensions are'required
to fit a sampling cage. The use of a sampling cage is generally
preferred to that of a weighted sampling beaker for volatile
extended—tube sampling, thief aamotng
products since loss of light ends is likely to occur when
transferring the sample from a weighted sampling beaker to
another container.
6.5 Sampling Apparatus is described in detail under each
of the specific sampling procedures. Sampling apparatus.
shall be clean, dry, and free of all substances that might
contaminate the material.
i
7. Precautions and Instructions
7.1 Sampling certain products requires a due amount of
caution for their handling. Refer to Annex Al for precau-
tionary statements regarding these products. '
7.2 Crude Petroleum and Heavy Fuel Oils usually are
nonhomogeneous. Automatic samplers are recommended
for sediment and water (S&W) and density measurement.
7.2.1 Tank samples may not be representative because:
7.2.1.1 The concentration of entrained water is higher
near the bottom. The running sample or the composite of the
upper, middle and lower sample may not represent the
concentration of entrained water.
7.2.1.2 The interface between oil and free water is difficult
to measure, especially in the presence of emulsion, layers, or
sludge.
7.2.1.3 Determining the volume of free water is difficult
because the free water level varies across the tank bottom
surface. The bottom is often covered by pools of free water or
water emulsion impounded by layers of sludge or wax.
7.3 Gasoline and Distillate Products usually are homoge-
neous but they are often shipped from tanks that have clearly
separated water on the bottom. Tank sampling is acceptable
under the conditions covered in 4.1.2.
7.4 When using tank samples, the S&W deduction is
usually the sum of the free water volume (usually determined
from a paste cut) and the entrained water volume deter-
mined from the S&W analysis of the tank sample. The
difficulty of determining the free water volume limits the
Page 66
-------
D4057
the most representative. Manual pipeline samples are less
representative than automatic pipeline samples. Manual
pipeline sampling is described in 10.4. Tank samples will
usually not be representative unless the tank is completely
homogeneous and contains no free water. .
7.6.1 Stationary or Shore Tanks:
7.6.1.1 Crude petroleum tanks may be sampled in the
following ways by mutual agreement: composite spot,
middle spot, all-levels, running samples or by sample cocks.
Additional samples may be taken as necessary.
NOTE 6—Where emulsions are in relatively higher concentration in
the bottom portions of the tank, the lower sample would not be
considered representative of this lower third. Automatic line sampling is
recommended in such cases. If this is not possible an outlet sample or
bottom sample should be required instead of the lower sample. In
addition, a bottom thieving should be made for both opening and
closing gauges so that any change in the S&W level at the bottom of the
tank may be observed and noted.
7.6.1.2 Where tank samples must be used for crude oil
fiscalization and the tanks do not have swing suction lines or
weirs, it is recommended that upper, middle, and outlet
samples be taken. These samples should be tested and
reported separately. The S&W deduction should be the
average of the three values. Other analyses should also be
averaged.
7.6.2 Ship or Barge Transfers—Samples of ship cargoes of
crude petroleum may be taken by the following methods by
mutual agreement
7.6.2.1 From the shore tanks before loading and both
before and after discharging as previously described.
7.62.2 From the pipeline during discharging or loading.
Pipeline samples may be taken either manually or with an
automatic sampler. If the pipeline requires displacement or
flushing, care must be taken that the pipeline sample
includes the entire cargo and none of the displacement
Separate samples may be required to cover the effect of the
line displacement on the prior or following transfer.
7.6.2.3 From the ship's tanks after loading or before
discharging. An all-levels sample or a running sample shall
be obtained from each compartment of the ship's cargo
tanks.
7.6.2.3 (/) Except where specifically exempted, when
loading a ship, the shore tank sample or the pipeline sample
taken from the loading line shall be official. However, ship's
tank samples may also be tested for sediment and water, and
for other quality aspects when required. The results of these
ship's tank sample tests, together with the shore tank sample
tests, may be shown on the cargo certificate.
7.62.3 (2) When discharging a ship, the pipeline sample
taken from a properly designed and operated automatic line
sampler in the discharge line shall be official. Where no
proper line sample is available, the ship's tank sample will be
official except where specifically exempted.
7.7 Finished Products—When loading or discharging fin-
ished products, taken samples from both shipping and
receiving tanks, and from the pipeline if required.
7.7.1 Ship or Barge Tanks—Sample each product after
the vessel is loaded or just before unloading.
7.7.2 Tank Cars—Sample the product after the car is
loaded or just before unloading.
7.7.3 Package Lots (Cans. Drums. Barrels, or Boxes)—
Take samples from a sufficient number of the individual
packages to prepare a composite sample that will be repre-
sentative of the entire lot or shipment. Select at random the
individual packages to be sampled. The number of such
random packages will depend upon several practical consid-
erations, such as:
7.7.3.1 The tightness of the product specifications:
7.7.3.2 The source and type of the material and whether
or not more than one production batch may be represented
in the lot; and
7.7.3.3 Previous experience with similar shipments, par-
ticularly with respect to the uniformity of quality from
package to package.
7.7.4 In most cases, the number specified in Table 3 will
be satisfactory.
7.8 Obtaining Samples:
7.8.1 Extreme care and good judgment are necessary to
ensure samples are obtained that represent the general
character and average condition of the material. Clean hands
are important. Clean gloves may be worn but only when
absolutely necessary, such as in cold weather, or when
handling materials at a high temperature, or for reasons of
safety. Select wiping cloths so that lint is not introduced,
which would contaminate samples.
7.8.2 As many petroleum vapors are toxic and flammable,
avoid breathing them or igniting them from an open flame
or a spark produced by static. Follow all safety precautions
specific to the material being sampled.
7.8.3 When sampling relatively volatile products (more
than 2 psi (13.8 kPa) Rvp). Fill and allow the sampling
apparatus to drain before drawing the sample. If the sample
is to be transferred to another container, also rinse this
container with some of the volatile product and then drain.
When the actual sample is emptied into this container,
upend the sampling apparatus into the opening of the sample
container and allow to remain in this position until the
contents have been transferred so that no unsaturated air will
be entrained in the transfer of the sample.
7.8.4 When sampling nonvolatile liquid products (2 psi
(13.8 kPa) Rvp or less), the sampling apparatus shall be filled
and allowed to drain before drawing the actual sample. If the
actual sample is to be transferred to another container, rinse
the sample container with some of the product to be sampled
and drain before it is filled with the actual sample.
NOTE 7—When taking samples from tanks suspected of containing
flammable atmospheres, precautions should be taken to guard against
ignitions due to static electricity. Metal or conductive objects such as
gage tapes, sample containers, and thermometers, should not be lowered
TABLE 3 Minimum Number of Packages to be Selected for
SftfltpnfflC)
Number of Packages
in lot
1 103
41064
6510125
126 to 216
217 to 343
34410512
513 to 729
730101000
1001 to 1331
Number of
Packages to
Be Sampled
all
4
5
6
7
8
9
10
11
Number of Packages
nlot
1332101728
1729 to 2197
2198 to 2744
2745 to 3375
3376 to 4096
4097104913
4914 to 5832
5833 to 6859
6860 or over
Number of
Packages to
Be Sampled
12
13
14
15
16
17
78
19
20
Page 67
-------
D40S7
into or suspected in. a compartment or tank which is being filled or
immediately after cessation of pumping. A waiting period will generally
permit a substantial relaxation of the electrostatic charge.
7.9 Handling Samples:
7.9.1 Volatile Samples—It is necessary to protect all
volatile samples of petroleum and petroleum products from
evaporation. Transfer the product from the sampling appa-
ratus to the sample container immediately. Keep the con-
tainer closed except when the material is being transferred.
When samples of more than 16 psi (110 kPa) Rvp are being
obtained, be sure to use containers strong enough to meet
local safety regulations. After delivery to the laboratory,
volatile samples should be cooled before the container is
opened.
7.9.2 Light-Sensitive Samples—It is important that sam-
ples sensitive to light, such as gasoline containing tetra-
ethyllead, be kept in the dark, if the testing is to include the
determination of such properties as color, tetraethyllead and
inhibitor contents, sludge-forming characteristics, stability
tests, or neutralization value. Brown glass bottles may be
used. Wrap or cover clear glass bottles immediately. It is a
definite advantage to use covered cardboard canons into
which the sample bottles may be placed immediately after
collection.
7.9.3 Refined Materials— Protect highly refined products
from moisture and dust by placing paper, plastic, or metal
foil over the stopper and the top of the container.
7.9.4 Container Outage—Never completely fill a sample
container, but allow adequate room for expansion, taking
into consideration the temperature of the liquid at the time
of filling and the probable maximum temperature to which
the filled container may be subjected.
7.10 Shipping Samples—To prevent loss of liquid and
vapors during shipment, and to protect against moisture and
dust, cover the stoppers of glass bottles with plastic caps that
have been swelled in water, wipe dry, place over the tops of
the stoppered bottles, and allow to shrink tightly in place.
Screw the caps of metal containers down tightly and check
for leakage. Appropriate regulations applying to the ship-
ment of flammable liquids must be observed.
7.11 Labeling Sample Containers:
7.11.1 Label the container immediately after a sample is
obtained. Use waterproof and oilproof ink or a pencil hard
enough to dent the tag, since soft pencil and ordinary ink
markings are subject to obliteration from moisture, oil
smearing, and handling: Include the following information:
7.11.1.1 Date and time (the period elapsed during contin-
uous sampling and the hour and minute of collection for
dipper samples);
7.11.1.2 Name of the sampler
7.11.1.3 Name or number and owner of the vessel, car, or
container
7.11.1.4 Brand and grade of material: and
7.11.1.5 Reference symbol or identification number.
8. Sampling Procedure
8.1 The standard sampling procedures described in this
method are summarized in Table 1. Alternative sampling
procedures may be used if a mutually satisfactory agreement
has been reached by the panics involved. Such agreement
shall be put in writing and signed by authorized officials.
8.2 Bottle or Beaker Sampling:
8.2.1 Application—The bottle or beaker sampling proce-
dure is applicable for sampling liquids of 16 psi (110 kPa)
Rvp or less in tank cars, tank trucks, shore tanks, ship tanks.
and barge tanks. Solids or semiliquids that can be liquefied
by heat may be sampled by this procedure, provided they are
true liquids at time of sampling.
8.2.2 Apparatus—A suitable sampling bottle or beaker, as
shown in Fig. 2, is required. Recommended uses and
diameter of openings in the bottle or beaker are given in
Table 4.
8.2.3 Procedure:
8.2.3.1 All-Levels Sample (One-Way)—Lower the
weighted, stoppered bottle, or beaker as near as possible to
the draw-off level, pull out the stopper with a sharp jerk of
the line and raise the bottle at a uniform rate so that it is
about three-fourths full as it emerges from the liquid. For
light products or deep tanks, a restricted opening may be
needed to avoid filling the bottle.
8.2.3.2 Running Sample (Two-Way)—Lower the unstop-
pered bottle or beaker as near as possible to the level of the
bottom of the outlet connection or swing line and then raise
the bottle or beaker to the top of the oil at a uniform rate of
speed such that it is about three-fourths full when withdrawn
from the oil. For light products or deep tanks, a notched cork
or other restricted opening may be needed to avoid filling the
bottle.
8.2.3.3 Top, Upper. Middle. Lower and Outlet Samples-
Lower the weighted, stoppered bottle to the proper top,
upper, middle, lower, and outlet depths (see Fig. 1).
NOTE 8—Where emulsions are in relatively higher concentration in
the bottom portions of the tank, the lower sample would not be
considered representative of this lower third. Automatic line sampling is
recommended in such cases. If this is am possible an outlet sample or
bottom sample should be required instead of the lower sample. In
addition, make a bottom thieving for both opening and closing gages so
that any change in the S&W level at the bottom of the tank may be
observed and noted.
8.2.3.3 (7) At the selected lex-el, pull out the stopper with
a sharp jerk of the line and allow the bottle or beaker to fill
completely. When judged full, raise the bottle or beaker,
pour off a small amount, and stopper immediately.
8.2.3.3 (2) Where tank samples must be used for crude
oil fiscalization and the tanks do not have swing suction lines
or weirs, it is recommended that upper, middle, and outlet
samples be taken. Test these samples and report separately.
The S&W deduction should be the average of the three
values. Also average other analyses.
8.2.3.4 Multiple Tank Composite Sample—Prepare a
composite sample in the laboratory (not in the field) by
i ADUC * wwgiraa aampimg DOIDB or BWMI
NOTE — See wrong or ill levels samples (92.3).
Material
Light lubricating oris. kemnes. gambles.
transparent gas ott. o**el fuels, and Osntttes
Light cruoe ots (less man 200 s Saytxtt Unversal
Viscosity at 100*F)
Heavy crude and fuel otfs
Diameter of
Opermg. n.
V4
I'/i
V.
IVk
(mm)
(19)
(38)
(19)
(38)
Page 68
-------
D40S7
mixing portions of the all-levels samples as specified in
3.1.16.
8.2.3.5 Composite Spot Sample—Prepare a composite
spot sample by mixing spot samples in equal proportions as
specified in 3.1.19, using either the three-way or the two-way
' procedure, whichever applies.
8.2.3.6 Middle Spot Sample—Obtain this sample in the
manner specified in 3.1.20.
8.2.3.7 Handling—Stopper and label bottle samples im-
mediately after taking them, and deliver to the laboratory in
' the original sampling bottles. There is an advantage to using
a bottle and a sampling cage instead of a weighted beaker for
volatile products. The loss of light ends is likely to occur
.when transferring the sample from a weighted beaker to
another container.
8.3 Tap Sampling:
8.3.1 Application—The tap sampling procedure is appli-
cable for sampling liquids of 26 psi (179 kPa) Rvp or less in
• tanks that are equipped with suitable sampling taps or lines.
This procedure is recommended for volatile stocks in tanks
of the breather and balloon-roof type, spheroids, etc.
(Samples may be taken from the drain cocks of gage glasses.
if the tank is not equipped with sampling taps.) The assembly
for tap sampling is shown in Fig. 3.
NOTE 9—If Rvp is more than 16 psi (110 kPa) but not more than 26
psi (179 kPa), a cooling bath as shown in Fig. 10 shall be used between
the tank tap and the sample container to cool the sample and container
and prevent volatilization of low-boiling components.
8.3.2 Apparatus:
8.3.2.1 Tank Taps—The tank should be equipped with at
least three sampling taps placed equidistant throughout the
tank height and extending at least 3 ft (0.9 m) inside the tank
shell. A standard V* in. pipe with suitable valve is satisfac-
tory.
8.3.2.2 Sample Cocks—Samples of crude petroleum may
be taken through sample cocks properly placed in the shell of
the tank. The upper sample cock shall be located 18 in. (457
mm) below the top of the tank shell; the lower sample cock
RG. 3 Assembly for Tap Sampling
shall be located level with the bottom of the outlet elbow or
other similar fitting if installed on the outlet connection: and
the middle sample cock shall be located halfway between the
upper and lower sample cocks. An additional cock for the
clearance sample should be located 4 in. (102 mm) below the
bottom of the outlet connection to determine whether the
level of merchantable oil is at least below this point. The
sample cocks should be located a minimum of 6 ft (1.8 m)
distant circumferentially from the pipeline outlet and drain
connection or at the top of upturned connections, and 8 ft
(2.4 m) from the filling line connection. The sample cocks
should be of V* in. size, and the lines should be of V* in.
nominal diameter for crude oil of 18* API gravity or less. For
lighter oil, '/>in. size cocks, with '/z-in. nominal diameter
lines, should be used. The lines should extend a minimum of
4 in. (102 mm) inside the tank shell, except on floating-roof
tanks, where flush installations are necessary. All sample
cocks should be equipped with scalable valves and plugged
inspection tees. ,
8.3.22 (7) On tanks of more than 10 000-barrel capacity,
at least two sets of sample cocks shall be installed, located.
equidistant around the circumference of the tank. Five or
more sample cocks should be installed per set, evenly spaced
between lower and upper sample levels.
8.3.2.3 Tube—A delivery tube that will not contaminate
the product being sampled and long enough to reach to the:
bottom of the sample container is required to allow sub-
merged filling. When a cooling bath is used while tap
sampling, a similar suitable tube should be used between the
tank tap and the cooler inlet.
8.3.2.4 Sample Containers—Use clean, dry glass bottles of
convenient size and strength to receive the samples. If the
vapor pressure of the product to be sampled is between 16
and 26 psi (110 and 179 kPa) Rvp, protect the bottle with a
metal cover until the sample is discarded. In some cases,*
such as the sampling of crude petroleum, metal containers
may be used instead of glass bottles.
8.3.3 Procedure:
8.3.3.1 Before a sample is drawn, flush the tap (or gage
glass drain cock) and line until they are purged completely.
Connect the clean delivery tube to the tap. Draw upper,
middle, or lower samples directly from the respective taps
after the flushing operation. Stopper and label the sample
container immediately after filling, and deliver it to the
laboratory.
8.3.3.2 When a sample cooler is used during the tap
sampling operation, flush the tap (or gage glass drain cock).
Then, using a section of clean tubing, connect the tap to the
cooler inlet. Flush the cooler thoroughly, after which connect
the clean delivery tube to the cooler outlet and proceed with
the sampling operation.
8.3.3.3 In the sampling of crude petroleum, check for
merchantable oil at the clearance sample cock. Flush each
sample connection until all oil from the previous run has
been removed and the sample lines are filled with fresh oil
from the tank.
8.3.3.3 (7) On tanks of 10 000-barrel capacity or smaller.
take samples of equal amounts from the lower, middle, and
upper sample connections. A measuring cup of proper size
may be used to assure the drawing of the proper quantity
from each sample cock.
Page 69
-------
8.3.3.3 (2) On tanks of more than 10 000-barrcl capacity
take samples of equal amounts from each of the sample
connections at each set of sample connections.
8.3.3.3 (3) Mix all samples in equal proportions for a
composite sample, or test the samples separately and averaee
the results. ' K
8.3.3.3 (4) When crude oil in a tank fails to reach the
upper or middle sample cocks on a tank equipped with three
sample cocks, it is suggested that the sample for the run be
obtained as follows: if the level of the oil is nearer the upper
sample cock than the middle, take two thirds of the sample
from the middle sample cock and one third from the lower
If the level of oil is nearer the middle sample cock than the
.upper, take one half of the sample from the middle and one
half from the lower. If the level of the oil is below the middle
sample cock, take all of the sample from the lower cock
8.4 Manual Sampling of Pipeline Streams:
8.4.1 Application—This manual line sampling procedure
is applicable for sampling liquids of 16 psi (110 kPa) Rvp or
less and semiliquids in pipelines, filling lines, and transfer
lines. Where custody transfer is involved, continual auto-
matic sampling is the preferred method. In the event of
automatic sampler failure, manual samples may be needed
Take such manual samples as representatively as possible
8.4.2 Apparatus:
8.4.2.1 Sampling Probe—Tin function of the sampline
probe is to withdraw from the flowing stream a portion that
will be representative of the entire stream. Probe designs that
are commonly used are shown in Fig. 4. These are-
hJ^lA'U* tUbf ?tt:nding to the cemer °f the line and
beveled at a 45 angle facing upstream (Fig. 4(a))
8.4.2.1 (2) A long-radius forged elbow or pipe bend
extending to the center line of the pipe and facing upstream
The end of the probe should be reamed to give Ysharo
entrance edge (Fig. 4 (&)). * ^^
8.4.2.1 (J) A closed-end tube with a round orifice spaced
near the closed end that should be positioned in such Twav
that the onfice is in the center of the pipeline and is facine
the stream as shown in Fig. 4(c).
8.4 3 Probe Location—Since the fluid to be sampled mav
not always be homogeneous, the location, the position and
the size of the sampling probe should be such as to minimize
any separation of water and heavier panicles that would
make their conccmnuion different in the gathered sample
than in the main stream.
8.4.3.1 The probe should always be in a horizontal
position to prevent drainback of any pan of the sample to
the mam stream.
8.4.3.2 The sampling probe should be located preferably
in a vertical run of pipe where such a vertical run can be
provided. The probe may also be located in a horizontal run
of pipe provided the flowing velocity is high enough to
provide adequate turbulent mixing. While adequate flowing
velocity may not eliminate a concentration difference be-
tween the bottom of the pipe and the top of the pipe, it may
provide an average concentration at the mid-pipe probe
location that will be representative of the entire stream at the
sampling station.
8.4.3.3 Where adequate flowing velocity is not available, a
suitable device for mixing the fluid flow should be installed
upstream of the sampling tap to reduce stratification to an
acceptable level. If flow has been vertical for a sufficient
distance as in a platform riser, such a device may not be
necessary even at low-flow rates. Some effective devices for
obtaining adequate mixing are: a reduction in pipe size; a
series of baffles: an orifice or perforated plate; or a combina-
tion of any of these methods. The design or sizing of the
device is optional with the user, as long as the flowing stream
is sufficiently well mixed to provide a representative sample
from the probe.
8.4.3.4 The sampling point should be as near as practi-
cable to the point where the oil passes to the sample receiver.
8.4.3.5 The sampling lines should be as shon as practi-
cable and should be cleared before any samples are taken.
8.4.3.6 To control the rate at which the sample is with-
drawn, the probe should be fitted with valves or plug cocks.
8.4.4 Procedure:
8.4.4.1 Adjust the valve or plug cock from the sampling
probe so that a steady stream is drawn from the probe.
Whenever possible, the rate of sample withdrawal should be
such that the velocity of liquid flowing through the probe is
approximately equal to the average linear velocity of the
stream flowing through the pipeline. Measure and record the
rate of sample withdrawal as gallons per hour (or litres per
hour). Divert the sample stream to the sampling container
continuously or intermittently to provide a quantity of
sample that will be of sufficient size for analysis.
8.4.4.2 In sampling crude petroleum, samples of'/: pt (0.2
L) or more should be taken every hour or less, whichever is
thought necessary. By mutual agreement, the sample period
or sample size, or both, may be varied to accommodate the
parcel size. It is important that the size of the samples and
the intervals between the sampling operations be uniform for
a uniform-flow rate. When the mainstream flow rate is
variable, vary the sampling rate accordingly so that the
amount of sample is always withdrawn from any given
-1/4-PIPE
TO RECEIVER
Qtt % AMPI.CM
BCVCU
< At
„ — . ORIFICE IN ftioc or »*oae
i
u>
1
/=
1C
\M
\\o.\\
iidll
l/«»-l/4- PIPE
TO RECEIVER
OR SAMPLER
a-
•ENO REAMED ro
A SHARP EOCC
s
M rrm» vnm mum* w MM (MM.
FIG. 4 Probes for Continuous Sampling
Page 70
-------
D4057
amount of fluid passing the sampling point in the main-
stream, in practice, this is difficult to accomplish.
8.4.4.2 (7) Place the sample of crude petroleum in a
closed container and at the end of the agreed-upon time
period, mix the combined samples and take a composite
sample for test purposes. Refer to 7.4 for mixing and
handling. Store the sample container in a cool, dry place;
avoid exposure to direct sunlight.
8.4.4.2 (2) Alternatively, line samples may be taken at
regular intervals and individually tested. The individual test
results may be arithmetically averaged, adjusting for varia-
tions in flow rate during the agreed upon time period.
8.4.4.2 (J) Either composited or arithmetically averaged
results are acceptable.
8.4.4.3 With either procedure, always label the sample
and deliver it to the laboratory in the container in which it
was collected.
NOTE 10—When sampling semiiiquids. heat the sampler lines,
valves, and receiver to a temperature just sufficient to keep the material
liquid and to assure accurate sampling and mixing.
8.5 Dipper Sampling:
8.5.1 Application—The dipper sampling procedure is ap-
plicable for sampling liquids of 2 psi (13.8 kPa) Rvp or less
and semiiiquids where a free- or open-discharge stream
exists, as in small filling and transfer pipelines (2-in. diameter
or less) and filling apparatus for barrels, packages, and cans.
8.5.2 Apparatus:
8.5.2.1 Dipper—Use a dipper with a flared bowl and a
handle of convenient length, made of material such as tinned
steel that will not affect the product being tested. The dipper
should have a capacity suitable for the amount to be
collected and must be protected from dust and dirt when not
being used.
8.5.2.2 Sample Container—Use a clean, dry container of
the desired size.
8.5.3 Procedure—Insert the dipper in the free-flowing
stream so that a portion is collected from the full cross
section of the stream. Take portions at time intervals chosen
so that a complete sample proportional to the pumped
quantity is collected. The gross amount of sample collected
should be approximately 0.1 %. but not more than 40 gal
(151 L), of the total quantity being sampled. Transfer the
portions into the sample container as soon as collected. Keep
the container closed, except when pouring a dipper portion
into it. As soon as all portions of the sample have been
collected, close and label the sample container, and deliver it
to the laboratory.
8.6 Tube Sampling:
8.6.1 Application—The tube sampling procedure is appli-
cable for sampling liquids of 2 psi (13.8 kPa) Rvp or less and
semiiiquids in drums, barrels, and cans.
8.6.2 Apparatus:
- 8.6.2.1 Tube—Either a glass or metal tube may be used,
designed so that it will reach to within about'/«in. (3.2 mm)
of the bottom and have a capacity of approximately 1 pt (0.5
L) or 1 qt (0.9 L). A metal tube suitable for sampling 50 gal
(190 L) drums is shown in Fig. 5. Two rings soldered to
opposite sides of the tube at the upper end are convenient for
holding it by slipping two fingers through the rings, thus
leaving the thumb free to close the opening.
ACT
8.62.2 Sample Containers—-Use clean, dry cans or glass
bottles.
8.6.3 Procedure:
8,6.3.1 Drums and Barrels—Place the drum or barrel on
its side with the bung up. If the drum does not have a side
bung, stand it upright and sample from the top. If detection
of water, rust, or other insoluble contaminants is desired, let
the barrel or drum remain in this position long enough to
permit the contaminants to settle. Remove the bung and
place it beside the bung hole with the oily side up. Cose the
upper end of the clean, dry sampling tube with the thumb
and lower the tube into the oil for a depth of about 1 ft (0.3
m). Remove the thumb, allowing oil to flow into the tube.
Again close the upper end with the thumb and withdraw the
tube. Rinse the tube with the oil by holding it nearly
horizontal and turning it so that the oil comes in contact
with that pan of the inside surface that will be immersed
when the sample is taken. Avoid handling any part of the
tube that will be immersed in the oil during the sampling
operation. Discard the rinse oil and allow the tube to drain.
Insert the tube into the oil again, holding the thumb against
the upper end. (If an all-levels sample is desired, insert the
tube with the upper end open.) When the tube reaches the
bottom, remove the thumb and allow the tube to fill.
Replace the thumb, withdraw the tube quickly, and transfer
the contents to the sample container. Do not allow the hands
to come in contact with any part of the sample. Close the
sample container replace and tighten the bung in the drum
Page 71
-------
D4057
or barrel. Label the sample container and deliver it to the
laboratory.
8.6.3.2 Cans—Obtain samples from cans of 5-gal (I9-L)
capacity or larger in the same manner as from drums and
barrels (8.6.3.1), using a tube of proportionately smaller
dimensions. For cans of less than 5-gal (I9-L) capacity, use
the entire contents as the sample, selecting cans at random as
indicated in Table 3 or in accordance with agreement
between the purchaser and the seller.
8.7 Thief Sampling:
8.7.1 Application—The thief sampling procedure is appli-
cable for obtaining bottom samples (Fig. 1), or of semiliquids
in tank cars and storage tanks.
8.7.1.1 The thief is also widely used in sampling crude
petroleum in storage tanks. In this application it may be used
for taking samples at different levels as well as for bottom
samples of nonmerchantable oil and water at the bottom of
the tank.
8.7.2 Apparatus:
8.7.2.1 Thief—The thief shall be designed so that a
sample can be obtained within '/: in. (13 mm) of the bottom
of the car or tank. Two types of thiefs are illustrated in Fig. 6.
One type is lowered into the tank with valves open to permit
the oil to flush through the container. When the thief strikes
the bottom of the tank, the valves shut automatically to trap
a bottom sample. The other type has a projecting stem on the
valve rod which opens the valves automatically as the stem
strikes the bottom of the tank. The sample enters the
container through the bottom valve and air is released
simultaneously through the top. The valves snap shut when
the thief is withdrawn. A core-type thief similar to that
shown in Fig. 6(6), with a uniform cross section and bottom
closure and with a capacity depending upon the size of the
sample required, may be used for sampling crude petroleum.
The thief should be capable of penetrating the oil in the tank
to the required level, mechanically equipped to permit filling
at any desired level, and capable of being withdrawn without
undue contamination of the contents. The thief may be
equipped with the following:
8.7.2.1 (/) Sample cocks for obtaining samples for the
determination of water and sediment;
8.7.2.1 (2) Extension rods for use in obtaining samples at
levels corresponding with requirements for high connections
or for samples to determine high settled water and sediment
levels;
8.7.2.1 (J) Water and sediment gage for determining the
height of water and sediment in the thief;
8.7.2.1 (4) Windshield to be used when taking the gravity
and temperature of the oil;
8.7.2.1 (5) Opener to break the tension on the valve or
slide at any desired level;
8.7.2.1 (6) A thief cord marked so that sample can be
taken at any depth in the vertical cross section of the tank;
and
8.7.2.1 (7) Hook to hang the thief in the hatch vertically.
8.7.2.2 Sample Containers—Use clean, dry cans or glass
bottles.
8.7.3 Procedure:
8.7.3.1 Lower the clean, dry thief through the dome of the
tank car or tank hatch until it strikes the bottom. When full
remove the thief and transfer the contents to the sample
container. Close and label the container immediately, and
deliver it to the laboratory.
8.7.3.2 In the sampling of crude petroleum lower the
LIKE F0»
LOWERING
t-i 314- _
DIAMETER
— 4 LUGS
•*- HIGH
TYPE SAUPUNO TMEF
B COMC THKF. THAT TYPE
RG. 6 Sampling Thtofs
Page 72
-------
4SII) D4057
clean, dry thief slowly into the oil to the desired depth, trip
the thief to secure the sample and raise slowly to avoid
agitation. The proper depths for sampling are described in
4.1.
8.8 Extended-Tube Sampling:
8.8.1 Application—The extended-tube sampling proce-
dure may be used only for obtaining bottom water samples.
NOTE 1—The user should be aware that the procedure is used for
sampling bottom water primarily on ships and barges. The procedure
. may be used for sampling bottom water in shore tanks, but no specific
guidelines for such use are available.
8.8.2 Apparatus:
8.8.2.1 A typical extended-tube sampling assembly is
shown in Fig. 7. The extended-tube sampler consists of a
flexible tube connected to the suction of a manually operated
pump. For support purposes and to establish a known
sampling point the tubing is attached to the weighted end of
a conductive wire or tape such that the open end of the tube
is located approximately '/z in. above the tip of the weight.
The tubing and wire (or tape) shall be long enough to extend
to the bottom (reference height) of the vessel or storage tank
from which the sample is to be obtained. A grounding cable
shall be provided for the assembly.
8.8.2.2 In addition to the sampler, a clean, dry bottle or
other appropriate container is required to collect each
sample.
8.8.3 Procedure:
8.8.3.1 Assemble the extended-tube sampler.
8.8.3.2 Lower the weighted end of the sampling tube into
a bucket of water and prime the sampler by operating the
pump. When the assembly is satisfactorily primed, close-off
(Ensure it's not vented to the atmosphere.) the top end of the
assembly to prevent the loss of priming water as the sampling
tube is removed from the bucket Remove the sampling tube
from the bucket, connect the grounding cable to the ship or
barge tank, and lower the weighted sampling tube to the
bottom of the tank.
8.8.3.3 Begin the sampling operation by slowly and
steadily operating the manual pump. To reduce the possi-
bility of capturing a contaminated sample, collect a volume
greater than twice the sampling assembly's capacity in a
graduated cylinder or other measuring device and discard
this volume (Note 2). Collect the sample(sj directly in a
clean, dry bottle(s) or other appropriate containers).
NOTE 2—The capacity of the sampler may be calculated from the
sampler dimensions or may be determined by measuring the volume of
liquid discharge from a completely full assembly.
8.8.3.4 If a sample at a different level within the bottom
water layer is required, raise the weighted bob and tubing to
the new level above the bottom. Purge the residual water
from the assembly and collect the new sample(s) as in-
structed above. 1'
8.8.3.5 After each sample has been collected, immediately
close and label the bottle (or container) in preparation for
delivery to the laboratory.
8.8.3.6 When the sampling operation is complete, clean
and disassemble the sampler components. ,
8.9 Boring Sampling: "~
8.9.1 Application—The boring sampling procedure is ap-
plicable for sampling waxes and soft solids in barrels, cases,
bags, and cakes when they cannot be melted and sampled as
liquids. ,'
8.9.2 Apparatus: '.
8.9.2.1 Ship Auger—Use a ship auger V* in. (19 mm) in
diameter, similar to that shown in Fig. 8, and of sufficient
length to pass through the material to be sampled. \
8.9.2.2 Sample Containers—Use clean, wide-mouth,
metal containers or glass jars with covers.
8.9.3 Procedure—Remove the heads or covers of barrels'
and cases. Open bags and wrappings of cakes. Remove any
dirt, sacks, string, or other foreign substances from the
surface of the material. Bore three test holes through the
body of the material, one at the center, the other two halfway
between the center and the edge of the package on the right
Suppo
FIG. 7 Typical Extended-Tube Sampler
Page 73
-------
D4057
FIG- 8 Ship Auger for Boring Procedure
and left sides, respectively. If any foreign matter is removed
from the interior of the material during the boring operation
include it as part of the borings. Put the three sets of borings
in individual sample containers, label and deliver them to
the laboratory.
8.9.4 Laboratory Inspection—If there are any visible dif-
ferences in the samples, examine and test each set of borings
.at the laboratory. Otherwise, combine the three sets of
borings into one sample. If subdivision of the borings is
desired, chill, pulverize (if necessary), mix, and quarter the
borings until reduced to the desired amount.
8.10 Grab Sampling:
8.10.1 Application:
8.10.1.1 The grab sampling procedure is applicable for
sampling all lump solids in bins, bunkers, freight cars.
barrels, bags, boxes, and conveyors. It is particularly appli-
cable for the collection of green petroleum coke samples
from railroad cars and for the preparation of such samples
for laboratory analysis. Refer to Method D 346 when other
methods of shipping or handling are used.
8.10.1.2 Place of Sampling— Petroleum coke may be
sampled while being loaded into railroad cars from piles or
after being loaded into railroad cars from coking drums.
8.10.2 Apparatus:
8.10.2.1 Sample Container— A polyethylene pail of ap-
proximately 10-qt (9.5-L) capacity.
8.10.2.2 Scoop, stainless steel or aluminum. No 2 size
8.10.3 Procedure:
8.10J.I Samplings-Lumpy solids are usually heteroge-
neous and therefore are difficult to sample accurately. It is
preferable to take samples during the unloading of cars, or
during transit of the material by conveyors. From material in
transit, obtain a number of portions at frequent and regular
intervals and combine them.
8.10.3.2 When sampling from railroad cars, use one of the
following procedures:
8.10.32 (/) Being Loaded from a ftVe—Take a full scoop
of sample at each of the five sampling points shown in Fig 9
and deposit in the polyethylene pail. Cover the sample and
deliver to the laboratory. Each sampling point shall be
located equidistant from the sides of the railroad car
8.10.3.2 (2) Railroad Cars After Direct Loading from
Coking Drums—At any five of the sampling points shown in
M-
•I'lCNITM Or CM-
•
-V —
XI*
—»-
oo oo
FIG. 9 Location of Sampling Points at Different Levels of Car
Ftg. 10, take a full scoop of coke from about 1 ft (0.3 m)
below the surface and deposit it in the polyethylene pail.
Cover the sample and deliver to the laboratory.
8.10.3.3 When sampling from conveyors, take one scoop
for each 8 to 10 tons (7.9 to 9 Mg) of coke transported. These
samples may be handled separately, or composited after all
samples representing the lot have been taken.
8.10.3.4 When sampling from bags, barrels, or boxes,
obtain portions from a number of packages selected at
random as shown in Table 3, or in accordance with the
agreement between the purchaser and the seller.
8.10.3.5 Quartering—Carefully mix the sample and re-
duce it in size to a convenient laboratory sample by the
quartering procedure described in Method D 346. Perform
the quartering operation on a hard, clean surface, free from
cracks, and protected from rain, snow, wind, and sun. Avoid
contamination with cinders, sand, chips from the floor, or
any other material. Protect the sample from loss or gain of
moisture or dust Mix and spread the sample in a circular
layer, and divide it into quadrants. Combine two opposite
quadrants to form a representative reduced sample. If this
sample is still too large for laboratory purposes, repeat the
quartering operation. In this manner, the sample will finally
be reduced to a representative, suitable size for laboratory
purposes. Label and deliver the sample to the laboratory in a
suitable container. Subdivision may be made in the labora-
tory by using a riffle sampler as Method D 346.
8.11 Grease Sampling:
8.11.1 Application—This method covers practices for ob-
taining samples representative of production lots or ship-
ments of lubricating greases, or of soft waxes or soft
bitumens similar to grease in consistency. The procedure is
necessarily quite general to cover the wide variety of condi-
tions encountered, and may require modification to meet
individual specifications. Proceed in accordance with 3.1 to
7.7. particularly those pertaining to precautions, care, and
cleanliness, except where they conflict with 8.11.2 to 8.11.4.
8.11.2 Inspection:
8.11.2.1 If the material is a lubricating grease and inspec-
tion is made at the manufacturing plant, take samples from
the finished shipping containers of each production batch or
lot. Never take grease samples directly from grease kettles,
cooling pans, tanks, or processing equipment. Do not sample
the grease until it has cooled to a temperature not more than
15T(8.3*C) above that of the air surrounding the containers
and until it has been in the finished containers for at least 12
h. When the containers for a production batch of grease are
of different sizes, treat the grease in each size of container as
a separate lot. When inspection is made at the place of
delivery, obtain a sample from each shipment If a shipment
consists of containers from more than one production batch
(lot numbers), sample each such batch separately.
8.11.2.2 If the material being inspected is of grease-like
consistency, but is not actually a lubricating grease, but some
mixture of heavy hydrocarbons such as microcrystalline
waxes or soft bitumens, it will be permissible to take samples
Page 74
-------
D4057
from pans, tanks, or other processing equipment, as well as
from containers of the finished product. The grease sampling
method shall be applicable to such stocks only if for some
reason it is not possible to apply heat and convert the
. material into a true liquid.
8.11.3 Sample Si:e—Select containers at random from
each lot or shipment to give the desired quantity specified in
Table 5.
8.11.4 Procedure:
. 8.11.4.1 Inspection—Examine the opened containers to
determine whether the grease is homogeneous, comparing
the grease nearest the outer surfaces of the container with
that in the center, at least 6 in. (152 mm) below the top
"surface, for texture and consistency. When more than one
container of a lot or shipment is opened, also compare the
grease in the respective containers.
8.11.4.2 Sampling—If no marked difference in the grease
• is found, take one portion from the approximate center and
at least 3 in. (76 mm) below the surface of each opened
container, in sufficient quantity to provide a composite
sample of the desired quantity (Table 5). Withdraw portions
with a clean scoop, large spoon, or spatula and place them in
a clean container. Very soft, semifluid greases may be
sampled by dipping with a Mb (0.45-kg) can or suitable
dipper. If any marked difference in the grease from the
various locations of an opened container is found, take two
separate samples of about 1 Ib (0.45 kg) each, one from the
top surface adjacent to the wall, and the other from the
center of the container, at least 6 in. (152 mm) below the top
surface. If any marked variations are noted between different
containers of a lot or shipment, take separate samples of
about 1 Ib (0.45 kg) from each container. When more than
one sample of a batch or shipment is taken because of lack of
uniformity, send them to the laboratory as separate samples.
8.11.4.3 Handling Samples—If more than one portion is
required to represent a lot or shipment of grease softer than
175 penetration (see Method D217), prepare a composite
sample by mixing equal portions thoroughly. Use a large
spoon or spatula in a clean container. Avoid vigorous mixing
or working of air into the grease. As grease samples become
partially "worked" in being removed from containers, the
procedure is not suitable for obtaining samples of greases
softer than 175 penetration on which unworked penetration
is to be determined. For greases having a penetration less
than 175, cut samples from the container with a knife in the
form of blocks about 6 by 6 by 2 in. (152 by 152 by 51 mm).
If required, make unworked penetration tests on blocks as
TABLE 5 SUe of Grease Samples
Container
Tubes or packages, less
- than 1 Ib
1-ibcans
5 or 10-lb cans
Larger man 10lb
Larger than 10 Ib
Larger than 10 Ib
Lot or Shipment
all
all
all
less than 10 000 It)
10 000 to SO 000 Ib
more than SO 000 Ib
Mmmum Sample
enough units tor a 2-lb
sample
three cans
one can
2 to 3 Ib from one or more
containers
2 to 5 Ib from two or more
cofttttners
2 to 5 to from three or
more containers
procured, and other inspection tests on grease cut from the
blocks.
9. Sampling Industrial Aromatic Hydrocarbons
9.1 Application—For obtaining samples of industrial aro-
matic hydrocarbons (benzene, toluene, xylene. and solvent
naphthas), proceed in accordance with Section 6 to 8.
particularly those pertaining to precautions, care, and clean-
liness.
10. Sampling for Specific Tests
10.1 Special Precautions—Special sampling precautions
and instructions are required for some ASTM test methods
and specifications. Such instructions, 11.2 to 11.3. supple-
ment the general procedures of this method and supersede
them if there is a conflict. ASTM methods in this category
are as follows:
ASTM Test Methods
D2I6
D323
D523
D873
0268
D 1856
D244
Sections
11.2
11.3
11.4
11.4
11.5
10.6
10.7
10.2 Distillation of Natural Gasoline—When obtaining
samples of natural gasoline that are to be tested using
Method D216. the bottle sampling procedure. 8.1.3 is
preferred. Before obtaining the sample, precooi the bottle by
immersing it in the product, allow it to fill, and discard the
first filling. If the bottle procedure cannot be used, obtain the
sample by the tap procedure, and with the use of the cooling
bath, as described in 8.2.3. Do not agitate the bottle while
drawing the sample. After obtaining the sample, close the
bottle immediately with a tight-fining stopper and store it in
an ice bath or refrigerator at a temperature of 32 to 40T (0 to:
4.5'C).
10.3 Reid Vapor Pressure:
10.3.1 When sampling products that are to be tested using
Method D 323. observe the following precautions and in-
structions:
10.3.1.1 Precautions—Vapor pressures are extremely sen-
sitive to evaporation losses and to slight changes in compo-
sition. When obtaining, storing, or handling samples, ob-
serve the necessary precautions to ensure samples
representative of the product and satisfactory for Reid vapor
pressure tests. Official samples should be taken by. or under
the immediate supervision of. a person of judgment, skill.
and sampling experience. If sampling or sample require-
ments for other tests differ from those described in 10.3.1.2
to 10.3.1.9. obtain a separate sample for the Reid vapor
pressure test. Never prepare composite samples for this test.
Make certain that containers which are to be shipped by
common carrier conform to I.C.C., state, or local regula-
tions. When flushing or purging lines or containers, observe
the pertinent regulations and precautions against fire, explo-
sion, and other hazards.
10.3.1.2 Cooling Bath—k bath (Fig. 11) of sufficient size
to hold the sample container and a cooling coil of about 25 ft
(7.6 rn) of copper tubing (% in. (9.5 mm) or less in outside
diameter) shall be required when using the procedure de-
scribed in 10.3.1.7. One end of the coil is provided with a
Page 75
-------
D4057
connection for attaching it to the tank sampling tap or valve.
The other end is fitted with a suitable valve (outlet) of good
quality. A removable copper tube of V* in. (9.5 mm) or less
m outside diameter and of sufficient length to reach the
bottom of the sample container shall be.connected to the
open end of the outlet valve.
10.3.1.3 Sample Containers—Use containers of not less
than 1 qt (0.9 L) nor more than 2-gal (7.6-L) capacity, of
sufficient strength to withstand the pressures to which they
may be subjected, and of a type that will permit replacement
of the cap or stopper with suitable connections for transfer-
ring the sample to the gasoline chamber of the vapor pressure
apparatus. Open-type containers have a single opening which
permits sampling by immersion. Closed-type containers have
two openings; one in each end (or the equivalent thereof),
fitted with valves suitable for sampling by water displace-
ment or by purging.
10.3.1.4 Transfer Connections—The transfer connection
for the open-type container consists of an air tube and a
liquid delivery tube assembled in a cap or stopper. The air
tube extends to the bottom of the container. One end of the
liquid delivery tube is flush with the inside face of the cap or
stopper, and the tube is long enough to reach the bottom of
the gasoline chamber while the sample is being transferred to
the chamber. The transfer connection for the closed-type
container consists of a single tube with a connection suitable
for attaching it to one of the openings of .the sample
container. The tube is long enough to reach the bottom of
the gasoline chamber while the sample is being transferred.
10.3.1.5 Sampling Open Tanks—Use clean containers of
the open type when sampling open tanks and tank cars. An
all-level sample obtained by the bottle procedure, 8.2.3, is
recommended. Before taking the sample, flush the container
by immersing it in the product to be sampled. Then obtain
the sample immediately. Pour off enough so that the
container will be 70 to 80 % full and close it promptly. Label
the container and deliver it to the laboratory.
10.3.1.6 Sampling Closed Tanks—Containers of either
the open or closed type may be used to obtain samples from
closed or pressure tanks. If the open type is used, follow the
cooling bath procedure described in 10.3.1.7. If the closed
type is used, obtain the sample using the water displacement
procedure. 10.3.1.8, or the purging procedure, 10.3.1.9. The
water displacement procedure is preferable because the flow
of product involved in the purging procedure may be
hazardous.
10.3.1.7 Cooling Bath Procedure—When using a con-
tainer of the open type, keep it at a temperature of 32 to 40*F
(0 to 4.5*C) during the sampling operation by using the
cooling bath (Fig. 11). Connect the coil to the tank sampling
tap or valve and flush it with a sufficient amount of product
to ensure complete purging. When obtaining a sample,
throttle the outlet valve so that the pressure in the coil will be
approximately the same as that in the tank. Fill the container
once to wash and cool it, and discard the wash product. Then
draw the sample immediately. Pour off enough so that the
container will be 70 to 80 % full and close it promptly. Label
the container and deliver it to the laboratory.
10.3.1.8 Water Displacement Procedure—Completely fill
the closed-type container with water and close the valves.
The water should be at the same temperature or lower than
-fe-r-fr—a
i
i
1
r-pr-rw-r-is-
i 1
i
i
"TOT-To-— Tff-
! !
i
I
•w-
»- s
w 8
i.i 1 T
Tf
»;• *,
RG. 10 Location of Sampling Points from Exposed
Surface of Car
that of the product to be sampled. While permitting a small
amount of product to flow through the fittings, connect the
top or inlet valve of the container to the tank sampling tap or
valve. Then open all valves on the inlet side of the container.
Open the bottom or outlet valve slightly to allow the water to
be displaced slowly by the sample entering the container.
Regulate the flow so that there is no appreciable change in
pressure within the container. Close the outlet valve as soon
as gasoline discharges from the outlet; then in succession
close the inlet valve and the sampling valve on the tank.
Disconnect the container and withdraw enough of the
contents so that it will be 70 to 80 % full. If the vapor
pressure of the product is not high enough to force liquid
from the container, open both the upper and lower valves
slightly to remove the excess. Promptly seal and label the
container, and deliver it to the laboratory. The above is not
applicable to LPG sampling.
10.3.1.9 Purging Procedure—Connect the inlet valve of
the closed-type container to the tank sampling tap or valve.
Throttle the outlet valve of the container so that the pressure
in it will be approximately equal to that in the container
being sampled. Allow a volume of product equal to at least
twice that of the container to flow through the sampling
system. Then close all valves, the outlet valve first, the inlet
valve of the container second, and the tank sampling valve
last, and disconnect the container immediately. Withdraw
enough of the contents so that the sample container will be
70 to 80 % full. If the vapor pressure of the product is not
high enough to force liquid from the container, open both
the upper and lower valves slightly to remove the excess.
Promptly seal and label the container and deliver it to the
laboratory.
10.4 Oxidation Stability:
10.4.1 When sampling products that are to be tested for
oxidation stability in accordance with Method D525.
Method D873, or by equivalent methods, observe the
following precautions and instructions:
10.4.1.1 Precautions—Very small amounts (as low as
0.001 %) of some materials, such as inhibitors, have a
considerable effect upon oxidation stability tests. Avoid
contamination and exposure to light while taking and
handling samples. To prevent undue agitation with air which
promotes oxidation, do not pour, shake, or stir samples to
any greater extent than necessary. Never expose them to
temperatures above those necessitated by atmospheric condi-
tions.
10.4.1.2 Sample Containers—Use only brown glass or
wrapped clear glass bottles as containers, since it is difficult
to make certain that cans arc free of contaminants, such as
Page 76
-------
D4057
rust and soldering flux. Clean the bottles by the procedure
described in 8.1.3.3 and. if possible, finish with a cleaning
solution of sulfuric acid and potassium dichromate. Rinse
thoroughly with distilled water, dry. and protect the bottles
from dust and dirt.
10.4.1.3 Sampling—An ail-levels sample obtained by the
bottle procedure, 8.2.3.1. is recommended because the
sample is taken directly in the bottle. This reduces the
possibility of air absorption, loss of vapors, and contamina-
tion. Just before sampling, rinse the bottle with the gasoline
to be sampled.
10.5 Lacquer Solvents and Diluents:
10.5.1 When sampling bulk shipments of lacquer solvents
. and diluents that are to be tested using Method D 268,
observe the following precautions and instructions:
10.5.1.1 Tanks and Tank Cars—Obtain upper and lower
samples (Fig. 1) of not more than 1 qt (0.9 L) each by the
bottle procedure, 8.2.3.2. Prepare in the laboratory a com-
posite sample of not less than '/z gal (1.9 L) by mixing equal
parts of the upper and lower samples.
10.5.1.2 Barrels, Drums, and Cans—Obtain samples
from at least five percent of the number of containers in any
shipment. The number of containers to be sampled may be
increased at the discretion of the purchaser. In the case of
expensive solvents which are purchased in small quantities, it
is recommended that each container be sampled. Withdraw
a portion from the center of each container to be sampled
with a clean tube (8.6.3) or weighted bottle (8.2.3.2). (A
smaller bottle may be used.) Prepare a composite sample of
at least 1 qt (0.95 L) by mixing equal portions of not less
than 1 pt (0.47 L) from each container sampled.
10.6 Asphaltic Materials—When sampling asphaltic ma-
terials that are to be tested using Method D 1856 or Method
D2172, obtain samples by the boring procedure (8.9.3) or
the grab procedure (8.9.3). A sample of sufficient size to yield
at least 100 g of recovered bitumen is required. About 1000 g
of sheet-asphalt mixtures usually will be sufficient. If the
largest lumps in the sample are 1 in. (25.4 mm), 2000 g will
usually be required, and still larger samples if the mixtures
contain larger aggregates.
10.7 Emulsified Asphalts—It is frequently necessary to
test samples in accordance with the requirements of Specif!-'
cations D 977, and Method D 244. Obtain samples from
tanks, tank cars, and tank trucks by the bottle sampling
procedure, 8.2.3, using a wide-mouth (I'/i-in. (38.1-mm) or.
greater) bottle. Use the dipper procedure, 8.5.3, to obtain
samples from filling or discharge lines. Sample packages in
accordance with Table 3. If the material is solid or semisolid,;
use the boring sampling procedure, 8.9.3. Obtain at least 1.'
gal (13.8 L) or 10 Ib (4.5 kg) from each lot or shipment Store.
the samples in clean, airtight containers at a temperature of.
not less than 4*C until tested. Use glass or black iron.
containers for emulsified asphalts of the RS-1 type.
OUTLET
VALVE
3 TO TANK
-PURGING VALVE
COPPER TUBING
SOS FT. • 1/4 00»
TOP VIEW
FIG. 11 Cooling Bath for Reid Vapor Pressure Sampling
Page 77
-------
IfllP 04057
ANNEX
(Mandatory Information)
Al. PRECAUTIONARY STATEMENTS
A 1.1 The following substances may be used throughout
the course of this standard test method. The precautionary
statements should be read prior to use of such substances.
A 1.1.1 Benzene
• • Keep away from heat, sparks, and open flame.
Keep container closed.
Use with adequate ventilation.
Use fume hood whenever possible.
Avoid build-up of vapors and eliminate all sources of
ignition, especially non-explosion proof electrical apparatus
and heaters.
Avoid prolonged breathing of vapors or spray mist
Avoid contact with skin and eyes. Do not take internally.
A 1.1.2 Diluent (Naphtha)
Keep away from heat, sparks, and open flame.
Keep container closed.
Use with adequate ventilation. Avoid build-up of vapors
and eliminate all sources of ignition, especially non-ex-
plosion proof electrical apparatus and heaters.
Avoid prolonged breathing of vapors or spray mist.
Avoid prolonged or repeated skin contact
A 1.1.3 Flammable Liquid (general)
Keep away from heat, sparks, and open flame.
Keep container closed.
Use only with adequate ventilation.
Avoid prolonged breathing of vapor or spray mist
Avoid prolonged or repeated contact with skin.
Al.1.4 Gasoline (White)
Harmful if absorbed through skin.
Keep away from heat, sparks, and open flame.
Keep container closed. Use with adequate ventilation.
Avoid build-up of vapors and eliminate all sources of
ignition especially non-explosion proof electrical apparatus
and heaters.
Avoid prolonged breathing of vapor or spray mist.
Avoid prolonged or repeated skin contact.
A 1.1.5 Toluene and Xylene
Warning—Flammable. Vapor harmful.
Keep away from heat, sparks, and open flame.
Keep container closed.
Use with adequate ventilation. Avoid breathing of vapor
or spray mist.
Avoid prolonged or repeated contact with skin.
ThtArntt
r tor Testing tnOMuttrmHttH* no pw
with tny mm menaoma in tht* ttunova. Ustn el this tttndtra are expntuy ettmea ttin
(mat riyta. «tt trm n*k a mtringtmtM of suet, rights, tn if***? the* own rtspom&Hy.
ot the vmMity ot tny tucli
Thl$ tunatnt is cutyacr to revision tt eny time oy the responsive technCM/ commmee end mutt oe reviewea every five yeert tnH
tna thouH M *Hnss*i to ASTH HMOqumn. Your comment wit receive cveM comamuon tt a meeting of 1 h» responsive
tectmctl committee, wmcti you muy Utena. n you leel tras your comments ntve not nctnea • tot nevmg you snouM mane your
view* Known to ln» ASTM Committee on SUrutras. 1916 flac* St.. PMua^pha, PA 19103.
Page 78
-------
Designation: E 145 - 68 (Reapproved 1987)
Standard Specification for
Gravity-Convection And Forced-Ventilation Ovens1
This standard is issued under the fixed designation E I45: the number immediately following ifie designation indicates the vear of
original adoption or. in the case ot" revision, the vear of last revision. A number in parentheses indicates the year of last rcapproval. A
superscript epsnon u> indicates an editorial change since the last revision or reapproval.
1. Scope
1.1 This specification covers the performance require-
ments for general-purpose air ovens ordinarily used in testing
operations, which have a testing chamber up to 0.6 m3 (25
ft3) in volume. It is applicable to gravity-convection ovens
designed to operate over all or pan of the temperature range
from 20"C above ambient temperature to 200*C and to
forced-ventilation ovens designed to operate over all or pan
of the temperature range from 20*C above ambient temper-
ature to 500'C.
NOTE I—Ovens are designed for maximum operating temperatures
of about 200*C. 300*C. and SOO'C. the thermal insulation and cost of the
oven being dependent on the maximum temperature required.
1.2 This specification does not include safety require-
ments that are essential for ovens used in the presence of
combustible vapors or £««•<
1.3 The values stated in inch-pound units are to be
regarded as the standard. The metric equivalents of inch-
pound units may be approximate.
2. Types
2.1 This specification covers the following four types of
air ovens:
2.1.1 Type I A—An oven ventilated by gravity convection
having a uniformity of temperature within ±2 % of the
differential between oven and ambient temperatures.
2.1.2 Type IB—An oven ventilated by gravity convection
having a uniformity of temperature within ±5 % of the
differential between oven and ambient temperatures.
2.1.3 Type IIA—An oven having forced ventilation and a
uniformity of temperature within ±1 % of the differential
between oven and ambient temperatures.
2.1.4 Type IIS—An oven having forced ventilation and a
uniformity of temperature within ±2.5 % of the differential
between oven and ambient temperatures.
3. Performance Requirements
3.1 The temperature within the testing chamber shall be
controllable by an automatic device, and shall be uniform
within the tolerances given in Table 1 for the particular type
of oven when tested in accordance with Section 4.
3.2 The "time constant" is an arbitrary measure of the
rate at which a standard specimen is heated following the
procedure prescribed in Section 5. The value of the time
constant shall not exceed the maximum value given in Table
This specification is under the jurisdiction of ASTM Comramee E-«l on
Laboratory Apparatus and is the direct responsibility of Subcommittee E 41 (T on
Metalware.
Current edition approved Sept. 13. 1968. Published November 1968 Ongmallv
published as E 145 - 59 T. Last previous edition E 145 - 65 T.
1 for the particular type of oven.
3.3 The rate of ventilation of the testing chamber shall
conform to the requirements specified in Table 1 for the
panicular type of oven when measured in accordance with
the procedure given in Section 6.
TEST METHODS
4. Temperature Uniformity
4.1 Place nine calibrated thermocouples (Note 2) made
from iron or copper-constantan wire, approximately 0.5 mm
in diameter (No. 24 gage) and having a junction size of not
more than 2 mm (0.08 in.), in the empty testing chamber
with shelves in place and vents open. Locate one thermo-
couple in each of the eight corners of the oven approximately
5 cm (2 in.) from each wail and place the ninth thermo-
couple within 2.5 cm (1 in.) of the geometric center of the
chamber. A minimum length of 30 cm (12 in.) of lead wire
for each thermocouple shall be inside the oven to minimize
the conduction of heat from the thermocouple.
NOTE 1—If calibrated thermocouples are not available, nine thermo-
couples made from the same spool of wire may be used provided they
give the same value for temperature when placed adjacent to one
another in the testing chamber at the temperature of test
4.2 Bring the oven to the specified temperature and allow
it to reach a steady state (Note 3). Record the temperatures
of the nine thermocouples for a period of at least 24 h. and
determine from the record the maximum deviation of each
point from the desired temperatures. The ambient room
temperature shall vary by not more than a total of IO*C. and
the line voltage for electrically heated ovens shall vary by not
more than a total of 5 % during the test.
NOTE 3—Some ovens may require as much as 24 h to reach a steady
state. If a steady state does not exist, there is a drift m the temperature
toward the steady-state condition.
5. Time Constant
5.1 Heat the oven to within 10*C of the maximum
operating temperature for which it is designed and allow it to
stabilize for at least I h. Prepare a standard specimen
consisting of a smooth brass cylinder 9.5 ± 0.1 mm (0.375 ±
0.005 in.) in diameter and 57 ± 1 mm (2.25 ± 0.05 in.) in
length, and solder one junction of a differential thermo-
couple to it.
5.2 Open the door of the oven for 1 min while the
standard specimen and differential thermocouple are being
suspended in the testing chamber. Suspend the specimen
vertically within 25 mm (1 in.) of the geometric center of the
chamber by means of an asbestos cord of fine wire (0.3 mm
maximum diameter. No. 30 gage). Place the free junction of
the differential thermocouple in the air space of the chamber
Page 79
-------
4jj|» E 145
TABLE 1 Performanc* RequtrwnMts tor OVMW
Chwunnc
Deviation trom mated temoariturt of test tnrougnout tesong cnamoer ounng
50*C or wss. max. *C
Mora man SO'C. max. percent at ttHannim
Time constant, max. s
Rate of vernation of testing cnamoer. aw oranges per nour
mm
max
TypelA
1
2
600
10
TypeB
2.5
5
720
10
Typed*
0.5
1
480
100
200
Type MB
175
2.5
660
100
200
at least 75 mm (3 in.) removed from the specimen. Then
close the door and either record or measure the temperature
differential even- 10 s. Determine the time in seconds
required for the temperature difference to decrease to one
tenth of the original or maximum value (for example, from
120*C to I2*C) and consider this to be the time constant of
the oven.
6. Rate of Ventilation (Note 4)
6.1 Seal the ventilation ports, door, and all apertures of
electrically heated ovens with adhesive tape or by other
means to prevent any air from passing through the oven
(Note 5). Connect a wan-hour meter, with the smallest
division reading in 0.01 Wh in the electrical supply line to
the oven.
NOTE 4—This method is only applicable to electrically heated ovens.
Methods are being developed by the committee for determining the rate
or ventilation of ovens that are not electrically heated and for deter-
mining the uniformity of air-flow within the testing chamber.
NOTE 5—In forced-ventilation ovens, the space around the motor
shaft where it enters the oven must be dosed, but the fan speed must not
be affected by the closure.
6.2 Heat the oven to a temperature of 80 ± 2*C above the
ambient room temperature, and while at this temperature
measure the consumption of electrical energy for a period of
at least'/: h. Stan and stop the test at corresponding points of
the "on-ofT heating cycle, that is. at the moment when the
heaters are switched on by the thermostat.
6.3 Then remove the seals, open the ventilation ports, and
measure the consumption of electrical energy in the same
manner. The ambient room temperature measured at a
point approximately 2 m (6 ft) from the oven, approximately
level with its base and at least 0.6 m (2 ft) from any solid
object, shall be the same within 0.2'C during the two tests.
6.4 Calculate the number of changes per hour of the air in
the test chamber from the following equation:
\-3590 (X- Y)/l'D*T
number of air changes per hour.
average power consumption during ventilation. W.
obtained by dividing the energy' consumption deter-
mined from the wan-hour meter readings by the
duration of the test in hours.
average power consumption with no ventilation, com-
puted in the same manner. W,
volume of the testing chamber, cm3,
density of the ambient room air during the test, g/cnr.
and
difference in temperature between the testing chamber
and the ambient room air. *C.
where:
.V =
X =
Y
V
D
Tfce American Senary lor Totting ana Marwiatt raftat no pennon naptdtng me vafcoWyofa/iy*
ngnoaaaerM
mm any «am menoonfO m Out aranoara. I/sen of rfus ttinova are e*pr«^ aovaetf mar eVremmeMn of r/w vaMftv of an? aucft
Tni* anntfartf «tuottct to reman at any long py me respomiBM Mcftnca/ commtot ma mat ou r*n***i *
ariMmarawn. four oommenn are
-------
Designation: E 180-85
Standard Practice for
Determining the Precision of ASTM Methods for Analysis and
Testing of Industrial Chemicals1
under the tixed destination E ISO: the number immedutelv following the designation indicates the vear of
in the case of revision, the vear of last revision. A number in parentheses indicates the year or last reapprovai. A
i indicates an editorial chance since the last revision or reapprovai.
1. Scope Conducting an Interlaboratory Study of a Test Method (577
1.1 This practice establishes uniform standards for ex-
pressing the precision and accuracy of test methods for
industrial chemicals. It includes an abridged procedure for
developing this information, based on the simplest elements
of statistical analysis. There is no intent to restrict qualified
groups in their use of other techniques.
1.2 This standard mav involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the safety problems associated with us use. It is
the responsibility of whoever uses this standard to consult and
establish appropriate safety and health practices and deter-
mine the applicability of regulatory limitations prior to use.
2. Referenced Documents
2.1 ASTM Standards:
D1013 Test Method for Total Nitrogen in Resins and
Plasticr
D1727 Test Method for Urea Content of Nitrogen ResinsJ
E 178 Practice for Dealing with Outlying Observations4
3. Significance and Use
3.1 All test methods require statements of precision and
bias. The information for these statements is generated by an
interlaboratory study. This practice provides a specific design
and analysis for the study, and specific formats for the
precision and bias statements. It is offered primarily for the
guidance of task groups having limited statistical experience.
3.2 It is recognized that the use of this simplified proce-
dure will sacrifice considerable information that could be
developed through other designs or methods of analyzing the
data. For example, this method does not afford any estimate
of error to be expected between analysts within a single
laboratory. Statements of precision are restricted to those
variables specifically mentioned. Task groups capable of
handling the more advanced procedures are referred to the
literature (1. 2.3)* and specifically to the .-l5r.V/ Manual for
J^fi ^H" 'V"!?" 'he 'unsdictlon °< *STM Commmee E- 1 3 on Industnal
" ***"»">"'«> of Subcomnmtee E1J.04 on Precision
Current edition approved Mav »|. 1983. Published Julv 1985
pubhshed as E 180-6. T. Last previous edmon E ISO - "8
- inmiat Bixik 01 ASTM Slanaaras. Vol 06 0"
' Ductmnmiia. See I9SJ Annum Book o, ASTM Stanaards Vol 06 (P
4 {nnuul Bixik ,,i ASTM Slanaanu Vol 14 0"
numbetl In parenines" "* " "" "" «
3.3 The various parts appear in the following order
Pan A—Glossary.
Part B—Preliminary Studies.
Pan C—Planning the Interlaboratory Study.
Pan D—Testing for Outlying Observations.
Pan E— Statistical Analysis of Collaborative Data.
Pan F— Format of Precision Statements.
Pan o—Bias (Systematic Error).
Pan //—Presentation of Data.
PART A—GLOSSARY
Al. Scope
A 1.1 The following statistical terms are defined in the
sense in which they will be used in presenting precision and
accuracy information. These definitions have been simplified
and are not necessarily universally acceptable.
A2. Definitions and Description of Terms
A2.1 error—in a statistical sense, any deviation of as
observed value from the true value. When expressed as a
fraction or percentage of the value measured, it is called a
relative error. All statements of precision or accuracy should
indicate clearly whether they are expressed in absolute ot
relative sense.
A2.2 random error—the chance variation encountered in
all experimental work despite the closest possible control of
variables. It is characterized by the random occurrence of
both positive and negative deviations from the mean value
for the method, the algebraic average of which will approach
zero in a long series of measurements.
A2.3 bias—a constant or systematic error as opposed to a
random error. It manifests itself as a persistent positive or
negative deviation of the method average from the accepted
reference value.
A2.4 precision—the degree of agreement of repeated mea-
surements of the same property. Precision statements in
ASTM methods for industrial chemicals will be derived from
the estimated standard deviation or coefficient of variation of
a series of measurements and will be expressed in terms of
the repeatability and reproducibility of the method (see A2.7,
A2.13. All4).'
A2.5 accuracy—the agreement between an experimen-
tally determined value and the accepted reference value. In
* Out ol pnnt. Available Irom Universitv Microfilms, inc.. 300 North Zeeb Rd.
Ann Arbor. MI 48106.
Page 81
-------
chemical work, this term is frequently used to express
freedom from bias, but in other fields it assumes a broader
meaning as a joint index of precision and bias (4, 5). To
avoid contusion, the term "bias" will be used in appraising
the systematic error of test methods for industrial chemicals.
A2.6 variance—a measure of the dispersion of a series of
results around their average. It is the sum of the squares of
the individual deviations from the average of the results.
divided by the number of results minus one.
A2.7 standard deviation—a measure of the dispersion of a
series of results around their average, expressed as the square
root of the quantity obtained by summing the squares of the
deviations from the average of the results and dividing by the
number of observations minus one. It is also the square root
of the variance and can be calculated as follows:
n- I
where:
5 = estimated standard deviation of the series of results.
X, = each individual value.
.? = average (arithmetic mean) of all values, and
n = number of values.
The following forms of this equation are more convenient
for computation, especially when using a calculator
or
where:
•? ^
S-V2 ^
(I.V)-
'" '
ntn- l)
estimated standard deviation.
sum of the squares of all of the individual values
square of the total of the individual values, and
number of values.
NOTE I —Care must be taken in using either ot these equations that a
sufficient number of decimal places is earned in the sum of the values
and in the sum ot' their squares so that senous rounding errors do not
occur. For best results, all rounding should be postponed until after a
value has been obtained for s.
In this practice, the standard deviation is obtained from an
analysis of variance of the results of an interlaboratory test
program (Part E\
A2.8 coefficient of variation — a measure of relative preci-
sion calculated as the standard deviation of a series of values
divided by their average. It is usually multiplied by 100 and
expressed as a percentage.
\2.9 mntfe— the absolute \alue of the algebraic difference
between the highest and the lowest values m a set of data.
A2.10 duplicate s— paired determinations performed by
one analyst at essentially the same time. This concept also
applies to other such multiple determinations.
A2.1 1 95 % confidence interval or confidence limits— that
interval or range of values around an observed value which
E180
will, in 95 % of the cases, include the expected value. The
expected value is defined as the average of an infinite series
of such determinations.
A2.12 95 % confidence level—this term is commonly used
in establishing the probability of precision statements and
means that there are 95 in 100 chances of being correct, and
5 in 100 chances of being wrong, in predicting that the
expected precision (or expected value) will fall within the
specified limits or range.
A2.13 repeatability—the precision of a method expressed
as the agreement attainable between independent determina-
tions performed by a single analyst using the same apparatus
and techniques. (This term is further defined and limited in
C2.1.6. E2.1.andE2.2.9.2).
A2.14 reproduability—the precision of a method ex-
pressed as the agreement attainable between determinations
performed in different laboratories.
PART B—PRELIMINARY STUDIES
Bl. Scope
B1.1 This part covers the preliminary work that should be
earned out in a few laboratories before undertaking a full
interlaboratory evaluation of a method.
B2. Discussion
B2.1 When a task group is ased to provide a specific test
procedure, there may be available one or more methods
from the literature or from laboratories already performing
such analyses. In such cases, these methods have usually
been the subject of considerable research and any additional
study of variables, at this stage, would be wasteful of
available task group time. It is recommended that such
methods be rewritten in ASTM format, with full descriptions
of the equipment and procedure, and be evaluated in a pilot
run by a few laboratories on selected materials. Three
laboratories and at least three such materials, using one or
two analysts performing duplicate determinations on each of
two days, by each method, constitutes a practical plan which
can be analyzed by the procedures described in Part E—
Statistical Analysis of Collaborative Data. Such a pilot study
will confirm the adequacy of the methods and supply
qualitative indications of relative precision and bias.
B2.2 When the method to be evaluated is new. or
represents an extensive modification of an available method.
it is recommended that a study on variables be carried out by
at least one laboratory to establish the parameters and
conditions to be used in the description of the method. This
should be followed by a three-laboratory pilot study before
undenaking a full interlaboratory evaluation.
B2.3 Detailed procedures tor executing such preliminary
studies are not described in th« practice but are available in
the general statistical literature/
" Talk group chairmen are referred specificallv to Youden W J. •Experimental
Design and ASTM Committees.- Matenais Rnrarch
-------
PART C— PLANNING THE f.NTERLABORATORY STUDY
Cl. Scope
tinn! f ^h" ^ COVCrS S°mC common«nse recommenda-
tions tor the planning ot interiaboratory studies.
C2. Variables
C2.1 The major variables to be considered are the fol-
lowing: methods, materials or levels, laboratories, apparatus
analysts. days and runs. These are discussed as follows: '
C2 1 1 Methods— -ftx preliminary studies of Pan B
should lead to agreement on a single method, which can then
be evaluated in a full interiaboratory study. If it is necessary
to evaluate two or more methods, the complete program
must be earned out on each such method. In either case, it
Sn£ia"?Xd thatnthC melhod vanabies have be"
explored and that a well-standardized, fullv-detailed proce-
dure has been prepared. Nothing short of this w.11 justify the
time and expense required for an extensive precision studv
C21.2 Materials or Levels-Hie number of samples
adequaie^- (In^«ng the number of
S—3* s'gniflcantly the Agrees of freedom
for predicting the reproducibility of the method
This can be achieved only by increasing the number of
laboratories.) Some imerlaboratory studies can be limited to
a single sample as m the case of preparing a specific standard
solution. Methods applicable to a single product of hS
puntv can usually be evaluated with one or two sampS
When different concentrations of a constituent or values of a
physical property are involved, the samples should represent
StSS^h10""- middle' and top leveis °f STS
S3? JSS £ Va2 over a *** range- the numbw °f
evels should be increased and spaced to cover the range If
technical grade products are used in a prerision studv die
2S H f mCth0d may ta und«'™»»ble unless the
accepted reference value and its limits of error are known
from other sources. For this reason, it is well to include™
ormoie samples ot known purity in the interlaboratort
C2.1.3 Laboratories-To obtain a reliable precision esti-
mate. .t is recommended that the interiaboratorv studv
include approximately ten qualified laboratories. When this
number of independent laboratories cannot be recruited
advantage can be taken of a liberalized definition ofco £
oratmg laboratories, quoted as follows from the 4STM
'
an mtegrated sequence of operations usmg
and
H a
uons mav be set up m the same area or -laboratory- Each such
participating mstallatton should be considered as a collaborate Jtora
tory so far as this procedure is concerned. S.m.lartv leu ^ «,
obtained w,th different pamapams or undeTSrenf
calibration would ,„ general constuute results from
E 180
rating laboratones even though they »ere obtained on the same sets a
equipment.
This concept makes it possible to increase the available
"laboratones" by using two analysts (but not more than two,
m as many laboratories as needed to bring the total to the
recommended minimum of ten. In such cases the two
analysts must evaluate the method independently in the
fullest sense of the word, interpreted as using different
samples, different reagents, different apparatus where pos-
sible, and performing the work on different calendar days.
(In the design in Section C8. laboratories using two analysts
are designated as A-1. A-2. B-l. B-2. etc.) The most desirable
laboratones and analysts are those having previous experi-
ence with the proposed method or with similar methods. It is
essential that enough experience be acquired to establish
confidence in the performance of a laboratory before starting
the interiaboratory test series. Such preliminary work must
be done with samples other than those to be used in the
formal interiaboratory test program.
C2.1.4 Apparatus—The effect of duplicate setups is not
often a cntical vanable in chemical analysis. In instrumental
methods, however, apparatus can become an important
factor because the various laboratories may be using different
makes or types of equipment for example, the various
colorimeters and spectrophotometers used in photometric
methods. In such cases, the effect of apparatus becomes
confounded with between-laboratory variability, and special
care must be used to avoid misinterpreting the results. Of
course, if enough laboratories have instruments of each type,
"apparatus" can be made a planned variable in the study.
C2.1.5 Analysts—The use of a single analyst in each
"laboratory" (as defined in C2.1.3) is adequate to provide the
information needed for calculating the repeatability and
reproducibilitv of the method as defined in this practice. It is
essential that all analysts complete the entire interiaboratory
test program. With regard to analyst qualifications, an
analyst who is proficient in the method should be selected.
C2.1.6 Am—As defined in A2.13. the repeatability of
the method shall be evaluated in terms of independent
determinations by the same analyst. To achieve this, all
scheduled determinations must be performed on each of two
days (see Sections C8 and E2).
NOTE 2—As used in this practice, the term "days" represents
replication of a set of determinations performed on any day other than
thai on which the lint set was run. It may become a systematic variable
to the extent that it is desirable that a given laboratory run the entire set
of samples on one day and repeat the enure set on another. Although
this ma> introduce a bias for that laboratory, there appears to be little
chance that such a bias would be common to all laboratones. Where
preitminarv studies suggest that instability may result in an over-all
systematic "davs" dTect. special planning will be required to take care of
this proDlem.
C2.1.7 Runs—The multiple determinations performed at
the same time or within a very short time interval, on each
day.
C3. Number of Determinations
C3.1 Each analyst is required to perform duplicate deter-
minations on each sample on each of two days. If one
Page 83
-------
determination of a paired set is accidentally ruined, another
pair must be run. An odd or unusual value does not
constitute a "ruined" determination. In such cases, an
additional set of duplicate determinations should be run and
all values reported, with an assignable cause if at all possible.
C4. Samples
C4.1 One person should be made responsible for accumu-
lating, subdividing, and distributing the materials to be used
in the test program. Extra samples should be held in reserve
to permit necessary replacement of any that may be lost or
damaged in transit. Proper techniques in packaging and
sampling should be followed, particularly with corrosive or
otherwise hazardous materials. It is recommended that: all
liquid samples be tested for closure leakage by laying the
bottles on their side for 24 h prior to packaging, sample
bottles be packed in boxes with strict attention to right side
up labels, sample bottles be enclosed in plastic bags with
plastic ties, packing of severely corrosive liquids be super-
vised by a technically trained person, and that stria attention
be paid to O.O.T. regulations. If a collaborating laboratory
should receive a sample which shows evidence of leakage, or
which is suspect for any other reason, the recipient should
not use it but should immediately request a replacement.
C4.2 The most important requirement is that the
subsampies be representative and homogenous. Single-phase
liquids usually present no problem unless thay are hygro-
scopic or unstable. Solid mixtures, in which the components
vary in particle size, should be ground, sieved, and recom-
bined to give a homogeneous product, and. then checked
(microscopically, or by any other available means) to con-
firm its homogeneity.
C4.3 In the case of stable, homogeneous materials, one
subsample can be distributed to each collaborating labora-
tory. If the material is hygroscopic, or otherwise unstable.
individual subsampies should be provided for each day's run
by each analyst.
C4.4 Instability of any type may impose other restrictions
on the execution of a planned program. It is the responsi-
bility of the task group chairman to include in the plans for
the interiaboratory study specific instructions on selecting.
preparing, storing, and handling of the standard samples.
180
C4.5 The samples distributed for the formal interiabora-
tory test program should not be used for practice runs.
Where "dry-runs" are performed to develop proficiency in
an inexperienced analyst or laboratory, this must be done on
samples other than these.
C5. Scheduling and Timing
C5.1 Interiaboratory studies fail occasionally because no
timetable had been established to cover the program, partic-
ularly in cases where the materials have changed in storage.
after opening the container, etc. The instructions to the
collaborators should cover such points as the time between
receipt of samples and their testing, time elapsing between
stan and finish of the program, the order of performing the
tests, etc.. with particular attention to randomizing as a
means of avoiding systematic errors.
NOTE 3—A discussion of randomizing is beyond the scope of this
practice. Refer to standard textbooks on statistics and specifically to the
indicated references (12. 13).
C6. Instructions and Preliminary Questionnaire
C6.1 Having decided on the variables and levels for each.
the task group chairman should distribute to all participants
a complete description of the planned collaborative study.
emphasizing any special conditions or precautions to be
observed. A detailed procedure and description of equip-
ment, prepared in ASTM format, must be included. A
questionnaire similar to the one in Table 1 will aid materially
in the successful execution of the interiaboratory study.
C7. Report Form
C7.1 A form for reporting the essential data should be
prepared and distributed (in duplicate) to all collaborators.
who should be instructed on the number of decimal places to
be used. It is recommended that interiaboratory studies be
reported to one decimal place beyond that called for in the
"Report" instructions of the method under study. Any
subsequent rounding-off should be done by the task group
chairman or the data analyst.
C8. Design for an Interiaboratory Test Program
C8.1 The plan given in Table 2 should cover most cases
where laboratories and levels (or materials) are the principle
ni» of Mtthoa (attacneoi:
TABLE 1 QuMtionnaire on Interiaboratory Study
laooratory w«n« to panopate « me cooper.,™ testng o. tn,s metnoo tor prec^nlall
2 As a particioant. we understand mat-
'o i
refluiremen« *<*°*° "> » metnoo must oe avaiaOe m our laooratory wt*n the
start1"9 oate' ^ " testmg speomens - "— ' Miei <* - "•"" must
idl Samms must oe nanowa in accordance witn instruction ana
iei A Quanted operator must oertorm tr.e tests.
* "" aOPraiSal * our CaMOlMles « •«*•"«•
•* «• «- <*
3. We can supply — auatfted operators.
4 Comments:
YES' ~ NO- • • -
aeons.
'
tor coooeraBve tesang of
YES- = NO' ' =
-Signature
-Company
Page 84
-------
E180
TABLE 2 Single Method. Singl. Analyst Ten Laboratories. N Levels or Materials
Lave or Matenai i
a
Leva or Matenai N IN - 3 or Greater)
itenal H
E
C
F
0
G
E
H
F
I
G
J
H
variables. It calls for each analyst to perform two determina-
tions in parallel on each of two days, at each level. Where
additional variables must be included, the proposed program
should be referred to a statistician, the Subcommittee on
Precision and Accuracy, or to Committee E-l I on Statistical
Methods for a specific recommendation.
PART D—TESTING FOR OUTLYING OBSERVATIONS
Dl. Scope
D1.1 This pan covers some elementary recommendations
for dealing with outlying observations and rejection of data.
Lacking a universally accepted practice for the rigid applica-
tion of available statistical tests, considerable technical and
common-sense judgment must be exercised in using them.
Accordingly, the following procedures are offered only as
guides for the data analyst and all decisions to exclude or to
include any suspect data shall be subject to the approval of
the task group concerned.
NOTE J—The test for outlying observations should be applied onlv
once to a set of interiaboratory test data. Although two or more values
can be rejected simultaneously, in no case should the remaining data
again be tested for outliers.
D2. Principle of Method
D2.1 The tests for outliers among the "multiple runs" and
"different da\s" data are based on control chart limits tor the
range, as described in the ASTM Manual on Quality Control
ui Materials*
D2.2 The test for outlying observations among laboratory
averages is that described in Practice E 178.
D2.3 The choice of probability levels for each of the three
tests is based on practical experience gained from a number
of interlaboratory studies involving chemical or physical
properties.
NOTE 5—In choosing probability levels, there are two alternatives:
(/) use of a high probability level, accepting the divergent data, inflating
variances, and perhaps failing to find significant differences. or(.') use of
a lower probability level, rejecting the divergent data, deflating vari-
ances, and perhaps finding significance where none exists, in the case of
multiple runs in an interlaboratory test program, the choice of the
99.93 level is based on the premise that only a high degree of
divergence should justify rejection of data from a laboratory for this
reason. The 99.0 "i level for days also reflects this premise, the 95 1
level for laboratories is frequently used and is chosen here because an
outlying laboratory average, even at this probability level, may have a
pronounced etTect on the claimed reproducibthtv of the method (see also
D7.:>.
D2.4 The procedures are illustrated by data developed in
an interlaboratory study on the determination of hydroxyl
number (see Table 3).
O3. Outliers Between Runs
D3.I Using the data of Table 3. tabulate the results of the
duplicate runs on each of two days, in each of the eleven
laboratories. Calculate the individual ranges and the average
range as shown in Table 4.
D3.2 Multiply the average range by the factor 3.488 to
obtain the critical range at a 99.9 ^ probability level 10.1 °J
significance level). For the four materials in question, these
values are:
1 Although st.il ava,lable. ih., manual has been supcrceded bv STP /.< D.
Matenai
Dodecanot
Ethvtene gJvcoJ
Nonvlphenol
Pemaervthniol
Average Range
1.63
18.69
1.5:
-2J1
Critical Ranee
< i
63.J
5.3
-74
Page 85
-------
E180
D
avg
2 a
Q
avg
gthytene giyca 1 a
b
avg
2 a
b
avg
Nonylpnend 1 a
b
avg
2 a
b
avg
PfintAOfyihntoi 1 3
b
avg
2 a
b
avg
292.0
2946
"2333
291.2
293.4
292.3
1767.0
1790.0
1778.5
1777.2
1787.0
1 782.1
248.8
250.0
"5157
247.2
248.3
"2473
1555.0
1541.9
1548.4
1550.8
1555.5
T3532
Laos
292.1
288.0
290.0"
287.2
287.2
"2873
1767.9
1801.5
1784.7
1706.4
1798.4
1752.4
243.8
244.7
244.2
245.2
247.7
-2457
1551.0
14491
1500.0
1468.6
1516.0
M92.3
290.3
291 1
290.7
291.6
2892
"2507
1798.0
1809.0
T5033
1783.0
1786.0
T78T5
261.8
263.4
"26TB
273.0
271 1
17215
1566.9
1561 7
T35O
1567.1
1558.3
297.1
296.9
297.0
296.6
301.4
"33815
1818.1
1830.7
1824.4
1817.4
1848.6
1833.0
250.1
252.1
"25TT
249.7
250.4
~23o!T
1469.5
14843
1476.3
1579.8
1566.3
309.0
311.0
"37oT5
305.0
303.0
"30T5
1783.0
1787.0
T7S53
1785.0
1785.0
T75sl5
248.0
251.0
249.5
245.0
246.0
"2153
1553.0
•550.0
TS3T3
1531.0
1628.0C
289.8
288.7
"2853
289.4
289.6
"1853
1716.1
1717.2
17l6~5
1725.7
1721.7
1 723.7
245.0
2447
"2473
245.2
246.4
"2753
1492.2
1492.7
1492.4
1487.2
1482.5
295.9
294.9
"2557
294.2
293.5
"2515
1782.0
1760.0
1771.0
1777.0
1761.0
T7B53
246.7
248.7
"2477
249.7
2472
-2437
1559.0
1550.0
T5573
1560.0
1560.0
296.2
296.7
"2557
292.3
2948
"2515
1782.7
1836.5
13555
1801.6
18176
T8B53
249.3
249.6
249.4
246.5
246.8
"2433
1611.2
1566.6
T35B3
1548.6
1555.6
294.8
295.8
295.3
296.3
294.0
"2553
1805.4
1789.3
1797.4
1769.3
1784.3
1776.8
246.9
247.5
"2473
247.7
245.8
246.8
1528.6
1533.5
1531.0
1540.3
1533.7
291.4
292.2
"25T3
297.6
293.4
"2553
1776.2
1782.8
T7753
1781.7
1783.7
T7B27
244.3
247.1
"2437
247.8
245.3
"27B3
1537.1
1530.6
T533H
1536.9
1533J
1535.1
291.2
289.9
"2553
289.5
290.6
"25oU
1778.3
1755.8
1767.0
1743.5
1759.4
1751.4
242.3
245.0
243.6
2432
242.8
"2T33
1579.6
1523.5
T35T3
1565.3
1529.6
1 547.4
"n° OUsne0 "" WM aadM
Rec
*** w «*w»n oral samples.
Indoong wncr. piae,, rt Figun88 ^ to ^ conj^^ s^nrtx^ n Specified Utnrt«ig Values. whrt
SUnauras. Vot 14.02,
c Temperature may nave ncreaaea dumg mnnon.
NOTE 6—The faaor 3.488 is the Dt value used to calculate the upper
control limit for the range and is derived by the equation:
Dt - I + «/,/
-------
E 180
TABLE 4 Outliers Between Rum
Lab-
ora- Day
tory
A 1
2
8 1
2
C 1
2
0 1
2
E 1
2
P 1
2
S 1
2
H 1
2
1 1
2
J 1
2
K 1
2
— — — — — — _
Total
Number of runs
Avg range
Run a
•^ HM_ _
292.0
291.2
292.1
287.2
290.3
291.6
297.1
298.6
309.0
305.0
289.8
289.4
295.9
294.2
296.2
292.3
294.8
296.3
291.4
297.6
291.2
289.5
>^^^M^M«M«
Runb
— ^••i—
294.6
293.4
288.0
287.2
291.1
289.2
296.9
301.4
311.0
303.0
288.7
289.6
294.9
293.5
296.7
294.8
295.8
294.0
292.2
293.4
289.9
290.6
M - 35.8
n-22
ff- 1 63
Ethytene Glyccx
Range
2.6
2.2
4 J
0.0
0.8
2.4
0.2
2.8
2.0
2.0
1 J
0.2
1.0
0.7
0.5
2.5
1.0
2.3
0.8
42
1.3
1 1
•^M^H^n^H
Run a
1767.0
17772
1767.9
1706.4
1798.0
1783.0
18181
18174
1783.0
1785.0
1716.1
1725.7
1782.0
17770
1782.7
1801.6
1805.4
1769.3
1776.2
1781.7
1778.3
17435
~«^^—^— ~^_
^
Runb
1790.0
17870
1801.5
1798.4
1809.0
1786.0
1830.7
1848.6
17870
1785.0
17172
1721.7
1760.0
1761.0
1836.5
1817.6
1789.3
'7843
'782.8
1 783.7
1755.8
•7594
IR m 41 1 2
n-22
9 - 18.69
Range
23.0
9.8
33.6
92.0
11 0
3.0
12.6
31.2
4.0
00
1.1
40
22.0
16.0
53.8
16.0
16.1
15.0
6.6
2.0
22.5
'59
Run a
248.8
2472
243.8
245.2
261.8
273.0
250.1
249.7
248.0
245.0
245.0
245.2
246.7
249.7
249.3
246.5
246.9
2477
244.3
2478
242.3
243.2
Nonylpnenoi
Runo
250.0
2483
244.7
247.7
263.4
271.1
252.1
250.4
251.0
246.0
244.7
246.4
248.7
247.2
249.6
246.8
247.5
245.8
247.1
245.3
245.0
242.8
Ifl-33.4
n»22
fi- 152
Range
1 2
1.1
09
2.5
1.6
19
2.0
0.7
3.0
10
0.3
1.2
2.0
2.5
0.3
0.3
0.6
1.9
2.8
2.5
2.7
04
Run a
1555.0
1550.8
1551.0
1468.6
1566.9
1567.1
1469.5
1579.8
1553.0
1531.0
1492.2
1487.2
1559.0
1560.0
1611.2
1548.6
1528.6
1540.3
1537.1
1536.9
1579.6
1565.3
Pentaerytnmoi
Runb
1541.9
1555.5
1449.1
1516.0
1561.7
1558.3
1484.3
1566.3
1550.0
1628.0
1492.7
1482.5
1550.0
1560.0
1566.6
1555.6
1533.5
1533.7
1530.6
1533.3
1523.5
1529.6
ZR- 488.6
/r-22
fl -22.21
Range
13.1
47
101.9
474
S.2
8.8
14.8
13 .5
3.0
97.0
0.5
4.7
9.0
0.0
446
7.0
49
6.6
6.5
3.6
56.1
35.7
D5.4 From Table 7 (Table 1 of Practice E 178). obtain the
critical value of Tat the f °5 significance (95 % probability)
level for /7 = II. Comparing the observed with the critical
values, the data show:
D6. Summary
D6.1 The data of Sections D3. D4. and D5 can tx
summarized as follows:
Material
Dodecand
Ethylene gjycol
Nonylphenol
Pemaerythmol
Suspect
Laboratory
Critical T Obsened Tn or T,
136 i49
136 (ilj. maul
2J6 :.88 C
136 11.86. maxi none
The indicated laboratories are suspect as rejectable at a
95.0 % confidence level.
D5.5 Practice £ 178 also indicates, in 4.3. that an alterna-
tive system based entirely on ratios of simple differences
among the observations is given in the literature (9.10). This
system may be used if it is felt highly desirable to avoid
calculation of 5.
Laboratono Suspect as Rejecuonabte
Material
Dodranoi
Elhvtene glycol
Nooylphenoi
Penuemhmol
Runs at 999 «
none
B
none
BandE
Days at
99.0%
Laboratory
Averages at
93.01
E
B
C
D
C
none
D7. Discussion
D7.1 When the above operations show any set of data
from a laboratory to be suspect, even- effort should be made
to find an assignable cause that will justify rejection.
D7.2 As Practice E 180 does not provide procedures for
TABLE 5 Putter. Between Day Average*
Labora-
tory
A
B
C
0
E
F
G
H
1
J
K
Total
Number of runs
Avg range
o«yi
293.3
290.0
290.7
2970
3100
289.2
295.4
296.4
295.3
291.8
290.6
Oooecanoi
Day 2
292.3
287.2
290.4
300.0
3040
289.5
293.8
293.6
295.2
295.5
290.0
Sfl- 22.2
n- 11
" • 2.02
. EftyleneGlveol Nnnvmnrnm
Range
i~5
2.8
0.3
30
60
0.3
16
2.8
0.1
3.7
0.6
—"——•••
Day 1
1778.5
1784.7
1803.5
18244
17850
1716.6
17710
1809.6
1797.4
1779.5
1767.0
™«™— ~^»™
Day 2
1782.1
1752.4
1784.5
1833.0
17850
! 723.7
1769.0
1809.6
1776.8
1782.7
1751.4
IH- 112.0
nm n
fl • 10.18
Range
3.6
32.3
19.0
8.6
0.0
7 1
2.0
0.0
20.6
3.2
15.6
Day 1
2494
244.2
262.6
251.1
2495
2448
2477
249.4
247.2
245.7
243.6
Oay2
2478
2464
2720
2500
245.5
2458
2484
246.6
246.8
246.6
243.0
ZR - 24.7
n- 11
fl • 2.25
Range
16
2.2
94
1 1
40
l 0
0.7
2.8
0.4
0.9
0.6
Oayl
1548.4
1500.0
1564.3
1476.9
1551.5
1492.4
1554.5
1588.9
1531.0
1533.8
1551.6
Day 2 Range
1553.2
1492.3
1582.7
1573.0
1579.5
1484.8
1560.0
1552.1
1537.0
1535.1
1547.4
ZR • 199.6
n- 11
K - 18.15
48
7.7
1.6
96.1
28.0
76
5.5
36.8
6.0
1.3
42
Page 87
-------
E180
UDorau
1 i •
A
S
c
D
E
F
G
H
1
J
K
IX'
IX2-
(ZX?-
(rxp/n -
*-
i:
r,-
r, -
yy ™
Actual
— •
292.8
288.6
290.6
298.5
307.0
289.4
294.6
295.0
2952
293.6
290.0
3235.6
952006.02
10469107.36
951737.03
V93ZOO6.0Z - 951737 03
11 - 1
5.19
294.1
307.0-294.1
TT3 * 2-49
3.19
294.1 - 288.6
— rrs '06
* To avow nancang large numoer* ana trtus smptty me e
values were used to calculate the standara deviaoon oireciiy.
fxamo/e:
Actual x - 17001*
1780.3 803
1768.6 68.6
1794.0 94.0
1828.7 128.7
1785.0 85.0
17202 202
1770.0 70.0
1809.6 1Q9.6
17«7.1 87.1
1781.1 81.1
17592 592
883.8
78767 20
781102.44
71009.31
11 - I
27.9
1780.3
1828.7 - 1780.3
27.9 "''J
1780.3 - 1720.2
27.9 •-'•'•'
aicuiations. me oata nave oeen 'ox
Tne mean. x. is ootamea oy me foil
Jf • SX/n * K
X"- 883.8/11 - 1700-
Nonyipneriu
Actual X - 200*
248.6 48.6
245.3 45.3
267.3 67.3
250.6 50.6
247.5 47 5
245.3 45.3
248.0 48.0
248.0 48.0
247.0 47.0
246.2 46.2
243.3 43.3
537.1
26638.37
288476.41
26225.13
V 26638.37 - 26225.13
11-1
6.43
248.8
267.3 - 248.8 , ..
643 -2M
248.8 - 243.3 .
6.43 C1
owing equaoon:
1780.3
Pentaenrmmoi
Actual X - 1400'
1550.8 150.8
1496.2 962
1563.5 163.5
1525.0 125.0
1565.5 165.5
1488.6 88.6
1557.2 1572
1570.5 170.5
1534.0 134.0
1534.4 134.4
1549.5 149.5
15352
22174424
2356839.04
214258.09
/221744 24 - 214258.09
V 11-1
27.4
1539.6
1570.5 - 1539.6
27.4 -M3
1539.6 - 1488.6 . ..
' 27.4 * ' M
the analysis of data in which values are missing, rejection in
any one of the three categories (runs. day. or laboratories)
makes it necessary to exclude from the analysis of variance
all of the data from that laboratory peninem'to the material
or sample in question.
„ N°V-?n'r the oudie« between runs need be eliminated from
the checking hmiu for duplicates calculations, as illustrated in
TABLE 7 Critical Value* for T When Standard Deviation to
Calculated from Present Sample
More—Based on avattbto literature (11). thaw prooaMty leva* nave Oeen
oouBted to take account ot the (act mat n actual pracoce the ontenen * appted to
«Her me sfnanatt or ma laroast oDsarvawn (or Doth) as trie case happens to oe.
AdiuMnient o« MM value* was atso made tar dmsnn oy n - 1 nsteaoofnn
caKUatng*.
D7.3 Although rejected data are usually excluded before
performing the analysis of variance, it is advisable to perform
the analysis using the entire set. as well as after the
elimination of the suspect data. With a desk calculator this
will entail relatively little additional work and the compara-
tive data are often helpful in appraising the results ofthe
entire program, as well as in deciding whether or not the
rejection is justified.
PART E— STATISTICAL ANALYSIS OF COLLABORATIVE
DATA
El. Scope
EI.l This pan demonstrates the statistical analvsis of
typical data obtained with the design of Section C8.
El. 2 The abridged analysis of variance eives the basic
information needed for calculating repeatability and reoro-
ducibility as defined in this practice. It determines the
between-laboratones and within-laboratones variances for
each level and combines them to give the two pertinent
standard deviations or coefficients of variation.
El. 3 Because it disregards interactions, this simplified
procedure sacrifices information that could be developed bv
using conventional methods for the analysis of variance
tanfetr of Obs*yv>oont. n
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
At S K Signficance
Level
1.15
1.48
1.71
1.89
2.02
2.13
221
229
2.36
2.41
2.46
2.51
2.55
2.59
2.62
2.65
2.68
2.71
2.73
2.76
2.78
2.80
2.82
All XSgreteanc*
Lave)
1.15
1.50
1.76
1.97
2.14
227
2J9
2.48
2.56
2.64
2.70
2.76
2.81
2.85
2.89
2.93
2.97
3.00
3.03
3.08
3.09
3.11
3.14
Task groups capable of handling such procedures are referred
to the literature (1, 2, 3) and specifically to the ASTM
Manual for Conducting an Interlaboratory Studv of a Test
Method (STP 335).*
Page 88
-------
E 180
JTABLE 10 An«ty«ia of Variance—€iamprt A
Source of Variance
en lafioratones
Between eiys. witnm laboratories
Total
Sum of Squares.
SS
Eq 2 - Eq 3
Eq 1 - Eq 2
Degrees of
Freedom.
OF
Mew Squirt
Expected Mean
Square
SS/DF
SS/OF
Eq 1 - Ea 3
m — numoer of columns f laboratories I.
• n - numoer m eacn column (daysi.
SS - sum of squares.
DF(w degrees of freedom.
s2,, - variance aue to differences oetween columns ilaBoratonesi.
and
s2. • variance due to differences wimm columns
(days).
Source of Variance
Between laboratories
Between days, witrun laboratories
Sum of Squares. SS
3483.8200 - 3307.5920 - 176 2280
3505.0600 - 3483.8200 - 21.2400
3505.0600 - 3307.5920 - 197.4680
rra-rinm f\C M»fl OQl
r'BVUUMj. i/r
10-1-9 1767280/9 -
10(2-11-10 217400/10-
ExpactedMean
"** Square
19.5809 s*. f 2s*.
2.1240 i2.
E2.2.6 Pooling of Data—The tabulated values should
exhibit one of the following three patterns: (/) the sa or the
sa+b values, or both, in good agreement for the four
materials. (.?) the coefficients of variation agreeing for the
four materials, or (3) neither showing the desired uniformity.
In Table 11. it is evident that the standard deviations differ
widely and. therefore, cannot be pooled. The coefficients of
variation for the between-days. within-laboratories data are
in excellent agreement and an over-all coefficient can be
calculated by pooling them as follows:
CVa 1 (overall)
v^
- oy *-,) •*-...
x cr,
/>/>... DF.
( 10 X 0.50^) + « 10 x Q.S3-) + «S x Q.63j) -t- < 10
10
The between-laboratones data show good agreement in the
coefficients of variation for dodecanol and nonvlphenol as
well as good agreement between those for oentaervthritol
and ethylene glycol. but there is a significant spread between
the two groups and most task groups would hesitate to
combine such data for the entire set. Therefore, the proper
action is to report separate coefficients of variation for the
two groups.
NOTE Jt-The following statistical tests are useful for determining
»nether or nut the standard deviations can be pooled:
Cochran Test: Eisenhard. C. Hastax. M. W. and Wallis. W A
-Techniques of Statistical Analysis.-McGraw-Hill Book Co Inc New"
York. NY. 1947 p. J88.
Hartley Test: Bowker. A. H. and Lieberman. G. 1. "Handbook of
Industrial Statistics." Prentice-HaiL Inc. Engiewoad Cliffs. NJ. 1955 p.
952.
The coefficient of variation for hydroxyl values in the 250 to
300 range is calculated as follows:
x l.!3-)-t-<9x0.9l-)
9 + 9
- 1.03 %
Similarly, the coefficient for values in the 1500 to 1800 range
is calculated as follows:
/(7x l.72-)f (9x |. 66-)
" V 7T9
-1.69%
NOTE 9— If the ;„ and 50.» values i rather than the coefficients)
should show good agreement, the mathematical procedure for pooling
them is analogous to that shown in E2.3J.
E2J.7 Checking Limits for Duplicates—* useful preci-
sion estimate can be obtained from the values for the
duplicate determinations in the form of the permissible range
for such paired determinations. The standard deviation for
duplicates can be calculated from the original data for paired
determinations as illustrated for dodecanol in Table 12.
s (from duplicates) - -U'-
'sum ol the squares of all differences
2 x number of sets
187.40
1 x ""
1.41. based on 11 degrees of freedom
Matenai
Nonyiprwnoi
Average OH Numoer
247.0
1543.6
1781.5
Between Oays. Witnm La
Degrees of
10
10
8
10
1.46
1.32
9.76
768
Doratones
Coefficient of
QJ5D
053
0.63
Snow Result. Any Laooratorv
Degrees of
9
9
7
e
«>•«•
379
275
2653
29.59
Coefficient of
1.13
0.91
1.72
Page 90
-------
Results of Duplicate Runs—EMTIIOI* A
292.0
291.2
292.1
287.2
290.3
291.6
297.1
298.6
309.0
305.0
289.8
289.4
295.9
2942
2962
292.3
294.8
296.3
291.4
297.6
291.2
289.5
294.6
293.4
288.0
287.2
291.1
2892
296.9
301.4
311.0
303.0
288.7
289.6
294.9
293.5
296.7
294.8
295.8
294.0
2922
293.4
289.9
290.6
22
41
0.0
0.8
2.4
02
2.6
2.0
2.0
1.1
02
1.0
0.7
0.5
2.5
1.0
2.3
0.8
4.2
13
1 1
0.64
5.76
0.04
7.84
400
400
1.21
0.04
1.00
0.49
0.25
6.25
1.00
529
0.64
17.64
169
1 21
The data for the other three materials are analyzed similarly
after eliminating outliers between runs '. These opera-
uons are not illustrated but the results are summarized in
2£i K •?" £case in EL2A ** fuu « « not *
pooled, but the coefficients of variation for dodecanol and
nonylphenol can be combined to give an over-ail value for
the 250 to 300 range, and the pentaerythritol and ethylene
glycol coefficients can be combined for the 1500 to 1800
range. Using the first pair as an example.
CV% - L<~
402
- 0.49 n
E2JLS Degrees of Freedom—Calculation of the exact
number of degrees of freedom applicable to the pooled
coefficient of variation .or to the pooled standard deviation)
is a complex procedure that is beyond the scope of this
practice. To permit making predictions concerning the
reproducibihty in a universe of laboratories based on a studv
TABLE 13 Standard Da*
cimts of Variation trc
A
Mannai
OH
Pecjeu 01
Standard
Of
EthytenB giyctx
24*44
iS9B.se
178167
22
22
20
21
1.41
124
15.53
14.00
0.48
0.50
1.01
E180
TABLE 14 Factors for Calculating Rang* of Two Results (95 %
Confidence Lave<>
Degrees ot
Freedom.
DF
1
2
3
4
5
6
7
8
9
10
11
12
Factor at
95 (Con-
fidence
Level
17.97
6.06
4.50
3.92
3.64
3.46
3.34
326
320
3.15
3.11
3.08
Degrees ot
CJ^^HWM
rrvBooni.
OF
13
14
15
16
17
18
19
20
21
22
23
24
Factor at
95 * Con-
Msnot
Level
3.05
3.03
3.01
3.00
2.98
2.97
2.96
2.95
2.94
2M
2.93
2.92
Degrees of
Freedom.
OF
25
26
27
28
29
30
40
50
60
120
*
Factor, at
95 X Con.
ftdence
Level
2.91
2.91
2.90
2.90
2.89
2.89
2.86
2.84
2J3
2.80
2.77
among m laboratories, a conservative estimate of (m — I)
degrees of freedom is used. For an estimate of the repeat-
ability of the method, the available degrees of freedom can be
approximated from the following equation:
DF = v materials or levels x m laboratories x in - I) days
In view of the fact that tests for outlying observations may
reject some data and result in different values of m for each.
level of material it is more correct to calculate the total'
degrees of freedom by adding the DF values for the pertinent
materials or levels. For the example cited, the between-dayv
witbin-iaboratories DF values of Table 11 are used. With
regard to checking limits for duplicates, the available DF can
be approximated as follows:
DF - * materials or levels x m laboratories
x A days x (r — I) multiples
For the reasons given in the preceding paragraph, it is also
more correct to total the DF values for the applicable levels,
as listed in Table 13.
E12.9 Calculation of Precision Estimates—The following
precision estimates should be calculated from the penmen
coefficients of variation of the preceding paragraphs, as
illustrated below.
E12.9.1 Checking Limits for Duplicates (95% Confi-
dence Level)—Multiply the coefficient of variation for dupli-
cate runs by the factor tor the applicable degrees of freedom
obtained from Table 14 (6,7.8). For the range of two results,
these factors can be calculated as follows:
Factor- vfxf00,
For the example cited in E2.2.7. where CV?c - 0.49 ^ and
DF - 22 + 22 « 44.0.49 x 2.85 - 1.4 %. relative, at the 250
to 300 level, the maximum range for duplicate values
acceptable at a 95 °J confidence level.
E2.2.9.2 Repeatability (Wi Confidence Level)—Simi-
larly, multiply the over-all coefficient of variation for the
between-days. within-laboratories data by the indicated
factor. In this case, where CVa % « 0.52 and DF « 10 + 10
+ 10 + 8 - 38.0.52 x 2.87 - 1.5 % relative, the maximum
range between two values (each the average of duplicates
obtained by the same analyst on different days) acceptable at
a 95 5 confidence level.
E12.9.3 Reproducibiiiiy—Thcx values are calculated as
Page 91
-------
E180
Summary of Data (or Three Levela—ExampM 0
**•"» i-ev« Comoonem x.
2*3
12.1
0.2
Between Days. With*, laooratones
Degrees of
Freedom. OF
10
10
s.
0.20
0.14
Coefficient of
065
165
70
Single ftestn. Any Laooratory
Degrees of
9
9
9
s^.
0.39
0.30
0.34
Coeffloent of
1.5
2.5
17
TABLE 16
Summary of PraenMon E«tiinatM~6«ampto I
Precision Estimates
Stanoara
Degrees
of Free-
dom. OF
95 * Range
Cneoung units (ouoicate)
Reqeataoutty
j x
Factor,
absolut
0.22
0.17
0.35
60
30
9
2.83
2.89
3.20
0.6
0.5
1 1
shown in E2.2.9.2 except that the over-ail coefficient of
variation for the between-iaboratohes data is multiplied by
the factor for the appropriate degrees of freedom. For tne
example cited at the 250 to 300 level, where the pooled
coefficient of variation = 1. 03 Vo relative, and DF « m - I *
9. the range at 95 % confidence level « 1.03 x 3.20 = 3.30 %
relative.
NOTE 10 — In the above examples, the coefficients of variation were
multiplied by the applicable factor because these had been pooled in
E2.2.6. if the standard deviations had proven poolable. the over-all sa
and sa+b values would have been used. These operations are illustrated
m E2J.
E2.3 Example B — The following example illustrates a
case where the standard deviations are in good agreement
and are pooled to give over-all standard deviations and
precision statements on an absolute basis.
E2.3.1 Specific Example — Three materials containing 24.
12. and 0 % levels of Component X were analyzed by one
analyst in each of ten laboratories, who performed duplicate
determinations and repeated the entire series one day later.
E2.3.2 Summary of Data- — To conserve space, the indi-
vidual results and the analysis of variance are not shown.
The results are summarized in Table 15.
E2.3.3 Pooling of Data—It is obvious that the standard
deviations show excellent agreement. Accordingly, the over-
all standard deviations are obtained by pooling as follows:
,„. overall)
*'*'
DF,
f(lO X 0.16-1 •* MU x U.20-) -MIU x 0.
10+10-*- 10
4-t
'0.17
.H(overall)
DF, - .. DFn
V(9 x 0.39-t - <* x 0 JO-) * (V x 0.34-
9 + 1*-*- V
•0.35
E2.3.4 Calculation of Precision Estimates—The precision
fstimates are calculated as shown in E2.2.9. except that the
standard deviations are used instead of the coefficients of
variation. These estimates and the pertinent data are shown
in Table 16.
PART F—FORMAT OF PRECISION STATEMENTS
Fl. Principle
F1.1 The formal statements of repeatability and reproduc-
ibility of methods for industrial chemicals should include the
estimated standard deviations or coefficients of variation, the
degrees of freedom, and the acceptable ranges at the 95 %
confidence level.
NOTE 11 —Although the recommended format uses the ranfe of two
results teach the average of duplicates) at a 95 % confidence level, other
combinations or confidence levels may be used if necessary (6.7.8).
F1.2 These estimates should be obtained by the proce-
dures outlined in Pan E or by equivalent statistical methods.
F2. Examples (Using the Data of Table 16)
F2.1 Precision—The following form is recommended for
such precision statements:
F2.1.1 Precision—The following criteria should be used
for judging the acceptability of results:
FL1.1.1 Repeatability (Single Analyst)—The standard
deviation of results (each the average of duplicates), obtained
by the same analyst on different days, has been estimated to
be 0.17 % absolute at 30 DF. Two such averages should be
considered suspect (95 % confidence level) if they differ by
more than 0.5 % absolute.
F2.1.1.2 Reproducibility (Muliilaboratoryt—'Tiic stan-
dard deviation of results (each the average of duplicates).
obtained by analysts in different laboratories, has been
estimated to be 0.35 % absolute at 9 DF. Two such averages
should be considered suspect (95 % confidence level) if they
differ by more than 1.1 % absolute.
F12 Checking Limits for Duplicates—Include this crite-
rion in the Report section of the method, using the following
typical wording:
Report—Report the percentage of Component X to the nearest
O.I ri. Duplicate runs which agree within 0.6 ^ absolute are acceptable
for averaging (95 "i confidence Jew)).
F3. Example (Using Data from Table 11 and Sections
E2.2.6. £2.2.9.2 and E2.2.9 J)
F3.1 Repeatability (Single Anaivst)—The coefficient of
variation of results (each the average of duplicate determina-
tions), obtained by the same analyst on different days, was
estimated to be 0.52 % relative at 38 DF. Two such averages
should be considered suspect (95 % confidence level) if they
differ by more than 1.5 % relative.
Page 92
-------
F3.2 Reproducibiiiiy (Muliilaboraiory)—The coefficient
of variation of results (each the average of duplicate determi-
nations), obtained by analysts in different laboratories, has
been estimated to be 1.03 % relative at 9 DF. Two such
averages should be considered suspect (95 % confidence
level) if they differ by more than 3.3 % relative.
PART C—BIAS (SYSTEMATIC ERROR)
Gl. Principle
G 1.1 In testing chemicals, the true or exact value is
seldom known and appraisals of systematic error often are
based on an expected value, such as a theoretical value
calculated for a purified or standard sample. In other cases.
the bias of a method is evaluated by comparing the deter-
mined average with the average obtained using a standard or
referee method. Again, the recoveries of known amounts of
the constituent in question from a prepared series of stand-
ards may be used for this purpose. The following are
suggested ways of expressing the expected bias of analytical
methods:
G2. Examples
G2.1 Example No. 1—Examples of expressing the ex-
pected bias referring to Test Method D 1013. are as follows:
The average value obtained in the analysis of a National Bureau of
Standards standard sample of* acetaniiide was 10.29 ± 0.04 *l'" versus a
theoretical nitrogen content of 10.36 %.
The average value obtained in the analysis of a purified melamine
sample was 66.28 ± 0.11 ** versus a theoretical nitrogen content of
66.67 *Z.
G2.2 Example No. 2—An example referring to Test
Method O 1727. is as follows:
The determined values for urea content averaged 0.2 ** absolute
higher than the expected values based on the total nitrogen content of
the urea resin solution, as determined by Test Method DIOI3. This was
true tor all three levels (0. 12. and 24 «6) used in the interlaboratory tea.
"' The limns 01' uncenamty at the avenges were calculated bv the procedure
given in UK ASTM Manual an Qtiatuv Com/at of MaienaU. STP If C. Pan i p.
41 (19511.
E180
G2.3 Example Mo. 3—An example referring to a hypo-
thetical case is as follows:
Recoveries oi' known amounts of Constituent X in a series of
prepared standards were as follows:
Amount Added, ppm Recovery, percent relative
10.0 91
<0.0 97
100.0 9S
The limit of detectability was found to be 2 ppm.
PART H—PRESENTATION OF DATA
HI. Experimental Data
Hl.l When a method is submitted to a letter ballot for
acceptance as an ASTM standard, the collaborative data
used in determining its precision and bias should be sent to
ASTM Headquarters. The precision and bias statement in
the standard should have a footnote that informs the reader
that the supporting data is on file in the Research Reports file
at ASTM and that copies are available by request to ASTM
(see Footnote 9).
H2. Statistical Data
H2.1 Details of the statistical analysis should not be
included in the draft, but should be referred to the Subcom-
mittee on Precision and Accuracy when the method is
submitted for editorial review. However, the draft of method
should contain a brief statement describing the interlabora-
tory study in sufficient detail so that the design will be
apparent to anyone statistically interested. This can be done
conveniently by adding a note to the section on Precision, as
in the following example:
NOTE 12—The above precision estimates are based on an interlabo-
ratory study on three samples, containing approximately 24. 12. and
0 % of Component X. One analyst in each often laboratories performed
duplicate determinations and repeated one day later, for a total of 120
determinations.1' Practice E 180 was used in developing these precision
estimates.
11 Supponiitf data are available from ASTM Headauanen. Request RRMOOS.
Rage 93
-------
E180
REFERENCES
(I) Finkner. Moms D.. "The Reliability of Collaborative Testing for
AOAC Methods." Journal. Assn. of Official Agricultural Chemists.
Vol40. 1957. pp. 882-892.
(2> ASTM Practice D1749. Imeriaboratorv Evaluation of Test
Methods Used with Paper and Paper Products. Annual Book of
ASTM Slumlord*. Vol. 15.09.
(3) Youden. W. J.. "Graphic Diagnosis of Interiaboratorv Test
Results." Industrial Qualnv Cururot. 1QCOA Vol XV. No. II.
May. 1959. pp. 24-28.
(4) Murphy. R. B.. "On the Meaning of Precision and Accuracy."
Materials Research A Slanaards. Am. Soc. Testing Mats.. MTRSA
Vol. I. No. 4. April. 1961. p. 264.
ASTM Recommended Practice E 177 Use of the Terms Precision
and Accuracy as Applied to Measurement of a Property of a
Material. 1964 Book of ASTM Standards. Pans 13. 14. 15. 35. and
41.
Pearson. E. S.. and Hartley. H. O.. "Tables of the Probability
(5)
(6)
Integral of the Studenttzed Ranee.' Biometnka. B1OKA Vol 33.
1943pp. 89-W
(7) Mav. Joyce M.. "Extended and Corrected Tables of the Upper
Percentage Points of the Studenuzed Range." Biometnka. BIOKA
Vol 39. 1952 pp. 192-3.
(8) Bennett. C. A., and Franklin. N. L. "Statistical Analysis in
Chemistry and the Chemical Industrv." Table 5.8. John Wiley
-------
Designation: E 203 - 75 (Reapproved 1986)*1
Standard Test Method for
Water Using Karl Rscher Reagent1
This standard u issued under the fixed destination E 203; the number immediately following the designation indicates the vear of
original adoption or. in the case 01" revision, the vear of Ian revision. A number in parentheses indicates the year 01" last reapprevai. A
superscript cpsUon «l indicates an editorial change since the last revision or napproval.
THu method has been approves liar use i>v agencies of the Depanmeiu of Defense and for iuimr in the DoD Index of Specifications and
Standards.
" Non—Editorial changes were nude throughout in January 1986.
I. Scope
1.1 This test method is intended as a general guide for the
application of the Karl Fischer reagent method for deter-
mining free water and water of hydration in most solid or
liquid organic and inorganic compounds. Samples that are
gaseous at room temperature are not covered (see Appendix
X4). By proper coice of sample size. Karl Fischer reagent
concentration, and apparatus, the method is suitable for the
measurement of water over a wise concentration range, that
is. pans per million to pure water. Both visual and electromet-
ric methods are described for ascertaining the end point
12 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the safety problems associated with its use. It is
the responsibility of whoever uses this standard to consult and
establish appropriate safety and health practices and deter-
mine the applicability of regulatory limitations prior to use.
Specific precautionary statements are given in Notes 1 and
15.
2. Referenced Documents
2.1 A list of existing ASTM Karl Fischer methods, their
applications to various products, and the sponsoring com-
mittees is given in Appendix X3.
22 ASTM Standards:
D1152 Specification for Methanol (Methyl Alcohol)1
D1193 Specification for Reagent Water3
DI744 Test Method for Water in Liquid Petroleum
Products by Karl Fischer Reagent4
E 200 Practice for Preparation. Standardization, and Stor-
age of Standard Solutions for Chemical Analysis1
3. Summary of Method
3.1 The sample, containing a maximum of 300 mg of
1 This test method is under the jurisdiction of ASTM Committee E-15 on
Industrial Chemicals and is the direct responsibility of Subcommittee El 3.24 on
Water.
Current edition approved Apnl 25. 19?!. Published June 1975. Onguullv
published as E 203 - 62 T. Lass previous edition E 203 - 641197H
: Annual Book of ASTM Standards. Vol 06.03.
1 Annual Book of ASTM Standards. Vol 11.01.
4 Annual Book of ASTM Standard!!. Vol 05.01.
5 Annual Book of ASTM Standards. Vol 15.05.
water, is dissolved or dispersed in a suitable liquid and
titrated with Karl Fischer reagent, which is a mixture of
iodine, sulfur dioxide, pvridine. and methanol or glycol
ether. As long as any water is present, the iodine is reduced to
colorless hydrogen iodide. The end point is the first appear-
ance of free iodine, determined either visually or electromet-
ricaJly. In some cases it may be desirable to add excess Karl
Fischer reagent and then back-titrate with methanol con-
taining a known concentration of water.
3.2 Fundamental equations are as follows:
C,H,N-I, + QH.N-SO, •*•
2O — 2C5H ,N-HI -l- C,H5N-SOj
ROH — C,H ,N-HSO4R
NOTE 1: Cute— Karl Fischer reagent contain! four toxic earn'
pounds, namely, iodine, sulfur dioxide, pyridine. and methanoiorglycol
ether. The reagent should be dispensed m a well-ventilated area. Care
must be exercised to avoid unnecessary inhalation of the reagent or
direct contact of the reagent with the skin. Following accidental spillage,
wash with large quantities of water.
4. Significance and Use
4.1 Titration techniques using Karl Fischer reagent are
one of the most widely used methods for the determination
of water.
4.2 Applications can be subdivided into two sections: (/)
organic and inorganic compounds in which water may be
determined directly, and (2) compounds in which water
cannot be determined directly, but in which interferences
may be eliminated by suitable chemical reactions or modifi-
cations of the procedure. Further discussion of interferences
is included in Appendix X2.
4.3 Water can be determined directly in the presence of
the following types of compounds:
Aeeuls
Adds (Note:)
Acvlhaudes
Alcohols
AMehvdes. stable (Note 3)
Amide
Amines, weak
-------
0» E 203
INORGANIC COMPOUNDS
Acids (Notes 61 Cupnc oxide
Acid oxides (Note >) Desicaais
Aluminum oxides Hydnzine sultaie
Anhvdnde* SUB Of a,,,,,* j^ iamvmic „.& (N , 7)
Banum dioxide
Calcium carbonate
NOTE -—Some acids, such as formic, acetic, and adipic and are
slowly estenfied. For high accuracy, use 30 to 50 5 pyndme in methanol
as the solvent.
, NOTE 3—Examples of stable aldehydes are formaidehvde. sugars.
chloral, etc. Formaldehyde polymers contain water as methylol groups
This combined water is not titrated. Addition of an excess of NaOCH,
in methanol permits release and titration of this combined water, after
approximate neutralization of excess base with acetic acid (see Note 10)
^NOTE 4—Weak amines are considered to be those with Kb value
NOTE 5—Examples of stable ketones are diisopropyl ketone cam-
phor, benzophenone. benzil dibenzolacetone. etc.
NOTE 6—Sulfunc acid up to a concentration of 92 % may be titrated
directly: for higher concentrations see Note 14.
NOTE 7—Compounds subject to oxidation-reducuon reactions in an
iodine - iodide system interfere.
5. Interferences
5.1 A number of substances and classes of compounds
interfere in the determination of water by this mrimetnc
method, complete descriptions of which are found in the
literature (I).6 This interference is associated with condensa-
tion or oxidation-reduction reactions.
5.2 Interferences of many classes of compounds can be
eliminated by chemical reactions to form inert compounds
prior to titration. The following are in this category:
Aldehydes and ketones. active (Note 8)
Amines, stront (Note 9)
Ammonia (Note 10)
Feme salts (Note 11)
Hydrant* derivatives (Note 10)
Hydroxyuumne sans (Note 12)
Metcaptans (Note 13)
Sodium methyiaie (Note 10)
Sulfunc acid (Note 14)
Thioacidj (Note 13)
Thiourca (Note 13)
NOTE 8—This interference may be reduced by use of pyndme rather
than methanol as solvent for the same or by the use of Karl Fischer
reagent and solvent prepared with ethylene glycol monomethyl ether in
place of methanol. The cyanhydrin reaction may be used to eliminate
the interference (I).
NOTE 9—Strong amines are considered to be those with A', value
>2.4 x 10°. Use salicylic add-methanol solution (Section 7) Glacial
acetic acid is applicable in certain cases.
NOTE 10—Addition of acetic aod eliminates the interference
NOTE II—Ferric fluoride does not interfere. Reaction with 8-
nydroxyquinoline is reported to eliminate this interference (7).
NOTE 12—Add 1 .V SO, in I-H pyridme-methanol or spent Karl
-ischer reagent.
NOTE 13—Olefm addition reaction eliminates imenerences (I) Oxi-
iation with neutral iodine solution eliminates the interference of
nercaptans 18).
NOTE 14—Sulfunc acid, above 92 rs. Add the sample (10 g) to a
arge excess of pyndine 135 mL). swirl to dissolve precipitate, and titrate
Addition of 8 mL of l-t-l pyndine - dioxane/1 g of sample also is
ausfactory. maintaining a homogeneous solution throughout the utra-
lon.
5.3 Many materials react stoichiometncaily with Karl
Fischer reagent. When their concentration is known, suitable
corrections can be applied. A list of such materials is given in
Appendix X2.
6. Apparatus
6.1 A suggested assembly of the apparatus, to provide a
closed system during titration. is shown in Appendix XI,
Fig. XI.1.
6.2 This equipment, without the end point detector, may
be used for visual titration.
7. Reagents
7.1 Reagent grade chemicals or equivalent as specified in
Practice E 200. shall be used in all tests.
7.2 Unless otherwise indicated, references to water shall
be understood to mean reagent water conforming to Type III
ofSpecificauon D 1193.
7.3 Karl Fischer Reagent—The Karl Fischer reagent may
either be prepared in the laboratory or purchased. Two types
of reagent are commonly used. Directions for preparing these
(Caution, see Note 15) and diluting if necessary, along with
commercial sources of supply, are as follows:
NOTE 15: Caution—Follow standard precautions for handling toxic
gases in preparing reagents (/) or (2} as described in the following
sections.
All operations should be earned out in a hood. Rubber gloves and a
face shield should be worn when handling pyndine and sulfur dioxide
and when mixing chemicals. Special precautions must be observed when
dispensing sulfur dioxide to prevent drawback of the solution into the
gas cylinder, which might cause an explosion. This is best accomplished
by placing a trap in the line between the gas cylinder and absorption
vessel.
7.3.1 Karl Fischer Reagent fEthvlene Glycol Monomethyl
Ether Solution. I mL - 6 mg H-,6) (2)—For each litre of
solution, dissolve 133 ± 1 g of iodine in 425 ± 5 mL of
pyridine in a dry glassstoppered bottle. Add 425 ± 5 mL of
ethylene glycol monomethyl ether. Cool to below 4*C in an
ice bath. Bubble 102 to 105 g of gaseous sulfur dioxide (SO,)
into the cooled mixture. Determine the amount of SO,
added by the change in weight of the SO, cylinder or the
increase in volume (about 70 mL) of the reagent mixture.
Alternatively, add about 70 mL) of freshly drawn liquid SO,
in small increments. Mix well and set aside for at least 12 h*
before using. (Caution: See Note 15.)
7.3.2 Karl Fischer Reagent (Methanol Solution. 1 mL = 6
mg H,O»—For each litre of solution, dissolve 133 ± 1 g of
iodine in 425 ± 5 mL of pyridine in a dry. glass-stoppered
bottle. Add 425 ± 5 mL of methanoi. Cool the mixture in an
ice bath to below 4*C. Bubble 102 to 105 g of gaseous sulfur
dioxide (SO,) into the cooled mixture. Determine the
amount of SO, added by the change in weight of the SO,
cylinder or the increase in volume (about 70 mL) of the
reagent mixture. Alternatively, add about 70 mL of freshly
drawn liquid SO, in small increments. Mix well and set aside
for at least 12 h before using. (Caution: See Note 15.)
7.3.3 Karl Fischer Reagent (Ethylene Glycol Monomethyl
Ether Solution, stabilized. 1 mL = 6 mg H;O).7
6 The boldface numbers in parentheses refer to the list of reterences appended at
ie end of this method.
7 Fisher Scientific Co. Catalog No. So-K-3 has been found saustactory.
Page 96
-------
7.3 5 tor/ ftafer *«7?eW. Z>/7u/e— Prepare more dilute
solutions of the Karl Fischer reagent bv dK£ £th S
proper solvent as follows: ' g e
Denied Siren**, mt
H.O/mL
LJUB of Diluent to AdoVUn of 6 mtymi.
KFReatem.
0.85
L6
These dilute solutions cannot be prepared bv simple propor-
.uon. since water added with the diluent must £ TccounTed
for. The volumes to add. indicated above, are calculated
assuming the diluent contains 0.05 % water.
7.4 Methanol. Standard (1 mL = I mg H,O)9-This
soluuon can be stored conveniently in a bottle with rubber
cap and pornons removed with a hypodermic svringe.
7.5 Sodium Tanrate Dihydrate—Gnnd certified material
(watercontent 15.61 to 15.71 %) to a fine powder" refcraS
H •""" ml
a stoppered bottle If doubt exists as to its water content.
dry a 2 to 3-g sample in an oven at 155 ± 5T to constant
weight (minimum 4h). (See Note 16.) constant
7.6 Solvents:
7.6.1 Acetic Acid,
• (See Note i->
to the single solvents
*
7.7.1 Methonol- Chloroform
°f
volume
7.7.3 Pyndine - <«« c/>ro/ ( , +4>_Mix , volume f
pyndine with 4 volumes of cthylene glycoL Use for com-
pounds containing carbonyl groups
7.7.4 Pyndine -Methanol (1 + 4>_Mix 1 volume of
pyndine with 4 volumes of methanol. Use for organic acids.
78 Sulfur Dioxide, anhydrous grade. (See Notes 1
8. Drying of Solvents
8J If it is necessary to prepare dry solvents in
laboratory, the following three methods can be used-
8.1.1 Distillation of Methanol from \tagnestum. to reduce
l° a°°5 r°- according to
the
8.1.2 A:eotropic Distillation using Benzene, to reduce the
moisture to 0.05 -.. Add I volume of benzene to 19 volumes
' Ktallnickiadt Cauiof No. 3651. his been found umfectorv
20
30 to SO
10.2 Add 25 or 50 mL of methanol to a clean, dry
titration flask containing a stirring bar. Close the neck of the
flask with a two-hole-rubber cap. Adjust the magnetic stirrer
to give a smooth stirring action. Turn on toggle switch S2 of
the end point detector (see Fig. X2.1). Set the hi-low toggle
switch SI to the desired sensitivity. With the electrodes out
of the soluuon or immersed in a wet soluuon. the meter
should read approximately 75 % of full scale. If the meter
reads below 50 °e. the batteries are weak and should be
replaced. Titrate with Karl Fischer reagent until a reading of
approximately 25 is retained for about 30 s.
10.3 Alternatively, titrate with Karl Fischer reagent to the
color end point. 9.1.
' Liade Type 4A Molecular Stew has been found aiufcctorv for Urn purpose.
" Aquameten. Models KF-2 and KF-3. Beckman Instruments. Inc.: Precision.
Dow Recordomauc Timor. Precision Scientific Co.: Metrohm Potemioanpn.
Bmkmann Instruments. Inc.: Tiuator. tnnsmonxed. Karl Fischer Matne-Mauc.
Annur H. Thomas Co.. have been found uusiaciorv tor Ura purpose.
Page 97
-------
miililitres of reagent required to titrate the sample.
B « miililitres of reagent required to utrate the solvent
blank.
F - water equivalent, in milligrams of water per miiiiiitre
of KF reagent.
W" « grams of sample, and
R = aliquot factor.
13. Precision and Bias
13.1 Sensitivity, precision, and accuracy depend on sev-
eral factors, for example, concentration of the Karl Fischer
reagent, titration technique, apparatus, quanuty of water
titrated, and nature of material being analyzed.
"Annual Book of JST.\fSianaants. Vol 15.04.
Paae 9R
-------
13.2 The sensitivity is about 0.1 mg of water for visual
titrations. Less than 0.02 mg can be measured by eiectro-
metnc titration.
13 J The following is an example of the precision attained
in an interiaboraiory (Note 19) study on two samples of
acetone containing 0.1 % and 0.4 <5 water and two samples
of methyl ethyl ketone containing 0.05 ?S and 0.17 % water.
13.3.1 Repeatability—Two results (each the average of
duplicate determinations) obtained by the same analyst
should be considered suspect if they differ by more than
0.013 %. absolute (95 % confidence level). Duplicate deter-
. initiations which agree within 0.008 % are acceptable for
averaging (95 % confidence level).
13.3.2 Reproducibility—Two results (each the average of
duplicate determinations) obtained by analysts in different
E203
laboratories should be considered suspect if they differ by
more than 0.028 %. absolute (95 % confidence level).
NOTE 19—The imeriaboratory study was earned out by ASTM
Committee D-l on Paint. Varnish. Lacquer, and Related Products.
Subcommittee V on Solvents. Plasticizen. and Chemical Intermediates.
Seven laboratories participated, with a single analyst performing dupli-
cate determinations on each of two days, using two methods on the lour
samples described above. ASTM Method O 1364 Test for Water in
Volatile Solvent (Fischer Reagent Titration Methodr5 was the subject of
the test program being compared with each laboratory's own version of
a Karl Fischer method. As neither the means nor the variances of the
two sets of data proved significantly different, all of the results were
pooled to give estimates of the repeatability based on 55 degrees of
freedom and reproduability based on 47 degrees of freedom.
13.4 The bias of this test method has not been deter-
mined.
APPENDIXES
(Nooinandatory Information)
XI. SUGGESTED APPARATUS FOR KARL FISCHER METHOD
X1.1 Scope
as^mhiv ^P3™05- W
assembly, and design of a dual-transistorized end point
detector There are available a number of assembli« of
similar design which are equally suitable for storing and
dispensing the Karl Fischer reagent, for containing »e
sample, and for detecting the end point. See Append* X3
and References (1,6). HI~UU« AJ
XI .2 Titration Assembly
r ,, ThC St°iage and d»Pensin8 assembly shall consist
of the following pare (see fig. X 1 . 1 ):
X 1.2.1.1 Bum. automatic, with TFE-flurocarbcm resin
plug and automatic zero, reservoir bottle, and connecting
tube.- Setect the size buret and bottle needed. An oveS
reservoir with micro buret may also be used.
Xl.2.1.2 Tube. Drying, calcium chloride, one bulb ->00
ram long.
Xl.2.1.3 Bottle. Aspirator. with outlet for tubing connec-
tions. 500-mL capacity.
XI .2. 1.4 Stirrer. Magnetic, with stirring bar coated with
TFE-fluorocarbon resin.
"te CU» Co. Caulo, No. .7.:4F. has been found sausfanory for this
' ta been
XI.2.1.5 Flask. Titration. 250-mL capacity.16
Xl.2.1.6 Electrodes, with connecting cord and plugs.17
X 1.2.1.7 Rubber Cap. 1.5 in. (38 mm) in outside diam-
eter.11 Punch two holes. 3 to 4 mm in diameter, through the
cap.
XIJ End Point Detector
XI J.I Figure X2.1 shows the wiring diagram of a sensi-
tive end point detector and a pans list This end point
detector is a dual-range constant-current type detector. The
electrode current is switched to select either 5 uA or 100 uA.
This makes possible Karl Fischer titrations using a weak
reagent for low pom water determinations or the usual
titrations using the stronger reagents.
XI .4 Assembly of Apparatus
Xl.4.1 Assemble the apparatus as shown in Fig. X 1.1. Fill
the drying tube and aspirator bottle with desiccant" Insert
the buret tip through one hole in the rubber cap. Use the
other hole for inserting a pipet or hypodermic syringe
containing liquid samples. Under humid conditions, keep
the second hole plugged except when introducing a sample.
or pass a slow stream of dry nitrogen into the flask. Connect
the plugs on ends of the leads on the titration flask to the
jacks of the end point detector.
'* V H. Thomas Co. Catalog No. 968MC-M umnon cdl has been found
lausfecunv for tats purpose.
ir V H. Thomas Co. Caulof No 9682-K70 electrode has been found
atn&norv for thu purpow.
c «',^V°I Rubter Ca C*ttto« No" :?0 w E- H S"*"' Co. Catalog No.
S-73U5. l.i m- have been found satisfactory for this pupae.
Indicauag-iype Dnente has been found lausbetory for Uus purpose.
Rage 99
-------
E203
TABLE X1.1
Bone «CKJ. H,8O,
HBO,
"fldne oxnt. 8A
Cupnctaw
MfOH),
CW. MgO. ZnO. A&O. HgO. Cu^). MnO^. P6O-.
1
1
0.5
1
2
3
2
3
2
3
1
1
OJ
0.5
t
2
1
3
1
(U
1
1
7
0.5
CataunareanMi
'u*** fluonav
wim KF iwgwi.
Net to Seal*
Bulb
Raogonr
trani«roriz«d
End Point
PKL X1.1
™»*>n Apparatus AsMmMy
Page 100
-------
E203
X2. INTERFERING COMPOUNDS THAT REACT STOICHIOMETRJCALLV WITH KF REAGENT THEREBY
ENABLING FREE WATER TO BE CALCULATED AFTER APPLYING CORRECTION
X2.1 Many interfering subsunces react stoichiometricaily
with constituents of the KF reagent. Consequently, when
independent analyses can be made for these compounds.
suitable corrections can be applied to the apparent water
results. Also in many cases moisture can be separated from
the interfering substance by extraction with a water-miscible
liquid, in which the sample is insoluble or by distillation.
preferably using a carrier that forms a homogeneous azeo-
trope. for example, dioxane. ethanol • benzene. Materials in
this class are given in Table Xl.l.
X2.2 Some compounds react only partially with KF
reagent when titrated under normal conditions. These in-
clude the following:
Methyloiurea-"
Peroxides, diacvl-'
Periods-'
Quinone
Anentous oxide
Chromaies
Dicnromatcs
Ironoudc
Nickel oxide
Sodium peroxide
Sodium sulfide
20 Interference of methylolurea can be eliminated by mranon at -WC (I).
:i Diacvl peroxides and pencids fairly rapidly oxidize the HI of spent KF
reagent. After a short time interval following addition of KF reagent this reaction
mav be quantitative III.
TO
ELECTRODES D,£'N459A
OguF^ i^ILQ1/
CI' ' T-
-9V.
PARTS LIST
METER SIMPSON MODEL 324
(4<* in.: 0-200 ItiAl
Amplifier 74IC
Diode IN459A
DiodeMZ:36l
Transmor^N 3641
Capacitor CI. 0.05 uF
Capacitor C2. 0.05 uF
ResotorRl. IIKO. "4 * carbon
Resmor Rl IOOKQ. ". W carbon
Resistor R3. 3.3 K.O.
-------
E203
X3. OTHER ASTM KARL FISCHER REAGENT WATER METHODS
Dvngnnon
0 789
0 890
01123
01348
01364
0 1457
01S33
. 01S68
• •01631
01744
02072
02849
03401
03621
E 700
E1084
Sponconng
COTvnttM
0-20
0-17
0-15
0-23
0-1
0-20
0-9
0-12
0-16
0-2
0-1
O-20
0-26
0-1
E-15
E-15
Tilto Of Method
: Test Metnoa tor Water n Uqud Naval Suns*
Test Method tar Water n Engne Coolant Concentrate by me Kart Footer Reagent Metfioa*
Ten Method tar Moisture n CeUaee"
Sp-Htictton tor PTFE Mattng and Extrusion Maim.*22
Test Mcmod tar W«v n irauwng LJQUK» (Kart Rsetwr Mtmod)34
Test Method tor Wrar »i Uqud PMoMun Products by Kart Fisefter RMgtnf*
Test MMhoo tar Wanr n F«tty t*tragen Camoounos2
MeBiom of Tcnng urnnm fom Peiyat R»w Mnmto21
PiAL4Hj9 tar tM OfliMiiradon of Wcttr n AofttM Esm2
Test M«noa tar Wiur n OMM Usng Kart Fochar fletgwit*3
Test Method tar Witer m Orgmc uouOs by Couomewc Kart Fnoner ritntwrr"
X4. DETERMINATION OF WATER IN GASES
X4.1 Procedures for determining moisture in gases are
described in the literature (1,5, 6, 10, II).
X4.2 As mentioned in Section 1, this test method does
not include procedures for samples that are gaseous at room
" Annual Book of ASTM Sunder*. Vol 01.01.
" Annual Book of ASTM Siantiards. Vol 13.03.
"Annual Book of ASTM Standards. Vol 10.03.
"Annual Book of ASTM Standard!. Vol 08.02.
temperature. The safe handling and analysis of gases require
a thorough knowledge of their properties and also the use of
special apparatus and techniques. The moisture content may
range from 1000 down to 2 to 3 ppm.
X4.3 The manufacturers of gases have developed very
precise Karl Fischer procedures for measuring moisture
down to a few parts per million (10,11). They should be
consulted when need arises. Also, there are available com-
mercial instruments that operate on the dew point infrared,
conductance, electrolysis principle, etc.. which are rapid and
accurate for determining moisture in gas samples (1,6).
Page 102
-------
E203
REFERENCES
(1)
(2)
Mitchell. J.. Jr.. and
Publishers. Inc. 1948.
Smith. D. M.. Aqtiametry. Intenoeaee
Peters. E D.. and Jungnickel. J. L. "Improvements in Karl Fischer
Method for Determination of Water.* Anaiviical Chemistry.
ANCHA. Vol 27, 1955 p. 450.
(3) Foulk. C. W.. and Bawden. A. T.. -A New Type of End Point in
Etectrotnetric Titration and its Appbcauon to lodimetry." Journal
of the American Chemical Society. JACSA. Vol 48. 1926. p. 2045.
(4) Basun. E L. Siegel, H- and Bullock. A. B.. •Microdetenninauon
• of Water by Titration with Fischer Reagent." Analytical Chemistry.
ANCHA. Vol 31. 1959. p. 467.
(5) Jones. A. G.. "A Review of Some Developments in the Use of the
Karl Fischer Regent," Analyst. Vol 76. 1951, p. 5.
(6) Mitchell. J.. Jr.. Treatise on Analytical Chemistry," Pan II. Vol 1.
1961. p. 69. Intencwnce Publishers. Inc.
(7) Laurene. A. H.. "Determination of Water by Karl Fischer Titration
in the Presence of Ferric Salts." Analytical Chemistry. ANCHA.
Vol 24. 1952, p. 1496.
(8) BrickelL W. F.. "Determination of Water Vapor in Natural Gas by
Direct Chemical Method." Petroleum Engineer. PENGA. Vol 24.
1952. p. 58.
(9) Freedman. R. W.. -Transistorized Dead-Stop End Point Detector."
Analytical Chemistry. ANCHA. Vol 31. 1959 p. 1287: see also
correction, p. 1686.
(10) Morton. J. D.. and Fuchs. L. K~. "Determination of Moisture in
Fluorocarbons." presented at a meeting of the American Society of
Heating, Refrigeration, and Air-Condiuoaing Engineers. June
13-15. 1960.
(11) E. I. du Pont de Nemours it Co.. Freon Technical Bulletin B-23.
"Moisture Determination in 'Freon' Fluorocarbons by Karl Fischer
Titration." June 1961.
(12) Card. L. N.. and Butler. R, C. "Determination of Moisture in
Sodium Bicarbonate—Karl Fischer Method." Analytical Chem-
istry. ANCHA. Vol 26. 1954. p. 1367.
(13) Beasiey. T. H.. Ziegler. H. W.. Charles. R. U and King, P..
•Critical Evaluation of the Karl Fischer Water Method." Analytical
Chemistry. ANCHA. Vol 44, 1972. p. 1833.
not
mtmonea n tra mnova. uttn oi ma tanatra v»
otunt nana. ma tfm m* of ttintytnm* or tucn nom. v» **my tntr own ntoonutntty
omtnniimKin at ir» vmaiiy et »ny uai
tnautouuot
to ASTU HtuoautrHrt. Your commtixt wm nc*v» etrtM t
tnuen m • mttang el th» /•coonMW
none* comnmit. winch you am? tnwna. H you tout to* your eommtm MV» nor netma » tor /WOTIO you ttnuU mttn your
v**»*no»mtotMASniComm«MtonSunatfas. 1916 Ruet St.. Phumpn*. PA 19103.
Page 103
-------
Page 104
-------
ASTM PROCEDURES REFERENCED IN SECONDARY ASTM
REFERENCES AND OTHERS
Page 105
-------
Designation: £ 300 - 66
Standard Practice for
Sampling Industrial Chemicals1
, E 300: the number «"»*«•»
-^
deamauon indicates the vear of
1. Scope
1.1 This practice covers procedures for sampling several
classes of industrial chemicals. It also includes recommenda-
tions for determining the number and location of such
samples, to ensure their being representative of the lot in
accordance with accepted probability sampling principles.
1.2 Although this practice describes specific procedures
for sampling various liquids, solids, and slurries, in bulk or in
packages, these recommendations only outline the principles
to be observed. They should not take precedence over
specific sampling instructions contained in other ASTM
product or method standards.
1.3 These procedures are covered as follows:
Statistical Consideration
Simple Liquids
Solids
Slum*
Sections
5 to 9
10 to 23
35tt>40
NOTE l--lt is intended to idd sections on viscous liquids: multiphase
liquids, parity solidified solids and liquefiable solids? iSSS£S
slurries: and gases, when available. —-* umwucu gases:
1.4 This standard may involve hazardous materials, oner-
"£"*• "If ?Wpment. This standard does not purport to
address all of the safety problems associated with its use It i
the responsibility of whoever uses this standard to consult and
establish appropriate safety and heaiih practices and deter-
mine the applicability of regulatory limitations prior to use
,F?£eEeamy?ttry mxaaam *« given in Sections 4.18*
17. 29. 33 and 36.
2. Referenced Documents
2.1 ASTM Standards:
D270 Method of Sampling Petroleum and Petroleum
Products- uwura
D 2234 Methods for Collection of a Gross Sample of Coal3
E 180 Practice for Determining the Precision of ASTM
Methods for Analysis and Testing of Industrial Chem-
icals
3. Significance and Use
3.1 This practice outlines the principles to be observed
when sampling several classes of industrial chemicals. This
1 This practice is under the joint luradKnon of ASTM Committee £.| j on
Industml Chemicals and u the direct responsibUity of Subeomimae7£|Ioj on
Cunwit edition appimcd Jan. 31. 1986. Published Maid, 19(6. OminaUy
published as £300-66. Last previous edition £300-73 (1913) •»••«»
: Discontinued, see 1933 Annual Book of ASTM Staaaants. Vol 05.01
1 mm*/ Book ot ASTM Stand**. Vol 05.05. ^^TOl Vo*mal-
4 \nnuai Book 01 ASTM Siangan*. Vol 15.05.
practice also covers the statistical considerations in the
sampling of industrial chemicals whether they are liquids.
solids, slurries, and in bulk or packages.
4. Safety Precautions
4.1 This practice covers procedures and sampling equip-
ment used to sample industrial chemicals that may be
potentially hazardous to personnel. Accordingly, it is empha-
sized that all applicable safety rules, regulations, and proce-
dures must be followed in handling and processing the
chemicals.
4.2 The characteristics of the material to be sampled will
govern the type of protective equipment required. Since
sampling may present such hazards as splashing or spilling,
protective clothing should be worn when the chemical is
capable of producing eye or skin irritation or bums. During
such potential exposures, chemical-type goggles or face shield
and protective gloves, or combination thereof, should be
worn.
4.3 Respiratory protection, where required, must be in
good condition and must be suitable to protect against
chemicals being handled.
4.4 When sampling chemicals that may be dangerous to
life by skin absorption, oral ingestion. or by breathing the
vapor, unusual precautions will be indicated. In such cases.
full-body protection such as supplied by a gas-tight or
one-piece air-supplied suit should be worn. A second man
should be continuously present to summon help and render
aid in the event of an emergency.
STATISTICAL CONSIDERATIONS*
5. Objectives
5.1 The sampling and testing of industrial chemicals may
have one or more of the following objectives:
5.1.1 The objective may be to estimate the average quality
characteristic of a given lot of material and to establish
confidence limits for this average. This would be the main
objective, for example, if a dollar value is to be placed on the
material for customs purposes or for sale.
5.12 The objective may be to decide whether the average
value for the lot meets a specification. This calls for an
acceptance sampling plan with the criterion being related to
the estimated mean of the lot.
5.1.3 The objective may be to estimate or make decisions
about the variability of a quality characteristic within the lot.
5.1.4 The objective may be to obtain simultaneous esti-
mates of the mean and variance or to make decisions about
1 Prepared on an Ad Hoc Committee ol ASTM Committee £-11 on Statistical
Methods.
Page 107
-------
E300
some loint combmauon of these estimates.
5.1.5 If the matenai comes in containers or can be viewed
as coming m clearly demanced units, the objective raaVoe
that pt estimating the number of such units outside of
speciiicauons. that is. the "fraction defective."
NOTE :-Proceaures are given beiow for estimating avenge
and for applying acceptance sampling inspection "- J
6. General Sampling Considerations
6 1 To obtain samples that are representative m a statis-
tical sense, one must consider such factors as physical form
- unuormity. type and number of containers, etc. All of these
influence the choice of method for performing the mechan-
ical sampling operation, as well as the number and location
of the required samples. Two commonly used practices for
selecting the sequence or location of the'individual samples
are descnbed.
6.2 Random Sampling is achieved when everv pan of the
A", ar\!?Ual CfaanCC Of bda» ***« imo "« sample
6.2.1 Designate all units in the lot. choosm* numbers m
sequence or other serial code so that sampling bv ranaom
numoers can be employed.
6.12 Preferably, this sequence should be m direct relation
to order 01 manutacture of packaging as an aid to observing,
from the sampie results, any evidence of stratificauon
6.13 Random selection of the numbers should be accom-
plished by chance or preferably by the use of a table of
random numbers.
6J Stratified Sampling can be emploved to estimate
average quality when it is known or suspected that the value
of a property of the material varies in nonrandom fashion
Uirougbout the lot for the following typical reasons: (a) the
lot may contain several production batches, (b) the lot mav
contain units produced by different procedures, equipment
shifts, etc- or (c) the lot may be nonuniform taSTrf
subsequent size segregation, moisture pickup, surface oxida-
tion, etc. If the assumed pattern is correct, the variance of the
population mean estimate will be less than that based on
random sampling. If the assumptions are incorrect, the
estimate of the mean may be biased. A stratified sampie can
be obtained as follows:
6.3.1 Based on the known or suspected pattern, divide the
lot into a number of real or imaginarv strata.
6.32 If these sections are not equal in size, the number of
samples to be taken from each stratum should be propor-
tional to the size of the various strata.
6.3.3 Further subdivide the major strata into real or
imaginary subsections and select the required number of
samples by chance or preferably by means of a table of
random numbers.
> 7. Estimate of Average Quality
7.1 Determination of the Variance of a Sample Mean—U
the matenai comes in. or can be viewed as comine in
-realizable pnmary units, each of which are to be divided into
realizable secondary units, then if nh pnmary unils
selected at random from a lot of N pnmarv units, and if n
secondary units are selected from each primary unit with k
tests being made on each secondary unit drawn, then the
variance of the mean of the results is given as follows (Notes
3 and 4):
Page 108
-------
variability as well as test variability), using Eq 5:
V-XUT-.r, )*/(«i-I) (5)
where .T, is the mean of the individual test results on the n,
primary units, with one secondary unit per primary unit and
one test per secondary unit.
7.3.2 Decide to estimate the mean of the lot from single
tests on single secondary units from n: primary units where
AN > n, and the n; units include the n, preliminary units, the
value on n: being determined from Eq 6:
«j - *,z/Ts* (6)
where r5f2 is the target value of an estimate of the variance
of X. The target value Ts^ will depend on the width of the
desired confidence interval. If it is hoped to have a 0.9S
confidence interval of width 2A. then for n, > 30. TStl
should be taken as (A/1.96)2. For smaller values of n-,, the
1.96 should be replaced by the 0.022 values from a /-table.
7.3.3 Estimate the variance of the mean after n: tests from
Eq7:
V - XX - T)J/n,
7.4 A Confidence Limits for the Mean of the Lai:
7.4.1 If the basic variances are known and two-stage
sampling (primary and secondary units) is employed, then
0.9S confidence limits for the mean of the lot u are given oy
Eq8:
0.95 confidence iimiu for u » .? ± 1.96 *t (8)
where V* is obtained from the o>: value given by Eq 1.
7.4.2 If the basic variances are unknown an|1 the variance
ofX is estimated as in 6.3 («, sample primary units with one
secondary unit per sample primary unit and one test per
secondary unit), then 0.9S confidence limits for the mean of
the lot u are given by Eq 9:
0.95 confidence limits for u - .F ± r,,^ st (9)
E300
rating both a buyer's and seller's nsk. The following proce-
dures are based on this concept.
8.2 Single Lower Specification Limit (LV Simple Random
Sampling from a Large Lot:
8.2.1 Procedure:
8.2.1.1 Step i—Note the value of the lower specification
limit for average lot quality and designate it by L, Assume
this value to represent a quality level for which the proba-
bility of acceptance should be high and the risk of rejection
low. In this procedure, the seller's risk is taken to be O.OS.
8.2.1.2 Step 2—Establish a lower value for the barely
tolerable lot quality for which the level of acceptance should
be low and designate it by L - A. Here, this buyers risk is
taken to be 0.10.
8.2.1.3 Step i—Take a preliminary sample of n, (equals
10 or more) units at random from the lot and compute
where Jt is obtained from the stz value given by Eq 7 and
/0 023 can be taken as equal to 1 .96 if n3 is greater than 30. but
otherwise should be taken from a table of r-values for n; - i
degrees of freedom.
8. Acceptance Sampling for a Lot Mean— Bask Variances
Unknown
MOTE 6"~ Tnis
the acceptance
descnbes a sii&pie randon mnpling plan for
of an isolated lot and provides for buyer's ud
seller's risks of maJdn« a wrong rirrniop IT a tenet of lots is to be
inspected and knowtedte of the bare vanaaces is available. «fM~™
savings nay be •"•'**"** by iftinij composites.
8.1 Introduction—If a specification requires, for example
that the average purity or assay of a lot be no less than
98.0 %, it it sometimes assumed that the sampling and
testing plan will accept all lots of 98.0 % or higher, but will
detect or reject any lot tailing below this value. This ideal
situation is not statistically realistic, as the required degree of
discrimination can be approached only if the lot units are
essentially uniform and the test procedure is capable of
attaining a very high level of precision. It is necessarv,
therefore, that the contracting parties realize that any sam-
pling plan based on a low probability of rejecting a lot which.
in fact, is 98.0 % or higher in purity, may also permit
acceptance of some lots below this specification minimum.
Accordingly, such specifications must be viewed as incorpo-
,.*•,/«,. and
- n
Seta,
(10)
(ID
(12)
8.2.1.4 Step 4—Note the value of A agreed to in Step 2.
Compute X, - d/J, and find from Table i the value of n that
comes closest to that given by the computed value of X,. Call
this/»-..
8.2J.5 Step J—Randomly select n, - n, additional units
from the lot Compute „,
03)
8.2.1.6 Step 6—Check on the adequacy of n, by taking *:
- 52. Compute A, - A/i,. Enter Table 1 and find the value of
n corresponding to X;. Call this n3. If n3 is much greater than
HI, for example, more than 20 %. randomly select n, - n;
additional units from the lot and return to Step 5. If n, is not
much greater than n^, proceed with Step 7.
8.2.1.7 Step 7—Using the final values obtained above.
calculate the following and accept the lot if
TABLE 1 V«*JM" of
Sin
-------
1JJIH E300
where n - «,. /:,. or n3, whichever is applicable. /„,,, is the
upper 0.05 point of a /-distribution for „ - t fcgreeV of
freedom, and s - j, or j, whichever is applicable. Otherwise.
reject the lot.
8.2.2 Example:
8.2.2.1 Assume that a contract covered the purchase of a
«S?Bima?tXai *"? a minimum P"nty specification of
98.0 %. The buyer and seller agreed that the probability of
rejecting a lot ot 98.0 * punry should be no greater than
0.05 and that ot accepting a lot as low as 97.0 % should be no
greater than 0.10. In this case, the pertinent levels are:
L - 98.0
L - d - 97.0
A- 1.0
8.2.2.2 On testing samples from ten units, selected at
random, the lot standard deviation was estimated to be:
»i - i, - 0.8
The values for X and A, were also calculated:
*- 97.5 To
X, - A/gs, - 1.0/0.8- 1.25
8.2.2.3 Entering Table I. the sample size n for X. = i 'e ;s
found to be 7. Accordingly, no further sampling is requirea
8.2.2.4 Substituting the above values in Eq 1 5:
(L - AVU/v/i) - (98.0 - 97.5)/(0.8/vTO)
-(OJx v'id)/0.8« 1.97
Since 1.97 is greater than 1.833 (the value for the upper 0 05
point of the /-distribution for 9 degrees of freedom) the "lot
should be rejected.
8.3 Single Upper Specification Limit (U); Simple Random
Sampling from a Large Lot— The procedures of 8 2 will
apply here except that U wiU replace L and U + A will
replace L — A. The criterion for acceptance will be:
M
8.4 Both Lower and Upper Specification Limits- Simple
Random Sampling jrom a Large Lot—Ust the following
sampling plan: Determine n, X. and s as in 8.2.1. Accept the
lot if
(L - XWs/Jn) S /aw. and (IT)
for n — I degrees of freedom. Otherwise, reject the lot.
8.5 General Remarks:
8.5.1 If A is small relative to the lot standard deviation, a
large sample size will be required to attain the low 0.10
consumer's and 0.05 producer's risks.
8.5.2 If the estimate of the lot standard deviation is less
than the true lot standard deviation, the sample size given bv
the above procedures will produce a sampling plan whose
risks will be different from those planned for. There will be a
greater sellers risk of having a lot rejected whose mean is
equal to the desired L level. Also, the buyer's risk of
accepting a lot. whose mean is below the L - d level for
barely acceptable quality, will also be greater than 0.10 (how
much greater depends on how far off the estimate of the lot
standard deviation may be).
8.5.3 If the estimate of the iot standard deviation is greater
than the true lot standard deviation, then the above proce-
dures wiij give a sample size iw) that is greater than necessarv
to yield the agreed upon risks. It will thus unnecessarily
increase sampling costs.
8.5.4 The risks stated in this practice are based on the
assumption that variability among units of the iot follows a
normal distribution and that the total quantity of material in
subsamples taken for testing does not exceed 10 % of the
total quantity m the lot. If vanability among units shows
evidence of considerable skewness. the logarithms of the data
-------
E 300
procedure employed in the inspection of the current lot.
Recommended procedures for estimating the batch variances
and the reduction and testing variances are given in the
Annex. In the sections that follow, it will be assumed these
estimates have been made.
9.1.4 A. Word of Advice—Before a particular program is
instituted, it would be desirable to review it with a statistician
to be sure that the recommendations of Section 9 are
. thoroughly understood.
9.2 Acceptance Tests Based on Current Samples:
9.2.1 Introduction—With knowledge of the basic vari-
ances for the product and for the method of reduction and
testing, the acceptability of a current lot from the given
stream of material can be determined as follows:
9.12 Formation of Composite Samples—For the purpose
of determining the acceptability of a current lot from the
given stream of lots, proceed as follows: Let the lot consist of
n, batches of material where n, is an integer. Presumably n,
is determined by the needs of the purchaser with respect to
his inventories, production, etc. (Note 7). Let n-, increments
ot material be taken at random from each of the n, batcnes
that make up the given lot and let n, be an even numoer.
(The determination of n* is discussed in 9.2.4). If the batcnes
are not distinct, take n,/t2 increments at random from the
lot. Form a composite of all the odd numbered increments
and another composite of all the even numbered increments.
Call the first composite A, the second composite B. Reduce
each composite separately and under uniform conditions run
two tests on each composite.
NOTE 7—A fraction of a batch should be treated as a whole batch in
determining n\.
9.2.3 Variance Formula—The variance formula for the
mean ( -~*i*.
9.2.9 The Acceptance Test when there is a Single Upper
Specification Limit (U)
9.2.9.1 Step /—Compute
Xv, + U + 1.645 (i6V/j, * '.-/"i"! * »r"/2 + »,2/4)'/2 .. .(4)
9.2.9.2 Step 2—Accept the lot if.? < .?t'«-
9.2.10 Acceptance Test when there are both a Lower
Specification Limit (L) and an Upper Specification Limit
(U):
9.2.10.1 Step /—Note whether U - L is greater than
If it is. continue to Step 2. If it is not. do not conunue.
Page 111
-------
E300
9.2.10.2 Step .'-Compute .F^ and .Ft. as m 9.2.8 and
7 •••*?• '
9.2.10.3 .Step .j—Accept the lot if .Fia < .f s .Ft.,
SIMPLE UQUIDS
10. Scope
10.1 This procedure covers the sampling of industrial
chemicals which are single-phase liquids under the condi-
tions ot sampling.
• NOTE 9—This procedure is based on Method D 270.
11. Summary
11.1 Samples of simple liquids are examined using var-
ious ASTM methods for the determination of physical and
chemical characteristics. It is accordingly necessarv that the
samples be truly representative of the simple liquids in
question. The precautions required to ensure the representa-
tive character of the samples are numerous and depend upon
the type of product being sampled, the tank, the earner or
container rrom which the sample is being obtained, the tvpe
and cleanliness of the sample container, ana the samoime
procedure that is to be used. A summary of the samounE
procedures and their application is presented in Table *"
Each procedure is suitable for sampling a number of specific
products under definite storage, transportation, or container
conditions. The basic principle of each procedure is to obtain
a sample or a composite of several samples in such manner
and from such locations in the tank or other container that
the sample or composite will be truly representative of the
product. Although single-phase liquids are homogeneous bv
definition, it may be desirable to check for this condition bv
sampling from various sections of the container.
12. Definitions
12.1 simple liquid—a single-phase liquid having a vapor
pressure of less than 16 psi Reid vapor pressure at 100'F (830
mm Hg at 37.8'Q and a Saybolt viscositv of less than
10 000 s (2160 cSt) at 25'C. '
IZ2 average sample—one that consists of proportionate
parts from all sections of the container.
12J all-levels sample—one obtained by submerging a
closed sampler to a point as near as possible to the draw-otF
level, then opening the sampler and raising it at a rate such
that is about three fourths full as it emerges from the liquid
An all-levels sample is not necessarily an average sample
because the tank volume may not be proportional to the
depth and because the operator may not be able to raise the
sampler at the variable rate required for proportionate filling.
The rate of filling is proportional to the square root of the
depth of immersion.
TABLE 2 Summary ol Samptoiq Pit
Type ot Container
"ype ot Samonno
Secnon
Storage tames (trucks, can. snps.
barges, stationary)
Storage tanks (trucks, can.
stationary)
Pipe »nes. filling mes. transfer
Drums, canov. cans. DOOMS
Free or ooerHMcnarpe streams
Bone samgang. tnef samotng 21.22
Taosanomg 23
Conenoous samodng 24
TUwaameano. 25
Jarsamcang 26
NOTE 10—The tuoe sampan? proceaure. 10.3. mav be useo to
obtain an all-levels sample tram a drum.
12.4 upper sample—*jne obtained from the middle of the
upper third of the tank contents (Fig. 1).
NOTE 11 —The taking ot' samples from various levels 01' the tank
permits the detection 01 variation in composition ot the contents caused
by stratification, if it is known that the contents are noc subject to trus
variation, the taking 01 samples at multiple levels mav be eliminated.
12.5 middle sample—one obtained from the middle of
the tank contents (Fig. 1) (Note 10).
12.6 single-tank composite sample—a blend of the upper.
middle, and lower samples. For a tank of uniform cross
section, such as an upright cylindrical tank, the blend
consists of equal pans of the three samples. For a horizontal
cylindrical tank, the blend consists of the three samples in
the proportions shown in Table 3.
12.7 compartment-tank composite sample (ship, barge.
etc.}—A blend of individual all-levels samples from each
compartment, which contains the product being sampled, in
proportion to the volume of material in each compartment.
!2.S ton sample—one normally obtained 6 in. (152 mm)
oeiow tne top sunace of the tank contents (Fig. 1).
J 2.9 »i
-------
TABLE 3 Sampling InatfucaotM tor Horizontal Cylindrical Tanks
(joudOnm.
Percant of
Diameter
Samomg Level. Percent of
Dimmer Aooy« Bottom
M*dia
Comootne Same
Piouciuunm Parts of
100
90
80
70
60
50
40
30
'.20
10
81
7i
71
3 50
5 50
3 50
50
50
40
• . . «
•
.
20 3
20 3
20 2
20 1
20
20
20
15
10
5
• 4
i 4
5
s
5
4
.
3
3
3
4
c
3
g
10
10
• V
10
10
12.14 drain sample—one obtained from the draw-off or
discharge valve. Occasionally, a drain sample may be the
same as a bottom sample, as in the case of a tank car.
12.15 bottom sample—one obtained from the material on
the bottom surface of the tank, container, or line at its lowest
point (Fig. 1). (Drain and bottom samples are usually taken
10 check for water, sludge, scale, etc.).
13. Sampling Equipment
13.1 General Requirements—all sampling apparatus and
closures shall be clean, dry, free of contaminants, and
constructed to materials that are inert to the product to be
sampled. The sampling container and closure shall be clean.
dry, and inert to the material being sampled.
132 Bottles and Jars—Bottles and jars may be made of
clear or brown glass or polyethylene with necks shaped to
receive a glass stopper or a screw cap made of metal or plastic
material Use of unprotected corks as closures is not recom-
mended for general use. Where safety indicates (such as for
peroxides) use corks covered with materials inert to the
sample, such as cellophane, polyethylene, or aluminum foiL
Clear glass is advantageous because the container mav be
examined visually for cleanliness and the sample may be
visually inspected for foreign matter. Brown glass affords
some protection for light-sensitive materials. Before using a
bottle or jar. examine it to see that it is scrupulously clean. A
variety of methods for cleaning glass containers may be used:
washing with detergents, chromic acid cleaning solution.
water, acetone, etc. The specific method used will depend
upon the material to be sampled. Care should be taken that
all of the cleaning agents are removed from the container
pnor to use. Close containers as soon as they are dry
13.3 Screw-Neck and Press-Cover Cans—Casaof tin plate
with seams soldered on the outside should be used. The neck
should be shaped to receive a screw cap or pressed cover
Care should be taken to ensure that cans are clean, even
when new. They may be cleaned by washing with low-
boiling, nonflammable solvents and blowing dry with clean
air. Cap the containers as soon as they are dry.
14. Time and Place of Sampling
14.1 Finished Products'—When loading or discharging
finished products, take samples from both snipping and
receiving tanks, and from the pipeline, if required.
14.2 Ship or Barge Tanks—Sample each product imme-
diately after the vessel is loaded, or just before discharging.
E300
14.3 Tank Caw—Sample the product immediately alter
the car is loaded, or just before unloading.
15. Number and Location of Samples
15.1 Bulk Containers (Tanks. Tank Cars etc.)—Simple
liquids in bulk containers are frequently found to be homo-
geneous and only limited sampling is usually required.
Upper, middle, and lower samples (21.3) or top and outlet
samples (21.5) can be individually tested to confirm this, by
means of simple physical tests such as refractive index.
specific gravity, viscosity; etc. Complete testing can then be
performed on a composite prepared as described in 21.4.
15.2 Packaged Materials (Drums, Cans. Bottles, etc.)—In
the case of lots of drums, bottles, and cans, the homogeneity
of the lot cannot be assumed, and the required number of
samples should be determined in accordance with Sections 5
and 6. The specific containers to be sampled for individual
testing should be chosen by means of a table of random
numbers.
16. Sampling Operations
16.1 Procedures for sampling cannot be made explicit
enough to cover ail cases. Extreme care and good judgment
are necessary to ensure samples which represent the general
character and average condition of the material. Gean hands
are important Gean gloves may be worn but only when
absolutely necessary, such as during cold weather, or for
reasons of safety. Select wiping cloths so that tot is not
introduced, contaminating samples.
16.2 When sampling relatively volatile products, the sam-
pling apparatus «MI be filled and allowed to drain before
drawing the sample. If the sample is to be transferred to
another container, this container shall also be rinsed with
some of the volatile product and then drained. When the
actual sample is emptied into this container, the sampling
apparatus should be upended into the opening of the sample
container and ITT""'" in this position until the contents have
been transferred so that no unsaturated air will be entrained
in the transfer of the sample.
16.3 When sampling nonvolatile liquid products, the
sampling apparatus shall be filled and allowed to drain
before drawing the actual sample. If the actual sample is to
be transferred to another container, the sample container
shall be rinsed with some of the product to be sampled and
drained before it is filled with the actual sample.
16.4 A sample shall be considered suspect under any of
the following circumstances and should be referred to the
appropriate supervisor before analysis:
16.4.1 The sample container is damaged or defective.
16.4.2 There is any doubt as to the nature of the contents
of the sample container for example, because of the presence
of an old label, incorrect markings, or insufficient identifica-
tion.
16.4.3 There is evidence of an unexpected lack of unifor-
mity, for example, a separate layer or suspended matter.
16.4.4 Obvious and unusual variations are apparent in the
sample.
16.4.5 The container closure is loose, whether or not there
is evidence of leakage.
16.4.6 Evidence that the closure or liner has been at-
tacked.
Page 113
-------
17. Size of Sample
17.1 The quantity of sample should be as specified bv the
test instructions, or at least three times greater than the
minimum necessary for the actual tests.
18. Precautions
r (2 to 16 psi Reid . __ , tJJM,e ut
to 830 mm ffg at 37.8'C))—h is necessarv to
protect volatile samples from evaporation. Transfer'the
product trom the sampling apparatus to the sampie con-
tainer immediately. Keep the container closed except when
* material is being transferred.
18.2 Light-Sensitive Samples—It is important that sam-
ples sensitive to light be kept in the dark if testing is to
include the determination of such properties as color
inhibitor content, stability tests, or neutralization values
Brown glass bottles may be used. Wrap or cover clear glass"
bottles immediately. It is a definite advantage to use covered
metal or cardboard containers into which the sampie bottles
may be placed immediately after collection
18.3 Materials ot High Purity— Protea highly rerinea
products trom moisture and dust by placing paper, piastic or
metal foil over the closure and the top of the container
18.4 Container Outage—Hevcr completely fill a sampie
container, but allow adequate room for expansion, takine
into consideration the temperature of the liquid at the time
of filling and the probable maximum temperature to which
the filled container may be subjected.
19. Shipping Precautions
19.1 To prevent the loss of liquid during shipment and to
protect against moisture and dust cover the closure of the
glass bottle with plastic caps which have been swelled in
water, wiped dry, placed over the top of the stoppered bottle.
and allowed to shrink tightly in place. Screw-top bottles are
recommended. The cap should be lined with material inert
to the sampie. The screw caps should be secured by use of
adhesive tape or similar material.
NOTE 12-Shipping of any chemical must comply w,« current
federal, state, and local reguiuoos for the specific materialiS
shipped. •"•«•»
20. Labeling Sample Containers
20.1 Label the container immediately after a sampie is
obtained. Use waterproof and oil-proof ink or a pencil hard
enough to dent the tag, since soft pencil and ordinary ink
markings are subject to obliteration from moisture, oil
smearing, and handling. If gummed labels are used, thev
should be further secured with transparent sealing tape
Sufficient detail should be written on the label to completely
identify the sample. The following information is frequently
-desirea:
20.1.1 Date and time (and for continuous and dipper
samples the hour and minute of collection).
„ 20.1.2 Name of sampler.
20.1.3 Name or number and owner of the vessel, car. or
container.
20.1.4 Brand name, grade of material, and code number
and
20.1.5 Reference symbol and necessarv identification
number.
Page 114
E300
21. Bottle Sampling
21.1 The bottle sampling procedure is applicable for
sampling simple liquids in tank cars, tank trucks, shore
tanks, ship tanks, and barge tanks. A suitable sampling
bottle, as shown in Fig. 2. is required. The diameter of the
openings in the bottles should be -ft-in. (19-mm). Stopper
and label bottles immediately after taking them and deliver
them to the laboratory in the original sampling bottle.
NOTE 13—The designs and dimensions which follow are intended
only as guides to the form that the sampling apparatus may take. When
metal is required for construction of the sampling apparatus, a corro-
sion-resistant type steel should be selected (Type j 16L may be suitable).
If flammaole materials are to be sampled, a nonmagnetic low-spark
generating stainless steei is required. When sampling flammable liquids.
extreme care should be exercised not to shaipiy strike the container
being sampled with the sampling apparatus. Alternative procedures may
be used if a mutually satisfactory agreement has been reached by the
parties involved.
21.2 All-Level Sampie— Lower the weighted, stoppered
bottle as near as possible to the draw-off level, pull out the
stopper with a sharp jerk of the twine or chain (spark-proof)
attached to the stopper, and raise the bottle at such a rate
that it is about three-fourths full as it emerges from the
liquid.
21.3 Upper. Middle, and Lower Samples—Lower the
weighted, stoppered bottle to the proper depths (Fig. 1).
which are as follows:
Upper sample
Middle ample
Lower sample
ituddle of upper tluid of the taok
miririlf of the tank content
nuddle of lower Hurt of the tank
1 out the stopper with a sharp jerk of the twine or chain
(spark-proof) attached to the stopper and allow the bottle to
ALTERNATE RIO
FKL 2 AiMfltoty for Beta* Sampttng
I LITER (I OUAHT1
WBOHTED BOTTLE SAMPLER
(CM 1C MIMCMIO TO W •"»
•OTR.CI
-------
E300
fill completely at the selected level, as evidenced by the
cessation of air bubbles. When fulL raise the bottle, pour off
a small'amount, and stopper immediately.
21.4 Composite Sample—Prepare a composite sample in
the laboratory (not in the field) by mixing portions of
all-levels samples as specified in 12.7 or by mixing portions
of the upper, middle, and lower samples as specified in 12.6.
21.5 Top and Outlet Samples—Obtain these samples (Fig.
• 1) in the same manner as specified in 12.13. but at the
following depths:
Top'sampie
Outlet sample
6 in. (152 mmi below the top suriace or the tank
opposite me tank outlet (either fixed or swing Une
outlet)
22. Thief Sampling
22.1 The thief sampling procedure is applicable for ob-
taining bottom samples (Fig. 1). of liquids of 2 psi Reid
vapor pressure at 100T (105 mm Hg at 37.8'G or less, in
tank cars and storage tanks.
22.2 Thief—The thief shall be designed so that a samoie
can be obtained with '/• in. 113 mmi of the bottom or the car
or tank. Two types of thiefs are illustrated in Fig. 3. One type
is lowered into the tank with valves open to permit the liquid
to flush through the container. When the thief stnkes the
bottom of the tank, the valves shut automatically to trap a
bottom sample. The other type has a projecting stem on the
valve rod which opens the valves automatically as the stem
3—-CHAIN FOR
UOWCRING
-VLUGS
1*101633
t-J 1/2-CB.39 cm) -
OU.
stnkes the bottom of the tank. The sample enters the
container through the bottom valve and air is released
simultaneously through the top. The valves snap shut when
the thief is withdrawn.
22.3 Procedure— Lower the clean, dry thief through the
dome of the tank car or tank hatch until it strikes the
bottom. When full, remove the thief and transfer the content
to the sample container. Cose and label the container
immediately, and deliver it to the laboratory.
23. Tap Sampling
23.1 The tap sampling procedure is applicable for sam-
pling simple liquids in tanks which are equipped with
suitable taps or lines. The assembly for tap sampling is
shown in Fig. 4.
23.2 Tank Taps—The tank should be equipped with at
least three sampling taps placed equidistant throughout the
tank height and extending at least 3 ft (9 m) inside the tank
shell. A standard '/i-in. (6-mm) pipe with suitable valve is
satisfactory.
23.3 Tube—\ delivery tube which will not contaminate
:he product being sampled and long enough to reach to the
bottom of the sample container is required to allow sub-
merged filling.
23.4 Procedure—-Before a sample is drawn, flush the tap
(or gage glass drain cock) and line until they are purged
completely. Connect the clean delivery tube to me tap-Draw
upper, middle, or lower samples directly from the respective
isi'MO.ocm)
(•) Bomb-TypM Samoanq Thief
(6) Core runt. Tap-Type
FIG. 3 Samptmcj Thtoti
Page 115
-------
E300
FIG. 4 AsMmttv tar Tap Sampling
taps after the flushing operation. Stopper and label the
sample container immediately after filling, and deliver it to
the laboratory.
24. Continuous Sampling
24.1 The continuous sampling procedure is applicable for
sampling simple liquids in pipe lines, filling lines, and
transfer lines. The continuous sampling may be done manu-
ally or by using automatic devices.
24.1.1 Precaution—The sample line should be purged
three times before the sample is taken and special precau-
tions should be taken to minimize exposure to the chemical
being sampled.
24.2 Sampling Probe—The Junction or' the sampling.
probe is to withdraw from the tlow stream portion that wiil.
be representative of the enure stream. The apparatus as-
sembly for continuous sampling is shown in Fig. i. Probe
designs that are commonly used are as follows:
24.2.1 A tube extending to the center of the line and
beveled at a 45* angle facing upstream.
24.2.2 A long-radius elbow or bend extending to the
center line of the pipe and facing upstream. The end of the
probe should be reamed to give a sharp entrance edge.
24.2.3 A tube extending across the pipeline with holes or
slots facing upstream. The position and size of the probe
should be such that it will minimize stratification and
dropping out of heavier panicles within the tube.
NOTE 14—Although this discussion is limited to simple liquids which
are assumed to be uniform in compostuon. it is possible that under
certain conditions, temporary stratification (caused by pressure, temper-
ature gradients, eic.i may exist and. therefore, certain precautions are
advisea to ensure obtaining representative samples.0 . ,
24.2.4 To control the rate at which the sample is with-
drawn, the probe or probes should be fitted with valves or
plug cocks.
24.2.: A clean, dry container of convenient size shall be
used to receive the sample. All connections from the sample
probe to the sample container must be free of leaks. The
container shall be constructed in such a way that it retards
evaporation loss and protects the sample from extraneous
material such as rain, snow, dust and trash. The construc-
tion should allow cleaning, interior inspection, and complete
mixing of the sample prior to removal The container should
be provided with a suitable vent.
24.3 Automatic Sampling Devices:
• Rinhian. 1. H_ and Hiltaud. J. G. "SaauAni of Nonhoino«eneoui flow in
Pipes.' Preprint No. 52-64. frtxttauin. Amencan Petroleum liumutt. PPTTA.
VoL 44. Section 3. 1964. pp. JI7-534.
. •»• If VCl
(HO *C*HIO TO
- VM*» to*-:
NOUS o« tiers
r*cme i "
!/•--•
T
TO RECEIVER
OR SAMPLER
tA|
TO RECEIVER
OR SAMPLER
(II
TO RECEIVER
OR SAMPLER
CCI
e"
> MoMieNTAkLT en «c*ne«kkr.
P^OBCS FOR CONTIMUQUS SAMPLING
Xno»
r
AUTOMATIC
| SAMPLING DEVICE I
I (IF USED!
PLUSH OR
DRAIN
•CTUMN LINE
OR CHAIN
SAMPLE
RECEIVER
IDJ TYPICAL ASSEMBLY rnp CONTINUOUS
S PfMNUtorCom
Page 116
-------
E300
«0* (1016 em) —
FK3. 6 Sampting Tub*
24.3.1 Time Cycle (Nonpropomonal) Types—A sampler
designed and operated in such a manner that it transfers
equal increments of liquid from the pipeline to the sample
container at a uniform rate of one or more increments per
minute is a continuous sampler.
24.3.2 Intermittent Sampler—A sampler that is designed
and operated in such a manner that it transrers equal
increments of liquid from a pipeline to the sample container
at a uniform rate of less than one increment per minute.
24.3.3 Flow-Response (Proportional) Type—A sampler
that is designed and operated in such a manner that it will
automatically adjust the quantity of sample in proportion to
the rate of flow is a flow-response (proportional) sampler.
Adjustment of the quantity of sample may be made either by
varying the frequency or transferring equal increments while
maintaining a constant frequency of transferring the incre-
ments to the sample container.
24.4 Procedure:
24.4.1 Nonautomatic Sample— Adjust the valve or plug
cock from the sampling probe so that a steady stream is
drawn from the probe. Measure and record the rate of
sample withdrawn as gallons per hours. Divert the sample
stream to the sampling container continuously or intermit-
tently, to provide a quantity of sample that will be sufficient
size for analysis. Label the sample and deliver it to the
laboratory in the container in which it was collected.
24.4.2 Automatic Sampling— Purge the sampler and the
sampling lines immediately before the start of a sampling
operation. If the sampler design is such that complete
purging is not possible, circulate a continuous stream from
the probe past or through the sampler and back into the line.
Withdraw the sample from the side stream through the
automatic sampler using the shortest possible connections.
Adjust the sampler to deliver not less than 1 and not more
than 40 gal of sample during the desired sampling period.
For time-cycle samplers, record the rate at which sample
increments were taken per minute. For flow-responsive
samplers, record the proportion of sample to total stream.
Label the samples and deliver them to the laboratory in the
containers in which they were collected.
NOTE 15—For time-eyrie samplers, derations in quantity of the
sample taken should not exceed ±5 % of the average rate for a given
setting. For flow-responsive samplers the deviation in quantity of sample
taken per 42 000 gal of flowing stream should not exceed ±5 % of the
chosen average.
25. Tube Sampling
22.1 The tube sampling procedure is applicable for sam-
pling liquids in drums and cans.
22.2 Tube—Either Type 316L stainless steel or other
material suitable for the particular liquid may be used. The
tube should be designed so that it will reach to within about
s in. (3 mm> of the bottom and have a capacity of
approximately I pt (5 m) or 1 qt(9 m). A metal tube suitable
tor sampling 55-gal (208-ml drums is shown in Fig. 6. Two
nngs, attached to opposite sides of the tubes at the upper
end. are convenient for holding it by slipping two fingers
through the rings—thus leaving the thumb free to dose the
opening. An alternative tube sampling apparatus is shown in
Fig. 7. This tube is ak" designed to reach within ft in. of the
bottom.
25 J Procedure for Drums:
253.1 Stand the drum upright and sample from the top.
If the drum dog not have a top bung, place the drum on its
side with the bung up. Thorough mechanical agitation of the
drum prior to sampling will ensure that its contents are
uniform. If detection of water, rust, or other insoluble
contaminants is desired, let the drum remain in the sampling
position long enough to permit the contaminants to collect
«.••» I V 00 STMLCM STKL
ON CUSS TutMO
•OUCTNIICNC
•UM
OHM* on
CM TO
M MMfUO
FIG, 7 AttwiMttv* Tub* Sampling AaftamMr
Page 117
-------
it the top or bottom, and take a top and a bottom sample.
Remove the bung and place it beside the bung hole with the
wet side up. Close the upper end of the clean, dry sampling
tube with the thumb, and lower the tube into the liquid for a
depth of about 1 ft (304 mm). Remove the thumb, allowing
the liquid to How into the tube. Again close the upper end
with the thumb and withdraw the tube. Rinse the tube with
the liquid by holding it nearly horizontal and turning it so
. that the liquid comes in contact with that pan of the inside
surface which wiii be immersed when the sample is taken.
Avoid handling any pan of the tube that will be immersea in
the liquid duhng the sampling operation. Discard the nnse
liquid and allow the tube to drain. Insert the tube into the
liquid again, holding the thumb against the upper end. (If an
ail-levels sample is desired, insert the tube with the upper end
open.) When the tube reaches the bottom, remove the thumb
and allow the tube to fill. Replace the thumb, withdraw the
tube quickly, and transfer the contents to the sample
container. Do not allow the hands to come in contact wuh
any pan of the sample. Cose the sample container replace
and tighten the bung in the drum. Label the samote
container and deliver it to the laboratory.
25.3.2 In using the alternative sampling device. :r.e
sample shall be pumped directly into the sample bottle by
means of a double-valve aspirator bulb. Samples at various
levels may be obtained by adjusting the depth of the tube in
the drum or can. Before collecting the sample, thorouenly
flush the device with the material being sampled.
25.4 Procedure for Cans—Obtain samples from cans of
5-gal (I9-L) capacity or larger in the same manner as from
drums (25.3.1) using a tube of proportionately smaller
dimensions. For cans of less than 5-gal capacity, use the
entire contents as the sample, choosing cans as prescribed by
the selected sampling plan section or in accordance with
agreement between the purchaser and the seller.
26. Jar Sampling
26.1 The jar sampling procedure is applicable for sam-
pling liquids where a free or open-discharge stream exists as
in small filling and transfer pipelines (2 in. (51 mm) in
diameter or less) and filling apparatus for bottles and cans.
NOTE 16—Jar sampling is parbcularry subject to contamination of
the material being sampled. Great care should be exercised to be sure
that foreign matter is not introduced into the sample from the air or
surroundings.
26.1.1 yar—Use a dean, dry, glass jar with screw cap. The
cap should be lined with material inert to the sample.
26.1.2 Procedure—-Insert a jar in the free-flowing stream
so that a ponton is collected from the full cross section of the
stream. Appropriate safety measures should be observed.
Take pontons at time intervals chosen so that a complete
sample propomonal to the pumped quantity is collected.
Samples collected may be analyzed individually or com-
posited to provide an average sample of the material
pumped.
SOLIDS
27. Scope
27.1 This practice covers equipment and procedures for
sampling materials that are solids (see 28. n at the time of
E300
sampling. The equipment and procedures that are described
in these sections are intended to supplement the experience
of the sampler as a guide in selecting methods that are
applicable to the material being sampled.
27.2 Subjects covered in these sections appear in the order
shown in Table 4.
28. Description of Terms
28.1 solid—a state of matter in which the relative motion
of molecules is restricted and in which molecules tend to
retain a definite fixed position relative to each other. A solid
may be said to have a definite shape and volume.
28.2 sampling—the process of extracting a small fraction
of material from a larger bulk, so that it will be sufficiently
representative of the bulk for the intended purpose.
28.3 lot—a discrete quantity of material. It may contain a
single batch or several batches, or be the product of
continuous process broken into units on the basis of time or
shipment. It is very desirable that individual batches in a lot.
be specifically identified so that they may become individual
or stratified units for inspection.
2S.4 increments—portions of material selected from var-
ous parts of a lot. which may be tested individually or
composited and tested as a unit.
28.5 gross sample—a composite prepared by mixing the
increments.
28.6 subsampte—a smaller sample produced in a speci-
fied manner by the reduction in volume or quantity of the
gross sample.
28.7 laboratory sample—that portion of the subsampie
which is sent to the laboratory for testing.
29. General Principles and Precautions
29.1 Every sample should be collected and piepared in
stria accordance with a specified procedure.
29.2 Because of many variations in the conditions under
which solids must be sampled, and in the nature of the
material being sampled, it is essential that the samples be
collected by a trained and experienced sampler. Because of
variations in the manner of handling the solid, it is impos-
sible to specify rigid rules describing the exact manner of
sample collection. Correa sampling principles must be
applied to conditions as they are encountered.
29.3 To be able to make probability, or confidence
statements about the property of a lot the sampling proce-
dure must allow for some element of randomness in selec-
tion because of the possible variations in the quality of the
TABLE 4 Stannary of ProcattJiM
Gcnra Pnncnies ana Precmuons
Juno Scoop
Stream Samping Cup
ShoMSamoMr
ThMSarnoMn
Sol Samp* Auger
Madra SamMrB
Apptcaoon at Sarmnq Eouormm
Pniparauon of Heducoon of Samow
Laboratory Sanaa ana Storaq* free
Uoamo Samoa Comarara
28
29
30
30.1
3O2
30J
30.4
3O5
3O6
31
32
33
34
Page 118
-------
material. Generally, where segregation is known to exist, and
random variation of quality is not possible, the sampling
shouid.be designed to allow tor this. The sampler should
always be on the alert for possible biases arising from the use
ot' a particular sampling device or from unexpected segrega-
tion in the material. Generally, where sampling, is to be
applied to the output of a given process on a continuous
basis, it will be desirable before adopting a particular
sampling plan, to undertake an extensive preliminary study
' of variation in the material and possible biases in sampling
instruments and methods of reduction.
29.4 The statistical principles governing the number and
location of the samples taken from packaged lots ot* solid
materials are essentially those outined in Sections 6. 7. and 8.
on statistical considerations.
29.5 Whenever possible, nonpackaged. bulk materials
should be sampled while the material is in motion rather
than in static piles, carloads, etc. Such occasions are fre-
quently ideal for the application of falling-stream samplers.
29.6 Sampling of bulk solids from boxcars, barges, etc..
introduces additional problems because of possible
nonuniformity in panicle size, moisture, impurities, etc. The
statistical treatment is complex and beyond the scope ot this
practice. For a typical example, see Methods D 2224. ana
Ref(7).7
29.7 All auger methods and all scoop methods used on
materials not being loaded or transferred fail a prime
sampling requirement—that of random selection of the
panicles or portions selected as samples. Scoops and shovels
are limited to use at or near the top surface. Augers and thiefs
are normally inserted in a preset pattern. Consequently,
panicles on the bottoms or along certain sides of containers
never have an opportunity to be included in a sample. For
heterogeneous or valuable material, this alone may furnish
sufficient reason to go to a falling-stream sampler.
29.8 Because of the above factors, the recommended
procedures that follow are limited to the mechanical opera-
tions of taking the required number of increments called for
in another standard or in a purchase contract (1,2).
29.9 The sampling equipment, sample preparation equip-
ment, containers, etc.. used in sampling must be clean, dry,
uncontaminated. and inert to the material being sampled.
and protection from heat. cold, light, loss or gain of moisture
may be necessary.
30. Sampling Equipment
30.1 Hand Scoop, for sampling powders from containers
and conveyors:
30.1.1 This implement is used for taking small equal
portions at either random or regular intervals from the mass
of material to be sampled. It is most frequently used to
sample drums, bags, barrels, or other containers, but may
also be used to take pontons trom a ilowing stream, such as
a belt conveyor, in a chute, etc.
30.1.2 The scoop can be 01" any suitable size or shape.
depending, in pan. on the size and shape of the panicles in
the matenai to be sampled and the quantity of sample
required.
E300
30.1.3 A sample of a flowing stream should be taken by a
single mouon of the scoop in such a way as to take a
complete cross section of the stream. The scoop should not
overflow during this single motion.
30.1.4 Scoop sampling of static material consists of taking
samples at or near the surface, and requires nearly perfect
homogeneity, a condition that rarely exists for all character-
istics of the material. The larger panicles, especially if they
approach the size of the scoop, will frequently be rejected in
the sample taking.
30.2 Stream Sampling Cup. for sampling powders from
conveyors and chutes:
30.2.1 The cup is used for selecting samples from a
(lowing stream, such as a conveyor, a chute, or a belt
30.2.2 The size of the cup depends upon the diameter of
the panicle being sampled and the width of the stream of
powder. The mouth width of the sampling cup should be at
least three times the diameter of the largest panicles being
sampled. The mouth length of the cup should be sufficient to
cut the entire stream of material as the material drops from a -
transfer belt. Figure 8 indicates a design of a suitable cup.
30.2.3 The cup is passed through the entire stream of
matenai as it drops from a belt or a chute. The approximate
discharge time should be predetermined in order to secure a
minimum of ten alternating, and equally timed, spaced cuts.
The cup should be passed through the entire stream in a
uniform motion, at the predetermined intervals throughout
the loading operation regardless of the size of the sample or
number of passes required. The stream samples is not
recommended normally for many materials unless a uniform
continuous flow of materials is maintained for at least 3 min
while the lot is sampled.
30.3 Shovel, for sampling large bulks:
30.3.1 A shovel is used for taking samples from larger
bulk shipments such as freight cars, boats, and truck loads. It
is most advantageous when material is being loaded or
unloaded, or moved by shoveling. It sutlers the same
disadvantages as the hand scoop.
30.4 Thiel Samplers:
30.4.1 Split Tube Thief:
30.4.1.1 This instrument is essentially a tube, usually %
in. (19 mm) in diameter, with a slot running the entire length
' The boldface numbers to
to this pncttce.
to we bst of' references appended
HO. 8 Strarni Sampling Cup
Page 119
-------
E300
FIG. 9 Split Tub* Thief
angfed
,„ 3ri4rh TK lhlC " i^Xmd im° the matenal r enoueh
u> reach the opposite side (or the bonom» of the container
The thief «s then carefully wuhdrawn. and the mc«m?m Ts
extruded into the sample container. >-«™«u is
30.4. 1 .3 The split tube thief is especially smtable for stickv
material, m which case the sampie mav need to be removed
with a spatula or other suitable device. removea
30.4.2 Concentric Tube Thiets:
304.2.1 This equipment is 'used for taking samnies or
free-rtowng materials like grain, from drum' ans
and other containers. Two types are described.
-0.4.2.2 Multi-Sot Tube r/««~This apparatus
rotating the inner tube (Fig. 101.
• 30A2::3 J*" lhief is insened «n the material with the
inner tube holes closed. The inner tube is rotated to u own
position to extract a sampie of the matenal and it '£25
position before the thief is withdrawn from the container
30.4.2..! U^fcSfa, n*r TW-TTiis
of two tubes fitting snugly into each other
has a slot running lengthwise and has a pointed
°ne to °
30.4.2.5 The sample thief is closed so that the lower end
of the outer tube rests on the shoulder at the bottimTf the
mner tube. and the inner tube is locked in position Sto he
thumo screw. The sampie thief is then pushed into the
material diagonally or horizontally, as applicable. The
tube is then unlocked and raised a few SteJ Texpc
slot of the inner tube to the material. The slot is
upward. The drum is shaken or jarred
enter the thief at the level of the slot o
is shaken while the sampie thief is opened
allow material from all levels to enteTSe
sample is in the inner tube, the outer tube should be pushed
down to its original position. The thief is then removed from
the matenal and inverted so that the sample drops into the
sample bottle through the open end. It may be neceSSv w
rap the thief sharply in order to dislodge the powder "
30.4.16 These concentnc tube samplers have 'limited
applicability. Matenal that ,s not free-tlowing or is hard
FIG. 11 Single-Slot Tube Thief
FIG. 12 Grain Probe
packed is excluded, thus usually eliminating tine powders.
On the other hand, the sampling of matenal containing
granules or particles exceeding one third of the slot width
should not be attempted, or bridging and resulting bias in
favor of the small panicles may result. Because of their
pointed enos. these devices cannot sampie the bottoms of the
containers. If material has been vertically segregated into
-.onzontal strata through vibration, or any other reason, the
jwest strata will be inadequately represented. These prob-
ems are common to both tubes.
30.4.3 Companmcmai Thiets (Triersr
30.4.3.1 This equipment is used for taking samples of
free-flowing materials like fertilizers, grain, and other pow-
ders trom bags, drums, cans, piles, carloads, and bins. Two
types are descnbed.
30.4.3.2 Grain Probe—This apparatus consists of two
tubes, one fitting snugly inside the other. One end of the
outer tube may be tapered or fitted with an auger point. The
trier is 63-in. (1600-mm) long, with an outside diameter of
1H in. (34.9 mm): an inside diameter of 1 'A in. (2&5 mm)
with eleven compartments 3'/:-in. (88.9-mm) long; separated
by 1%-in. (34.9-mm) long plugs (Fig. 121 The outer tube
consists of slots that correspond to the compartments of the
inner tube. The outer tube slides over the inner tube.
30.4.3.3 The trier shall be insened into the material
vertically and should not be pointed toward the center of the
load. Open the tube with the slots facing upward, then close
the tube, and withdraw the sampie. The sampie shall be
discharged into a receiver as long as the sampling tube.
30.4.3.4 Afissoun Trier—This apparatus consists of two
tubes, one fitting snugly inside the other. The trier is an
imemipted core-companmental double tube. The trier is
59-in. (1498-mmi long, with an outside diameter of 1% in.
(28 mm), an inside diameter of % in. (22 mm), and with
eight compartments 3 in. (76 mm) in size. The outer tube
consists of slots that correspond to the compartments of the
inner tube. The tner operates in the same tashion as the gram
probe. The tner shall be msened as specified in 30.4.3.3. The
slot width shall be at least three times the diameter of the
largest particles to be sampled (Fig. 13).
10 MuW-Slot Tube Thief
Page 120
-------
E300
rteo
16 Gnwtyftow Ai
flowing mass. The material captured in the sample tuoe is
augered out of the tube by an internal worm screw A
soienoid switch actuates the motor-driven auger at preset
intervals and simultaneously engages a dutch to rotate the
auger tube. The combination of auger pitch and rotation
must be such as to remove the collected material to a
collection chute before the sample can fall out on the
opposite side through the more slowly rotating slot
30.6.5.3 This sampling device has the advantages of
relative simplicity and little occupied space. Another^ria-
n^*^?8*811 " °M " ****** °Pen slot is alwavs
upward and does not rotate, in which case the rotating aucer
D!fL,?ITy away thc coUcct«* sample before bridjJngTr
overfilling can occur. For both designs, slot width andlengih.
auger pitch and variation of pitch along axis, rotational
speeds, flow rate of the bulk mass, and the amount of sample
required must ail be properly matched for accurate samotine.
The disadvantage is that such a device cuts only pan of the
stream pan of;the time. Therefore, if the flowing stream is at
all segregated in its cross section, a nonrepresentative sample
wdl result unless all segregated layers are pnnwrnonaSycut
\"»
30.6.6 Falling-Stream Samples:
30.6.6.1 The most reliable method of removing a sample
from a bulk mass employs a tailing-stream system wherea
moving cutter removes all of the falling stream pan of the
time. Such cutters fall into the two general categories of
arc-path and straight-path samplers. Slot widths of the cutter
should be at least three times the diameter of the largest
panicles to be sampled: four times or more is preferable
Obviously, the speed of travel through the stream is one
control of the sample size collected, but the speed of the
cutter should not be so great as to knock the parades awav
(Fig. 17).
30.6.6.2 Arc-Path Samples—The most popular and prob-
ably the best performer of the arc-path samplers is the
Vezin-type shown in the left half of Fig. 17. The material
falls from a belt or vibratory feeder or is fed through a chute
as a vertically falling stream that is cut by the radiattv
fCO
KCJtCT
FKL 17 FalUng-Strawn Sa
rotating oriented slots of the cutter. Such a device will have
one to usually not more than four slots. Material collected by
the slots falls into the sample chute while the bulk of the
material falls by into the reject stream. Mechanically, the
Vezin sampler has the advantage of simple rotary motion.
but it will not cut equal percentage from all pans of a stream
if the siot sides are not perfect radii. The quanuty of sample
collected is controlled by siot width, number of slots.
trequency of the slots passing through the stream (rotational
velocity), and the rate of stream flow.
30.6.6.3 Straight-Path Samplers--With a straight-path
sampler, the bulk material falls from a moving belt or other
feeder, in a vertical stream through which passes a rectan-
gular siot as shown in the right half of Fig. 17. Sample
collected is usually diverted through an angled chute into a
sample receptacle, and the gross reject material falls directly
downward. The amount of sample rollfctpd is controlled by
feed rate, slot width, cutter speed, and frequency. This
sampler cuts every pan of any shaped stream proportion-
ately, and is potentially the most accurate type. Many
variations occur in slot design and orientation and in the
drive mechanism. The common Geary-Jennings type has a
drive in which the cutter carriage is actuated by a heavy,
motor-driven screw.
30.6.6.4 The cross-cut sampler is a specific modd of a
straight-path sampler intended for installation in a spout or
chute as shown in Fig. 18 (4). Because of the limiting
enclosure, the material sampled must be free-flowing. The
apparatus consists of an air-actuated head in a box. a control
box, and an airline connection. Collected sample is dis-
charged through a flexible tube at the bottom of the sampler.
31. Example of the Application of Sampling Equipment
31.1 Thief Sampling from a Container (S):
31.1.1 Remove a thief sample from each of the shipping
containers selected for sampling in accordance with Section
5 on statistical considerations.
31.12 Nearly all containers are filled in such a way that
segregation occurs in the filling. For example, the large or
heavier panicles roil to the outside and the small or light
particles remain under the pouring spout where they fall.
Additional segregation will probably result from the vibra-
tion of shipping. Therefore, sampling patterns are devised so
that samples are taken in locations to represent as accurately
as possible the segregated layers or regions. Cylindrical
containers, or structures such as solidified metal pours, will
Page 121
-------
E300
FK2. 13 Missouri Trier
30.4.3.5 It has been found that these triers secured sam-
ples that were closely comparable and most nearly represen-
tative of the material being sampled. These triers have the
tendency to secure samples that are biased to varying degrees
in selecting more of the smaller size particles and less 01" the
larger panicles fraction. The triers are at the present time
being used by the fertilizer industry (3).
30.4.3.6 Because of the close clearances, double-tube
thiefs and triers will impart a grinding action to the material
being sampled. Soft granules are affected by such action, and
thiefs should not be used for such material if product sizing is
important.
30.5 Soil Sample Auger, for sampling compact materials:
30.5.1 This is a screw-or-worm-type instrument useful for
taking samples of compacted materials (Fig. 14).
30.5.2 The auger is turned into the material and then
pulled straight out The sample is removed from the auger
with a spatula or other suitable device. The process is
repeated at different locations as dictated by the sampling
plan.
30.6 Machine Samplers, for sampling powders from con-
veyors, bins, and containers:
30.6.1 Vacuum Probe Samplers, for large bulk containers:
30.6.1.1 This equipment can be used for extracting large
samples from freight cars, barges, bins, boats, and truckloads.
but only where air exposure does not affect significant
properties of the material, such as moisture content. This
type of sampler develops bias, if sizing is important. It
preferentially selects fines.
30.6.1.2 The apparatus (Fig. 15) consists of a combination
cyclone separator and motor driven blower, a probe and
connection tubing (4).
30.6.1.3 This equipment works the same way as a vacuum
cleaner does. The probe burrows its way into the material
being sampled and sucks the material into the sample
collector.
FKL 14 Auger Sampler
FKL 15 Vacuum Sampler
30 6 2 In general, augering probably offers the best com-
bination of economy, penetration ability, and sample repre-
sentation, if the material is packaged in drums or similarly
sized containers that are to be moved or transhipped without
dumping. Although there are many designs, augers tall into
the two general categories of open and enclosed **&**•
30.6.3 Powered Open Auger—One of the most useiui
varieties of the open type is a ship auger aboutl%«n.
(30.16-mm) diameter, powered by a ^-operaed A-m.
(19.05-mm) drill The augering is performed through_ a hole
in a catch pan that collects the sample brought to the top
Contents of the pan are then dumped into a sample
container. Open augering may not give .^^"""JE
sentauon of the container because material at ttar»p maybe
preferentially removed at the expense of the_tower layers.
Since many materials are frequently segregated vertically, a
biased sample may result.
30.6.4 Enclosed Auger u^_..«f«-
30.6.4.1 Enclosed augers may either ^ the ship-auger
type or have a central shaft with one or more foghtsJto'either
case; it will be surrounded by a *W^ *heath
which does not rotate. Material removed
discharged through a side hose at the top, or tt
in the sheath for discharge by reversing the auger alter
withdrawal from the drum. •«-.«»;««
30.6.4.2 Because of the power required for the penetraaon
drive and withdrawal, as well as the rotary «onon, a fixed.
permanent installation is required for an enclosed auger-
Therefore, it is applicable only when a large number ot
similar drums or containers are to be sampled ow a tang
period of time. An enclosed auger will obtain much^ im-
proved vertical representation over an open auger, alfcough
it is also deficient in sampling the bottom 1 to 2 in.
-------
E300
FIG. 18 Cross-Cut Samottr
commonly exhibit radial segregation and occasionally an-
gular segregation (variation is observed along the circular
path around the center). Sampling positions are calculated so
as to represent annular rings of constant volume in pro-
ceeding from the center to the periphery. Angular or
pie-shaped segments would be preferred but are usually
impractical.
31.1.3 Except where a definite sampling pattern as previ-
ously described is to be followed, the sample equipment
should usually be inserted diagonally into the container (3).
31.1.4 Individual samples from a single container may be
composited if necessary to obtain a sample of adequate size
for that container.
31.2 Machine Sampling from a Flowing Stream:
31.2.1 Sampling a material in motion, especially in a
free-falling stream, is the preferred method for obtaining the
most representative sample.
31.12 Arc-path or straight path samplers may be com-
bined to give a series of two or more stages of sampling. In
the design and operation of such a system, care must be
taken to avoid air flows for dust collection, etc^ which might
bias the sample.
31.2J If operations are short term so as not to justify
installation of a complete falling-stream system, the falling
stream sampling may be attempted manually. The place
should be accessible and sate for the person taking the
sample. A scoop or slot (see 30.2) with parallel sides should
be swept through the stream at a steady but sufficiently rapid
rate so that u does not overfill on one pass. Passes should be
timed and made at exactly regular intervals. Sampling under
gondola cars is particularly difficult and should be replaced
with sampling from a conveyor belt if possible.
31.3 Auger of Shovel Sampling from Cars. Ships, Barges.
etc.:
31.3.1 The following is a typical example of the top
sampling of an open railroad car. For a more detailed
discussion and other procedures, please refer to the general
literature.
TO. 19 Location o« Swnptin? Point* from tfw fiptwsrt Surtaw
otttwCtr
3132 Superimpose an imaginary grid a
and take samples at the intersections (fig. 19), P"*?"*? oy
auger or thief if practical, or by digging a series f J^to (pick
and shoveU below the surface of the material before any
portion of the contents has been removed. _
3133 Collect and identify the individual increments.
3 1.4 Pattern Sampling of Bulk Matenai: ^^
31 A A Pattern sampling was developed * J«*f* ™J™
sampling of material in bulk form. Tms method of samphng
takes into consideration the vanauon ot parade ******
composition around the loading point (3), for a particular
^Tl 4 2 The core locations of sampling patterns shall be as
follows: I and 2 within 15 in. (381 mm) of ^^^^
4. 5, 6. midway between loading point and side or end. and
7. 8 9 and 10 within 18 in. (457 mm) rf«"
toward bottom center (fig. 20). The sampli
inserted vertically in all locations.
32. Preparation and Reduction of Sample
32,1 Appearance-Vtoul inspection
recommended to determine if the material conMiiis gross
contamination, or if it is equal to the standard. It may show
if the material has picked up «^^m
laboratory processing is required to reduo >
more uniform panicle size. Unusual
determine if the contaminant should be
the sample through an appropriate screen.
TO. 20 PMMI* ilMMtt
Page 123
-------
32J Griname—Coaae or nonumiorm samples may re-
quire grinding in a monar. a milL or other suitaole mecnan-
icai devices to obtain a more uniform sample. The enure
sample may be subjected to grinding; or it may be more
efficient to screen off the oversize, grind it. and then blend all
portions together.
32.4 Minimum Sample Size—Where analysis of a com-
posite sample is specined or permitted, individual sample
increments are combined and reduced. Many gross samples
are unsuitable for laboratory handling or analysis, because
they may be too heterogeneous or too large tor the analyst to
obtain good representation with his small sample. For every
bulk solid, with its particular size distribuuon. there is a
minimum amount of material which must be taken in the
sampling operation in order for the sample to adequately
represent the solid. This minimum quantity is called the
minimum sample size, and the goal of any sample prepara-
tion is to make this minimum sample size at least as small as
the smallest quantity that wiU be taken for any single
analysis. Although a thorough discussion of minimum
sample size is beyond the scope of this practice, an excellent
presentation by Benedem-Pichier may be found in Ref. (6).
32.5 Sample Preparation Scheme—In general, a samme
preparation scheme will consist of panicle size reouction.
blending, splitting, and a repeat of this senes of operations
until the desired minimum sample size is attained. It is
difficult to write a general scheme for the reduction of a
sample for all types of material because of the nature of the
material and the purpose of the sample. It is important.
however, that any splitting operation be immediately pre-
ceded by blending. Two standard operations are given by the
following procedures:
32J.1 Cone-and-Quaner Method—Dump the individual
increments onto a dean canvas, and shovel into a pile.
placing each shovelful on top of the pile. Flatten the apex of
the cone with a shovel or a board until it is about one fourth
its original height Divide the pile into four equal pans bv
drawing a board twice through the center of the pile, making
right-angle cuts. Discard the opposite quarters, chosen at
random, and combine the "*""ning quarters into a cone*
shaped pile. Repeat the above operation until the desired
quantity of sample is obtained. Place the final sample on a
clean canvas, and mix by alternately raising opposite comers
of the canvas.
32J.2 Sample Splitter or Riffle:
32J.2.1 The sample splitter or riffle (Fig. 21) should be
FKL 21
E300
constructed of a matenai suitable for use with material unaer
test It consists of a senes of chutes that are directed
alternately to opposite sides. The slot width should be at least
three or more times the diameter of the largest parade to be
passed through to prevent bridging and. therefore, biased
splitting.
32.5.2.2 Pass the composite or gross sample through the
riffle to divide it into two approximately equal portions. Pass
one of these pontons, selected at random again through the
riffle. Continue this operation until the sample size is
reduced to either that required or the minimum sample size
beyond which additional grinding is necessary.
32.6 Blending—Sample homogeneity must be assured by
thorough blending prior to analysis. This operauon should
be performed on ail samples in such a way that they will not
be changed because of light sensitivity, nygroscopidty, etc.
No sample container should be tilled more than approxi-
mately half full, and no container should be opened for
sample removal until it has been tumbled on a mechanical
blenaer designed for the purpose or rotated by hand, end
over end. at least 25 revolutions. On any sample, blending
snouid be oone after screening or grinding, or both.
33. Final Laboratory Sample and Storage Precautions
33.1 At least four times as much reduced sample should
be prepared as is required for one laboratory to perform a
complete analysis. Retain one portion of the well-blended
sample for the manutaciurer or seller, one for the purchaser.
one for the umpire, if necessary, and one reserve to replace
breakage or loss.
332 Samples that are to be stored over long periods, that
may be affected by atmospheric exposure, or that may
become seriously contaminated in contact with paper or
cardboard should be packaged in widemouth. home-canning
type mason jars having two-piece, metal caps. Best results are
obtained if the sample is compatible, by vacuum sealing such
bottles (5). Widemouth. screwcapped glass jars with caps and
liners of suitable inen material are generally satisfactory.
33 J For materials in which water content is important or
composition is subject to change upon atmospheric expo-
sure, plastic containers are generally unsuitable because of
their permeability. In other cases, tight leakproof paper
sample envelopes or cardboard canons with or without
plastic liners or coatings, or even tin cans, may be used to
hold samples.
33.4 Where corrosion or atmospheric exposure cause
problems it is usually better to use widemouth glass jars with
suitable screw caps and liners (see 332).
34. Labeling Sample Containers
34.1 Label the container immediately after a sample is
obtained. Use waterproof and oil-proof ink or penal hard
enough to dent the tag, since soft pencil and ordinary ink
markings are subject to obliteration from moisture, oils
smearing, and handling. If gummed labels are used, they
should be further secured with transparent sealing tape.
Sufficient detail should be written on the label to complete y
identify the sample. The following informauon is frequently
desired:
34.1.1 Date and time.
34.1.2 Name of supplier.
Page 124
-------
34.1.3 Name or number and owner of the vessel, car. or
container.
34.1.4 Brand name, grade of material, and code number.
34.1.5 Reference symbol and necessary identification
number.
SLURRY SAMPLING
35. Scope
• 35.1 This practice describes equipment and procedures
for sampling materials which are slumes at the time of
sampling. A slurry is considered to be a suspension of solid
particles in a liquid which can be separated by filtration or
sedimentation (does not include emulsions). The equipment
and procedures that are described in this practice are
intended to supplement the experience of the sampler and to
serve as a guide in selecting methods that are applicable to
the material being sampled.
36. General Principles and Precautions
36.1 Quite often the value or quality of matenai being
tested in the sample is related to panicle size. When tnis is
the case, any segregation of the particles tends to arfect tnese
values. Liquids that carry solid particles must have a certain
velocity to keep the solids in suspension. To overcome the
problem of segregation of materials by size or weight requires
application of certain fundamentals of good sampling prac-
tice. The slurry should be stirred rapidly before sampling to
assure uniform distribution of the solids.
362 At the time the sample is taken, all panicles should
be uniformly distributed throughout the liquid carrier. This
will help to obtain a uniform sample.
36 J The sampling of slurries with any degree of accuracy
is quite difficult. This is particularly true when sampling a
normally static system such as storage tank or vat. Arrange-
ment must be made to agitate thoroughly the content of such
storage units prior to sampling. The most desirable and
convenient place to sample a slurry is from a pipeline as the
material moves through the line. Even here it is difficult to
obtain an accurate sample, because slurries subjected to
shearing will tend to change in composition due to the loss of
the liquid. Fittings, bends, and other constructions in the line
will tend to create nonuniformity in solids content. Lines
that are smaller than 1 in. in diameter are usually not
suitable for handling slurries because of frequent plugging.
The use of a continuous running sample line provided with
an orifice to reduce slurry velocity seems quite satisfactory.
36.4 If only a portion of any slurry sample can be used for
analysis, the sample should be shaken and a portion dumped
out. Attempts to pour out a predetermined volume are
unsatisfactory because the solids have time to separate
during the pouring.
36.5 Slurry solids should be washed only with the filtrate,
unless it has been proven that the proposed wash liquid does
not dissolve out any fraction of the solids. Large errors can
be introduced by washing out soluble fractions of a slurry.
36.6 Sampling practice adhering to above techniques will
produce a reliable sample. The sample is accepted as
representing the entire stream at the time it was taken. The
more frequently the subsampies are taken, the more accu-
rately will the sample represent the total stream.
E300
37. Continuous Sampling
37.1 Sampie Cutter of a Slurry Stream—Continuous
samples are taken at various locations in the plant by a
properly designed sample cutter. The opening in the cutter
should be sufficiently large that collision of panicles will not
restrict their entrance into the cutter. The cutter should hold
all of the sample without overflowing, and should move
completely through the stream at a uniform speed.
37.2 Stationary Sampling Probe. Horizontal Pipe:
37.2.1 A continuous sample may also be taken in pipes by
a stationary sampling probe which should be located at 20
pipe diameters (PD) and preferably 40 PD or more down-
stream from any elbow, valve, or other fitting.
37.22 The probe opening should be placed at the center
of the cross-section of the pipe and pointed precisely
upstream.
37.2.3 The sample should be withdrawn at a rate such
that the velocity of flow (feet per second) through the probes
opening is equal to the centeriine velocity (isokinetic).
However, for practical purposes, the sample can be with-
drawn at 1.2 multiplied by the average velocity of flow.
37.2.4 The average concentration in the pipe is calculated
by dividing the composition of the sample by a value V
(determined from Fig. 22X
37.2.5 Openings flush with the pipe wall elbow wall (Fig.
23) or pump wall do not yield reproducible results for
systems that are difficult to suspend. Such systems are those
whose settling ratios. £ are above 1.0 (S is the ratio of
bottom to top concentration in a settling device). For
systems whose settling ratios. S are below 1.0, and whose
concentration gradient, -m is less than 0.1. a side-wall tap
will give satisfactory results (8).
372.6 Use of a circular port probe under the conditions
described in the preceding paragraphs, (see 372,1 through
372.4) will result in samples whose reproducible average will
be within 8 % of a stream composition for a wide variety of
systems and within 2 % for a large majority of suspensions
likdy to be encountered in petroleum operations (8).
37.3 Sampling in a vertical pipe, upward flow, pipe
precisely vertical.
37J.I The sample probe opening must be pointed down-
ward, precisely vertical, and at least 3 PD above any elbow or
fitting.
37J2 The probe opening should be placed at the center
of the pipe crass-section.
cul
04 Q» OS 10
OONCtNTNMION CW4OSNT,-m
Sawiilt tfirttat br »'. wall •vtrt^t iiokiMtic «am-
« ct«Mr wiir. isakiaetM v«toci»r u HIM •» i-
attract vekaaty.
FKJ. 22 A«
i Sampto Conowmtion
Page 125
-------
E300
POUT
CLBOW
TAP IN EL8CW
FKL 23 Sampling Probes and Taps
1
37.3.3 The sample should be withdrawn at a rate such
that the velocity of How through the probe opening is equal
to the centeriine velocity of the flowing stream. It is
satisfactory to calculate centeriine velocity as 1.2 multiplied
by the average velocity of How. '
37.3.4 Use of a circular probe under the conditions
described m 37.3.1 through 37.3.3 will result in samples that
will equal the average composition within ±0.05 absolute
volume percent It is not necessary to use an adjustment
factor as is the case for the condition described in 37 •> 4 lor
the horizontal pipe (8).
37.3.5 The withdrawal probe, used as recommended, will
give a sample that can be accurately related to the average
composition that flows through the pipe, and deviations in
position and withdrawal velocity will result in a change of
sample composition. Such changes are primarily the result of
the settling rate of the dispersed phase, the rate of withdrawal
of the sample, and the rate of flow in the pipe
37.4 In order to reduce the volume of this continuous
primary sample so that the amount of material is correct for
the analysis, the primary sample is reduced in size by an
automatic sampler to provide an increment sample This
increment is discharged into a small agitator to keep the
solids in suspension until the sample is analyzed (7).
NOTE 17—The isokinetic sampling (where the linear veiocitv
through the opening of the sampling probe is equal to the linear velocity
in the pipe in front of the opening) is recommended for sampling in both
vertical and horizontal pipes. Nonuotaneuc sampling can bedone wnh
equally accurate results, but a knowledge of concemnnon gradient
(which is a function of sealing vetacity, pipe sire, and rate of ftowTb
necessary so that the rano between ^^ composition and average
pipeline composition can be determined. To this end. a method has
been found and partially developed whereby , sealing ratio can be
determinedIby a static test, and this in turn related to the distribution of
solids, or the concentration gradient, from top to bottom of horanul
pipe cross-section.
38. Sampling of a Slurry in Tanks, Tank Cars. Drums, and
Other Storage Containers
38.1 .\fixing—Mix the full tank containing the siurrv for
1 -h by using mechanical agitation with a four-arm and a
rake-bottom sweep agitator moving at 4 rpm. so that the
slurry is thoroughly uniform. Stop the mixing and take the
samples immediately while the contents of the tank are still
in motion.
38.2 Sampling:
38.2.1 The sampling bottle is a 32-oz (1000-mL) weighted
bottle attached to a chain and a stopper with an attached
chain (see Fig. 24).
FIG. 24 Missouri and Indians Weighted Restricted-fin Fluid
Fertilizer Sampling Bottles Designed to Fill While Being Lowered
(and Raised) In Storage Tanks
38.22 Sample bottles are three 32-oz (1000-mL) wide-
mouth polyethylene jars with proper seals, labeled 1.2, and
3. *
382.3 Fill each numbered 32-oz (1000-mL) sample bottle
one third full from the sampling jar (see Fig. 24) in the
numerical order of 1.2, and 3.
38.2.4 Refill the sampling jar and fill the sample bottles
about two thirds full in the numerical order of 2. 3. and 1.
38.2.5 Fill the sampling jar and fill the sample bottle
within about 2 in. (51 mm) of the top in the numerical order
of 3, I,and2.
382.6 Determine the temperature of the slurry to the
closest 1*C during the final sampling step.
382.7 Take the three sample bottles to the laboratory,
wash them, wipe the lips and seals clean and dry, and reseai.
382.8 Seal the sample kept as a retainer with a new cap,
and then further seal by a plastic wrap, covering the entire
cover.
382.9 Drop the weighted stoppered sampling bottle into
the slurry to a depth well under the surface and well away
from the side of the tank near the center point if practical.
Pull the stopper, allow the bottle to fill, and then pull to the
surface. Initially, run other location checks to determine
uniformity of the slurry.
39. Settling Rate Test (8)
39.1 The settling rate test will distinguish between those
suspensions which are easy or difficult to suspend and.
consequently, those which are easy or difficult to sample. A
suspension can be withdrawn from a flowing stream and
used in the test Low values of the settling ratio means that
the suspension is insensitive to the method and rate of
sampling, whereas high values show that the recommenda-
tions given above must be adhered to.
Page 126
-------
40. Consistency of Slurry Suspensions «9)
40.1 Scope—The method is applicable to aqueous slurry
suspensions containing 0 to 1. 1 to 4. 4 to 15. and 15 to 25?£
of" dry- solids. Slurry consistency of more proper "concen-
tration'* is defined as the weight of oven-dry solids in 100 g or'
the slurry.
40.2 Apparatus:
40.2.1 Sampling Cup of about 200-mL capacity with a
height approximately equal to its diameter and with a
smooth lip. if the slurry to be sampled is to be taken from a
source where it is being well mixed, it is preferable to use a
larger-sampling cup or jug having a capacity of about 1 L.
40.2.2 Beakers. 600 to 1500-mL. tared to the nearest 0.1
g-
40.2.3 Containers—* 10-L (3-gal) bucket and a 40-L
container, both tared.
40.2.4 Mixing Device, for the 10 and 40-L containers.
preferably a portable electric stirrer.
40.2.5 Balances. 40-kg capacity, accurate to 50 g. a 2-ke
capacity accurate to 0.1 g.
40.2.6 Buchner Funnel and Flask. 150-mm.
40.2.7 Filler Paper. 150-mm diameter, coarse texture.
40.2.8 Drier or Steam Cylinder, with wire mesn co%'er.
large enough to accommodate 15-cm filter papers, controlled
within a range from 110 to 150*C (230 to 302T).
40.2.9 Laboratory Drying Oven with Balance, oven main-
tained at 105 ± 3*C the balance having a capacity of at least
100 % and accurate to 0.01 g.
40.3 Sampling—'Take the sample at the point of where
the sample is uniform and in such a manner as to be
representative of the solid water mixture.
40.4 Procedure:
40.4.1 .\fbaures Containing Less Titan I % of Solids:
40.4.1.1 Use the sampling cup to withdraw five represen-
tative portions of approximately 100 g each. Each time.
deposit the entire contents in the tared 600-mL beaker.
Carefully dry the outside of the beaker and weigh it and its
contents to the nearest 0.1 g to determine the net weight of
the specimen.
40.4.1.2 Place a tared filter paper in the Buchner funnel
moisten with water, then apply suction to the flask and filter
the slurry. Remove the resulting pad and filter paper and
heat on the dryer until it ceases to steam. Place the paper and
pad on the weighing pan of the laboratory oven balance and
make successive measurements after additional drying until
a constant weight is obtained. Weigh to the nearest 0.01 g.
40.4.1.3 The percentage consistency concentration of the
sample is then:
Kw-.'VfJx 100
E300
where:
iv » weight of the moisture-free mat and filter paper, g,
/ « weight of the moisture-free filter paper, g, and
? — net weight of the original sample in the 600-mL beaker.
g-
NOTE 1 8— Alter removing filter paper and pad from Buchner tunnel.
be sure all solids are wiped clean from the inside surface ot the funnel
and deposited onto the pad. This can be done with the linger.
NOTE 19 — Drying can be sped up if the pad is pressed between
blotters in a hydraulic press before drying.
NOTE 20 — if the pad tends to suck to the cylinder, place the pad
between dry blotters. The surface of the cylinder may be treated with a
silicone spray or TFE-fluorocarbon to prevent sucking.
40.4.2 Mixtures Containing I to 4% Solids:
40.4.2.1 Use the sampling cup to withdraw ten consecu-
tive representative ponions of solids slurry: till the cup each
time and empty the entire contents into the tared 1500-mL
beaker. Weigh the contents to the nearest 0.5 g and deter-
mine the weight of the specimen.
40.4.2.2 Deposit the specimen into the tared 10-L bucket
and dilute to 0.5 % consistency concentration or less, using
some of the water to rinse all the solids from the beaker.
Weigh and determine the net weight of the contents to the
nearest 10 g.
40.4.2.3 Determine the percentage consistency concentra-
tion of the stock in the bucket as in 40.4.2. stirring the stock
vigorously with the sampling cup before withdrawing a
portion. The percentage consistency of the original sample is
then
where:
p « percentage consistency concentration of the diluted
stock.
W - net weight of the contents of the bucket, g and
w «* weight of specimen, g.
40.4.3 Mixtures Containing 4 to 15% Solids:
40.4.3.1 With the sampling cup. withdraw ten consecutive
ponions of solids slurry, filling the cup approximately half
full each time and emptying the contents into the tared
1500-mL beaker. Weigh the beaker and its contests to the
nearest 0.5 g and determine the weight of the specimen.
40.4.32 Deposit the specimen into the tared 40-L con-
tainer and dilute to less than 0.5 % consistency using some of
the water to rinse all the solids into the beaker. Insert and
adjust the electric mixer for thorough agitation of the
suspension.
40.4.3.3 Proceed as in 40.4.2.
40.4.4 For mixtures containing 15 to 25 •& solids, proceed
as in 40.4.3. except instead of ten. take five representative
ponions of the original stock of approximately 100 g each.
REFERENCES
(1) Duncan. A. J. "Bulk Sampling: Problems and Lines of Attack -
Technometncs. TCMTA. Vol 4. No. 3. August. 1962. p 319
(2) Bicking. C. A. 'The Sampling of Bulk Materials.- Maenais
Researcn and Standards. MTRSA. Vol 7. No. 3. March. 1967. p.
(3) Gehrke. C. W.. Baker. W. L- Knuse. G. F. and Russell. C H
-Sampung of Bulk Feralizen.- Journal ot the Assoaauon ot
Official Anatmcai Chemists. JANCA. Vol 50. Apnl. 1967.
(4) Available from Automated Sampling Systems. P. 0. Box 2706. Des
Moines.lA503l5.
(5) Jones. R. J.. Ed- "Selected Measurement Methods for Plutonium
and Uranium in the Nuclear Fuel Cycle." Division of Technical
Information. U. S. Atomic Energy Commission. 1963. pp. 49-50.
(6) Bed. W. G~ Ed- Phvsieal Methods in Chemical Analvns. Aca-
demic Press, inc. New York. 1956. Vol 3. pp. 183-201.
Page 127
-------
(7) Cook. P. E.. "Continuous Sampling tor X-Rav Anaivsis." Denver
Equipment Co.. .Bulletin No. S1-B12) UOO 17th St.. Denver. CO
(8) Rushton. J H. and Hillestao. J. G.. "Sampling of Nonho-
mogeneous Row ,„ Ptpes.~ Papcr lor Presentauon to a ^^ on
General Engineenng durmg the 29th Midvear Meeung of the
E300
American Petroleum institute Division of Refining in Chase Park
Plaza Hotel. St. Louis. MO. Mav 13. 1964.
(9) "Consistency i Concentration i of Pulp Suspensions." T-240. SU-67.
TAPPI Standard 360. Technical Association of Pulp and Paper
Industry. One Dunwood Park. Atlanta. GA 30341.
(10) Penbennv. Division of Houdaille Industries. Inc.. P. 0. Box 112.
Prophetstown. IL61277.
ANNEX
(Mandatory Information)
Al. DETERMINATION OF THE BASIC VARIANCES FROM AN INITIAL PILOT STUDY OF THE
MANUFACTURING PROCESS
Al.I Introduction
Al.1.1 The procedures of this practice assume that the
manufacturing process turns out matenal in batches and that
±rn«VrT WIthinbatChKand««"«" batc"«"t also
assumes that the vanauon within batches and the variation
between batches are both random. Funhermore. it assumes
that the variance ot the within-batch vanauon ,s the same tor
all batches and the variance of the between-batch vananon°
Constant over nme. A preliminary step, therefore ° to
determine whether these hypotheses are acceptable for a
liven process. If they are accepted, then the next step is to
*timate the withm-batch and between-batch variant If
he hypotheses are rejected, the next step is either (IT to
nake engmeenng changes in the process that will lead to the
ZEH ft*?**-:«" or <*> ^ find a more *pm£!
ated model for descnbing the process.
il.2 Determination of Randomness of Within-Batch and
Between-Batch Variation
A 1.2.1 Step /-Take 2 "increments of material" at ran-
om tor each of 25 consecutive batches produced bvtoe
iven process. If the material comes in packaeed form 'the™
icrements ot matenal could be 2 randomly selected' padf-
8r' t ? ,C,°meS '? ^k forau lhe 2 Cerements could be ?
shovel fuUs" or the like. For packaged material this is piled
i order, the 2 packages should be selected by the use of
mdom numbers. Matenal that comes in bulk is probablv
sst sampled as it is moved, on a conveyor belt or otherwise^
3d the increments should then be selected at random
itervals of time, random numbers again being employed In
•e latter case, the precise nature of the sampling instrument
hovel, etc.) must be specified and also the size of me
crement it selects.
A 1.2.2 Step .'—Prepare for laboratory testing each of the
) mcrcmems taken as called for in Step I and make a smde
easuremem on each under as uniform conditions "as
)ssible (same laboratory, same analyst, same day. if pos-
A 1.2.3 Step ^-Prepare 3 control charts for the 50 mea-
rements called for above. These should be:
(J) A range chan for the 2 increments from each batch.
(•?) A chan of the means of the 2 tests made on each batch.
' See anv standard book on control clum. tor eumwt. Duiuan. A J Quo/uv
•uro, ana Sndunruu Siatuuu. or Grant £. 1_ SuuutKOj Qualuv Conmi
(3) A moving range chart of the batch means.
A 1.2.4 Step 4—If there is no point above the upper limit
on Chan i /) and no run of 7 or more points above or below
the center line or other evidence of nonrandom variation.
accent the hypothesis that the basic within-batch variability
is random with the same variance.
A 1.2.5 Step 5—If on Chan (2) there is no trend, no runs
of 7 or longer above the average or no other evidence of
nonrandom variation and also if on Chan (3) there is no
point above the upper control limit or no run of 7 or more
above the center line or other evidence of nonrandom
variation, then accept the hypothesis that the variation from
batch to batch is random with constant variance.
A 1.2.6 If the process passes all of the above tests without
exception, the data can be used without modification to
measure the basic variances as described in A 1.3. If any
exceptions occur, a statistician should be consulted as to
what information about basic variances may be obtained
from the data and just how this should be done. In fact it
might be helpful even prior to this to have the advice of a
statistician in interpreting the control charts, in Steps 3 and
4.
Al J Determination of Basic Material Variances
A U.I Given that the process meets the randomness
requirements of Section A 1.2. a "components of variance
analysis" should be performed to estimate the within-batch
variance and the between-batch variance. The procedure is
as follows:
A 1.3.1.1 Step /—Compute
j^-^r/^-.i-^^ffi-n (5)
sh- = 2 ZfX, - I' )2/(25 - I) (6)
where:
A", is the mean of the test made on the 2 increments from
the i'* batch.
and
•V is the mean of the whole (25) (2) = 50 tests.
A 1.3.1.2 Step .'—Take ov = S,' - ff,: as an unbiased
estimate of the within-batch variance where
-------
A1.4 Determination of Basic Variances of Reduction and
Analysts
A 1.4.1 Step 1—Take 20 increments from each of 5
batches. From physical composites of the 1st 2nd. 3rd
20th increments from each of the 5 batches making a total of
20 composites. Reduce each composite to the size of a
laboratory sample by whatever method of reduction is
standard for the given material and prepare each of the 20
laboratory samples for testing. Run two tests on the labora-
tory sample from each composite.
A 1.4.2 Step 2—Construct a control chart on the differ-
ences between the two tests for each of the 20 composites.
E300
This is to check on the uniformity of the testing procedure.
Call the chan Control Chan (4).
A 1.4.3 Step j — Construct a moving range chan on the
means of the two tests for each of the 20 composites to check
on the uniformity in the reduction procedure.
A 1.4.4 Step 4 — If the hypothesis of uniformity is satisfied
in each case, compute
V - 2 I^ - X )2/(20 - I) (8)
A 1.4.5 Step j— Take at~ a y,~ as an unbiased estimate of
the testing variance.
A 1.4.6 Step 6 — Take a? » ((sr- - v)/2] as an unbiased
estimate of the variance of reduction.
« ««, n^. M ^^,^ wm
Page 129
-------
Excerpts from Standard Practice for
* *
Units
-------
E 380
TABLE 1 BSM SI Units
Quantity1 um
lengtn mew
mass " keagrem
time secona
ewcme current ampere
mermooynsme temperature-4 ketwn
amount of suDstsncs mow
lurranous mensity csnoeis
Symbol
m
*5
s
A
K
, moi
cd
TABLE 3 Dewed SI Units wrtn Soectsl Nsmss
- For • Discussion at Celsus tsmoerMure «ee 3.4.2.
Ousnmy3
TABLE 2 Supptemsnisry SI Units
Urw
Symdot
puns snow
sou snow
rad
sr
sioniess derived quantities. Therefore, the supplementary
units radian and steradian are to be regarded as dimension-
less derived units which may be used or omitted in the
expressions for derived units.
2.4 Derived Units:
2.4.1 Derived units are formed by combining base units.
supplementary units, and other derived units according to
the algebraic relations linking the corresponding quantities.
The symbols for derived units are obtained by means of the
mathematical signs for multiplication, division, and use of
exponents. For example, the SI unit for velocity is the metre
per second (m/s or m«s~'), and that for angular velocity is
the radian per second (rad/s or rad •$"').
2.4.2 Those derived SI units which have special names
and symbols approved by the CGPM are listed in Table 3.
2.4.3 It is frequently advantageous to express derived
units in terms of other derived units with special names: for
example, the SI unit for electric dipole moment is usually
expressed as C-rn instead of A-s-ra.
2.4.4 Some common derived units are listed in Table 4
2J 57 Prefixes (see 32 for application):
2J.1 The prefixes and symbols listed in Table 5 are used
to form names and symbols of the decimal multiples and
submultiples of the SI units except for kilogram.
2J.2 Unit of Mass—Among the base and derived units of
SL the unit of mass (kilogram) is the only one whose name.
for historical reasons, contains a prefix. Names of rf^T"ai
multiples and submultiples of the unit of mass are formed by
attaching prefixes to the word gram (g).
2JJ These prefixes or their symbols are directly attached
to names or symbols of units, forming multiples and
submultiples of the units. In strict terms these must be called
-multiples and submultiples of SI units." particularly in
discussing the coherence of the system (see Section 5). In
common parlance, the base units and derived units, along
with their multiples and submuitiples. are all called SI units.
3. Application of the Metric System
3.1 General—SI is the form of the metric system that is
preferred for all apptications. It is important that this
modernized form of the metric system be thoroughly under-
i "Qutnnv" as used in the Headings of the tables of this nandmi
mesMiaMe attribute of phenomena or miner.
Quantity*
urn
Symtxx
Formula
frequency tiumnenoe
activity (of a raoonucKMI
aotoneo oose'
3ose eouvaiem
lumen
lux
grsy
HZ
N
Pa
j
w
c
F
a
s
wo
T
H
•C
Ix
Bq
Gy
Sv
N/rrr1
N-m
J/8
A-t
W/A
CIV
V/A
A/V
V-S
Wo/A
K(SIS 3.4.21
cover
JflV
A tnckMon n me taow of denvea SI ants wnn toecw names sopreveo oy me
CIPM * 1976. _
• Related dummies tang me tuns unt are: spscftc enemy mpsneo. Mima.
ano aosofDeo one noex.
stood and properly applied. Obsolete metric units and
practices are widespread, particularly in those countries that
long ago adopted the metric system, and much usage is
improper. This section gives guidance concerning the limited
number of cases in which units outside SI are appropriately
used, and makes recommendations concerning usage and
style.
32 Application of SI Prefixes:
32.1 General— In general the SI prefixes <2J) should be
used to indicate orders of magnitude, thus eliminating
nonsignificant digits and leading zeros in decimal fractions.
and providing a convenient alternative to the powers-of-ten
notation preferred in computation. For example:
12 300 mm becomes 12J m
12J x I01 m becomes 12J km
0.00123 uA becomes 1.23 nA
322 Selection— When expressing a quantity by a numer-
ical value and a unit, a prefix should preferably be chosen so
that the numerical value lies between 0.1 and 1000. To
variety, it is recommended that prefixes repre-
,
senting 1000 raised to an integral power be used. However.
three factors may justify deviation from the above:
322. 1 In expressing area and volume, the prefixes hecttx
deka-, deck and centi- may be required, for example, square
hectometre, cubic centimetre.
3.L2.2 In tables of values of the same quannty. or in a
discussion of such values within a given context, it is
generally preferable to use the same unit multiple through-
out
3,2.2.3 For certain quantities in particular applications.
one particular multiple is customarily used. For example, the
millimetre is used for linear dimensions in mechanical
engineering drawings even when the values lie far outside the
range 0. 1 to 1000 mm: the centimetre is often used for body
Page 131
-------
E380
TABLE 4
aoscfOed oosa rate
angular vatocffy
concarttraoon (o* amount ot suosuncei
etacmc cnarpa danativ
. • etacmc fietd ttranQin
eMctnc tux oanarty
energy oentny
exposure (X and gamma rays)
rwucspacny
Some Common Derived Units o! SI
Unt
gray per secono
metre oer second souarea
rattan per seoora souareo
radian per second
squire metre
ampere per souare metre
KrioQiajn per f^ttm metre
couono per QJOC metre
vott per ^DAU^
couiomo per souare nieue
joue per cuac metre
fouMperketvm
couiomo per MoQiaiTi
lOuMperKemn
Symoa
Gy/s
m/*2
runs1
m*
A/m2
kg/nr1
C/nr1
V/m
C/m»
J/m*
J/K
C/kg
J/K
hestfluxaensffy
magnetic flea strenqtn
«n«if energy
motartmrooy
moment of forea*4
permeaoj«v (msgneoci
watt per souare metre
candeta per square metre
ampere per metre
lOuMpermow
joule per mow keivn
loue per mwe netvm
newton metre
w/m*
cd/m2
A/ht
J/mtt
J/fmoi-K)
PMMTCMIMy
raaant wtensrtv
speofc energy
specific entropy
sosoflc volume
surface tenenn
tnniM emucmty
tarao per metre
wan per souare metre
watt per souare metre steratsan
wan per steraoan
joule per Mogram MMVI
loule per kdogram
joule per tutoqram kewn
cuttc metre per kilogram
newion per meve
wejt per meve kehnn
N-
Hfi
F/m
W/»f
J/W'K)
J/XO
JAkg-K)
m/i
Pi-i
maw par aaoond
maw
SM 3.4.4.
measurements and clothing sizes.
3.2J Prefixes in Compound Units*— It is recommended
that only one prefix be used in forming a multiple of a
compound unit. Normally the prefix should be attached to a
unit in the numerator. One exception to this is when the
kilogram occurs in the denominator.
Examples:
V/m, not mV/mm. and MJ/kg, not kJ/g
3.2.4 Compound Prefixes— Compound prefixes, formed
by the juxtaposition of two or more SI prefixes are not to be
used. For example, use
I nm. not I mum
1 pF, not I
m*
i/m
If values are required outside the range covered bv the
prefixes, they should be expressed by using powers of ten
applied to the base unit.
3 .2.5 Powers 0/t/mu— An exponent attached to a svmbol
containing a prefix indicates that the multiple or submultipie
of the unit (the unit with its prefix) is raised to the power
expressed by the exponent. For example:
• * compound unit is a derived unit that is expnssea in terms ol two or more
units ntner wan bv a single special name.
Page 132
I cm3 -(lO-'mr1 -l
-------
380
TABLE 5 S! Prettm
a.^. ^i.^ . C^tfM*^
•wumuBjcmon rvcwr
J 000 000 000 000 000 000 - 10'*
1 000 000 000 000 000 - 10**
1 000 000 000 000 - 10"
1 000 000 000 • 10*
1 000 000 - 10*
1 000- 10*
100- 10*
10 - »0'
0.1 - 10"
0.01 •
0.001 •
0.000 001 •
0.000 000 001 •
0.000 000 000 001 •
0.000 000 000 000 001 •
0.000 000 000 000 000 001 >
• 10~*
• 10~*
• 10~*
« io-»
. io-«*
. 10-'*
. io-««
P^tjroi SytnOQ
me
C
MM P
HM ^
ttrs T
QXtt G
jfTa» *•
maca M
Mo k
hacto* n
oana* da
dacr* a
caMr* c
ma) m
fflRSiO }i
ntno n
pno p
fjJH.Ul f
IVTHIO i
ano a
the minute and second is discouraged except for special fields
such as cartography.
3.3.2.3 Area—The SI unit of area is the square metre
(m-). The hectare (ha) is a special name for square
hectometre (hnri. Large land or water areas are generauv
expressed in hectares or in square kilometres ikm-i.
3.3.2.4 Volume—The SI unit of volume is the cubic
metre. This unit, or one of the regularly formed multiples
such as the cubic centimetre, is preferred. The special name
litnr (L)6 has been approved for the cubic decimetre, but use
of this unit is restricted to volumetric capacity, dry measure.
and measure of fluids (both gases and liquids). No prefix
other than raiili- or micro- should be used with litre.
3J.2J Mass—The SI unit of mass is the kilogram. This
unit, or one of the multiples formed by attaching an SI prefix
to gram (g), is preferred for all applications. The megagram
(Mg) is the appropriate unit for measuring large masses such
as have been expressed in tons. However, the name ton has
been given to several large mass units that are widely used in
commerce and technology—the long ton of 2240 Ib. the
short ton of 2000 Ib. and metric ton of 1000 kg (also called
the tonne]. None of these terms are SI. The term memo ion
should be restricted to commercial usage, and no prefixes
should be used with it. Use of the term tonne is deprecated.
'Sea Appendix X 1.11.1.
•ThtCGPM in October 1979
Since tte letter symbol I can
symbol L is fBrmnmcnrted for USA
Land las i
with the i
bob far toe.
I. only the
TABLE 6
Untt
Unte in UM with si
wna
mnuia
hour
day
WMK* nionvi. etc*
h
a
i m
1 h
i d
mawta"
60 a
60 mn. 3600*
24 h - 66 400 s
WlMlng
o/wr
-------
E380
multiples and submuiupies of SI units are to be avoided
except for the Hire 13.3.2.4), metric ton (3.3.2.5), and hectare
(3J.2J). For exampie. do not use:
fenni
micron
millimicron
are .
(magnetic flux densnv)
•v (must
X ( volume t
mho . . .
candle
candlepower
fermi
millimicron
are
nmma
I -V
I A
mho
candle
I candleoower
—
—
-
m
^
m
—
a
^^ m-is _ •
um ** 1 0™* m
nm » I0~* m
dam* m ' 1 00 rn*
nT
,
ui* "" 1 mm*
s
cd
cd
3.3.4.4 Miscellaneous Units—Other non-Si units that are
deprecated include the following:
calorie
grade (t grade - «-r/200) rad)
kilogram-force
langiey(— I cai/cnr)
metric carat
tnetnc horsepower
millimetre of mercury
millimetre, centimetre, metre of water
standard atmospnere
(I aim- 101.325 kPa)
technical atmosphere
(t at- 08.0665 kPa)
torr
3.4 Other Recommendations Concerning Units:
3.4.1 .\fass. Force, and Weight:
3.4.1.1 The principal departure of SI from the gravimetric
system of metric engineering units is the use of explicitly
distinct units for mass and force. In SI. the name kilogram is
restricted to the unit pf mass, and the kilogram-force (from
which the suffix force was in practice often erroneously
dropped) should not be used. In its place the SI unit of force.
the newton. is used (see Fig. 1). Likewise, the newton rather
than the kilogram-force is used to form derived unia which
include force, for example, pressure or stress
-------
E380
the force that, if applied to the body, would give it an
acceleration equal to the local acceleration of free fail. The
adjective -local" in the phrase "local acceleration of free fall"
has usually meant a location on the surface of the earth: in
this context the -local acceleration of free .fall" has the
symbol ? (commonly referred to as "acceleration of gravity")
with observed values of g differing by over 0.2 ?5 at various
points on the earth's surface. The use offeree of gravity (mass
times acceleration of gravity) instead of weight with this
. meaning is recommended. Because of the dual use of the
term weight as a quantity, this term should be avoided in
technical practice except under circumstances in which its
meaning is completely clear. When the term is used, it is
important to know whether mass or force is intended and to
use SI units properly as described in 3.4.1.1, by using
kilograms for mass or newtons for force.
3.4.1.3 Gravity is involved in determining "ia« with a
balance or scale. When a standard mass is used to balance
the measured mass, the effects of gravity on the two masses
are equalized, but the effects of the buoyancy of air or other
fluid on the two masses are generally not equalized. When a
spring scale is used, the scale reading is directly related to the
force of gravity. Spring scales graduated in mass units may be
properly used if both the variation in acceleration of gravity
and the buoyancy corrections are cot significant in their use.
3.4.1.4 The use of the same name for units of force and
mass causes confusion. When the non-Si units are used, a
distinction should be made between force and moss, for
example. Ibf to denote force in gravimetric engineering units
and Ib for mass.
3.4.1.5 The term load means either mass or force, de-
pending on its use. A load that produces a vertically
downward force because of the influence of gravity acting on
a mass may be expressed in mass units. Any other load is
expressed in force units.
3.4.2 Temperature—Toe SI unit of thermodynamic tem-
perature is the keivin (K.X and this unit is properly used for
expressing thermodynamic temperature and t
:_.___!. »KJ- : • * - -
Nommi
. • - ' -——»»~«»*«»»fc ouu icmpenxure
intervals. Wide use is also made of the degree Celsius CO.
which is the SI unit used for expressing Celsius temperature
and temperature intervals. The Celsius scale (formerly «.»~l
centigrade) is related directly to thermodynamic temperature
(kdvins) as follows:
The temperature interval one degree Celsius equals one keivin
exactly. Celsius temperature nr/(rad*s). In the solution of problems that involve
rotation, the use of radian in these units will retain all
advantages of ^'*"^"*»"nfll analysis.
3.4.4.4 The use of the unit N-ra for torque and bending
moment may result in confusion with the use of the unit
N-m for mechanical energy. If vectors were shown, the
distinrrinn hunuMn mwHa ntr>ai energy i"i^ torque would be
obvious since torque is the product of moment arm and
force perpendicular to the moment arm. white mechanical
energy is the product of force and displacement in the
direction of the force. It is important to recognize this
difference when N-m is used as the unit for torque. The
joule, which is a special name for the unit of energy, should
not be used for the torque unit N • m but may be used in the
torque unit J/rad when rotation is involved and work occurs.
3.4.5 Impact Energy Absorption—This quantity, often
incorrectly called impact rmmanee or impact strength, is
measured hi terms of the work required to break a standard
: the proper unit is joule.
3.4.6 Pressure and Vacuum—Gags pressure is absolute
pressure minus ambient pressure (usually atmospheric pres-
sure). Both gage pressure and absolute pressure are properly
expressed in pascals, using SI prefixes as appropriate. Abso-
lute pressure is never negative. Gage pressure is positive if
above ambient pressure and negative if below. Pressure
below ambient is often called vacuum: whenever the term
vacuum is applied to a numerical measure it should be made
dear whether negative gage pressure or absolute pressure is
meant. See 3 J J for methods of designating gage pressure
and absolute pressure.
3.4.7 Dimenstoniess Quantities:
Page 135
-------
147.1 The values of so-called dimenstoniess quantities.
™!??Ple relracnve «nd« and relative permeability, are
n2^ 1 PUrB """^ In th« cases the corresponding
SI unit is the ratio of the same two SI units and mav be
expressed by the number 1. y
n,i™ Tt^°S SUCfa ? P6"8"1- pam Per 'housand. and
pans per million may also be "spd.
Exniin Inrf ^ "*, memin8 must be unequivocal.
Expressions like -The mole fraction of CO, in the sample
was 1.2 parts per million" or -The mass fraction of TO, in
.the sample was 1.2 pare per million" are permissible, but
would not be permissible if the word "mole^Tthe first
expression or "mass" in the second expression were not
present.
3.5 Style and Usage-Case must be taken to use unit
symbols properly, and international agreement provides
uniform rules. Handling of unit names varies because of
language differences, but use of the rules included here w,il
improve communications in the United States.
3.5.1 Rules for Writing Unit Symbols-
3.5.1.1 Unit symbols should be printed in upright upe
regareUess of the type style used in the surrounding text.
3.5.1.2 Unit symbols are unaltered in the plural
3.5.1.3 Unit symbols are not followed by a period except
when used at the end of a sentence.
3.5.1.4 Letter unit symbols are written in lower-case (for
example, cd) unless the unit name has been derived from a
proper name, in which case the first letter of the symbol is
t^tmtn\{ffri (for raamni* U7 D.\ -n. ^T -/«•*«« u
—-.uTT ,- f^Sf* ' *•*• The exception is the
symbol for litre, L. Prefix symbols use either lower-case or
upper-case letters as shown in 2J.1. Symbols retain their
prescribed form regardless of the surrounding typography.
For symbols for use in systems with limited
E380
3.5.1.8 Symbols, not abbreviations, should be used for
units. For example, use -A" and not "amp" for ampere.
3J.2 Rules for Writing Names:
3.5.2.1 Spelled-out unit names are treated as common
nouns in English Thus, the first letter of a unit name is not
capitalized except at the beginning of a sentence or in
capitalized material such as a title.
3.5.2.2 Plurals are used when required by the rules of
English grammar and are normally formed regularly, for
example, henries for the plural of henry. The following
irregular plurals are recommended:
mte caracttseoL
refer to ANSI X3 JO or ANSI/IEEE 260. as applicable. The
symbols in ANSI X3 JO are intended for applications in the
field of information processing, where unambiguous trans-
mission of information between computers is required. The
symbols in ANSI/IEEE 260 are generally consistent with
±± r51 ^ "ir* intended for «£SJ2
between human bangs. The symbols for limited character
sets must never be used when the available character set
permits the use of the proper general-use symbols as given in
tois st&odflrcL
3J.1J When a quantity is expressed as a numerical value
and a unit symbol, a space should be left between them. For
example, use 35 mm, not 35mm, and 2.37 1m (for 2 37
lumens), not 2.371m.
Exception: No space is left between the numerical value
and the symbols for degree, minute, and second of plane
angle, and degree Celsius. For example, use 45* ">0*C
3.5.1.6 When a quantity expressed as a numbeTand a unit
is used in an adjectival sense, it is preferable to use a hyphen
instead of a space between the number and the unit name or
between the number and the symbol. Examples: A three-
metre pole. . . The length is 3 m. . . A 35-mm film. The
width is 35 mm. However, per 3.5.1.5 Exception, a 90*
angle ... an angle of 90*.
3.5.1.7 No space is used between the prefix and unit
symbols.
Page 136
Singular
lux
hertz
Plural
lux
hertz
Siemens
3.5.2.3 No space or hyphen is used between the prefix and
unit name. There are three cases where the final vowel in the
prefix is commonly omitted: megohm, kilohm, and hectare.
In all other cases where the unit name begins with a vowel
both voweis are retained and both are pronounced.
3.5.3 Units Formed by Multiplication and Division:
3.5.3.1 With unit names:
Product, use a space (preferred) or hyphen:
newton metre or newion-nietre
In the case of the wan hour the space may be omitted.
thus:
wutnour
Quotient, use the word per and not a solidus:
metre per second, not metre/second
Powers, use the modifier squared or cubed placed after the
unit name:
metre per second squared
In the case of area or volume, the modifier may be placed
before the unit name:
squire millimetre, cubic metre
This alternative is also allowed for derived units that include
area or volume:
wut per square metre
NOTE—To avoid ambiguity in complicated expressions, symbols are
preferred over words.
3 J J.2 With unit symbols:
Product, use a raised doc
N-m for newton metre
In the case of W- h. the dot may be omitted, thus:
Wh
An exception to this practice is made for computer print-
outs, automatic typewriter work, etc.. where the raised dot is
not possible, and a dot on the line may be used.
Quotient, use one of the following forms:
-i ™
m/scrm-s ' or —
In no case should more than one soiidus be used in the same
-------
expression unless parentheses are inserted to avoid ambi-
guity. For example, wme:
but not
J/(mol • K) or J • mor' K." or (J/moD/K.
J/moi/K.
3.5.3.3 Symbols and unit names should not be mixed in
the same expression. Write:
joules per kilogram or J/kg or J • kg-'
but not
joules/kilogram nor joules/kg nor joules-kg"
3.5.4 Numbers:
3.541 The recommended decimal marker is a dot on the
line. When writing numbers less than one. a zero should be
written before the ^^Trmi marker
15.4.2 Outside the United States, the comma is often
used as a decimal marker. In some applications, therefore
the common practice in the United States of using the
comma to separate digits into groups of three (as in 23 478)
may cause ambiguity. To avoid this potential source of
contusion, recommended international practice calls tor
separating the digits into groups of three, counting from the
decimal point toward the left and the right, and using a small
space to separate the groups. In numbers of four digits on
either side of the decimal point the space is usually not
necessary, except for uniformity in tables.
Examples:
2.141 596 73 722 7372 0.1335
Where this practice is followed, the space should be narrow
(approximately the width of the letter "i"), and the width of
the space should be constant even if. as is often the case in
printing, variable-width spacing is used between words.
Exceptions: In certain specialized applications, such as
engineering drawings and finatntaj statements, the practice
of using a space for a separator is not customary
3.5.4.3 Because billion means a thousand million (nrefix
giga) in the United States but a million million (prefix ww»
m most other countries, this term and others, such as trillion-
should be avoided in technical writing.
3.5.4.4 Use of M to indicate thousands, as in MCF for
thousands of cubic feet or in MCM for thousands of circular
mils, of MM to indicate millions, of C to «ti^TP hundreds.
etc_ is deprecated because of obvious conflicts with the SI
prefixes.
3.5.5 Attachment—Attachment of letters to a unit symbol
as a means of giving information about the nature of the
quantity under consideration is incorrect. Thus MWe for
-megawatts electrical (power)," Vac for "volts ac." and Ut
for "kilojoules thermal (energy)" are not acceptable. For this
reason, no attempt should be made to construct SI equiva
lents of the abbreviations "DM" and "psig," so often used to
distinguish between absolute and gage pressure. If the
context leaves any doubt as to which is meant, the word
pressure must be qualified appropriately. For example:
"... at a gage pressure of 13 kPa"
E380
Where space is limited, such as on gages, namepiates. graph
labels, and in table headings, it is permissible to use the unit
symbol followed by a space and the modifier m parentheses.
For example: V (ac) and V (dc): kPa (gage) and kPa
(absolute).
3.5.6 Pronunciation—Some recommended pronuncia-
tions in English are shown in Table 8.
4. Rules for Conversion and Rounding
4.1 General:
4.1.1 Conversion factors to change a value of a quantity
expressed in non-Si units to the corresponding value of that
quantity expressed in the International System of Units may
be exact or approximations adequate for the particular task.
The rules in this section are based on using either exact or
approximate factors such as those of seven-digit factors listed
in Appendix X3. In some cases the quantity is such that
factors with fewer digits are appropriate.
4.1.2 Conversion of quantities should be handled with
careful regard to the implied correspondence between the
accuracy of the data and the given number of digits, in ail
conversions, the number of significant digits retained should
be sucn that accuracy is neither sacrificed nor exaggerated.
(For guidance concerning qgnifieam digjB see 4.3.) For
example, a length of 125 ft converts exactly to 38.1 m. If.
however, the 125-ft length had been obtained by rounding to
the nearest 5 ft. the conversion should be given as 38 m: and
if it had been obtained by rounding to the nearest 25 ft the
conversion should be given as 40 m.
4.1.3 Proper conversion procedure is to multiply a value
by a conversion factor that is more accurate than is required:
the result is then rounded to the appropriate number of sig-
nificant digits. For example, to convert 3 feet 2*/i« inches to
metres: (3 x OJ048) + (2J625 x 0.0254) - 0.979 487 5 m.
which rounds to 0.979 m. Do not round either the conver-
sion factor or the quantity before performing the multiplica-
tion, as accuracy may be reduced. After the conversion, the
or
TABLE 8
IUSAT
.«* •(••*»MM)
.KTon
.IWCK'tO*
.«*'•<• a*
'on
. Inn' toa I ftii9 as
.as n anatomy
ItMS
"... at an absolute pressure of 13 kPa"
itstaanmy.
on M «m
Page 137
-------
£380
A dimension is expresses in inchesu.
The limits are .
Conversion of the two limiu into miUunetres
"
1.950 ±0.016
1.934 ana 1.966
49.1236 and 49.9364
49.12 and 49.94
49.13 and 49.93
Method A—The tolerance equals 0.032 in and
thus lies between 0.004 and 0.04 in (see Table
9). Rounamf these values to the nearest 0.01
mm. the values in millimetres to be emptoved
tor these two limns are
Method B—Rounding toward the interior 01' the
tolerance, millimetre values for these two
limits are
This reduces the tolerance to 0.80 instead of 0.82 mm given by Method A.
4:5.2 Special Method for Dimensions with Plus and Minus
Deviations—In order to avoid accumulation of rounding
errors, the two limits of size normally are converted sepa-
rately: thus, they must first be calculated if the dimension
consists of a basic size and two deviations. However (except
when Method B is specified) as an alternative, the basic size
may be convened to the nearest rounded value and each of
the deviations converted toward the interior of the tolerance.
This method, which sometimes makes conversion easier.
gives the same maximum guarantee of accuracy as Method
A. but usually results in smaller convened tolerances.
4.5.3 Special Methods for Limitation imposed by Accu-
racy oj Measurements—If the increment of rounding for tne
tolerances given in Table 9 is too small for the available
accuracy of measurement limits that are acceptable for
interchangeability must be determined separately for the
dimensions. For example, where accuracy of measurement is
limited to 0.001 mm, study shows that values convened
from 1.0000 ± 0.0005 in can be rounded to 25.413 and
2SJ87 mm instead of 25.4127 and 25.3873 mm with little
disadvantage, since neither of the two original limits is
exceeded by more than 12 % of the tolerance.
4.5.4 Positional Tolerance—IT the dimensioning consists
solely of a positional tolerance around a point defined by a
nontoleranced basic dimension, the basic dimension must be
converted to the nearest rounded value and the positional
variation (radius) separately convened by rounding down-
ward.
4.5.5 Toleranced Dimension Applied to a Nontoleranced
Position Dimension—If the toieranced dimension is located
in a plane, the position of which is given by nontoleranced
basic or gage dimension, such as when dimensioning certain
conical surfaces, proceed as follows:
(a) Round the convened reference gage arbitrarily, to the
nearest convenient value.
(b) Calculate exactly, in the convened unit of measure-
ment, new maximum and minimum limits of the specified
tolerance zone, in the new plane defined by the new basic
dimension.
(c) Round these limits in conformity with the rules in 4.4.
For example, a cone of taper 0.05 in/in has a diameter of
. 1.000 ± 0.002 inch in a reference plane located by the
nontoleranced dimension 0.9300 in. By virtue of the taper of
the cone, the limits of the tolerance zone depend on the
„ position of the reference plane. Consequently, if the dimen-
sion 0.9300 in « 23.6220 mm is rounded to 23.600 mm (that
is. a reduction of 0.022 mm), each of the two original limits.
when convened exactly into millimetres, must be corrected
by 0.022 x 0.05 = 0.0011 mm. in the appropriate sense.
before being rounded.
4.5.6 Consideration of Maximum and Minimum Material
Condition—The aoiiitv to assemoie mating pans depends on
a "go" condition at the maximum material limits of the
pans. The minimum material limits, which are determined
by the respective tolerances, are often not as critical from a
functional standpoint. Accordingly, it may be desirable to
employ a combination of Methods A and B in certain
conversions by using Method B for the maximum material
limits and Method A for the minimum material limns.
Alternatively, it may be desirable to round automatically the
convened minimum material limits outside the original
limits to provide greater tolerances for manufacturing.
4.5.7 While the technique described in 4.5 provides good
accuracy of conversion, it will often result in dimensions that
are impractical for actual production use. For conversions
intended for production, it is usually necessary to round to
fewer decimal places and apply design judgment to each
dimension to assure interchangeability.
4.6 Other Units:
4.6.1 Temperature—General guidance for convening tol-
erances from degrees Fahrenheit to kelvins or degrees Celsius
is given below:
Conversion 01 Temperature
Tolerance Rwiurernena
Tolerance.
T
2 (±11
4 (±2)
10 <±i>
20 (±101
30 (±13)
40 (±20)
50 (±22)
Tolerance.
Kor'C
I (±0.3)
2(±i)
6 (±3)
17 (±13)
22 (±11)
2S(±I4|
Normally, temperatures expressed in a whole number of
degrees Fahrenheit should be convened to the nearest 0.5
kdvin (or degree Celsius). As with other quantities, the
number of significant digits to retain will depend upon
implied accuracy of the original dimension, for example:
IOO±5T: implied accuracy estimated to be 2T.
37.7777 ± 2.7777*0 rounds to 38 ± 3*C
1000 ± SOT: implied accuracy estimated to be 20T.
537.7777 ± 27.T7TTC rounds to 340 ± 3QTC.
4.6.2 Pressure or Stress— As with other quantities, pres-
sure or stress values may be convened by the principle given
above. Values with an uncertainty of more than 2 7o may be
convened without rounding by approximate factors:
1 JW/lird pa) - 7 kN/m- - 7 IcPa
5. Terminology
5.1 To help ensure consistently reliable conversion and
rounding practices, a clear understanding of the related
nontechnical terms is a prerequisite.
TABLE 9 Rounding Tolvranen tnehe»
Orignti TotcranoQ,
finm*so*Roirang,
athmt imiran """
0.000 04
0.0004
0.004
0.04
0.4
0.000 4
0.004
04)4
0.4
0.0001
0.001
0.01
0.1
1
Page 140
-------
E380
5.2 Certain terms used in this standard are defined as
follows:
accuracy (as distinguished from precision)—the degree of
conformity of a measured or calculated value to some
recognized standard or specified value. This concept involves
the- •systematic error of an operation, which is seldom
negligible.
approximate vaine—a value that is nearly but not exactly
correct or accurate.
coherent system of units—a system of units of measure-
ment in which a small number of base units, defined as
dimensionaliy independent, are used to derive all other units
in the system by rules of multiplication and division with no
numerical factors other than unity (see Appendix XI.9).
deviation—variation from a specified dimension or design
requirement, usually defining upper and lower limits (see
also tolerance).
digit—one of the ten arabic numerals (0 to 9).
dimension—a geometric element in a ^y*ien, such as
length or angle, or the """g^nrdP of such a Quantity.
feature—an individual characteristic of a part, such as
screw-thread, taper, or slot.
figure (numerical)—an arithmetic value expressed by one
or more digits.
inch-pound units—units based upon the yard and the
pound commonly used in the United States of America and
defined by the National Institute of Standards and Tech-
nology. Note that units having the same names in other
countries may differ in magnitude.
nominal value—a value assigned for the purpose of con-
venient designation: existing in name only.
precision (as distinguished from accuracy)—the degree of
mutual agreement between individual measurements.
namely repeatability and reproducibility.
significant digit—any digit that is necessary to define a
value or quantity (see 4.3).
tolerance—the total amount by which a quantity is
allowed to vary; thus the tolerance is the algebraic difference
between the rnaximuro and ny"'"*0"1 limits.
Page 141
-------
E380
To convert from
E 380 SELECTED CONVERSION FACTORS
to
atmosphere (760 mm Hg)
board foot
Btu (International Table)
Btu (International Table )/h
Btu (International Table)-in./s- ft2-T (k, thermal con-
ductivity)
-. calorie (International Table)
centipoise
cenustokes
circular mil
degree Fahrenheit
foot
ft2
ft3
ft-lbf
ft-lbf/min
ft/s2
gallon (U.S. liquid)
horsepower (electric)
inch
in.2
in.3
inch of mercury (60*F)
inch of water (60T)
kgf/cm-
kip(lOOOlbf)
kip/in.2 (ka)
ounce (U.S. fluid)
ounce-force
ounce (avoirdupois)
oz (avoirdupoisVft2
oz (avoirdupoisVyd2
oz (avoirdupoisVgal (U.S. liquid)
pint (U.S. liquid)
pound-force (IbO
pound (Ib avoirdupois)
lbf/in2 (psi)
lb/in.J
lb/ft3
quart (U.S. liquid)
ton (short. 2000 Ib)
torr (mm Hg, 0*Q
W-h
yard
yd2
ydj
•Exact
pascal (Pa)
cubic metre (m3)
joule (J)
watt(W)
wan per metre kelvin (W/(m-K)]
joule (J)
pascal second (Pa- si
square metre per second (nr/s)
square metre (nr)
degree Celsius
metre (m)
square metre (nr)
cubic metre (m3)
joule (J)
watt(W)
metre per second squared (m/s2)
cubic metre (mj)
watt
metre im)
square meter (m-)
cubic metre inr5)
pascal (Pa)
pascal (Pa)
pascal (Pa)
newton (N)
pascal (Pa)
cubic metre (m3)
newton (N)
kilogram (kg)
kilogram per square metre (kg/nr)
kilogram per square metre (kg/nr)
kilogram per cubic metre (kg/m3)
cubic metre (m3)
newton (N)
kilogram (kg)
pascal (Pa)
kilogram per cubic metre (kg/m3)
kilogram per cubic metre (kg/m3)
cubic metre (m3)
kilogram (kg)
pascal (Pa)
joule (J)
metre (m)
square metre (nr)
cubic metre (m3)
multiply by
1.013 25 x 10*
2.359 737 x I0-i
1.055 056 x IO3
2.930 711 x 10"'
5.192 204 x IO2
4.186 800"
1.000 000' x I0~3
1.000 000' ID"6
5.067 075 x 10"°
3.048
9.290
2.831
1.355
2^59
3.048
3.785
7.460
2.540
6.451
1.638
3.376
2.488
9.806
4.448
6.894
2.957
1780
1834
3.051
3J90
7.489
4.731
4.448
4.535
6.894
1767
1.601
9.463
9.071
1J33
3.600
9.144
8.361
7.645
000" x 10-'
304* lO'2
685 x ID'2
818
697 x lO'2
000* x 10"
412 x lO'3
000* x IO*2
000* x IO'2
600* x IO-4
706 x 10~J
85 x 103
4 x IO2
650* x IO4
222 X IO3
757 x IO6
353 x 10~J
139 x 10~'
952 x 10'2
517 x 10-'
575 x JO'2
152
765 x 10-*
222
924 x 10'1
757 x IO3
990 x IO4
846 x 10
529 x IO"4
847 x IO2
22 X IO2
000* x IO3
000* x ID'1
274 x 10"
549 x 10"
Page 142
-------
Designation: £ 691 - 87
Standard Practice for
Conducting an Interiaboratory Study to Determine the
Precision of a Test Method1
TT^MMMdi* awed under the fixed deosniuon £ 691: the number unmedatdv foUowint the deodiuion indicates the yev of
^^ ^SfUL^' I" *? *"* "t™0**- «•* ** «>« «*«««>. A number miwemaeMuuttcuaineyev 01 last mppiov^. A
i epmoa I,) incucaiei an editonai chante wee the ta revnwn or lappnvn.
INTRODUCTION
Tests performed on presumably identical materials in presumably identical circumstances do
not. in general, yield identical results. This is attributed to unavoidable random errors inherent in
every test procedure: the factors that may influence the outcome of a test cannot all be completely
controlled. In the practical interpretation of test data, this inherent variability has to be taken into
account. For instance, the difference between a test result and some specified value may be within
tnat which can be expected due to unavoidable random errors, in which case a real deviation from
r UC haS not been demonstr»«i. Similarly, the difference between test results from
« r . ,
two batches of material will not indicate a fundamental quality difference if the difference is no
more than can be attributed to inherent variability in the test procedure.
Many different factors (apart from random variations between supposedly identical specimens)
may contribute to the variability in application of a test method, including: a the operator, b
equipment used, c calibration of the equipment, and d environment (temperature, humidity, air
pouutton. etc.). It is considered that changing laboratories changes each of the above factors. The
variability between test results obtained by different operators or with different equipment will
usuauy oe greater than between test results obtained by a single operator using the same
equipment. The variability between test results taken over a long period of time even by the same
operator will usually be greater than that obtained over a short period of time because of the greater
pOMLbility of changes in each of the above factors, especially the environment
rne general term for expressing the closeness of test results to the "true" value or the accepted
reference value is accuracy. To be of practical value, standard procedures are required for
determining the accuracy of a tea method, both in terms of its bias and in terms of its precision.
rais practice provides a standard procedure for determining the precision of a test method.
™^I- "1 ***"**»* «st methods, is expressed in terms of two measurement concepts.
repeatability and reproducibility. Under repeatability conditions the factors listed above are kept or
remain reasonably constant and usually contribute only minimally to the variability. Under
reproduability conditions the factors are generally different (that is. they change from laboratory to
taboratory) and usually contribute appreciably to the variability of test results. Thus, repeatability
andreproduabuity are two practical extremes of precision.
The repeatabiutymeasure, by excluding the factors a through d as contributing variables, is not
intended as a mechanism for verifying the ability of a laboratory to maintain "in-connol"
conditions for routine operational factors such as opei^r-tcM^erator and eqmpment difference
or •nyfffecBof longer time intervals between test results. Such a control study is a separate issue
foreacn laboratory to consider for itself, and is not a recommended pan of an interlaboratory
study.
The reproducibility measure (including the factors a through d as sources of variability) reflects
wnat precision might be expected when random portions of a homogeneous sample are sent to
random "in-contror laboratories.
To obtain reasonable estimates of repeatability and reproducibility precision, it is necessary in an
imenaboratory study to guard against excessively sanitized data in the sense that only the uniquely
best operators are involved or that a laboratory takes unusual steps to get "good" results. It is also
important to recognize and consider how to treat "poor" results that may have unacceptable
assignable causes (for example, departures from the prescribed procedure), the inclusion of such
of Suboomnunee £11.04 on Devehmmciu Cunem edmon ipprovcd Nov. 16. 1987. Pubbstad Jwwwy 1988. Oricwilv
pubtahed u £ 691 - 79. Laa wewom cdmoa £ 691 - 79.
Page 143
-------
E691
results in the final precision estimates might be questioned.
An essential aspect of collecting useful consistent data is careful planning and conduct of the
study. Questions concerning the number of laboratories required for a successful study as well as
the number of test results per laboratory affect the confidence in the precision statements resulting
from the study. Other issues involve the number, range, and types of materials to be selected for the
study, and the need for a well-written .test method and careful instructions to the participating
laboratories.
To evaluate the consistency of the data obtained in an interiaboratory study, two statistics may
be used: the "/c-value". used to examine the consistency of the within-laboratory precision from
laboratory to laboratory, and the "/i-value". used to examine the consistency of the test results from
laboratory to laboratory. Graphical as well as tabular diagnostic tools help in these examinations.
1. Scope
1.1 This practice describes the techniques for planning,
conducting, analyzing, and treating the results of an interiab-
oratory study (ILS) of a test method. The statistical tech-
niques described in this practice provide adequate informa-
tion for formulating the precision statement of a test method.
1 .2 This practice does not concern itself with the develop-
ment of test methods but rather with gathering the informa-
tion needed for a test method precision statement after the
development stage has been successfully completed. The data
obtained in the interiaboratory study may indicate, however.
that further effort is needed to improve the test method.
1.3 Since the primary purpose of this practice is the
development of the information needed for a precision
statement, the experimental design in this practice may not
be optimum for evaluating materials, apparatus, or indi-
vidual laboratories.
1.4 Field of Application — This practice is concerned ex-
clusively with test methods which yield a single numerical
figure as the test result, although the single figure may be the
outcome of a calculation from a set of measurements.
1.4.1 This practice does not cover methods in which the
measurement is a categorization, such as a go-no-go alloca-
tion (two categories) or a sorting scheme into two or more
categories. For practical purposes, the discontinuous nature
of measurements of these types may be ignored when a test
result is defined as an average of several individual measure-
ments. Then, this practice may be applicable, but caution is
required and a ytatmirian should be consulted.
1 .5 The information in this practice is arranged as follows:
Sunmufy of PFKOCB
Significance ud Use
Planning tne Interlaboratonr Studv (ILS)
ILSMembenhtp
Teat Method
Laboratories
Materials
Number of Test Results per Material ..
Protocol
Conducting tne Testa* Phut of the ILS
Pilot Run
Full Scale Run
Calculation and Display of Statistics
Calculation of the Statistics
Tabular and Graphical Display of Statistics .
Data Consistency
Flatting inconsistent Results
I
2
3
4
5
6
7
8
9
10
II
12
13
14
15
16
17
Section
IS
19
20
21
Al
A2
Table
1-7
8-11
12
Rg.
1-5
6-10
Investigation • • •
Task Croup Actions
Examples of interiaboratory Studies
Precision Statement Information
Repeatability and Reproducibility
Annexes
Theoretical Considerations
Index to Selected Terms
References
rabies ana Figures
Tables
Glucose in Serum Example .
Pemosans in Pulp Example
Critical Values of Consistency Statistics, b and k
Figures
Glucose in Serum Example
Pemosans in Pulp Example
1.6 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the safety problems associated with its use. It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use.
2. Referenced Documents
2.1 ASTMStandards: .
E 177 Practice for Use of the Terms Precision and Bias in
ASTM Test Methods2
E 456 Terminology for Statistical Methods*
E 1169 Guide for Conducting Ruggedness Tests*
3. Terminology
3.1 Definitions—For formal definitions of statistical
terms, see Terminology E 456.
32 Descriptions of Terms Specific to This Standard:
3.2.1 Test Method and Protocol—In this practice, the
term "test method" is used both for the actual measurement
process and for the written description of the process, while
the term "protocol" is used for the directions given to the
laboratories for conducting the ILS.
3.2.2 Observations. Test Determinations and Test Results:
322.1 A test method often has three distinct stages, the
direct observation of dimensions or properties, the arith-
metic combination of the observed values to obtain a test
determination, and the arithmetic combination of a number
of test determinations to obtain the test result of the test
method. In the simplest of test methods a single direct
: Annual Book of ASTM Standaras. Vol 14.02.
Page 144
-------
E691
2.5
1.5
S o
J-.5
-1
-1.3 f
-2
I
T^+**»'
2.15
L
TT
-2.13
-2.5 j-
H*T: A»CDE ABCOE ABCDE ABODE ABCDE ABCOE ABCDE ABCDE
"»' I 2 3 4 5 6 7 8
FKL 1 GIuco»«m8wum:/i—Materials wMMn LaboratoriM
observation is both the test determination and the test result
For example, the test method may require the measurement
of the mass of a test specimen prepared in a prescribed way.
Another test method may require the measurement of the
area of the test specimen as well as the mass, and then direct
that the mass be divided by the area to obtain the mass per
unit area of the specimen. The whole process of measuring
the mass and the area and calculating the mass per unit area
is a test determination. If the test method specifies that only
one test determination is to be made, then the test determi-
nation value is the test result of the test method. Some test
methods require that several determinations be made and
the values obtained be averaged or otherwise combined to
obtain the test result of the test method. Averaging of several
determinations is often used to reduce the effect of local
variations of the property within the material.
- 3.2.2.2 In this practice, the term "test determination" is
used both for the process and for the value obtained by the
process, except when "test determination value" is needed
for clarity.
3.2^.3 The number of test determinations required for a
test result should be specified in each individual test method.
The number of test results required for an interlaboratory
study of a test method is specified in the protocol of that
studv.
3.2.3 Test Specimens and Test Units—In this practice a
test unit is the total quantity of material needed for obtaining
a test result as specified by the test method. The portion of
the test unit needed for obtaining a single test determination
is called a test specimen. Usually a separate test specimen is
required for each test determination.
3.2.4 Precision, Bias, and Accuracy of a Test Method:
3.2.4.1 When a test method is applied to a large number
of portions of a material, that are as nearly alike as possible.
the test results obtained nevertheless will not all have the
same value. A measure of the degree of agreement among
these test results describes the precision of the test method
for that material.
3.2.4.2 Numerical measures of the variability between
such test results provide inverse measures of the precision pf
the test method. Greater variability implies smaller (that is.
poorer) precision and larger imprecision.
3.2.4.3 This practice is designed only to estimate the
precision of a test method. However, when accepted refer-
ence values are available for the property levels, the test
result data obtained according to this practice may be used in
estimating the bias of the test method. For a discussion of
bias estimation and the relationships between precision, bias.
and accuracy, see Practice E 177.
3.2J Repeatability and Reproducibility—These terms
deal with the variability of test results obtained under
Page 145
-------
V
TABLE 1 GlucoM in SmumlLS T«st Aa*un Data
A B C o E
1 41.03 7828 13236 193.71 292.78
41.45 78.18 13333 19339 29409
41.3T 78.49 133.10 19335 29239
2 41.17 77.78 13232 19038 29227
4230 8038 13630 200.14 30940
41.15 7934 13640 19430 29538
3 41.01 79.18 13231 192.71 29533
4038 79.72 13530 19328 290.14
4236 8031 13536 19028 29234
4 «•£ 51-08 13"° 19S-85 *»•«
4237 7830 14830 19836 29544
4233 8132 13539 19943 29633
5 41.88 78.16 13130 19239 29333
41.19 7938 134.14 191.44 292.48
4132 7833 133.76 195.12 29428
6 £S £2 137-21 19SJ4 »™
«i2 I?« 135-14 198-26 »*"
4228 81.75 13730 198.13 29033
7 41.08 79.76 13037 194.66 28729
«« «« 131'59 191'99 »7o-
3932 7735 13432 187.13 28936
8 4336 8044 13546 19736 298.46
42.65 8030 135.14 19539 29528
41.72 7930 13333 20032 296.12
jecified laboratory conditions. Repeatability concerns the
inability between independent test results obtained within
single laooratory in tne soonest practical period of time by
single operator with a specific set of test apparatus using
st specimens (or test units) taken at random from a single
oantity of homogeneous material obtained or prepared for
ie ILS. Reproducibility deals with the variability between
ngle test results obtained in different laboratories, each of
hich has applied the test method to test specimens (or test
aits) taken at random from a single quantity of homoge-
nous material obtained or prepared for the ILS.
3.2.5.1 Repeatability Conditions— The within-laboratory
mditions specified above for repeatability The single-
particular step in the measurement process the «™»
imbinauon of operator and apparatus is used for every test
suit and on every material. Thus, nnf nrvrritni- 'm™,.
'pnarr thp tMt vn^nmem s uwmH •••••••••« *it^ ^i -
btAUK uic Wat TIH i inirini a anAjuu measure IHB "Munition?
id a third measure the breaking force. "Shortest practical
•nod of time" means that the test results, at least for one
aierial. are obtained in a time not less than in normal
nine and not so long as to permit significant chanen in twr
atenal. equipment or environment.
3.3 For further discussion of the term* rfjscussed above
: Practice E 177, and the formal definitions m Practice
456.
Summary of Practice
4.1 The procedure presented in this practice consists of
ree basic steps: planning the imeriaboratory study, guiding
: testing phase of the study, and analyzing the test result
E691
TABLE 2* IntMtaboratory Study WofkshMt for Oluco** in Swum
UOdfaWV TestRaSUttS.Jt
NWMMT ; 23
1 4133 41.4S 41.37 41.2833 02230 -02350 -039 021
2 41.17 4230 41.15 41.4400 0.4851 -0.0783 -0.13 0.46
3 41.01 40.68 42.66 41.4500 1.0608 -0.0683 -0.11 1.00
4 3937 4237 4233 41.4567 1.8118 -0.0618 -0.10 1.70
5 4138 41.19 41.32 41.4633 03667 -0.0550 -0.09 0.34
6 4328 4030 4228 42.0200 1.4081 03017 033 132
7 4138 4127 3932 40.4567 12478 -1.0816 -1.75 1.17
8 4336 4235 41.72 423767 0.8225 1.0584 1.75 0.77
* Awaraoa of eat awraoas. J - 41 3183
Standard dawaaen of out avarages. *r • 0.6061
napaaiaoaliv sunautt dsvwaon. », - 1.0632
tMtNTK
x - mdmdual last rasut
H
t » CM average- Zx/n wmren • number ottnt mutts par cat • 3.
0
i - avwaoe of can avenge* - £ T/p whera p - numdar ot ttboraianaa - 8.
/"
a » ce« oevtaoan • « - i
&r m vafMaiadriity standard oswaaon • V 2 M /p
A -«*/»„ and
* -«/»^
data. The analysis utilizes ta^'iar, graphical . and siaiifi"*J*
diagnostic tools for evaluating the consistency of the data so
that unusual values may be detected and investigated, and
also includes the calculation of the numerical measures of
precision of the test method pertaining to both within-
laiv^ratftffv rpt^iitahilftv 9nH lytiifrrn»lnhnratflrv 1 CUlDdUC-
ibility.
5? Signifinntr ••«* Use
test methods in terms of repeatability niu* repmriiifahility.
This practice may be used hi obtaining the needed informa-
tion as simply as possible. This information may then be
Practice E 177.
PLANNING THE INTERLABORATORY STUDY (ILS)
6. ILS Membership
6.1 Task Grout? — Either the task group that developed
the test method, or a special task group appointed for the
purpose, must have overall responsibility for the ILS, in-
cluding funding where appropriate, staffing, the design of the
ILS, and decision-making with regard to questionable data.
1 To (acitittte toe ptcpmaaa of the foul upon on the ILS. the tt* araop can
Page 146
-------
E691
TABLE
• should specify any special calibration procedures and the
rjeatab^conditions to be spewed fn thfprotoS? £
6.3 Statistician:
6.3.1 The test method task group should obtain th»
assistance of a person familiar withth I statical SS±
available, the task group should obtain
who has experience in practica
The
materials and the test method involved
statistical knowledge, (see 15.1 J i 15
6.4 DataAnaiyst—Tbi* individual should be
who » carefid in making calculations and can
directions in Sections 15 through 17
6.5 L^orarory /LS Supervuor-Each laboratory must
TABLE 4
- Laboratory
1
2
3
4
5
6
7
8
021
0.46
1.00
1.70
044
142
1.17
0.77
B
0.11
049
046
145
042
149
148
044
Matanai
022
0.79
0.63
0.44
0.47
0.77
046
0
042
1 73
041
0.74
0.72
0.63
1.45
oiii
049
022
024
1.03
044
Laboratory
1
2
3
4
5
6
7
8
TABLE
A
-0.39
-0.13
-0.11
-0.10
-0.09
0.63
-1.75
1.75
5* GIUCOM in Sanmi-ft '
e
-1.36
-0.45
0.22
1.85
-0.99
021
-0.16
0.67
Matenai
C
-0.88
0.39
-0.08
1.59
-0.64
1.09
-128
0.01
0
-0.41
0.15
-1.01
0.96
-0.64
047
-1.33
1.31
E
-0.46
1.64
-0.68
0.49
-0.34
0.17
-1.62
0.79
_^_ after oorracong Get C4.(aee 20.1.4 and 20.14).
•Critical value-2.15.
have an ILS supervisor to oversee the conduct of the ILS
within the laboratory and to communicate with the ILS
Coordinator. The name of the supervisor should be obtained
on the response form to the "invitation to participate" (see
7. Basic Design
7.1 Keep the design as simple as possible in order to
obtain estimates of within- and between-laboratory vari-
ability that are free of secondary effects. The basic design is
represented by a two-way classification table in which the
rows represent the laboratories, the columns represent the
materials, and each ceil (that is. the intersection of a row with
a column) contains the test results IT""** by a particular
laboratory on a particular material (see Table 1).
8. Test Method
8.1 Of prime importance is the existence of a valid,
well-written test method that has been developed in one or
more competent laboratories, and has been subjected to a
niggedness test prior to the ILS.
8-2 A niggedness test is a screening procedure for investi-
gating the effects of variations in environmental or other
conditions in order to determine how control of such test
conditions should be specified in the written description of
the method. For example, the temperature of the laboratory
or of a heating device used in the test may have an effect that
cannot be ignored in some cases but may be much less in
others. In a nimedness test, deliberate variations in temper-
ature would be introduced to establish the allowable limits
on control of temperature. This subject is ^iy"**"** more
rally in Refs (U and 3) see also Guide E 1169.
8 J As a result of carrying out the screening procedure.
and of some experience with the test method in the spon-
soring laboratory and one or two other laboratories, a written
TABLE 6** Glucoa* in Serum*
Laboratory
1
2
3
4
5
6
7
8
A
021
0.46
1.00
1.70
044
142
1.17
0.77
B
0.11
049
046
145
042
149
148
044
Matenai
C
048
1.40
1.12
142
0.78
043
148
0.63
0
0.02
1.78
041
0.74
0.72
043
1.4S
044
E
0.18
(233)
049
022
024
1.03
044
0.42
* Rwantataa vamaa anar oorracarq eat C4. (Sea 20.1.4 ana 20.14).
• Critical vakia* 2.06.
Page 147
-------
E691
2.5 +-
2.15
-.5 f-
* -I f-
-1.3 f-
-2
-2.5 +-
-2.15
12345678 12345678 12345678 12345678 12345678
ABODE
MAI:
FNL 2 GlucoMinS
version of the test method must have been developed (but
not necessarily published as a standard method). This draft
should describe the test procedure in terms that can be easily
followed in any properly equipped laboratory by competent
personnel with knowledge of the «"«*»nah and the propeuy
to be tested. The test conditions that affect the test results
appreciably should have been identified and the proper
degree of control of the test conditions specified in the
description of the test procedure. In addition, the test
method should specify how closely (that is, to how many
digits) each observation in the test method is to be measured.
8.4 The test method should specify the calibration proce-
dure and the frequency of calibration.
.9. Laboratories
9.1 Number of Laboratories:
9.1.1 An ILS should include 30 or more laboratories but
this may not be practical and some ILS have been run with
fewer. It is important, that enough laboratories be included
in the ILS to be a reasonable cross-section of the population
of qualified laboratories: that the loss or poor performance of
a few will not be fatal to the study, and to provide a
reasonably satisfactory mimatc of the reprodudbility.
9.1.2 Under no circumstances should the final statement
of precision of a test method be based on acceptable test
results for each material from fewer than 6 laboratories. This
would require that the ILS begin with 8 or more laboratories
in order to allow for attrition.
9.1 J The examples given in this practice include only 8
and 7 laboratories, respectively. These numbers are smaller
than ordinarily considered acceptable, but they are conven-
ient for illustrating the calculations and treatment of the
ufltft*
92 Any laboratory considered qualified to run thetoi
routinely (including laboratories that may not be members of
ASTM) should be encouraged to participate in the ILS, if the
preparatory work is not excessive and enough suitably
homogeneous material is available. In order to obtain an
adequate number of participating laboratories, advertise the
proposed ILS in where appropriate (for example, trade
magazines, meetings, circulars, etc.).
9.3 "Qualified" implies proper laboratory fatalities and
testing equipment, competent operators, familiarity with the
test method, a reputation for reliable testing work, and
sufficient time and interest to do a good job. If a laboratory
meets all the other requirements, but has had insufficient
experience with the test method, the operator in that
laboratory should be given an opportunity to familiarize
himself with the test method and practice its application
before the ILS starts. For example, this experience can be
obtained by a pilot run (see Section 13) using one or two trial
Page 148
-------
E691
moo CD >
41.5183
79.6796
134.7264
194.7170
294.4920
3.6061
10027
1.7397
2.5950
2.6931
1.0632
14949
1.5434
2.6251
3.9350
1.0632
1.5796
2.1462
3.3657
41923
2.98
4.19
4.33
7.35
11.02
2.98
4.42
6.02
9.42
11.74
samples provided by the task group and returnine the raw
^"™ 3nd " "M to ***** *e work to a
-
i
°Ut i
TTie ILS should not be restricted to a group
exceptionally qualifiS
h o ° Kliniates for i
test method should be obtained
qualified laboratories and personnel
in
10. Materials
10.1 Material.^ j-u-uwnnapiopertv
be measured. Different materials having the same'
may be expected to have different property levels, meaning
higher or lower values of the property. Different rtlSTof
the same material or compound to be assayed are considered
different materials" for the purpose of this practice S
. terminology "different levels of material" may be used! if
appropriate. J ^^ u
'Si "^f Jmmb?r and type of matcnals to be included in
, an ILS will depend on the range of the levels in the daks of
materials to be tested and likely relation of precision toleve
over that range, the number of different typSof^ateriabTo
which the test method B to be applied, the difficSwand
expense involved in obtaining, processing, and distributing
samples, thedifficuity of, length of time requiredfor S
expense of performing the test, the commercial or legal nod
for obtaining a reliable and comprehensive esumatcTf
TABLE 8 PentouiM in PutpHLS Teat Rtsutts
Laboratory
1
2
3
4
5
6
7
A
0.44
0.49
0.44
0.41
0.41
0.41
0.51
0.51
0.51
0.40
0.38
0.37
0.49
0.49
0.49
0.43
0.41
0.40
0.186
0.171
0.153
8
0.96
0.92
0.82
0.83
0.83
0.84
0.92
0.93
0.92
0.96
0.94
0.94
0.82
0.82
0.84
0.88
0.92
0.88
0.866
0.900
0.831
C
123
1.88
1.24
1.12
1 12
1.12
1.11
1.13
1.11
1.15
1.13
1.13
0.98
0.98
0.98
1.11
1.12
1.11
1.05
0.962
0.927
0
125
1.25
1.42
1.25
1.25
1.26
1.35
1.35
1.35
1.29
1.29
1.29
1.23
1.23
1.23
1.31
1.30
1.31
1.13
1.15
1.15
E
1 98
1.92
1.80
1.99
1.94
1.95
2.05
2.08
2.03
2.05
2.04
2.04
1.94
1.96
1.96
2.01
1.99
1.98
1.98
1.93
1.98
F
4.12
416
416
410
411
4.10
4.11
416
4.16
420
420
4.22
4.61
463
4.53
3.93
3.92
3.84
4.21
418
416
G
5.94
5.37
5.37
576
S.26
576
5.16
5.16
571
570
5.20
570
5.00
5.00
4.96
485
4.87
4.91
577
5.32
5.10
H
10.70
10.74
10.83
10.07
10.05
9.82
10.01
10.17
10.17
10.98
10.67
10.52
10.48
1077
1048
9.57
9.57
9.62
11.5
10.8
11.5
1
17.13
16.56
16.56
16.06
16.04
16.13
16.01
15.96
16.06
16.65
16.91
16.75
15.71
15.45
15.66
15.05
14.73
15.04
18 M
187
18.1
precision, and the uncertainty of prior information on any of
these points.
10.2.1 For example, if it is already known that the
precision is either relatively constant or proportional to the
average level over the range of values of interest, a smaller
number of materials will be needed than if it is merely
known that the precision is different at different levels. The
niggedness test (see 8.2) and the preliminary pilot program
(see Section 13) help to settle some of these questions, and
may often result in the saving of considerable time and
expense in the full ILS.
10.2.2 An ILS of a test method should include at least
three materials representing different test levels, and for
development of broadly applicable precision statements, six
or more materials should be included in the study.
10.2.3 The materials involved in any one ILS should
differ primarily only in the level of the property measured by
the test method. When it is known, or suspected, that
different classes of materials will exhibit different levels of
precision when tested by the test method, consideration
should be given to conducting separate intexiaboratory
studies for nch class of material.
10.3 Each material in an ILS should be made to be or
selected to be as homogeneous as possible prior to its
subdivision into test units or test specimens (see 3.2.3). If the
randomization and distribution of individual test specimens
(rather than test units) does not conflict with the procedure
for preparing the sample for test, as specified in the test
method, greater homogeneity between test units can be
achieved by randomizing test specimens. Then each test unit
would be composed of the required number of randomized
test specimens. (See Section 11 and 14.1 for the quantity of
each material needed, its preparation and distribution.)
NOTE—Jt may be convenient to use established lefercncc materials.
since their homogeneity has been demonstrated.
Page 149
-------
E691
2.54-
o
u
2
1.5 •
1
.5
0 -
•
•
.III,
1
SOI: AICDE ABCOE ABCOL ABODE ABODE ABCOE ABCDC ABCDE
2.06
1 2 3 4 5 6 7 8
HO. 3 OhiooMinSMunefc-
11. Number of Test Results per Material
11.1 In the design of an ILS a sufficient total number of
test results on each material must be specified to obtain a
good estimate of the measure of repeatability, generally the
repeatability standard deviation. In many cases, the standard
deviation in question will be a function of the property level
being measured. When this occurs, the standard deviation
should be determined separately for each level. It is generally
sound to limit the number of test results on each material in
each laboratory to a small number, such as three or four. The
minimum number of test results per laboratory will normally
be three for a chemical test and three or four for a physical or
optical test The number may be as small as two when there
is little danger that a test unit will be lost or questionable ten
results obtained, or as many as ten when test results are apt
to vary considerably. Generally, the time and effort invested
in an ILS is better spent on Mamming more materials across
more laboratories than on recording a large number of test
results per material within a few laboratories.
12. Protocol
12.1 In the protocol cite the name, address, and tele-
phone number of the person who has been dmgnated US
coordinator (see 6.2). Urge the laboratories to call the
"•wrfTffittrr when any questions arise as to the conduct of
thettS. f w
122 dearly identify the specific version of the test
method being studied. If the test method allows several
options in apparatus or procedure, the protocol should
specify which option or options have been selected for the
TABLES P-h->
1 O46
2 005
3 093
4 -0.19
5 O7S
6 008
7 ±m
a
O35
-1.14
088
1.40
-128
021
-0.41
C
-------
E691
2.5 4-
2
u
••«
*4
•1
*il
«
*>
(/>
» 1. 5 •
U
e
V
*»
*
91
!
u
, 1
Jt
.5
0 ' ( - - r I I L I • • • u -.J.J._1J I^L±^ia
-
-
**
(|
LLLiiii
1
II
Ul: 12345678 12345678 1234S678 12345678 12345678
M*T: A B C 0 E
2.06
ROL 4
GlucoM in SMURB *—UboiatoriM wMMn Utotarals
* adibraiion Procedures are required
before every determination or every test result, theyVhould
be described specifically in the test method tf thTtett
method specifies calibration only daily or less frequently, the
before obtaining each test result While doine
ehniinatt caJibnuion drift and help ensure relative indepen-
dence ofthe testresute, changes in calibration may^re^e
the variability between test results. w
addJlS^e,any ***? ^"aaa** that must be
addressed in implementing the repeatability conditions, such
as the penod of time between obtaining L test reSts for
TABLE 10 P«ntOMn« m Pulp**
Latxnury •
2
3
4
S
6
7
040
040
142
040
1.02
1.10
B
0.18
0.18
0.36
046
0.72
1.07
0.00
048
048
040
0.04
0.44
0
<2S>
0.15
040
040
040
0.15
0.31
Mttmi
E
<22>
0.67
0.64
0.15
029
043
0.73
F
0.71
0.18
049
046
1.83
1.S2
0.77
(2J2>
0.00
022
040
0.17
023
044
0.72
0.48
1.21
044
0.15
(O5>
, "~
1.53
021
023
041
0.64
044
the same material; that is. not less than in normal testing and
not so long as to likely permit significant changes in test
material, equipment or environment.
12J Specify the required care, handling, and conditioning
of the materials to be tested. F«piain the coding system used
in identifying the ma"rmfc and the distinction between test
units and test specimens, where appropriate.
12.6 Supply data shet-fy for fa**** P^T**1"!*! fnr leuQiding the
raw data as observations are "«"1» Give instructions on the
number of g»gtiifii-a«t Higfo to be recorded, usually one
more, if possible, than required by the test method. Also,
supply test result sheets on which test results can be
calculated and reported. In many instances this can be
combined with the raw data sheet. Specify the number of
significant digits to be reported, usually two more than
required by the test method. Request the laboratories send
raw data and test result sheets as soon as the testing is
completed, and at least weekly if testing will continue over
several weeks.
12.7 Request that each laboratory keep a record (or log) of
any special events that arise during any phase of the testing.
This record, to be sent to the ILS coordinator, will provide a
valuable source of information both in dealing with unusual
data and in making improvements in the test method in
future revisions.
Page 151
-------
E691
4.5 4
3.5-4
I, 4
1.5 4
.5 4
0
•
FKL5
12.7.1 Instruct the laboratories to notify the ILS coordi
natoi -promptly whenever an error in test pr^u
that a decision can be made as to whether a new s
12.8 Enclose with toe protocol a questionnaire
mformanon on specific aspects of the
pparat
calibration. or procedure, as well as anv other
that might assist in dealing with data'
or
ng wt ata
ensure the task group that tne laboratory compfcd
current requirements of the test method. Also obtain any
^information that may -be needed in preparing the fiS
research report on the ILS (see Footnote 3).
CONDUCTING THE TESTING PHASE OF THE ILS
13. Pilot Ron
13.1 Before investing laboratory time in the full scale ILS.
u is usually wise to conduct a pilot run wi
pilot run with oiuone,
TABLE 11
A
B
C
0
E
F
G
H
1
0.4048
04841
1.1281
14886
14809
4.1814
5.1843
10.4010
164610
0.1131
04447
0.1871
04878
04838
04071
04172
04830
14901
04180
04322
0.1429
04375
04396
04325
0.1330
0.1936
04156
0.1137^
04519
0.1957
04742
04628
04088
04426
04646
1.1042
044~
049
0.40
0.11
0.11
049
047
044
040
«
042
0.14
045
041
0.16
048
046
144
349
200 290 100 390 400
(O) and ftepMtttofltty (•) Versus Avwcg*
perhaps two, materiaKs) to determine whether the test
method as well as the protocol and all the ILS procedures are
dear, and to serve as a familiarization procedure for those
without sufficient experience with the method (see 9.3). The
results of this pilot run a|y give the task group an indication
of how well »arh laboratory will perform in terms of
promptness and following the protocol. Laboratories with
poor performance should be encouraged and helped to take
corrective action.
13.2 All steps of the procedures described in this practice
should be followed in detail to ensure that these directions
are understood, and to disclose any mi lit nmn in the
protocol or the test method.
14. Full Scale Ron
14.1 Material Preparation and Distribution:
14.1.1 Sample Preparation and Labelling—Prepare
enough of each material to supply 50 % more than needed
by the number of laboratories committed to the ILS. Label
each test unit or test specimen with a letter for the material
and a sequential number. Thus, for ten laboratories and two
test results for tticfa laboratory the test units for material B
would be numbered from Bl to B30, or, if five test
specimens per test unit are required, the test specimens may
be numbered Bl to B1SO.
14.1.2 Randomization—for each material independently,
allocate the specified number of test units or test specimens
to each laboratory, usjng a random number table, or a
suitable computerized randomization based on random
numbers. See Ref. (4) for a discussion of randomization.
Page 152
-------
E691
2.5
1.5 4-
1 4-
.5
§-.5
u
-I
-1.5
-2
2.05
111
-2.05
-2.54-
M*T: ABCDEFGHI ABCOEFCBZ ABCOEFCHI ABCOEFCHI ABCDEFGBI ABCOEFGBX ABCOEFGHI
2 3
F1GL 6 POTtmw in Pulp
6
itori«
14.1.3 Shipping— Ensure that the test units are packaged
property to arrive in the desired condition. When the
material is sensitive to the conditions to which it is exposed
(light, heat, humidity, etc), place special directions for
opening the package on a label outside the package. Clearly
indicate the name of the person who has been dgffgimfd as
ILS supervisor at the laboratory on the address of rnrh
package. Follow each laboratory's instructions for ensuring
prompt delivery of the
14.1.4 Follow-up — Once the test units have been shipped,
the ILS coordinator should call each laboratory US super-
visor within a week to ten days to confirm that all test units
have arrived safely. If the task group has decided to inter-
mingle test units from different materials in the order of
testing, the testing should not start until all the test units
have arrived at the laboratory so they can be tested in the
specified order.
14.1.5 Replacement Sets of Test t/niu— AS the ILS
progresses, a laboratory may discover that the test method
was not used properly on some test units. The laboratory ILS
supervisor should discuss this with the ILS coordinator, who
may send a replacement set of test units, replace the miyiyd
test units, or dp nothing, as may seem desirable.
14.2 Checking Progress — From time to time, at intervals
appropriate to the magnitude of the ILS. the coordinator
should call each ILS supervisor to ascertain how the testing is
progressing. By comparing the progress of all laboratories,
the coordinator can determine whether some laboratories are
lagging considerably behind the others and so advise these
laboratories. , ,.
14.3 Dau Inspection—The completed data sheets should
be examined by the coordinator imiPfriigte|y uPpn receipt in
order to detect unusual values or other deficiencies that
should be questioned. Replacement sets of test units or of
specific test units may be sent when there is missing or
obviously erroneous data. The task group can deade later
whether or not the additional data should be used in tne
estimation of the precision of the test method.
CALCULATION AND DISPLAY OF STATISTICS
IS. Calculation of the Statistics
15.1 Overview—The analysis and treatment of the ILS
test results have three purposes, to determine whether the
collected data are adequately consistent to form the basis for
a test method precision statement, to investigate and act on
any data considered to be inconsistent and to obtain the
precision statistics on which the precision statement can be
based. The statistical analysis of the data for estimates ot tne
precision statistics is simply a one-way analysis of variance
(within- and between-laboratories) carried out separately tor
each level (material). Since such an analysis can be invau-
Page 153
-------
E691
2.5
l.S
ta
>>
o
u
-I
-i.s 4-
-2 _t
2.05
-2. OS
-2.5 4-
Ukl:
SUX:
1234567 1234567 1234567 1234567 1234567 1234567 1234567 1234567 1234567
* B CDEFCHI
RQ. 7 P««o.«w«if>utp:h-Cabor«ortowitWnlyUt«wi«
dated by the presence of severe outliers, it is necessary to first
examine the consistency of the data. The following para-
graphs show, in terms of a numerical example, how the
entire program is carried out:
15.1.1 The calculations are illustrated with test results
from an ILS in which the concentration of glucose in serum
(see Table 1) was measured at five different concentration
levels by eight laboratories. Each laboratory obtained three
test results at each concentration level.
15.1.2 For extended calculations it is usually necessary to
retain extra significant digits in order to ensure that statisti-
cally important information is not lost in calculation by
rounding off too soon. As a general rule, retain at least two
more digits in the averages than in the reported test results
and at least three significant figures in the standard devia-
tions.
15.1.3 While the calculations described in this section are
arranged for use of a hand calculator, they also can be readily
programmed for the computer. If necessary, contact Com-
mittee E-ll for advice on computational matters, (see
15.4.2).
152 Table of ILS Test Results—The test results received
from the laboratories are usually best arranged in rows and
columns as in Table 1. Each column contains the data
obtained from all laboratories for one material, and Mrh row
contains the data from one laboratory for all materials. The
test results from one laboratory on one material constitute a
cdL Thus, the cell for Laboratory 2 and Material C contains
the test results 132.92.136.90 and 136.40. This cell is called
C2, by material and laboratory. It helps in the interpretation
of the rfgtq to arrange the materials in increasing order of the
measured values.
15 J JfohfesAeoi—Generally, it facilitate the calculations
to prepare a separate calculation worksheet for each material
using Table 2 as a model but making appropriate changes for
different numbers of laboratories, and ten remits per mate-
rial. Enter the test result data for one material (from one
column of Table 1) on a worksheet Also enter the results of
the following calculations for that material on the same
worksheet, as illustrated in Table 2. Work on only one
material at a time.
15.4 Cell Statistics:
15.4.1 Cell Average, .T—Calculate the cell average for each
laboratory using the following equation:
7-2-t/n
i
where:
Y» the average of the test results in one ceil.
JT - the individual test results in one ecu. and
n »the number of test results in one cell
(I)
Page 154
-------
Thuyrom Table 2 for Matenai A. Laboratory I (that is. for
*- (41.03 + 41.43 i.4|.37)/3 -4L2833.
15.4.2 Cell Standard Deviation. s-CAcabte the standard
USmg ^ f°UowinS
E691
15.6.1 Repeatability Standard Deviation, s,—Calculate
this statistic using the following equation:
The symbols have the same meaning as for Eq 1. Thus
CellAl:
(2)
tor
j- / (41.03 - 41.2833)* -I- (41.45 - 41.2833V
V + (41.37 - 4l.2833)2J/(3 - 1)
- 0.2230
While Eq 2 shows the underlying calculation of the cell
standard deviation, inexpensive pocket calculators are avail-
able that calculate both the average and the standard
deviation directly. Check to be sure the calculator uses (n -
1) as the divisor in Eq 2, not n. and has adequate precision ot
calculation.
15.5 Intermediate Statistics:
15.5.1 Average of the Cell Averages, x—Calculate the
average of all the ceil averages for the one material using Eq
3.
(3)
I
where:
x - the average of the cell averages for one material,
x * the individual cell averages, and
p «the number of laboratories in the ILS
Thus for material A:
x - (41.2833 * 41.4400 + 41.4500 + 41 4567
+ ,41;4633 * 42-0200 * *°-4S67 + 4iS767V8
— 41.5153
13.5.2 CM Deviation.
-------
E691
Critical
vauaof
1 49
1 74
1.92
2.05
2.15
223
229
;. 2J4
2.38
2.41
2.44
2.47
2.49
241
243
244
248
247
248
249
2.60
241
2.62
242
243
244
244
TABUE 12 Critical Valuaa of nai** at »• 04 %Soni«canc« Lav*'
p
3
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Critical vaWM 01*
1.72
1.95
2.11
222
2.30
2.36
2.41
2.45
2.49
241
244
246
247
249
240
241
242
243
244
245
246
2.66
247
247
248
248
249
249
3
1.67
1.82
1.92
1.98
2.03
246
2.09
2.11
2.13
2.14
2.15
2.16
2.17
2.18
2.19
220
220
221
221
221
222
222
223
223
223
223
224
224
Kroonawaneystn
1.61
1,73
1.79
1.84
1.87
1.90
1.92
1.93
1.94
1.96
1.96
1.97
1.98
1.98
1.99
1.99
2.00
2.00
2.00
2.01
2.01
2.01
2.01
2.02
2.02
2.02
242
1.56
1.66
1.71
1.75
1.77
1.79
1.81
1.82
1.83
1.84
1.84
1.85
1.86
1.86
1.86
1.87
1.87
1.87
1.88
1.88
1.88
i 88
1.86
1 89
1.89
1.89
1.89
6
1.52
1.60
1.65
1.68
1.70
1.72
1.73
1.74
1.75
1.76
1.76
1.77
1.77
1.77
1.78
1.78
1.78
1.79
1.79
1.79
1.79
1.79
1.79
1.80
1.80
1.80
140
1.49
1.56
1.60
1.63
1.65
1.66
1.67
1.68
1.69
1.69
1.70
1.70
1.71
1.71
1.71
1.72
1.72
1.72
1.72
1.72
1.72
1.73
1.73
1.73
1.73
1.73
1.73
1.47
1.53
1.56
1.59
1.60
1.62
1.62
1.63
1.64
1.64
1.65
1.65
1.66
1.68
1.66
1.66
1.67
1.67
1.67
1.67
1.67
1.67
1.67
1.68
1.68
1.68
148
1.44
1.50
1.53
145
1.57
1.58
1.59
149
1.60
1.60
1.61
1.61
1.62
1.62
1.62
1.62
1.62
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.64
10
1.42
1.47
1.50
1.52
1.54
1.55
146
146
147
147
148
1.58
146
148
149
149
149
149
149
149
149
1.60
1.60
1.60
1.60
1.60
1.60
am wwa caicuataa from Studanrs r ana ma f*ano mm ma taaMwio mauonmaK
*«vp/(H-*v AIMS Al .2 tar cfthvivnons of
16. Tabular and Graphical Display of Statistics
lii •V/are"fl/ °«fer—It is often useful to arrange the
worksheets in order of increasing values of Jt, the material
averages. This order may facilitate interpretation.
16.2 Tables— From the Table 2 results for each material.
prepare tables of h and k as shown in Tables 3 and 4 for the
glucose in serum example.
16 J Graphs—Prepare bar graphs for h and /tin two ways:
raatenais grouped by laboratory as in Figs. 1 and 3, and
laboratories grouped by material as shown in Figs. 2 and 4
Arrange the laboratories and materials within and between
each grouping in the same order as used in Table 1. Thus the
materials will be arranged in order of increasing i from left
to right, and the laboratories in order of laboratory code
number.
DATA CONSISTENCY
17. Flagging inconsistent Results
17.1 Critical Values of the Consistency Statistics—Table
12 lists critical values of the A and A: consistency statistics at
the 0.5 % significance level. The critical values for h (first
column) depend on the number of laboratories (p, second
column) participating in the 1LS and the critical values for k
(columns headed 2 through 10) depend both on the number
of laboratories (p) and on the number of replicate test results
(«) per laboratory per material. The 0.5 % level was chosen
based on the judgment and experience that the 10 %
resulted in too many ceils being flagged and the 0.1 %'level
in too few. For further discussion see Annex Al.
17.1.1 Obtain from Table 12 the appropriate critical
values. For the glucose in serum example, the respective
critical A and Jt values are 2.15 and 2.06. In Tables 3 and 4
cirde those values that exceed the critical values and
underline those values that approach the critical values. On
each graph draw a horizontal line for each critical value: two
for/i. since there are both positive and negative values of A.
and one for k. as shown in Figs. 1 to 4.
17.1.2 The h and k graphs and the marked tables give a
picture of the overall character of the variability of the test
method as well as «"cl'"C out particular laboratories or cells
that should be investigated.
17.2 Plots by Laboratory—la order to evaluate the differ-
ences between laboratories, use the following guidelines.
17.2.1 h Graph—Then are three general patterns in these
plots. In one. all laboratories have both positive and negative
h values among the materials. In the second, the individual
laboratories tend to be either positive or negative for all
"mtfriab and the number of negative laboratories equals the
number of positive laboratories, more or less. Neither of
these patterns is unusual or requires investigation, although
they may tell something about the nature of the test method
variability. In the third pattern, one laboratory, with all h
values positive (or negative), is opposed to all the other
laboratories, with substantially all the h values negative (or
positive). Such a pattern calls for an investigation of that
laboratory.
Page 156
-------
691
2.34-
2 T=
X . .
u 1.5
s
u
2.03
MAT:
ui:
ABCDZFCHI AaCDEFCHI ABCDEFGHI AiCDEFCHI AJCDEFGHI AJCDEFGHZ ABCDEFGHI
1 2 3 4 5 6 7
FKL 8 Pmtosans in Pulp: *—Materials within Labotatortaa
17.2.1.1 Another kind of pattern to look for occurs within
one laboratory, in which the h values for low property levels
are of one sign, and for high property levels are of the
opposite sign. If the values are extreme, this behavior should
be investigated.
17.22 k Graph—Hen the primary pattern to look for is
that of one laboratory having large k values (or very small k
values) for all or most of the materials. High A: values
represent within-laboratory imprecision. Very small k values
may indicate a very insensitive measurement scale or other
measurement problem.
17.3 Plots by Material—When a plot by laboratory shows
several // or k values near the critical value line, look at the
corresponding plot by material to see how that laboratory
differs from the rest for a given material. Often a vertical line
that seems strong in the plot by laboratory, because of its
relation to the lines for the other materials, will turn out to
be reasonably consistent with the other laboratories tor the
same material. Contrarywise. the h or k value for the one
laboratory may be revealed as strongly different from the
values for the other laboratories in the plot by material. If so.
this behavior should be investigated.
18. Investigation
18.1 Clerical and Sampling Errors—Examine the labora-
tory report for each flagged cell. Try to locate where each test
result in the flagged cell begins to deviate from the others. Is
it in the original observations? Are the data rounded prema-
turely? Are the calculations correct? Then, look for signs of
mislabeling of test units such that the test result for one
material was reported as belonging to another material.
Check these errors with the laboratories: do not assume them
to be so.
18.2 Procedural Errors:
18.2.1 Study the laboratory reports again looking lor
deviations from either the test method or the protocol. For
instance, variations in the number of significant digits
reported in the test results may be a sign of incorrect
rounding, or that the equipment in one laboratory is
different from the rest. Also, study the event log for special
comments relating to the flagged cells.
19. Task Group Actions
19.1 General—If the investigation disclosed no clerical.
sampling or procedural errors, the unusual data should be
retained, and the precision statistics based on them should be
published. If. on the other hand, a cause was found during
the investigation, the task group has several options to
consider. If the laboratory clearly and seriously deviated
from the test method, the test results for that laboratory must
be removed from the 1LS calculations. However, despite the
danger of the recalcitrant laboratory having prior knowledge.
Page 157
-------
E691
3 4-
S
u
2.5
1.5
1
.5
0 -
•
•
•
-
II
LU: 1234567 1234
II 1
>567 123456;
1 1
r 1234S&
7 1
,|
1
21AS&7 123A4
«7 1?^
||
!&<£? 153A5
1 II
67 1234S67
2.03
B C 0 t r C H I
9 PmtMara in Pulp: Je-
ll may be appropriate to ask the laboratory to retest one or
more materials following the correct procedure, and then
include the new set of test results in the ILS calculations. Of
course, if the data have changed, recalculation of the h and k
values must be made and the data consistency Maminfd
again.
192 Exception—When a large number of laboratories
have participated in the DLS and no cause for some unusual
cell values have been found during the investigation, it may
be appropriate to delete a cell tram the study if all of the
other laboratories are in substantial agreement. The number
of laboratories that can be considered large enough to
support deletion of data without an identified cause cannot
be stated exactly. Any action which results in discarding
more than five percent of the ILS data likely will lead to the
presentation of precision data that the test method cannot
deliver in routine application.
19.3 Test Method Vagueness—One of the important
things to be on the alert for during a laboratory investigation
is for vagueness in the test method standard that permits a
wide range of interpretation leading to loss of precision.
Particular elements to check are lack of measurement
tolerances, diversity of apparatus and insufficient direction
for operator technique. These problems can be the hq«is for a
revision of the standard.
20. Examples of Iota
ltd
vy Studies
20.1 Glucose in Serum—The ILS is described in 15.1.1.
20.1.1 A Statistic—The overall impression given by Figs.
1 and 2 and Table 3 is one of reasonable consistency for
variation among laboratories. Only Laboratory 4 stands out
with large values for Materials B and C The graph for
Material C, in Fig. 2, shows that Laboratory 4 is distinctly
different from the other laboratories. The graph for Material
B, however, does not single out Laboratory 4.
20.1.2 k Statistic— Laboratories 2 and 4 stand out in Fig.
3 and Table 4. The laboratory plot, in Fig. 3, indicates
Laboratory 4 has three high values, out a look at the material
plots (Fig. 4) for/I and 5 suggests that Laboratory 4 is not
out of line for these two materials. On the other hand, the
plot for Material C shows Laboratory 4 is different. Simi-
larly, the plot for Material E shows Laboratory 2 is different
for this material.
20.1.3 Cells and Ten Resulu—Cells C4 and E2 should be
investigated. A look at Table 1 reveals that the second test
results of 148.30 in C4 and of 309.40 in E2 are the particular
values to be investigated.
20.1.4 Action—If the data from Laboratory 4 were typed,
the result 148.30 in CeUC4 could have been a typographical
error. We have no way of knowing this today, many years
after this study was made. We will suppose, however, that the
task group did indgfd call the laboratory and did find that
Page 158
-------
E691
1.2
I .6 4
I
•* -4
s
44
«o
.2 4
o 4
o°2
FKL 10
10 12
14 16
U 20
*•*»•». in Pulp: Stem o^irton. of H^codudbUHy (O)««
(« v««» A»«^,
the number should have been 138.30. However, let us
suppose that for Cell £2 the task group
case they should retain the value.
20.1.5 Recalculation—Tables 5 and 6 show the recalcu-
lated consistency statistics resulting from correcting Cell C4.
The riicMie««w« Q( fjjg glucose in serum d^tfl is continued in
20J Pentosans in /^p-Seven laboratories tested nine
*•—-u, obtaining three test results ^^
202.1 hSuuistic-M first glance no one laboratory is
angled out for attention by Fig. 6 or Table 9. For MatS A
Laboratory 7 is different and for Material C UboraSry I
On further inspection of the laboratory plot for Laboratory 7
we note the first five materials are negative andIteSfour
positive. Keeping in mind that the first five materials are
close together in property level while the last tw^emuch
higher in property level, one can see that Laboratory 7 has a
different response to property level than the other laborato-
slronglj " 6 ShOWS ^ reVWSC reSp0nse' but not «
.h^'Hf SMrw/KT-From Fig. 8 and Table 10. it is obvious
that Laboratory 1 is different from the rest with five
materials greatly exceeding, and one near, the critical value
line In addition. Laboratory 7 is different for Material H
20 2.3 Cells and Tea *«uto-Both I^ratorSTand 7
rhould be investigated m depth. Examination of the cell data
' Laboratory I in Table 8 suggests special attention should
given to test results2 in A. 3 in B. 2 in C. 3 in D 3 in E.
and 1 in G, but there appears to be an overall problem in
within-laboratory variability. On the other h^nri. Laboratory
7 has a different problem in not agreeing with the other
laboratories at the two extremes of property IcveL For
Material A the Laboratory 7 test results arc less than half the
values obtained by the other laboratories, while for Material
I the test results are about 10 % higher than the rest This
variation with property level should be explored.
20.2.4 Action—Note that Laboratory 7 reported test re-
sults to three significant digits for all property levels while all
the other laboratories reported to two decimal places. This
difference in reporting would have been a good place to start
the inquiry of this laboratory. It might be the indication of
apparatus differences, or perhaps a sign that the laboratory
may have disregarded other requirements of the test method
or interiaboratory protocol. The apparently poor within-
laboratory precision of Laboratory I, if determined to be due
to improper test equipment or poor maintenance of test
environment might have required omitting this laboratory's
data from the analysis, but with so few laboratories in the
ILS and no physical evidence, the task group should retain
this laboratory's data in the analysis.
PRECISION STATEMENT INFORMATION
21. Repeatability and Reproductbility
21.1 General—QBXX the task group has concluded which
ceils are sufficiently inconsistent to require action, and
action has been taken, the statistics of 15.4 through 15.6 are
recalculated (see also 20.1.5). Using the corrected statistics.
Page 159
-------
calcutatt tor each material the 95 % repeaiabilitv and
rcproduabiiity limits (see Pracnce E 177) according xo the
following Eqs 10 and 11: ,
2.8
(10)
R-2.lsK (11)
21.2 Prepare a table for the corrected precision statistics as
shown in Tables 7 and 11.
21.3 Variation of Precision Statistics with Property Level:
. 21.3.1 Quite often the values ofsr and SK will be found to
vary with the values of the property level x. This type of
response is the case for both examples as can be seen in Figs.
5 and 10. that are based on Tables 7 and 11 respectively. The
manner in which the statistics vary with the property level
should be shown in presenting the precision information in
the precision statement of the test method. The statistician
should recommend the most appropriate relationship to
present, using Practice E 177 as a guide.
21.4 Precision Statement—Table 7 or 11 (with the
column for sx omitted) is a useful format for the presenta-
tion of the precision statement of the test method as required
by Section A21 of the "Form and Style of ASTM Stanoaros
(Blueboofc)". Having obtained the required precision mior-
E691
mauon in accordance with this practice, the final form of the
precision statement may be prepared in accordance with
Pracnce E 177.
21.5 Conclusion—The precision statistics obtained by an
ILS such as described in this practice must not be treated as
exact mathematical quantities which are applicable to all
circumstances and uses. The small number of laboratories
and of materials included in the usual ILS guarantees that
there will be times when differences greater than predicted by
the ILS results will arise, sometimes with considerably
greater or smaller frequency than the 95 % probability limit
would imply. The repeatability limit and the reproducibility
limit should be considered as general guides, and the
associated probability of 95 % as only a rough indicator of
what can be expected. If more precise information is needed
in specific circumstances, those laboratories directly involved
in a material comparison must conduct interiaboratory
studies specifically aimed at the material of interest4
4 FoUowiof me ASTM Raeuca ft
jo tte 1LS to be filed « ASTM Hi
ANNEX
(Mandatory Information)
Al. THEORETICAL CONSIDERATIONS
A1.1 Underlying Assumptions of ILS
Al.1.1 Within-Laboratory Variability—The cell standard
deviation is a measure of the within-laboratory variability of
each individual laboratory. All laboratories are assumed to
have essentially the same level of variability when following
the specified repeatability conditions. This assumption is not
always fulfilled. However, the shorter the period of time in
which the test results for .a particular material are to be
obtained by the laboratories the more likely the validity of
this assumption. Therefore, the laboratory cell variances can
generally be pooled by averaging the squares of the ceil
standard deviations. The square root of this average within-
laboratory variance is the repeatability standard deviation j,.
A1.L2 Between-Laboratory Variability:
A1.L2.1 Variability of Laboratory Means—The test re-
sults obtained on a particular material at any particular
laboratory are considered pan of a population having a
normal distribution with a standard deviation equal to the
repeatability standard deviation but wnh a mean that may be
different for each laboratory. The laboratory means are also
assumed to vary according to a normal distribution, whose
mean is estimated by the average of all ILS test results for a
given material, and whose standard deviation is d^gnated
by SL. (The effect of a single outlying laboratory on this
assumption will be less if there are enough laboratories.) For
the ILS calculations. SL is estimated from the standard
deviation of the cell averages. *„ and the repeatability
standard deviation. sr is as follows:
(Al.l)
Where (sf is the pooled variance for the cell averages of one
manual, and it is the number of test results per odL (sjr is
the observed variance of the average of the cdl averages.
When (j J2 calculates to less than zero. SL is taken equal to
zcxo*
Al.1.12 Reproducibility Standard Deviation— Tbe vari-
ance among individual test results obtained in different
laboratories is the sum of the within-laboratory variance and
the between-laboratory variance of the laboratory means.
Thus, the reproducibility variance is given by Eq ALZ as
follows:
Substituting Eq Al.l Into Eq A1.2 produces Eq Al J:
(A1J)
Simpurying and taking the square root gives Eq A1.4 as
follows (and Eq 7):
When SK calculates to less than s* SK is set equal to s,.
A1.2 Consistency Statistics
AIJ.1 Critical Values— The derivation of the equations
for Mtfyiat^g critical values of h and k are given in M.L2
and A1.2J. In each case critical values were tafoJ^J at
three «g™fi«m«. levels. 1 %, 0.5 %, and 0.1 %. Of these
three only the 0.5 % critical values were chosen for flagging
as described in Section 17. This choice is based on the
judgment from experience that the 1 % values are too
Page 160
-------
sensitive (flag too many) and the 0.1 % values are not
I"10"1* ^ fl38gmg ade"uattly in «ta analysis of
A 1.2.2 Between-Laboratory Consistency:
Al.2.2.1 The consistency statistic h is an indicator of how
one laboratory's cell average, for a particular material
compares with the average of the other laboratories. The
cnticai values for the comparison are calculated with an
equauon derived from an unpaired /-test as given bv Eq Al 5
I/(/>-!)] (Al.5)
where:
7 = observed Student's t value.
xe = cell average being tested.
x* = average of all cell averages except the one being tested.
V = standard deviation of all the cell averages except the
one being tested, and
p = number of laboratories in the ILS.
In this relationship t has p-2 degrees of freedom. Three
further equations are required in order to express h in terms
oft from Eq Al.5. These follow as Eqs A 1.6. A 1.7. and A 1.8:
n (-M.6)
E691
repeatability conditions, on a particular material, compares
with ail of the laboratories combined. Values oik larger than
1 indicate greater within-laboratory variability than the
average for all laboratories. Since such variation among
laboratories is expected, critical values of k have been
calculated to aid in the decision of whether the cell standard
deviation of one laboratory is sufficiently different from the
rest of the laboratories as to require investigation.
A 1.2.3.2 A valid test for determining whether a particular
cell variance is inconsistent relative to the variances of the
other laboratories is to calculate the /"-ratio of the one ceil
variance to the pooled variance of all the other laborato-
ries—excluding the variance being tested. This is shown in
Eq A 1.10 as follows:
Each of these equations is derived by simple algebraic
operations from the definitions of symbols accompanying Eq
Al.5 and Table 2. Combining them with Eq Al.5 results in
Eq A 1.9 as follows:
A - (P - I V/Vri/» + p - 2) (Al.9)
A 1.2.12 The critical values of h were calculated by Eq
Al.9 using published values of Student's t at the 0.5 %
two-tailed significance levels (6). The values obtained are
given in Table 12.
A 1.2.3 Within-Labortuory Consistency:
A12J.I The consistency statistic. ^ is an indicator of
how one laboratory's within-laboratory variability, under
where:
s2 = cell variance of cell being tested.
- = summation of all other variances.
I
l j,)2 = cell variances other than the one being tested, and
p = the number of laboratories.
The consistency statistic k is defined by Eq A 1.1 1 and the
repeatability variance by Eq A 1.1 2 as follows:
k-s/s,
(Al.ll)
Combining Eqs. Al. 10, Al. 11 and Al. 12 results in EqAl. 13
as follows:
A12.3.3 The degrees of freedom for Fin Eq A1.10 are n
- 1 and (p - 1) (n - 1). The upper critical values of £ are
calculated from the upper critical values of fat the 0.5 %
significance level for selected combinations of numbers of
test results and laboratories. The values of k given in Table
12 were obtained using SAS's BETAINV (inverse beta func-
tion) and using IMSL's routine MDF1 (for the F cdf inverse).
Page 161
-------
E691
A2. INDEX TO SELECTED TERMS
Term
accuracy
average of the ceil averages, x
bias
cell
ceil average, .r
cell deviation, d
cell standard deviation, s
fftl WWMMM*a>y ^%JiHpitii W**
reproducibiiity standard deviation. SK
standard deviation of the ceil averages, s,
tmft fmutttinrt*
wai viriniiiivni
test nfiT^^^^^^"^^^^
tnt method
test result, x
test specimeQ
test unit
Section
Intro.. 3.2.4.3
15J.1
Intro.. 3.2.4.3
7.1
15.4.1
15.52
15.4.2
1C A
ijA
15.7
Intro.. 15.7.1
Intro.. 15.7.2
9.
10.
322
Intro.. 3.2.4
15.6
10.1, 20.3
32.1, 12.
Intro*. 3.2.5
Intro.. 32.5.1
21.1
15.6.1
Intro.. 32^
f«*vn
intro.
21.1
15.62
82
8.2
Vumm
15JJ
82
^ •* t
33~i
32.1 8
WMM&* U*
322.
32J
32J
REFERENCES
(1) Youden. W. L. "Experimental Design and ASTM Committee." (3) Pante. R. C, Marinenko. G- KnoenteL M, and Koch. W. F..
Maunais Research and Standards. ASTM. November 1961 pp. "Rntjednm Testm»—Pan £ RT-""t inttftrrimn " Journal
*6i-«67. of Research of the NBS. VoL 91.1986. pp. 9-15. „._.___.
(2) Pante. R. C, Marinenko. G» KnoerdeL M^ and Kocfa. W F W Duncan, A. J, Quality Control and Industrial Statisais. KdarA
~ ' ~ : Ignoring Interactions.- Journal rf D. Irwin Ine, Homewood tt. 5th edition. 1986.
^u^uuncaons. Jo«ma/o/ (J) !,„„„. O.J. and OaifcV. A»^ii«/)«w^Kwia«e«irfJ»ef«no«.
* W' John VWtey and Sou. 1974.
«vmeft you IIMV MMM. tt^utttittm your eamniMiMw nor
to «• ASTM ComnMTM on SMncwn. »9» 0 Am fit. PH*****. PA 19103.
Page 162
-------
RELATED MATERIAL
^^Te^SSfS^^S, ?firniflked "P™"08*" ^Published for information oniy.Tliey have
officially accept* 5? SB &STSI ^ comminee for P«Wic«ion as "proposed" but nave no? been
Philadelphia, Pa. 19103 Comments are solicited and should be addressed to ASTM, 1916 Race Su
Page 163
-------
Designation: £ 1267 - 88
Standard Guide for
ASTM Standard Specification Quality Statements1
wpenmpt cpaton I.)
E 12«:i
of icvnoa. ibe year of last revmon. A aunterap
idkaia the year of
.
u> editorial chime ante toe last rrvmoo or itappiovat.
t&eycarofittieappfwl. A
1. Scope
1.1 . This guide is intended as a reference to assist ASTM
technical committees and subcommittees addressing quality
statements in product specifications under their jurisdiction.
1.2 It is recognized that quality provisions are not re-
quired in every standard specification. Lack of a quality
provision does not indicate a deficiency in the standard.
1.3 This guide addresses the following areas and provides
a check list of factors to be considered for each topic
calibration and measurement; inspection and testing; ban-
dling, storage, preservation, and shipping; nonconformine
materials; and documentation.
1.4 This standard may invoke hazardous materials oper-
ations. and equipment. This standard does not purport to
address all of the safety problems associated with its use It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use.
2.1 ASTM Standards:
E lH£ys* for Use «f ^ Ter°» Precision and Bias in
ASTM Test Methods^
12 ANSI/ASQC Standard:
M-l Calibration Systems3
2.3 Military Standards:
MIL-MS208 Inspection Systems Requirements4
MIL-Q-9858 Quality Program Requirements*
MIL-Q-21549 Product Quality Program Requirements
for Fleet Missile Weapon Systems Contractors4
MIL-STD-45662 CaUbnhon System Reauin
ft i-t?^ ___ 99 ___ l«_ _ _ •_ •* ... I|HH*
Military Handbook 52A (Interpretation of MIL-STD-
45662)*
3. Terminology
3.1 Definitions—Refer to Terminology E 456 for defini-
tion of terms other than those listed in 3.2 which are used in
this standard guide.
3.2 Descriptions of Terms Specific to This Guide:
3.2.1 calibration—comparison of a measurement stand-
ard or instrument of known bias with another standard or
1 This sunk is under the junadkMra of ASTM ComimneeE.il oo Quality and
Siauma in ASTM Standard! and is the daect respoaability of SubenmnmM
EU. 10 on Quality Document Prepanaon. n.mmnc
Cuncm caaon approved Nov. I. I9M, Pabfebtd January |«I9
2.4amM< Book of ASTM Stmutardt. Vol 14.02.
1 Available from Society for Quabty Coatrat 310 W. Wavm A»e_
Milwaukee. Wl 53203. ^~
• Available from Naval Pubucanoas and Fonm Ccater. iMI Tabor
Ph.Uddphia. PA 19120. "*
instrument to detect, report, and/or minimize by adjustment
any unacceptably large bias of the item being compared.
3.2.2 traceability (calibration sense)—the ability to relate
individual measurement results to national standards or
nationally accepted measurement systems through an un-
broken chain of comparisons (see MIL-STD 45662).
3.2.3 raw material—any material intended to undergo
change when introduced into the process.
3.2.4 component—any material that is incorporated into
the final product without undergoing any significant change
in the manufacturing process.
NOTE I—Components, (for example, electrical cables, plastic con-
nectors), generally require control measures equivalent to those needed
for the final product.
3.2.5 nonconformonce—deviation of a material to some
degree from one or more of the technical requirements of a
standard.
NOTE
du
y deviate
nation, without retard to the degree to which a pi
from speofied omits. An individual item may be nonconwnning. and
must be treated as such, even if h comes from a tot which meets an
agreed sampling plan acceptance omit
3.2.6 fitness for use—suitability of a product for its
intended
Nora 3—Fttnen for use is a soinewhat siiojecnve concept in which
of deviation from opnmum become* important, and may not
4. Significance and Use
4.1 In view of the great diversity of ASTM specifications.
it is not feasible to recommend a single set of suitable quality
statements, nor even to develop a small number of state-
ments with rules for selection. F"*** committee or subcom-
mittee must consider the need for quality statements for its
specifications. This guide is intended to simplify the process
by ruling attention to the considerations that should enter
into development of specification quality statements.
5. Calibration and Measurement Precision and Bias
5.1 Check list for calibrauon and measurement precision
and bias:
5.1.1 Identify the most important measurements (with
respect to control and assurance of quality).
5.1.2 Determine limits of acceptable measurement error
in light of product tolerances, and specify maximum accept-
able measurement uncertainty for the measurement uncer-
tainty can be shown to be negligibly small compared to the
tolerance being »«"«•** it should not be neglected in setting
acceptance limits. That is. the tolerance band should be
narrowed by the uncertainty. For example, if an ASTM
standard specification fr>llt for a dimension that is 20 cm ±
Page 164
-------
E1267
0.1 cm. items falling in the range 19.9 cm to 20.1 cm would
be accepted as being within specification if the measurement
uncertainty were negligibly small. If the uncertainty of
measurement were estimated to be 0.05 cm, the range of
acceptable dimensions should be shrunk to 19.95 cm to
20.05 cm to allow for the measurement uncertainty.
5.1.3 Identify the most likely sources of bias and impreci-
sion in the measurement process and consider providing
guidelines on how to minimize these sources of error.
5.1.4 Consistent with the desired uncertainty limits, rec-
ommend calibration procedures where it is appropriate to do
so and specify how traceability to national standards is to be
realized (if applicable).
5.1.5 Identify particularly important measurements for
which special measurement assurance procedures might be
specified and consider specifying suitable procedures.
5.2 Discussion:
52.1 Defining Measurement Uncertainty for Important
Measurements—Measurements important with respect to
controlnnf* assurance of quality should be flagged during the
standards development process. Some variation in product
parameters is inevitable, so upper limits of permissible
variation (product tolerances) are established. It must be
iBcognited that, in the general case, the measurement system
used to quantify these parameters will also have some
imprecision and bias, and limits to this variability and bias
must also be established. Measurement errors generally
consist of a random emu component (imprecision) and a
systematic error component (bias). The uncertainty of a
measurement is the best estimate of the upper bounds to a
suitable combination of these two error sources. (See Practice
E 177, the Appendix to ANSI/ASQC M-l, and Ref (I)5 for a
more comprehensive treatment of this topic.)
5.2,1.1 Measurement uncertainty should be small com-
pared to the tolerances in the product parameters being
ac«*wd for compliance with the specification. For example,
a frequently-used rule of thumb is to limit measurement
uncertainty to no more than 10 % of the tolerance of the
parameter being measured. Unless the measurement uncer-
tainty can be shown to be negligibly small compared to the
tolerance being assessed, it should not be neglected in setting
acceptance limits. That is, the tolerance band should be
narrowed by the uncertainty. For example, if an ASTM
specification standard calls for a dimension that j$ 20 cm ±
0.1 cm, items falling in the range 19.9 cm to 20.1 cm would
be accepted as being within specification if the measurement
uncertainty were negligibly small. If the uncertainty of
measurement were estimated to be O.OS cm, the range of
acceptable dimensions should be shrunk to 19.95 cm to
20.05 cm to allow for the measurement uncertainty.
5.22 Traceability—Where national measurement refer-
ence standards exist, measurements should be consistent
with those standards, that is, bias should be acceptably small
(As a simple example, where an ASTM specification calls for
measurements in metres or kilograms, all parties using the
standard must use the same metre or kilogram unit if the
specification is to be meaningful.) To help ensure tM*
consistency is achieved, procurement requirements or regu-
lations sometimes contain provisions to the effect that
"measuring and test equipment must be calibrated with
standards that are traceable to the National Bureau of
Standards" or that "measurements must be traceable to
NBS." The intent of such citations is to ensure accurate
measurements by avoiding measurement errors due to biases
relative to national standards Ref (2).
522.1 ASTM committees should not include statements
along the lines of "all measuring equipment must be
calibrated traceable to the U.S. National Bureau of Stand-
ards. NBS. or its equivalents in other countries," without
further explanation as to what constitutes adequate trace-
ability. A preferred approach is to discuss required measure-
ment uncertainty and provide suggestions on how to achieve
an acceptably low level Requiring traceability of calibration
to NBS or the use of NBS Standard Reference Materials may
be appropriate in some circumstances while the use of
commercial reference materials and/or commercial calibra-
tion services may be appropriate in other cases. It should be
noted that national standards for common measurements
such as mass and length are highly compatible around the
world, whereas national standards for some more obscure
measurements may vary considerably from country to
country. In such cases, it may be necessary to specify which
national standards constitute an acceptable reference tor
traceability purposes.
52J Specifying Measurement Techniques-Oatx the
measurements to be controlled are identified and the permis-
sible uncertainty specified, attention should torn to methods
to ensure that measurement errors are less than the thresh-
olds specified. It may be helpful to prepare a table such as
Table 1 to display such information. A table may be
included in the standard itsdf. but to avoid stifling innova-
tion, a committee should avoid mandating particular calibra-
tion methods, calibration intervals, etc. to the exclusion of all
others unless there are sound technical reasons for doing so.
As outlined in ASTM Standard E691, interiaboratory
comparisons (round-robins) should be carried out to assess
TABU 1
IbcUMdbyABTM
(Accuracy
±OXXXani
ircfertocbBof
the tad of
±XXBrM
ASTMSanaw*
XYZ
Page 165
-------
E1287
typically-achieved precision and bias of the measurements of
interest unless such information is already available. ASTM
Committee E-l 1 can assist in the design of such studies and
should be asked to hdp.
5.2.4 Conditions of Calibration—Instruments are often
calibrated to manufacturer's specifications, but this may not
be cost-effective because the manufacturer's specifications
may be more stringent than the application calls for, and
high-accuracy calibration is expensive. Calibration to lesser
accuracy is appropriate as long as adequate measurement
accuracy is attained for the measurement of interest and the
actual calibration uncertainty is apparent to the user of this
instrument. Multi-range instruments need only be calibrated
in the range(s) to be used, but the calibration conditions
must be readily apparent to the user. Where appropriate.
committees vM consider recommending a minimum
number of calibration points. While calibration is normally
an activity designed to Quantify an^ ptmfyntrK bias, calibra-
tion ^g*»gpf that involve sufficient redundancy may be used
to a***** and monitor precision also.
3.2.4.1 When an organization uses a contract calibration
service, it is necessary to ensure that the service is competent
and M**ng calibration procedures an fraction of m***"**™^?*** whose uncertainty
exceeds the acceptable limit*. In the absence of other
information concerning suitable calibration intervals, the
recommended intervals ^yuousnoo oy the L^eDarcment of
Defense Ref (4) can be a good starting point
5 2ffi Recalls—When instruments are found to be out of
tolerance upon recalibranon, corrective action may or may
not be necessary, depending on the nature of the measure-
ment and how much the instrument is found to be out of
tolerance. When an instrument used to niMMirp a parricu-
larry important parameter is found to be appreciably out of
tolerance, a produce recall might be justified for any product
measured uy»flg th'* instrument sine* the last intolerance
verification. Committees should consider providing guide-
lines for product recalls resulting from the discovery of
out-of-tolerance measuring instruments.
5.2.7 Special Measurement Assurance Programs—Since
recalls or product liability lawsuits can be costly, consider-
ation should be given to defining a more elaborate measure-
ment quality assurance program for measurements of partic-
ularly important parameters (such as. those affecting health
and safety). Such programs should include the regular use of
stable check standards, bund tests, and control charts to
monitor the performance of the measurement system in real
time. It is possible to develop a measurement quality
assurance program that has a built-in -go/no-go" decision
for each important measurement Ref (5).
6. Inspection and Testing
6.1 Check List for Inspection and Testing:
6.1.1 Identify characteristics to be inspected or tested with
reference to importance.
6.1.2 Define the method of sampling.
6.1.3 Consider need for special inspection and test proce-
dures. and any special preparation procedures.
6.1.4 Consider environmental requirements for inspec-
tion and testing..
6.1.5 Identify equipment by type, and
6.1.6 Establish precision, calibration, etc. (see Section 5)
and need for personnel certification.
6.2 Discussion:
6.2.1 Important Characteristics— Characteristics impor-
tant to the normal end use(s) of the product or those most
likely to fluctuate should be identified in order to focus
attention where it is most beneficial. A standard may
recommend increased attention to such characteristics.
6.2^ Method of Sampling— When the inspection or
testing is not to be performed on 100 % of the material, the
unit of sample should be defined. Where possible, it should
be related to a logical process such as a manufacturing
process, a source of materials or storage and handling
operation. For example, a batch is a logical and powerful
sampling lot for batch manufacturing processes. For contin-
uous processes, or for products not traceable to batch,
sampling techniques representative of the process 1»>u'dJ~
used. Statistical sampling plans are often the most powerful
tool for representative sampling for both lot-by-lot and
continuous processes. Reference material is available to aid
in determination of the type and amount of sampling
required and to provide Tatnp*i"g plans to fit various needs.
ASTM Committee E-ll should be consulted to identify
reference material to meet the require-
«pm i
ments of a standard or for reference within the standard, or
both. When it is advisable for a sampling plan lo be designed
by a TPtwki»«v the standard should so recommend.
Non
of acceptance criteria for
pj«ni in a
touted fern ana
aandaid is pntibned by the ASIM «yie
Style far ASTM Standards. (Blue Book).
6 23 Special Procedures— Any special or non-standard
inspection and test procedures must be described in detail.
but standard procedures may be simply referenced. Special
pie-test preparation procedures, including cleaning, surface
preparation, temperature soak, rnachining, and mounting
must be described in detail
6 .2.4 Environmental Conditions— Temperature, humid-
ity, lighting, cleanliness, and safety cooperations must be
addressed as necessary to ensure that the inspection results
property reflect the characteristics covered by the istandard.
6.2J fip«pm«tf-Equipmenl should be adequate to
achieve required precision. Specifications should leave as
much flexibility as appropriate to users, but at least one
acceptable type of equipment should be mentioned where
there is doubt as to what is adequate. Generic descriptions
are to be used in preference to brand names. Handungor
equipment should be consistent with the inspection needs
and nature of the equipment and, any special handling
required must be adequately described. •
Page 166
-------
6.2.6 Personnel—Standards should address minimum in-
spection and testing personnel qualifications and training
where appropriate. Where personnel certification is required.
it should be referenced to the certifying organization or
document, or should be described in detail.
7. Handling, Storage, Preservation, and Shipping
7.1 Check List for Handling Provisions:
7.1.1 Consider need for measures to maintain quality and
•' condition of the products.
7.12 Consider need for measures to prevent mixups.
72 Check List for Storage and Preservation Provisions:
7.2.1 Consider need for specification or control of envi-
rQplpgntai COOCUuOHS*
7.22 Consider need for protective measures to maintain
quality and conditions of materials in storage.
7.2J Determine whether quality or fitness for use of the
product deteriorates over time, and whether control mea-
sures are required.
7.2.4 Consider need for measures to prevent mixups.
7.3 Check List for Shipping Provisions:
7.3.1 Consider need for protective measures to maintain
quality ntir* condition in shipment.
732 Consider regulatory req
rements
to
7JJ Consider whether inclusion of technical
measures to ensure their inclusion.
7.4.11 Handling Provisions—Products produced to ASTM
spmal packaging, instructions for propeTmSiariS of
orboth,
of
lots of product, or between
prevent mixupsjuch procedures may
segregauon of product, but this should not be
the only acceptable means to prevent mixup
mental conditions that may require control include
IT S*" J^""*"*1* "*•••*» or
humidity, and intensity or type of light.
E1267
7.4.4 Storage Provisions— Maintenance of satisfactory
conditions in storage generally requires consideration of all
the measures discussed in 7.4.1. There may also be a need to
control clearance between the material or packagCi or both.
and warehouse floors, walls, or ceilings. Committees should
also consider the need to both prevent and eliminate
infestation by rodents, insects, bacteria, etc., and where
appropriate, to specify suitable materials, methods, and
procedures.
7.4.5 Deterioration with Time—ASTM committees
should consider the likelihood and possible extent of deteri-
oration with time for products included in their Trap**3"4*
Where such deterioration is severe, establishment of a stock
rotation system to assure consumption of the oldest product
first or establishment of expiration dates may be required.
7.4.6 Shipping /"roro/orn—Shipping conditions may be
more severe than the conditions encountered in prior
handling or storage. The considerations of 7.4.1 and 7.4.3
apply, but in addition, special attention may be required to
product containers and shipping containers to ensure that
product protection is adequate under any anticipated condi-
tions. Container testing programs and monitoring of actual
shipping conditions may be necessary for sensitive products.
7.4.7 Regulatory Requirements—Regulatory require-
ments may apply to shipping, though normally only where
some environmental or safety hazard is involved. If appro-
priate, measures to ensure compliance with regulatory ship-
ping requirements may be incorporated into standards.
7.4.8 Document /ncfanon—Some products may require
inclusion of tf**twnA documents, instructions for '^»«n»iia.
tion, instructions for use, etc* to ensure proper and safe use
of the product If any such documents are required to be
jnrfoidfxf with the product, a
-------
*tef*?»t
products with a multipiicixy of uses, the standard should
require users to make those determinations.
8.22 Identification of Nonconforming Materials—When
materials are found to be nonconforming, they must be
positively identified as such to prevent inadvertent ship-
ments or \\Tt while fa"*! disposition is being determined.
This can be a very difficult task for some continuous
processes. Identification of nonconforming materials can
take many forms, including but not limited to physical
segregation, marking of the material. anr* identification of
the nonconformance by means of documents. If a committee
wishes to establish a particular means of identifying noncon-
forming material or to emphasize the need for some positive
identification scheme, the standard should so state. A
specification statement regarding identification of noncon-
formance may be particularly important for products manu-
factured by continuous processes,
8.2.3 Material Renew Procedures— Nonconforming ma-
terial is nearly always reviewed for final disposition, whether
the review is formal or informal Possible dispositions
include soap, reworking to specification, assignment to
less-demanding applications, or use as is. This last disposi-
tion, to use nonconforming material for its originally in-
tended purpose, requires considerable care and thought.
Such dispositions can only be based on fitness for use. Often
the disposition is based on the fact that the nonconformance
will no longer exist by the point of end use, as with
ofl-dimension in a T"Ttiil mill product to be ^Mniiw1^ by
or other PTPP^WTIB^ At other times a noncon-
forming product may be used when it is determinedtnat the
nonconformance does not make the product unfit for use. It
is essential that the user take pan in any decision to use
nonconforming matrriat, especially when the noncon-
formance will still exist in the finished product. Certification
of material accepted on this basis must indicate the noncon-
formance. Products not conforming to standards should not
be supplied without approval of the purchaser.
8.2.3.1 An ASTM committee may wish to consider spec-
ifying a particular procedure for reviewing nonconforming
materials A traceabutcy system for such mati-ti^k and even
for rjresumed-satisfactory products, so that problems arising
from oonconformances may be more easily tracked, may
fCOOOUDCDOCu
8.2.4 Corrective Action—\\ is simply good quality assur-
ance and quality control practice to attempt to identify an
cause for any significant recurring noncon-
formance. and then to take corrective action to eliminate this
assignable cause. It is only in this manner that high quality
levels can be attained and maintained economically.
8.2.4.1 Normally identification and correction of noncon-
formance causes is purely a management function, and not a
concern of a standards-writing committee. However, a high
level of nonconformances due to failure to take corrective
action can affect the user, due to nonconforming materials
escaping detection (no inspection is actually 100 % reliable),
delivery schedule slippage, or economic pressures. Therefore,
for sensitive products or applications, a committee may
consider requiring at least the existence of a corrective action
system, though it is preferable not to specify details. Provi-
sions for corrective action are a pan of many military
is tote
m ttaetracfcsr power iuduaUfyccar
9. EX
9.1 Check List for Documentation:
9.1.1 Consider need for documenting quality of supplies
or raw materials, or both.
9.1.2 Determine need for documentation of in-process
quality testing.
9.1.3 Determine need for documentation of final product
testing. . - .
9.1.4 Conridet documentation of product quality to be
provided to the user, and ,
9.1.5 Consider documentation of audits of the vendors
facilities by the buyer or other interested party when re-
quired.
9.2 Discussion: .
9.2.1 Supplies and Raw Materials—V quality of supplies.
components, or raw materials (see 3.2J) is considered
important to the product covered by a specification, then tne
quality pf such supplies or materials should be documented.
In general documentation should include proper specifica-
tions in purchase agreements and either a vendors cernnca-
tion of quality (perhaps backed by audit), or results of testing
in the user's facility. It may also be necessary to contro
identification. handling, or storage of such material until
incorporation in a final product. An ASTM committee may
choose only to indicate the important of such documenta-
tion, or may specuy its form. .
922 In-Process Testing—It may be miportantto deter-
mine product quality at one or more
before going on to further stages.
ifiifif* fi Docu
location and test results may be required,
92J Final Product Testing—Most ASTM
specify final product testing and mspecnon
Good practice requires estiMHhment of the
bias of ten and inspection methods and <*OCL_
the calibration status of measurement equipment, as dis-
cussed above. Test and inspection
C1BXB aUUVC. 1C3V OUM mi|»n •""- •—- - - ..
/ton>TT"'"tcd, and records kept for a suitable penoa.
mhtees may wish to specify the documentation "^
9.2.4 Certification to User—* manufacturer may oe ex-
pected to provide some certification or report, orbotn. ottesi
results to the user. Certification may range from a sunpie
statement of compliance to a specification »J™"«
documentation of all tests and inspections. Ccmmtttees
should consider and specify the degree of documentation to
the user required. Committees should attempt to specuy oniy
the documentation actually beneficial to the user, and avoid
requiring extensive documentation. If the ^numufitturer•*
required to maintain documentation in his nies tor rarer
reference as needed, then documentation to the user can be
m.n.m.T^ for example, results of dimensional inspection
of mill products are rarely reported to users though they may
be recorded in the manufacturer's internal 'tocumentauon.
923 Audits—An ASTM committee should consider
whether an audit of the manufacturer's
required. Such audits are time^onsummg and
can be useful especially for technically sophisticated
Page 168
-------
E1267
ucts in crucial applications. If an audit is to be required, then documented and this documentation will be used to deter-
the manufacturer's quality assurance program must be mine compliance.
TABLE 2 CtMddurt to Oatanmn* V* Conalatancy of Any ASTM Standard WHh 6-4« OuktoUnM and ASTM Fonn and Styt* Manual
(MM Book)
wan tt» omocm el Mi goat.
("tnoMDiHy to NBS i UvitfiBd. •••* —• ^HB ^~*
I tonHrMnga en packagng and I
> or irwBueBona wmi antnama la appnju«i» i
(a) OaowmnangquMiyod
W
(C| OOGURHMMIBn Of 11111 1
i and/or raw
ef product quMly to &• provian to tr« UHT
REFERENCES
(1) "Mewuitmem Assurance Prognms," NBS Special Technical Pub-
licaoon 676: Pan I General Introduction. May 1984. Bdaaaer B.
C: Pan II Deveiopment and implementation. April 1984,
Crouton. M. C '
(2) Bdanaer. B. C. Traceabiltty. An Evolving Concept,- ASTM
StaadndiBBioB Newi. January 1980. pp. 22-28
(3) Bfmmmmrted Pncoce Number 2. National Confem^ of *,.-*-
La
bomonea. "Evaluanon of Measurement
(4) "Calibration and Repair Requirements for the Maintenance of
Amy Material." DcpBtment of the Anny Technical Bulletin
TB4J. 180-1; Technical Manual: "TMDE Interral Calibration and
Repair Tech Older Reference Guide and Work Unh Code
Manual.- Air Force Techmcai Order Number TO-33K-1-IOO:
"Maroiofy Reqmremeaa Ua." Department of the Navy Publica-
tion NAVAIR I7-35MTH. SPAWAR P4734-31-0001.
NAVSEA OD 4S84S. and Marine Corps TM733-15/13.
..,
for Gage Blocks," NBS Moooanpb 163. February 1979
Page 169
-------
E1267
M tna mat O» wwMwtf «*»y ft* »•*» tna
or
-------
Designation: D 3960-89
Standard Practice for
Determining Volatile Organic Compound (VOC) Content of
Paints and Related Coatings1
This standard is issued under the fixed designation D 3960: the number immediately following the designation indicates the year of
original adoption or. in the case of revision, the year of last revision. A number in parentheses indicates the year of last itapproval. A
superscript epsilon (0 indicates an editorial change since the last revision or reapproval.
1. Scope
I.I This practice covers the use of available ASTM
methods for determining the volatile organic compound
(VOC) content of paints and related coating materials and
serves as a guide for the selection of the proper ASTM test
method for determining VOC.
1.2 This practice does not describe the actual procedure
for measuring total volatile organic materials volatilized
from a coating film or for determining the composition of
the volatile fraction. It does offer a simple method of
calculating VOC.
1.3 This practice includes all organic materials volatile
under the test conditions. Allowances have been made for
organic materials that are considered to be negligibly photo
chemically reactive. For the purpose of brevity in this
practice these materials are referred to as exempt
NOTE 1—For applicable exempt solvents, see appropriate Federal,
state, and local regulations.
1.4 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the safety problems associated with its use. It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use.
2. Referenced Documents
2.1 ASTM Standards:
D1475 Test Method for Density of Paint, Varnish, Lac-
quer, and Related Products2
D2369 Test Method for Volatile Content of Coatings2
D2697 Test Method for Volume Nonvolatile Matter in
dear or Pigmented Coatings2
D2832 Guide for Determining Volatile and Nonvolatile
Content of Paint and Related Coatings2
D3792 Test Method for Water Content of Water Reduc-
ible Paints by Direct Injection into a Gas Chro-
matograph2
D4017 Test Method for Water in Paints and Paint
Materials by Karl Fisher Method2
D4457 Test Method for Determination of Dichloro-
methane and 1,1,1-Trichloroethane in Paints and Coat-
ings by Direct Injection into a Gas Chromatograph2
1 This practice is under the jurisdiction of ASTM Committee D-1 on Paint and
Related Coatings and is the direct responsibility of Subcommittee D01.2I on
Chemical Analysis of Paints and Paint Materials.
Current edition approved Jan. 27. 1989. Published March 1989. Originally
published as D 3960-81. Last previous edition D 3960 - 87".
2 Annual Book of ASTM Standards. Vol 06.01.
E 180 Practice for Determining the Precision of ASTM
Methods for Analysis and Testing of Industrial
Chemicals3
3. Definitions
3.1 nonvolatile content—of a coating, the solid material
remaining after the volatiles have been driven from the film
under specified test conditions. The total percentage volatile
present is obtained by subtracting the nonvolatile content
from 100.
3.2 volatile organic compound (VOC) content—in coat-
ings, any compound of carbon that evaporates from a paint
or coating film under specific test conditions. Water and
exempt solvents that are volatile are not included as VOC.
VOC may be determined at a specified temperature for a
specified baking or air dry time and at a controlled film
thickness.
4. Summary of Practice
4.1 The nonvolatile content and density of the coating are
determined in accordance with specific methods. For solvent
reducible coatings, shown by tests to contain no water, the
volatile content is equivalent to the VOC For water-
reducible coatings, the water content is determined sepa-
rately and subtracted from the total volatile content The
percent water should be determined on all water-reducible
coatings and on all solvent-reducible coatings thought to
contain water since water is volatile under the conditions
used for determining nonvolatile content and, if present,
would be included in the volatile organic content The VOC
is calculated as grams per litre of paint or coating.
4.2 Since some solvents are exempt they also should be
determined and subtracted from the total volatile content
and not included in the VOC calculations (see Note 1).
4.3 As the applicable ASTM standards show, different
types of coatings are heated at different temperatures and for
different times in determining nonvolatile or volatile content
and results may vary with test conditions. It is imperative,
therefore, to define clearly all test conditions used to
calculate VOC.
5. Significance and Use
5.1 This practice discusses applicable ASTM test methods
needed to determine the volatile organic compound content
of paints and related materials and provides the equations for
calculating the VOC. VOC results are required by various
regulatory agencies.
3 Annual Book of ASTM Standards, Vol 15.05.
Rage 171
-------
D3960
6. Nonvolatile Content
6.1 In Guide D2832, the methods for determining the
nonvolatile content of most raw paint materials, finished
paints, varnishes, lacquer and shellac are d'sqi'sscd and
referenced.
6.2 For general paints Guide D 2832 recommends the use
of Test Method D 2369.
6.3 When reporting the nonvolatile content, also report
the baking schedule.
7. Water Content
7.1 To determine the water content of water-reducible
coatings two test methods are available:
7.1.1 In Test Method D 3792, a paint specimen is diluted
with dimethyl formamide, an internal standard (2-propanol)
is added, and an aliquot of the mixture is injected directly
into a gas chromatograph.
7.1.2 In Test Method D 4017, a paint specimen is diluted
with pyridine and a catalyst (1-ethylpiperidine) is added to
obtain a rapid and definitive titration end point.
8. Density
8.1 The density of the paint or coating at 25'C is
determined in accordance with Test Method D 1475. Al-
though both pycnometer and weight-per-gaUon cup are
covered by the test method, and the former is more accurate
and precise, the weight-per-gallon cup is recommended be-
cause of its speed and ease of use.
9. Exempt Solvents
9.1 In Test Method D 4457 an internal standard (1-propa-
nol) is added to a paint specimen, and then injected directly
into a gas chromatograph.
10. Calculations of VOC
10.1 Solvent-Reducible Coating—For solvent-reducible
coatings that are known or believed to contain water, use the
calculations in 10.2. Calculate the VOC of solvent-reducible
coatings in grams per litre according to the following
equation using results obtained from ASTM test methods:
where:
K, = total VOC, g/L (organic volatiles),
N - % nonvolatile,
Dm = density of the coating, g/mL, at 25*C, and
10 = factor for converting results to g/L.
To convert grams per litre to pounds per gallon, multiply
by 8.345 x 10~3.
10.1.1 When exempt solvents are present they should be
treated as water and not be in VOC calculations in accord-
ance with the following:
B-(y2-Ex)(DJlO
where:
B = VOC in coating containing exempt solvent g/L, and
Ex = exempt solvent weight %.
This is the same equation used in 10.2.3 except W(% water)
is replaced by Ex.
10.1.2. The volatile organic content of a solvent-reducible
coating containing exempt solvent can also be calculated on
the basis of the liquid coating without its exempt solvent
content Use the following equation to determine VOC of
solvent-reducible coating minus exempt solvent:
5100
100 -
where
B
= VOC, g/L, coating minus exempt solvent,
- density of exempt solvent at 25*C, and
* VOC in coating containing exempt solvent, g/L
(from 10.1.1).
This is the same equation used in 10.2.4 except Dw is
replaced by Z>CT; A by B and w by Ex.
10.2 Water-Reducible Coating:
10.2.1 Calculate the amount of volatile in the coating in
grams per litre as specified in 10.1.
10.2.2 Determine the weight percent of water in the
specimen in accordance with 7.1.1 or 7.1.2.
10.2.3 Calculate the total VOC content as follows:
where:
A - VOC in the water-reducible coating, g/L,
V-i = total volatiles (including water), weight %, and
W = water found in the coating, weight %.
To convert grams per litre to pounds per gallon, multiply
A(VQQ by 8.345 x 10~3.
10.2.4 The volatile organic content of a water-reducible
coating can also be calculated on the basis of the solids and
organic volatile contents alone, that is, the liquid coating
without its water content The intent of such a calculation is
to place the VOC of a water reducible coating on a basis
similar to that of a solvent reducible coating. To avoid
confusion with VOC defined in 3.2, this quantity is desig-
nated y^ VOC of a water reducible coating minus .water
can be determined by the following calculations:
v -4'OQ ' '
-
where:
V^ - VOC, g/L coating minus H2O,
£H,O - density of H2O at 25*C (0.997),
NOTE 2—The derivation of the above is
f m, (VOC of coating minus water)
KIOO-AWAJl-
x 1000
total volatile* (g)-water (g)
total paint (mL)-water (mL)
x 1000
10.3 The volatile organic content of solvent and water re-
ducible coatings can also be calculated as mass of VOC
emitted per unit volume of applied coatings solids. This
quantity will be designated as Cm.
10.3.1 Calculate Cm using the organic volatile content
density of the coating, and volume % solids of the coating
using the following equation:
where:
Cm = VOC content as VOC mass/applied coatings solids,
H'v « organic volatile content weight %,
Wy . p, _ w,
V2 m total volatiles (including water), weight %, and
Page 172
-------
D3960
W m water found in coating, weight %.
density of the coating, and
volume % nonvolatile content of the liquid coating as
determined by Test Method D 2697.
To convert grams per litre to pounds per gallon, mulitply
by 8.345 x 10~3.
APPENDIX
(Nonmandatory Information)
XI. AUTOMOTIVE COATINGS SUPPLIERS ROUND ROBIN
Xl.l A round robin was conducted at the laboratories of
automotive coatings suppliers for determination of VOC
using Practice D 3960. The analysts involved were persons
experienced in running all the test methods involved in VOC
determination. The data was analyzed statistically in accord-
ance with Practice £ ISO. As was suspected from previous
round robins conducted to evaluate Practice D 3960 (which
involved some laboratories not familiar with these test meth-
ods), when well experienced analysts conduct the tests, the
precision data is much improved.
XI.2 The interlaboratory study involved four laboratories
and six samples; four solvent-reducible automotive topcoats
and two water-reducible automotive topcoats. One operator
in each of the four laboratories analyzed the sample in dupli-
cate on 2 different days. The following duplicates, repeata-
bility, and reproducibUity coefficients of variation were ob-
tained.
Automotive Topcoats Duplicates, %
Repeatability. % Reproducibility,
(Within (Between
laboratory) 1 aboratory)
Soivent*roducibte
Water-reducible
0.86
3.94
1.62
5.29
9.75
The American Scatty lor Jesting and Materials fakes no position respecting the validity of any patent rights asserted m t
HriBi any turn mentioned in this standard. Users of this standard an expressly advised that determination el the validity at any such
patent nffhts, and the nsk of inhuiyeiHeiit of such rights, are entirely their own responsibility.
This standard is subject to revision at any time by the responsible technical ce
It r&rev»ed. either reepproved or withdrawn. Your cornmeru are imind either tcf revision ot this standard or tor edi
and must be renewed every five yean and
I standards
and should be addressed to ASTMheadduafters. Your comments win receive careful connderetlen at a meeting el the responsible
technical committee, which you may attend. If you teal that your comments have not received a laJr hearing you should malte your
views Known to the ASTM Committee on Standards. 1H6 Race St.. PhOadelphlt. PA 10103.
Page 173
-------
Designation: 0 1475 - 85
Standard Test Method For
Density of Paint, Varnish, Lacquer, and Related Products1
™»»*^««'« «w*r the fiud dowunon D I47S: the number tmmcm^y foltow,n, tfic vcwm«er—Any type, or weight-per-gallon cup,
having a capacity of from 20 to 100 mi. may be used,
provided that it may be filled readily with a viscous liquid.
adjusted to exact volume, and covered to exclude loss of
volatile matter.
62 Thermometers, graduated in 0.1*C such as are sup-
plied with glass pycnometers.
6.3 Constant-Temperature Bath, held at 25 ± 0.1%. »s
desirable.
6.4 Laboratory Analytical Balance.
NOTE 1—The usual weignt-per-gallon cup and
pycnometen may have filled weights that exceed the capacor of the
anal laboratory aaatyticai frilir~ la such cases, use of a hantint pan.
triple-beam *«»••>• with KilT graduated to 0.01 ( has been found to
provide results the mean of which was comment with the overall
prrrnioo and accuracy of the method.
6J Desiccator and Desiccated Balance, or a room of
reasonably constant temperature and humidity are desirable.
7. Calibration of Pycooineter or Cup
7.1 Determine the volume of the container at the sped-
fied temperature by employing the following steps:
7.1.1 Oean and dry the container and bring it to constant
weight Chromic acid (see 7.1.1.1) cleaner and nonresiduai
solvents may be used with glass containers and solvents with
metal containers. For maximum accuracy, conttnuennsing,
drying, and weighing until the diflerence L^w^a two
successive weighings does not exceed 0.001 % of the weight
of the container. Fingerprints on the container will change
the weight and must be avoided. Record the weight .w. m
grams,
7.1.1.1 Chromic acid cleaning solution is corrosive to
skin, eyes and mucous membranes and can cause severe
burns. Avoid contact with eyes, skin or clothing. In making
dilute solution, alwavs add acid to water with care. In case of
contact flush skin with water, using a shower if exposure is
severe. Flush eyes for 15 minutes with copious amounts of
water, immediately call a physician. Remove dothing imme-
-------
4S1)) D 1475
TABLE 1 Absolut* D«fi»ity of Water. g/mL
•c
15
16
17
18
19
20
21
22
23
24
25
28
27
28
29
30
Omny
0599127
0.998971
0.998772
0.998623
0.998433
0.998231
0.998020
0.997798
0.997566
0.997324
0.997072
0.996811
0.996540
0.996260
0.995972
0.995684
diately and wash before reuse. Chromic acid cleaning solu-
tion is a strong oxidizer. Avoid contact with organic or
reducing substances as a fire could results. See supplier's
Material Safety Data Sheet for further information.
7.1.2 Fill the container with freshly boiled distilled water
at a temperature somewhat below that specified. Cap the
container, leaving the overflow orifice open. Immediately
remove excess overflowed water or water held in depressions
by wiping dry with absorbent matenaL Avoid occluding air
bubbles in the container.
7.1.3 Bring the container and contents to the specified
temperature using the constant-temperature bath or room if
necessary. This will cause further slight flow of water from
the overflow orifice due to the expansion of the water with
the rise of the temperature.
7.1.4 Remove the excess overflow by wiping carefully
with absorbent material, avoiding wicking of water out of
orifice, and immediately cap the overflow tube where such
has been provided. Dry the outside of the container if
necessary, by wiping with absorbent material. Do not re-
move overflow that occurs subsequent to the first wiping
after attainment of the desired temperature (Note 2). Imme-
diately weigh the filled container to the nearest 0.001 % of its
weight (Note 3). Record this weight, N, in grams.
NOTE 2—Handling the container with bare hands will increase the
temperature and cause more overflow from the overflow orifice, and will
also leave fingerprints; hence, handling only with tongs and with h"Kfr
protected by dean, dry, absorbent material is T^mmmrirrl
NOTE 3—Immediate and rapid weighing of the filled container is
recommended here to minirnirr toss of weight due to evaporation of the
water through orifices, and from overflow subsequent to the first wiping
after attainment of temperature where this overflow is not retained by a
cap.
7.1.5 Calculate the container volume as follows:
V-(N-M)/P
where:
V »
N =
M -
volume of container. mi_
weight of container and water, g (7.1.4),
weight of dry container, g (7.1.1), and
= absolute density of water at specified temperature,
g/mL (see Table 1).
7.1.6 Obtain the mean of at least three determinations.
8. Procedure
8.1 Repeat the steps in Section 7. substituting the sample
for the distilled water and a suitable nonrcsiduai solvent for
the acetone or alcohol (see 7.1.2 and Note 4). Record the
weight of the filled container. W, and the weight of the empty
container, w. in grams.
NOTE 4— Trapping of paint liquids in ground glass or metal joints is
likely to result in high values of density that appear to increase with the
viscosity and density of the material: such errors should be nunimrrrri
by firm seating of the joints.
8.2 Calculate the density in grams per miililitre as follows:
where:
Dm = density, g/mL.
8.3 Calculate the density in pounds per gallon as follows:
where:
D = density, Ib/gal,
K = 8.3454 (Note 5), and
V = volume of container. raL (see 7. 1.6).
NOTE 5— The factor K, 8.3454. is calculated from volume-weight
relationship as follows:
8.345404 - K2J4)" x (231.00)*]/(4S3J9237)C
is the conversion factor for rnillflitres to cubic inches.
* 231.00 is the conversion factor for cubic inches to gallons.
c 453.59237 is the conversion factor for grams to pounds.
9. Report
9.1 In reporting the density, state the test temperature to
the nearest 0.1'C, the units, and the value calculated to the
third place to the right of the decimal point (for example, D
- xaxx Ib/gal at 25'Q; state the mean, the range, and the
number of replicate determinations.
10. Precision •*"* Bias
10.1 The precision estimates are based on an interiabora-
tory study in which one operator in each of six different
laboratories analysed in duplicate on two different days five
samples of paint ranging in density from 8.5 to 12^ Ibs/gaL
The results were analyzed statistically in accordance with
Practice E 180. The within-laboratory coefficient of variation
was found to be 020 % relative with 25 degrees of freedom
and the between-laboratory coefficient of variation was
0.61 % relative with 20 degrees of freedom. Based on these
coefficients, the following criteria should be used for judging
the acceptability of results at the 95 % confidence level:
10.1.1 Repeatability—Two results, each the mean of du-
plicate determinations, obtained by the same operator on
different days should be considered suspect if they differ by
more than 0.6 % relative.
10.1.2 Reproducibility—Two results, each the mean of
duplicate determinations, obtained by operators in different
laboratories should be considered suspect if they differ by
more than 1.8 % relative.
10.2 No data on bias has been generated for this method.
Page 175
-------
D1475
». PA 19103.
Page 176
-------
Designation: D 2369 - 87'2
Standard Test Method for
Volatile Content of Coatings1
This standard is issued under the fixed designation O 2369: the number immediately following the designation indicates the year of
original adoption or. in the case of revision. the year of last revision. A number in parentheses indicates the year of last itapprovai. A
supencript epsiion (<) indicates an editorial chance since the last revision or itapprovai.
aapnm for use bvagemei of the Depanmem of Defense 10 repfae Method 404 I.J of Federal Test Method
Standard No. 141 A ana for laiuig in tne DoD Inaac oj Spectftautons ana Stanaards.
" NOTE — Paragraph 7.3 was editorially corrected and the designation date was chanted June 10. I9S7.
NOTE— The references to notes and sections in 1.4. 7.2. and 7.3 were editorially corrected in December 1988.
1. Scope
1.1 This test method describes a procedure for the deter-
mination of the weight percent volatile content of solvent-
reducible and water-reducible coatings. Test specimens are
heated at 110 ± S'C for 60 min.
1.2 Sixty minutes at 110 ± 5*C is a general purpose test
method based on the precision obtained with both solvent-
reducibie and water-reducible coatings (see Section 9). These
coatings (single package, heat cured) are commonly applied
in factories to automobiles, metal containers, flat (coil) metal
and large appliances, and many other metal parts.
NOTE 1—Testing at 110 ± S'C for 20 min was utilized for the
establishment of the original ten method in 196S. Precision data are not
available and may not have been property generated at the time. The
nine paints tested then were ail solvent-reducible. These conditions. 20
min at 110 ± S'C, are no longer sausbctoiy for the determination of the
volatile content of many coatings being tested at the present time.
Water-reducible and solvent-reducible coalings were tested in the
development of Test Method D 2369 using 110 ± S'C for 60 min and 20
min for which precision data have been generated. See Appendix for
precision statements on the 20-rain oven residence "T?c
1.3 This test method does not cover multi-package coat-
ings wherein one or more pans may, at ambient conditions,
contain liquid coreactams that are volatile until a chemical
reaction has occurred with another component of the
multipackage coating.
1.4 This test method may not be applicable to all types of
coatings such as printing inks, and other procedures may be
substituted with mutual agreement of the producer and user
See Note 3.
1.5 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address ail of the safety problems associated with its use. It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use. A specific
hazard statement is given in 7.3.1.
2. Referenced Documents
2.1 ASTM Standards:
D343 Specification for 2-Ethoxyethyl Acetate (95%
Grade)2
D 362 Specification for Industrial Grade Toluene3
D1193 Specificauon for Reagent Water3
E 145 Specification for Gravity Convection and Forced-
Ventilation Ovens4
E 180 Practice for Determining the Precision Data of
ASTM Methods for Analysis and Testing of Industrial
Chemicals5
3. Summary of Test Method
3.1 A designated quantity of coating specimen is weighed
into an aluminum foil dish containing 3 mL of an appro-
priate solvent, dispersed, and heated in an oven at 110 ± 5*C
for 60 min. The percent volatile is calculated from the loss in
weight
4. Significance and Use
4.1 This test method is the procedure of choice for
determining volatiles in coatings for the purpose of calcu-
lating the volatile organic content in coatings under specified
test conditions. The inverse value, nonvolatile, is used to
determine the weight percent solids content This informa-
tion is useful to the paint producer and user and to
environmental interests for determining the voianles emitted
by coatings.
5. Appsntns
5.1 Aluminum Foil Dish. 58 mm in diameterby 18 mm
high with a smooth (planar) bottom surface. Precondition
the dishes for 30 min in an oven at 110 ± 5*C and store in a
desiccator prior to use.
52 Forced Draft Oven. Type IIA or Type Iffi as specified
in Specification E 145.
5.3 Syringe, 5-mL. capable of properly dispensing the
coating under test at sufficient rate that the specimen can be
dissolved in the solvent (see 7.2).
5.4 Test Tube, with new cork stopper.
5.5 Weighing or Dropping Bottle.
1 This test method is under the jurisdiction of ASTM Committee D-l on Paint
and Related Coatings and-Materials and is the direct responsibility of Subcom-
mittee D01.21 on Chemical Anatvu of Paints and Paint Materials.
Current edition approved June 10. I9S7. Published August 1917. Originally
published as O 2369 - 65 T. Last previous edition O 2369 - 86.
1 Discontinued: see 19S2 Annual Book of ASTM Standards. Pan 29.
1 Annual Book of ASTM Standards. Vol 06.03.
4 Annual Book of ASTM Standards. Vol 14.02.
9 Annual Book of ASTM Standards. Vol 15.03.
Page 177
-------
6. Reagents
6.1 Purity of Reagents—Reagent grade chemicals shall be
used in all tests. Unless otherwise indicated, it is intended
that all reagents shall conform to the specifications of the
Committee on Analytical Reagents of the American Chem-
ical Society, where such specifications are available.6 .Other
grades may be used, provided it is first ascertained that the
reagent is of sufficiently high purity to permit its use without
lessening the accuracy of the determination.
6.2 Punty of Wafer—Unless otherwise indicated, refer-
ences to water shall be understood to mean Type II of
Specification D 1193.
6.3 Toluene, technical grade. Specification D 362.
6.4 2-Ethoxyethyl Acetate, Technical grade. Specification
D343.
7. Procedure
7.1 Mix the sample, preferably on a mechanical shaker or
roller, until homogeneous. If air bubbles become entrapped.
stir by hand until the air has been removed.
7.2 Using an appropriate weighing container (5.3. 5.4. or
5.5. with the syringe preferred for highest precision), weigh to
0.1 mg, by difference, a specimen of 0.30 ± 0.10 g for
coatings believed to have a volatile content less than 40
weight % or a specimen of 0.50 ± 0.10 g for coatings believed
to have a volatile content greater than 40 weight %, into a
tared aluminum foil dish (5.5) into which has been added 3
± 1 mL of suitable solvent (6.2, 6.3, or 6.4). Add the
specimen dropwise, snaking (swirling) the dish to disperse
the specimen completely in the solvent. If the material forms
a lump that cannot be dispersed, discard the specimen and
prepare a new one. Similarly prepare a duplicate.
NOTE 2—If the specimen cannot be dispersed in the solvents listed
(6.2.6.3. or 6.4) a compatible solvent may be substituted provided it is
no less volatile than 2-ethoxyethyi acetate (6.4).
7.3 Heat the aluminum foil dishes containing the dis-
persed specimens in the forced draft oven (5.2) for 60 rain at
110 ± S"C
7.3.1 Warning—Provide adequate ventilation, consistent
with accepted laboratory practice, to prevent solvent vapors
icai Soc. Waduaawa DC For
the American Chemical Society. ate
OB it* mat oft
02369
from accumulating to a dangerous level.
7.4 Remove the dishes from the oven, place immediately
in a desiccator, cool to ambient temperature and weigh to 0.1
rag.
NOTE j—If unusual decomposition or degradation of the specimen
occurs during heating, the actual time and temperature used to cure the
coating m practice may be substituted for the time and temperature
specified in this test method, subject to mutual agreement of the
producer and user.
8. Calculations
8.1 Calculate the percent volatile matter. V, in the liquid
coating as follows:
V, % - 100 - {((W, - w,)/.S) x 100]
where:
Wi » weight of dish.
Wj = weight of dish plus specimen after heating, and
5" a specimen weight
8.2 The percent of nonvolatile matter. N, in the coating
may be calculated by difference as follows:
.V, % «• 100 - volatile matter
9. Precision and Bias
9.1 The precision estimated for tests at 60 rain at 110 ±
5'C are based on an interiaboratory study7 in which 1
operator in each of 15 laboratories analyzed in duplicate on 2
different days 7 samples of water-based paints and 8 samples
of solvent-based paints containing between 35 and 72 %
volatile material. The paints were commercially supplied.
The results were analyzed statistically in accordance with
Practice E 180. The within-laboratory coefficient of variation
was found to be 0.5 % relative at 213 degrees of freedom and
the between-iaboratories coefficient of variation was 1.7 %
relative at 198 degrees of freedom. Based on these coeffi-
cients, the following criteria should be used for judging the
ability of results at the 95 % confidence leveL
9.1.1 Repeatability—Two results, each the mean of dupli-
cate determinations, obtained by the same operator on
different days should be considered suspect if they differ by
more than 1.5 % relative.
9.1.2 Reprodudbility—Two results, each the mean of
duplicate determinations, obtained by operators in different
laboratories should be considered suspect if they differ by
more than 4.7 % relative.
9.2 fffar Bun has not been determined.
001*1026.
data arc available from ASTM HeadouaMn.
RR:
Page 178
-------
Designation: 0 2697 - 86
Standard Test Method for
Volume Nonvolatile Matter in Clear or Pigmented Coatings1
This standard a nraed under the fixed designation D 2697: the number immediately following the designation indicates the year of
originaladoption or. tn the case ot revision, the year of laa revision. A number in parentheses indicates the year of last teapprovai. A
supenenpt epsilon (<) indicates an editorial change since the last revision or reapproval.
1. Scope
• 1.1 This test method is believed to be applicable to the
determination of the volume of nonvolatile matter of a
variety of coatings. An interlaboratory study to establish the
precision of this test method included a water-reducible
exterior latex paint and three automotive coatings that
included a solvent-reducible primer surfacer, water reducible
pnmer surfacer, water reducible enamel topcoat, and acrylic
dispersion lacquer topcoat. Eariier collaborative studies "in-
cluded a gloss enamel, a flat wall paint a gloss house enamel
an industrial baking enamel, an interior latex paint, and an
exterior latex paint.
1.2 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the safety problems associated with its use It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use.
2.' Referenced Documents
2.1 ASTM Standards:
D1475 Test Method for Density of Paint, Varnish, Lac-
quer, and Related Products2
D2369 Test Method for Volatile Content of Coatings2
D3925 Practice for Sampling Liquid Paints and Related
Pigmented Coatings2
D3980 Practice for Interlaboratory Testing of Paint and
Related Materials2
3. Summary of Method
3.1 The weight and volume of a stainless steel disk are
determined. After the disk is coated with the material being
tested, the weight and volume of the disk plus dried coating
is determined by weighing in air and then by weighing in a
«id of known density, the volume being equal to the
: otient of the weight loss of the coated disk (due to the
Archimedes buoyancy effect) divided by the density of the
liquid displaced. The liquid may be water, organic liquid
such as low-solvency mineral spirits or kerosine, or with
special modifications not covered specifically in this method,
mercury. The choice of liquid depends upon the nature of
the coating tested.
NOTE 1— Distilled water is suitable for most paints. Exceptions are
coatings that contain ingredients that are readily leached out of the dry
1 This test method i* under the jurisdiction of ASTM Committee CM on Paint
and Related Coatings and Materials and is the direct mpombihty of Subcom-
mittee D01.21 on Chrmicil Analysis of Paints and Paint Materials.
Current edition approved April 25. 1916. Published June 1986. Originally
published as D 2697 -68. Laa preview edition D 2697 -73 (1979). ^^
2 Animal Book of ASTM Standard. Volume 06.01.
film by the water and low-gloss coatings, the surface of which is poorly
wet by water even with surfactant added. (Note 2) Low-solvency
hydrocarbon solvent (KB below 36) is also practical for most paints and
is preferred by some workers.J It is considered to be particularly good for
paint films not readily wet by water. Analogously, organic solvents must
not be used if the coating to be tested contains ingredients that will be
dissolved readily by the solvent Lacquers containing monomenc
plasucizers would be examples where hydrocarbon solvents should
definitely not be used. Coatings formulated much above the CPVC
present a special problem, where mercury might be the desired "sus-
pending" liquid (Note 3), and for solvent-reducible paints hydrocarbon
solvent might be considered the poorest (unless it is the objective to
obtain values closer to "theoretical" spaces between pigment parades
not filled with binder, becoming partially filled with solvent during the
test).
Note 2—Concentration of surfactant must be kept very low or
literature values for the density of the water cannot be used.
Note 3—Details of the mercury displacement techniques can be
found in the literature.4
3.2 From the measured weights and volumes of the disk
before and after coating, the weight and volume of the dried
coating film are calculated. Based on the density of the liquid
coating and the weight percent nonvolatile matter, the
volume of the liquid coating deposited on the coated disk is
rairnfavvt, The volume of the dried coating divided by the
volume of liquid coating, multiplied by 100, provides the
volume percent nonvolatile matter in the total liquid
coating.
4. Significance and Use
4.1 This test method is intended to provide a measure of
the volume of dry coating obtainable from a given volume of
liquid coating. This value is useful for comparing the
coverage (square feet of surface covered at a specified dry
film tt|fcirn,ps5 per unit volume) obtainable with different
coating products.
4.2 For various reasons the value obtained may not be
equal to that predicted from simple additivity of the weights
and volumes of the raw materials in a formulation. One
reason is that the volume occupied by a solution of resin in
solvent may be the same, greater, or less than the total
volume of the separate ingredients: such contraction or
expansion in resin solutions is governed by a number of
factors, one of which is the extent and direction of spread
between solubility parameters of the resin and solvent
4.3 The spatial configuration of the pigment particles and
the degree to which the spaces between the pigment panicles
are filled with the binder also affect the volume of a dry
1 Bissey, J. E- Official Digtst. Federation of Paint and Vannsh Production
dubs. Vol 35. 1963. p. 1071 and Athlon. H. £. Maunoii Aettmdi and
Suudanis. Vol 1.1961. p. 349.
* Cole, ILl^ Journal. Oil Colour Chenusn' Assn. VoL 45.1962. p. 776.
Page 179
-------
D2697
coating formulation. Above the critical pigment volume
concentration, the apparent volume of the dry film is
significantly greater than theoretical due to the increase in
unfilled voids between pigment panicles. The use of volume
nonvolatile matter values in such instances should be care-
fully considered as the increased volume is largely due to air
trapped in these voids.
5. Apparatus
5.1 Analytical Balance.
• 5.2 Steel Disk, preferably stainless steel. 2Vi in. (60 mm)
in diameter and 22 gage (0.65 mm) in thickness with a small
hole near the circumference. A fine wire, such as Chrome! A.
28 gage (0.32 mm), is attached through the hole and made
the appropriate length for subsequent suspension of the disk
in a liquid. The wire should have a small loop on the upper
end so the disk and wire can be hung by this loop on the
balance.
NOTE 4—-Instead of sted disks, some analysts use aluminum tubes.
In the round-robin results, essentially no difference was found m the
precision obtained by both methods. Source and dimensions of these
tubes are described in the annex.
5.3 Counterweight, to be placed on the balance stirrup
after hanger bow and pan are removed.
5.4 Beaker, 1-L—For easier manipulation during the
weighing of disk in liquid it is advisable to cut the beaker to
a height of 4% in. (115 mm).
5.5 Support for holding the beaker under the balance
stirrup without jamming the pan damper in the floor of the
balance. A cork or neoprene ring is suitable when a
single-pan balance is used.
5.6 Weight per Gallon Cup, aerometer, or other suitable
means for determining the density of the coating material
and the suspending liquids if not known.
6. Volume Determination of Uncoated Disks
6.1 Dry the disk in an oven at 110 ± 5*C for 10 min. Cool
and weigh the disk in air.
6 J Weigh the disk in the liquid to be used for suspension
of the coated disk. If water is used as the suspending liquid, a
few drops of wetting agent (Note 2) added to the liquid will
help to ensure rapid and thorough wetting of the disk. Be
careful that no air bubbles form on the disk or wire. Mark
the level of liquid in the 1-L beaker necessary for complete
immersion of the disk which should be at least V* in. (20 mm)
above the disk. Maintain this level in subsequent weighings
when the disk is coated.
6.3 Record the temperature of the liquid. Obtain the
density of the liquid at the temperature used, from a table,
such as is found for pure water in Handbook of Chemistry
and Phrsics.* or determine it to 0.001 g/mL.
6.4 Calculate the volume of the disk. C, in miililitres as
follows:
G-<.v, - w.\/D
where:
»v, = weight of disk in air. g
ws = weight of disk in liquid, g, and
* CRC Press, inc. West Palm Beach. FL. 1986.
D « density of liquid at temperature of test. g/mL.
7. Procedure
7.1 Take a representative sample of the liquid coating in
accordance with Practice D3925. Mix thoroughly before
taking specimens for the individual tests.
7.2 Determine the weight nonvolatile of the liquid coating
by drying 1 h at 1 10* ± 5*C in accordance with Method B of
Test Method D 2369.
NOTE 5 — If this method does not apply, then the method used
should be agreed upon between producer and user.
7.3 Determine to 0.001 g/mL the density of the liquid
coating in accordance with Test Method D 1475.
7.4 Dip the disk in the liquid coating and allow the liquid
to come up on the wire a distance from 'A to Vi in. (5 to 15
mm). Allow about 10 min for draining, and blot the coating
material off the bottom edge of the disk so that beads or
drops do not dry on the bottom edge of the disk.
NOTE 6 — In some cases the paint or vanish may be of such
consistency that the amount of solid matter remaining on the disk after
drying is too small for an accurate volume determination. The use of a
flat pan with a sidewail about 10 mm in height in place of the disk
enables the operator to obtain a more desirable volume of solid matter.
However, extra care must be observed to prevent trapping of air at the
point where the sidewail meets the bottom of the pan. In no case should
bubbles be allowed to be present in can films. This procedure has not
been evaluated and no precision statement is available.
7.5 When beads or drips stop forming, hang the disk in
the oven for 1 h at 1 10*C (Note 5). Remove and cooL Weigh
the coated disk in air.
7.6 Weigh the coated disk in the chosen medium in the
same manner as for the uncoated disk, recording the
temperature of the liquid at the time of the test
8. Calculations
8.1 Calculate the volume of the coated disk, H, in
miililitres. as follows:
where:
Wj •* weight of coated disk in air, g,
w4 m weight of coated disk in liquid, g, and
D •density of liquid at temperature of test
8.2 fa'ff'liitf the volume of the dried coating, F, in
miililitres. as follows:
F-H-G
8J Calculate the volume of wet coating, V, in miililitres,
from which the dried coating was obtained, as follows:
where:
iv - nonvolatile matter in 1 g of wet coating, g, and
p - density of liquid coating material.
8.4 Calculate the percent volume nonvolatile content in a
liquid coating as follows:
(F/V) x 100
NOTE 7— The displacement liquid used should be reported with
volume percent nonvolatile results. The method of drying the films
should also be stated if different from that specified.
Page 180
-------
D2697
9. Precision6
9.1 Precision—In an interiaboratory study of this test
method in which one operator in cadi of five laboratories
analyzed in duplicate on two days four coatings (two
solvent-reducible and two water-reducible) with nonvolatile
contents ranging from 24 to 35 volume %. the pooled
within-laboratory standard deviation was found to be
0.444 % with 17 degrees of freedom (DF) and the pooled
between-laboratories standard deviation 1.195% with 16
DF. after discarding one day's results from two laboratories
•Supporting dau are available from ASTM Headquarters. Request
RR.-DOI-I052.
on one sample, one day's results from one laboratory on
another sample, and one duplicate result from one labora-
tory on a third sample. Based on these standard deviations
the following criteria should be used for judging the accept-
ability of results at the 95 % confidence level:
9.1.1 Repeatability—Two results, each the mean of dupli-
cates, obtained by the same operator on different days should
be considered suspect if they differ by more than 1.32 %
absolute at volume nonvolatile contents of 24 to 35 %.
9.1.2 Reproducibility—Two results, each the mean of
duplicates, obtained by operators in different laboratories
should be considered suspect if they differ by more than
3.59 % absolute at the same levels.
9.2 Bias—Bias has not been determined.
ANNEX
Nonmandatory information
Al. Aluminum Tubes
Al.l Aluminum tubes,7 uncoated. plain with no cap or A1.2 Cut two 3-in. (75 mm) lengths of tube from the
liner. *I6 neck and orifice. 1'A by 6'A in.. (32 mm by 160 aluminum tubing. Make a 3A-in. (20 mm) cut on the
nun). flattened end of the tube about 'A in. (6 mm) from the end.
Slip the tube over a short length of 1 in. (25 mm) inside
diameter electrical conduit and return the tube to a round
condition. Remove the tube from the pipe and press 1 in.
wide strip at an end of the tube toward the center to serve as
a hangar attachment.
7 Tubes manufactured by Tefedynr. 2290 W. Townsend Su Cluster. PA
19016. have been found satisfactory for this purpose.
W
m***»>*l m tfia tunaurg. UMH of tmt «t«no«fO in ufpnttly aovaed urn 0**mnX>on el tt» vuuaitr el vy tuen
ma mutt t» n****a •*•
TJ* tanatn a suoper ro rwnswn mt tny ttom oy rrw mponsXM Mcrracaf corn
«
-------
MM IK Des'9nation: O 2832 - 83
Standard Guide for
Determining Volatile and Nonvolatile Content of Paint and
Related Coatings1
UBder *" fiwd d8*"*8"0" O 2J32: the numoer tmmeoiatelv following the designation indicates the vear of
> «ne cue 01 revwon. the *e» « last revision. A number in putnineja mdiota the ye«r ot Ua rewprovu. A
nipencnpi epnion i«) indicates an editorial change since the l*st revision or itapproval.
1. Scope
1.1 This guide is intended to aid in the selection of the
proper ASTM standard for determining the volatile and
nonvolatile content of paint and related coatings.
NOTE'—Test methods for determining the composition of the volatile
fraction are not covered by this guide.
1.2 The standards included are as follows:
Type of Count
Aerosol coauncs
Architectural watt coauatv intenor. faith
Section Oeuanauon
Asphalt tool
AaphaUrafi
Bitumens, emulsified
Bleached lac varnish
Coil
House puna. *la
ladusuul baking
Ucquervdetraad
USCX Ptt
Magnet witei
rlittici* cmtinp for
Powder
Solvem-vedoabie
Traffic
Vi
Wall
Wall pamu. flat
___ • --- *-'-
4.1
4.J
4.4
4.3
4.6
4.7
44
4.9
4.10
4.10
4.11
4.10
4.10
4.12
4.13
4.14
4.7
4.2
4.13
4.16
4.17
4.10
4.2
D3062
01644
D2S23
D2S24
O2939
OI630
OI333
O2697
0113
O2697
02697
D1644
O333
D2697
O2697
D32S1
OI644
03431
01630
02369
02203
DI644
D2697
02697
02369
-.^ ___ --- -- _______
ncMcoucme coaaap 4J D2369
1 J This standard may involve hazardous materials, oper-
ations. and equipment. This standard does not purport to
address ail of the safety problems associated with its use It is
the responsibility of whoever uses this standard to consult and
establish appropriate safety and health practices and deter-
mine the applicability of regulatory limitations prior to use.
2. Referenced Documents
2.1 ASTM Standards:
D 1 1 5 Methods of Testing Varnishes Used for eectrical
Insuiation-
D 333 Test Methods for Gear and Pigmcnted Lacquers
D 1353 Test Method for Nonvolatile Matter in Volatile
Solvents for Use in Paint. Varnish. Lacquer, and Re-
lated Products4
D1644 Test Methods for Nonvolatile Content of
Varnishes3
D1650 Methods of Sampling and Testing Shellac
Varnish*
D 2205 Guide for Selection of Tests for Traffic Paimr
D2369 Test Method for Volatile Content of CoatingsJ
D2697 Test Method for Volume Nonvolatile Matter in
Clear or Pigmented Coatings3
D2823 Specification for Asphalt Roof Coatings6
O2824 Specification for Aluminum-Pigmented Asphalt
Roof Coatings6
D2939 Method of Testing Emulsified Bitumens Used as
Protective Coatings6
D3062 Test Method for Solids Content of Aerosol
Coatings7
D3288 Method of Testing Magnet- Wire Enamels*
D3451 Practices for Testing Polymeric Powders and
Powder Coatings^
3. Significance nmi Use
3.1 The nonvolatile content of paint and related coatings
is useful to producers and users and to environmental and
health and safety interests in comparing the coverage of
competing products and in estimating the volatile organic
content.
4.1 Aerosol Coatings— Test Method D 3062 covers the
determination of solids content (weight %) in aerosol coat-
togs.
42 Volatile Content of Coatings (Test Method
D 2J69;— This test method covers the determination of the
volatile content of coatings. It is considered to be applicable
to most solvent-reducible and water-reducible V^-
4.3 High Performance Interior Architectural Wall Coat-
ings fHIPAC)— Determine the nonvolatile content of
HIPAC coatings in accordance with Test Methods D 1644.
Calculate the volatile content (weight %) by difference.
1 This imde u under the mmdienaa of ASTM Committee D-l on Paint and
Related Coaunai and Mam* and a the difeci lanoaability of Siibamumnee
D01.21 on Cheimcml Anatvm of Paint and Paint Matenais.
Cun^2inM«*2TlLMfICtl '• l983' Pubtahed Se««o«» «983. Or*.
oally published as D2S32-69. LaaptcvMus edition 02132-69(1980).
2 AIUUMU Book ofASTMSunOam. Vol 10.01.
> Amaui Book of ASTM Sundays. Vol 06.01.
* Annual took of ASTM Stanoants. Vol 06.03.
>A*oiai Book of ASTM Standards. Vol 06.01
* Annual Book of ASTM Standards. Vol 04.04.
' Animal Book of ASTM Standards. Vol 13.09.
•Amaui Book of ASTM Standards. Vol 10.01
Page 182
-------
D2832
4.3.1 Method A—3 h at 105*C for paints with nonsoivent
components that decompose at higher temperature.
4.3.2 Method B—10 min at 149*C for most paints with
nonsoivent components that are reasonably stable at 149*C.
4.4. Asphalt Root Coatings—Determine the nonvolatile
content (weight %) of asphalt roof coatings of brushing or
spraying consistency in accordance with 8.2 of Specification
D 2823.
* 4.5 Aluminum-Pigmented Asphalt Roof Coatings—The
nonvolatile content (weight %) of asphalt-based aluminum
roof coatings suitable for application to roofing or masonry
surfaces by brush or spray is determined in accordance with
8.2 of Specification D 2824.
4.6 Emulsified Bitumens Used as Protective
Coatings—Section 8 of Test Methods D 2939 contains a
method for determining residue by evaporation (weight %)
of emulsified bitumens used in relatively thick films as
irotective coatings for metals and built-up roofs.
.'' Shellac Yarnish—Determine the nonvolatile matter
r. orange shellac and bleached lac varnishes in accordance
wirh Sections 14 through 16 of Methods D 1650.
4.8 Coil Coatings—Although stated to be for solvents.
determine the nonvolatile matter (weight %) in accordance
with Test Method D 1353. Determine volume solids in
accordance with Test Method D 2697.
4.9 Electrical Insulation famishes:
4.9.1 Sections 18 through 22 of Methods D 115 on
nonvolatile matter by weight, are applicable to the following
classifications of varnishes used for electrical insulation:
alcohol-soluble varnishes, oxidizing air-drying varnishes,
thermosetting varnishes, oxidizing baking varnishes, air-
drying asphaltic varnishes, silicone varnishes, and thermo-
setting laminating varnishes.
4.9.2 Determine nonvolatile matter in electrical insulating
varnishes intended for electrical equipment operating at
180*C and above in accordance with Methods D 115 except
that the temperature used shall be 275 ± 5.5T (135 ± 3*Q
or at a temperature agreed upon between the producer and
the user.
4.10 Volume Nonvolatile Matter in Clear or Pigmented
Coatings (Test Method D 2697)—This test method is appli-
cable to the determination of the volume nonvolatile matter
of coatings. A gloss enamel, a flat wall paint, a gloss house
paint, an industrial baking enamel, an interior latex paint,
and an exterior latex paint included in formal collaborative
studies of this test method.
4. 1 1 Lacquer Coatings— Determine the nonvolatile con-
tent of clear and pigmented lacquers as described in Test
Methods D 1644. As an additional requirement, the spec-
imen shall be reheated and reweighed until the weight is
constant to within i mg. Method A of Test Methods D 1644
is preferred since Method B is potentially dangerous when
used with lacquers.
4.12 Magnet Wire Enamels— Sections 17 through 23 of
Test Method D 3288 cover the determination of nonvolatile
content (weight %) in magnet wire enamels.
4. 13 Coatings for Plastics— Determine nonvolatile matter
in clear and pigmented coatings designed for use on rigid or
semirigid plastic substrates in accordance with Test Methods
D 1644.
4.14 Polymeric Powders and Powder Coatings — Deter-
mine nonvolatile content (weight %) in accordance with
Section 12 of Practices D 3451. Determine volatile content
at baking or fusion temperature in accordance with Section
13 of Practices D 3451.
4.15 Traffic Paints — Determine the nonvolatile content
of traffic paints, ready-mixed, of spraying consistency of the
premix. drop-in, or combination type in accordance with
Test Methods D 1644. and state any necessary larger spec-
imen size for beaded paint. Either of the two methods can be
used as follows:
4.15.1 Method A— 3 h at 105*C for paints with
nonsoivent components that decompose at higher tempera-
ture.
4.15.2 Method B— 10 min at 149'C for most paints whh
nonsoivent components that are reasonably stable at 149*C.
4.16 Varnishes— Nonvolatile content (weight %) of var-
nishes is determined using Test Methods D 1644. These test
methods may give high results due either to incomplete
elimination of volatile matter or to absorption of oxygen by
oxidizing-type varnishes.
4.17 Solvent-Based Interior Semigloss Wall and Trim
Enamels— Use Test Method D2369 to determine volatile
content. Volume nonvolatile matter is determined in accord-
ance with Test Method D 2697.
5. Precision
5.1 Some of the referenced ASTM standards have preci-
sion limits. Reference to the individual standards for preci-
sion statements is recommended.
.
"~ooumn. your eommtm w,u raewv* evthH eonsmrmen u t mtfiing oi me resoonUM
Mcfiwcif OMWUIM M«cn row muy crana. H you tttt tntt your comment /wv» nor ncmwa • Ivr /MOT* you snotaa mm your
vwws Known to «• ASTM CammuM an Sunatna. 79T6 Race Sf.. PMMOWPM. PA 1S103.
Page 183
-------
Designation: O 3792 - 86
Standard Test Method for
Water Content of Water-Reducible Paints by Direct Injection
Into a Gas Chromatograph1
oi
1. Scope
1.1 This test method is for the determination of the total
water content of water-reducible paints. It has been evalu-
ated for latex systems (styrene-butadiene. poltfvinylacetate)-
acrync,acryuc). It has not yet been evaluated for other
water-reducible paints but is believed to be applicable. The
established working range of this method is from 40 to 55 %
water. There is no reason to believe that it will not work
outside of this range.
1.2 This standard may involve hazardous materials oper-
ations. and equipment. This standard does not purport 10
address all of the safety problems associated with its use It is
the responsibility of the user of this standard to establish
appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use For
specific hazard statements, see Section 8.
2. Referenced Documents
2.1 ASTM Standards:
D 1 193 Specification for Reagent Water
DJ^64uT(? Mcthod for Water in
(Fischer Reagent Titranon Method)2
E 180 Practice for Determining Precision Data of ASTM
Methods for Analysis and Testing of Industrial Chem-
icals3
D 3792: the number immeoiaidy following the detonation indicate the year of
the year oi last revision. A number in parentnan indicates the year of last reapprovat. A
the last icvmon or leapprovai.
TABLE 1 Inattumtflt Condition*
Solvents
3. Summary of Test Method
3.1 A suitable aliquot of whole paint is internally stan-
dardized with anhydrous 2-propanoi. diluted with dimethyl.
formamide, and then injected into a gas chromatographic
column containing a porous polymer packing that separates
water from other volatile components. The water content is
determined from area calculations of the mat^fr „„.
ducing peaks on the chromatogram.
4. Significance mid Use
4.1 With the need to calculate volatile organic content
(VCQ of water-reducible paints, it is necessary to know the
water content. This gas chromatographic test method pro-
vides a relatively simple and direct wav 10 determine water
content.
?£??"** il»*|?tte "««'«•«» of ASTM Committee D.I on Paint
Related Catnap and Materials and is the dincc
1 Amml Book of ASTM Standards. VoJ 06.03.
1 Arnuai Book of ASTM Sudanis. Vol 15.05.
Temperaum. *C:
Samp* met
Cctmt*
MttaJ
final
Gamer Gas
Flow rats. rnL/rran
Oatactor currant
Spacvnan siza
200
240
80
170
30/mn
hetourn or flrtnopjan
so
150mA
For isothermal i
in Mt me column tampamm ai 140*C. Atwr m«
2-preoana nu cmno me column aamst tne temoaraturt to 170% untt OMF
attn in* oommn. fleaat ow Mmpvnra «o 140% ftx tUMMuam mm.
5. Apparatus
S.I Gas Chromatograph — Any gas-liquid chromato-
graphic instrument having a detector may be used. Temper-
ature programming capability is preferable, but isothermal
operations may be adequate. See Table 1.
3.2 Qrfumn— The column should be 4 ft (1.22m) of VMn.
(3.2-mm) outside diameter tubing of stainless steel, or other
suitable material packed with 60/80 mesh (180 to 250 urn)
porous polymer packing material.4 A replaceable glass sleeve,
glass wool plug, or other suitable material may be placed on
the entrance end of the column to retain any nonvolatile
This will minifniT^ sludge buildup in the column.
5.3 Recorder— A recording potentiometer with a rail-scale
deflection of 1 to 10 mV, full-scale response time of 2 s or
less and sufficient sensitivity and stability to meet the
5.4 Liquid Charging Devices— Micro syringes of 10 or
25-uL capacity.
6. Column Conditioning
6.1 /Vocerfure^-The packed column is installed in the gas
chromatographic unit leaving the exit end disconnected from
the detector. This will prevent any contamination of the
detector with the column bleed. Set the helium or nitrogen
flow rate at 20 to 30 mL/rain if a Vi-in. (3 .2-mm) outside
diameter column is used. Purge the column 5 or 10 nun
before heating. Heat the column from room temperature to
200*C at 5'C/min and hold this temperature for at least 12 h
(overnight). At the end of this time, heat the column at
4 Panpak Q». a tndemark of Wawi AMOC. Inc. Mfltori. MA, tan tan found
rpackrai or other column, »»«• equivalents
'.Any otter |
or i
may be wad. Toe* oraducu an available from
Page 184
-------
5'C/mm to 250"C (the maximum temperature for this
packing; and hold for several hours. Cool the column to
room temperature and connect the column detector. Reheat
the column to 250*C at 5*C/min to observe if there is
column bleed. Optimum conditioning of this column may
take several cycles of the heating program before a good
recorder baseline is achieved.
6.2 Before each calibration and series of determinations
* tor daily) condition the column at 200'C for 1 h with earner
gas flow.
91. Reagents and Materials
7.1 Pieriiy of Reagents—Reagent grade chemicals shall be
used in ail tests. Unless otherwise indicated, it is intended
that all reagents shall conform to the specifications of the
Committee on Analytical Reagents of the American Chem-
ical Society, where such specifications are available.5 Other
grades may be used, provided it is first ascertained that the
reagent is of sufficiently high purity to permit its use without
lessening the accuracy of the determination.
7.2 Purity 01' W'oier—Unless otherwise indicated, refer-
ence to water shall be understood to mean reagent water
conforming to Type II of Specification D 1193.
7.3 Comer Gas—Helium of 99.995 % or higher punty.
High-purity nitrogen may also be used.
7.4 Dimethylformamide (DMF) (Anhydrous) gas chroma-
tography. spectrophotometric quality (Note 1).
7.5 2-Propanol (Anhydrous) (Isopropanof)—See Note 1.
7.6 Septum Sample Vials, 10-mL capacity with fluorocar-
bon-faced septa are preferred.
NOTE 1—Determine the water content of the DMF and 2-propanol
by Karl Fischer utranon in accordance with Ten Method D 1364. Dry
the 2-propanol if water is found in it or replace with anhydrous grade.
8. Hazards
8.1 Dimethylformamide is harmful if inhaled or absorbed
through skin. It is suspected to be embryotoxic. Use only
with adequate ventilation. Avoid contact with skin. eyes, and
clothing. Refer to suppliers Material Safety Data Sheet.
9. Preparation of Apparatus
9.1 Install the column in the chromatograph and establish
the operating conditions required to give the desired separa-
tion (see Table 1). Allow sufficient time for the instrument to
reach equilibrium as indicated by a stable base line. Control
the detector temperature so that it is constant to within 1'C
without thermostat cycling which causes an uneven baseline.
Adjust the carrier-gas flow to a constant value.
10. Calibration
10.1 Using the information in Table 1 (as a guide), select
the conditions of temperature and earner gas flow that give
the necessary resolution of the components.
10.2 Determination of Relative Response Factors—Anhy-
drous 2-propanol is used as an internal standard. The
D3792
response factor to water relative to the standard is deter-
mined by means of the following procedure. See fig. 1 for a
typical chromatogram. It is good practice to determine the
relative retention time daily or with each series of determi-
nations.
10.2.1 Weigh about 0.2 g of water and 0.2 g of 2-propanol
to 0.1 mg into a septum sample vial. If it has been
determined that a correction for the water content is
necessary, weigh 2 mL of dimethylformamide (DMF) into
the vial. If the DMF is anhydrous, simply add 2 mL of it as
weighing is not necessary.
10.2.2 Inject a 1-uL aliquot of the above solution into the
column and record the chromatogram. The retention order
and approximate retention times after the air peak are (/)
water, about 0.7 min: (2) 2-propanol. about 2.8 min: and (3)
DMF. about 7 min.
10.2.3 The preferred procedure to obtain the water con-
tent of the DMF is the Karl Fischer titration (Note 1). If this
has been determined, calculate the response factor for water
by means of the following equation:
1 'Reagent Chemicals. American Chemical Society Specification!- Am Chem-
ical Soc.. Washington. DC. For suggestions on tne testing 01 reagents not usted by
the American Chemical Society, see -Reagent Chemicals and Standards.- bv
Josepn Rosin. D. Van Nosnand Co. Inc. New York. NY and the -United States
Pharmacopeia.''
(fFH,0-i-PW,)A,
where:
R
- response factor.
= weight of 2-propanol.
weight of water added,
weight of dimethylformamide,
^H,O • area of water peak.
A, » area of 2-propanol peak, and
P » weight % water in DMF
100
10.2.4 If Karl Fischer titration is not available, the fol-
lowing procedure may be used to obtain a reasonable
FIQ. 1 Typical Chrematognm
Page 185
-------
estimate of the response factor
10.2.4.1 Inject the same size aliquot of DMF and 2-
propanoi mixture, but without added water, as a blank. Note
the area of the water peak in the blank.
10.2.4.2 The response factor for water is calculated by
means of the following equation:
.vnere:
response factor.
weight of 2-propanol.
weight of the water.
area of 2-propanol peak.
area of the water peak, and
area of the water peak in the blank.
03792
wi • weight of 2-propanol added.
W, » weight of paint, and
R m response factor determined in 10.2.
12.3 Correction for Water Content of Solvent:
12.3.1 If the blank indicates the presence of a detectable
peak for water in the dimethylformamide used as solvent.
make a correction in the calculation.
12.3.2 The water content of the dimethylformamide de-
termined by either chromatography (10.2.4) or, preferably.
Karl Fischer titration (10.2.3) is used to make the correction.
Calculate the water content due to the solvent by using the
following equation:
1. Procedure
i«I'IiiWe5hnl? °'1 ?& °-6 g of wawr-reducible paint (see
i 2) a"d 0.2 g of 2-propanol into a septum viaL Add
-mL of DMF into the viaL Seal the vial. Prepare a blank
jntaining the 2-propanol and DMF but no paint.
c* *yaan lo * analyzed for «»«*"»•
Coalescing igents do not imenere with ttau detenninauon.
11.2 Shake the vials on a wrist action shaker or other
•itable device for 15 min. To facilitate settling of solids
low the vials to stand for 5 min just prior to injection into
e cnromatograph. Low-speed centrifugation may also be
ed.
1 1.3 Inject a 1-uL sample of the supernatant from the
epared solutions into the chromatographic column.
*ord the chromatograms using the conditions described in
.ble 1.
. Calculations
12.1 Measure the area of the water peak and the 2-
jpanoi internal standard peak and multiply each area by
; appropriate attenuation factor to express the peak areas
a common basis. Use of an electronic integrator is
ommended to obtain the best accuracy and precision
'wever, triangulation, planimeter, paper cut out. or ball
i disk integrator may be used.
23. Calculate the water concentration in the paint by
ans of the following equation:
^--
:re:
o =
area of water peak.
area of 2-propanol peak.
where:
Wt = weight of dimethylformamide.
Wp =» weight of paint, and
P =* weight % water in DMF
fiJO
12.3.3 The water content of the paint in this case is the
difference between the total percent determined in 122 and
the correction tor the solvent water content as determined in
12.3.2.
13. Precision and Bias6
13.1 The precision estimates are based on an inter-
laboratory study in which.nine different laboratories ana-
lyzed in duplicate on two days four samples of water-
reducible paints containing from 40 to 55% H2O
(theoretical). The results obtained were analyzed statistically
in accordance with Practice E 180. The within-laboratory
coefficient of variation was found to be 1.0 % relative at 34
degrees of freedom and the bctween-laboratories coefficient
of variation 2.6 % relative at 30 degrees of freedom. Based on
these coefficients, the following criteria should be used for
judging the acceptability of results at the 95 % confidence
level.
13.1.1 Repeatability—Two results, each the mean of du-
plicate deteminations. obtained by the same operator on
different days should be considered suspect if they differ by
more than 2.9 % relative.
13.1.2 Reproducibiluy—Two results, each the mean of
duplicate determinations, obtained by operators in different
laboratories should be considered suspect if they differ by
more than 7.5 % relative.
132 Bias—Bos has not been determined.
•Supporunt data arc available (ram ASTM Hcadqtunm. ftcquat RR: 001 -
1022.
Page 186
-------
Designation: O 4017 - 38
Standard Test Method for
Water in Paints and Paint Materials by Karl Fischer Method1
This standard a issued unoer Me fixed designation O 40! 7: the numoer immetuatetv following ine designation indicates Oie vear 01
onttnal adoption or. in the case of revision, me year 01" last revision. A numoer in parentheses indieaia ine vev m last reaoprovai. A
uipencnot epsuon iti indicates an editorial change since tne last revision or reapprovai.
1. Scope
1.1 This test method is applicable to all paints and paint
materials, including resins, monomers, and solvents, with the
exception of aldehydes and certain active metals, metal
oxides, and metal hydroxides. While the evaluation was
limited to pigmented products containing amounts of water
in the 30 to 70 % range, there is reason to believe that higher
and lower concentrations can be determined by this test
method.
\2 This standard may involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
address all of the safety problems associated with its use. It is
the responsibility of the user of this standard to estaoiish
appropriate safety and health practices and determine me
applicability of regulatory limitations prior to use. Specific
hazard statements are given in Section 7.
2. Referenced
2.1 ASTM Standards:
D 1 193 Specification for Reagent Water
E 180 Practice for Determining the Precision of ASTM
Methods for Analysis and Testing of Industrial
E 203 Test Method for Water Using Karl Fischer Reagent3
3. Summary of Test Method
3.1 The material is dissolved in pyridine. or another
appropriate solvent, and titrated directly with ttanriarriip^j
Karl Fischer Reagent, to an elecuumemc end point The
sluggish reaction with water in pyridine is accelerated with a
3.2 Pyridine is used as a solvent to mini™*,- interference
problems caused by ketones. It is also used because the more
commen solvent, methanol, will not dissolve many common
resins and because methanol reacts with some resins to
produce water.
4. Significance and Use
4. 1 Control of water content is often important in control.
ling the performance of paint and paint ingredients, and it is
critical in controlling volatile organic compound (VOQ
content.
4.2 Paim materials are often insoluble in common Karl
Fischer solvents such as methanoL Pyndine has been found
to be a nearly universal solvent for these materials: however.
the Karl Fischer reaction is too slow in that solvent at room
temperature. To speed it up, l-€thyl-piperidine is added at
5 % as a buffer, or "catalyst".
5. Apparatus
5.1 Karl Fischer Apparatus, manual or automatic, encom-
passed by the description in Test Method E 203. Apparatus
should be equipped with a 25-mL buret. Class A. or
equivalent.
5.2 Svnnge, 100-uL capacity, with needle.
5.3 Syringes, 1-mL and 10-mL capacity, without needle.
but equipped with caps.
6. Reagents
6.1 Purity of Reagents—Reagent grao^ chemicals shaUbe
used in ail tests. Unless otherwise indicated, it is ""ended
that all reagents shall conform to the specifications ot toe
Committee on Analytical Reagents of the American^Chem-
ical Society, where such specifications are available, Other
grades may be usgd. provided it is ascertained that the
reagent is of sufficiently high purity to permit its use without
lessening the accuracy of the determination. __
62 Purity of Water—Unless otherwise indicated, refer-
ences to water shall be understood to mean reagent grade
water conforming to Type II of Specification D 1193.
6J Karl Fischer Reagent.*
6.4 Pyridine.
6.5 1-Ethylpiperidine.
6.6 Hydrochloric Acid (Hd) Concentrated.
7. Hazards
7.1 Karl Fischer reagent contains four toxic compounds.
namely iodine, sulfur dioxide, pyridine. and metnanoi or
giycol ether. The reagent should be prepared and dispensed
in a hood. Care must be exercised to avoid inhalation or son
contact. Following accidental contact or spillage, wash witn
large quantities of water. .
12 Pyridine and metnanoi solvents should be treated witn
the same care as Karl Fischer reagent.
' This test method is under the ninsdieuon at'ASTM Committee D-l on Paint
and Related Coatings and Materials aad is the direct reipotmeitnv of Subeom.
mmeeDOI.21 on Chemical Analysts of Mot* and Paint Materials.
Current edition approved Oct. 31.19SS. Pubbshed Decemeer I9SS. OhtjnaUv
published as O 4017 - 81. Las prcvms* aamoa O 4017 - 81 (19171"
: Amna fools ot ASTM Sum**. Vol O6.03.
1 Annual Book ot ASTM SumaaiOt. Vol 13.05.
icrfSoc. WfftfltftP,DCForsuaeodnid.•!*«•«•«
the Am«ema Chans* Socwy. ste -R«tf« Chawed* and
. Van Nosnnd Co. Inc. New York. NY and me
1 Sd-K*3 available mn
tunable for this purpote.
i Fischer Soenanc Co- or edtnvateM tat
Page 187
-------
D4017
TABLE 1 Specimen Guidelines
ExpMM
WMftT.
0.5-1.0
1-3
3-10
10-30
30-70
>70
Aperoxvnaw
Soeovnen
Wexjnt.
9
5
2-5
1-2
0.4-1 0
0.1-0.4
0.1
Appcoxvnaft
Titram volume
at 5 mg/mL
titrt. mL
5-10
10-20 .
10-20
20-25
15-25
20
7.3° 1-ethylpiperidine is of unknown toxicity and. there-
fore. should be handled with the same care as the above
materials.
8. Procedure
8.1 Standardization of Karl Fischer Reagent:
8.1.1 Add enough fresh pyridine to cover the electrode tip.
plus I mL of 1-ethylpipendine catalyst per 20 mL of
pyridine. Catalyst performs best at a concentration of about
5 % of the volume present.
8.1.2 Fill the 100-uL synnge to about half full with
distilled water and weigh to the nearest 0. 1 mg.
8.1.3 Pretitrate the pyridine to the end point indicated by
the equipment manufacturer, by adding just enough Karl
Fischer Reagent 1 (KFR) to cause the end point to hold for at
least 30 s.
8.1.3.1 The use of the catalyst greatly increases the reac-
tion rate between water and Karl Fischer reagent To obtain
reliable results, increase the electrode sensitivity and reduce
titration rate to a minimum Most instruments have controls
for these functions. Consult the instructional manual for
information on these controls.
8.1.4 Empty the contents of the syringe into the titrator
vessel. Immediately replace the stopper of the sample port
and titrate with KFR to the end point as described in 8.1.3.
8.1.5 Repeat standardization until replicate values of F
agree within 1 %. Determine the mean of at least two such
determinations. Carry out calculations retaining at least one
extra decimal figure beyond that of the acquired data. Round
off figures after final calculations.
8.1.6 Calculation:
8.1.6.1 Calculate the KFR titre F as follows:
where:
J - water added, g, and
P - KFR used. mL.
The value for F should be recorded to the four significant
digits and should be the mean of at least two determinations.
Typical values are in the range of 0.004000 to 0.006000
g/tnL.
8.2 Analysis oj Samples \llth Afore Than 0.5 7c Water.
8.2.1 The titration vessel should already contain
preutrated pyridine and catalvst. as described in steps 8.1.1
and 8.1.3 in the standardization procedure. Best results are
obtained with fresh solvent, that is. contain no previously
titrated specimen in the vessel.
8.2.2 With a 1-mL or 10-mL synnge, draw the amount of
material indicated in Table 1 .
8.2.2.1 Remove the synnge from the specimen, pull the
plunger out a little further, wipe the excess material off the
synnge, and place a cap on the synnge tip. Weigh the filled
synnge to the nearest 0.1 mg.
8.2.3 Remove the cap. and empty the syringe contents
into the preutrated pyndine vessel. Pull the plunger out and
replace the cap. Titrate the specimen with KFR to the end
point described in 8.1.3.
8.2.4 Reweigh the emptied syringe, and calculate the
specimen weight by difference.
8.2.5 Calculation:
8.2.5.1 Calculate the percent water L as follows:
L-\P*Fx IOOJ/S
8.3 Analysis of Materials With Less Than 0.5 % Water.
8.3.1 For 0.1 to 0.5%. follow procedure in 8.2 (1-g
specimen), except substitute a 1-mL microburet for the
25-mL buret in the Karl Fischer apparatus.
8.3.2 For less than 0.1 %, use a 1-mL microburet and
increase specimen size as much as needed, up to 10 g. It
should be possible to measure moisture levels down to 1 ppm
(0.0001 %) by this approach.
NOTE—Specimens with less than 0.1 % water may require special
nandlinE tecnmques to prevent pickup 01" atmospnenc moisture. The
precision ot this test method was determined with specimens containing
higher water levels.
9. Maintenance
9.1 C/«w«p^-Clean the titration vessel by rinsing with
fresh pyridine. Do not use methanol or other solvents.
92 Dnmess—Check frequently to be sure that all drying
tubes are in good condition and tightly connected, w8**
dessicant when indicator color changes through nay of the
tube.
9.3 Electrode Performance--If electrode response is slug-
gish or otherwise off standard, take the following steps, in
turn, to correct the problem. Test the electrode with a
titration after each step, to determine if the next step is
required.
9.3.1 Wipe the electrode tip with a dean paper towel.
9.32 Wash the electrode by dipping in concentrated
hydrochloric acid for at least 1 min. Rinse first with distilled
water, then with methanoL __.
9.3.3 Follow manufacturer's instructions on resetting end
point meter.
9.3.4 Replace power source. See manual for replacement
procedure.
9.3 J Replace the electrode.
10. Precision and Bias
10.1 The precision estimates are based on an inter-
laboratory study in which one operator in each of seven
different laboratories analyzed in duplicate, on two different
days, seven samples of water-based paints of various types
containing between 25 to 75 % water. The results were
analyzed statistically in accordance with Practice E 180. The
within-laboratory coefficient of variation was found to be
1.7 % relative at 98 df. and the between-laboratory coeffi-
cient of variation was 5.3 % relative, at 42 df. Based on these
coefficients, the following criteria should be used for judging
the acceptability of results at the 95 % confidence level.
10.1.1 Repeatability—Two results, each the mean of du-
plicate determinations, obtained by the same operator on
Page 188
-------
D4017
different days shouid be considered suspect if they differ by 102 Bias—Bias has not been determined for this test
more than 4.7 % relative. method.
10.1.2 Reproductbiiity—Two results, each the mean of
duplicate determinations, obtained by operators in different ll' Index Terms
laboratories shouid be considered suspect if they differ by 11.1 This test method is indexed under the following
more than 15.0 % relative. terms: Kari Fischer reagent method: moisture content: water
content (paints/related coatings;.
*-«ftaffyinmnMnmnwMma*r«no«ro. Ustnottnatttnavov»•tons**tavatavatomtmmmaonatirwvtuaayattmf tucn
prntM norn. ana m» nut of mtnnqumunt at tuen ngnts. «• «nr»wy (Mr own rMaamwy.
• ««>•« w mtuon m toy imu OY f/»e
raion Mcnmew common* ma mat O» iwntntua uvwr
_
eemmatu. wneti you may MMM. ff you ww mat your commtnis nav* nor netuna • tar Mtnng you tnoua HUM your
«no»n re nw ASTM Comnnm* en Stanairas. I9T6 fl«c« Si.. PMtataim. PA 19103.
Page 189
-------
«
Designation: D 4457 - 65"
Standard Test Method for
Determmation of Dichlorpmethane and 1,1,1-Trichloroethane
by Direct
1. Scope
.nf u coven lhe de«nninauon of total
amount of dichloromethane or l.U-trichioroethane or
both in paints and coatings. It has been evaluated for
ceUulose man. allcyd, vinyl, and styrene-butadi™e1v«ems
h has not yet been evauiated for other formulation?^
believed to be applicable. The established
S^t^?^1 "^ % for '.-n
and 32 to 78 % for dicbiorometnane. There is
believe it will not work ou^S^e^ e
of 1-propanol in paints and coatings requires the uf
difrereminternal standard. (See also Practice E 260)
L2 Thisstandard ^yinv^ve hazardous materials. oper.
towns, and equipment. This standard does not
yo^o^usesst
establish appropriate safety and health
mine the applicability of regulatory lim
Speonc precaution statements are given in Section 7
2. Referenced Documents
2.1 ASTM Standards:
^Dcten?ining *** Precision Data of
for Analysis and Tesdng of Industrial
E 260 Practice for Packed Column Gas Chromatograpfav3
3. Summary of Method
3.1 Anhydrous 1-propanol (see 10.3) is added as an
internal standard to suitable aliquot of the whole paint. The
aliquot is then diluted with dimethyUormamide andiniecied
onto a gas chromatographic column containing a porous
polymer packing that separates dichloromethane and I 1 1
mchloroethane from other volatile compounds.
•<• Significance and Use
4.1 1 Use of 1.1.1-trichloroethane and dichloromethane
which do not measurably contribute to the atmospheric
of ASTM Commune o-l
and ReUttd Onimp ud Mttcnte aad is the (tect i^StoTaf
minee DOI^l on Cheouoi Aulyuof Pasat and PaiatKtai^r
Cuncm edmott «purovui July 16.198S. Pubfataed October
2 IniMtt Book of ASTM Staitonit. Vd 11M. ^™w
1 Annual Book of ASTM Stewards. Vol 14.01.
Page 190
oxidant level, is a way for industry to meet government or
other regulations on volatile organic compounds. This test
method is designed to determine the content of these
haiohydrocarbon solvents in paints and coatings. That con-
tent can subsequently be used in calculating the volatile
organic compound content of a coating.
5. Apparatus
3.1 Chromatograph, any gas-liquid chromatographic in-
strument equipped with a thermal conductivity detector and
capable of being temperature programmed (see Table 1).
Optionally, a flame ionizanon detector may be used if the
sample is diluted so that no more than 1000 ppm each of
dichloromethane and 1,1,1-thchioroethane is piesent in the
injected specimen.
5.2 Recorder, a recording potentiometer with a fufl-scaie
deflection of 1 0 m V, a rail-scale response time of 2 s or less.
and a maximum noise of ±0.03 % of full scale.
5J Pre-Column. 40 in. (100 mm) long by Vi in. (3.2 mm)
outside t^atncicrya'Tif<^B STPCJ, packed with c*a** wool, fitted
on the entrance end of the column to retain any nonvolatile
materials and mtmTmr» sludge buildup in the column.
2.4 Column, 4 ft (L22 m) long by Vt in. (3.2 mm) outside
diameter stainless steel packed with 80/100 mesh (150 to
180 um) porous polymer packing matg"al* or other suitable
5.3 Liquid Charging Devices, such as microsyringes of
5-fiL or 10*uL capacity, rlfanrx? with acetone or other
suable solvent. Visually inspect for plugs or cracks before
and after each injection.
5.6 Vials. 25-mL. to minimm head space, capable of
being septum sealed.1
6. Reagents and Materials
6.1 Purny of Reagents — Reagent grade chemicals shall be
used in all tests, unless otherwise specified (as in 6.7). Unless
otherwise indicated, it is intended that ail reagents shall
conform to the specifications of the Committee on Analyt-
ical Reagents of the American Chemical Society, where such
4 taw* R». rafebfe from Wtten Aaocuta. Inc. MUfenL MA. fan been
fonadowfiaoryibrttapunxae.
»Mmnatvilva. rattle from ThePiaeeCbeniMiCo-flai HT.Rockfari.
IL 61IOJ. haw ben fend aufeaorr for ton
-------
D4457
speculations are available." Other grades may be used
provided it is first ascertained that me reagent is of sutFicieni
high punty to permit its use without lessening the accuracy
of the determination.
6.2 Carrier Gas. helium of 99.995 % or higher puntv
High punty nitrogen may also be used.
6.3 Dimetnviformamide (DMFi. reagent erade.
, 6.4 I-Propanol. gas chromatography spectrophotometnc
quality (see 10.5).
6.5 l.l.l-TrichJoroei(ianeixc6.7\
6:6. Dichloromeinane (see 6.7).
» 6.7 Halogenated Hydrocarbon Stabilizers— All commer-
cial grades of these haiogenated hydrocarbons contain stabi-
lizers. Either obtain the same solvent used in the coating ibr
use as the standard, or find the type and quantity of stabilizer
specified for use in the solvent of interest and add the
appropriate quantity to the pure solvent.
7. Safety Precautions
7.1 Dimeihylfonnamide is harmful if inhaled or afasoroea
through skin. Use only with aaequate ventilation Avoia
contact with skin. eyes, and clothing.
8. Preparation of Apparatus
8.1 Column Conditioning—The packed column is in-
stalled in the gas chromatographic unit leaving the exit end
disconnected from the detector. This will prevent anv
contamination of the detector with the column bleed. Set the
helium flow rate at 30 mL/min if a Vt in. (3.2 mm) outside
diameter column is used. Purge the column 5 to 10 min
before heating. Heat the column from room temperature to
200'C at S'C/min and hold this temperature for at least 12 h
(overnight). At the end of this period of time, heat the
column to 240'C at a 5'C/min rate and hold this tempera-
ture for several hours. The maximum temperature for this
ESS8 *.:~°'C- C°01 the column to 100*C an<* «heat to
240 C at a C/mm to observe the column bleed. Optimum
conditioning of this column may take several cycles of the
heating program before a good recorder baseline is achieved.
Conditioning of any column other than that suggested (5 4)
should be in accordance with the manufacturer's recommcn.
dations.
8.2 Install the column in the chromatograph and use the
information in Table 1 as a guide to establish the conditions
required to give the desired separation. Allow sufficient time
for the instrument to reach equilibrium as indicated bv a
stable recorder baseline. Adjust the carrier-gas flow to a
constant rate. Before each calibration and series of determi-
nations (or daily), condition the column at 200*C for I h
with earner-gas flow.
»
9. Calibration
9.1 Preparation of Standards—All standards, as well as
samples and blanks, should be at a constant temperature.
•Reagent Chemicals. Amencan Qu
it So
Soc_ Washington. DC. For sutsesuons OB me testing 01 reas^s°Mi toed bvTe
Amman Chemical Society, see -Reagent Chemteaii and Sunoaitis.- bv Jojeon
Rojin. D. Van Nostnnd Co_ Inc. New York. NY. ana the -Uuica
Pharmacopeia." ^^
TABLE 1 Typical instrument Conditions
Detector
Column
thermit conoucovnv
4 it (1.22 mi oy v* m. 13.2 mmi outsne
Temoerature. 'C
iniecoonoon
Detector DUCK
Column
inmal
Rnai
•Clmn
Camergas
Flow rate, mc/nvn
Soeamen size. uL
ui»i»i»i uacKN wiin au- mu mmi
porous poiymer oaaung
200
250
100
230 (tor 6 mmi
8
hetum
30
1
The given order of ingredient addition should be observed to
minimize loss of volatile ingredients.
9.1.1 Weighing to 1.0 mg, add 16.0 g of dimethyl-
formamide to a vial capable of being septum sealed. Add 2.0
g of l.l.l-tnchloroethane. 2.0 g of 1-propanol (see 10.5) and
2.0 g or' dichloromethane. Seal the vial with a crimp-on or
septum seal.
0 2 Determine the retenuon time of each component by
Lmeeting small amounts either separately or in known
mixtures. The components should eiute close to the typical
retention times given m Table i and the chromatograms
should closely approximate those shown in Fig. 1.
9.3 The area under each peak of the chromatogram is
considered a quantitative measure of the corresponding
compound. The relative area is proportional to concentra-
tion if the detector responds equally to all the sample
components. The response to different components is gener-
ally significantly different for both flame ionization and
thermal conductivity detectors and especially for flame
u
2 \
Li
PKL 1 Typical Chrematogram* ot Pttatt
Page 191
-------
D4457
.onnauon detectors. This difference m detector response
may be corrected by use of relative response factors obtained
by injecting and measuring the response of known blends.
For precise and accurate determination of the halogenated
hydrocarbons inject a 1 nL specimen of the standard in
accordance with the preparation in 9.1. Calculate the re-
sponse factors relative to unity for the halogenated hydrocar-
bons. iyurocar
10. Procedure
tO. 1 Keep all samples, blanks, and standards at a constant
temperature. Observe the given order of ingredient addition
to minimize loss of volatile ingredients. Shake paints, then
sample from the middle of the container
10.2 Weighing to 1.0 mg. add 16.0 g of dimethvi-
forraamide and 5.0 g of the paint to a vial capable of bein£
septum sealed. Add 2.0 g of 1-propanol (see 10.5). Seal the
vial with a cnmp-on or septum seal
10.3 Shake the vial Then to facilitate settling, centrifuge
using a low speed centrifuge at 1000 rpm for 5 rain.
I0.41nject a 1-uL specimen of the supernatant from the
prepared solution onto the chromatograpnic column m
accordance with the conditions established in 8.2. Record
the peaks of all components. ^cuuru
,. ia5-Ifthe ""Ponnon of the paint is unknown, test for
the presence of 1-propanoL Prepare a blank, omittinis the "> 0
M i~P£?^01 'V? Z aad injcct a ^ specimen. To this
bfcnk add 2.0 g of I-propanol and inject a 1-uL specimaZ
Then compare peak response to that from the test solution. If
1-propanoI is present in the paint, substitute a different
internal standard. Other possible internal standards include
alcohols, esters, and hydrocarbons.
10.6 If the composition of the paint is unknown, establish
whetherpeaksmteifenngwith I -propanol or the halogenated
hydrocarbons are present by using both the column specified
m 5.4 and a second column that yields different retention
times.
11. Calculation
1 1.1 Measure the area of all peaks (Note) and multiply
each area by the appropriate attentuation factor to express
the peak areas on a common
is recom-
«as may be deterauned by «ny method that meets the
preosoii reqmremeno of Section 12. Electron
mended for best icnin.
where:
RF, = correaed peak response for /'* component, area
units.
Wt = weight of i* component in the standard solution, g,
' Padonp from the Ponpik t
and the Chrem
lorUuipunxHe. Howrvrr. u a UK
chromatograpnic peak area for the i:H component m
the standard soluuon. area units.
then:
where:
CH wt
RFCH
CHwt%
ism
x 100
Vs-
IV
chlorinated hydrocarbon, weight %.
• response factor for the chlorinated hydro-
carbon in the standard solution, area units.
area of the chlorinated hydrocarbon peak in the
test solution, area units.
weight of internal standard added to the paint.
g.
response factor for the internal standard in the
standard solution, area units.
area of the internal standard peak in the test
solution, area units, and
specimen weight, g.
12. Precision and Bias8 (see also Practice £ 180)
12.1 Precision:
12.1.1 l.LJ-TriMoroethane—On the basis of an inter-
laboratory test of this test method in which one operator in
each of eight laboratories tested three coatings containing
from 31 to 65% 1,1,1-trichloroethane (theoretical), the
within-laboratory coefficient of variation was found to be
1.01 % relative at 20 degrees of freedom and the between-
laboratories coefficient of variation was found to be 2.72 %
relative at 17 degrees of freedom. Based on these coefficients.
the following criteria should be used for judging the accept-
ability of results at the 95 % confidence level:
12.1.1.1 Repeatability—Two results, each the mean of
duplicate runs, obtained by the same operator should be
considered suspect if they differ by more than 3.0 % relative.
12.1.1.2 Jleprodudbility—'Two results, each the mean of
duplicate runs, obtained by operators in different laborato-
ries should be considered suspect if they differ by more than
8.1 % relative.
12.1.2 Dichioromahane—Qn the basis of an inter-
laboratory test of this test method hi which one operator in
each of eight laboratories tested two coatings containing
from 32 to 78 % dichloromethane (theoretical), the within-
laboratory coefficient of variation was found to be 0.98 %
relative at 14 degrees of freedom and the between-labora-
tories coefficient of variation was found to be 5.16 % relative
a 12 degrees of freedom. Based on these coeffirirrm, the
following criteria should be used for judging the acceptability
of results at the 95 % confidence level:
12.1.2.1 Repeatability-'Two results, each the mean of
duplicate runs, obtained by same operator should be consid-
ered suspect if they differ by more than 3.0 % relative.
12.1.22 Reproduabiiiiy—Two results, each the mean of
duplicate runs, obtained by operators in different laborato-
ries should be considered suspect if they differ by more than
17.92 % relative.
122 Bias—Determination of a bias statement for this test
method is not practical at this time.
choose* column tan frasrmaieincaipeua.
iSupponraid«»«it«*mitobkftOTJASTMHeadauin«n.Reoa«lRJtDOI -
104].
Page 192
-------
D4457
Tne American Society lor Testing via Materials MMS no oeaaion resoecang me vauoav at any patent rignts asseneo in connection
**" *")'''*"' menaonea H> rm» stanoara. user* at ma stanoara art Marasaiy acmsao mai oaramwiaiioB of wa vaMMy erf any tucn
pmm ngnu. ana ffw nwt w mtrrngtomnr at suen ngm. vt vnmv thttt own nsoomtoiiity.
This sttnaara a suower to revision ar an/ rrrrw oy me resoonsioie tecnmeu committee ana must oe rawewao every five yean ana
rf nor retnsea. eiffler reaoorovea or witnartwn. Your comments are invitea eitner tor revision of rrw* stanava or tor aaaaionai stanoaros
ana snouta oe aaoresseo to ASTM Heeaauanen. Your comments wm receive careful consioeranon ar a meeting ot me resoonsioie
tecnmcai committee, wmeti you may amno. // you teal mat your comments nave not recenwo a /air Hearing you snouu mane your
view* Known to me ASTM Committee on Stanaaras. 7976 flace St.. Phitaaetonie. PA 19103.
Page 193
-------
SECTION 4
CITATIONS OF RM 24/24A IN 40 CFR 60
40 CFR PART 60-STANDARDS OF PERFORMANCE
FOR NEW STATIONARY SOURCES
Subpart A-General Provisions
60.8 Performance tests.
60.17 Incorporations by reference.
Subpart EE-Standards of Performance for Surface Coating of Metal Furniture
60.313 Performance tests and compliance provisions.
60.316 Test methods and procedures.
Subpart MM-Standards of Performance for Automobile and Light-Duty Truck Surface
Coating Operations
60.393 Performance test and compliance provisions.
60.396 Reference methods and procedures.
Subpart QQ-Standards of Performance for the Graphic Arts Industry: Publication
Rotogravure Printing
60.435 Test methods and procedures.
Subpart RR-Standards of Performance for Pressure Sensitive Tape and Label
Surface Coating Operations
60.443 Compliance provisions.
60.446 Test methods and procedures
Page 195
-------
Subpart SB-Standards of Performance for Industrial Surface Coating: Large
Appliances
60.453 Performance test and compliance provisions.
60.455 Reporting and recordkeeping requirements.
60.456 Test methods and procedures.
Subpart 77 - Standard of Performance for Metal Coil Surface Coating
60.463 Performance test and compliance provisions.
60.466 Test methods and procedures.
Subpart WW-Standards of Performance for the Beverage Can Surface Coating
Industry
60.493 Performance test and compliance provisions.
60.495 Reporting and recordkeeping requirements.
60.496 Test methods and procedures.
Subpart BBB-Standards of Performance for the Rubber Tire Manufacturing Industry
60.543 Performance test and compliance provisions.
60.547 Test methods and procedures.
Subpart FFF-Standards of Performance for Flexible Vinyl and Urethane Coating and
Printing
60.581 Definitions and symbols.
60.583 Test methods and procedures.
Page 196
-------
Subpart SSS-Standards of Performance tor Magnetic Tape Coating Facilities
60.713 Compliance provisions.
60.715 Test methods and procedures.
Subpart TTT-Standards of Performance tor Industrial Surface Coating: Surface
Coating of Plastic Parts tor Business Machines
60.723 Performance test and compliance provisions.
60.724 Reporting and recordkeeping requirements.
60.726 Delegations of authority.
APPENDIX A-REFERENCE METHODS
Page 197
-------
Page 198
-------
SECTION 5
BIBLIOGRAPHY
ASTM Procedures in Order of ASTM Number
1. Standard Specification for Industrial Grade Toluene, ASTM D362-84. 1990
Annual Book of ASTM Standards, Volume 6.03. Philadelphia, PA, 1990.
2. Standard Specification for Reagent Water, ASTM D1193-77. 1990 Annual
Book of ASTM Standards, Volume 6.03. Philadelphia, PA, 1990.
3. Standard Test Method for Water in Volatile Solvents (Fischer Reagent Titration
Method), ASTM D1364-87. 1990 Annual Book of ASTM Standards, Volume
6.03. Philadelphia, PA, 1990.
4. Standard Test Method for Density of Paint, Varnish, Lacquer, and Related
Products, ASTM D1475-85. 1990 Annual Book of ASTM Standards, Volume
6.01. Philadelphia, PA, 1990.
5. Standard Test Method for Volatile Content of Coatings, ASTM D2369-87.1990
Annual Book of ASTM Standards, Volume 6.01. Philadelphia, PA, 1990.
6. Standard Test Method for Volume Nonvolatile Matter in Clear or Pigmented
Coatings, ASTM D2697-86. 1990 Annual Book of ASTM Standards, Volume
6.01. Philadelphia, PA, 1990.
7. Standard Guide for Determining Volatile and Nonvolatile Content of Paint and
Related Coatings, ASTM D2832-83. 1990 Annual Book of ASTM Standards,
Volume 6.01. Philadelphia, PA, 1990.
8. Standard Specification for 2-Ethoxyethyl Acetate (99% Grade), ASTM D3728-
88. 1990 Annual Book of ASTM Standards, Volume 6.03. Philadelphia, PA,
1990.
9. Standard Test Method for Water Content of Water-Reducible Paints by Direct
Injection Into a Gas Chromatograph, ASTM D3792-86. 1990 Annual Book of
ASTM Standards, Volume 6.01. Philadelphia, PA, 1990.
10. Standard Practice for Sampling Liquid Paints and Related Pigmented Coatings,
ASTM D3925-81. 1990 Annual Book of ASTM Standards, Volume 6.01.
Philadelphia, PA, 1990.
Page 199
-------
11. Standard Practice for Determining Volatile Organic Compound (VOC) Content
of Paints and Related Coatings, ASTM D3960-89. 1990 Annual Book of ASTM
Standards, Volume 6.01. Philadelphia, PA, 1990.
12. Standard Practice for Intel-laboratory Testing of Paint and Related Materials,
ASTM D3980-88. 1990 Annual Book of ASTM Standards, Volume 6.01.
Philadelphia, PA, 1990.
13. Standard Test Method for Water in Paints and Paint Materials by Karl Fischer
Method, ASTM D4017-88. 1990 Annual Book of ASTM Standards, Volume
6.01. Philadelphia, PA, 1990.
14. Manual Sampling of Petroleum and Petroleum Products, ASTM D4057-81.
1990 Annual Book of ASTM Standards, Volume 5.03. Philadelphia, PA, 1990.
15. Standard Test Method for Determination of Dichloromethane and 1,1,1-Trichlo-
roethane in Paints and Coatings by Direct Injection into a Gas Chromatograph,
ASTM D4457-85. 1990 Annual Book of ASTM Standards, Volume 6.03.
Philadelphia, PA, 1990.
16. Specification for Gravity Convection and Forced-Ventilation Ovens, ASTM
E145-68. 1990 Annual Book of ASTM Standards, Volume 14.02. Philadelphia,
PA, 1990.
17. Practice for Determining the Precision Data of ASTM Methods for Analysis and
Testing of Industrial Chemicals, ASTM E180-85. 1990 Annual Book of ASTM
Standards, Volume 15.05. Philadelphia, PA, 1990.
18. Standard Test Method for Water Using Karl Fischer Reagent, ASTM E203-75.
1990 Annual Book of ASTM Standards, Volume 15.05. Philadelphia, PA, 1990.
19. Standard Practice for Sampling Industrial Chemicals, ASTM E300-86. 1990
Annual Book of ASTM Standards, Volume 6.02. Philadelphia, PA, 1990.
20. Excerpts from Standard Practice for Use of the International System of Units
(SI) (the Modernized Metric System), ASTM E380-89. 1990 Annual Book of
ASTM Standards, Volume 6.01. Philadelphia, PA, 1990.
21. Standard Practice for Conducting an Interlaboratory Study to Determine the
Precision of a Test Method, ASTM E691-87. 1990 Annual Book of ASTM
Standards, Volume 6.01. Philadelphia, PA, 1990.
Page 200
-------
22. Standard Guide for ASTM Standard Specification Quality Statements, ASTM
E1267-88. 1990 Annual Book of ASTM Standards, Volume 6.01. Philadelphia,
PA, 1990.
Other Literature/Bibliographical References
23. Federal Test Method Standard No. 141c, Paint, Varnish, Lacquer, and Related
Materials; Methods of Inspection, Sampling, and Testing. General Services
Administration, January 24,1986.
24. EPA-340/1-84-001 a, VOC Sampling and Analysis Workshop, Volume I VOC
Reference Methods. U.S. EPA, OAQPS, SSCD, Washington, DC, September
1983.
25. EPA-340/1-86-016, A Guideline for Surface Coating Calculations. U.S. EPA,
OAQPS, SSCD, Washington, DC, July 1986.
26. EPA 340/1-88-003, Recordkeeping Guidance Document for Surface Coating
Operations and the Graphics Arts Industry, U.S. EPA, SSCD, Washington, DC,
May 1989.
27. EPA-450/2-77-008, Control of Volatile Organic Emissions from Existing Station-
ary Sources - Volume II: Surface Coating of Cans, Coils, Paper, Fabrics,
Automobiles, and Light-Duty Trucks. U.S. EPA, OAQPS, Research Triangle
Park, NC, May 1977.
28. EPA-450/3-84-019, Procedures for Certifying Quantity of Volatile Organic
Compounds Emitted by Paint, Ink, and Other Coatings. U.S. EPA, OAQPS,
ESED, Research Triangle Park, NC, December 1984.
29. EPA-600/4-77-027b, Quality Assurance Handbook for Air Pollution Measure-
ment Systems: Volume III. Stationary Source Specific Methods. U.S. EPA,
EMSL, Research Triangle Park, NC, November 1976.
30. EPA-600/9-76-005, Quality Assurance Handbook for Air Pollution Measure-
ment Systems: Volume I. Principles. U.S. EPA, EMSL, Research Triangle Park,
NC, December 1984.
31. Code of Federal Regulations, General Provisions, Performance Tests, 40 CFR
60.8 Subpart A, Washington, DC, 1989.
32. Code of Federal Regulations, General Provisions, Incorporation by Reference,
40 CFR 60.17 Subpart A, Washington, DC, 1989.
Page 201
-------
33. Standards of Performance for Surface Coating of Metal Furniture, 40 CFR 60
Subpart EE, Washington, DC, 1989.
34. Standards of Performance for Automobile and Light-Duty Truck Surface
Coating Operations, 40 CFR 60 Subpart MM, Washington, DC, 1989.
35. Standards of Performance for the Graphic Arts Industry: Publication Rotogra-
vure Printing, 40 CFR 60 Subpart QQ, Washington, DC, 1989.
36. Standards of Performance for Pressure Sensitive Tape and Label Surface
Coating Operations, 40 CFR 60 Subpart RR, Washington, DC, 1989.
37. Standards of Performance for Industrial Surface Coating: Large Appliances, 40
CFR 60 Subpart SS, Washington, DC, 1989.
38. Standards of Performance for Metal Coil Surface Coating, 40 CFR 60 Subpart
TT, Washington, DC, 1989.
39. Standards of Performance for the Beverage Can Surface Coating Industry, 40
CFR 60 Subpart WW, Washington, DC, 1989.
40. Standards of Performance for the Rubber Tire Manufacturing Industry, 40 CFR
60 Subpart BBB, Washington, DC, 1989.
41. Standards of Performance for Rexible Vinyl and Urethane Coating and Print-
ing, 40 CFR 60 Subpart FFF, Washington, DC, 1989.
42. Standards of Performance for Magnetic Tape Coating Facilities, 40 CFR 60
Subpart SSS, Washington, DC, 1989.
43. Standards of Performance for Industrial Surface Coating: Surface Coating of
Plastic Parts for Business Machines, 40 CFR 60 Subpart TTT, Washington, DC,
1989.
44. Code of Federal Regulations, Reference Methods 24 AND 24A, 40 CFR 60,
Appendix A, Washington, DC, 1989.
45. Manual on Determination of Volatile Organic Compounds in Paints, Inks, and
Related Coating Products, J. John Brezinski, ed., ASTM Manual Series: MNL
4, 1989.
Page 202
-------
TECHNICAL REPORT DATA
(flease read Instructions on the reverse before completing/
1. REPORT NO.
EPA 340/1-91-012
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
£PA Reference Methods 24 and 24A
Compilation of Procedures/References
5. REPORT DATE
Mav 1991
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Bruce A. Olson, Melinda K. Wood, John T. Chehaske
8. PERFORMING ORGANIZATION REPORT NO.
91-133-T4/R
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Pacific Environmental Services, Inc. (PES)
560-Herndon Parkway, Suite 200
Herndon, Virginia 22070-5225
10. PROGRAM ELEMENT NO.
WA 91-133
11. CONTRACT/GRANT NO.
68-02-4464
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Stationary Source Compliance Division
Washington. D.C.
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
IS. SUPPLEMENTARY NOTES $QQ 3lSO:
SSCD, Organic Chemicals Section, 401 M St., S.W. (EN-341-W) EPA-340/1-91-010 Sampling
Washington. D.C. 20460 EPA-340/1-91-011 Analysis
16. ABSTRACT
Documented procedures for use with coating and graphics arts samples taken for
analysis of coating materials by Reference Methods 24 or 24A (RM 24/24A) as found in
40 CFR 60, Appendix A are presented in this compilation report. This document is to
be used as a source of reference information for EPA compliance determinations and
enforcement activities related to VOC regulations.
Contents of this report include: 1) a reproduction of EPA Reference Methods
24 and 24A from 40 CFR 60, Appendix A; 2) a compilation of ASTM Procedures pertaining
to the application of sampling and analytical methods relevant to the use of RM 24/24A;
3) a list of citations of RM 24/24A from the Standards of Performance for New
Stationary Sources (NSPS); 4) an EPA report titled: "Procedures for Certifying
Quantity of Volatile Organic Compounds Emitted in Paint, Ink, and Other Coatings"
(EPA-450/3-84-019, December 1984, Revised June 1986); and 5 a list of references.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN SNDED TERMS
c. COSATi Field, Group
Air Pollution
Coatings
Analysis
Solvents
Sampling
VOC
ASTM Methods
Sampling of Coatings
and Inks
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS I This Report)
Unclassified
21. NO. OF PAGES
202
20. SECURITY CLASS {Tliiipafr:
Unclassified
22. PRICE
EPA Form 2220-1 (R«». 4-77) PMKVIOUS COITION is OBSOLETE
------- |