United States EPA-600/2-88-026b
Environmental Protection
A9encv April 1988
v>EPA Research and
Development
DEVELOPMENT OF PROPOSED
STANDARD TEST METHOD FOR
SPRAY PAINTING TRANSFER EFFICIENCY
Volume II. Verification Program
Prepared for
Office of Air Quality Planning and Standards
Prepared by
Air and Energy Engineering Research
Laboratory
Research Triangle Park NC 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-S8-02Sb
April 1988
DEVELOPMENT OF PROPOSED STANDARD TEST METHOD
FOR SPRAY PAINTING TRANSFER EFFICIENCY
VOLUME II. VERIFICATION
PROGRAM
BY
K. C. KENNEDY
CENTEC Corporation
Reston, Virginia 22090
EPA Contract Number 68-03-1952
EPA Project Officer
Charles H. Darvin
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Prepared for:
U. S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C. 20460
-------
ABSTRACT
Over the past 5 years, the Environmental Protection Agency Air
and Energy Engineering Research Laboratory has been working to
develop a standardized laboratory test method for determining the
transfer efficiency of spray painting operations. This document
describes the final phase of laboratory experiments conducted to
characterize the interlaboratory precision of the transfer
efficiency test method developed in earlier efforts.
The test program included extensive experiments conducted at
eight industrial spray painting laboratories. Three types of
spray equipment (conventional air spray, electrostatic air spray,
and airless) were tested at each laboratory. Six replicate
transfer efficiency measurements were made for each equipment
type at each laboratory.
The results of these experiments document the maturity of the
draft standard transfer efficiency test method and the expected
ruggedness of the res.ults to differences within and between
laboratories. As anticipated from earlier research efforts, the
transfer efficiency results for each spray system were differ-
ent. However, the results for each spray system demonstrated
exceptional consistency when expressed as within-laboratory
standard deviation. (Standard deviation is expressed in units of
transfer efficiency. It can be used for estimating precision at
various confidence intervals.) The within-laboratory standard
deviation across eight laboratories was:
Conventional air spray .... 1.52
Electrostatic air spray ... 1.91
Airless spray ............. 1.10
These within-laboratory standard deviations clearly demonstrate
the capability of the test method to produce consistent results
within a particular laboratory. The within-laboratory standard
deviations were well below the value predicted at the onset of
this project, 2.5.
The total standard deviations were considered from two
standpoints. The first standpoint included the deviations
directly attributable to differences among the spray guns used
for the tests. These values were based on the total variance,
including the within-laboratory portion of the variance (shown in
the table above), gun-to-gun portion of the variance, and
between-laboratory portion of the variance. The total standard
deviations are reported below.
Conventional air spray .... 6.79
Electrostatic air spray ... 9.42
Airless spray 5.82
ii
-------
The second standpoint included the within-laboratory portion of
the variance and the between-laboratory portion of the variance.
However, it excluded the gun-to-gun portion of variance (that is,
the part attributable directly to differences among the spray
guns). The total standard deviation (excluding gun-to-gun
differences) are presented below.
Conventional air spray .... 6.72
Electrostatic air spray ... 8.70
Airless spray 5.26
The exclusion of the gun-to-gun portion of the variance
marginally improves the total standard deviation.
While arguments can be made towards including and excluding the
gun-to-gun differences in the overall analysis, it makes little
impact on the results of this study. Between-laboratory
variances accounted for the vast majority of the total standard
deviation for all equipment types at all ranges of transfer
efficiency observed.
For the convenience of the reader, total standard deviations are
referred to in this report as either including gun differences or
excluding gun differences.
In classical interlaboratory programs there are two measures of
the quality of the method: accuracy and precision. Precision is
the measure of variability. The precision goals and results of
this research have been discussed above and are presented in
detail herein. Accuracy is the measure of how far off the
observed values of transfer efficiency are from the true transfer
efficiency. In this research there is no known true measure of
transfer efficiency; therefore, accuracy cannot be addressed.
However, since accuracy is a measure of the bias encountered in
estimating the value of a parameter (and because there is no
reason to believe that we have a significant bias for the spray
system, laboratories, and targets examined), it is believed that
the draft transfer efficiency test method is reasonably accurate.
The absence of evidence regarding bias may be interpreted as an
absence of bias.
111
-------
TABLE OF CONTENTS
Abstract n
List of Tables vi
Abbreviations and Symbols vii
Acknowledgements viii
1. Introduction 1
2. Conclusions and Recommendations 2
3. Background 3
4. Phase I - Airless Transfer Efficiency Test Method 5
5. Phase II - Design of Interlaboratory Experiments 6
6. Performance of Experiments 19
Test Site 1 22
Test Site 2 30
Test Site 3 36
Test Site 4 39
Test Site 5 42
Test Site 6 45
Test Site 7 48
Test Site 8 . . 51
7. Test Results and Statistical Analysis 54
Appendices
A Proposed Standard Transfer Efficiency Test Method A-l
B Quality Assurance/Quality Control Plan B-l
C Detection of Outliers by Means of Nalimov's Test . C-l
D Statistical Analysis Addendum D-l
E Operating Conditions and Raw Data Summary .... E-l
-------
LIST OF TABLES
Page
5-1 Level of Support Required from Participating
Laboratories
5-2 Paint Specifications .......«•••••• ^'
6-1 Order of Airless Gun Testing ......•••• 24
6-2 Order of Electrostatic Air Spray Gun Testing . . 26
6-3 Order of Conventional Air Spray Gun Testing . . 27
6-4 Laboratory 1 Test Results ........... 28
6-5 Laboratory 2 Test Results ........... 32
6-6 Number of Laboratories (b) Required to be P%
Sure that the Estimated Total Variance is
Within K% of the True Value ......... 35
6-7 Laboratory 3 Test Results ........... 37
6-8 Laboratory 4 Test Results .....,• 40
6-9 Laboratory 5 Test Results ........... 43
6-10 Laboratory 6 Test Results ........... 47
6-11 Laboratory 7 Test Results ........... 50
6-12 Laboratory 8 Test Results ........... 53
7-1 Transfer Efficiency Results, % Transfer efficiency.
Airless ....... 55
7-2 Transfer Efficiency Results, % Transfer efficiency,
Conventional air spray ........... 55
7-3 Transfer Efficiency Results, % Transfer efficiency,
Electrostatic air spray 57
7-4 Transfer Efficiency Results .......... 58
VI
-------
ABBREVIATIONS AND SYMBOLS
LIST OF ABBREVIATIONS AND UNIT CONVERSIONS
ABBREVIATIONS
AOAC — Association of Official Analytical Chemists
ASTM — American Society for Testing and Materials
EPA — United States Environmental Protection Agency
Fan air — shaping air or horn air
FP — flat panel (target configuration)
PSIG — pounds per square inch, Ib/in^, gauge
O&M — operating and maintenance
QA/QC — quality assurance/quality control
VOC — volatile organic compounds
vii
-------
ACKNOWLEDGEMENTS
The authors wish to acknowledge the assistance and cooperation of
the manufacturers and suppliers of the paint (Glidden) and spray
equipment (Graco Inc. and Wagner Spray Tech) used in this
project. A special acknowledgement is made of the participating
spray painting laboratories where the testing was conducted for
the many courtesies extended to the project team:
The DeVilbiss Company
Exxon Chemical Company
Graco Inc.
Kremlin Incorporated
Nordson Corporation
PPG Industries Incorporated
Ransburg Electrostatic Equipment Inc.
BASF - Inmont (formerly United Technologies - Inmont)
Acknowledgement is made of the interest of the Chemical Coaters'
Association in this research; their assistance in locating spray
painting laboratories is appreciated.
Special acknowledgement also is made of the contributions of the
members of the Transfer Efficiency Project Steering Committee.
viii
-------
SECTION I
INTRODUCTION
Spray painting transfer efficiency is a measurement of that
quantity of paint solids which actually coats a surface compared
with the total paint solids sprayed. Transfer efficiency measurements
can be used to optimize on-line spraying or to develop more
efficient spray equipment. More recently, the need to determine
transfer efficiency has taken on a new aspect: transfer
efficiency can be used to quantify volatile organic compound
emissions from spray painting operations.
During the past 5 years, the U.S. EPA has been conducting
extensive research on transfer efficiency. The objective of this
research has been to develop a laboratory transfer efficiency
measurement method. Many companies have developed their own
methods for determining efficiency; however, these methods
vary widely in capability although most share common elements.
The EPA research program was designed and initiated to develop
the necessary background and research data to permit development
of a standardized laboratory transfer efficiency test method.
To ensure as broad participation in -the program as possible,
numerous sources in the industry were contacted and their
assistance solicited where possible.
Earlier research developed a laboratory transfer efficiency
measurement method. The approach used was to develop the test
method by studying concepts for transfer efficiency determination
and methods currently in use and then using the best features of
each. Laboratory tests were conducted to provide supporting data
and to establish the precision of the formulated method. The
tests established the standard deviation (repeatability) of the
transfer efficiency results as less than 2.5 within a given
laboratory. (Standard deviation is always expressed in the units
being measured by the test. In this research it is always
expressed in units of transfer efficiency.)
The research program described in this document was conducted to
establish the efficiency of the transfer efficiency method using
the preliminary draft method defined earlier in this program.
Eight field-based laboratories participated in this test program.
Electrostatic air spray, conventional air spray, and airless
spray systems were tested at each laboratory.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
The preliminary draft transfer efficiency test method was used at
eight laboratories. Within each participating laboratory the
results were repeatable at levels well below the standard
deviation goal of 2.5 set by previous research (CENTEC
Corporation, 1982).
A statistical analysis of the results showed that the gun portion
and the within-laboratory portion of the total variance were
small. The between-laboratory portion of the variance was over
six times larger than the within-laboratory portion of the
varianceo This ratio implies that the differences between
laboratories were real and resulted in detectable, quantifiable
differences in transfer efficiency that can be attributed to
differences between participating laboratories.
The toal standard deviations observed from laboratory to
laboratory ranged from 5.82 (airless spray) to 9.42 (electro-
static spray). The preliminary results of this research indicate
that the probability that the transfer efficiency measured at a
random qualifying laboratory would fall within 7.3 to 12.0
transfer efficiency units of the true transfer efficiency,
provided that the assumption that bias is nil is correct.
These results must, however, be considered preliminary. A
sufficient number of laboratories was not used in the program
to comply with an 80 percent probability criteria to establish
for method precision.
-------
SECTION 3
BACKGROUND
The EPA has been attempting to develop regulatory strategies to control
the emissions of VOC from metal coating processes of the metal finishing
industry. In those processes where the emission streams are easily
defined and contain relatively high concentrations of VOC, development of
regulations has been straight-forward. Those regulations are based
primarily upon the ease with which presently available control technologies
can be used to control high VOC concentration emission streams. Those
technologies include carbon adsorption and incineration.
Although high concentration emission streams can be controlled effectively
with available technologies, as VOC concentrations decrease the available
control system's economic and technical feasibility also decreases.
Questions have also been voiced by representatives of industry and
government about the relationship of various metal coating techniques to
their VOC emissions potential.
To date no method has been certified as a standard to measure transfer
efficiency. Numerous companies and technical organizations, however,
have proposed methods for measuring transfer efficiency. These have yet
to be shown to have sufficient precision and accurary to define transfer
efficiency for most painting scenarios. Thus, tranfer efficiency measurements
have not become a part of present control strategies for compliance with
VOC emissions regulations. Only when a simple yet accurate and precise
method of measuring painting transfer efficiency is developed can TE be
used in VOC control strategies.
Beginning in 1982, a contract was authorized by EPA to develop a
preliminary transfer efficiency test method. Laboratory studies
and evaluations were conducted to define a procedural method to
accurately and precisely measure transfer efficiencies from
various spray painting equipment. This equipment included
-------
conventional air spray, electrostatic air spray, and rotating
bell spraying equipment. Results of these studies are presented
in Volume I.
The current project was conducted in two parts. The first part
addressed the status of the transfer efficiency method for
airless spray equipment^ which was not researched during previous
phases of this program. The second part involved the implemen-
tation and validation testing at multiple laboratories to
estimate the precision of the transfer efficiency test method.
-------
SECTION 4
PART I - AIRLESS TRANSFER EFFICIENCY 'TEST METHOD
In the first part of this contract, background information on
existing methods for determining the transfer efficiency of
airless spray painting equipment was developed. Information
from publicly available sources was obtained through a manual
and computer literature search. A wide spectrum of private
industry sources was contacted, including major spray painting
equipment manufacturers, industrial metal finishers and spray
painters, and paint formulators. These sources were questioned
about different methods for determining the transfer efficiency
of airless equipment. The general consensus was that there
were no characteristics unique to airless spray which would
make the current test method unacceptable except the need to
determine mass flow with a meter rather than using scales and
a stopwatch. In airless spraying, the paint pot is connected
via high pressure hose to a reciprocating pump. This connection,
and the action of the pump, make the scales vibrate. This
vibration can be severe enough to make reading the scales
difficult if not impossible.
Preliminary test results of the transfer efficiency procedure
defined in appendix A, found that a reasonable degree of precision
could be achieved. The standard deviation of replicate flat
panel target test runs using airless spray guns was 1.31 and the
standard deviation of replicate vertical cylinder test runs was
0.03, both well within the range of 2.5 that was specified in the
preliminary draft test procedure. Therefore, it was concluded
that the existing draft transfer efficiency test method was
appropriate for airless spraying equipment even though it was not
as well developed as for other spray painting systems.
To document this conclusion, a conventional airless spray system
was selected as one of the three spray equipment types to be used
for the second part of this study. The selection of airless
spray equipment also helped to ensure that the preliminary draft
transfer efficiency test method would be thoroughly developed
across a wide range of transfer efficiencies.
-------
SECTION 5
PART II - DESIGN OF INTERLABORATORY EXPERIMENT
PROGRAM DESIGN
The purpose of the experimental strategy of this effort was to build an
interlaboratory test program that allowed estimation of the precision of
the draft standard transfer efficiency test method. For this purpose, no
operating parameters were intentionally varied during the program.
The strategy was divided into four major components: establishing the
number of laboratories for the test program, deciding what equipment
types were to be tested at each laboratory, estimating the number of
replicates to be made for each equipment type at each laboratory, and
establishing the gun portion of the variance. The experimental strategy
called for replicate transfer efficiency determinations to be made at
each laboratory participating in the program. These determinations were
to be made using the same type of equipment (in similar condition) at the
same spray conditions (controlled and held constant as provided in the
QA/QC plan in Appendix B). These conditions were established during the
first laboratory experiment and were held constant for all subsequent
laboratory experiments in this program. Equipment operating conditions
were set at the first laboratory to provide a good spray pattern and
reasonable finish. The spray environment (i.e., the configuration of the
laboratory, proximity of grounds, airflow patterns, and so forth)
necessarily was less controlled from laboratory to laboratory than were
the spray conditions. Ideally, interlaboratory method verification
experiments would be conducted at nearly identical laboratories using
identical spray equipment and identical operating conditions. However,
from a practical standpoint it is impossible to have 100 percent identical
laboratories. Thus, it might be expected that the resulting transfer
efficiency valves will be somewhat different between laboratories.
ESTABLISH THE NUMBER OF LABORATORIES
The number of laboratories required to determine the precision of a test
method within certain estimated confidence limits may be estimated from
existing data generated from the same analysis performed at different
laboratories. In this program, however, the existing data base could not
be used for estimation of the number of laboratories because previous
data were not obtained at the same conditions. Known and quantifiable
differences existed between transfer efficiency tests conducted at four
different laboratories under prior research efforts. Further, other
documented transfer efficiency determinations performed by industry could
not be documented as meeting required project QA/QC plans; their results
also could not be used to estimate the number of laboratories required.
-------
The results of transfer efficiency determinations made during
this program at the same laboratories as prior studies could not
be compared to the results of the prior research. The spray
equipment, paint system, and operating conditions for this
program were not the same as for prior efforts; thus, the results
are not comparable. They are from different sets.
Since previous data were obtained at inconsistent spray painting
conditions, an experimental design was undertaken that was not
dependent upon prior data and was most efficient of resources.
The experimental methodology meeting these criteria was the
sequential experimental design. In the planned sequential
experimental design, the same test was conducted at two
laboratories. The data from these two laboratories served as a
basis for further sample size projection. The results from these
two laboratories truly reflected the most current level of
knowledge.
Appropriate computations of number of laboratories required were
made from the transfer efficiency results of the first two
laboratory tests. The data from the first two test laboratories
served as a portion of the final data set to be analyzed. This
approach was consistent with the current trend in the field of
experimental design, i.e., design the experiment in stages, using
preliminary stages to make decisions regarding design parameters
for the final stage (Steinberg and Hunter, 1984). This
sequential experimental design allowed the program to avoid
relying too heavily on the use of preliminary data that might be
suspect or incompatible with the conditions anticipated in this
study.
Based on the preliminary calculation described above, it was
expected that this sequential methodology would result in selec-
tion of approximately eight laboratories for testing in the
interlaboratory program. (The number of laboratories partici-
pating under this contract was fiscally limited to eight.) This
estimate was also based on recommendations of the American
Society for Testing and Materials in their Standard Practice for
Conducting an Interlaboratory Test Program to Determine the
Precision of Test Methods (ASTM E 691-79, Part 41). This
estimate was further supported by the Association of Official
Analytical Chemists in their statistical manual (Steiner and
Youden, 1975).
Selection of laboratories for participation in this program was
based primarily upon the availability of necessary facilities;
this selection process was necessarily assumed to be random for
purposes of the statistical analysis.
ESTIMATE THE NUMBER OF REPLICATES
The number of replicates (n) for each gun type must be sufficient
to reduce the variance of the test precision to an acceptable
value. In this program, parameters were chosen for an 80 percent
-------
confidence interval that the total variability would be within a
certain range. In that confidence interval, the element that
drives down" the total variability is the number of laboratories.
The number of replicates at each laboratory (based on within-
laboratory estimates of precision from previous experimental
data) is not as important. As shown in the equation below, the
total variance (excluding gun variance, which is addressed
separately) is much more sensitive to the number of laboratories
than to the number of replicates.
2aw4 2(aw
2
_ _
b(n-l) n(b-l)
where aT = total variance (excluding gun variance)
CTW = within laboratory variance
crt = between-laboratory variance
n = number of replicates at each laboratory
b = total number of laboratories
Based on this equation, one can see that changing the number of
replicates from six to ten hardly affects the confidence
interval, while increasing the number of laboratories from six to
seven shrinks the confidence interval significantly.
Six replicates were considered adequate to meet this criterion
based on previous experiments. Thus, six replicate transfer
efficiency measurements were made for each equipment type in
each laboratory under equivalent conditions for equivalent
equipment types, as best as could be controlled.
ESTABLISH EQUIPMENT TYPES FOR TEST PROGRAM
Although the precision of the test method may be estimated
by testing a single spray system across a number of independent
laboratories, this approach would leave some questions as to the
consistency of interlaboratory precision over a wide range of
transfer efficiencies. It was important to establish that the
test method was as precise at low transfer efficiencies as it
was for high transfer efficiency values. To avoid this question,
three spray types (having a wide range of anticipated transfer
efficiency values) were selected for this research. The spray
system types were electrostatic air spray, conventional air
spray, and airless. As documented in prior research (cited
above), electrostatic air spray and conventional air spray
painting systems typically have considerably different
transfer efficiencies even when operating in the same booth,
using the same test paint and targets. For instance, during
testing conducted in June 1983, conventional air spray
equipment exhibited transfer efficiencies ranging from 10
8
-------
percent to 60 percent, while electrostatic air spray equipment
exhibited transfer efficiencies from 75 percent to 95 percent.
Both spray systems used the same test paint, in the same spray
booth, with the same sets of target configurations.
Electrostatic air spray and conventional air spray systems
operate on different principles: conventional air spray
equipment relies on the paint particle size and mass velocity to
carry it to the desired target, while electrostatic air spray
equipment uses an electrical attraction in addition to conven-
tional attractive forces to draw the paint to the desired
target. These differences make the selection of electrostatic
air spray and conventional air spray systems desirable for this
research. It allowed demonstration of the ruggedness of the test
method to different spray mechanisms across a number of labora-
tories .
An additional advantage of selecting electrostatic air spray and
conventional air spray equipment was that equipment costs were
minimized. As recommended by the Spray Painting Transfer
Efficiency Project Steering Committee in March 1985,
electrostatic air spray equipment was used both as electrostatic
equipment and to simulate conventional air spray equipment by
turning off the electrostatics. In fact, these types of spray
equipment (as offered by one manufacturer) were virtually
identical except for the electrode on e'lectrostatic spray guns.
The decision to simulate conventional air spray equipment by
using electrostatic air spray with the electrostatics turned off
saved over $5,200 in equipment costs. It resulted in further
savings by eliminating the amount of time needed to change
equipment types (twice) at each laboratory.
A third equipment type, (conventional) airless, was also included
in the test program as a result of Part I recommendations.
Airless spray equipment does not have the test history of other
equipment types. In order to ensure that the test method was a's
well developed for airless spray equipment as it was for other
equipment types, airless equipment was included in this test
program.
Automatic guns were recommended by the Spray Painting Transfer
Efficiency Project Steering Committee, since they had not been
fully tested using the draft standard transfer efficiency test
method; however, automatic spray equipment was not available
within the time frame and budget available for this program.
Manual spray equipment was used for all tests in this program.
The spray guns were fixed in position using a mounting pole, and
were triggered manually with the exception of an actuator
(automatic triggering device) used at one laboratory on the
manual spray guns.
-------
ESTABLISH THE GUN PORTION OF THE VARIANCE
By nature, the interlaboratory test program involved almost
simultaneous testing at up to eight sites. This requirement
necessitated the use of different spray guns (of the same make
and model) from laboratory to laboratory* It was desirable to
know the portion of the between-laboratory variance that was due
strictly to differences in spray guns. This variance is called
the "gun-to-gun portion of the variance" or "gun portion of the
variance." The gun portion of the variance was considered part
of the between-laboratory variance? it is never possible to test
the identical gun at identical conditions at two laboratories.
Even the same gun would have undergone additional wear in the
process of being tested.
The gun portion of the variance had to be established to
determine whether it was significant and, if so, to determine
what portion of the total variance it represented.
The gun portion of the variance was established during the first
laboratory testo Eight different spray guns (of same make and
model) for each equipment type were subjected to replicate
transfer efficiency determinations at the same laboratory. (These
same guns were used in subsequent laboratory tests in this pro-
gram.) Six replicates were conducted for each gun. The data
were analyzed to determine how much difference in transfer
efficiency was attributable to using different guns.
At the first test, mean transfer efficiencies and standard
deviations from each of the three sets (one set for each
equipment type) of 48 transfer efficiency determinations (eight
guns per equipment type multiplied by six replicates for each
gun) were compared. The gun portion of the variance was derived
as described in Section 7 - Test Results and Statistical
Analysis.
LABORATORY REQUIREMENTS AND SELECTION
Spray painting laboratories were required to meet certain minimum
criteria in order to participate in the interlaboratory test
program. These criteria included their ability to provide the
laboratory conditions and support listed in Table 5-1, their
willingness to participate in a timely manner, their laboratory
rental cost, and their level of interest in the project. Best
professional judgment and previous test experience were used to
determine which laboratory conditions could reasonably be
controlled from test to test. The Steering Committee contributed
to defining laboratory conditions for testing* Part of the
between-laboratory variance, then, accounted for uncontrollable
differences from laboratory to laboratory. If more laboratory
conditions had been specified (such as relative humidity), or
condition ranges had been more closely defined (such as booth air
rate at 95-105 fpm instead of 80-120 fpm), no known laboratories
would have qualified to perform the test. Laboratory selection
10
-------
TABLE 5-1. LEVEL OF SUPPORT REQUIRED FROM PARTICIPATING
LABORATORIES
o Two technicians, knowledgeable and proficient with spray
system use and maintenance (minimum)
o Back-draw water wash spray booth or equivalent, with
80-120 fpm air velocity in middle at plane of target
(if dry filter booth is used, laboratory must provide
sufficient filters to maintain these conditions),
associated chemicals and operating costs
o Adjustable speed (20 to 40 fpm) overhead conveyor system
capable of hanging EPA targets as specified
o Conveyor speed measurement equipment
o Utilities
o Paint mixing equipment and facilities for documenting
paint characteristics (contractor provided Ford viscosity
cup)
o Curing oven of sufficient size for curing EPA targets,
with temperature control at 375*F
o Curing rack
o Timer
o Cleaning solvents
o Packing and shipping to next facility
o Air rate measurement equipment
o Humidity and temperature measurement equipment
o Foil cutters
o Foil handling facilities
o Laboratory scales (O.OOlg accuracy)
o Hangers for targets
o Compressed air supply in laboratory
o Temperature control in laboratory (65 - 75°F)
Relative humidity was recorded but could not be
controlled at available laboratories.
o Spray equipment assembly tools
11
-------
TABLE 5-1. LEVEL OF SUPPORT REQUIRED FROM PARTICIPATING
LABORATORIES (Continued)
o Copying machine (for completed data sheets)
o Work space for record keeping and calculations
o Security for test equipment and supplies
o Miscellaneous hose and fittings
12
-------
was assumed to be random within the defined sample frame,
although there was no absolute method to confirm the assumption.
Participating laboratories were required to supply at least two
knowledgeable and conscientious technicians for the test pro-
gram. Technicians were required to be familiar with the test
equipment and capable of performing the tests precisely, as
described in the Test Plan supplied to each laboratory.
In previous testing it was found that the transfer efficiency was
influenced by the linear booth air velocity. In order to have
consistent results, it was necessary to control the air velocity
as much as possible. At the suggestion of the Spray Painting
Transfer Efficiency Project Steering Committee, it was decided
that a qualifying air rate of 100 fpm +_ 20 fpm at the plane of
the target in the center of the booth would be required for all
participating laboratories.
An adjustable speed overhead conveyor system, a curing oven, and
hangers large enough to handle the EPA targets were required to
be supplied by the laboratories. The EPA targets consisted of
ten 15.24 cm (6 in) wide metal panels mounted 15.24 cm (6 in)
apart, and hanging 121.9 cm (48 in) in length (see Volume I).
The importance of an adjustable speed overhead conveyor system
can not be overemphasized because it provided the means to
control the speed of the targets passing through the spray area.
Since the conveyor speed is a controlled parameter, it
necessitated the need to have conveyor speed measurement
equipment at each participating laboratory- The degree of
sophistication required was such that using a stopwatch and
timing marks was an acceptable measurement method as allowed in
the procedure.
Because this testing involved a standard test method, measurement
and/or control of other pertinent parameters was essential. The
temperature of the testing area could affect the Transfer
efficiency and therefore was controlled to within a few degrees
centigrade (that is, +5*F). Air rate measurement equipment
(anemometer), humidity measurement equipment, and laboratory
scales with O.OOlg accuracy were included as part of the level of
support needed from participating laboratories.
Finally, all interested laboratories were required to have paint
mixing equipment, cleaning solvents, foil handling facilities,
miscellaneous hose and fittings, spray equipment assembly tools,
and a copying machine available during the testing period.
Utilities, including compressed air and electricity, were also
required.
EPA and its representatives, and cooperating equipment
manufacturers and suppliers, supplied other test equipment,
including but not limited to spray painting equipment and
auxiliaries, test method, data books, and technical guidance.
13
-------
The laboratories agreed that none of the test equipment supplied
would be used for any purpose other than was directly related to
the EPA transfer efficiency research program. This agreement
specifically forbade using, tampering, dismantling, or otherwise
examining this equipment except as required for the test program
or for proper maintenance. An appropriate amount of security was
necessary to ensure that this requirement was enforced.
Each laboratory was required to allow EPA representatives and EPA
contract engineers full and complete access to the test area.
The laboratories agreed to allow Spray Painting Transfer
Efficiency Project Steering Committee members into their facility
as observers of the research program.
Suitably equipped industrial laboratories were recommended
through the Spray Painting Transfer Efficiency Steering Com-
mittee, through outside listings of available laboratories
(Thomas' Register of Manufacturers, state-by-state industrial
directories. Products Finishing 1984 Directory, and Metal
Finishing Guidebook and Directory and Metal Finishing Guidebook
and Directory for 1984.)
Participating laboratories were asked to schedule two weeks each
for conducting tests except in the case of the first test, which
required four weeks of laboratory time. Several additional
experiments were conducted during the first test.
TEST EQUIPMENT
Overview of Test Equipment
In this part of the program, three types of spray equipment were
tested at up to eight laboratories. One spray equipment type,
electrostatic air spray, was used to simulate conventional air
spray equipment as well,, Thusf two complete spray systems were
obtained for electrostatic air spray and for airless. Eight
spray guns of the same make and model were obtained for use on
the two electrostatic air spray and two airless systems. These
equipment types are described in detail below. Spray systems and
guns were supplied to participating laboratories in new condition
to ensure as nearly identical spray equipment conditions as
possible. Each component of the spray system, from the gun tip
to hoses and supply tanks, was required to be of the same size,
make, and model. This requirement enhanced control of variances
from spray system to spray system. Each spray gun was mounted in
the proper spray position on a pole. Spray gun operation was
manual (hand-triggered), except at one laboratory where a remote
actuator was employed.
Other test supplies and equipment described in this section
include instrumentation, paint, targets, and foil. The supply of
some test equipment was required of participating laboratories.
These were described earlier in this section.
14
-------
Electrostatic/Conventional Air Spray Equipment
A manual mid-range (75 kV) electrostatic air spray gun was
selected for use as electrostatic spray equipment and to
represent conventional air spray equipment. (As previously
described, the same equipment was used but without voltage
applied.) The electrostatic air spray gun was equipped with a
1.2 mm (0.047 in) fluid tip. This was the same model spray gun
used during a previous test effort conducted to determine how
operating and maintenance variables affect transfer efficiency
(U.S. EPA, 1985). It was considered representative of manual
electrostatic air spray guns available from other manufacturers,
although differences exist from manufacturer to manufacturer.
A 75 kV adjustable power supply was used to support the
electrostatic air spray gun. This is a mid-range power supply
equipped with appropriate cable and connections. A kV meter and
ammeter were built into the power supply control panel. Power
cables were supplied in 762 cm (25 ft) lengths for each power
supply.
Pressure tanks were used to supply paint to the gun through 1524
cm (50 ft) fluid lines. A dual air regulator pressure tank with
agitator (to minimize stratification of the paint) was supplied
with each electrostatic air spray system. Air hoses were
supplied, in 1524 cm (50 ft) lengths for each system.
Airless Spray Equipment
A manual airless spray gun was selected for the interlaboratory
test program. The spray gun included a 0.38 mm (0.015 in)
diameter tip.
An electric pump was the fluid supply mechanism. The small, 1.89
liter per minute (0.5 gpm) pump was used. Fluid hose was
obtained in 762 cm (25 ft) lengths for airless transfer
efficiency determinations. Two fluid hoses were used in each
test, with one segment spanning from the pump to the control
panel (see below) and one segment spanning from the control panel
to the spray gun.
Instrumentation
Control Panel
Two control panels were constructed to provide adequate process
control equipment to participating spray painting laboratories.
Each control panel housed air and fluid regulators, an air
rotameter, a mass flow meter, and a mass flow totalizer on a
movable cart. The regulators had a pressure range of 6.9 to
344.7 kPa (1 to 50 psig). The recommended supply pressure for
the regulators was 827.3 kPa (120 psig), with a maximum supply
pressure of 1034.2 kPa (150 psig). All connections were 1/4 NPT.
15
-------
Five pressure gages were mounted in each control panel. These
pressure gages monitored air and fluid pressure throughout the
spray system. All gages were dead-weight calibrated prior to
testing. Dead-weight calibration is a standard method for
calibrating pressure gages by applying a static ("dead")
pressure. In this case, a series of static pressures (framing
the expected test pressures) were applied to each gage being
calibrated. Calibration curves were developed and used for all
gages used in this research.
A direct scale reading rotameter was included on each control
panel to indicate air flow through the system.
A mass flow meter was utilized to indicate the total mass of
paint sprayed during each test. Each control panel was equipped
with a mass flow meter. This meter was fitted with a digital
totalizer, enabling test personnel to read paint mass flow
quickly and easily.
Other Instrumentation
An anemometer was available to laboratories not having their own
equipment to document the spray booth air velocity -
A sling psychrometer was also available to laboratories not
having their own relative humidity or thermal measurement
equipment.
A calibrated #4 Ford viscosity cup was supplied for use at each
laboratory.
Paint
Certain desirable paint characteristics were necessary for the
test program. The same paint was to be sprayed by all spray
systems, except that the viscosity of the paint was higher for
airless spray equipment. Thus, the paint had to be of relatively
high viscosity when uncut but mid-range viscosity when cut. It
had to have a simple solvent system, for ease in obtaining and
simplicity in cutting (thinning) at test sites. The paint had to
be readily available from a large manufacturer's batch to
encourage homogeneous characteristics. It had to adhere well to
aluminum foil without cracking, peeling, or breaking. It was
required to have at least a 6-month shelf life so that a single
batch could be used in all experiments in this program. Paint
resistivity must be high for electrostatic spraying, or the paint
may ground the system,, Finally, the paint had to be reasonably
priced and readily available.
The selected paint met these criteria as summarized in Table
5-2. It was a highly flexible alkyd base paint commonly used for
painting light fixtures. As shown in Table 5-2, the paint was
cut to a wide range of viscosities using xylol. Nitration grade
xylol was used as the solvent.
16
-------
TABLE 5-2. PAINT SPECIFICATIONS
Supplier: Glidden 447-W-02133
Resistivity: 360 Megohms/cm
Nonmetallic
Alkyd based
Xylol solvent (titration grade)
Part of a high volume batch
Adheres well to aluminum foil
Cure: 900 s (15 min) at 190.5°C (375°F)
Approximately 52 percent solids by weight, after cutting
A sample of the test paint was sent to the first laboratory for
field confirmation of its properties. It met all test
requirements.
During the tests, paint weight percent solids were determined
several times daily as required by the test method and by the
project QA/QC plan (Appendix A and Appendix B, respectively).
ASTM Method D-2369, Standard Test Method for Volatile Content of
Coatings, was used for all paint weight percent solids determi-
nations, except that the vendor-recommended cure schedule was
followed.
Foil
A medium-temper aluminum alloy foil was required by the draft
transfer efficiency test method (Appendix A). As described in
the test method, 0.0037 cm (1.5 mil) thick foil, 38.1 cm (15 in)
wide and on rolls approximately 21336 cm (700 ft) long, was
supplied to participating laboratories by the contractor. Over
227 kg (500 Ibs) of foil was supplied to test laboratories, or
about 28.4 kg (62 Ibs) per laboratory.
Targets
As directed by the draft transfer efficiency test method, one EPA
target consisted of a set of ten galvanized steel 121.9 cm (48
in) by 15.2 cm (6 in) panels mounted on 30.48 cm (12 in) centers
(see Appendix A). A set of ten panels thus configured and hung
on an adjustable speed conveyer was used for each transfer
efficiency determination. The first two and last two panels in
17
-------
each set of ten were scavengers. Selection of these targets over
other configurations is discussed later in this section.
Twelve sets of targets (consisting of ten panels each) were
supplied to each participating laboratory to help expedite
transfer efficiency testing. By having extra targets on hand,
one member of the research team could be weighing and wrapping
panels in preparation of testing while other researchers were
setting up spray equipment. During testing, as each spray
painting pass was completed and the targets removed for curing,
another set of prepared targets was ready for mounting and
spraying.
Targets were suspended so that the spray pattern fell across the
middle of the target, with a minimum 30.5 cm (1 ft) clear space
between the edge of the spray pattern and the target top and
bottom. This requirement was to prevent excessive overspray.
Previous transfer efficiency test results using target
configurations ranging from large flat panels (almost prohibiting
overspray) to small vertical cylinders hung on wide centers
(almost totally overspray) demonstrated that either extreme
design tended to desensitize the transfer efficiency test method.
That is, target configuration affected the ability of the test
method to detect changes in transfer efficiency. The target
configuration selected for this research was determined to be the
most sensitive target type tested. Therefore, the selected
target configuration was most likely to detect changes in
transfer efficiency within or between laboratories.
18
-------
SECTION 6
PERFORMANCE OF EXPERIMENTS
FIRST LABORATORY
Special tests were conducted at the first laboratory in this
program. First, base spray conditions were established for each
type of equipment. Once established, these conditions were used
for all subsequent transfer efficiency determinations in this
program. Second, the gun portion of the variance was established
through a prescribed set of experiments. Finally, the first
interlaboratory test was conducted.
The draft transfer efficiency test method (Appendix A) was used
for all transfer efficiency determinations made in this program.
The test method as presented in Appendix A has been modified to
include base spraying conditions from the first laboratory as
part of the test criteria. These criteria were a necessary part
of the interlaboratory test program, and are not part of the
generic draft transfer efficiency test method.
Establish Base Spray Conditions
Prior to conducting transfer efficiency determinations at the
first laboratory, standardized spray conditions were
established. (These same conditions were used at all subsequent
laboratories.) Spray conditions were established through a
trial-and-error procedure to determine an acceptable spray
pattern and finish. Extensive previous transfer efficiency
testing using the same types of spray equipment and similar paint
was used in setting the first rough approximations of spray
conditions and in adjusting conditions. Disposable paper targets
were used for the rough first approximations of spray
conditions. Foil-covered EPA targets were painted at tentative
spray conditions to demonstrate their appropriateness for use
with the actual test targets.
Of the spray conditions, paint viscosity was one of the hardest
to set and control. Paint viscosity was to be set at mid-range
(15 s on a #4 Ford cup) for electrostatic air spray and
conventional air spray transfer efficiency determinations. Paint
viscosity was set at 30 s on a #4 Ford cup for airless spraying.
Nitration grade xylol was used to cut the paint to specified
conditions. ASTM D-1200-70 was the method used to determine
paint viscosity.
Test booth conditions were documented prior to setting spray
conditions and prior to conducting transfer efficiency
determinations. If booth conditions did not meet test require-
ments, or if process control systems (booth rate, regulators,
19
-------
mass flow meter, etc.) were not operating, they were repaired
prior to setting spray conditions.
Actual spray system operating conditions were set for airless
equipment first. Gun-to-target distance, fluid pressure, and
paint viscosity adjustments were set at manufacturer's recom-
mendations. Test patterns were shot onto paper and operating
conditions were adjusted until a good spray pattern was
realized. Once a good spray pattern was realized, foil-covered
EPA targets were painted and cured. When the resulting coating
was acceptable, spray operating conditions were documented.
These spray conditions were fixed for the remainder of airless
equipment testing in this program.
Airless testing for gun portion of the variance and for the first
interlaboratory test was conducted at base operating conditions.
Electrostatic air spray conditions were set next. Since
electrostatic air spray and conventional air spray equipment were
physically identical in this program, no equipment changes were
required. The same procedure for establishing spray conditions
was followeds setting paint characteristics* then making
trial-and-error adjustments to operating conditions (including
voltage, shaping air, and atomizing air) until an acceptable
spray pattern was realized. Once an acceptable spray pattern was
achieved, foil-covered test panels were painted and cured to
assure an acceptable pattern and finish under the tentative spray
conditions. This process was repeated until an acceptable
pattern and finish were realized on test panels. Then spray
conditions for electrostatic air spray and conventional air spray
equipment were documented, fixing them at the same level for all
subsequent transfer efficiency determinations in this program.
Establish Gun Portion of the Variance
The following procedure for establishing the gun portion of the
interlaboratory variance was performed for each equipment type at
the first laboratory. The serial number of the spray gun was
recorded for each transfer efficiency determination made. Once
standard spray conditions were determined as above, the order of
testing was randomized to provide six replicate transfer
efficiency determinations—one for each gun and each type of
spray system.
The data analysis for determining the gun portion of the inter-
laboratory variance is described in Section 7 - Test Results and
Statistical Analysis.
First Interlaboratory Test
As described above, much of the work conducted at the first
laboratory involved start-up operations including: setting up
the three different spray systems, paint cutting and
documentation, establishing standard operating conditions for
20
-------
each spray system, and conducting tests to determine the gun
portion of the variance for each spray system. The first
interlaboratory test consisted solely of conducting six replicate
transfer efficiency tests on each of the three spray systems. No
operating or maintenance variables were altered from test to
test.
SUBSEQUENT INTERLABORATORY TESTS
Transfer efficiency tests at subsequent laboratories each
consisted of three discrete experiments corresponding to three
equipment types—electrostatic air spray, conventional air spray,
and airless. Six replicate transfer efficiency determinations
were conducted for each type of equipment. Subsequent test
series in the interlaboratory program took 5-8 days in each
laboratory. The summary of each of these test series, the
problems encountered, and the test results follows.
21
-------
TRANSFER EFFICIENCY TEST AT
LABORATORY NO. 1 - June 3-28, 1985
The transfer efficiency tests were performed at the first
laboratory to define and establish spray conditions to be
utilized as a standard for subsequent testing, to calculate the
gun portion of the variance, and to help determine the number of
subsequent sites required to statistically prove the precision
requirements of the draft transfer efficiency test method.
At the first laboratory, six replicate transfer efficiency
determinations were made for each spray gun (eight each of
conventional airless, electrostatic air spray, and conventional
air spray). The order of the transfer efficiency determinations
was determined using random number tables.
Test Set-Up
Several items were established prior to the testing including
paint viscosities, the cure schedule, and the numerous operating
pressures. In addition, scales, gages, and meters were
calibrated to within the EPA-approved standards (see Appendix B).
Two Ford viscosity cups were calibrated against standardizing
oil.
The paint cure schedule was chosen after several weight percent
solid tests were performed. The cure schedule was set at 330°F
for 11 minutes. It was apparent, however, that this time and
temperature did not achieve complete curing., At the suggestion
of the manufacturer, conditions were adjusted to 375 9F for 15
minutes. Although the painted surface was apparently cured by
the new schedule, paint weight continued to decline with
additional curing. According to the manufacturer, this
phenomenon was caused by the breakdown of a chemical cross-
linking mechanism in the coating. This phenomenon was an
additional concern in trying to adequately control curing of test
targets.
The paint manufacturer also recommended the viscosities for
airless and electrostatic air spraying using this type of paint.
With the higher weight percent solids, the use of 15 seconds
(No. 4 Ford cup) for electrostatic air spraying and conventional
air spraying, and 30 seconds for airless spraying was advised.
The pressure gages to be used for the program were high-precision
gages. They were dead-weight tested prior to TE testing.
Once the pressure gages were tested, the mass flow meter was
calibrated. Several major problems existed. The airless system
caused high vibrations due to the airless pump pulsations. The
control panel vibrated, causing an even greater problem with the
meter stabilization.
22
-------
After reviewing the mass flow meter installation instructions, it
was decided to place steel plates on the control panel to give
the meter a more rigid base. After several trial runs, it was
determined that these steel plates did dampen the vibration.
The mass flow meter was tested against the paint capture method
at various flow rates. It met the accuracy requirement of +0.9
percent defined in the QA/QC Plan (see Appendix B.)
Airless Testing
The airless test actually began one week after the project team's
arrival on site. Using random number tables, the order in which
the airless tests would be performed was derived. The order and
the resultant foil numbers and transfer efficiencies are included
as Table 6-1.
The airless gun test was completed in two days. The results are
included in Table 6-1. The average transfer efficiency was 44.4,
with a standard deviation of 2.50 across all gun types. These
numbers do not have any particular value of their own; the values
take on significance only when examined in conjunction with the
results of the other laboratories.
Electrostatic/Conventional Air Spray Testing
This portion of the test was set up and run according to the
operating parameters set up in Appendix A and recorded in the raw
data sheets in Appendix E. No major problems were encountered
during the testing; however, it appeared that the mass flow meter
was becoming increasingly less accurate. This problem forced on-
site engineers to request a change in the QA/QC requirements to
+_ 10 grams, consistent with other acceptable mass flow
measurement methods in Appendix B.
These transfer efficiency determinations were made during the
last three weeks of June 1985.
Test Results
The objectives of the first test were to establish operating
conditions for the eight-laboratory study, to determine the gun
portion of the variance, and to perform the first part of an
eight-laboratory study. These objectives were achieved.
Base Operating Conditions
As described in the preceding section, base operating conditions
for all three equipment types were established during the first
week of testing. They are presented in Appendix E.
23
-------
TABLE 6-1. ORDER OF AIRLESS GUN TESTING
GUN
61632
60838
61449
61449
60838
61632
60829
60429
61632
61629
60829
61629
61649
60865
61449
61632
60865
61449
61632
61449
61629
61629
61644
60838
61644
60865
61449
61629
61649
60865
__
61649
61644
61649
61644
61644
61644
60865
61632
60865
60838
60838
60829
60829
60838
60829
61629
FOIL NOS.
31-36
37-42
43-48
49-54
55-60
61-66
67-72
79-84
85-90
91-96
97-102
103-108
109-114
115-120
121-126
127-132
133-138
139-144
145-150
151-156
163-168
169-174
176-181
182-187
188-193
194-199
200-205
206-210
212-217
218-223
230-235
236-241
242-247
248-253
254-259
260-265
266-267
272-277
278-283
284-289
290-295
296-301
302-307
314-319
320-325
326-331
332-337
TRANSFER EFFICIENCY
42o2
45.7
47»1
46,1
44.5
40.6
42.4
43el
43.5
44.9
44.2
43.8
43.3
44.8
45.9
44.1
45.1
47.1
44.8
45.3
45.2
44.2
46.1
45.1
42.2
44.7
45e6
44.2
43.4
42.6
OBOD
43.5
43.4
42.8
43.3
42.8
43.9
44.3
43.2
45 . 5
45,0
45,0
42.4
43.0
45ol
43.2
44.0
24
-------
Gun Portion of the Variance
The gun portion of the variance was statistically detectable and
significant. Summaries of the order of gun testing, foil
numbers, and the transfer efficiency results are presented in
Tables 6-1, and 6-3. The variance (not standard deviation)
attributable to differences between guns used during this
research for the three equipment types was:
Airless 6.22
Conventional air spray ... 0.88
Electrostatic air spray ... 13.01
The variances for airless and conventional air spray guns were
considered unexpectedly low. In the final analysis, they
contributed little to the total (overall) interlaboratory
variance. Electrostatic air spray variance was higher, but
still accounted only for 15 percent of the total interlaboratory
variance for electrostatic air spray transfer efficiency
determinations.
Thus, it was concluded that the differences between guns must
be accounted for even though they were not a major contributor
to the total interlaboratory variance.
The determination of the gun portion of the variance is discussed
in more detail in Section 7 - Statistical Analysis.
First Laboratory Test Results
After each test was completed, an outlier analysis was performed
on the data (see Appendix C.)
Airless test results from the first laboratory had a mean
transfer efficiency value of 44.2, with a standard deviation of
0.821. The data were tightly grouped as shown in Table 6-4.
These numbers do not have any extraordinary value of their own;
the values take on significance only when examined in conjunction
with the results of the other laboratories. They do, however,
demonstrate the capability of the draft transfer efficiency test
method to produce highly repeatable data at a given facility.
The electrostatic data showed a significantly greater degree of
dispersion (see Table 6-2). The reason for this trend could
not be technically pinpointed nor statistically modeled; however,
it was apparent that there was a constant factor influencing the
testing, especially the electrostatic air spray gun tests. Test
results are presented in Section 7. There was speculation onsite
that the electrical supply changed during the day as heating demands,
office lighting demands, and other electrical demands were made on
the power supply. This speculation could not be documented.
25
-------
TABLE 6-2. ORDER OF ELECTROSTATIC AIR SPRAY GUN TESTING
GUN
C1378
C1320
C1356
C1336
C1356
C1290
C1320
C1382
C1290
C1290
C1290
C1336
C1290
C1378
C1356
C1355
C1365
C1378
C1365
C1290
C1290
C1356
C1356
C1320
C1355
C1320
C1356
C1382
C1382
C1336
C1336
C1382
C1336
C1365
C1365
C1382
C1355
C1382
C1355
C1336
C1365
C1365
C1368
C1378
C1378
C1320
C1355
C1355
C1320
C1320
FOIL NOS.
573-578
585-590
597-602
609-614
621-626
633-638
561-566
717-722
729-734
741-746
747-752
759-764
771-776
783-788
417-422
429-434
441-446
453-458
465-470
477-482
489-494
501-506
513-518
525-530
537-542
549-554
1016-1021
944-949
956-961
968-973
980-985
992-997
1004-1009
886-891
863-868
875-879 (+901)
908-913
920-925
932-937
795-800
807-812
819-824
832-837
844-849
857-862
356-361
369-374
381-386
393-398
405-410
TRANSFER EFFICIENCY
65.27
71.04
75.50
73.70
73.64
71*65
79.63
70.56
64.39
62.46
62.86
72.20
64«68
70.32
74.60
77*83
79.24
73 = 75
79.04
77.06
77.37
75.10
76*99
76.45
70.58
74.54
77«98
73,75
72.30
77.35
74*74
74,28
77,73
74,41
76.48
76.15
66.50
74.90
66.98
70.03
73,89
73,71
70.76
71.68
72,11
77.61
69.71
74.10
80.30
78.83
26
-------
TABLE 6-3. ORDER OF CONVENTIONAL AIR SPRAY GUN TESTING
GUN
C1378
C1320
C1356
C1336
C1356
C1290
C1320
C1355
C1355
C1320
C1320
C1356
C1355
C1365
C1378
C1365
C1290
C1290
C1356
C1356
C1320
C1355
C1320
C1320
C1382
C1290
C1290
C1336
C1290
C1378
C1336
C1365
C1365
C1378
C1378
C1378
C1365
C1365
C1382
C1355
C1382
C1355
C1382
C1382
C1336
C1336
C1382
C1336
C1356
FOIL NOS.
567-572
579-584
591-596
603-608
615-620
627-632
349-355
363-368
375-380
387-392
399-404
411-416
423-428
435-440
447-452
459-464
471-476
483-488
495-500
507-512
519-524
531-536
543-548
555-560
705-710
723-728
735-740
753-758
765-770
777-782
789-794
801-806
813-818
825-830
838-843
851-856
880-885
892-897
869-874
902-907
914-919
926-931
938-943
950-955
962-967
974-979
986-991
998-1003
1010-1015
TRANSFER EFFICIENCY
32.72
39.09
29.80
37.29
34.44
41.26
36.67
36.44
37.80
39.30
40.40
34.29
35.56
34.25
36.86
34.65
39.92
39.98
31.46
33.20
38.19
33.02
38.20 -
38.18
33.43
37.28
39.09
35.91
38.79
35.41
36.63
30.41
32.31
34.86
33.96
33.30
30.50
31.75
34.99
34.87
31.22
34.62
35.18
34.96
35.06
34.85
34.79
35.08
29.23
27
-------
TABLE 6-4. LABORATORY 1 TEST RESULTS
Airless
Transfer efficiency (%): 44,9
43,8
45,2
44,2
42c9
44.0
Mean: 44.2
Standard deviation: 0.821
Electrostatic air spray
Transfer efficiency (%): 65.3
70.3
73.8
70.8
71.7
72.1
Means 70„7
Standard deviation: 2..S9
Conventional air spray
Transfer efficiency (%) 32.7
36.9
35.4
34o9
34.0
33.3
Mean: 34.5
Standard deviation: 1.53
28
-------
Electrostatic air spray equipment produced a mean transfer
efficiency of 70.7 with a standard deviation of 2.89. These
numbers do not have any extraordinary value of their own; the
values take on significance only when examined in conjunction
with the results of the other laboratories. They do, however,
demonstrate the capability of the draft transfer efficiency test
method to produce repeatable data at a given facility.
Conventional air spray equipment produced a mean transfer
efficiency of 34.5, with a standard deviation of 1.53. Again,
these numbers do not have any extraordinary value of their own;
the values take on significance only when examined in conjunction
with the results of the other laboratories. They do, however,
demonstrate the capability of the draft transfer efficiency test
method to produce highly repeatable data at a given facility.
29
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 2
Test Facilities
The second laboratory transfer efficiency tests were conducted in
a Sinks water-wash spray booth in the spray painting laboratory.
The laboratory provided the spray painting laboratory, techni-
cians, conveyor system, curing oven, and other associated test
materials.
The engineering laboratory was roughly 1220 cm by 915 cm with 600
cm ceilings (about 40 ft by 30 ft, with 20 ft ceilings). The
booth maintained the linear air velocity at 40.6-61.0 cm/s
(80-120 ft/min), as required by the proposed laboratory selection
criteria. The booth area temperature was controlled at 22.2°C
(72°F) during test runs. Paint and solvent were kept in the
booth area and their temperatures closely matched booth ambient
temperatures. Relative humidity was controlled at 50 percent
during the test.
A Mayfran Cableway overhead conveyor system equipped with a
Lab-line digital timer was used in all experiments at this site.
A Despatch oven was used for curing painted targets.
The laboratory was supplied with the spray gun, 5-gallon pressure
tank with agitator, hoses, and 75 kV power supply. The same gun
also was used to simulate conventional air spray painting by
conducting transfer efficiency determinations with no voltage
applied.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
September 24, 1985. Transfer efficiency testing began on
September 24, 1985. Six replicate transfer efficiency
determinations were made for each equipment type. Airless tests
were completed the first day of testing. Conventional air spray
tests and electrostatic air spray tests were completed the
following day. Operating conditions are presented in the raw
data sheets in Appendix E.
The purpose standard transfer efficiency test method and spray
conditions, as set forth in Transfer Efficiency Method Evalu-
ation Plan, were followed for all transfer efficiency determi-
nations. All operating conditions are presented in the raw
data sheets in Appendix E. All QA/QC requirements were met,
and no special problems were encountered during the experiment.
30
-------
Test Results
Transfer efficiency results for these tests are presented in
Table 6-5. Mean and standard deviations are also summarized in
Table 6-5.
Airless test results from the second laboratory had a mean
transfer efficiency value of 33.1, with a standard deviation of
1.73. The data were tightly grouped as shown in Table 6-5.
These numbers do not have any extraordinary value of their own;
the values take on significance only when examined in conjunction
with the results of the other laboratories. They do, however,
demonstrate the capability of the draft transfer efficiency test
method to produce highly repeatable data at a given facility.
Electrostatic air spray equipment produced a mean transfer
efficiency of 66.2 with a standard deviation of 0.95. These
numbers do not have any extraordinary value of their own; the
values take on significance only when examined in conjunction
with the results of the other laboratories. They do, however,
demonstrate the capability of the draft transfer efficiency test
method to produce repeatable data at a given facility.
Conventional air spray equipment produced a mean transfer
efficiency of 24.4, with a standard deviation of 0.57. Again,
these numbers do not have any extraordinary value of their own;
the values take on significance only when examined in conjunction
with the. results of the other laboratories. They do, however,
demonstrate the capability of the draft transfer efficiency test
method to produce highly repeatable data at a given facility.
A standard outlier test, using Nalimov's criteria, was conducted
on transfer efficiency results for each spray system. One
outlier was identified (Run No. 10) in the electrostatic air
spray results. No reason for the outlier could be identified.
The run was repeated as per the requirements of the QA/QC Plan.
According to the QA/QC Plan, outlier transfer efficiency results
were replaced by the results of a replacement transfer efficiency
determination. Thus, outliers were not included in the final
data set from each laboratory and had no effect on the results of
the study.
Sequential Analysis Estimating Number of Laboratories
Results of transfer efficiency determinations from the first and
second laboratory were used to estimate the number of labora-
tories required to estimate the method's precision satisfac-
torily, i.e., to ensure that the probability of the estimated
variance differing from the actual variance by less than a
specified amount is high. In terms of a mathematical
relationship, this means:
~ 2 2
Pr[ a - a < <5 ]>P (1)
31
-------
TABLE 6-5. LABORATORY 2 TEST RESULTS
Airless
Transfer efficiency (%): 35.3
34.4
33.0
31.3
33.6
30.8
Mean: 33.1
Standard deviation: 1.73
Electrostatic air spray
Transfer efficiency (%): 65.4
65.8
65.2
67.8
66.6
66.2
Mean: 66.2
Standard deviation: 0.95
Conventional air spray
Transfer efficiency (%) 24.9
25.2
24.0
24.0
23.8
24.2
Mean: 24.4
Standard deviation: 0.57
32
-------
where o2 is the square of the method precision, 52 j_s the
estimated variance from the interlaboratory test program, and 6
is the degrees of freedom. For estimation of the number of
laboratories based on the results of the first two tests, a2 is
the total variance between the first two laboratories. The
probability (P) was specified as 80 percent by the contract
sponsoring this research. Fiscal limitations allowed up to eight
laboratories to be tested; if the estimated number of
laboratories (based on the first two laboratories' results)
exceeded eight to obtain 80 percent confidence, then only eight
laboratories could be tested.
The following analysis of variance was performed for each of the
three equipment types tested at the first two laboratories:
Analysis of Variance
First Two Laboratories
Source SS DF MS
Between laboratories SSa 1 MSa
Between guns, labora-
tory 1 SSfc 7 MSb
Reproducibility within
guns, laboratory 1 SSC 32 MSC
Reproducibility, labora-
tory 2 SSa _2 MSd
49
The sum of squares (SS) divided by the number of degrees of
freedom (DF) results in a value for the mean square (MS). The
following equations have been derived, based on expected values
of the mean squares, to determine the variance components from
the test results:
§w2 = 32 SS + 9 SS^ (2)
c~41 d
33
-------
The desired estimate of test precision is then
= 2 * 2
o =
where the ~ notation indicates the estimate is based on results
in the first two laboratories.
The final estimates of the within- and between-laboratory
variance components, based on testing at all b laboratories, were
given by
°w2 = MSw with b(n-D degrees of freedom
5b2 = MSb ~ MSw / n with b~1 Degrees of freedom
(5)
Thus the final estimate of the test precision was
V
0 -.JMSW A - 1\ + MSb
I n n
Confidence limits may be placed on this estimate of the test
precision based on the fact that the ratio of the variance
estimate to the population variance, when multiplied by the
number of degrees of freedom, is distributed as chi-square. This
allows the variance of the estimate to be written as;
Var (o2) = 2 °w4 + 2(°w2 + n V)2 (7)
b (n-1) n (b-1)
The initial estimate of Var (a2) was based on the results of
testing in the first two laboratories, equations (2) and (3):
o = a • o =o
w w' b ub (8)
Assuming that ttje final estimate of the precision is normally distributed
with the mean o and the variance given by equation (7), then equation (1)
can be used to determine the number of laboratories (b) required. This
was done to determine the approximate number of laboratories that would be
required to varify the efficiency of the metho to .a defined level. How-
ever, the actual number of laboratory test that could be conducted was govern-
ed by fiscal constraints. Table 6-6 tabulates the number of laboratories
needed to validate the method for n=6, crw=l, cfb=2.
34
-------
Table 6-6. Number of laboratories (b) required to be P% sure that the
estimated total variance is within K% of the true value.
P%
20
30
Percent relative error (K%)
40 50 60 70 80
Airless conventional spray systems (ALC)
90
80
70
60
50
>30
31
21
14
10
23
15
10
7
5
14
9
6
5
3
9
6
5
3
3
7
5
4
3
2
5
4
3
2
2
Electrostatic spray systems (ALE)
90
80
70
60
50
>30
33
22
>30
>30
23
15
10
32
20
13
9
7
21
13
9
7
5
15
10
7
5
4
11
7
5
4
3
4
3
3
2
2
9
6
4
3
3
90
Air atomized conventional spray systems (AAC)
90
80
70
60
50
>30
>30
23
>30
>30
24
16
11
33
21
14
10
7
22
14
9
7
5
15
10
7
5
4
12
8
6
4
3
9
6
5
4
3
4
3
2
2
2
7
5
4
3
2
8
5
4
3
3
100
3
3
2
2
2
6
4
3
3
2
7
5
3
3
2
The results presented in Table 6-6 indicate that the ALE and AAC require
approximately twice the number of laboratories as the ALC to achieve the
level of confidence criteria. That criteria had an 80 percent probability
of being within 2.5 transfer efficiency units of the mean. As previously
indicated, however, the actual number of laboratory tests that could be
conducted was governed by fiscal constraints. Thus, only the ALC program
approached that criteria. For b=8, i.e., the number of tests conducted,
the estimates indicate only approximately 65 percent chance of being
within 50 percent of the true values. This lack of laboratory experience,
naturally, reduced the scope of conclusions that can presently be made about
the defined method.
35
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 3
Test Facilities
Tests were conducted in a Protect Aire back draw spray booth in
the spray painting laboratory. The laboratory provided the spray
painting laboratory, technicians, conveyor system, curing oven,
and other associated test materials.
The spray painting laboratory was roughly 1220 cm by 915 cm with
600 cm ceilings (about 40 ft by 30 ft, with 20 ft ceilings). The
booth was a back draw, water-wash type, which maintained the
linear air velocity at 41-61 cm/s (80-120 ft/min), as required by
the proposed standard transfer efficiency test method. The booth
area temperature was controlled at 22.1-23.3°C (71-74°F) during
test runs. Paint and solvent were kept in the booth area; paint
and solvent temperatures closely matched booth ambient
temperatures. Relative humidity ranged from 44 to 56 percent
during the test.
A Rapistan overhead conveyor system equipped with a Century
E-Plus speed control was used in all experiments at this site. A
Michigan oven was used for curing painted targets.
The laboratory was supplied with a manual airless spray gun,
5-gallon pressure tank'with agitator, hoses, and 75 kV power
supply. The electrostatic air spray gun also was used to
simulate conventional air spray painting by conducting transfer
efficiency determinations with no voltage applied.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
September 30, 1985. Transfer efficiency testing began on October
1, 1985. Six replicate transfer efficiency determinations were
made for each equipment type. Operating conditions are detailed
in Appendix E.
The proposed standard transfer efficiency test method and spray
conditions, as set forth in Transfer Efficiency Method Evalu-
ation Plan, were followed for all transfer efficiency determi-
nations. All QA/QC requirements were met, and no special
problems were encountered during the experiment.
Test Results
Transfer efficiency results for these tests are presented in
Table 6-7. These transfer efficiency values are only applicable
to the particular system (equipment and paint) under test at
specified operating conditions. Mean and standard deviations
are also summarized in Table 6-7.
36
-------
TABLE 6-7. LABORATORY 3 TEST RESULTS
Airless
Transfer efficiency (%): 49.4
47.5
44.9
48.5
46.4
50.5
Mean: 47.9
Standard deviation: 2.06
Electrostatic air spray
Transfer efficiency (%): 68.7
71.2
70.2
70.8
72.6
71.6
Mean: 70.9
Standard deviation: 1.33
Conventional air spray
Transfer efficiency (%) 39.6
39.4
39.0
39.2
39.1
38.6
Mean: 39.2
Standard deviation: 0.35
37
-------
A standard outlier test, using Nalimov's criteria, was conducted
on transfer efficiency results for each spray system. One
outlier was detected (Run No. 6) during the airless test. No
reason for the outlier was determined. The run was repeated as
required by the QA/QC Plan; the result was substituted for the
original outlier as prescribed by the QA/QC Plan. Thus, the
outlier had no effect on the results of the study.
38
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 4
Test Facilities
Tests were conducted in a spray painting area roughly 427 cm by
305 cm, with 305 cm ceilings (about 14 ft by 10 ft, with 10 ft
ceilings). The booth was a back-draw dry filter type, which
maintained the linear air velocity at 61 cm/s (120 fpm), as
required by the proposed standard transfer efficiency test
method. The booth area temperature was 22.2-26.6°C (72-80°F)
during test runs. Paint and solvent were kept in the booth area;
paint and solvent temperatures closely matched booth ambient
temperatures. Relative humidity was 51-58 percent during the
test.
An Econo overhead conveyor system was used in all experiments at
this site. A Despatch oven was used for curing painted targets.
The laboratory was supplied with a manual Wagner G-10 airless
spray gun, associated hoses, high-pressure paint pump, and 015
spray tip. For electrostatic air spray transfer efficiency
determinations, the laboratory was supplied with a manual spray
gun, 5-gallon pressure tank with agitator, hoses, and 75 kV power
supply- The electrostatic air spray gun was also used to
simulate conventional air spray painting by conducting transfer
efficiency determinations with no voltage applied.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
October 23, 1985. Transfer efficiency testing began on October
24, 1985. Six replicate transfer efficiency determinations were
made for each equipment type. Airless tests were completed the
first day of testing. Electrostatic air spray and conventional
air spray tests were completed the following day. Operating
conditions are presented in Appendix E - Raw data.
The proposed standard transfer efficiency test method and spray
conditions, as set forth in Transfer Efficiency Method Evalu-
ation Plan, were followed for all tansfer efficiency determi-
nations. All QA/QC requirements were met, and no special
problems were encountered during the experiment.
Test Results
Transfer efficiency results for these tests are presented in
Table 6-8. These transfer efficiency values are only applicable
to the particular system (equipment and paint) under test at
specified operating conditions. Mean and standard deviations are
also summarized in Table 6-8.
39
-------
TABLE 6-8. LABORATORY 4 TEST RESULTS
Airless
Transfer efficiency (%): 39.9
42,1
38.5
36.2
35.5
36.5
Mean: 38.1
Standard deviation: 2.54
Electrostatic air spray
Transfer efficiency (%)s 70.6
67.1
68.9
68.7
69.7
69.4
Mean: 69*1
Standard deviation: 1.17
Conventional air spray
Transfer efficiency (%): 34.9
35.7
35.3
35o6
34.2
36.3
Mean: 35.3
Standard deviation: 0.723
40
-------
A standard outlier test, using Nalimov's criteria, was conducted
on transfer efficiency results for each spray system. One
outlier was identified in these tests. The booth exhaust fan was
not running during painting of Run No. 11. The results of this
flawed test run were replaced as per the QA/QC Plan, and thus had
no effect on the final analysis of the results.
41
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 5
Test Facilities
Tests were conducted in a Sinks spray booth. The laboratory
provided the spray painting area, curing oven, and assistance
required to complete the transfer efficiency tests.
The spray painting area was roughly 549 cm by 1219 cm with 427 cm
ceilings (about 18 ft by 40 ft, with 14 ft ceilings). The booth
was a dry back filter type, which maintained the linear air
velocity at 56 cm/s (110 ft/min), as required by the proposed
standard transfer efficiency test method. The booth area
temperature was 23-24°C (73-75°F) during test runs. Paint and
solvent were kept in the booth area; paint and solvent tempera-
tures closely matched booth ambient temperatures. Relative
humidity was 65-77 percent during the test.
An overhead conveyor system was used in all experiments at this
site. A Despatch oven was used for curing painted targets.
The laboratory was supplied with a manual airless spray gun,
associated hoses, high-pressure paint pump, and 015 spray tip.
For electrostatic transfer efficiency determinations, the
laboratory was supplied with a manual spray gun, 5-gallon
pressure tank with agitator, hoses, and 75 kV power supply- The
electrostatic air spray gun was also used to simulate conventional
air spray painting by conducting transfer efficiency determina-
tions with no voltage applied.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
November 11, 1985. Transfer efficiency testing began on November
12, 1985. Six replicate transfer efficiency determinations were
made for each equipment type. Airless and electrostatic air
spray tests were completed the first day of testing. Conven-
tional air spray equipment transfer efficiency determinations
were completed the following day. Operating conditions are
presented in Appendix E - Raw data.
The proposed standard transfer efficiency test method and spray
conditions, as set forth in Transfer Efficiency Method Evaluation
Plan, were followed for all transfer efficiency determinations.
All QA/QC requirements were met, and no special problems were
encountered during the experiment.
Test Results
Transfer efficiency results for these tests are presented in
Table 6-9. These transfer efficiency values are only applicalbe
to the particular system (equipment and paint) under test at
42
-------
TABLE 6-9- LABORATORY NO. 5 TEST RESULTS
Airless
Transfer efficiency (%): 30.1
29.9
29.5
30.5
30.1
30.7
Mean: 30.1
Standard deviation: .427
Electrostatic air spray
Transfer efficiency (%): 50.4
52.1
51.2
50.3
54.5
54.6
Mean: 52.2
Standard deviation: 1.94
Conventional air spray
Transfer efficiency (%) 27.6
27.8
29.0
28.6
27.2
27.9
Mean: 28.02
Standard deviation: .664
43
-------
specified operating conditions. Mean and standard deviations are
also summarized in Table 6-9.
A standard outlier test, using Nalimov's criteria, was conducted
on transfer efficiency results for each spray system. No
outliers were identified.
44
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 6
Test Facilities
Tests were conducted in a DeVilbiss spray booth. The laboratory
provided the spray painting area, conveyor system, curing oven,
and assistance required to complete the transfer efficiency
tests.
The spray painting area was roughly 427 cm by 305 cm, with 305 cm
ceilings (about 14 ft by 10 ft, with 10 ft ceilings). The booth
was a water wash type, which maintained the linear air velocity
at 56 cm/s (110 fpm), as required by the proposed standard
transfer efficiency test method. The booth area temperature was
23.9°C (75QF) during test runs. Paint and solvent were kept in
the booth area; paint and solvent temperatures closely matched
booth ambient temperatures. Relative humidity was 45-70 percent
during the test.
An overhead conveyor system was used in all experiments at this
site. A DeVilbiss oven was used for curing painted targets.
The laboratory was supplied with a manual airless spray gun,
associated hoses, high-pressure paint pump, and 015 spray tip.
For electrostatic air spray transfer efficiency determinations,
the laboratory was supplied with a manual spray gun, 5-gallon
pressure tank with agitator, hoses, and 75 kV power supply. The
electrostatic air spray gun was also used to simulate
conventional air spray painting by conducting transfer efficiency
determinations with no voltage applied.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
November 4, 1985. Transfer efficiency testing began on November
5, 1985. Six replicate transfer efficiency determinations were
made for each equipment type. Airless tests were completed the
first day of testing. Airless and electrostatic air spray
equipment transfer efficiency determinations were completed the
first day of testing. Conventional air spray transfer efficiency
determinations were completed the following day- Operating
conditions are presented in Appendix E - Raw data.
The proposed standard transfer efficiency test method and spray
conditions, as set forth in Transfer Efficiency Method Evaluation
Plan were followed for all transfer efficiency determinations.
All QA/QC requirements were met, and no special problems were
encountered during the experiment.
45
-------
Test Results
Transfer efficiency results for these tests are presented in
Table 6-10. These transfer efficiency values are only applicable
to the particular system (equipment and paint) under test at
specified operating conditions. Mean and standard deviations are
also summarized in Table 6-10.
A standard outlier test, using Nalimov's criteria, was conducted
on transfer efficiency results for each spray system. One
possible outlier was identified in run No. 15 of the conventional
air spray tests. In accordance with the test plan, no repeat of
the run was conducted. No explanation for the possible outlier
could be identified.
46
-------
TABLE 6-10. LABORATORY 6 TEST RESULTS
Airless
Transfer efficiency (%): 43.1
43.4
42.1
42.8
42.5
43.7
Mean: 43.0
Standard deviation: 0.547
Electrostatic air spray
Transfer efficiency (%): 70.4
71.3
71.8
71.3
74.6
69.2
Mean: 71.4
Standard deviation: 1.80
Conventional air spray
Transfer efficiency (%): 31.1
31.2
31.8
30.9
30.7
30.5
Mean: 31.0
Standard deviation: 0.455
47
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 7
Test Facilities
Tests were conducted in a DeVilbiss spray booth. The laboratory
provided the spray painting area, conveyor system, curing oven,
and assistance required to complete the transfer efficiency
tests.
The spray painting area was roughly 101.6 cm by 76.2 cm with 25.4
cm ceilings (about 40 ft by 30 ft, with 10 ft ceilings). The
booth was a dry back filter type, which maintained the linear air
velocity at 254-302.26 cm/s (100-119 fpm), as required by the
proposed standard transfer efficiency test method. The booth
area temperature was 18.3-23.9°C (65-75°F) during test runs.
Paint and solvent were kept in the booth area; paint and solvent
temperatures closely matched booth ambient temperatures.
Relative humidity was 46-60 percent during the test.
An overhead conveyor system was used in all experiments at this
site* A Gehnrich oven was used for curing painted targets.
The laboratory was supplied with a manual airless spray gun,
associated hoses, high-pressure paint pump, and 015 spray tip.
For electrostatic air spray transfer efficiency determinations,
the laboratory was supplied with a manual spray gun, 5-gallon
pressure tank with agitator, hoses, and 75 kV power supply. The
electrostatic air spray gun was also used to simulate conven-
tional air spray painting by conducting transfer efficiency
determinations with no voltage applied.
In addition to supplying spray painting systems, the contractor
provided paint, solvent, mass flow meter, pressure gauges,
control panel, scales, weight percent solids equipment, viscosity
cups, 60 test targets, foil, and other miscellaneous test
equipment. An engineer and technician were present during all
transfer efficiency determinations at Laboratory 7.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
December 17, 1985. Transfer efficiency testing began on December
18, 1985. Six replicate transfer efficiency determinations were
made for each equipment type. Airless tests were completed the
first day of testing. Electrostatic air spraying and
conventional air spraying tests were completed the following day.
Operating conditions are presented in Appendix E - Raw Data.
The proposed standard transfer efficiency test method and
spray conditions, as set forth in Transfer Efficiency Manual
Evaluation Plan were followed for all transfer
48
-------
efficiency determinations. All QA/QC requirements were met, and
no special problems were encountered during the experiment.
Test Results
Transfer efficiency results for these tests are presented in
Table 6-11. These transfer efficiency values are only applicable
to the particular system (equipment and paint) under test at
specified operating conditions. Mean and standard deviations are
also summarized in Table 6-11.
A standard outlier test, using Nalimov's criteria, was conducted
on transfer efficiency results for each spray system. No outliers
were identified.
49
-------
TABLE 6-11. LABORATORY 7 TEST RESULTS
Airless
Transfer efficiency (%): 48.8
46.5
48.7
47.7
48.2
44.2
Mean: 47.4
Standard deviation: 1.76
Electrostatic air spray
Transfer efficiency (%): 66.3
69.9
68.0
72.0
68.7
69.1
Means 69.0
Standard deviation: 1.91
Conventional air spray
Transfer efficiency (%): 39.7
38.3
37o4
38.5
39.9
38.6
Mean: 38.7
Standard deviation: 0.903
50
-------
TRANSFER EFFICIENCY TEST AT LABORATORY NO. 8
Test Facilities
Experiments were conducted in a DeVilbiss dry filter booth in the
engineering laboratory. Laboratory No. 8 provided the spray
painting laboratory, technicians, conveyor system, curing oven,
and other associated test materials.
The engineering laboratory was roughly 2300 cm by 1100 cm with
600 cm ceilings (about 75 ft by 36 ft with 20 ft ceilings). The
booth was a back-draw dry filter type, which maintained the
linear air velocity at 40.6-61.0 cm/s (80-120 ft/min), as
required by the proposed standard transfer efficiency test
method. The booth area temperature was controlled at 21.7-24.48C
(71-76°F) during test runs. Paint and solvent were kept in the
booth area; paint and solvent temperatures closely matched booth
ambient temperatures. Relative humidity ranged from 58 to 80
percent during the test.
A Unibuilt overhead conveyor system equipped with a Nordson
Countamatic timer was used in all experiments at this site. A
Grieve (Model SC 550) oven was used for curing painted targets.
The•laboratory was supplied with a manual airless spray gun,
associated hoses, high-pressure paint pump,, and 015 spray tip.
For electrostatic air spray transfer efficiency determinations,
the laboratory was supplied with a manual spray gun, 5-gallon
pressure tank with agitator, hoses, and a 75 kV power supply.
The electrostatic air spray gun was also used to simulate
conventional air spray painting by conducting transfer efficiency
determinations with no voltage applied.
Transfer Efficiency Tests
Equipment set-up and instrument calibrations were completed on
December 15, 1985. Transfer efficiency testing began December
16, 1985. Six replicate transfer efficiency determinations were
made for each equipment type. Airless and conventional air spray
tests were completed December 18, 1985. Electrostatic air spray
tests were completed on the subsequent day- Operating condi-
tions are presented in Appendix E - Raw data.
All pressure gages not previously calibrated were calibrated on a
dead weight tester. Calibration curves were developed and used
for pressure measurements. The mass flow meter was zeroed and
calibrated, then checked against paint capture to assure its
accuracy. At each flow rate a constant 1.2 percent difference
was detected between the meter and paint capture methods. Thus,
a calibration adjustment of 1.2 percent was applied to all mass
flow measurements from the meter. Although it was not required
by the test method or the QA/QC plan, a velocity profile was
51
-------
developed for the test booth. The booth air velocity was well
within QA/QC requirements.
The proposed standard transfer efficiency test method and spray
conditions, as set forth in Transfer Efficiency Method
Evaluation Plan were followed for all transfer efficiency
determinations. All QA/QC requirements were met, and no special
problems were encountered during the experiment.
Test Results
Transfer efficiency results for these tests are presented in
Table 6-12. These transfer efficiency values are only applicable
to the particular system (equipment and paint) under test at
specified operating conditions. Mean and standard deviations
are also summarized in Table 6-12.
52
-------
TABLE 6-12. LABORATORY 8 TEST RESULTS
Airless
Transfer efficiency (%): 46.0
44.6
45.7
44.8
44.8
46.0
Mean: 45.3
Standard deviation: 0.652
Electrostatic air spray
Transfer efficiency (%): 87.6
88.1
82.6
84.0
86.6
88.7
Mean: 86.3
Standard deviation: 2.44
Conventional air spray
Transfer efficiency (%): 25.9
25.6
25.4
25.3
25.0
25.9
Mean: 25.5
Standard deviation: 0.354
53
-------
SECTION 7
TEST RESULTS AND STATISTICAL ANALYSIS
TEST RESULTS
The results of the interlaboratory testing of the transfer
efficiency test method at eight field laboratories are presented
in Tables 7-1, 7-2, and 7-3 for airless, conventional air spray
and electrostatic spray equipment respectively. A summary of
the statistical results for all test sites and equipment types
is presented in Table 7-4.
STATISTICAL ANALYSIS
Introduction
The following statistical analysis is based on three assumptions.
First, it is assumed that laboratories were selected randomly
from the sample frame. This assumption cannot be tested, since
there is no definitive list of members of the sample frame
available. Second, it must be assumed that measurements were
unbiased and made independently. Finally, it is necessary to
assume that the within-laboratory variances are the same from
laboratory to laboratory. There is certainly evidence that this
assumption does not hold for test results for two equipment
types. One possible alternative is to eliminate laboratories
whose variances differ significantly from the others. Another
is to proceed while noting this anomaly. The statistical
analysis has been performed using all laboratories for each
equipment type. A special discussion is devoted to this
assumption on page 65.
54
-------
TABLE 7-1. TRANSFER EFFICIENCY RESULTS, AIRLESS
(% Transfer efficiency)
LABORATORY: # 1
RUN NO. 1: 44.9
RUN NO. 2: 43.8
RUN NO. 3: 45.2
RUN NO. 4: 44.2
RUN NO. 5: 42.9
RUN NO. 6: 44.0
MEAN: 44.2
STAND. DEV.: 0.821
SUMMARY
Within-lab variance:
Between-lab variance
(without gun variance):
Between-lab variance
(including gun variance)
Total variance
(without gun variance):
Total variance
(including gun variance)
f 2
35.3
34.4
32.9
31.3
33.6
30.8
33.1
1.73
2.30
42.93
: 43.81
45.23
: 46.11
t 3
49.4
47.5
44.9
48.5
46.4
50.5
47.9
2.06
#4 #5
39.9 30.1
42.1 29.9
38.5 29.5
36.2 30.5
35.5 30.1
36.5 30.7
38.1 30.1
2.54 0.427
Standard deviation:
Standard deviation:
Standard deviation:
# 6
43.1
43.4
42.1
42.8
42.5
43.7
43.0
0.547
1.53
6.72
6.79
17 #8
48.8 46.0
46.5 44.6
48.7 45.7
47.7 44.8
48.2 44.8
44.2 46.0
47.4 45.3
1.76 0.652
55
-------
TABLE 7-2. TRANSFER EFFICIENCY RESULTS, CONVENTIONAL AIR SPRAY
(% Transfer efficiency)
LABORATORY:
RUN NO
RUN NO
RUN NO
RUN NO
RUN NO
RUN NO
STAND.
. 1:
. 2:
3:
. 4:
. 5:
. 6:
MEAN:
DEV.:
* 1
32.7
36.9
35.4
34.9
34.0
33.3
34.5
1.53
#
24.
25.
24.
24.
23.
24.
24.
0.
2
9
2
0
0
8
2
4
565
# 3
39.6
39.4
39.0
39.2
39.1
38.6
33,2
0.345
#
34.
35.
35.
35,
34.
36.
35.
0.
4
9
7
3
6
2
3
3
723
f 5
27.6
27.8
29.0
28.6
27.2
27.9
28.02
0.664
* 6
31.1
31.2
31.8
30.9
30.7
30.5
31.0
0.455
# 7
39.7
38.3
33.4
38.5
39.9
38.6
38.7
0.903
# 8
25.9
25.5
25.4
25e3
25.0
25.9
25.5
0.354.
SUMMARY
Within-lab variance:
1.22
Between-lab variance
(without gun variance): 25.41
Between-lab variance
(including gun variance): 3} 53
Total variance
(without gun variance): £6 63
Total variance
(including gun variance): 32 34
Standard deviations
Standard deviations
Standard deviation: 5 73
56
-------
TABLE 7-3. TRANSFER EFFICIENCY RESULTS, ELECTROSTATIC AIR SPRAY
(% Transfer efficiency)
LABORATORY: * 1
RUN NO. 1: 65.3
RUN NO. 2: 70.3
RUN NO. 3: 73.8
RUN NO. 4: 70.8
RUN NO. 5: 71.7
RUN NO. 6: 72.1
MEAN: 70.7
STAND. DEV.: 2.89
SUMMARY
Within-lab variance:
Between-lab variance
(without gun variance):
Between-lab variance
(including gun variance)
Total variance
(without gun variance):
Total variance
(including gun variance)
#2 #3
65.4 68.7
65.8 71.2
65.2 70.2
67.8 70.8
66.6 72.6
66.2 71.6
66.2 70.9
0.95 1.33
3.63
79.44
: 85.03
83.07
: 88.66
#4 #5
70.6 50.4
67.1 52.1
68.9 51.2
68.7 50.3
69.4 54.5
69.7 54.6
69.1 52.2
1.17 1.94
Standard deviation:
Standard deviation:
Standard deviation:
# 6
70.4
71.3
71.8
71.3
74.6
69.2
71.4
1.80
1.905
8.70
9.42
#7 #8
66.3 87.6
69.9 88.1
68.0 82.6
72.0 84.0
68.7 86.6
69.1 88.7
69.0 86.3
1.91 2.44
57
-------
TABLE 7-4. TRANSFER EFFICIENCY RESULTS
AIRLESS ELECTROSTATIC CONVENTIONAL
VARIANCE STD. DEV. VARIANCE STD. DEV. VARIANCE STD. DEV.
WTTHIN-LAB 1.22 1.1 3.63 1.905 2.3 1.5165
BETWEEN-LAB 26.41 72.02 42.92
GUN 6.22 13.01 0.88
TOTAL 33.85 5.82 88.66 9.42 46.1 6.79
58
-------
Laboratory, Gun, and Within-laboratory Variance Components
For each equipment type, eight laboratories were used and six
runs were made for each laboratory- An attempt was made to
estimate the "between-laboratory" portion of variance and the
"within-laboratory" portion, assuming, of course, a homogeneous
within-laboratory variance. In addition, in a single laboratory
experiment, the between-gun portion of variance was computed.
With the eight laboratory experiments, a one-way random effects
model was assumed, with random laboratories. In other words, it
is assumed that the transfer efficiency result is produced from a
random effect due to the laboratory and a random effect
representing the random "within-laboratory" effect. This model
assumption is made simply because there are two components of
variation in light of the way the experiment was constructed.
This one-way random effects model then assumes that if we call
yu the /th transfer efficiency measurement by the rth laboratory,
ytj = n + t/ + 60
where T, is a random component contributed because of the ith
laboratory, and c,y is the "within laboratory" error. The reader
can view the T, and c^ as having distributions, each having a
population mean of zero and some variance. The variances will be
called a*w and ol for within laboratory and between laboratory
respectively. Thus, there are two variances that require
estimation from the data.
A standard analysis of variance procedure is used to estimate
these variances and thus the standard deviations. Following
elementary procedures in many statistics texts including Ott
(1984) and Walpole and Myers (1985), the estimates of these
"variance components" come by doing the analysis of variance of
the total variation in the data. The sample variance for the two
sources of variation is viewed in the experiment as
MSL (mean square between laboratories, the sample variance
between laboratory means)
MS^ (mean square within laboratories, the sample variance
within laboratories)
Now the estimates of the two variances o^ and a[ are obtained
from these two mean squares. Elementary manipulation leads to
estimates given by
"2
= MSW
— MSW
07, = - -— - —
*• n
59
-------
where " is the number of runs taken in each laboratory. The 'A'
type notation signifies that the quantity is an estimate. The
main objective of determining the estimates of a{ and a*w is to
ascertain a total variance or standard deviation of the method of
measuring transfer efficiency.This total variance is made up of
the sum of the "components of total variance," the latter being
o[ and a'w. The total variance is featured as the standard
deviation of a single observation of transfer efficiency taken at
a random laboratory. The percentage of the total variance
attributed to laboratory and within laboratory is reported. In
addition, the coefficient of variation is given, the latter being
the precision or "total standard deviation" expressed as a
percent of the mean transfer efficiency.
The computation of the estimated variance components was made
through the use of the Statistical Analysis System (SAS) PROC VAR
COMP- The total variance, of course, is computed as the sum of
the two components. In addition* the coefficient of variation is
given. The coefficient of variation is defined as
C.V. =
and expressed in percentage units. Its purpose is merely to be
able to express this standard deviation in a unitless way. The
fact that it is expressed as a percent has absolutely nothing to
do with the fact that transfer efficiency is expressed as a
percent. The purpose of the coefficient of variation is to
account for the fact that the precision of the measurement of
transfer efficiency may very well depend on the average size of
the transfer efficiency. The C.V. is intended as a method of
"expressing precision as a function of the size of the
measurement."
What is the interpretation of the Total Standard Deviation?
As was suggested earlier, the total standard deviation is a
measure of the precision of the transfer efficiency method with a
random laboratory«, It should be emphasized that the quantity
computed is only an estimate. However, if the true standard
deviation were known, then it could be said that roughly 95% of
transfer efficiency measurements would deviate from the mean by
± 2ar • Based on the computations made, the bounds
A
mean ± 2ar
represent estimates of bounds that cover 95% of the transfer
efficiency measurements. But it cannot categorically be stated
that ± 2ar covers 95% of transfer efficiency measurements simply
because values depend on a finite number of data points.
60
-------
Airless
The average transfer efficiency for the airless equipment is
41.1167 transfer efficiency units.
o2H. (within-laboratory component of variance) = 2.300
oi (between-laboratory component of variance) = 43.8116
o£ (gun component of variance) = 0.8777
oi* (between-laboratory variance, minus gun variance) = 42.9339
The gun portion of the between-laboratory variance is 2.01%.
of (total variance) •= 46.116.
95.01% of this variance is laboratory variance.
o'r (minus gun variance) « 45.2339
or (total standard deviation) = 6.7906 transfer efficiency
units
Thus the standard deviation in transfer efficiency at a random
laboratory is 6.7906 transfer efficiency units, which is 16.52%
of the mean transfer efficiency of the experiment. As indicated
earlier, the most important statistic is or, the estimated total
standard deviation. An estimate or an approximation of bounds on
transfer efficiency measurements is given by +_ 1.28ar = +9.1
transfer units. The purpose of the coefficient of variation was
indicated earlier.
Thus, if the true standard deviation was known (we have developed
an estimate), 80 percent of all transfer efficiency measurements
at qualifying laboratories using the same type of spray
equipment, paint, targets, and operating conditions would fall
within 9.1 units of the true transfer efficiency.*
At a random laboratory.- one is concerned with how close the
measured transfer efficiency is to the true transfer efficiency.
The results of this research indicate an 80 percent probability
that the measured transfer efficiency would fall within 9.1 of
the true transfer efficiency.*
*Provided that our bias assumption is correct.
61
-------
Conventional
Average transfer efficiency is 32 transfer efficiency units.
oj (within-laboratory component of variance) = 1.2165
o[ (laboratory component of variance) = 31.6277
02. (gun component of variance) = 6.2190
The gun portion of the laboratory variance is 18%.
oi* (laboratory component of variance - gun variance) = 25.4087
o:r (total variance) = 32.8442
96.3% of this variance is laboratory variance.
o^ (minus gun variance) = 26.6252
or (total standard deviation) = 5.731 transfer efficiency
units.
Thus the standard deviation in transfer efficiency units at a
random laboratory is 17.91% of the mean transfer efficiency of
the experiment. It is estimated that + 1.282 cr = + 1.282(5.731)
= ^7.33 transfer efficiency units covers roughly 80% of the
transfer efficiency readings around the mean.
Thus, if the true standard deviation was known (we have developed
an estimate), 80 percent of all transfer efficiency measurements
at qualifying laboratories using he same type of spray equipment,
paint, targets, and operating conditions would fall within 7,3
units of the true transfer efficiency-*
At a random laboratory, one is concerned with how close the
measured transfer efficiency is to the true transfer efficiency.
The results of this research indicate an 80 percent probability
that the measured transfer efficiency would fall within 7.3 of
the true transfer efficiency.*
*Provided that our bias assumption is correct,
62
-------
Electrostatic
Average transfer efficiency is 69.45 transfer efficiency units
o^ (within-laboratory component of variance) = 3.6302
o[ (laboratory component of variance) = 85.0295
o£ (gun component of variance) = 5.5912
The gun portion of the laboratory variance is 6.58%.
02* (laboratory variance - gun variance) is 79.4383
oJ7 (total variance) = 88.6597
95.91% of this total variance is laboratory variance.
o'j. (minus gun variance) « 83.0685
or (total standard deviation) = 9.4159 transfer efficiency
units
Thus the standard deviation in transfer efficiency units at a
random laboratory is 13.56% of the mean transfer efficiency of
the experiment. As before, +_ 1.282ar around the mean covers
roughly 80% of the transfer efficiency measurements. In this
case + 1.282or » + 12.03 transfer efficiency units.
Thus, if the true standard deviation was known (we have developed
an estimate), 80 percent of all transfer efficiency measurements
at qualifying laboratories using the same type of spray
equipment, paint, targets, and operating conditions would fall
within 12.0 units of the true transfer efficiency-*
At a random laboratory, one is concerned with how close the
measured transfer efficiency is to the true transfer efficiency.
The results of this research indicate an 80 percent probability
that the measured transfer efficiency would fall within 12.0 of
the true transfer efficiency.*
*Provided that our bias assumption is correct,
63
-------
Confidence Interval on Mean Transfer Efficiency Based on This
Experiment
The previous results regarding the precision of the transfer
efficiency method give an estimated standard deviation of a
single measured transfer efficiency value at a random laboratory.
To provide more information regarding accuracy of the method
combined with precision or reproducibility, a computation was
made which produced an estimate (for all three equipment types)
of the transfer efficiency for the conditions of this experiment,
along with a standard error of that estimate and an 80%
confidence interval on the transfer efficiency.
Based on the one factor analysis of variance, random effects
model, an estimate of the mean transfer efficiency for the
present experiment is y_ , the average TE over the entire
experiment, while the standard deviation of this average is
.= id
>- V e tn
where t is the number of laboratories and n is the number of
runs per laboratory. For a sketch of the proof of the above
result, see Appendix D.
Now, the standard deviation of y. is estimated by V in , where
MSL is the laboratory mean square in the experiment. (See
Appendix D). This produces a t-type confidence interval on the
mean transfer efficiency which is the parameter u in the
experiment. Thus an 80% confidence interval on the mean transfer
efficiency based on the results ^ of this experiment (and for the
conditions of this experiment, i.e., point-type, etc.) is given
by
These confidence intervals are as follows:
Airless - For the conditions of this experiment, the mean TE is
between 37.709 and 44.493 TE units, with 80% confidence.
Conventional - For the conditions of this experiment, the mean TE
is between 29.178 and 34.822 TE units, with 80% confidence.
Electrostatic - For the conditions of this experiment, the mean
TE is between 64.825 and 74.084 TE units, with 80% confidence.
64
-------
The purpose of this work is to get an impression of how much
error might be associated with an estimate of transfer efficiency
(actually, mean transfer efficiency) from an experiment such as
this. It is obvious that if there is a "true transfer
efficiency," and it was estimated from a sample mean from an
experiment such as this one, a confidence interval on the
"population mean transfer efficiency" is a clear way of
determining the accuracy of the estimate of "true transfer
efficiency-" Surely, if the between-and within-laboratory variance
is very large, it could be expected that the width of the confidence
interval in conjunction with the standard regarding how "tight"
the estimate should be is very important. For example, in the
case of the conventional equipment, the width of the 80 percent
confidence interval is 32.0 + 2.822. The issue then centers
around whether missing by + 2.822 is good enough.
Importance of the Homogeneous Variance Assumption
As was indicated earlier, it is assumed that the within-
laboratory variance is constant from laboratory to laboratory.
It is clear from the sample data that this, indeed, may not be
true. It is important that this be acknowledged and that the
impact of this be addressed. The fact that the within-laboratory
variances differ for at least two equipment types results in the
following conclusions:
(a) Laboratories are not equal in the precision with which
they measure transfer efficiency. Certainly, any
future experimental effort in this area should focus
on this.
(b) It was suggested earlier that there are alternatives
that could be used. The laboratories could be divided
into homogeneous groups, with each group containing
laboratories whose precision for measuring transfer
efficiency do not differ significantly. Thus, there
would be two sets of answers, one for the high
precision laboratories and one for the low precision
laboratories. This is clearly undesirable. Since the
intent of this work was to determine the precision of
the method, it must be interpreted as one result and
view the answer obtained as one result which has, in
a sense, been averaged over laboratories, even
though they do not perform with equal capability.
Thus, it is felt that the results given are as
reasonable as can be produced in the given situation.
65
-------
REFERENCES
U. S. EPA (1985). Air and Energy Engineering Research Laboratory
Transfer " i" ficiency of Improperly Maintain 1 or Operated
Spray Painting Equipmeac: T^nsitivi.ty Studies. EPA-
obo/2-35-!07 (NTIS PB86-108 271/AS). Novem'o•*.r 1935.
Steinberg, J. S. and W. Hunter (1984). "Experimental Design,"
Technometrics, May.
SteJner, ". H. and W. J. Youden (1975). Gtatia tio \1 "lanual of
the Association of Official Analytical Chemists.
Association of Official Analytical Chemists.
Ott, R. Lyman (19H4). An Tatroduction to Statistical Methods
and Data Analysis, 2nd edition, Boston: PWS.
Walpole, Ronald E. and Raymond H. Myers (1985). •Probability
and Statisti.c.s for EajLaeers and Scientists, 3rd edition,
"-Te^ York: MacMillan Publishing Company.
-------
APPENDIX A
PROPOSED STANDARD TRANSFER EFFICIENCY
TEST METHOD
A-l
-------
APPENDIX A
PROPOSED STANDARD
TRANSFER EFFICIENCY TEST METHOD
1.0 SCOPE
1.1 This method covers method verification testing at multiple
laboratories to define the interlaboratory characteristics
of the existing method.
1.2 The testing will be accomplished at a number of industrial
sites under controlled defined conditions, i.e., the
identical test protocol for applicable equipment will be
used at different locations. The final evaluation will
result in a complete characterization of the interlaboratory
characteristics of the TE method.
1»3 A significant number of tests will be conducted to define
the performance of the method for three automatic spray
equipment types: air atomized conventional, air atomized
electrostatic, and airless conventional.
2.0 APPLICABLE DOCUMENTS
2.1 ASTM Standards;
o D-1200 - 70 Viscosity of Paints, Varnishes, and
Lacquers by Ford Viscosity Cup
o D-2369 - 81 Standard Test Method for Volatile Content
of Coatings
o D-1005 - 51 Measurement of Dry Film Thickness of
Organic Coatings
o D-1212 - 79 Measurement of Wet Film Thickness of Organic
Coatings ?
o D-1475 - 60 Density of Paint Varnish, Lacquer, and
Related Products
o D-3925 - 81 Sampling Liquid Paints and Related Pigmented
Coatings
A- 2
-------
3.00 TRANSFER EFFICIENCY TEST METHOD
3.01 Inspect all equipment listed on Data Sheet 1 - Equipment
Specifications, and complete the data sheet where appli-
cable. All equipment and materials must meet the require-
ments of the approved Environmental Protection Agency (EPA)
quality assurance/quality control (QA/QC) plan.
Note: Place "N/A" in all cells that do not apply.
Ensure that the data sheet is dated and
initialed by both the person recording the infor-
mation and the person checking the information.
"Type" refers to the design of a given piece of
equipment.
3.02 Set up paint supply and mass flow measurement equipment per
manufacturer's instructions.
Note:
Paint supply and mass flow measurement equipment
must be grounded to avoid problems with static
electricity.
3.03 Calibrate the mass flow measurement equipment once per week
or each time that it is moved, whichever occurs more
frequently.
3.04 Begin agitation of paint at least thirty minutes before any
paint samples are taken.
3.05 Using a small glass jar with an airtight lid, take a paint
grab sample from the paint pot.
3.06 Record test run number on label of jar. (Each pass of ten
targets is a run.)
3.07 Complete Data Sheet 2 - Paint Specifications.
Paint weight percent solids should be determined at the
start of each day, at the end of each day, and any other
time deemed appropriate.
3.08 Set up the conveyor speed measuring equipment consisting of
photoelectric cells or limit switches used in conjunction
with a digital timer, or timing marks on the conveyor used
in conjunction with a stopwatch.
3.09 Cut an appropriate number of strips of 0.0037 cm (1.5 mil)
thick aluminum foil to dimensions of 38.1 cm (15 in) by
approximately 127cm (50in) for the testing.
A-3
-------
3.10 Consecutively number each precut foil strip on the dull
side using a permanent marking pen.
3,11 Weigh each foil strip and record the foil number and mass
on Data Sheet 3 - Mass of Solids Applied in the MASS OF
FOIL COLUMN.
Note:
Data Sheet 3 - Mass of Solids Applied will hold
the information from six runs.
3.12 Attach preweighed labeled foil (dull side to the target)
to six targets using the method shown in Figure A-l.
Attach unlabeled foil on four scavenger targets. All seams
must face away from the spray equipment.
3.13 Mount the foil-covered targets in consecutive order from
right to left (facing the booth), as shown in Figure A-2,
with the foil seam on each target facing away from the
spray gun.
3.14 Adjust all equipment operating parameters to the values
desired for testing.
3.15 Complete Data Sheet 4 - Operating Conditions and Calcula-
tions.
Note:
Cure time and temperature should be set per manu-
facturer's instructions.
3.16 Recheck operating parameters to ensure that they are
correct.
3.17 For electrostatic spray equipment, measure the operating
voltage and adjust according to manufacturer's instructions
and record value on Data Sheet 4.
3.18 Inspect conveyor clock, stopwatch, and mass flow measure-
ment equipment to assure that all are prepared to operate.
3.19 Turn on spray booth and conveyor. As the leading edge of
the first scavenger target passes in front of the gun, turn
on paint spray equipment and simultaneously begin mass flow
measurement.
3.20 As the trailing edge of the last scavenger target passes in
front of the gun, stop the paint spray equipment and mass
flow measurement simultaneously.
A-4
-------
3.21 Record the mass flow measurement on Data Sheet 4 - Opera-
ting Conditions and Calculations.
3.22 Measure the wet film thickness on the trailing scavenger
and record on Data Sheet 4.
3.23 Remove the painted targets from the conveyor and ensure
that no paint is lost. Securely hang the coated targets on
oven racks so all painted surfaces are exposed for uniform
drying. Orient all targets in the same direction in the
curing oven.
3.24 Insert racks in oven and bake at recommended schedule per
Data Sheet 4. Oven door should be opened for minimum
amount of time to prevent cooling.
3.25 Remove targets from oven and record actual cure schedule on
Data Sheet 4. Cool foil to room temperature. Remove foil
from each target, weigh foil and record mass on each foil
and on Data Sheet 3 in the Mass of Foil Plus Paint column.
3.26 After weighing, store foils in appropriately labeled
plastic bags, with the appropriate test run number labeled
on it. The laboratory, shall retain all samples until data
analyses are complete. Check all data for correctness and
completeness.
3.27 Perform the transfer efficiency calculations indicated on
Data Sheet 4.
3.28 Repeat steps 3.05 through 3.27 for each test run.
Note:
Follow QA/QC plan for equipment calibration and for
weight percent solids determinations when multiple
runs are anticipated.
3.29 Make sure all data sheets have been checked, dated, and
initialed.
3.30 When approximately 70 percent of the runs have been
completed, an outlier analysis shall be performed. Data is
to be recorded on Data Sheet 5. Repeat any outlier runs.
3.31 As part of Quality Assurance requirements, a QA report is
to be submitted to the CENTEC Quality Assurance Officer at
the end of each day.
A-5
-------
4.0 SAFETY CONSIDERATIONS
4.1 According to Section 9.8 of NFPA 33, when using fixed
electrostatic apparatus, the resistance of the equipment
to ground should be measured at a resistance of less than
1 x 10^ exponent Ohms.
4.2 If electrostatic equipment is being used, the gun-to-target
distance should be at least twice the sparking distance.
This requirement is in accordance with Section 9-7 NFPA 33*
A-6
-------
DATA SHEET 1 - EQUIPMENT SPECIFICATIONS
Test Date!__/__/
Foil Numbers to
LABORATORY SCALE
PLATFORM SCALE
MASS FLOW METER
CONVEYOR TIMER
STOPWATCH
PAINT SUPPLY TANK
PAINT SPRAY EQUIPMENT
PAINT SPRAY BOOTH
CONVEYOR
FORCED DRAFT OVEN
PAINT HEATER
TYPE
MANUFACTURER
MODEL
NUMBER
SERIAL
NUMBER
RATED
CAPACITY
RATED
ACCURACY
AIR CAP
FLUID TIP
NEEDLE
Data Collected byt
Data Checked by:
-------
DATA SHEET 2 - PAINT SPECIFICATIONS
Test fete:
Poll Number;
to
Manufacturer
Paint Type
Resin Type
Solvent
Manufacturer ID No.
Int No.
Color
RESIST-
IVITY
DATE TIME (M"/on2)
VISCOSITY
sect _ cup
at
°C)
FTJLL
SYRINGE WT.
(g)
EMPTY NET
FULL
DISH WT.
(g)
EMPTY
NET
SOLIDS
i
oo
Data collected by:
Data checked by:
-------
DATA SHEET 3 - MASS OF SOLIDS APPLIED
FOIL
NUMBER
MASS or FOIL "*ss °r, "*ss OF
PLUS PAINT (9) rOIL <9> PAINT <9>
-
— •
— •
-
-
— •
RUN NUMBER TOTAL MASS -
• K
— •
-
— •
-
— •
RUN NUMBER TOTAL MASS «
-
-
— •
-
-
-
RUN NUMBER TOTAL MASS -
FOIL
NUMBER
MASS Of FOIL MASS OF MASS OF
PLUS PAINT (g) TOIL <9) PAINT
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
roiL
NUMBERS
to
to
to
to
to
to
PRESSURE AT GUN (kPa)
FLUID ATOM. MR
ROTATING RPS
HI. FLUID K/0 FLUID
OPERATING
VOLTAGE (kV) RESIST. (MO)
TEMPERATURE (°CJ
AMBIENT FLUID
BOOTH AIR
RATE (ca/o)
RELATIVE
HUMIDITY
GUN TO TARGET
DISTANCE (cm)
CURE
riMEU) TEMP (°C)
-
I
I—»
o
FOIL
NUMBERS
to
to
to
to
to
to
VERTICAL
OQWBRAGE (»(
FILM «cm>
MET DRlf
FLUID MASS rUW RATE DETERMINATION
INIT. (9) FINAL (9) & (9) TIME (8 » RATE (9/1)
CDOWVEYOR
SPEED (CTR/S)
KEIGHT «
SOLIDS
NET DRY
SOLIDS (9*
TRANSFER
EFFICIENCY (»)
Data Collected byt
Data Checked by:
-------
DIRECTION OF
WRAPPING
h5.24on-J
(6 in) '
121.92 cm
(48 in)
STEEL WNEL
M.DMINUM FOIL
Scale: 1:12
Figure 1. Foil Attachment Technique
A-ll
-------
CONVEYOR
V ft
S
A
V
E
N
G
E
R
J J
S
c
A
V
E
N
G
E
R
J J
T
A
R
G
E
T
1
«
T
A
R
G
E
T
J J
A
P
G
E
T
J J
T
A
R
£
T
J J
T
A
Q
E
J J
I
_
'
T
J J
S
c
A
V
E
N
R
J J
S
A
V
F
N
J
R
1
r
121 n cm
(48in)
1
(6">> (6 in)
(IMin )
F* fixed liming mork
M-meving liming mark
A/^*""*~^^
z^ rf^V»
/ yT ^p
I \ /
STOP WATCH
Figure 2. Target Configuration for Transfer Efficiency Determination
-------
APPENDIX B
QUALITY ASSURANCE/QUALITY CONTROL
PLAN
B-l
-------
APPENDIX B
QUALITY ASSURANCE/QUALITY CONTROL PLAN
TRANSFER EFFICIENCY METHOD VERIFICATION PROGRAM
CENTEC CORPORATION
Resto/j, Virginia 22090
CONTRACT NO. 68-03-1952, Phase 2
AIR AND ENERGY ENGINEERING
RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, NC 27711
APPROVAL SIGNATURES?
QjUit.&
Charles H. Darvin
EPA Project Officer
O>^MV 9
0
Kerri C. Kennedy
CENTEC Project Manager!
JudTth Ford
EPA Quaj^ty ifesuran/cf officer
Dr. E. Handel
CENTEC QA/QC Officer
May 1985
B-2
-------
TABLE OF CONTENTS
Page
SECTION 1
SECTION 2
SECTION 3
SECTION 4
SECTION 5
SECTION 6
SECTION 7
SECTION 8
SECTION 9
SECTION 10
SECTION 11
PROJECT DESCRIPTION B-l
PROJECT ORGANIZATION AND RESPONSIBILITY B-5
A OBJECTIVES FOR MEASUREMENT DATA IN TERMS
TERMS OF PRECISION, ACCURACY, COMPLETENESS,
REPRESENTATIVENESS AND COMPARABILITY B-7
SITE SELECTION AND SAMPLING PROCEDURES B-ll
SAMPLE CUSTODY. . B-13
ANALYTICAL PROCEDURES, CALIBRATION PROCEDURES,
AND FREQUENCY B~16
DATA REDUCTION, VALIDATION, AND REPORTING B-18
INTERNAL QUALITY CONTROL CHECKS B-22
RESULTS OF PERFORMANCE AND SYSTEM AUDITS B-23
PREVENTATIVE. MAINTENANCE B-25
SPECIFIC ROUTINE PROCEDURES TO ASSCESS DATA
PRECISION, ACCURACY AND COMPLETENESS B-26
SECTION 12 CORRECTIVE ACTION.. . B-28
B-3
-------
SECTION 1
PROJECT DESCRIPTION
INTRODUCTION
This quality assurance/quality control (QA/QC) plan assures
collection of high quality data and insures consistent quality
control measures for data developed under "Phase II - Method
Verification Program," Contract No. 68-03-1952. Under this
contract, CENTEC Corporation will be conducting tests to deter-
mine the capability of the draft standard transfer efficiency (TE)
test method to precisely measure transfer efficiency.
PROJECT DESCRIPTION
This QA/QC plan is designed to ensure collection of high quality
data for interlaboratory testing of the draft standard transfer
efficiency (TE) test method. It encompasses the determination of
the capability of the draft standard TE test method to precisely
measure TE. The draft standard TE method (Appendix A) will be
used for all tests in this program. The test method consists of
passing a prescribed set of preweighed targets in front of spray
equipment under rigidly controlled conditions in an industrial
laboratory spray booth. The cured painted targets are weighed,
and the original weight is subtracted from the final weight to
obtain the net dry solids deposited on the targets. The net dry
solids are divided by the total solids sprayed at the targets,
B- 4
-------
which is then multiplied by 100 percent to determine TE. A
complete description of the draft standard TE test method is
provided in Appendix A.
This objective will be accomplished by testing the method at a
statistically determined number of industrial sites under
controlled conditions, i.e., the same test protocol will be used
at each location. Operating conditions for each spray system
will be determined at the first laboratory using a trial and
error approach to achieve reasonable coating thickness and
finish. This technique is used by industrial finishers to set
spray conditions, so test conditions are expected to simulate
actual industrial practice. Once spray conditions are determined
at the first laboratory, the same conditions will be specified
for all subsequent laboratories in the test program. The final
evaluation will result in a complete characterization of the
performance of the draft .standard TE test method. This evalua-
tion includes statistical information.
TE results from previous studies using the draft standard test
method have revealed information about variables which affected
the TE of those particular spray systems. These variables have
been shown to affect the TE of different spray systems in varying
amounts; their effect is not consistent from spray system to
spray system. Thus one of the major factors affecting TE, the
B-5
-------
spray system, is being held constant in this test program. Three
spray systems of different design (air atomized conventional,
airless conventional, and air atomized electrostatic) will be
tested at each laboratory. A new set of three spray systems
(same make and model) will be used at each participating lab-
oratory to ensure all spray equipment is in the same condition.
The same coating will be used .for all spray systems in this test.
Some of the variables which have been found to significantly
affect TE for other spray systems include gun-to-target distance,
target design, conveyor speed, linear air velocity, paint mass
flow rate, atomizing air pressure, shaping air, gun condition,
tip voltage, lag discharge distance, and electrode position.
Test parameters which can be controlled at well-equipped painting
laboratories (including gun-to-target distance, conveyor speed,
paint mass flow rate, gun condition, and target design) will be
specified in the test method after they are set in the first
test. Special attention has been paid to ensure that diffi-
cult-to-control variables (including booth air velocity and
shaping air) are consistently controlled and monitored to the
extent possible. Parameters which cannot be reasonably con-
trolled (including laboratory temperature and laboratory relative
humidity) will be carefully recorded during testing. If para-
meters which cannot be controlled are later discovered have a
significant effect on TE, enough data will be available for
B-6
-------
assessing which variable(s) might have been the culprit. Part of
the objective for this test program is to determine if uncontrol-
lable differences between laboratories are significant.
Testing is scheduled to begin June 3, 1985, and to continue for
up to eight months.
B-7
-------
SECTION 2
PROJECT ORGANIZATION AND RESPONSIBILITY
This project is administered through CENTEC Corporation struc-
ture, as shown in Figure B-l. Day-to-day test program activities
will be managed on-site by a CENTEC Project Engineer in direct
contact with CENTEC QA management personnel.
At the test site, the CENTEC engineer is responsible for imple-
menting QA throughout the test program. The engineer conducts
onsite evaluations to verify the degree of implementation,
assures that appropriate QA records are kept, provides QA
direction to the laboratory staff, and reports regularly to the
Project Manager on the status of QA.
Dr. Ted Handel is the Quality Assurance Officer for this pro-
ject. Dr. Handel functions independently from Project Manage-
ment, reporting directly to the Vice President of CENTEC Applied
Technology. He continuously monitors the implementation of QA
and provides feedback to the CENTEC engineer onsite and to
CENTEC management. Daily QA records are kept by the engineer
(onsite) and submitted weekly to the Quality Assurance Officer
(offsite). These records serve as resources for preparing
reports and documenting adherence to QA procedures and specifi-
cations .
B-
-------
CHAIRMAN AND
PRESIDENT
PAUL S. MINOR
(Director)
VICE PRESIDENT
FINANCE AND
ADMINISTRATION
ROBERT D. SMITH
VICE PRESIDENT '
PROCESS SYSTEMS
L. THOMAS SNIDER
VICE PRESIDENT
CENTEC APPLIED
TECHNOLOGIES
ROBERT SCHAFFER
VICE PRESIDENT
MANAGEMENT SYSTEMS
CURT GRINA
00
i
QUALITY ASSURANCE
OFFICER
TED HANDEL
SENIOR PROFECT
ENGINEER
K. KENNEDY
Figure B-l. Project Organization as Related to Corporate Structure
-------
SECTION 3
QA OBJECTIVES FOR MEASUREMENT DATA IN
TERMS OF PRECISION, ACCURACY, COMPLETENESS,
REPRESENTATIVENESS AND COMPARABILITY
TE test conditions will be set according to a trial and error
approach at the first laboratory. Operating conditions will be
set at approximate appropriate levels according to experience or
manufacturer's recommendations (manufacturers' representatives
will be on site for this portion of the first test), then a spray
pattern will be taken. Spray conditions will be refined to
correct any irregularity in the pattern, then another pattern
will be taken. This procedure will be followed until the spray
pattern (shape, thickness, and finish) are of reasonable indus-
trial quality. Once an acceptable spray pattern is achieved,
spray conditions will be recorded. These spray conditions will
be used at all subsequent, laboratories in this test program.
Specified spray conditions include gun-to-target distance, fluid
pressure, air pressure, shaping air, conveyor speed, voltage,
pattern size, and paint viscosity. (Refer to AppendixE , Data
Sheet 4.)
This technique for determining spray painting conditions is the
same as industrial practice, and is therefore expected to be
B-10
-------
highly representative of operating conditions that might have
been set by industry for the systems being tested.
For each major measurement parameter, specific QA objectives for
precision, accuracy, and completeness are required. These
objectives are detailed in Table B-l. Not all test conditions
are measured directly as listed in Table B-l, for instance, mass
flow measurements may be derived from weight and time measure-
ments and cure conditions are a combination of time and tempera-
ture.
Care must be taken to assure that all measurements are repre-
sentative of the media (paint) and conditions (spray conditions)
being measured. Proven techniques or methods are therefore used
for all measurements.
•
Data quality objectives are based on accuracy and precision of
each measurement, as established in Table B-l. Data integrity
will be validated through a series of inspections and tests
described later in this plan.
Data completeness objectives are 100 percent. This objective
will be met by subjecting each data sheet to two reviews, one by
a laboratory representative and one by the CENTEC Project
Engineer at the test site. If a piece of data cannot be
B-ll
-------
Table B-l. Spray Painting Transfer Efficiency Precision,
Accuracy and Completeness Objective
Measurement Parameter
(Method)
o Weight
o Grounding
o Mbltage
o units
o Distance-length
o Time (stopwatch.
timer)
0 Met Film Thickness
o Dry Film Thickness
W o Viscosity (Ford cup)
S— '
ro
o Resistivity
o Pressure
o Relative Humidity
Reference
Method
IEEE Std 32-1972
ANSI/IEEE Std 142-1972
ANSI C2
IEEE Std 4-1978
ASTM E 380-76/
IEEE Std 268-1976
(See ASTM 1200-70)
ASTM D-1212-79
ASTM 0-1005-51(1079)
ASTM D 1200-70(1976)
(Sling pajchroroeter)
o Temperature (cure conditions)
o Linear Mr Wlocity
(rotating vane or
heated wire anenoneter)
o Density
o Wt% Solids
o Paint Sampling
o Condition in Container
o Conveyor Speed (derived
AOGIH iReoantnended
Practice, Section 9*
ASTM D 1475-60(1980)
ASTM D 2369-81
ASTM D 3925
ASTM D 3011-1
from Time and Distance)
o Mass Flow Measurement (mass flow meter method)
Experimental Conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
Laboratory conditions
In accordance with NFPA 33
To be determined
To be determined
To be determined
To be determined
Precision
(Std. Deviation) Accuracy
lab scale 0.01 g lab scale +0.01 g
plat, scale 5 g plat, scale +5 g
~ —
0.05 kV +0.1 kV
— —
0.08 on 0.04 cm
0.1 s 0.2 8
0.265 mil 0.85 mil
2% 2%
+0.1 mil
1.5 e 2 a
Ool MO 0.1 MO
+1% +1%
1»F 3%
0.1'C 0.1"C
3% +3%
+0.001 g/mL 0.002 g/faL
1.5% 4.7%
— —
_ _
_ _
+0.4% +0.9%
Conpleten
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
•Industrial Ventilation - A Manual of tecamnended Practice, American Conference of Governmental Industrial Hygenists, 1972.
-------
obtained, the CENTEC engineer will note the reason for fairing to
meet completeness objectives (i.e., power failure, broken
measurement apparatus) on the data sheet. Every effort will be
made to replace faulty measurement apparatus as quickly as
possible. The CENTEC Quality Assurance Officer will be notified
whenever completeness objectives are in jeopardy.
In the event that a test run is conducted without 100 percent
completeness, the QA Project Officer will determine if the
missing data are critical or if they are necessary to perform TE
calculations. If the missing data are considered critical or
necessary, the test run will be voided. Voided test runs will be
repeated as soon as the problem is corrected. If the missing
data are not integral to performing TE calculations, the.QA
officer will make a ruling about whether the test run should be
voided and repeated, or accepted pending passing an outlier
analysis.
B-13
-------
SECTION 4
SITE SELECTION AND SAMPLING PROCEDURES
SITE SELECTION
Previous tests have shown the importance of spray booth config-
uration for transfer efficiency determinations. To minimize this
variable, CENTEC has developed a set of requirements that each
laboratory must have in order to participate in the test pro-
gram. These requirements were developed in close consultation
with the spray painting transfer efficiency Steering Committee.
They represent our collective best professional judgement of how
to specify laboratory criteria well enough to control transfer
efficiency at a satisfactory precision, but without being so
stringent that only a few laboratories in the country would
qualify.
First, the laboratory must be equipped with a back-draw spray
booth, preferably of water wash design. Dry filter booths may be
considered if filters are changed frequently to maintain air flow
rates. Air velocity in the booth must be 100 fpm (plus or minus
20 fpm) in the center of the booth. Downdraft booths are not
acceptable for this test program.
Participating laboratories must have an adjustable-rate overhead
conveyor system capable of hanging the standardized targets as
B-14
-------
prescribed. Participating laboratories must provide adequate
laboratory balances, work areas for foil wrapping and data
reduction, hangers, cleaning solvents, utilities, two know-
ledgeable technicians, and security for test equipment. They
must also provide a curing oven capable of accomodating EPA
targets and curing them at controlled prescribed temperatures.
Given these restrictions, CENTEC began making contacts with spray
equipment manufacturers, coatings companies, other interested
manufacturers, paint associations, and other possible operators
of spray painting laboratories. Two qualifying laboratories have.
already been located; they are scheduled for tests in June and
July, 1985. Eight other laboratories have shown interest in
participating in the test program. Information about qualifying
requirements has been sent to these companies.
SAMPLING PROCEDURE
A description of the sampling procedure is provided in Appendix
A, draft standard TE method. It includes:
o A description of the test method, including references
to standard methods
o Figures illustrating specific operations
o Description of sampling and test equipment
o Data sheets
o Other special conditions and considerations in
performing the test
o Data reduction equations
B-15
-------
SECTION 5
SAMPLE CUSTODY
This test program does not generate "samples" in the usual
sense. This test program produces sets of data sheets and sets
of painted foil targets. The data sheets and painted foil
targets are considered "samples" for the purposes of this
section.
The onsite CENTEC engineer will be responsible for obtaining and
recording necessary information on the data sheets, and shall
retain all original data generated by the test program. Pro-
cedures and forms for data sheets are presented in Appendix A.
Data sheets will be stored in an orderly fashion in a Test
Noteboook during and after the test. The Test Notebook remains
in CENTEC custody from its hand delivery to the test site,
through the performance of all tests, and shall be hand carried
back to CENTEC's corporate offices in Reston, Virginia. In
addition to the Test Notebook, CENTEC engineers will maintain a
comprehensive Log Book for special notes and observations during
the test program.
Painted foils will be stored by the participating laboratory in
sealed plastic bags labeled to indicate the test date, site,
equipment type tested, and foil indentification numbers. Foils
B-16
-------
will be stored by the participating laboratory until all data
analysis is complete. Labels may be hand written in indelible
ink on the plastic storage bags.
Field tracking reports will be kept daily, and submitted weekly
to the CENTEC Quality Assurance Officer (see Table B-2). The
CENTEC engineer and laboratory technician will check and sign all
data sheets and tracking forms. The laboratory will retain all
weighed foils, as described in the draft test method, until the
data analysis is complete. Paint grab samples and target storage
bag identification numbers will be recorded on sample custody
sheets. CENTEC will retain all original data sheets.
B-17
-------
Contract No. EPA 68-03-1952 **09
FELD TRACKNC REPORT
Laboratory Location
Reid Sample Code
Foi Number
Brief Description*
~
Data
TImefo)
Sampler
*AAC, AAE, ALC
Table B2. Sample of field tracking report form
B-18
-------
SECTION 6
ANALYTICAL PROCEDURES, CALIBRATION PROCEDURES, AND FREQUENCY
ANALYTICAL PROCEDURES
Analytical procedures for determining transfer efficiency are
discussed in the draft standard TE test method, Appendix A. The
draft standard TE test method is not a traditional laboratory
procedure; it is performed at industrial spray painting facil-
ities using almost entirely equipment that is readily available
on site. The QA/QC plan makes certain accuracy and precision
requirements for instrumentation to measure test parameters, but
does not specify make or model for instrumentation. Therefore,
CENTEC cannot provide detailed operating and/or calibration
instructions for each piece of instrumentation at every labor-"
atory. The party responsible for conducting the TE test (in
this case CENTEC) must ensure that equipment and instrumentation
meet the requirements of -Table B-l. The responsible party must
also ensure that participating personnel follow manufacturers'
instructions regarding equipment calibration, frequency, and
use. If calibration or operating instructions are unavailable,
CENTEC will attempt to locate instructions from the manufacturer
or supplier. If unsuccessful, the engineer will report the
problem to the QA Officer for resolution.
B-19
-------
CALIBRATION PROCEDURES AND FREQUENCY
Participating laboratories will be provided with pre-calibrated
test equipment for determining test pressure, linear air vel-
ocity, relative humidity. These instruments will be calibrated
befora and after each test series, and as deemed appropriate by
either the CESTEC engineer (onsite) or the CENTEC QA Officer
(offsite).
Complete manufacturer's instructions for calibration of other
test equipment (including test equipment such as the mass flow
meter supplied by CEtfTEC to participants) shall be followed for
all other equipment.
B-20
-------
SECTION 7
DATA REDUCTION, VALIDATION, AND REPORTING
GENERAL
Data will be collected at the test laboratory under the guidance
of a CENTEC engineer. The data will be collected and documented
according to the requirements of the draft standard TE test
method. Equations for reducing the data are also contained in
the draft standard TE test method.
DATA REDUCTION, VALIDATION, AND REPORTING
Figure B-2 shows the responsible parties for each data validation
and reduction step. Data reduction will be performed using*:
(Weight of cured painted foil - Weight of clean foil)(100%)(Total spraying distance)
(Paint weight fraction solids)(Total solids sprayed)(Effective target width)~~~
which simplifies to:
•pg — 15 333 (Weight of cured painted foil-Weight of clean foil)
(Paint weight percent solids)(Total solids sprayed)
for the prescribed target configuration.
Any data generated by test runs with known discrepancies in per-
formance (i.e., a smudged test panel, or a test run at different
conditions than specified) will be labeled as suspect for later
evaluation. Duplicate data for all suspect runs will be obtained
whenever resources permit.
*SI Units
B-21
-------
CENTEC ENGINEER ON-SITE
CENTEC PROJECT MANAGER
DATA COLLECTION
TE, SDff COV .
CALCULATIONS
CD
I
ro
ro
OUTLIER
EVALUATION
STATISTICAL
ANALYSIS
Figure B-2. Data Validation Responsibilities
-------
For each experimental design, the reduced data will be subjected
to a series of tests using Nalimov's criteria to evaluate
outliers. This evaluation will be performed onsite when
a test series is complete, but before taking down the spray
equipment. Outliers will be replaced by duplicate runs as
resources permit. The CENTEC field engineer will report any
questions or problems to the Quality Assurance Officer on a daily
basis.
CONSIDERATION OF OUTLIERS
Outliers will be searched for both within the data of each
laboratory and among the results of the many laboratories.
Nalimov's test for outliers will be utilized. Each suspect data
point within a set of data generated at a single laboratory
will be tested against the mean of that data set according to
Nalimov's Factor*:
|x* - x|
2
n
s V(n - 1)
where x* is the suspect value, s is the standard deviation, x is
the mean of the set, and n is the number of observations in the
set. Outliers are classified as possible, probable, or definite
according to whether the value of R exceeds its 95 percent, 99
percent or 99.9 percent confidence limit respectively.
B-23
-------
Possible outliers will be retained unless a defect in the experi-
ment can be identified. Probable outliers will be identified and
replaced with a repeated run whenever possible. Definite
outliers will be rejected in every case and an experimental
explanation sought.
These considerations will govern the handling of outliers among
the various participating laboratories. The mean of all obser-
vations for each gun type will be tested against the mean value
obtained by all laboratories, and the Nalimov criteria applied*
B-24
-------
SECTION 8
INTERNAL QUALITY CONTROL CHECKS AND AUDITS
Internal quality control checks are incorporated into the
experimental design and draft standard TE test method. These
checks include a battery of six replicates for each type of spray
equipment to be tested. Replicates will be examined for outliers
as described in Section 8.
The use of blanks, spiked samples, and similar controls is not
appropriate for this test program. Due to the nature of con-
ducting transfer efficiency tests, these options are not prac-
tical. Since there are no reagents or calibration standards
directly applicable to TE determinations, these are ruled out as
well.
B-25
-------
SECTION 9
RESULTS OF PERFORMANCE AND SYSTEM AUDITS
The performance of the TE tests will be monitored constantly as
described in draft standard TE test method.
In addition, performance and system audits will be performed by
the Quality Assurance Officer or by his designated represen-
tative. This designation is intended to eliminate any question
of conflict of interest in the performance of audits.
After the spray painting system is operational, performance
audits will be conducted to assure continued acceptable precision
during testing. It is the nature of the experimental design for
this program that TE results cannot be tested for outliers until
a test series is complete. To minimize the likelihood of
obtaining poor TE results prior to outlier analyses, internal
audits are required twice daily for each major measurement
contributing to TE:
o Net solids on target, g
o Conveyor speed, cm/s
o Paint weight fraction solids
o Paint mass flow rate, g/s
o Effective target width, cm
o Target spacing, cm
B-26
-------
These measurements are subject to the precision, accuracy, and
completeness criteria in Table B-l. They will be examined for
precision and accuracy at the beginning and completion of each
test series. Instrumentation such as the mass flow meter will be
calibrated and checked against paint capture on a twice daily
basis. Periodic audits also may be conducted during the test day
as deemed appropriate by either the laboratory technician or
CENTEC engineer on site. Performance audit requirements are
detailed in Table B-l and in the draft standard TE test method.
B-27
-------
SECTION 10
PREVENTATIVE MAINTENANCE
Preventative maintenance practices in the program are those
recommended by the manufacturer to the spray equipment user.
These practices include keeping the spray equipment and spray
area clean, handling equipment carefully to avoid damage, and
using appropriate equipment for the given job. These general
practices must be observed to prevent inadvertent deterioration
of spray equipment condition and to minimize downtime.
In addition to these preventative maintenance practices, extra
electrodes and air caps should be kept on hand. Ample supplies
for performing TE tests should be available to avert shortages.
These include foil, paint, solvent, and other supplies outlined
in the test method.
B-28
-------
SECTION 11
SPECIFIC ROUTINE PROCEDURES TO ASSESS DATA PRECISION,
ACCURACY AND COMPLETENESS
The precision and accuracy of each component the total measure-
ment system will be documented at the beginning and end of each
test series. (Refer to Table B-l.) Problems identified by the
performance audit will be corrected before continuing with the
test program.
Accuracy is calculated based on comparison to a reference.
Pressure gages are calibrated against a laboratory-standard
gage. Voltage readings obtained with the experimental equipment
are calibrated against the laboratory standard, etc. Test
instrumentation is adjusted until accuracy criteria in Table B-l
are met, where accuracy is calculated as:
%Error = Indicated Value - True Value x 100
True Value
Precision is measured as the standard deviation of a series of
measurements, thereby determining the repeatability of the
measurement. Standard deviation is calculated as:
2
? (x.-x)
n-1
where x is the mean of the series of measurements, x is the
value obtained in each measurement, and n is the number of
observations making up the series.
B-29
-------
Determination of the overall precision of the test method is the
objective of the interlaboratory test program. The completeness
objective for all readings and data points is 100 percent. The
method of testing for outliers presented in Section 8 during the
test series, and the fact that outliers will be replaced by
duplicates, will insure that completeness objectives are meet.
Completeness requirements are audited continuously and auto-
matically by the dual check-off procedures required on each data
sheet in the draft standard TE test method.
B-30
-------
SECTION 12
CORRECTIVE ACTION
Performance audits are required twice daily for each measurement
contributing to TE. Should any measurement not meet the pre-
cision or accuracy requirements laid out in Table B-l, corrective
action must be taken. Corrective action includes recalibration,
repair, or replacement of the measurement system in question.
The CENTEC engineer on site is responsible for initiating the
appropriate corrective action, with concurrence from the parti-
cipating required in writing in the next QA report to management.
Corrective action may also be taken to replace data identified as
erroneous by .the required data outlier analysis. The CENTEC
Project Manager is responsible for initiating corrective action
to replace outlier data.
Other corrective action may be taken at the request of onsite
CENTEC or laboratory personnel whenever suspect or undocumented
conditions occur. The CENTEC engineer is responsible for all
such corrective actions.
B-31
-------
APPENDIX C
DETECTION OF OUTLIERS BY MEANS OF
NALIMOV'S TEST
C-l
-------
CONSIDERATION OF OUTLIERS
Outliers were searched for, both within the data of each
laboratory and among the results of the eight laboratories.
Nalimov's test for outliers was utilized (see below). Each
suspect data point within a set of data generated at a single
laboratory was tested against the mean of that data set according
to Nalimov's Factor:
R =
- i2
x - x
where x is the suspect value, s is the standard deviation, x is
the mean of the set, and n is the number of observations in the
set. Outliers were classified as possible, probable, or definite
according to whether the value of R exceeded its 95 percent, 99
percent, or 99.9 percent confidence limit, respectively.
Possible outliers were retained unless a defect in the experiment
was identified. Probable outliers were identified and replaced
with a repeated run whenever possible. Definite outliers were
rejected in every case and an experimental explanation sought.
Similar consideration governed the handling of outliers among the
various participating laboratories. The mean of all observations
for each gun type was tested against the mean value obtained by
all laboratories, and the Nalimov criteria applied.
C-2
-------
APPENDIX D
STATISTICAL ANALYSIS ADDENDUM
D-l
-------
Appendix D
The random effects model is given by
i = 1, 2, ... , t
yy = n + TI + e//
j = 1, 2, .... n
where T, is the effect of the rth lab and e,; is the /th random disturbance within each laboratory.
The average y__ is given by
. =
Now, Var(T,) = CT! and Vai^e,,) = a^ If one takes the variance of the right hand side of the above,
one obtains, after simplification
and thus
2 _2
L W
Now, an estimator of aj is given by MSt/nt. This is easily verified since standard methods reveal
that
£(MSL) = oV -
thus MSJnt is an unbiased estimator of -~- + •— .
D-2
-------
APPENDIX E
OPERATING CONDITIONS AND
RAW DATA SUMMARY
E-l
-------
LABORATORY 1
E-2
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
,W".'(U
(a Ik 3 2
ut'l
FOIL
NUMBERS
-)*£
i3to /£
n*-*i
?«_**:
5Mtoe£
SWU
PRESSURE AT GUN (Kpa)
FLUlb ATOH.JUR
1 ^2*7 ^^
I O^-^
\fcz£
(gOO
MM
\
?
ROTATING RPS
MI. FLUID M/0 FLUID
VJIA
1
Mlft
\
i/
OPERATING
VOLTAGE (kV) RESIST. (MO)
MM
f
k)/A
N
/
TEMPERATURE (°C)
AMBIENT FLUID
z«.et^
^•c
a.^c
,^*c
ZWL
as-c-
^
/
*
BOOTH AIR
RATE {cm/m}
(^^0^
>
/
RELATIVE
HUMIDITY
-^'/n
/'
, ./
•
_50*2
OTN TO TARGET
DISTANCE (cm)
**;£?
N
-
CURE
riME(«)TEMP (°C)
15^7^
V
/
i
x
l
Z
6/622
4653g
6/W
6IT-H
FOIL
NUMBERS
7*12
r^ig
Kf-zH
zst0^
V*&
%^\
VERTICAL
COVERAGE (%)
3^
-^
*>r)
-to
FILM (cm}
(JIET} DRT
(9>ru^
0(*^s
O'(d fOUA^
FLUID MASS FLOW RATE DETERMINATION
INIT. (g) FINAL (g) A (g) TIHE(i) RATE (g/«)
N*
1
\
/
t\J In
\
f
11,3.1
I7/.7
l(/)*n . O
1^.4,
r^/^
14,^
H-0
^•f?
ll .52
;;.^i
n. n
0.50
CONVEYOR
SPEED (o>/«)
^,^/A;f]
\
• /
/
WEIGHT %
SOLIDS
5o /c)
^--erjO A
O ^ '^-^
£~ ^y^/
5^/o
NET DRY
SOLIDS (g)
TRANSFER
EFFICIENCY (%)
Data Collected byt
Data Checked byt /^ *e = 54.68/bxd\
Vaxc/
or e = 15,833 / d \
Vcxfj
-------
SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN (KP«)
rLUIDA)J/ ] ATOM. MR
ROTATING RPS
HI. FLUID N/O FLUID
OPERATING
VOLTAGE (kV) RESIST. (HO)
TEMPERATURE (°C)
AMBIENT riJJID
BOOTH AIR
RATE (d«/«)
RELATIVE
HUMIDITY
(TUN TO TARGET
DISTANCE (cm)
CURE
TIME (•) TEMP (°C)
to
A/
A
Al/ft
37T
to
31 f
to
^75-
to
to
to
w
FOIL
NUMBERS
VERTICM,
COVERAGE $«
MASS FLON SATE DETERMINATION
g> & (9) TIME(») RATE
CONVEYOR
SPEED (cm/»»
NEtGHT \
SOLIDS
WET DRY
SOLIDS (9)
TRANSFER
EFFICIENCY (%)
to
.
to
to
332 33
to
to
to
Data Collected byt
Data Checked byi
:8/bxd\
Vaxc/
or e=15,833
\cxf
-------
DATA SHEET 4 - OPERATING CONDITIONS AMD CALCULATIONS
FOIL
HUHBERS
PRESSURE AT CUM (KP«)
FLUID ATOM.AIR
ROTATING DPS
HI.FLUID W/0 FLUID
OPERATING
VOLTAGE (kV) RESIST. (MO)
TEMPERATURE (°C»
AHBIEMT FLUID
BOOTH AIM
RATE (cm/f)
RELATIVE
HtmiDITT
SUN TO TARGET
3ISTANCZ (cm)
CURE
riME(B)TCNP
/OO^/fff
37^
JO
375
3
to
w
I
FOIL
NUMBERS
VERTICAL
COVERAGE (%)
FILM (CB)
&f DRT
' FLUID MASS FLOW RATE DETERMINATION
INIT. (9) FINAL (g) A (9) TIMEU) RATE td/s)
OONVETOR
SPEED (C*/B)
HEIGHT %
SOLIDS
NET DRT
SOLIDS (9)
TRANSFER
EFFICIENCY (%)
to
/4-f
to
2.0°!°
23.
to
to
i/l.l
42.4-
III.*
to
fo
57.
Data Collected
Data Checked byi
*e=54.68/bxd\
Vaxc/
or e = 15,833 / d \
\cxfj
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN (**•)-
FLUID ATOM.AIR
_J3BUU»ll»l. HI1!!"" OPERATING
•IiTMIT" M.fn rimn InriTTnrr (kv) RESIST. (Hfl)
TEHPERATURB !°C)
AMBIENT FLUID
BOOTH AIR
RATE (cm/t)
RELATIVE
HUMIDITY
GUN TO TARGET
)I STANCE (en)
CURE
rlME(»)TEHP
~7
H/ft
HA
[00
/Sn
S
Zi
ps,
300
51
I
71.
**>!
71 s'r
10, N
to*
, to
|00
i
en
C 17; SC
FOIL
NUMBERS
11
to a
VERTICAL
COVERAGE C*l
zl
•1)
FILM (cm)
MET DRT
o .M
INIT.
FLUID MASS
«q) FINAL «gj
MATE DETERMINATION
A (9) TIMECc) KATE (g/»J
353
351
H/A
28 ¥5
OOHVETOR
SPEED (d«/«)
/0
MEIGHT %
SOLIDS
. "2,
WET DRT
SOLIDS (q)
n.oi
TRANSFER
EFFICIENCT (%)
32.7^
Data Collected byi
Data Checked byt
*e=54.68/bxd\
Vaxc/
or e = 15,833 / d \
(cxf J
6.7?
-------
»>y ru\DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN (kP«)
FLUID ATOM.A1R
OPERATING
VOLTAGE (kV) RESIST.(MO)
TEMPEJUkTORB <°C)
AMBIENT FLUID
BOOTH AIR
RATE (o»/«)
RELATIVE
HUMIDITY
SOU TO TARGET
DISTANCE (en)
CURE
flME (•) TEMP (°C
15 -
X
Zl
I'll*
!(,&
.s r-
3 C
P'l
300
IS/-
f
\O IK
TOIL
NUMBERS
VERTICAL
COVERAGE (%)
(cat
DRT
*r<0"> 'FLUID MASS FLOW RATE DETERMINATION c
«U1. J
-------
fine
'ATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT CON «KP«»
FLUID ATOM. AIR
*8T»Tina RI a
OPERATING
VOLTAGE (kV) RESIST. (W7)
TEMPEFATURE «°C}
AMBIENT f&UID
BOOTH AIR
RATE (cm/>)
RELATIVE
HUMIDITY
GUN TO TARGET
31 STANCE (cm)
CURE
rtHE(»)TEMP
£1356
M, M
, to ,
i to
25.HC
DM
,5 C
IpSI
1 6
IS
25,56
70
2-5"
\ 0 , K
7^5 F
10 i K
M^JTF
M
I
O3
FOIL
NUMBERS
VERTIOU.
COVERAGE f*$
DRT
rLtno MASS n/m RATE DETERMINATION
INI?, fg) FINAL
(9) TIM£(«) RATE
TONVETOR
SPEED (cn/s)
HEIGHT %
SOLIDS
MET DRT
SOLIDS (9!
TRANSFER
EFFICIENCY (%)
c
C I
20
47-5
0
S7JO
^ r?
158
o
Data Collected byt
Data Checked byt
*e=54.68/bxd\
Vaxc/
or e = 15 8 8 3 3 / d
-------
CI
CI370
113S5
w
I
us
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN (KPa)
FLUID ATOH.AIR
10(7^1
HI'S
TTJ-LJlTn 1I| 1 1 IIIITtl
/0
160
OPERATING
VOLTAGE (kV) RESIST. (Ml))
TEMPERATURE (°C)
AMBIENT FLUID
7'-ojc
70.SF
BOOTH AIR
RATE
RELATIVE
HUMIDITY
5^*7.
TO TARGET
1ISTANCE (cm)
\ 0 \v\
\ 0
CURE
TIME (•) TEMP (°C
475 f
7 F
FOIL
NUMBERS
.
to
*
VERTICAL
COVERAGE (%)
n-5
n
I'o
LN (n)
DRY
. M
FLUID MASS FLOW RATE DETERMINATION
INIT. (9) FINAL (9) A (9) TIHE(l) RATE (?/•}
ni.S
ral
«o
Data Collected byt
CONVEYOR
SPEED (Ol/*)
HEIGHT %
SOLIDS
47,5
M7-5
M7-S
NET DRT
SOLIDS (g)
TRANSFER
ETFICIENCY (%)
Data Checked byt
*e=54.68/bxd\
Vaxc/
or e = 15,833 / d \
\cxfj
5,53
-------
^
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
t f
gto ^
\ <>
to
to
to
to
to
PRESSURE Kt GUN JKP«)
FLUID ATOM. MR
<%<>,
tl\ (v;i
-Hutu 11 IIS in ij
in ijjiTn UJ'B iiiiiin
305
^
OPERATING
VOLTAGE (kV) RESIST. (Hfl)
H/A
>
/
^//^
N
/
TEMPERATURE «°C)
AMBIENT FLUID
1^ F
•HM=
25°?^
BOOTH AIR
RATE (on/l)
l°° W
loo^py/
RELATIVE
HUHIDITT
ISl,
&%
OTN TO TARGET
DISTANCE «cm)
|0 i rv
I0i^
10m
IO,H
{0^
\ 0 «(A
CURE
TIME(«)TEMP J°C)
/9^
i1^^
S^^v
IS*
\s»^
1SVV
^7<7/
4^^r
M7^lr
ir^r
MTCf
^^^f
FOIL
NUMBERS
H
to
to
to
to
to
VERTICAL
COVERAGE g«|
1^5
MUI (ail
-------
Date;
DATA SHEET 4
oy;
w
I
Ho.
4
Gun Ho.
/&?!>
'390
Poll Ro.'i
7^7-
•t Qun
Plutd/fl Fluid
T)°F 2S^
Booth Air
tat* «e»/«)
Humidity
Control Panel Preaeuree
Supply Atom/' , \ Paint
Air Al
1VP-
inei rrei
lrP
Conveyor Speed
-------
Date:
DATA SHEET 4
Data By!
Data Checked By:
Run
No.
Gun No.
Foil *o.«»
Pmcur* at Gun
Fluid Hto». »lr
Operating
V»U«q«(fcv)
Tvifxr
cttirc
Booth* C> Fluid
Booth Kir
IUt« (c»/
taUtlv*
Hualdlty
Control Pan
Supply HtcMi
a>ur«*
Paint
Kir
7
_£
Bod
f
rob
812
rf
(L
60
)TF
rco
300
I
I—"
IN)
Run
No.
7
/o
'1
rhicSt.
•«; Pi
(•111)
4V
"7
•9
"
H»f Ms*
S
Convayor tp««d
Nt. %
Solid*
Met Dry
Solid, (at
?/
Efflol«ncf
36,63
67Z
-------
Date:
DATA SHEET 4
Data Checked Byt
w
i
*—*
CO
No.
1
n-
Gun Ho.
Poll No.'*
%7~
¥r
4&J
rlula
Ate*. Air
/o
Z
13
Operating
Volt*a«(kVl Currant ( .1
^
(
icl6A
Booth* C>Pluld
Booth Kir
IUt* <€•/•)
„
Itclatlv*
Humidity
/
IKt-
Mn«l r
Control
Supply Atoa.
Mr Air
d
»ur«*
Flint
Pot
/C3/
/O/
Air
Plcw
Hun
Ha.
V«rt.
niick.
«t riu
(•1UI
Fluid MM Plow
*••• <9> TlM («) IUt« (ql«)
CM«*yor Bp**d
(ci./.)
Nt. «
Solid*
H«t Dry
Solid* (g)
Tr*n«f«r Efficiency
C»t
'i
Ifr.
rt.to
t
•1
HL
.£>%
31.15
.7
In
tf
37
34,
/•D
15 3,
3 1,72
7
If 3
-------
Data
Date:
DATA SHEET 4
Data Checked Byt
i
t—<
4^
Ho.
S
f/
III,
Y
Gun Ho,
foil *>.'•
797
/
Fraaaur* at Cum " •
Mula Aton. Mr
!*>
2.0
Operating
Volt»ga(kV) Qirr.i>t<
Booth C} fluid
•ootb
lut*
\
talatlv*
HuBldlty
7t%
7^/
Control r«n«l
•upply Atoa.
Mr Mr
ft
Paint
Pot
i^-/
uR
Mr
flow .
(.Cf/ly
Vart.
tat film
H*Ur
Plov Rat*
lgl«l
ttotal
""•
ill Kata »ql«|
Ganvcyor
Mt. \
Solid!
Itot Dry
Solid, (at
Tr«n«f«E
*-10*tS
50%
6.
In
h
ri
'1
12-
-7
r
-------
DATA SHEET 4
ey:
w
i
Hun
Ho.
'//
Gun Ho.
/&b
roil HO.-*
fo/o-
/o/f
Cpsi)
Pressure «t Cm
Plula Man. Mr
^
~—t ~-r *~
^1 £. >
>if
Operating
Uoltaqe (kV) Curnnt ( .1
fi/1*
V/A
Teap«gatar«
Booth' C>Pluld
^6
Booth Mr
lut* (CH/I)
/o^/5
talatlv*
Humidity
7^/6
Control Hiicl rr«*«ur*»
Supply At on. Paint
Mr Mr Pot
A>O
^
°%,j
Mr
Plow
tocjprt.
#f
tun
Ho.
Vert.
niick.
at rii
(•ill)
Plow Rat*
-J«l*
Pluld
Tot«l
Plow
m» (e) Hate (gla)
Conveyor lp**d
(CM/I)
Nt. %
Solid*
IUt Dry
Bolld. !„)
Transfer Efficiency
7
-------
t-'
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUM (kPa)
FLUID ATOM.AIR
linn in'* BM
»T jrtifjn Mfn murim
OPERATING
VOLTKCB (HV) RESIST. (Ml)
TEMPERATURE «»CS
AMBIENT FLUID
BOOTH AIR
RATE (c*/i)
RELATIVE
HUMIDITY
OTN TO TARGET
DISTANCE (cm)
CURE
TIME (•) TEMP (°C)
-1 to
72 F
e
5 0
-?».
557
"
"
30
72 F.
-i
36O
»
10
Zl |)S)
567
FOIL
NUMBERS
VERTICAL
COVERAGE (*$
FILM 6
36,7
ns.'
21
•286
472
-75 .
0
Ifjfe.'S
42,70
75.10
4.10
46.04
-73-
<:
-------
^>vc
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT CUM (KP«)
FLUID ATOM. AIR
OPERATING
VOLTAGE
FOIL
NUMBERS
f-r
1 C,
to
to
to
to
to
VERTICAL
COVERAGE (%)
^
EH* {cm}
o,c*
•:
FLUID MASS FLON RATE DETERMINATION
INIT. f9) FINAL 19) A (9) TIME(s) RATE (g/«)
3T?
M/A
\
/
Kfl.S
•22>.tS
/|5.^
CONVEYOR
SPEED (o«/«)
/0|/r/^
\
I '
f
1
HEIGHT %
SOLID!
M"7 *^
•
(
NET DRY
SOLIDS (9)
A •? -? j
TRANSFER
EFFICIENCY («)
-7^3
6.7-0
Data Collected byt
Data Checked byt
*e = 54
.68/bxd\
Vaxc /
or e=15,833 / d \
\cxfj
-------
i
(— •
co
Date:
Mb
DATA SHEET 4
Data
Data Checked By:
r
-------
C
w
I
I—t
vo
C
/I/1
V
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN OtP«)
FLUID ATOM. AIR
4^
*
V'
IIUMUHU MS "
Ht.PLUIDN
3^5
OPERATING
VOLTAGE (fcV» RESIST. (KO)
250
TEMPERATURE (°C)
AMBIENT FLUID
71 SF
2S5C
ZS.5C
25
BOOTH AIR
RATE
10 ^
RELATIVE
HUMIDITY
TO TARGET
DISTANCE (en)
(O
.vv
\0 it\
\O i^
CURE
riME(«)TEMP
15
*/*
FOIL
NUMBERS
VERTICAL
COVERAGE («)
FILM (CB)
/WT\ CRT
FLUID MASS FLOW RATE DETERMINATION
INIT. («) FINAL (9) A (g) TIME(|) RATE (g/n)
351
N/A
"50
25 S
CONVEYOR
SPEED (c«/>)
HEIGHT %
SOLIDS
Ml-5
47,5
Ml. 5*
NET DRY
SOLIDS (9)
, 70
TRANSFER
EFFICIENOf (%)
-77. ofc
Data Collected byf
Data Checked byt
*e=54.68/bxd\
\axc/
or e = 15,833
-------
v -,"••',!
KX^" ^i:
I DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN (KPa)
FLUID ATOM. MR
OPERATINC £»'-'
VOLTAGE (ItV) RESIST. (Ml)
TEHPEJWTOME !°C|
AMBIENT FLUID
BOOTH AIR
RATE (m/>)
RELATIVE
HUMIDITY
OTM TO TARGET
DISTANCE (cm)
CURE
TIME (•) TEMP (°C
c
to
11.
73
70*%
r *i
w
I
ro
CD
Xi
uc,
73-$ F
27.0C
FOIL
NUMBERS
VERTSCM,
COVERAGE «%l
DRT
FLUID MASS F&OH RATE DETERMINATION
INIT. (9) FINAL $9) A Kg) TIME(c) RATE
OONVETOR
SPEED (»/•)
HEIGHT %
SOLIDS
MET DRT
SOLIDS ig
TRANSFER
EFFICIENOf (%)
3MS
P
3/5
f-i
0,
47.5-
IfeSZ
47.5
31. 1,/
25.6
-70
n .5
BSS
Data Collected byi
Data Checked byt
*e=54.68/bxd\
\ a x c /
or e = 15,8 3 3 /_d \
(cxfj
-------
Date;
Data Checked Byt
w
i
f r
Run
No.
'1,
Gun No.
13*
roll uo.'«
%$&
Fluid At on. Air
•*W
//
Operating
Voltta* (kV) Currant I .1
7f
-W
Booth1 C) fluid
£C,£
Booth Air
Kat« (CK/«>
X^/J
Hualdlty
**
X l) *lr
Control ranal rr«i»ur«« / OJ// plow
Supply Ato.. F.lntt^>1 -
Air Air Pot I "c>-l
/oo
^
^^
c^%?
tan
No.
't
V«rt.
99*«r«o«
Thick.
Mt riu
(•!!•)
..... . fluid Ha«« Mow
H«Ur totml
l".V* *"• '«' *»« <•» tat* (oil)
^r
151-0
o?'?'. 0
'' /
Conveyor fp*«d
(a/it
10. IG,
Mt. %
Solid*
H^i^S
m»t Dry
Solid, (a)
37-7
Trui«f*r Efflelmcy
(«l
77,^3
-------
Date:
Data By; ~
DATA SHEET 4
Data Checked By:
i
ro
Run
No.
1
i
ni-
Oun No.
Poll Mo.**
#*?
loo?
Pr«i«ur* »t Gun
rluld »to». Mr
/o
(0
,0
•z/.r
to
VH>U*g«(liVl Current<
Booth* C! fluid
ZW
-------
Date:
DATA SHEET 4
Data Checked By:
w
ho
Co
Ho.
fc
roll *>.«•
**<
^32.-
riiM"
/o
/o
to
On
Ktm. »lr
Opvratln)
VoU»«(liV) currant
7-
-rf-
J*
*
•oath1 C>rluld
Zl
it
(6
Booth Air
lUto
flow
tlm* M tat« |gl»»
folU*
Itot Dry
•olid, U
Tnnifar efficiency
'7
43
9.80
7
-
1
7
74-10
5" 37
7
-------
Date:
DATA SHEET 4
Data By;
Data Checked By:
w
I
ho
7
r
4-
JO
II
iZ
Gun Ho.
,36?
roll MO.**
1T4-*
miiur* *t Gw
rlula >to». Mr'
"?^7
/o
Opcrktlnf
Volttq«(liV) Curr«n
1/
c!
riuia
175'^-
Ut/d
•ooth Mr
IUt« (cm/m)
Ak
Ntuddlty
Control twi«&
*t
loo
|OO
Paint
rot
7,6
Kir
rio-
Ucf»>
500
X
u1
rhiek
Kat
iMlll)
...
Fluid Itacc Flow
Tot'1
Ctmnyor tp«*d
•t. «
Sollal
N«t Dry
Solid. 1.)
Tir«ntf«r ZffloUncy
7
ZIO,
•7
41.50
21,0
47- 3 z
.-7
319.
70,7^
11
0 .
//I/I3
•7
<-
-------
\*> DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUM (RPft)
rUlIO ATOH.AIR
RUIA'l'lNU M*S
HLJIUIB N/^U tCll
OPERATING
VOLTAGE IkV) DESIST. (MO)
TCMPEMTUM! (°C)
AMBIENT FLUID
BOOTH AIM
RATE (CM/*)
RELATIVE
HUHIDITT
SUN TO TARGET
1ISTANCE (
CURE
riNE(B)TEHP
IS
Q5S
•*"y*-i i
73
10
i*
10
3*1
->, 2
25,^6
(Al*
10
tol
I
K)
FOIL
NUMBERS
VERTICAL
COVERAGE |%)
DR»
FLUID MASS FLOW RATE DETERMINATION
INIT. (9) FINAL ig) A (9) TIME(*) RATE
CDNVETOR
SPEED (c»/«)
WEIGHT \
SOLIDS
NET DRT
SOLIDS (9)
TRANSFER
EFFICIENCT (%)
/A
'si .3
5,
'V
s
80.^0
0-C.
5.573/5
3 £-2*
73 -
Data Collected byi ft
Data Checked byi
or
>54.l
15,833 /_d_\
\cxfj
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE ATS, GUM (kp«)
s ATOM. AIR
ROTATING RPS
MI. FLUID N/O rUJID
OPERATING
VOLTAGE (KV) RESIST. (HO)
TEMPERATURE («>Cj
AMBIENT
BOOTH AIM
RATE (CB/I)
RELATIVE
HUMIDITY
OTH TO TARGET
31 STANCE (en)
CURE
riME(»)TEHP
1&2.5
H/A
N/A
H
fl
N/A
MA
ioo c» A-
iro
V
ho
CTv
FOIL
NUMBERS
VERTICAL
COVERAGE f%)
(cat
DOT
FLUID MASS ItOH RATE DETERMINATION
(9) A (q» TIMEC*} RATE (
-------
SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN (kPa)
FLUID ATOH.AIR
ROTATING M>S
HI. FLUID N/0 FLUID
OPERATING
VOLTAGE (kV) RESIST. (Ml)
TEMPERATURE (°C)
AMBIENT FLUID
BOOTH AIR
RATE icm/u)
RELATIVE
miMIDITT
SUN TO TARGET
)I STANCE (on)
CURE
riME(B)TEHP J°C
w
\'\ft
"fi
ML
j
Slit-
Z|.0°C
(/>!(, 3Z
-oz
1
1/0° Ib
FOIL
NUMBERS
VERTICAL
COVERAGE |%)
(CM)
DRT
FLUID MASS FLOM RATE DETERMINATION
4«l FINAL (9) A (9) TIME(c) RATE (g/»)
OONVETOR
SPEED Cc«/«)
WEIGHT %
SOLIDS
NET DRT
SOLIDS (9)
TRANSFER
EFFICIENCY i\)
11%
l»o,4
IG-53
7^° 7?
40
toO/
5-7,25"
33 -i
57.25"
1 10
14.3^5
MM
Data Collected
Data Checked byt
*e=54.68/bxd\
Vaxc/
or e=15,833
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT GUN IKP«»
ATOM,A1R
ROTATING M>S
WI. FLUID H/0 FLUID
OPERATING
VOLTAGE JkV) RESIST. CM))
TEMPERATURE t°C)
WffllEMT rtUID
BOOTH AIR
RATE fa*/*)
RELATIVE
HUMIDITY
GUN TO TARGET
3ISTANCZ (cm)
CURE
riME(»)TEMP
to
A///)
A/ /A
2-
V
VOv— )
to
to
l&S'C
to
to
to
M
l
NJ
CO
FOIL
NUMBERS
VERTICM-
OOVERAGe ««
DRT
MASS n-ow RATE DETERMINATION
FINAL Igl A IQ) TIME(l) KATE fg/a
CONVEYOR
SPEED (cm/u\
HEIGHT %
SOLIDS
NET DOT
SOLIDS (9)
TRANSFER
EFFICIENCY (%)
to
n%
-ZO.S3
II 4
( .
20 .q 7
to
15
- ''
T\
44*?
to
5^.40
. 40
Data Collected by 8
Data Checked byi
*e=54.68/bxd\
Vaxc/
or e=15#833
cxf
-------
DATA SHEET 4 - OPERATING CONDITIONS AMD CALCULATIONS
FOIL
NUMBERS
PRESSURE AT CUM fKPa)
FLUID ^ ) ATOM. MR
ROTATING UPS
MI. FLUID N/0 FLUID
OPERATING
VOtTACB (kV) RESIST. (HO)
TEMPERATURE C°C)
AMBIENT FLUID
BOOTH AIR
RATE (c*/«)
RELATIVE
HUHIDITT
OTH TO TARGET
)ISTANCB (cm)
CURE
Kijjft
/oop/s
?/i
to
wo
to
^
to
K)
vO
FOIL
NUMBERS
VERTICAL
COVERAGE \\\
Ceml
DRT
' rtUIO MASS FLON RATE DETERMINAnOM
llHTi (y) FINAL (9) A (9) TIME(l) RATE (9/1)
COHVETOR
SPEED (oi/*>
WEIGHT %
SOLIDS
NET DRT
SOLIDS (9)
TRANSFER
CFFICIENCT (%)
7
to
^
zi.
S7.40
Zl -Ob
to
/q-,3
to
*7,40
to
x/J
5-7.40
45.2%
to
( '
71.1
Data Collected
Data Checked byt
*e=54.68/bxd\
\
or e = 15,833
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT CUM ) RATE <9/«)
OONVETOR
SFEED (
to
£
57-^0
11%
57^0
.OS
to
13^-0
O
to
?/)
Stot
740
20-70
•42 .
Data Collected byi,
Data Checked byi
*e»54.68
\axc/
or e-15,833/_d_\
\cxfj
-------
DATA SHEET 4 - OPERATING CONDITIONS AND CALCULATIONS
FOIL
NUMBERS
PRESSURE AT CUM (KP«)
FLUID ATOM. MR
ROTATING MS
MI.PLUID N/0 FLUID
OPERATING
VDLTACB (fcV) RESIST. (MO)
TEMPERATURE |«C)
MtBIEHT FLUID
Boom MM
RATE (a»/«)
RELATIVE
HUMIDITY
OTW TO TARGET
31 STANCE (
CURE
FIME (•) TEMP |°C
to
ftjft
to
375-
to
/wo
to
3/.3L,
to
I&O
to
w
Co
/
I
FOIL
NUMBERS
VERTICAL
COVERAGE ft)
FILM (ca)
.^'
PITT. (
FLUID MASS FLOW RATE DETERMINATION
. (9) FINAL Ig) A (9) TIMEU) RATE <9/«)
COKVETOR
SPEED icm/»}
WEIGHT %
SOLIDS
NET DRT
SOLIDS {9)
TRANSFER
EFFICIENCT (»)
.1
A
Zl.
to
,
-------
DATA SHEET 4 - OPERATING CONDITIONS AMD CALCULATIONS
FOIL
NUMBERS
PRESSURE AT CUM tKP«)
rLUID/pt'AATOH.AIR
ROTATING M>B
MX.rLUID M/O rLUID
OPERATING
VOLTAGE «*V) RESIST. (Ml)
TEHPEAATVRB (°c)
AMBIENT rUlID
BOOTH AIR
RATE (CM/I)
RELATIVE
HUHIDITT
CUM TO TARGET
3ISTANCE (cm)
riHE(«)TEKP
to
/i/A
to
Zl'C,
to
to
t-i/u 21'/
to
27/13
°C
to
w
U)
NJ
*=-
roit
NtJMBERS
VERTICAL
. (oil
IIHTt ty» riMAfc
MTB DKTERNtMATION
TIMEjij MTB
OONVETOR
SPEED (cV«)
WEIGHT %
SOLIDS
NET DRT
SOLIDS (9)
TRANSFER
ETTICIENCT (%)
to
•zs .
to
to
/
-------
LABORATORY 2
E-33
-------
T7
A&
DATA SHEET 4
3
,,5" -'30
r
f 3 /7y? —
• "I />U
)(,*>•*-
roil »».'»
"ul«
•t om
Mat. Mr
WUm<'lYL.P"T»"*t
Al
Booth' C'rluld
•oottt
tat*
w
<2
3
f
n
n
1
7
Data
•j 0,3 ?
Control rwi«t rnilur**
•upply »to«. r«lnt
Mr Mr rot
I
jJ/A
ff
Data
By:
Mr
rlM
Uctal
44^
©™5 4• i
, 6 8/bxd\
Vaxc/
or e-15 ,833/_d \
I^TTTJ
-------
J0\ SO
O It?
DATA SHEET 4
roll *».••
rraitur* at Oun
rlula >te». Mr
VolttQ.Uvl _0irr.ntt
Booth' C>rluld
Itet*
Milctlv*
Hualdtty
Control Vwtcl rr*««ur««
•upply MOB. Mint
Mr Mr »ot
Uctal
DIM
tj/L
1
MM
"».
V«rt.
t
miok.
t rl
(•1U)
_.,._ rl«M HIM rlev
-l ' Total
Data By:
•7
A
ff
Data Checked By:
Ht. %
•olid*
Not tey
•olid* lot
30.C-3
Tractor effictency
1*1
*e-54.68/bxd\
Vaxc/
or e-15(833/_d_\
Icxf)
-------
Date:
DATA SHEET 4
Gun No.
Mo.9*
»t Oun
UlliS
VoU*i«(l.vl _0.rr«ntt
•ooth
•ootk Wr
IUt« «a«/
Humidity
Control F*n«l rra»ur««
•upply ktaa. r*ln£
Ur
MOM
Uct.1
w
u>
cr>
f
ton
¥«rt.
niick.
t MI
(•tui
In 2*1
Data By:
da
W
\
7%
Writs-/
/<£>
%& -/O O
toy*
'U
ri»ia
if/.a
TUM (•» tat* (ql«)
///c.
-k-6*
.<**
O
M. «
Sol MB
N«t Dry
•olldi (a)
/3.1?
'3 .
-. .1 7
*e-54.68/bxd\
Vaxc/
IData Checked By:
or e-15,833
-------
DATA SHEET 4
7
roll *>.'•
Mull.
•t Oun
»to». Ut^
0|»»tlnfl
VoltH.M.Vl Qirr.i.M
\IL
V
•oath' C)fluld
•ooth Mr
tat* (a*/*)
talitlv*
HuBidttr
Central r*n«l m«iur««
•upply Moa. f»lnt
Mi Wr rot
Mr
rlow
Uch)
-ffo
//.
m
Lo
^J
V«rt.
feMMO*
Itilek.
•tat rtla
(•1UI
DataByj^^
rlov MM
'•'•t
ttetal
&
(
Flon
Tta. I.I
tat*
Mt. «
•olid*
Met try
•olldi (at
frwufcr crricUncr
1*1
Data Checked By:
*e-54.6B/bxd\
Vaxc/
or e-15,833 / d \
I f^ U f I
-------
Date:
/ r
DATA SHEET 4
•35
/J
Oun Ho.
roll »>.••
9-91,
M
Lo
do
v.rt.
mick.
t rii
(•11*1
H
.I/
Data
"uiq
•t Oun
»to». *i«
VQU^tlliVi _0irr.ntt
"74
Of
Booth* CkPluld
15.6
H5
T65
•ootlk Air
Mt«
-------
LABORATORY 3
E-39
-------
Date:
w
J>
o
DATA SHEET 4
f.
roil *»."•
Hula »t<». Air
/U/t.
VttUHltW .^"T"*!
5"
*«rt.
nitck.
*t ri
{•11*1
S"
5-75
122.7
Data By:
rVil-r»
•oath1 C> fluid
33. a.
222
Bootll *i«'
Central r*n«l »r««»ur»e
•upply Moa. r*lnt
Mr Mr rot
c
M* %
•oltde
>*it Dry
•oll4« fat
IF.24-
Aft
Air
ri
-------
Date:
DATA SHEET 4
No.
?
Qon M>.
t*i9
roll ••.••
n^g
riula Atoa. Air
X C' ^ A*"
/ ^\ ^-^p
•
^
*
••
VDU4a«lkVl Oirnntl .]
•
,
fr
•
Ttapogctara
•ooth1 C|fUld
•ootk Alv
tot* lfW»)
-------
Date;
DATA SHEET 4
W
l
-f>
K>
-60
-72.
73-7Z
t Oun
rlaia Rtcn. Air
*«*•
10.0
if
tooth1 Clrlold
23?
25.3
Seotk Mir
tat*
5*5-. *S>
5V
Control r«n*l »r»»*«3i»e
•upplf »to«. ralnt
Wt Ur Pot
93S
22 .
22 ^j
Ur
r»«
Ucfml
V«rt.
mick.
*•«: Fit*
37
2.7
27
27
7
Total
"••• '«»
353-
hi
tat^
l.l
s:*?
c
M. %
•olidc
b
.tot Dry
•olidc (
Transfer CfficlMcy
ttl
f 8
*e-54.68(bxd\
Data By;
Vaxc/
-------
Date:
DATA SHEET 4
Lo
Pell *».'•
run
•t OUM
ttam. Mr
Voit4a«fhVl Dirnntl .1
Booth* C>PtaU
taUtin
Ontiel Van*! m**ur*«
•upply Horn. taint
Mr Mr rot
2.2.5"
Mr
rlow
toctal
Ifiich.
ri
total
C
M. %
•olldi
47.35-
b
••t Her
•olldi !•!
*e-54.68/bxd
Data By:
-------
Datet
DATA SHEET 4
roll *».••
iMt
*«<»• **'
VPitM.ttVl Current j
c'
ri»id
•ootn Mr
lute
Humidity
Control rwi*l
•upply Iktoa.
M( Air
r«lnt
rat
Air
no*
Ucf»l
25.3
5V
/7
25,?
fi-
2.5.
37
10 ft
41
5?
fl
««rt.
miek.
5"
J//
/57./
c
Nto «
h
H*t twy
•olid* (•!
e
Trmifcr
111
£-3.
33.7^
fte-54.68
Data Bys
axc
-------
LABORATORY 4
E-45
-------
Date;
CT.
DATA SHEET 4
,5"
*<*&?
roll *>«
30
Jp~
•t Out
Hula »t<». Air
/ft*
/
• fl> A
Booth* C'rl»W
•ooth *ijf
tat* lc»/«)
Control Pana
•upply ktoa.
Kir Ur
Vtort.
•*tnt*a>
ff.#
iMek.
*t rii
U1U)
2
*ur """
««•«*• ^'1,1
Mow
TlM (•)
7.7*
21 /.
c
Mo %
•olidc
b
i^t DTf
•olid* (a)
Data By:
r*l«t
»ot
?•
^
Ur
Mo.
Uctal
*e-54.68/bxd\
\H^/
-------
Datet
DATA SHEET 4
1
Ho*
7
ttm
7
OHA Htt*
(0fr AcA {
Vttrt.
krv*r40«
«»
^M=l
n>ick.
y
rraasiica at <*•»
Mula itoa. Mr
•
'tfft
' :
VolMaalkVI currant I .1
,
f}fi
•
?***' Total
IJT.V* "••• '•» »!»• «a» Data3!*!*!
-f«V
X/^-7
S.of
tooth' c>ruu
«3
^.({
A
GDfAAtyOV •pft'td
^^T dVV • C^
•ootk Mr
tata (c^/a)
^^^K
•
1
c
•t. «
•ollda
•5u./&
talatlva
Mwldlty
f5:
-------
Date;
w
4>
oo
DATA SHEET 4
A*
/z
/^r
11
roil ••.••
.run.
•t Qu»
A&Bj-AI
lift -
curr»ntt
A/
•oath* Cl Fluid
7?.$
Mr
Humidity
-7-f
Control r*n»l rm*iur*«
Supply «t
-------
Date: /
DATA SHEET 4
foil *».'•
tr«Miu* at Oun
Mala Ato». Air
VoltM.lkV) oirnittl
I rl
ath1 'fluid
Booth »tr
tot* •«/•)
Nwddlty
Oontret r«n«l Vn«*nn«
•upply MOB. r»»nt
Hi* Mr rot
Ucfml
/£>
9.7
7-f
?- ^
If
. o
7/
9.7
227 221
7/
n
f
It
«trt.
M ''
mtck.
»t riu
(•ltd
i?
S'
N.t.r
r*~.r"*
Total
l*1,t
39,1 #
as. ?/
7
c
«t. »
•olid*
4 $.7 7
b
••t Dry
tot Mi (•>
/1.03
Data By:
111
'tt./'f
7^-23
71.
*e»54.68/bxd\
-------
LABORATORY 5
E-50
-------
w
I
DATA SHEET 4
5
/I
ft
ft
It
»'-'• .•*«-• **'
nn
Ai/?
Oiri.Htl .1
•oath1 c'rl«U
u*
»•*• <«*/•
r
itilek.
ot rili
iBlUI
g
c
•t. %
•oIU*
••Utiv*
Control *•*•! PvM*«ra«
upplf MOM. Mint
Me Ur
fcU
ri
-------
Date 8
r
DATA SHEET 4
Data By;
Data Checked By:
Ln .
No \
Sf»
Ho.
'f
IT
ff
c(s?o
roll MO.'
vmiwr* «t Am
rlula MOB. »lr
V«U*«•
O*' _
t« w»>n
{0. BO
c.
M. «
•olid!
••t Dry
»ol|J« M
IffJcUncy
.'. r • • f
-------
w
U1
U)
^7
DATA SHEET 4
ft,
17
tf
01326
rat •».••
tol-16
•t 9m
»'"'• .**«•• ***
11
«*rt.
nttck.
.t rll
fBlU)
(0
/o
^ By:
,OtTfff*.i.
•eetk1 ClfUI«
1 70
MiUtlv*
Ocmtr*!
•Ufiply MOM.
Me Air
It!
MZ
HZ
r»int
££
t-
u*
n
-------
DATA SHEET 4
/o
17-
«*rt.
niick.
«t riu
yy
r
IP
X
0,/f
c
»7T
•olU*
d
••1 Brf '
•el 14* lit
Data
Data Checked
1*1
tl
X ~t>?.05-
5- U IS"
,6B/bxd\
\axc/
or e-15,633
-------
LABORATORY 6
E-5'5
-------
DATA SHEET 4
en
4
»t
Jill*.
f*
AJ/J
A)/)
Dfl-
•oath9
Mff
Mt.
Onttcl »MM| >l«»»«n«»
^rly MM. *•!•«
Me klr
*>/>
3
5
nilcft.
t ri»
fall*!
"•»•*
c
fT
*>**•
3
d
s
loll«* lat
17.93
.i_a
111
Data Checked Byj
u*
»|M
Uctel
Af/t-
axe/
or ,._-15,833/ d
-------
DATA SHEET 4
w
l_n
^J
f
Mil •».*•
67-7-Z.
M.I. It-. H»
!«*•(•••
IMck.
7
7
7
?>-?*>
By:
•2.0
2.0
-2,6
-2.0
«»U4~tWI OanHH .1
74-
+
^.•tl.' _>UU
•zz.%
?
22$
28.Z
••*• «"^
KkUtlw
•Mltfltf
**•
33,13
b .
r*u>*
IM
U(
ri««
Uctol
/
^7/.
••-54.6B/bxd
or 15,833
-------
m
001
I/
II
fl
ft
f»lt
79-V4
«*rt.
ntick.
S'
3-
7
7
Data By:
DATA SHEET 4
< •*-&
•t On
•*«•• "•
fljfl
0«*t««U
M
XX-
?$.<*
••*•
ItoUtlv*
*
Contral MM| »IMM««C
•upply MOB. *•!•«
Mr Ur »o«
1.1
^'"^Q
<£./
-------
LABORATORY 7
E-59
-------
DATA SHEET 4
w
i
cr>
o
7-/Z
3 /-
*!•*•-
«*rt.
7
7
7
7
thick.
7
7
7
ca By:
ML.
MA
vam««ftVl _Qun*l
227
Me Mr
flaw
9./S
b
to*/*)
2^,73
30.? /
c
lETT
d
•^sr
•olid* !•!
iti
X -
Data Checked By:
or
c»54.68/bxd\
laxc/
-15,833 / d
leir-r
-------
DATA SHEET 4
nil «••'•
i»- air
«*•*•*!»•
OUtMNtl .I
h* ClrUU
Itet* to*/*)
MUtlv*
OantMl »•*•!
•uff-ly MM.
Alt Air
Mint
Uf
flow
toctal
7
Zo
5V.
110
7
2.0
//*
7/
/•z
72-7%
to
ft,
3(0
niick.
*t rlla
33>7
7/
9*1
c
ItTT
•elM*
Brf
Ul
70.
Data Checked By:
I. If
S-i.
U)
(•U\
•e-54.68/bxd\
\17Z)
or 15,83!
-------
DATA SHEET 4
cri
ro
ft
-8*
»»•'•
fa
ntlck.
it ri
22, £
Z£££
•y
Data By:
Mm, *ir.
c>
ruu
"•*• ««^'»
77
77
7
23.
77
Mo*. *•>•«
M( Air tot
-ff
&/LL
6.7?
525
d
5
•el 14* lat
tra»f«r
lit
Jo
u*
MM
Ucfvl
/bxd
Data
or
\axc/
«-15,833/_d_^
-------
LABORATORY 8
E-63
-------
f,"
DATA SHEET 4
w
I
3
r
r*a »*<•'*
2 5-- 36
51-
C8&
MR
/Oft
o«t^tt
73"
TS"
•.t.
ll
It
1
3
7
5
«nrt.
r
7
7
7
7
Itilcft.
»t ri
(•1UI
9
7
"•*•'
i,t f
S2TA
M.I
1.1
15.
15.
3}
»to ttf •>
ill
3L3P
b
M«
W^
Ic^/.l
II
c
ItTT
•otU«
ItoUtl**
42^
•xpplf MOB.
Mr Mr
r*l*t
»ot
\
d
M Crf '
•alMa Ut
21.27
ZI.Sl
zus
tnMf*r imitate*
ftl
n
4-7,1
48 .
u*
ri
-------
J- ^ o "5 l
DATA SHEET 4
n
i
en
en
s
•».••
/o
^o
74
74
l .1
67
&-?
£7
7
/z.
/7
Itiick
f
f
3.00
c
•t. %
totlto
I a By:
Data Checked By: , '
Ontnl
•xppir •*<»
Me Air
4t
t-
t
/a.
Ur
riw
toe hi
e
TrauUr
-10
•e-54.6B/bxd\
vr^/
or -15,833 / d \
17771
-------
DATA SHEET 4
'
tfl
cr>
CTl
'y
'T
'7
ts-fc
«t
£>
10
Jo
lo
fg
/
7
is
*§rt.
/^
t
9
9
ff
¥olt«..IM» (VrpBtt
7Z-
,*iL.
Mo..
Ur
23
u«
C
ifT
d
»•* <*T
•alii* !•!
. -77
^ 3«.
A-
(I7'
Data. a«cked Bv:
•e-S4.69jbxd
Of ««15.Q33/ d
-------
TECHNICAL REPORT DATA
(Please read Inurucnons on the reverse before completing)
1. REPORT NO
EPA-600/2-88-026b
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Development of Proposed Standard Test Method for
Spray Painting Transfer Efficiency; Volume II.
Verification Program
6. REPORT DATE
April 1988
6. PERFORMING ORGANIZATION CODE
7. AUTHOR1S)
K. C. Kennedy
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING OROANIZAT.ON NAME AND ADDRESS
Centec Corporation
11260 Roger Bacon; Drive
Reston, Virginia 22090
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-03-1952
12. SPONSORING, AGENCY NAM? AND ADDRESS
EPA, tiff Ice of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle; Park, NC 27711
13. TYPE OF RE8»OHT «* 9 PEfllOD COVERED
Final: 1/82 - 1/87
14. SPONSORING AGENCY CODE
EPA/600/13
18. SUPPLEMENTARY NOTES
919/541-7633. Volume
AEERL project officer is Charles H. Darvin, Mail,Drop 62b,
ume I describes laboratory development of the method.
16. ABSTRACT
The two-volume report gives results of a program to develop and verify
a standardized spray-painting transfer-efficiency test method. .Both review of the
literature and laboratory research were conducted. Transfer efficiency measure.-
ment methods presently used by industry were .evaluated and compared. The besf
.characteristics of thesejmethods were incorporated into the 'final proposed standard
method. The resulting method was determined to be viable for laboratory evalua-
tions. It still awaits adaptation and verification for production line applications.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS
c. COSATi Field/Group
Pollution
Spray Painting
Tests
Pollution Control
Stationary Sources
Transfer Efficiency
13 B
13H
14 B
8. DISTRIBLH IO.N STATEMENT
Release to £»£
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
185
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form
E-67
------- |