United States                   EPA-600/2-88-Q26a
             Environmental Protection

             A8encv	April 1988	
<&EPA     Research and
             Development
             DEVELOPMENT OF PROPOSED

             STANDARD TEST METHOD FOR

             SPRAY PAINTING TRANSFER EFFICIENCY

             Volume I. Laboratory Development
             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-600/2-88-026a
                               April 1988
               DEVELOPMENT OF
        PROPOSED STANDARD TEST METHOD
                     FOR
     SPRAY PAINTING TRANSFER EFFICIENCY
      VOLUME I.  LABORATORY DEVELOPMENT
                     BY
                K. C. KENNEDY
             CENTEC CORPORATION
           Reston, Virginia  22090
   EPA Contract Number 68-03-1721, Task  2
             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

     This research proqram was initiated with the overall ob-
jective of developing a standardized spray painting transfer
efficiency determination methodology.  A Steering Committee made
up of industry and EPA representatives was assembled to provide
consultation and guidance for the TE program.

     Research into existing TE methods and an exhaustive litera-
ture search were conducted.  TE methodologies identified by this
research were compared and evaluated.  The best characteristics
of these methods were used in developing the standardized
laboratory TE method.

     In initial tests the standardized laboratory method
comprised three major equipment types, two paint types, and
specially designed, spray targets.  Paint was applied to the
targets under rigidly specified conditions.  The amount of
solids deposited on the target was divided by the net solids
sprayed at the target to arrive at transfer efficiency-

     According to ASTM 691-79, the first requirement for the
"existence of a valid, well-written test method [is that the
test method] has been developed in one or more competent
laboratories and has been subjected to a screening procedure or to
ruggedness testing."  To fulfill this requirement,  the test
method was tested at PPG Industries and at Ransburg Electro-
static Equipment.  In each laboratory the TE results were
tightly grouped, exhibiting high precision.  The standard
deviations were 2.5 and 1.8 for PPG and Ransburg,  respectively.

     The differences between TE's for the two tests indicated
that some important factors may not have been adequately
controlled.  A third test, performed at Nordson Corporation, was
designed to intentionally vary those factors suspected of not
having been adequately controlled in the two previous tests; the
objective of this design was to see if those factors were
responsible for the variation in TE's.  The results of the third
test were evaluated for their relative impact on TE.  Six
factors were determined to have a significant effect on TE for
at least one spray gun type.  The effect of these factors was
enough to have been responsible for the differences between the
first two tests.  Recommendations are made in Section 8 to
revise the test plan to properly specify all significant
factors.
                                11

-------
                            CONTENTS
Abstract	   in
Figures 	    v
Tables	   .Y1.
Abbreviations and Unit Conversions	v]1
Acknowledgments	   ix

     1.  Introduction 	     1
     2.  Literature Search  	     3
     3.  Steering Committee 	     4
     4.  Transfer Efficiency Test Methods . 	     6
              Industrial Test Methods 	     6
              Standardized EPA TE Test Method	     6
              Improvements to Test Procedure	   12
              Draft Transfer Efficiency Test Procedure   .   14
     5.  Phase I Laboratory Test	   35
              Facilities	   35
              Description of Paints	   35.
              Mass Flow Comparison Test	   40
              Test Parameters	   42
              Phase I Test Results	   49
              Transfer Efficiency Test Results  .....   53
              Weight Percent Solids Test Results  ....   61
              Mass Flow Comparison Test Results	   61
     6.  Phase II Laboratory Test	   67
              Facilities	   67
              Description of Paints 	   67
              Test Parameters	   70
              Test Sequence	   70
              Foil Handling Procedures  	   73
              QA/QC Procedures	   73
              Test Results	   73
     7.  Test Comparison	   91
              Effect of Foil Wrap Technique - Vertical
                Cylinder	   91
              Comparison of Interlaboratory Variances .  .   93
              Analysis of Variance for Paint,
                Laboratory	   95
              Test Reproducibility	   98
                              m

-------
                            CONTENTS
                           (Continued)


     8.   Third Laboratory Test  	 „.....«  102
              Facilities  	 .......  1°2
              Description of Paint	  102
              QA/QC Procedures  .......  	 •  102
              Test Design	  104
              Test Parameters 	  .....  106
              Test Sequence ..........«••••  106
              Solvent-Only Run  .........«•••  H7
              Data Analysis ..........	H7

     9.   Conclusions and Recommendations  ........  124

Bibliography  	 .......  	 •  126

Appendix A - Worth Assessment Model of  TE  Test Methods  .  130

Appendix B - Screening Procedure/Multiple  Linear
             Regression ..................  138

-------
                             FIGURES

Number                                                      Pag*

  1   Set-up for paint supply equipment and platform
        scales	   20

  2   Permissible methods for measuring conveyor speed .  .   24

  3   Target configurations for air atomized conventional
        and electrostatic spray guns	   28
  4   Target configuration for high speed bell	   29

  5   Ransburg vertical cylinder wrapping technique  ...   30

  6   Flat panel foil attachment technique 	   31

  7   Foil attachment techniques for vertical cylinder and
        flat panel targets, PPG test,  September 1982 ...   45

  8   Comparison of vertical cylinder  and semitubular
        target configurations	   57

  9   Foil attachment techniques for vertical cylinder
        (VC) and flat panel (FP) targets proposed for
        Ransburg test	   74

-------
                             TABLES

Number                                                      ?*21

  1   Summary of Measurement Methodologies for Paint
        Spray Transfer Efficiency 	 ....      7
  2   Summary of Paint Spray Transfer Efficiency Test
      Methods Used by Industry  ..... 	 ...      9
  3   Summary of Paint Spray and Peripheral Equipment
        Specifications	 .  - . .  «	    36
  4   Test Equipment Specifications	  .    37
  5   Measured Parameters and Rated Measurement Accuracy  .    38
  6   Summary of Recorded Paint Specifications ......    39
  7   Nomenclature for Spray Painting Transfer
        Efficiency Tests	    46
  8   Transfer Efficiency Test Sequence at PPG
        September 1982 ..... 	  .......    48
  9   Summary of Transfer Efficiency Test Results  ....    50
 10   Results of Bartlett's Test for Evaluation of Pool-
        ability of Variances at 95% Level of  Confidence   .    52
 11   Summary of Equipment Operating Conditions for Air
        Atomized Electrostatic and Conventional
        Spray Guns .-	    54
 12   Summary of Equipment Operating Conditions for
        High Speed Bell	 . . .  .    55
 13   Summary of TE Test Results for Air Atomized
        Electrostatic Spray Equipment  	  .....    56
 14   Summary of TE Test Results for Air Atomized
        Conventional Spray Equipment ...........    59
 15   Summary of TE Test Results for High Speed
        Bell Coating Equipment ..............    60
 16   Summary of Weight Solids Test Results  .......    62
 17   Test Equipment Specifications for Mass  Flow
        Comparison Tests 	 ...........    63
 18   Paint Specification for Mass Flow Comparison Tests  .    64
 19   Equipment Specifications and Operating  Conditions
        for Mass Flow Comparison Test  ..........    65
 20   Mass Flow Comparison Test Results	    66
 21   Summary of Reported Paint Spray and Peripheral
        Equipment Specifications for Ransburg Test ....    68
                             va

-------
                       TABLES (continued)

Number                                                      Pag6


 22   Summary of Paint Properties  	   69
 23   estimation of Required TE Test Sample Size for
        Future Laboratory Tests	   71
 24   Test Matrix for Phase II Laboratory Tests of
        TE Standard Test Method	   72
 25   Summary of TE Test Results - Ransburg Test	   77
 26   Results of Bartlett's Test for Evaluation of Pool-
        ability of Variances at 95% Level of Confidence   .   78
 27   Transfer Efficiency Data, Ransburg Laboratory Test  .   79
 28   AAE Paint Spray and Peripheral Equipment
        Specifications 	   81
 29   AAC Paint Spray and Peripheral Equipment
        Specifications 	 .....   84
 30   HSB Paint Spray and Peripheral Equipment
        Specifications 	   86
 31   Transfer Efficiency Data	   89
 32   Tests to Determine Effect of Foil Attachment Method
        AAC-67-VC: Tape Wrap'Method	   92
 33   Test to Determine Effect of Foil Attachment Method
        AAC-67-VC:  Crimp Method 	   92
 34   Comparison of Variances for Tests at each Laboratory   94
 35   Analysis of Variance - Transfer Efficiency 	   95
 36   ANOVA Results for Paint and  Laboratory  	   97
 37   Transfer Efficiency Comparison 	   98
 38   Paint Specifications	•	103
 39   AOAC Screening Test Design	  105
 40   Air Atomized Electrostatic Test Matrix and Results
        (Nordson)	107
 41   High Speed Bell Test Matrix and Results  (Nordson)   .  108
 42   Air Atomized Conventional Test Matrix and Results
        (Nordson)	109
 43   Nordson Test Equipment Specifications  	  110
 44   AAE Paint Spray and Peripheral Equipment
        Specifications (Nordson) 	  Ill
 45   AAE Equipment Operating Conditions  (Nordson) ....  112
 46   HSB Equipment Operating Conditions  (Nordson) ....  113
 47   HSB Paint Spray and Peripheral Equipment
        Specifications (Nordson) 	  „  114
 48   AAC Equipment Operating Conditions  (Nordson) ....  115
 49   AAC Paint Spray and Peripheral Equipment
        Specifications (Nordson) 	  116
 50   Screening Experiment Analysis Tabulated Values of
        Variance and Contrasts 	  118
 51   ANOVA Results for Screening Test	119
 52   Regression Models Derived from the Screening Tests  .  121
 53   Predicted Transfer Efficiency Results
        (PPG and Ransburg)	  123

                              vii

-------
           LIST OF ABBREVIATIONS AND UNIT CONVERSIONS

ABBREVIATIONS
AOAC
ASTM
AAC
AAE
EPA
FP
HSB
QA/QC
TE
VC
VOC
Association of Official Analytical Chemists
American Society for Testing and Materials
air atomized conventional paint spray equipment
air atomized electrostatic paint spray equipment
United States Environmental Protection Agency
flat panel (target configuration)
high speed bell paint spray equipment
quality assurance/quality control
transfer efficiency
vertical cylinder (target configuration)
volatile organic compounds
UNIT CONVERSIONS

To go from
cm
g
kg
kg/L
kPa
L
m
m
m/s
mVs
rps
s
To

op
in
Ib
Ib
Ib/gal

gal
ft
mils
ft/min (fpm)
ftvmin
rpm
min
                                   Multiply by
                                   1,
                                   2,
                                   0,
                                   2,
                                   8,
                                   0,
                                   0,
                                   3,
                                   3,
                                 196,
                                2118,
                                   0,
                                  60
8°C + 32
54
0022
204
328
145
264
281
937 x 10
86
8
017
kPa -14.7

-------
                         ACKNOWLEDGMENTS
     The contributions of PPG Industries, Ransburg Electrostatic
Equipment, and Nordson Corporation are gratefully acknowledged.
These companies donated laboratory facilities, test equipment
and supplies, and provided technical support for the tests
described herein.

     The Spray Painting Transfer Efficiency Steering Committee
has provided considerable help and constructive suggestions for
this project.  This committee's contributions have been
invaluable.
                               IX

-------
                            SECTION  1

                          INTRODUCTION


     Spray Painting Transfer Efficiency  (TE).  is  a measurement of
how much paint actually coats a surface  compared with  the  total
paint available to coat that surface.  Historically,  the  spray
painting industry has developed its  own  methods  for  determining
TE.  These methods vary from company to  company  and  in their
objectives.  The measurements are used by  the  coatings industry
to optimize on-line spraying; they are used by manufacturers to
develop more efficient spray equipment;  and they are  used  to
minimize losses throughout  the  industry.   More recently the need
to determine TE has a new aspect:  TE can  be  used  to  quantify
emissions from these sources.

     The U.S. Environmental Protection Agency (EPA)  has been
charged by Congress via the Clean Air Act  to  quantify and
control emissions of volatile organic compounds  (VOC).  Acting
in response to this congressional edict, EPA  has designed  a
program to standardize TE test  methods.  Phase I of  this  effort
examined the TE methods currently used.  Based on  these methods,
a  standardized procedure was proposed and  tested at  PPG Industries
in September 1982.  Three types of spray painting  equipment and
two  types of paint were tested.  The results  are documented in
Section 5.  The -PPG test established  the test method as a viable laboratory
procedure, with standard deviation of 2.5.
     A second  laboratory  test  was  performed  in March 1983, at
Ransburg Corporation  to further  develop  the  proposed test method
(Phase II).  The same paints and equipment configurations were
tested.  The results  of these  tests  are  included  in Section 6.
The standard deviation of the  second test was  1.8.

     The TE's  varied  considerably  between laboratories.  Analy-
sis of the test data  from Phase  I  and Phase  II showed some
factors were not adequately controlled from  test  to test.  (The
data comparison and explanation  is made  in Section  7 of this
report.)

-------
     A third  laboratory  test  was  conducted at Nordson Corporation
as a screening  procedure  to establish the relative effects of
previously  uncontrolled  factors.   In the screening test each of
these factors was  intentionally varied in such a way as to
facilitate  evaluation  of  the  effect  of each factor.  The evalua-
tion found  six  factors to have significant effects on TE for at
least one of  the  three spray  gun  types tested;

     •  Booth air  rate (at target plane), fpm

     •  Shaping air,  psig

     •  Atomizing  air, psig

     •  Voltage at tip,  kv

     •  Paint discharge  technique

     •  Paint mass flow,  g/s

     Recommendations  are  made in  Section 8 to revise the draft
test plan to properly  specify all significant factors.  The
draft TE test method  thus developed  is included in Section 4.

-------
                            SECTION 2

                       LITERATURE SEARCH
     An extensive literature search was performed to document
the state of the art in TE determination.  Considerable infor-
mation was collected on current TE methodologies, available
equipment, and necessary measurements.  This information was
collected from vendors, manufacturers, journals, DIALOG (on-line
search in Compendex), EPA Publications Bibliography, and others.
A list of the most useful references is in the Bibliography.

     Considerable research into current TE methods was required
in developing the standardized laboratory method presented in
Section 4.  As part of developing the test procedure, appro-
priate paint formulations and equipment configurations were
evaluated.  Personal visits were made to General Motors,
Ransburg, Nordson, and PPG.

     Some of the people contacted during the literature search
were selected to the Steering Committee for this project.   The
members of the Steering Committee represent divergent experience
and interest in the.TE project.  Their names and affiliations
appear in Section 3.

-------
                            SECTION 3

                       STEERING COMMITTEE
     CF.NTEC assembled a Steering Committee made up of  industry
and EPA representatives to provide consultation and  guidance  for
the TE program.   T"he industrial members of this committee  were
unpaid and contributed their time to the project at  their  own expense

     The Steering Committee was responsible for reviewing  the
technical aspects and results from this project.  The  participa-
tion of a diverse Steering Committee assures the interest  of
each group was recognized within the technical objectives  of  this
project.

     The following people are members of the Spray Painting
Transfer Efficiency Steering Committee:
     Charles H. Darvin
     Physical Scientist
     Industrial Processes Branch
     U.S. EPA,  Air and Energy Engineering
       Research Laboratory
     Research Triangle Park, NC  27711

     David I. Salman
     U. S. EPA, OAQPS, MD-13
     Research Triangle Park, NC  27711

     Charles M. Rowan, Jr.
     CENTEC Corporation
     11260 Roger Bacon Drive
     Reston,  VA  22090

     Gregory A. Conner
     Davidson Rubber Division
     Ex-Cell-O  Corporation
     Industrial Park
     Dover,  NH   03820

     John Lipscomb
     Milwaukee  Solvents &
     Chemicals  Corp.
     P- O.Pox 444
     Butler,  WI  53007
EPA Project Officer
Representing EPA
Representing
Project Contractor
Representing Industry,
Coatings Chemist
Representing the
Chemical Coaters
Association

-------
Albert Hungerford
Butler Manufacturing Co.
1020 South Henderson St.
Galesburg, IL  61401

Mike Tersillo
Project Leader
Applications Group
PPG Industries, Inc.
R&D Center, P- O. Box 9
Allison Park, PA  15101

Steven J. Gunsel
Nordson Corpation
555 Jackson Street
P. O. Box 151
Amherst, OH  44001

Robert Lukes
General Electric Co.
Bldg. 35, Room 1117
Appliance Park
Louisville, KY  40225

Eugene A. Praschan
Engineering Group Manager
Fisher Body
Division of General Motors Corp.
30001 Van Dyke
Warren, MI  48909

Al Chasan
Coatings, Paints & Preservation
Branch, Code 2841
Naval Ship Research
and Development Center
Annapolis, MD  21402

E. W. Drum
Director. Environmental Affairs
Ransburg Electrostatic Equipment
3939 West 56th Street
Indianapolis, IN  46254
Representing Industry,
Paint Applicator
Representing Industry,
Paint Manufacturer
Representing Industry,
Spray Equipment
Manufacturer
Representing Industry,
Paint Applicator
Representing Industry/
Paint Applicator
Representing the Navy,
High Performance
Representing Industry,
Spray Equipment
Manufacturer

-------
                            SECTION 4

               TRANSFER EFFICIENCY TEST METHODS
INDUSTRIAL TEST METHODS

     A number of TE test methods are currently used by industry.
In all of these test methods,  three basic measurements are
necessary:

     •  The amount of solids in the paint

     •  The amount of paint sprayed

     •  The amount of solids deposited on the target

     These amounts may be determined by weight or by volume.
Typical methods are shown in Table 1.   The paint may be
deposited on a target, shim, or work piece.   Some of the test
methods are strictly laboratory procedures,  while others are for
production line TE determination.

     CENTEC considered over a dozen test methods in developing
the standardized EPA test method.   A summary of paint spray TE
test methods considered and comments on these methods are
presented in Table 2.

STANDARDIZED EPA TE TEST METHOD

     The EPA test procedure was developed to have the best
qualities of previous methods.  Since  it was desirable to have a
controlled standardized procedure, production line tests were
ruled out.  Production line tests  are  usually made at production
conveyor speeds using foil-coated  dummy targets shaped like the
product to be coated.  These characteristics are all site
specif ic.

     The EPA test procedure was developed for ease of per-
formance in standard industrial spray  paint laboratories.  For
verification tests, three major types  of equipment (air atomized
conventional, air-atomized electrostatic, and high speed bell)
and two types of paint (65 and 52 weight percent solids) were
selected.

-------
 TABLE  1.   SUMMARY  OF  MEASUREMENT METHODOLOGIES  FOR  PAINT SPRAY  TRANSFER  EFFICIENCY
     Paint Solids Content (I)

             By Weight

I.  Measure VOC In nonwaterborne
    paints per AS7M D-2369-81.
2.  Measure VOC and water In
    waterborne paints per ASTM
    D-2369-81 and ASTM D-3792-79
    (or  D-40I7-8I).

3.  Measure density per ASTM
    D-1475-80.
4.   Measure weight solids per
    ASTM  0-1644-81. (2)
        Paint Spray Flow Rate

           By Weight

1.  Gun  nozzle capture (bucket and
    stopwatch).


2.  Mass flow meter (Micro Motion
    type) and time clock.
3.  Load cell or scales for  paint
    mix tank.
4.   Dipstick on or other level mea-
    suring device on paint mix tank
    (convert by density to mass flow).
         Paint Solids on Target

           By Weight

A.  Where target Is product
    I.   Target weight change (load cell
        or scales)

    2.   Foil on target weight change
       (scales).
    3. Shlm(s) on target weight change
       (scales and weight conversion
       via  area ratio)

    4.  Target solvent stripping, dis-
        tillation and weighing.
                                                                                B.  Where  target  Is simulated shape
                                                                                    (flat  plate,  vertical cylinders,
                                                                                    etc.).

                                                                                    I.  Target weight change (load cell
                                                                                        or scales).

                                                                                    2.  Foil  on target weight change
                                                                                        (scales).

                                                                                    3.  Shim  on target weight change
                                                                                        (scales and weight conversion
                                                                                        via area  ratio).

-------
                                                     TABLE 1  (continued)
           Paint Soilds Content (!)

                    By Volume
        Paint Spray Flow Rate

           By Volume
         Paint Solids on Target

            By Volume
       I.  Determine volume fraction of
          paint solids by calculation
          using manufacturer's formu-
          lation.

       2,  Measure volume solids per
          ASTM 0-2697-79.
I.  Dipstick on paint mix tank.
2.   Paint  Inventory change.
A.  Where target  Is  product
    1.  Target coating thickness and
        area.
    2.  Foil  on target  thickness area.
                                                                                            3.  Shlm(s) on target  thickness and
                                                                                                area (convert via  area  ratio).
CD
                                                                                            Where target Is simulated  shape
                                                                                            1,  Target coating thickness and
                                                                                                area.

                                                                                            2.  Foil on target thickness and
                                                                                                area.

                                                                                            3.  Shlm(s) on target thickness
                                                                                                and area (convert via  area
                                                                                               ratio).	
       (1) Reference Method 24  — Revision to Appendix A of 40 CFR Part 60,  Final Rule.

       (2) Method generally used  by  Industry.

-------
        TABLE  2.    SUMMARY  OF  PAINT  SPRAY  TRANSFER  EFFICIENCY
       	TEST METHODS  USED BY  INDUSTRY	
                      Summary of Methods

    Foil  on Simulated Shaoe; Weight Change (Lab) -Measure
    weight solids  per ASTM 0-1644-81 (or similar method);
    measure paint  spray  flow rate by capturing atomizer output
    in  anatomized  state; weigh foil before and after  drying
    (use  same  curing schedule as for paint samples).
    Shim on  Product; Thickness and Area (Plant) - Use theo-
    retical  mileage data  (ftvuncut gallon of paint)  from
    paint supplier; measure paint spray flow rate by  cap-
    turing atomizer output in anatomized state; measure shim
    dry film thickness  and overall product (part) area
    using GE magnetic thickness gage and flexible tape.
3.
Product Coating and Area  (Plant) - Use theoretical
mileage data (ft^/uncut gallon of paint) from
paint supplier; measure paint spray flow rate by
capturing atomizer output In anatomized state;
measure actual  product dry  film thickness and
overalI product (part) area.

Paint Inventory/Product Coating Thickness and Area
(Plant) - Use production  data (parts coated, surface
area, gallons of paint used); use theoretical mileage
data  (ft^/uncut gallon of paint) from paint
supplier; use QC data  for average film thickness.
                 Comments

   Ability  to accurately measure
   paint spray flow  is question-
   able (industry experience  with
   calibrated mass  flow meter  In-
   dicates  that this  method has
   inherent inaccuracy); overall,
   method looks good, e.g., dummy
   targets  used before and after
   main target to simulate pro-
   duction  situation.  Foil and
   simulated  shape  thermal  inei—
   tla — equivalent  bake
   schedule problem.

2. Potential  variability  in coat-
   ing thickness means deter-
   mination of overalI "average"
   film thickness may have  In-
   herent repeatability problem
   — multiple shims  helps
   minimize this problem;
   question of reliability on
   manufacturer's data for
   theoretical  paint  mileage;
   paint spray flow measurement
   in question (see comment I);
   questionable ability to mea-
   sure film thickness accurate-
   ly — repeated calibration
   may help minimize  this prob-
   lem; ability to  measure area
   accurately Is a  concern.

3. Same as  Comment  2; additional
   disadvantage is  limitation
   to ferromagnetic or aluminum
   substrate (magnetic or eddy
   current  applications).
                                                              4. Question of reliability on
                                                                manufacturer's data for theo-
                                                                retical paint mileage; ability
                                                                to measure film  thickness
                                                                accurately Is In question;
                                                                ability to measure area
                                                                accurately is a  concern.

-------
                                  TABLE  2  (continued
                            Summary of Methods

    Foil on Simulated Shape; Weight Change (Lab) -
    Measure weight solids per AS.TM D-1644-81;
    measure paint spray flow rate by mass flow meter
    (Micro Motion); weigh foil  attached to flat panel
    before and after drying (using same curing schedule
    as  for paint samples).
 6.   Foil on Simulated Shape; Weight Change (Lab) -
     Measure weight solids per ASTM 0-1644-81;
     measure paint spray flow rate by capturing
     atomizer output  In unatomlzed state; weigh
     foil attached to flat panel before and after
     drying — remove foil from rack prior to drying
     to  better simulate same bake schedule as paint
     samples.

 7.   Foil on Simulated Shape; Weight Change (Lab) -
     Same as Method 6 except that simulated shape
     is  2" x 4"  lumber.

 8.   Foil on Product; Weight Change (Lab) -Measure
     weight solids per ASTM 0-1644-81; measure
     paint use by weighing paint supply apparatus
     before and  after spraying; weigh foil before
     and after drying —  initial foil weight de-
     termined by subtraction process for amount
     of  foil remaining on roll.
 9.   Foil on Product; Volume (Lab) -Measure paint
     density per ASTM D 1475-60 (1980); measure
     paint use by weighing apparatus before and
     after spraying; measure film thickness on
     foil with micrometer; tape measure for area;
     measure volume fraction solids per ASTM D 2697-73.
10.   Product Coating; Weight (Plant) -Measure
     paint  solids on target by solvent stripping,
     distillation (evaporation, and weighing of
     dried  sol ids).
                 Comments

 5.  Only part of  foil  panel  is  coated
    and  thus  lab  test  does not  accurate-
    ly simulate plant  production  situ-
    ation —  this shortcoming,  however.
    Is true of essentially a I I  lab
    tests; other  problem  is  that  of
    thermal Inertia (foil  &  rack) and
    equivalence of  bake schedule.

 6.  Problem with  accurate measurement
    of paint  spray  flow (see Comment I);
    good approach on  paint drying
    procedure (target  and sample).
 7.  Same advantages and  disadvantages
    as Comment 6.
 8.  Problem with bake schedule since
    entire car  body dried  with foil
    attached (equivalalence problem
    with  sample drying method);
    hysteresis  (repeatability) con-
    cern  with use of load  cell.

 9.  Average film thickness determi-
    nation requires multiple
    (measurements to obtain good
    statistical  average);  complexity
    of  product  shape makes area
    determination a tedious process;
    load  cell hysteresis  problem.

10.  Load  cell hysteresis  (repeatability)
    problem; accuracy was  supposedly
    good  on load cell  (1,000 I b _+_ 0.02
    Ib),  however, as confirmed by  good
    correlation with lab TE weight
    method (foil on body  — refer  to
    Method 8).
                                             10

-------
                                  TABLE  2  (continued)
11.
               Summary of  Methods

Product Coating Thickness  i Area  (Plant) -
Measure paint volume solids per ASTM D 2697-73;
measure average film thickness with eddy
current Instrument; measure area  with flexible
tape; measure paint use be dipstick in mix tank.
12.
Product Coating; Weight Change  (Plant) -
Measure paint density per ASTM  0  1475-60
(1980); measure paint use by dipstick In
mix tank; measure weight gain of  car body
by load cell; measure paint weight solids
per ASTM D 1644-81.
               Comments

11.  Average  film thickness determined
    by  numerous measurements; car
    body  divided Into two general area
    categories to  help account for
    for variances  in film thickness
    for low  appearance (large variation
    in  film  thickness) and high appear-
    ance  (small variation In film thick-
    ness) areas.

12.  Incomplete method, however, concept
    may have some  merit for small parts
    applicatlons.
                                                        13. Same as Comment 4.
                                             11

-------
     Two target configurations were selected for the standard-
ized test.   A set of four flat panel targets was selected to
typify TE's for large, relatively flat industrial targets.  A
set of four vertical cylinder targets was selected to typify
TE's for coating smaller, more intricate targets.  These
targets, mounted in a prescribed configuration, constitute the
test panels for the EPA draft TE laboratory method.

     The following test procedure is similar to the one used for
the two laboratory tests described in Sections 5 and 6, but
reflects some improvements made to the procedure as a result of
subsequent test experience.

IMPROVEMENTS TO TEST PROCEDURE

     As a result of actual test experience, several recommenda-
tions can be made to clarify and standardize the test procedure.
First, it is recommended that the booth air flow rate be stand-
ardized.  Since most booth exhaust fans are not rate-adjustable,
booths with O.bl-m/s (100-ft/min) linear flow rates should be
selected for testing.  This  is a standard booth size that should
be available to testers.  The. exhaust fans should be used at all
times to keep air-borne solvent and paint levels at a minimum.

     Second, the test procedure must be modified to provide
identical paint delivery to  the targets.  The paint flow should
be  initiated from the same position (relative to the targets)
from test to test.  A consistent starting point and delivery is
expected to lead to more consistent TE's   between laboratories,
especially for electrostatic equipment.  It is recommended to
use the timing marks at the  first and last scavengers as paint
start/stop marks  (refer to the following section, "Draft
Transfer Efficiency Test Procedure").  The rationale for this
recommendation is in Section 7.

     Third, the electrostatic equipment voltages must be con-
sistent between laboratories.  Fixed power supplies may vary by
30 percent in output voltage, and cable, gun, and paint charac-
teristics vary considerably  between laboratories.  Variable
power supplies will be required to keep equivalent voltages on
subsequent tests.

     Fourth, calibration of  all pressure gages involved before
the test is advisable.  Gages on the paint supply tank, paint at
the spray gun, atomizing (or turbine) air pressure, and shaping
air pressure should be calibrated before the test.  The
requirement assures equivalent operating pressures between
laboratories.

     Fifth, spray guns should be equipped with new fluid tips at
the beginning of the test.

                               12

-------
     The test procedure should provide more guidance in accept-
able techniques for determining paint characteristics.   The
recommended technique must be readily available to labora-
tories.  Consistent documentation will avert confusion  during
tne test and uncertainty when comparing results.

     Finally, there has been some speculation about the effect
of having other grounded equipment near the target assembly.
Although this effect has not been directly observed,  it seems
prudent to keep all grounded objects away from the targets.  A
minimum 3 meters (10 feet) is recommended as the distance to
walls or other grounds.
                               13

-------
DRAFT TRANSFER EFFICIENCY TEST PROCEDURE1
Equipment Calibration

Perform calibration of the platform scale once per week or each
time that it is moved and leveled, whichever occurs more fre-
quently.  Perform calibration of the laboratory scale once every
3 months.  Calibrate all pressure gages.


Test Procedure

A.  Select test equipment.  Using Data Sheet 1, document
    the test equipment specifications.  Be sure to check the
    information and sign the form.
1Many conventional industrial units are used throughout the
 test procedure to accommodate participating laboratories and to
 minimize conversion errors on site.  Metric conversions are
 made as required as shown in the conversion list at the front
 of the report.
                               14

-------
                           Data Sheet 1

                   Test Equipment Specifications^

      Test Date:             Test No.:      Data by/Checked  by:
A. Vfeiyht Percent Solids Measurement Equipment
   1.  Laboratory Scales
       a. Manufacturer
       b. Model No.
       c. Serial No.
       d. Capacity, g
       e. Rated accuracy, g
   2.  Foil Dishes
       a. Type
       b. Size
   3.  Syringe
       a. Type
       b. Capacity, mL
   4.  Solvent Type
B. Conveyor Speed Measurement Equipment
   1.  Rule
       a. Type
       b. Graduations
   2.  Electronic Timer
       a. Type
       b. Manufacturer
       c. Model No.
       d. Serial No.
       e. Rated accuracy, s
C. Mass Flow Measurement Equipment
   1.  Platform Scales
       a. Manufacturer                       	
       b. Model No.                          _
       c. Serial No.                         _
       d. Capacity, kg                       _
       e. Rated accuracy, g                  _
   2.  Stopwatch
       a. Manufacturer                       _
       b. Model No.                          _
       c. Serial No.                         __
       d. Rated accuracy, s                  ~
D. Target Foil                               ~
   1.  Type                                  _
   2.  Nominal Thickness, mils
   3.  Temper
E.  Wet Film Measurement Equipment           ~
       a. Manufacturer
       b. Model No.                          ~
                                   15

-------
     B.   Select coating type to be used.  Using Data Sheet  2,
         document the paint characteristics.  Again, check  your
         information and sign the form.
Test Date:
                             Data Sheet 2

                         Paint Specifications
Test No. s
Data by/Checked by:
   1. Paint Type

   2. Resin Type

   3. Manufacturer

   4. Manufacturer's Paint ID No.

   5. Lot No.

   6. Color

   7. Recommended Cure Schedule

   8. Viscosity (uncut)

   9. Reducing Solvent

  10. Vol. of Solvent Put into
      Vol. Paint

  11. Viscosity - Spray (cut)*

  12. Wt./Gallon - Spray

  13. Wt. Solids - Spray

  14. Resistivity or Conductance
                   min. @
                sec. #
      Ford Cup @
                        (vol)  solvent  in
                        (vol) paint
               sec
      Ford Cup @
                           Ibs/gal
*Use ASTM D-1200-70, "Viscosity of Paints, Varnishes,  and  Lacquers
 by Ford Viscosity Cup."  Viscosity may also be determined by
 ASTM D-3794, Part 6 (Zahn Cup method) in_ addition _to  the  Ford Cup
 measurements.
                               16

-------
C.  Set up paint supply equipment and platform scale.
    Using Data Sheet 3, document the paint supply equip-
    ment specifications.  Be sure to check your informa-
    tion and sign the form.
                         17

-------
                          Data  Sheet  3

      Paint Spray and  Peripheral  Equipment Specifications

Test Date:                   Test  No.:            Data by/Chkd by;
A. Paint Supply Tank

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial  No.
   5. Rated Capacity,  gal

B. Paint Spray Equipment

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial  No.
   5. Rated Capacity,  cc/min
   6. Air Cap
   7. Fluid Tip
   8. Needle

C. Paint Spray Booth
   1. Type
   20 Manufacturer
   3. Model No.
   4. Serial  No.
   5. Rated Capacity,  cfm

D. Conveyor
   1. Type
   2. Manufacturer
   3. Model No.
   4, Serial  No.

E. Forced  Draft  Oven
    1. Type
   2. Manufacturer
   3. Model No.
   4. Serial  No.

F. Paint  Heaters
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial  No.
                                18

-------
D.  For electrostatic spray equipment only, ground paint
    supply equipment and platform scale per Figure 1.
    NOTE:  In accordance with Section 9-8 of NFPA 33 for
    fixed electrostatic apparatus, measure resistance of
    equipment to ground (conveyor frame) to insure resist-
    ance is less than 1 x 106 Ohm.

E.  Using a small glass jar with an airtight lid, take
    paint grab sample from paint pot.  Record test series
    number on label of jar.

F.  Measure weight solids from grab sample using syringe
    weight difference technique as described in ASTM
    D-2369-81.  Document the cure oven bake schedule and
    temperature on Data Sheet 4.  Be sure you use the cure
    schedule recommended by the manufacturer on Data Sheet
    2.  Record raw data and results on Data Sheet 5.
    Paint weight percent solids should be determined before
    each test series, at the start of each test day,  and
    periodically between tests.
                          19

-------
                         Arrangement A
                  Nonelectrostatzc Eauiomen
                             Paint

                             Suppl-_.

                             Tank
                         Arrangement B
                    Electrostatic  Equipment
                                                  -Digital  electronic
                                                  platform scale
      Grounding Cable
                                                 •Electrical Insulation Block*
                                                   Digital electronic
                                                   Platform Scale
*31oc.k must  be capable of preventing current flow from supply tank to
 ground through the platfora scale.
   Figure  1.   Set-up for Paint Supply  Equipment and
                 Platform Scales
                                  20

-------
                            Data Sheet 4

                   Equipment Operating Conditions

Test Date:                    Test No.:          Data  by/Chkd  by
A. Paint Spray Equipment
   1. Paint Pressure at Paint Pot, psig
   2. Paint Pressure at Spray Gun, psig
   3. Atomizing/Turbine Air Pressure at
       Spray Gun, psig
   4. Operating Voltage, kV
   5. Disk or Bell Speed, rpm
      a. With Paint Applied
      b. Without Paint Applied
   6. Shaping Air for Bell, psig
   7- Paint Temperature at Paint Pot,  °F
   8. Gun to target distance, cm
   9. Pump Setting

B. Paint Spray Booth
   1. Ambient Temperature, °F
   2. Relative Humidity, %
   3. Air flow Velocity, fpm
   4. Air Flow Direction

C. Target Parameters
   1. Average Wet Film Thickness, mils
   2. Average Dry Film Thickness
   3. Vertical Paint Coverage, cm (in)
   4. Target Height, cm (in)
   5. % Vertical Coverage
   6. Resistance to Ground, Ohm

D. Forced Draft Oven*
   1. Cure Time, minutes
      a. Foil Dish (sample)
      b. Target Foil

   2. Cure Temperature, "F
      a. Foil Dish (sample)
      b. Target Foil

E. Paint Heaters
   1. Temperature In, °F
   2. Temperature Out, °F

F. Conveyor Speed Setpoint, fpm (cm/sec)
*Same cure schedule as foils.

                               21

-------
                          Data Sheet  5

               Weight Solids Test  Data  &  Results
Test Date?
                                 Test  No.:
      Data by/Chkd by:
1.  Syringe Weight
    a.  Full, g
    b.  Empty, g
    c.  Net Wet Paint, g

2.  Dish Weight
    a.  After Drying, g
    b.  Empty, g
    c.  Net Dry Solids, g

3.  % Weight Solids  (2c/lc
                                Sample
                                  A
   Sample
     B
Average
                    A3
NOTES:

     1.  Actual Cure Schedule
min
Refer to ASTM  2369-81,  Procedure  B  of  "Standard Test Method for
Volatile Content of Coatings."
                                22

-------
G.  Set up the paint spray equipment.  Using Data Sheet 3,
    document specifications for the paint spray equipment
    and spray booth used in this test.

H.  Set up the conveyor speed measuring equipment.  This
    equipment may consist of photoelectric cells or limit
    switches used in conjunction with an automatic digital
    timer.  Alternatively, the conveyor speed may be mea-
    sured using timing marks (chalk marks) on the conveyor
    in conjunction with a hand held stopwatch.   Figure 2
    shows tne permissible methods for conveyor speed mea-
    surement.  Using Data Sheet 6a, record the horizontal
    distance between the photo cell or limit switch on/off
    positions.
                         23

-------
                        METHOD A
  Target
Target
Target


^

B
r~


c


                                                       Electronic
                                                         "imer
         A = Stationary photoelectric cell or  limit switch

         B = Stationary photoelectric cell or  limit switch

         C = Moving plate of known width



                        METHOD B
            Known Distance
                 .onvevor
                                                       Stop Watch
         E =  Fixed timing mark

         F =  Moving timing mark
Figure  2.    Permissible Methods for Measuring
              Conveyor Speed
                           24

-------
                           Data Sheet 6a

                     TE Test Data and Results


Test Date:                  Test No.:        Data by/Checked by:
A. Weight Percent Solids  (from Data Sheet 5)         	 A3


8. Total Solids Sprayed


   1. Paint Spray Flow Rate

      a. Beginning Weight, g           	.

      b. End Weight, g	

      c. Time Between Weighings, s     	
      d. Flow Rate, g/s                	Bid


   2. Conveyor Speed

      a. Distance  Between Marks, cm    	

      b. Time Between  Marks, s         	
      c. Speed,  cm/s                   	B2c
    3. Total  Effective Target
      Width,  cm*                         15.24    B3
    4. Total  Solids  Sprayed, g
       (A3 x  Bid  x B3/B2c)              	B4
     Total effective  target width is six  inches per  foil  on  flat
     panel target  (on 6" centers), and six  inches per cylinder
     on  vertical cylinder target  (also on 6" centers).  Six
     inches  =  15.24 cm.
                                25

-------
                                 Data Sheet 6b

                            TE Test Data and Results
Test Date:                     Test No.:        Data by/Checked by:
C. Total Solids on Target
   Flat Panel Target

   Foil Weight After Drying, g:

Foil #1               2               3               4           Total
   Foil Weight Before Spraying, g:
   Net Dry Solids, g:
    Foil vfeight Before Spraying, g:
    Net  Dry Solids, g:
 E.  Transfer  Efficiency  (by weight)^

    Flat  Panel Target

    Vertical  Cylinder Target
 1.    Net  Dry Solids         x   100  = TE
    Total  Solids Sprayed

                                     26
 D. Vertical Cylinder Target

   Foil Weight After Drying, y:

 Foil  *  1               2              3                4           Total

-------
I.  Set up targets in accordance with Figure 3 or 4, as
    appropriate.  Target configuration,  material, and
    spacing is critical.  Scavengers are metallic, as is
    the FP target.  Cut 6-inch-wide aluminum foil strips to
    required length for each target.  Label each foil strip
    with the appropriate nomenclature.   (See Section 5 for
    nomenclature.)  Weigh each foil strip and record value
    on foil and on Data Sheet 6b.

J.  Attach foils to the vertical cylinder and/or flat panel
    targets as shown on Figure 5 or 6,  as appropriate.
    Perform resistance check to verify  adequacy of ground-
    ing. Per NFPA 33 Section 9-8, resistance shall be less
    than 1 x 106 ohms.

K.  In accordance with Figure 3 or 4, attach shim stock
    to scavenger in order to measure wet film thickness.

L.  Adjust all equipment operating parameters, i.e., gun to
    target distance, atomizing air press, paint pot pres-
    sure, shaping air pressure, turbine air pressure, etc.,
    to desired values.  Record equipment operating parame-
    ters on Data Sheet 4.  NOTE:  In accordance with Sec-
    tion 9-7 of NFPA 33 for fixed electrostatic apparatus,
    the gun to target distance shall be at least twice the
    sparking distance.

M.  Check spray gun condition.  Install new fluid tip, air
    cap, and needle.

N.  For electrostatic spray equipment,  measure the gun tip
    operating voltage (with lines full  of paint, but gun
    not operating).  Adjust to desired  voltage and record
    on Data Sheet 4.

O.  Check conveyor clock, stopwatch, and platform scale to
    ensure that all have been zeroed (reset) and that the
    scales are in the tare mode.

P.  Turn on conveyor.  As the leading edge of the first
    scavenger passes in front of the gun, turn on paint
    spray equipment and initiate flow;  simultaneously,
    start stopwatch.

Q.  As the trailing edge of the last scavenger passes in
    front of the gun area, stop stopwatch and paint spray
    flow simultaneously.  Turn off conveyor.  Record
    platform scale, conveyor clock, and stopwatch readings
    on Data Sheet 6a.

R.  Measure wet film thickness on shim plate and record on
    Data Sheet 4, line C-l.

                          27

-------
                                             Target movement toward gun
    Wet Film
    Thickness
     Plate
NJ
oo
     Note:
    lin.=
    2.54cm.
                                            Conveyor
                       12"
                                   Foil No.
                                 4321
           Foil  No.
           3     2    1
f • • • 1




kl

K
0


1



















































o
V
c
o
>

o
if)



yl%MDia
aluminum
pipes
Stainless
St«el

^^





1 1
1 	 1






^



L










1




















































                           Vertical  Cylinder (VC)
                               Target
6"  6
       2-    2"    2"

Flat  Panel  (FP) Target
                       Figure 3   Target configuration for air atomized conventional
                                  and electrostatic  spray guns.
                                        36"

-------
                                  Target movement  toward gun
                     conveyor
                             Foil Number
                                                      Foil Number
T
 60"
\ / \ 4321 / \ 4 3 2 1 / /
I /







M
O
U'
c
O
£»
ID
U
W


t!"
r.




h
01
tr
C
OJ
^
ra
u
to







































































'1



^

ID
CP
c
a)
^
(d
u
CO



1

12"

6"
_
6"

6"

6"

6"
IV dia
aluminum
pipes
Stainless
\ 1




Steel
.
^^^
^^x^.








1
12"


























































1


















































































6"

6-


i'
-'il
T

f."
r-«-

J'

6""

G"


















6"
^ /





^
c>
Cn
C
O

flj
u
CO




8"

             Vortical Cylinder Taryot
                   (VC)
                                                                                         wet film
                                                                                         thickness
                                                                                           plate
Flat  Panel Target  (FP)
              Figure   4   .   Target  Configuration  for  High  Speed Bell

-------
           Leading  edge
            of foil
         Vertical  Cylinder
                                                      -.-.-ran
                                             Hold  edge: of foil in place against
                                             cylinder while wrapping leading edge
                                             arour.d  cylinder.
  Foil
       V.'rap vertical cylinder  targets with cylinders  mounted on target bracket
       (See Figure  3   and  4  j.  wrap so the leading edge  forras a seam away
       fron the direction of sorav.
                                                 •Gric"
'"rap
'•••'rap
                                           "Grio"
                                               "Grip"
                                                             •Grin"
                       "Grip"
                                                        -_-.  "Grip"
  As leading edge overlaps starting
  edge,  solidly "grip" foil into  place
  by grasping foil-covered cylinder.
Secure foil on  cylinder by gripping
the length of the  cylinder.  Foil will
have a uniformly wrinkled surface.
      Figure  5.    Ransburg Vertical Cylinder Wrapping  Technique

                                       30

-------
      FLAT PANEL TARGET
Flat Panel Target
   Foil-Ready for
   attachment
                                      Double-sided tape
 Figure  6.   Flat  Panel Foil  Attachment  Technique

                          31

-------
S.  Remove foils from targets and securely attach to oven
    racks so all painted surfaces are exposed for uniform
    drying.  Spring clips or tacks may be used to mount wet
    targets.  Insert racks in oven and bake at recommended
    schedule per Data Sheet 2.  Flash time (the time be-
    tween spraying and getting the targets into the oven)
    should be kept to a minimum.  Five minutes is con-
    sidered excessive flash time-  Set oven timer per
    recommended schedule.

T.  Remove foils from oven and record actual bake schedule
    on Data Sheet 4.  Weigh foils and record weight on each
    foil and on Data Sheet 6b.  After weighing, store foils
    in appropriately labeled plastic bags, i.e., bags
    that have test run number identified.

U.  Perform TE calculations using the average weight solids
    for all the samples taken for the test run.  Document
    results on Data Sheet 6b.

V0  After completing a test series, perform statistical
    analysis as indicated on Data Sheet  7.  Document
    results of analysis on Data Sheet 7.

W.  Repeat above steps (A through S) for each test series.

X.  After completing all 12 test series, transfer results
    of each TE test to Data Sheet 8.
                           32

-------
                                Data Sheet 7

                vfork Sheet for Calculating Standard Deviation
Test Series:

Calculation by:

Checked by:
Test Run No.
1
2
3
4
5
6
7
8
Totals:1 n =
TE|yi = B/n
s = [ C / (n-1)] °
TE, %








B

.5
(TE-TE,), %







•
=
=
=
(TE-TEtf)2, %2








C =
(1)
(2)
 1.  TEjj = mean value for transfer efficiency, %:

      vdiere B   = TEi + TE2 + ...  TEn

      and n     = Total number of  test runs for one
                  test series

      s         = Standard deviation, %

      where C   = [(TE^TE^)2 + (TE2-TEN)2 + ... +
                                     33

-------
                               Data Sheet 8

              Statistical Analysis Sunmary for TE Test Results

Prepared by:	

Checked by;  	

              Number of                   Standard           Coefficient of
Test Series   Test Runs   TE,^, %  (1)    Deviation,  %  (2)     Variation,  % (3)
   1.   TEM = mean value  for  transfer  efficiency
       (Refer to Data Sheet  7)
   2.   Refer to Data Sheet 7
   3.   (Standard Deviation / TE^)  x 100
                                     34

-------
                            SECTION 5

                    PHASE I LABORATORY TEST
FACILITIES

     The original laboratory test was performed September 24
through 30, 1982, at PPG Industries' Allison Park, Pennsylvania,
laboratory.  PPG donated the facilities, paints, and technicians
for the test.

     The Phase I laboratory tests were run in two booths.  The
air atomized conventional and electrostatic equipment was run in
a closed water-wash booth rated at 1.9 m3/s (4000 ft3/min).
The high speed bell equipment was run in a closed dry-filter
booth rated at 4.2 m3/s (9200 ft3/min).   Exhaust was normal
to the targets in both booths. Paint spray and peripheral equip-
ment specifications used for this test are shown in Table 3.
The equipment sensitivities are documented in Table 4 and 5. In
addition to this equipment, the following items were used for
measuring electrical resistance and electrostatic voltage
potential:

     •  Volt-ohm meter; Nordson 790-108 Hand kV Meter.

     •  High voltage DC meter; 0-100 kV; PCF Group, Inc.;
        5,000 megaohm probe; Model HV-100A.

DESCRIPTION OF PAINTS

     A 65-percent solids coating and 52-percent solids coating
were supplied for testing by PPG.  The paint properties are
summarized in Table 6.  The higher solids coating is typical of
a compliance solvent-borne coating for the metal finishing
industries.  The lower solids coating is an example of an
acrylic enamel used in the automotive industry.  Paint weight
percent solids content was determined by ASTM D-2369-81, except
for cure times.

     The ASTM method uses syringes to dispense the wet paint
into the aluminum weight dishes.  Initially, five grab samples
were processed for each coating using two weight dishes for each
grab sample.   At the end of the testing (Thursday, September 9,
1932), an additional grab sample was taken from the 5-gallon can


                               35

-------
TABLE  3.   SUMMARY  OF  PAINT SPRAY  AND  PERIPHERAL  EQUIPMENT SPECIFICATIONS  (1)


A. Paint Supply Tank
1. Type
2. Manufacturer
3. Model No.
4. Serial No.
5. Rated Capacity, gal
B. Paint Spray Equipment
'. Type
2. Manufacturer
3. Model No.
4. Serial No.
5. Rated Capacity, cc/mln
6. Air Cap
7. Fluid Tip
8. Needle
9. Power supply
C. Paint Spray Booth
1. Type
2. Manufacturer
3. Model No.
4. Serial No.
5. Rated Capacity, m-Vs
(ft3/mln) estimated
Oe Conveyor
1. Type
2. Manufacturer
3. Model No.
4. Serial No.
E. Forced Draft Oven
1. Type
2. Manufacturer
3. Model No.
4. Serial No.
Air Atomized
Electrostatic

Pressure Pot
Blnks
83-5501
Not aval lable
2

Air atom elec
Nor d son
AN-8
Not aval lable
Not available
987
Not available
Not ava 1 lable
EPU-8; 76 kV rated;

Water wash
Binks
Dy naprec 1 p 1 tor*
Not aval lable
1.9 (4,000)


Overhead
Not available
Not aval lable
Not available

Convect ion
Despatch
V-29 HD; V-15 HD (3)
Not aval lable
Air Atomized
Conventional

Pressure Pot
Blnks
83-5501
Not ava 1 lable
2

Air atom conv
Blnks
610
Not aval lable
Not aval lable
63 PB
63 C
Not aval lable
Not appl 1 cable

Water wash
Blnks
Dynapreclpltor®
Not aval lable
1.9 (4,000)


Overhead
Not available
Not aval lable
Not available

Convect Ion
Despatch
V-29 HD; V-15 HD (3)
Not aval lable
High Speed
Bell

Positive displacement pump
Ransburg
9966-9
062702
1 (2)

High speed bel 1
Ransburg
Turbobel 1
Not aval lable
Not aval lable
Not appl icabl e
Not appl (cable
Not appl Icable
Not available

Dry filter
Blnks
Not available
Not aval lable
4.2 (9,200)


Overhead
Not available
Not aval lable
Not available

Convection
Despatch
V-29 HD; V-15 HD (3)
Not aval lable
       (I)  No  paint heaters were required for these tests.
       (2)  A l-gailon can was used  to  provide paint to the  pump suction.
       (3)  Two different ovens used to dry foils.

-------
            TABLE 4.  TEST EQUIPMENT SPECIFICATIONS
A.
B.
C.
Percent Weight Solids
1. Laboratory Scales
a. Manufacturer
b. Model No.
c. Serial No.
d. Capacity, g
e. Rated accuracy
2. Foil Dishes
a. Type
b. Size
3. Syringe
a. Type
b. Capacity, raL
4. Solvent Type
Measurement Equipment
Mettler
P12UUN
558312
1,200
, Q + 0.01
Aluminum
Standard
Luer
5
High Solids - Solvesso® 100
Low Solids - Acetone
Conveyor Speed Measurement Equipment
1. Rule
a. Type Flexible Steel Tape
b. Graduations
2. Electronic Timer
a. Type
b. Manufacturer
c. Model No.
d. Serial No.
e. Rated Accuracy
Mass Flow Measurement
1/16 inch
Digital
Lab - Line Instruments, Inc.
1405
Not available
, s Not available (1)
Equipment
       Platform Scales
       a. Manufacturer                         NCI, Inc.
       b. Model No.                             5780
       c. Serial No.                          C790482
       d. Capacity, kg                     	90 (2)	
       e. Rated Accuracy, g                	 £ 5	
   2.  Stopwatch                                 ~"
       a. Manufacturer                     Markson Science,Inc.
       b. Model No.                           Digital	
       c. Serial No.                           26119	
       d. Rated Accuracy, s                	 + 0.001% (3 )	
D.  Target Foil                                ~"
   1.  Type                                  Aluminum	
   2.  Nominal Thickness, mils                 1.5
   3.  Temper                                Medium	
E.  Wet Film Measurement Equipment
   1.  Distributor                         Paul N. Gardner Co.
   2.  Model No.                              Nordson Type	

(1)  Per telephone conversation with manufacturer on 10/4/82.
(2)  Fitted with 45 kg load cells.
(3)  +0.001% of reading.
                               37

-------
TABLE 5.  MEASURED PARAMETERS AND RATED MEASUREMENT  ACCURACY,
     Parameter
   Test
  Device
   Weight

     - Aluminum dish

     - Syringe

     - Target foil

     - Paint supply tank

   Time

     - Paint capture

     - Conveyor speed

   Distance

     - Target width


     - Conveyor spe.ed
lab scale

lab scale

lab scale

platform scale



stop watch

digital timer



steel rule


steel rule
   Rated
  Accuracy
 +_ 0.01 g

 + 0,01 g

 _+ 0.01 g

 +    5 g



 _+ 0.001 %  (1)

 Unknown (2)
4- 0.001 m
(1/32 inch)  (3

i 0.001 m
(1/32 inch)
 (1) Of  reading.

 (2) Not available  from  manufacturer;  resolution is +_ 0.05 sec
     (1/2 scale reading  of  0.1  sec).                 ~~

 (3) Equivalent to  1/2 of smallest  reading.
                                38

-------
              TABLE  6.  SUMMARY OF RECORDED PAINT SPECIFICATIONS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Paint Type
Resin Type
Manufacturer
Manufacturer's Paint ID No.
Lot No.
Color
Reconmended Cure Schedule 10 min (
Viscosity (uncut paint)
Reducing Solvent
% Reduction to Spray
Viscosity - Spray
wt/gal - Spray
wt Solids - Spray
Resistance or Conductance
Paint 1
Enamel
Polyester
—
W4533D
Not available
Beige
§ 177°C (350°F)
28.0 sec (1)
Solvesso® 100
2700 cc/15. 1L
14.5 sec (1)
64.8% (4)
64.8% (4)
2.3 Mf>
Paint 2
Enamel
Acrylic
—
DXH82445
Not available
Gold
30 min @ 121°C (250°F)
27.2 sec (2)
DxD reduce (3)
1400 CC/11.4L
18.0 sec (2)
52.0% (5)
52.0% (5)
2.2 M!2
(1)   #3 Zahn cup @ 23°C (74°F).

(2)   *4 Ford cup @ 23°C (74°F).

(3)   DxD reducer composition:   5U% acetone, 30% Cellosolve® acetate,
       12% toluene, 8% Solvesso^ 150.

(4)   Mean value for 19 weight dish samples.

(5)   Mean value for 10 wt dish samples.
                                    39

-------
containing the cut 65-percent coating.  Another nine dishes were
processed from this grab sample and the following bake schedule
was used:

     *  3 dishes for 10 minutes at 177°C (350°F)

     •  3 dishes for 20 minutes at 177°C (350°F)

     •  3 dishes for 30 minutes at 177°C (350°F)

Nineteen (5x2-1-3x3) sample dishes were thus obtained
for the 65-percent solids coating.

MASS FLOW COMPARISON TEST

     On Monday, September 27, 1982, the scheduled laboratory
tests began with a comparison of two paint mass flow measurement
techniques:

     •  Method A - Pressurized paint pot weight change
                   (atomizing air on)

     •  Method B - Capture and weighing of unatomized paint flow

Method A would determine actual atomized paint flow during TE
test runs; Method B would predetermine a paint mass flow rate
for unatomized paint to be used in TE calculations.  Method B,
commonly used by industry, was suspected to be unrepresentative
of paint flows during spray operations.  The purpose of this
test was to determine if there was a statistically significant
difference between the two techniques.  If a significant
difference is detected, Method A will be selected as most
representative for TE testing.

     The following procedure was used to obtain the desired test
data:

     1.  Select test equipment.  Using Data Sheet D-l, document
         the test equipment specifications.

     2.  Select coating type to be used.  Using Data Sheet D-2,
         document the manufacturer supplied paint character-
         istics .

     3.  Select spray equipment type and paint supply equipment
         to be used.  Using Data Sheet D-3, document the equip-
         ment specifications.

     4.  Set up all equipment and adjust paint pot pressure and
         atomizing air pressure to desired settings.  Record
         these settings on Data Sheet D-3.
                               40

-------
     5.  Check stopwatch and platform scale to ensure that both
         have been zeroed (reset) and that the scale is in the
         tare mode.

     6.  Turn on paint spray equipment and initiate flow.
         Simultaneously, start stopwatch.  After approximately
         500 g have been sprayed, stop stopwatch and paint
         spray flow simultaneously.   Record platform scale and
         stopwatch readings on Data Sheet D-4.

     7.  Repeat steps 5 and 6 until a total of eight samples
         have been recorded.  Proceed to Step 8.

     8.  Turn off atomizing air pressure to spray gun.

     9.  Tare a paper cup on laboratory scale.

    10.  Check stopwatch to ensure that it has been zeroed.

    11.  Turn on paint spray equipment and initiate unatomized
         paint flow into paper cup.   Simultaneously, start
         stopwatch.  After cup is approximately 3/4 full, stop
         stopwatch and paint flow simultaneously.

    12.  Weigh filled (and previously tared) paper cup on
         laboratory scale.  Record weight gain and stopwatch
         readings on Data Sheet D-4.

    13.  Repeat steps 9 through 12 until eight samples have
         been recorded-.

    14.  Perform statistical analyses of test data using the
         following two techniques:

         •  Wilcoxon test for two independent samples
         •  "Prob-Value" using t-statistic

     In performing the above tests the platform scale was
initially calibrated and then checked for accuracy using a paper
cup that had been preweighed at 5 g on the laboratory scale.
The results indicated that the platform scale was capable of
measuring 1,000 _+ 5 g and 500 _+ 5 g and was thus within the
desired precision band.

     After checking out the platform scale, the first test run
was performed and aborted because the stopwatch was accidentally
stopped at about 300 g without shutting off the spray gun.  A
total of five test runs had been performed when the entire test
series was aborted.  This termination of testing occurred be-
cause it was discovered that there was insufficient paint
available of the same type already in the pot to complete the


                               41

-------
tests.   A switch was made to another spray booth and another
paint and the Method A test series was begun again.  At this
time it was suggested that a calculation be made to determine
the impact of displacement of the paint in the pot by pres-
surized air at 322 kPa (32 psig).  The calculation indicated
that 500 g of paint was displaced by about 1.4 g of air.  In
effect, therefore, about 501.4 g of paint was actually being
sprayed.  Since the accuracy and resolution of the platform
scale was _+ 5 g, the 1.4-g value was below this threshold limit
and for all practical purposes was undetectable. Therefore, no
adjustment was made in the spray flow rate measurements for
Method A  (paint pot weight change).

     After completing eight measurements for Method A, a ninth
measurement was made in the event any of the data was later re-
jected using an outlier analysis.  For example, it was noted in
the Log Book that data point A-8 may have been suspect due to
the paint line possibly having an air bubble in it.

     After completing the Method A test series, the testing
proceeded to Method B for capturing the unatomized output.  This
test series went smoothly until data point B-8 was ready to be
taken.  At this time it was noted that although the spray gun
had been shut off, a small stream of paint was leaking from the
nozzle.  The gun was taken apart, cleaned and reassembled for
data point B-8.  An additional test (B-9) was run in the event
8-8 or any other data point in the Method B test series was
later rejected on an outlier basis.

     The analysis of the data, which led to the selection of
Method A  for all tests, is discussed on p. 61.

TEST PARAMETERS

Determination of Sample Size

     The  selection of the recommended sample size for each test
series was based on the use of small sampling theory and the t-
statistic.  Small sampling theory uses the following terms and
relationships:

        U = X + t x s/(n)°-5

     where U = True population mean value of transfer
               efficiency, %

           X = Measured mean value of transfer efficiency for n
               samples, %

           t = t-statistic representing area under the
               t-distribution probability curve with n-1 degrees
               of freedom for some specified confidence interval

                               42

-------
           s = Standard deviation for n observations

           n = Number of TE sample values

     The last term in the above relationship represents the
error in estimating the true mean value of transfer efficiency
(U) using the measured mean (X).  To estimate the sample size it
was assumed that the error term should be less than or equal to
the standard deviation at the 95-percent level of confidence.
The sample size can then be evaluated using the following
relationship:

        t/(n)0-5 <_ l

     or n >_ t2

  where t = t-statistic at 95 percent confidence level for n-1
            degrees of freedom

     The results of the sample size evaluation are as follows:

        H    n-1    t @ 95%    t2

        3      2
        4      3
        5      4
        6      5
        7      6
        8      7

     This evaluation shows that a minimum of seven observations
should be selected to ensure that the measured mean TE values
are within one standard deviation of the true population mean at
the 95-percent level of confidence.  For these laboratory tests
a sample size of eight was selected for each test series.

Foil Handling Procedures

     Of all the steps in the transfer efficiency test procedure,
foil handling proved to be the most time consuming.  Foil
handling involves the following activities:

     •  Cutting foil to desired target length

     •  Labeling foil with appropriate test run nomenclature

     •  Preweighing foil and labeling foil with the beginning
        weight

     •  Mounting foil on flat panel or vertical cylinder targets
                               43
4.303
3.182
2.776
2.571
2.447
2.365
18.5
10.1
7.7
6.6
6.0
5.6

-------
     o  Removing foil from flat panel or vertical cylinder tar-
        gets

     o  Attaching foil to oven drying racks and inserting racks
        in oven

     o  Removing foils from oven and racks, weighing and
        labeling each foil with the ending weight

     Of all these activities the foil mounting and removal from
the targets proved to be the most difficult, particularly for
the 3.2 cm (1-1/4 inch)  diameter vertical cylinder targets.
Figure 7 shows the attachment techniques for the two target
configurations.  For the flat panel targets the foil attachment
and removal were relatively easy.  Double-sided tape was
attached to the flat panel targets to allow four sample foils to
be placed on the targets.

     In the case of the vertical cylinder targets, however, foil
attachment and removal were difficult.  First, a strip of mask-
ing tape was placed the entire length of one of the seams.  Next
a strip of double-sided tape was placed over the masking tape.
The foil was then lined up as vertically plumb as possible and
the front edge (refer to Figure 7) attached to the vertical
cylinder using a small piece of masking tape at the top, middle,
and bottom of the foil.   Then the foil was tightly wrapped until
the rear seam was barely touching the vertical cylinder.  At
this point a vertical mark was made to allow a piece of masking
tape to be lined up correctly and tjien attached to the vertical
cylinder.  The rear seam with double-sided tape was then se-
curely attached to the masking tape.  The purpose of the masking
tape was to minimize the potential for tearing the wet foil
while removing it prior to oven rack attachment.  Experimenta-
tion with the foil attachment had shown that the double-sided
tape was too sticky to allow its direct placement on the foil
surface.

QA/QC Procedures

     As part of the overall Quality Assurance Program Plan,
QA/QC procedures were developed and implemented for the
laboratory TE tests.  For example, to provide ease in data
handling, the nomenclature shown in Table 7 was adopted and used
extensively throughout the tests.  To facilitate recording of
data, three notebooks were kept during the tests:

     •  Log Book

     •  Master Data Book

     •  Weight Solids Data Book
                               44

-------
••.asking
  Tace
Oouble-sided tape
                                        -Double Sided Tape
                                                                    Foil
                                                              -Target

                                                            Expanded Top View
                                                          Foil "wrapping VC  Target
      Foil-Ready for attachment
         to VC target
              VC Target with
              Foil Attached
                                  FLAT PANEL TARGET
                                                          0.15 m (6 in.)
                           Flat Panel Target
                              Foil-Ready for
                              attachment	
                                                                 Double-sided tape
            Figure 7.   Foil Attachment Techniques  for Vertical
                         Cylinder and  Fl.-it  Panel Targets,  PPG  Test,
                         September 1982
                                        45

-------
              TABLE 7.   NOMENCLATURE FOR SPRAY PAINTING
                       TRANSFER EFFICIENCY TESTS
  Parameter                                         Nomenclature
I.    Spray Equipment Type

     A.   Air atomized conventional                      AAC
     B.   Air atomized electrostatic                     AAE
     C.   High speed bell                                HSB
II.  Coating Type (1)

     A.  Automotive  enamel                               52
     B.  General purpose high solids                     65
III.  Target Configuration

     A.  Flat panel                                     FP
     B.  Vertical cylinders                             VC
Test No.
    Example:   AAC - 52 - FP - 1

    rthere   AAC - 52 - FP = Test series number
      and               1 = Observation number
(1) For clarity, the numerical weight percent solids is used
    to identify the coating.
                                 46

-------
     The Log Book was used to keep track of miscellaneous data
and to record any test anomalies.  The Master Data Book
contained the following information:

     •  All data sheets for all tests

     •  Up-to-date copy of the Test and Evaluation Plan,
        including Appendices A through E

     •  Summary of test matrix components

     •  Test sequence

     •  Test flow charts

     The Master Data Book served as the primary reference docu-
ment for all the TE tests.  The height Percent Solids Data Book
contained all the necessary data sheets and a copy of ASTM
D-2369-81 (Standard Test Method for Volatile Content of Coat-
ings).  All of the weight percent solids analyses were recorded
in this book.

     To help ensure correctness, thoroughness, and completeness,
all data sheets were independently reviewed, checked, and signed
off.

     As part of the overall QA/QC Plan, equipment with high
rated accuracy was used wherever possible.  Table 5 summarizes
the major measured parameters and the associated equipment rated
accuracy-

Test Sequence

     Table 8 depicts the actual test  sequence used for the
transfer efficiency laboratory tests.  As this table shows, the
platform scale was initially set up on September 24, 1982.  A
quick check of the scale indicated that it was a 90-kg (200-lb)
model as opposed to the desired 45-kg (100-lb) model.  The scale
rental supplier advised that the desired 45-kg load cells had
been installed in the scale with the  desired _+ 5-g accuracy and
readability (resolution).  A paper cup was cut to size and
weighed on the Mettler laboratory scale (accuracy of _+ 0.01 g)
to obtain a "calibration" weight.  The scale was then loaded
with 1,000- and 500-g weights and the paper cup was added to
determine if the scale would read 1,005 and 505 g.  This
preliminary check-out was successful.

     After checking out the scale, the vertical cylinder targets
were assembled from the prefabricated pieces.  The target
assembly went smoothly and all preliminary test arrangements
were completed in preparation for the following week's tests
(9/27 - 9/30/82), described below.

                               47

-------
                     TABLE 8.   TRANSFER EFFICIENCY TEST SEQUENCE
                                AT PPG SEPTEMBER 1982
    DATE

Friday (9-24-82)
Monday (9-27-82)
         ACTIVITY

o Set up platform scale and
  made preliminary adjustments

o Set up vertical and flat panel
  targets and performed "dry run"
  check out

o Performed mass flow comparison
  test
                                                             TEST  RUNS
o Began TE test runs



Tuesday (9-28-82) o Continued TE test runs

o Performed additional TE
test run










Wednesday (9-29-82) o Changed to HSB booth

o Completed TE test runs

o Booth cleanup



Thursday (9-30-82) o Rechecked target foils
o Performed additional weight
solids tests
o
0
o
o
o
o
o
0
0
o
o
o
o
o
o
o
o
o
0
o
o
o
o
o
o
o



AAE -
AAE -
AAE -
AAE -
AAE -
AAE -
AAE -
AAE -
AAC -
AAC -
AAC -
AAC'-
AAC -
AAC -
AAC -
AAC -
AAE -
AAE -
HSB -
HSB -
HSB -
HSB -
HSB -
HSB -
HSB -
HSB -



65
65
65
65
52
52
52
52
52
52
52
52
65
65
65
65
65
65
65
65
65
65
52
52
65
65



- VC -
- FP -
- VC -
- FP -
- VC -
- FP -
- VC -
- FP -
- VC -
- FP -
- VC -
- FP -•
- VC -
- FP -
- VC -
- FP -
- VC -
- ST -
- FP -
- FP -
- VC -
- VC -
- FP -
- FP -
- VC -
- VC -



1
1
5
5
1
1
5
5
1
1
5
5
1
1
5
5
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
4
4
8
8
4
4
8
8
4
4
8
8
4
4
8
8
1A to 4A
1
1
5
1
5
1
5
1
5



to
to
to
to
to
to
to
to
to



3 (1)
4
8
4
8
4
8
4
8



  1) ST = semi-tubular target configuration
                                    48

-------
PHASE I TEST RESULTS

General

     Transfer efficiency measurements were made for a total of
12 test series.  These series encompassed a test matrix of three
types of spray equipment, two coatings, and two target config-
urations.  For each test series, eight foils were coated to
determine transfer efficiency.  The eight observations per test
series were obtained by coating two separate groups of four
foils per target assembly (refer to Figures 3 and 4).  For
comparative purposes, however, Table 9 summarizes the results of
the tests based on evaluating the data under the following two
scenarios:

    •  Each test series is composed of eight observations from
       two groups of four foils each

    •  Each test series is composed of two observations; each
       observation is the average of a group of four foils

     These two different scenarios for evaluating the data are
shown because of the question of whether or not a group of four
foils should be treated as four separate observations or as one
observation averaged from four foils.  Treating the test data
for each test series as two observations composed of the
averages of four foils each is a conservative approach.
Realistically, the data obtained in the tests are equivalent to
a sample size somewhere between two and eight.  The more conser-
vative approach (data treated as two observations of four foils
each) is the preferred  technique for estimating the overall
precision of the test method.

     In both scenarios, the measured mean TE values are equal;
however, as expected, the standard deviations and hence
coefficients of variation are different.  For example, the
measured mean TE values ranged from a low of 10.2 percent for
the air atomized conventional gun spraying 65-percent solids
paint at a vertical cylinder target, to a high of 93.9 percent
for a high speed rotating bell delivering 65-percent solids
paint to a vertical cylinder target.

     For the standard deviations, however, the two different
data handling scenarios produced noticeably different results.
In the first case (data treated as eight individual observa-
tions), the standard deviations ranged from 0.6 to 5.2 percent.
The standard deviations in the second case (data treated as two
observations of four foils each), however, ranged from 0.04 to
6.1 percent.  Similarly, the coefficients of variation, which
are a measure of the relative precision among test series,
ranged from 1.2 to 5.3 percent for the first case and 0.1 to 6.1
percent for the second case.

                               49

-------
      TABLE  9.   SUMMARY OF  TRANSFER EFFICIENCY TEST RESULTS
CASE 1: CASE 2:
Data treated as 8 Data treated as 2
individual observations observations of 4 foils each
Test Series
AAE-52-VC
52-FP
65-VC
65-FP
AAC-52-VC
52-FP
65-VC
65-FP
HSB-52-VC
52-FP
65-VC
65-FP
30.8
81.2
25.7
81.5
10.8
61.7
10.2
60.8
91.3
91.3
93.1
92.8
S
1.7
1.3
1.6
1.9
0.7
1.2
0.6
1.6
2.8
1.9
5.2
1.1
COV, %
5.5
1.6
6.2
2.3
6.5
1.9
5.9
2.6
3.1
2.1
5.3
1.2
S
1.7
0.2
0.4
1.7
0.1
0.5
0.04
0.04
2.4
0.5
6.1
0.6
COV, %
5.5
0.4
1.6
2.1
0.9
0.8
0.4
0.1
2.6
0.5
6.1
0.7
NOTES:




TEjYj = measured mean transfer efficiency, %




  S = sample standard deviation




COV = coefficient of variation (S/TEvj) x 100, %
                                         50

-------
     Having calculated the standard  deviations  for each test
series, the analyses proceeded to an evaluation of the
poolability of the variances  to determine  if  a  statement could
be made concerning the overall precision of the test  method.
Bartlett's Test was used  to examine  the homogeneity of the
variances of the test series  (i.e.,  are the variances of the
test series equal?).

   From the Bartlett's Test it was  found that the variances of
all 12 test series could  not  be considered equal for either of
the cases in Table 9.  The analyses  then proceeded to an
examination of the following  various groups of  data:

     •  Compare poolability of variances for  TE's obtained from
        the same spray equipment  type (4 TE's = 2 targets x 2
        coatings); 3 gun  groups to  be evaluated.

     »  Compare poolability of variances for  TE's obtained for
        the same target configuration (6 TE's = 3 spray equip-
        ment x 2 coatings); 2 target groups to  be evaluated.

     •  Compare poolability of variances for  TE's obtained for
        the same coating  (6 TE's  =  3 spray equipment x 2 tar-
        gets); 2 coatings groups  to  be evaluated.

     •  Several combinations  of the  first  group.

     These analyses were  performed  for-only the second case,
i.e.,  treating the data for each  test series  as two observations
of four foils each.  The  results  of  these  analyses are shown in
Table  10.  The pooled standard deviations  ranged from 0.3 to
3.3; however, the pooled  standard deviation for the maximum
sample size of 16 observations was  2.5.  This value is the
preferred value for use in estimating the  required sample size
for future laboratory tests.

     An examination of the results  in Table  10 shows that the
data were not poolable, primarily due to the  very large standard
deviation  (6.1) for one test  series (high  speed bell, 65-
percent-solids, vertical  cylinder)  coupled with the very low
variance exhibited by the AAC configuration.   As discussed
later, in this particular test series (HSB-65-VC), one sample
resulted in a mean TE of  102.4 percent and the other a mean of
93.9 percent, with a resultant average value  of 98.1 percent.
Since  transfer efficiencies greater than  100   percent are
impossible, the mean TE of 102.4  percent  is  clearly in error.
However, .the data exceeding  1QO percent could not  be  rejected on an outlier
basis.  Sufficient data points were hgt available  to  perform an outlier
evaluation to permit removal of the data from the  statistical analysis.
                                51

-------
     TABLE 10.  RESULTS OF BARTLETT'S TEST FOR EVALUATION OF POOLABILITY
                OF VARIANCES AT 95% LEVEL OF CONFIDENCE              	
Test Series
 Evaluated

  I. All test series
                                  Pooled
  No. of        No. of           Standard
Test Series  Observations (1)   Deviation
     12
24
 Not poolable
 II. Groups of test series

     A. Sane spray equipment type

        1. Air atomized electrostatic
        2. Air atomized conventional
        3. High speed bell

     B. Sane target configuration

        1. Vertical cylinders
        2. Flat panels

     C. Same coating

        1. 52% wt solids
        2. 65% wt solids
      4
      4
      4
      6
      6
      6
      6
 8
 8
 8
12
12
12
12
    1.2
    0.3
    3,3
Not poolable
    0.8
    1.2
Not poolable
III. Combinations of groups

     A. Group Al & A2
     B. Group Al & A3
     C. Group A2 & A3
      8
      8
16
16
16
Not poolable
    2.5
Not poolable
   (1)  Data treated as two observations of four  foils each.
                                     52

-------
Coating and Equipment Specifications and Equipment
Operating Conditions

     Tables 3, 11, and 12 show the specifications of the equip-
ment and the equipment operating conditions used in these tests.
Table 6 shows the coating specifications.

     Table 3 shows the specifications of the three types of
paint spray equipment used in this test program.  In addition,
the table provides information on the peripheral equipment used
in the tests.

TRANSFER EFFICIENCY TEST RESULTS

Air Atomized Electrostatic Spray Equipment Test Results

     Table 13 summarizes the results of all the tests for the
air atomized electrostatic paint spray equipment.  Transfer
efficiencies ranged from 23.5 to 33.7 percent for the foils on
the vertical cylinders to 78.5 to 84.4 percent for the foils on
the flat panel targets.  The resulting standard deviations were
low  and ranged from 0.2 to 1.7, indicative of very good
precision.

     In addition to these results, one supplementary test was
performed in which four additional vertical cylinder foils and
three semi-tubular target foils were coated using the high sol-
ids coating.  The supplementary test was performed because the
first sample, AAE-HS-VC-1 to 4, in particular, had not used the
foil attachment technique shown in Figure 7-  As a result, wet
paint was deposited on masking tape used to hold the rear seam
down.  This tape was then removed, subsequently affecting the
dry solids weight for the cured foil.  Figure 8 shows the
difference between the vertical cylinder and semitubular target
configurations.

     If this additional observation is included in the AAE-65-VC
test series, the new calculated results are as follows:

    •  TEM = 27.9%

    •    S =  3.9

    •  COV = 14.0%

     These results are based on the four foils in the third
observation having individual TE's of 33.9, 31.5, 32.1, and 32.1
percent.  The new standard deviation and COV are significantly
higher than the results of the other test series for the air
atomized electrostatic spray gun.  The bias introduced by the
tape removal problem (discussed above) clearly contributed to


                               53

-------
          TABLE  11.    SUMMARY OF  EQUIPMENT  OPERATING  CONDITIONS  FOR  AIR ATOMIZED
                            ELECTROSTATIC  AND  CONVENTIONAL  SPRAY  GUNS
                                                                  Air Atomized Electrostatic
                                                                                                            Air Atomized Conventional
 A.   Paint Spray Equipment (1)
     1.   Paint Pressure at Paint  Pot, kPa (psig)
     2.   Paint Pressure at Spray  Gun, kPa (psiy)
     3.   Atomizing Air Pressure at Spray Gun, kPa (psig)
     4,   Operating Voltage,  kV
     5.   Paint Temperature at Paint Pot, "C (°F)
     6.   Gun to Target Distance,  in (an)

B.   Paint Spray Hooth
     1.   Ambient Temperature,  °C  ("F)
     2,   Relative Humidity, %
     3.   Air Flow Velocity, m/s (fpm)
     4.   Air Flow Direction

C.   Target Parameters
     1.   Average Wet Film Thickness, 10~*» meters  (mils)
     2.   Vertical Paint Coverage, cm (in)
     3.   Target Height,  cm (in)
     4.   % Vertical Coverage
     5.   Resistance to Ground, Ohm

D.   Forced Draft Oven
     1.  Cure Time,  sec (min)

        a.   Foil Dish (sample)

        b.   Target Foil

    2.  Cure.Temperature,  °C (°F)
        a.   Foil Dish (sample)
        b.   Target  Foil

E.  Conveyor Speed  Setpoint, m/s  (fpm)
                                                              Low Solids

                                                              294 (28)
                                                             Not measured
                                                              446 (50)
                                                                   45  (2)
                                                                23 (74)
                                                          12 (30); 13  (33)
(3)
  High Solids

   252  (22)
  Not measured
   446  (50)
      45  (2)
    23  (74)
12 (30); 13  (33)
   Low Solids

   294 (28)
  Not  measured
  446-453  (50-51)
 Not applicable
    23 (74)
12 (30); 13  (33)
  High Solids

   225 (18)
  Not measured
  453 (51)
 Not applicable
    23  (74)
12 (30);  13  (33)
                                                                23 (74)               24 (75)
                                                                   52                   49
                                                       0.762-1.02 (150-200)  0.762-1.143 (150-225)
                                                            Normal to target      Normal to target
                                23  (73)  '
                                   54
                        0.762-1.02  (150-200)
                            Normal  to  target
                                               23  (73)
                                                54
                                       0.762-1.02  (150-200)
                                           Normal  to target
                                                        i.27-2.54 (1/2 to  1)
                                                             Not measured
                                                             86-89 (34-35)
                                                               75 (est)
                                                             < 1 million
                                                             1600 (30)

                                                             1800 (30)
                                                              121 (250)
                                                              121 (250)

                                                              0.15 (30)
3.81 (1  1/2)
  Not measured
  86-89  (34-35)
    75 (est)
  < 1 million
     600, 1200, 1800 (4)
       (10, 20, 30)
         600  (10)
         177  (350)
         177  (350)

         0.15  (30)
                          J.81  (1  1/2)     5.08-6.35 (2 to 2 1/2
                             Not  measured         Not measured
                             86-89  (34-35)        86-89 (34-35)
                                75 (est)             75 (est)
                              < 1  million          < 1 million
   1800 (30)

   1800 (30)


  12l'(250)
  121 ,(250)

  0.15 (30)
                                       600, 1200,  1800  (4)
                                          (10, 20,  30)
                                              10
                                           177 (350)
                                           177 (350)

                                           0.15 (30)
(1)  Paint heaters not used in these  tests.
(2)  Measured at gun tip with paint in lines.
(3)  12  inches for vertical cylinder  target; 13 inches for flat panel  target.
(4)  Some sample dishes were baked for 20 and 30 minutes (refer to discussion in this report).

-------
             TABLE  12.    SUMMARY  OF  EQUIPMENT  OPERATING  CONDITIONS  FOR  HIGH  SPEED  BELL
ui
01
Vertical Cylinders


Low Solids
High Solids
Flat
Low Solids
Panel
High Solids
A. Paint Spray hijuipment (1)
1.
2.
3.


4.
5.
6.
Turbine Air Pressure, KPa (psig)
Operating Voltage, M
Disk or Bell Speed, rps (rpm)
a. With faint Applied
b. Without Paint Applied
Shaping Air for Bell, kPa (psig)
Paint Temperature at Paint Pot, "C (°F)
Gun to Target Distance, cm (in)
308 (30)
80 (2)

Not measured
333 (20,000)
308 (30)
21 (70)
30 (12)
308 (30)
80

Not measured
333 (20,000)
308 (30)
21 (70)
30 (12)
308 (30)
80

Not measured
333 (20,000)
308 (30)
21 (70)
JO (12)
308 (30)
80

Not measured
333 (20,000)
308 (3U)
21 (70)
30 (12)
B.  Paint spray booth
    1.  Ambient Temperature,  °C (°F)
    2.  Relative Humidity,  %
    3.  Air Flow \telocity,  m/s (fpm)
    4.  Air Flow Direction

C.  Target  Parameters
    1.  Average Wit Film Thickness, 10~6 meter (mils)
    2.  Vertical Paint Coverage, cm (in)
    3.  Target Height, on (in)
    4.  * Vertical Coverage
    5.  Resistance to Ground, Ohm
         24 (75)               21  (70)              21 (70)              21 (70)
            50                   56                   56                  b6
0.635-0.889 (125-175)  0.508-0.762  (100-150) 0.508-0.762 (100-150) 0.508-0.762 (100-150)
     Normal to target      Normal  to target      Normal to target     Normal to target


  3.81  (1 1/2)      3.81-5.08 (1 1/2 - 2)   3.81-5.08 (1 1/2 - 2)     3.81 (1 1/2)
       Not measured           Not measured         Not measured         Not measured
        142 (56)               142 (56)             142 (56)             142 (56)
        90 (est)               90  (est)             90 (est)             90 (est)
     .< 1 million           <  1 million          < 1 million          <  1 million
          D.
              Forced Draft oven
              1.  Cure Time, s (mint






E.
a. Foil Dish (sample)

b. Target Foil
2. Cure Temperature, °C (°F)
a. Foil Dish (sample)
b. Target Foil
Conveyor Speed Setpoint, m/s (fpm)
1800 (30)

1800 (30)

121(250)
121(250)
9.1(30)
600, 1200, 1800 (3)
(10, 20, 30)
600 (10)

177(350)
177(350)
9.1(30)
1800 (30)

1800 (30)

121(250)
121(250)
. 9.1(30)
600, 1200, 1800 (3)
(10, 20, 30)
600 (10)

177(350)
177(350)
9.1(30)
          (1) Paint heaters not used in these tests.
          (2) Measured at bell high voltage take-off connection with paint  in lines.
          (3) Sooc sample dishes were baked for 20 and 30 minutes (refer to discussion in this report).

-------
U1
                 TABLE  13.    SUMMARY OF  TE  TEST  RESULTS  FOR  AIR  ATOMIZED  ELECTROSTATIC
                                   SPRAY  EQUIPMENT
                                                                Transfer Efficiencies, % (1)
                                                                                                                         Coeffi-
                                      Foil No.                             Foil No.                           Std Devi-   cient o£
 Coating      Target                                                                                          ation,    Variation,
Ht Solids   Configuration       1111   TElU)    ^      j>        II   TE^O) TCfi(4)     (5)         t (6)


    52     Vertical Cylinder   31.0   28.8   29.2   29,6    2«.7    33.7    32.2     30.3    31.8   32.0   30.8     1.66        5.5


    52     Hat Panel          82.5   82.5   81.4   79.2    81.4    80.5    80.5     80.5    82.8   81.1   81.2     0.23        0.2


    65     Vertical Cylinder   26.9   2-1.2   25.8   26.9    26.0    27.4    24.0     23.5    26.8   25.4   25.7     0.37        1.6


    65     Flat Panel          81.2   80.6   80.6   78.5    80.2    80.4    84.4     82.7    83.2   82.7   81.5     1.73        2.1


(1)  Values  are shown rounded to nearest 0.1%; statistical analyses were performed with values rounded to nearest  0.01%.

(2)  TEj  is  Observation No.  1 and represents the arithmetic average of foil Nos. 1, 2, 3 and 4.

(3)  TE2  is  Observation No.  2 and represents the arithmetic average of foil Nos. 5, 6, 7 and 8.

(4)  Ttfj  is  the neasurerl mean TE for each test series and represents the arithmetic average of Observation No. 1 (TEi) and Observation
    No.  2 (TE2>.

(5)  For  these tests which assume two observations,  see  test series? the sample standard deviation  is as follows?   S =  ((TEi -
           (6) The coefficient of variation  (COV) is determined as follows!  OOV = SA^M? standard deviations were rounded to nearest 0.1 before can-
              put ing the Cov.

-------
      3.2cm
       (I3* in.)
~H
      (l in.)
      Vertical.
    Cylinder (VC).
      Target
  Semitubolar (STi
     Target
Figure  8.   Comparison of Vertical Cylinder  and
            Semitubular Target  Configurations
                        57

-------
the calculation for the mean transfer efficiency and the
relatively large standard deviation for this test series.

     As expected,  due to the increased exposed surface area, the
semitubular target had a higher transfer efficiency than the
vertical cylinder  target.  The three foils for the semitubular
target had individual TE's of 34.5, 35.7, and 40.5, for an
average of 36.9 percent.

     The coefficients of variation for the test series in
Table 13 ranged from 0.2 to 5.5 percent, indicative of good
relative precision.  The COV for the AAE-65-VC test series using
three observations instead of two was a poor 14 percent.

Air Atomized Conventional Spray Equipment Test Results

     Table 14 summarizes the results of all the test series for
the air atomized conventional paint spray equipment.  Transfer
efficiencies ranged from 9.7 to 11.9 percent for the foils on
the vertical cylinders to 58.6 to 63.4 percent for the foils on
the flat panel targets.  The resulting standard deviations were
very low and ranged from 0.04 to 0.5, indicative of excellent
precision for these test series.  Similarly, the relative preci-
sion expressed by  the COV's was excellent and fell in the range
of 0.1 to 0.9 percent.

Hj.gh Speed Bell Spray Equipment Test Results

     Table 15 summarizes the results of all the test series for
the high speed bell coating equipment.  Transfer efficiencies
ranged from 86.7 to 106.3 percent for the foils on the vertical
cylinders to 89.3  to 94.4 percent for the foils on the flat
panel targets.  The resulting standard deviations were rela-
tively low and ranged from 0.5 to 6.1, indicative of good
precision for these test series.  Similarly, the COV's were
relatively low and ranged from 0.5 to 6.2 percent.

     Note that the transfer efficiency in one of the vertical
cylinder/65-percent solids test series was over 100 percent.
Transfer efficiencies over 100 percent are clearly erroneous and
most likely are the result of bias in one or more of the
following measurements:^

     •  Incorrect  end weight for platform scale (mass flow rate
        calculation)

     •  Incorrect  stopwatch reading (mass flow rate calculation)

     •  Incorrect  conveyor time reading (conveyor speed measure-
	ment calculation)
1Refer to Section 6 for further explanation.

                               58

-------
                 TABLE  14.    SUMMARY OF  TE  TEST  RESULTS  FOR  AIR  ATOMIZED  CONVENTIONAL
                                   SPRAY  EQUIPMENT
ui
Coating
Wt Solids

Foil No.
Target
Configuration 123
Transfer Efficiencies
4 TCi(2) 5
, % (1)
Foil No.
678 TE2(3) TEj)
Std Devi-
ation,
(4) (5)
Coeffi-
cient of
Variation,
» (6)
    b2     Vertical Cylinder   11.9   10.9   10.4   10.2    10.9    10.5    11.7    10.5    10.0   10.7   10.B     0.12        0.9


    52     Flat Panel          61.9   62.4   60.4   60.9    61.4    63.4    63.2    60.9    60.7   62.1   61.7     0.46        0.8


    65     Vertical Cylinder    9.8    9.8   10.8   10.3    10.2    11.3    10.2     9.7     9.7   10.2   10.2     0.04        0.1


    65     Flat Panel          61.3   61.9   58.6   61.3    60.8    58.6    61.3    60.8    62.4   60.8   60.8     0.04        0.1


(1)  Values are shown rounded to nearest 0.1»; statistical analyses were performed with values rounded to nearest O.OH.

(2)  TE^ is Observation No. 1 and represents the arithmetic average of  foil Nos.  1, 2, 3 and 4.

(J)  TE2 is Observation No. 2 and represents the arithmetic average of  foil Nos.  5, 6, 7 and B.

(4)  Tfc^ is the measured nean TE for each test series and represents the arithmetic average of Observation No. 1 (TEj) and observation No. 2
    (TE;;).

(5)  tor these tests which assuire two observations, see test series; the sample standard deviation is as follows:  S »  l(TEi - T
         (6) The coefficient of variation (COV) is determined as follows:  COV = S/Tk^; standard deviations were rounded to nearest  0.1 before com-
            puting the CUV.

-------
                   TABLE  15,    SUMMARY  OF  TE  TEST  RESULTS  FOR  HIGH  SPEED  BELL  COATING
                                     EQUIPMENT	
                                         	Transfer Efficiencies, % (1)	

                                                                                                                                     Coetti-
                                                 Foil No.                             Foil No.                           Std Uevi-   cient of
           Coatiny       Target                                                                                          ation.     Variation,
          hit Solids   Configuration        1111    TEj<2)    11        11  TE^P) TE^H)    (5)         %  (bj


              S2     vertical Cylinder   93.9   91.4   92.8    93.9    93.0    90.2    86.7     87.8    93.5    89.6   91.3     2.44        2.6


              52     Flat Panel         94.0   89,3   89.3    91.2    91.0    94.2    90.6     91.7    90.3    91.7   91.3     0.53        0.5


              65     Vertical Cylinder  100.0  100.9  102.5   106.3   102.4    92.8    92.8     91.9    97.9    93.9   98.1     6.07        6.2


              65     Flat Panel	93^   92.2   93.1    94.4    93.2    93.6    92.4     90.5    92.7    92.3   92.8     0.64	0^6	
OT>
°        (1) Values are shown rounded to nearest 0.1%; statistical analyses were performed with values rounded  to nearest 0.01%.

          (2) TEj is Observation No.  1 and represents the arithmetic average of foil Nos.  1,  2,  3 and 4.

          (3) TE;> is Observation No.  2 and represents the arithmetic average of foil Nos.  5,  6,  7 and 8.

          (4) Tt+j is the iteasured maan TE for each test series and represents the arithmetic  average of observation  No.  1 (TEj)  and Observation
              No. 2 (TE2).

          (b) For these tests which assume two observations,  see test series; the sample standard deviation is as follows!  S »  ((TEj - Tt^j)2  +
              (TE2 - TEM»2)0.5

          (6) The coefficient of variation (COV) is determined as follows!  COV = S/TEMi standard deviations were rounded to nearest 0.1 before com-
              puting the COV.

-------
     •  Incorrect foil weight before spraying (net solids
        deposited calculation)

WEIGHT PERCENT SOLIDS TEST RESULTS

     Table 16 summarizes the results of the weight percent
solids tests performed as part of the transfer efficiency
measurements.  The high solids coating used in these tests had
an average weight solids content of 64.8 percent and a standard
deviation of 2.4 percent for 19 samples.  The low solids coating
had an average weight solids content of 52.0 percent and a
standard deviation of 1.9 for 10 samples.  These average values
were used in the calculations for transfer efficiency.

MASS FLOW COMPARISON TEST RESULTS

     Tables 17, 18, and 19 summarize the specifications for the
test equipment, paint, and paint spray and peripheral equipment
used in the mass flow comparison test.  Table 20 summarizes the
results of the nine test runs performed for Method A (platform
scale weight change) and Method B (capture unatomized output).
Using this information, statistical analyses were performed to
determine if both mass flow measurement techniques yielded
results that could be considered essentially equal.  Two
independent statistical techniques showed that the two mass flow
measurement techniques yielded significantly different
results.

     The first method applied to evaluate the test data was the
Wilcoxon test for two independent samples.  In applying this
method, it was decided not to adjust the flow rate in Method A
to account for the air weight change in the pressure pot.  This
adjustment is not recommended since the air weight change is
approximately 1.4 g per 500 g paint sprayed and represents only
28 percent of the 5-g accuracy and resolution of the platform
scales.

     The second method applied to evaluate the test data was the
t-statistic.  The same data were used as for the Wilcoxon eval-
uation.  The results show that the measured mean flows, 2.902 x
10~3 and 3.021 x 10~3 kg/s, are significantly different with
a significance level of 0.0005, i.e., 0.05 percent.

     As a result of this comparison, Method A was selected as
the most representative mass flow determination method.  Method
A is used in the Draft Transfer Efficiency Test Procedure.
                               61

-------
TABLE 16.  SUMMARY OF WEIGHT SOLIDS TEST RESULTS
High Solids Coating (1)
Sample
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
(1)
(2)
Cure
Time, s (min)
1200 (20)
1200 (20)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
600 (10)
1200 (20)
1200 (20)
1200 (20)
1800 (30)
1800 (30)
1800 (30)
Ave.
S
Cure temperature =
Cure temperature =
Low Solids Coating (2)
Weight Cure vfeight
Solids, % Time, s (min) Solids, %
64.9
64.7
67.3
69.2
65.1
62.1
65.6
66<, 7
61.4
62.3
66.7
66.7
64.0
66.7
66.7
64.4
64.2
59.3
64.1
64.8%
2.4
177°C (350°F);
121°C (250°F);
1800
1800
1800
1800
1800
1800
1800
1800
1800
1800









(30) 51.9
(30) 54.2
(30) 48o2
(30) 50.0
(30) 53.7
(30) 52.0
(30) 53.7
(30) 53.7
(30) 51.1
(30) 51.9









Ave. = 52.0%
S = 1.9
all samples.
all samples.
                          62

-------
   TABLE 17.  TEST EQUIPMENT SPECIFICATIONS FOR MASS FLOW COMPARISON TESTS
A.  Laboratory Scale



    1.  Manufacturer



    2.  Model No.



    3.  Serial No.



    4.  Capacity/ g



    5.  Rated Accuracy, g





B.  Platform Scale



    1.  Manufacturer



    2.  Model No.



    3.  Serial No.



    4.  Capacity, g



    5.  Rated Accuracy, g






C.  Capture Container



    1.  Type



    2.  Approximate dry weight, g





D.  Stop vtotch



    1.  Manufacturer



    2.  Model No.



    3.  Serial No.



    4.  Capacity, g



    5.  Rated Accuracy, g
           Mettler
           P1200N
           558312
            1,200
              0.01
          NCI, Inc.
             5780
          C790482
               90 (1)
              + 5.
          Paper Cup
    Markson Science, Inc.



	Digital LCD



	26119	



	Not available	



	 + 0.001% of reading
(1) Fitted with 45-kg load cells.
                                     63

-------
 TABLE  18.   PAINT SPECIFICATION FOR MASS FLOW COMPARISON TESTS
 1.   Paint  Type






 2,   Resin  Type






 3.   Manufacturer






 4.   Manufacturer"r Paint ID No.






 5.   Lot No.





 6.   Color






 1.   Recommended Cure Schedule






 8.   Viscosity





 9,,   Reducing Solvent






10.   % Reduction to Spray
                  Enamel
                 Polyester
                    PPG
               AG452D1331-A
               Not Available
                  White
                Not Required
                Not Required
                  Xylene
                  Unknown
11.   Viscosity - Spray
12.   Wt/Gallon - Spray
13.  Wt Solids - Spray
14.  Resistance or Conductance
19 sec wi/»3 Zahn cup @ 23°C  (74°F)
       1.33 kg/L (11.1 Ib/gal)
                  Unknown
                  Unknown
                               64

-------
        TABLE 19.  EQUIPMENT SPECIFICATIONS  AND  OPERATING
                   CONDITIONS FOR MASS  FLOW  COMPARISON  TEST
A.  Paint Supply Tank

    1.  Type

    2.  Manufacturer

    3.  Model No.

    4.  Serial No.

    5.  Rated Capacity, gal


B.  Paint Spray Equipment

    1.  Type

    2.  Manufacturer

    3.  Model  No.

    4.  Serial  No.

    5.  Rated  Capacity,  cc/min

     6.  Air Cap

     7.   Fluid Tip

     8.   Needle


 C.  Equipment Operating Parameters

     1.   Atomizing Air Pressure at
         Spray Gun, psig

     2.   Paint Pressure at Paint
         Pot

     3.   Initial Pot Loading
    Pressurized Pot
        Binks
       83-5501
      Not Available
         AAE
       Nordson
        AN-8
     Not Available
      Not  Available
          987
          228
      Not Available
9.38 kPa (50 psig)
6.77 kPa (32 psiq)

 3.8L (1 gal)
                               65

-------
                       TABLE 20.  MASS FLOW COMPARISON TEST RESULTS
Method A:
Test Run
No.

A-l
A-2
A-3
A-4
A-5
A- 6
A-7
A- 8
A-9

Method Bs
Test Run
No.

B-l
B-2
B-3
B-4
B-5
B-6
B-7
B-8
B-9

(Platform scale;
Beginning
Wt, kg

0
0
0
0
0
0
0
0
0

(Laboratory scale
Beginning
Wt, kg

0
0
0
0
0
0
0
0
0

atomized.. output )
Ending
Wt, kg

0.505
0.505
0.505
0.500
0.505
0.505
0,505
0.500
0.500

I unatomized output)
Ending
Wt, kg

134.17
135.44
133087
134;71
135.49
131.30
131.03
140.82
137.85


Elapsed
Time, s

174.23
173.81
172.01
172.06
174.59
173.23
174.80
169.40
171.99


Elapsed
Time, s

44.33
44.99
44.97
45.08
45.14
43.81
44.09
45. 11
44.51


Flow Rate,
10-3 kg/s (1)
(g/min )
2.898 (173.91)
2.905 (174.33)
2.936 (176.15)
2.906 (174.36)
2.892 (173.55)
2.915 (174.91)
2.889 (173.34)
2.952 (177.10)
2.907 (174.43)
Mean Flow = 2.902 (174.68) (2)

Flow Rate,
10-3 kg/s (1)
(g/min)
3.027 (181.60)
3.010 (180.63)
2.977 (178.61)
2.988 (179.29)
3.002 (180.09)
2.997 (179.82)
2.972 (178.31)
3.122 (187.30)
3.097 (185.82)
Mean Flow = 3.021 (181.27) (2)
(1)  Method A results not corrected for pressurized paint pot air weight change.
(2)  Arithmetic mean values for nine observations.

-------
                           SECTION 6

                    PHASE II LABORATORY TEST
FACILITIES

     The Phase II Laboratory test runs were done using the
original test plan.  They were run in an open booth approxi-
mately 4.6 x 7.6 meters  (15 feet by 25 feet) with the exhaust
fan along the back wall normal to the targets.  Paint spray and
peripheral equipment specifications used for this test are shown
in Table 21.  Every effort was made to duplicate the equipment
specifications used in the original test.  The equipment rated
accuracy matched or exceeded the requirements of the Test Plan
(Table 5).

     For the voltage measurements, Ransburg supplied a Singer
Kilovolt Meter, which does not draw current while sensing the
voltage.

DESCRIPTION OF PAINTS

     PPG supplied two paints from the same batch as the first
test.  These paints were recut to test specifications at PPG
before shipment.  Table  22 documents the paint properties
measured at Ransburg prior to the test.  In the Phase I test,
the paint viscosities were 18 seconds (Ford C4 cup) and 14.5
seconds  (Zahn C3 cup) for the 51-percent and 67-percent solids
paints, respectively.  PPG reports identical results for the
paint they supplied for  Phase II.  Values of 22 and 20.5
seconds were measured for the corresponding paints at Ransburg.
Since agreement with specifications was documented by PPG before
shipment, the difference is attributed to small variations from
cup to cup and in measurement technique.

     A Ransburg Model 234 meter was used to measure the resis-
tance of the paint.  The resistance was 1.5 megohms for the 51
percent solids paint, and 2.5 megohms for the 67 percent solids
paint.  These resistances also differ from the original test
observations.

     Paint weight percent solids determinations were made
according to the Test Plan and ASTM D-2369-81 (see Section 4).
The paint manufacturers' recommended cure schedules were used
for the weight percent solids determination and for curing

                               67

-------
                    TABLE  21.   SUMMARY  OF REPORTED  PAINT SPRAY AND  PERIPHERAL
                                 EQUIPMENT SPECIFICATIONS  FOR RANSBURG TEST (1)
CD
A 1 r Atom 1 zed
E 1 ectrostat 1 c
A. Pa
1.
2.
3.
4.
5.
B. Pa
1.
2.
3.
4.
5.
6.
7.
a.
9.

C. Pa
1.
2.
3.
4.
5.
Int Supply Tank
Type
Manu f actur er
Model No.
Serial No.
Rated Capacity, gallons
Int Spray Equipment
Type
Ma nu f actur er
Model No.
Serial No.
Rated Capacity, cc/mln
A 1 r Cap
Fluid Tip
Need le
Power supply

Int Spray Booth
Type
Ma nuf act urer
Model No.
Serial No.
Rated Capacity, cfm
Press ur e
Bl nks
83-5508
Not ava 1
2

Air

a torn
Pot
fa b 1 e

el

ec
A 1 r Atom 1 zed
Convent 1 ona 1
Pressure
Bl nks
83-5508
Not ava 1 1
2

Air
Nordson

Not
Not


Not
EPU-8;
150 u


Not
Not
Not
Not
AN-8
ava 1
ava 1
987
228
ava 1

1 ab
lab

le
le


Not

a torn


High Speed
Bel 1
Pot Positive dlsplacemen
Ransbur g
9966-9
able 062702
2

conv
Bl nks
21 {
361 1
ava 1
2
3
1
)

able
63 PB

1 ab

1 e
76 kV rated ;
amp

Open
ava 1
ava 1
ava 1
ava 1

Not
Not
63 C
ava 1
appl

1
1

abl e
cable

High speed be 1
Ransbur g
Turbobe 1 1
Not ava liable
Not aval (able
Not appl 1 ca b 1
Not appl 1 cab 1
Not appl 1 ca b 1
Not ava liable

t pump

1




e
e
e

rated


1 ab
lab
lab
lab


le
le
le
le


Not
Not
Not
Not

Open
ava I
ava 1
ava 1
ava 1


1
1
1
1


able
able
able
able

Open
Not ava 1 1 ab 1 e
Not available
Not a va 1 1 ab 1 e
Not aval (able






             Conveyor
             t. Type
             2. Manufacturer
             3: Model No.
             4. Serial  No.
             Forced  Draft  Oven
             1. Type
             2. Ma nu factur er
             3. Model No.
             4. SerI a I  No.  	
    Overhead
 Rlchards-WIIcox
  Not available
  Not ava I table

   Electric
Young Brothers
  Not ava I IabIe
  Not aval table
    Overhead
 Rlchards-WIIcox
 Not avaI(able
 Not avaliable

   Electr Ic
Young Brothers
Not  ava I (able
Not  aval IabIe
    Overhead
R I chards-WI I cox
 Not ava liable
 Not ava I lable

     Electrl c
 Young Brothers
 Not a va I (able
 Not aval I a bIe
          (1) No paint  heaters  were  required  for  these  tests
          (2) Model  21  equivalent  to Model  610  except made of different materials.  All Interior
              dimensions  identical.

-------
                        TABLE  22.  SUMMARY OF PAINT PROPERTIES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Paint Type
Resin Type
Manufacturer
Manufacturer's Paint ID No.
Lot No.
Color
Recommended Cure Schedule
Viscosity (uncut)
Reducing Solvent
% Reduction to Spray
Viscosity - Spray
wt/gal - Spray
wt Solids - Spray
Resistance or Conductance
Paint 1
Enamel
Polyester
PPG
W4533D
Not available
Beige
600 s @ 177°C
(10 min e 350°F)
Not available
Solvesso® 100
Not available
20.5 s (2)
Not available
66.8% (4)
2.5 Mf2
Paint 2
Enamel
Acrylic
PPG
DXH82445
Not available
Gold
1800 s @ 121°C
(30 min @ 250°F)
Not available
DxD reduce (1)
Not available
24 s (3)
Not available
51.4% (5)
1.5 Mfi
(1)   DxD reducer composition:   50% acetone, 30% Cellosolve® acetate,
       12% toluene, 8% Solvesso® 150.

(2)   #3 Zahn cup @ 24°C (76°F).

(3)   #2 Zahn cup @ 25°C (77°F),  converts to 22 sec on Ford #4.

(4)   Mean value for 32 weight dish samples.

(5)   Mean value for 15 wt dish samples.
                                    69

-------
the targets.   Weight percent solids determinations were made
between each  test series,  and at the start of each day.

     During this test there was an insufficient supply of small
aluminum dishes for the paint weight percent solids determination,
The ASTM method was compared with Ransburg's own method using
the remaining dishes.  The two methods differ only in the
vehicle the paint is put into for drying and in that Ransburg
uses no solvent. The Ransburg method fabricates a small dish of
aluminum foil and places the paint in it as a thin film for
drying.  The  two techniques yielded almost identical results,
and since the subsequent TE's fell well within our method
standard deviation, the paint samples were not rerun.

TEST PARAMETERS

     Based on the results of Phase I testing (refer to Section
5), the standard deviation for this test method was determined
to be 2.5.  Based on an allowable difference of 2 (absolute)
between the measured mean TE and population mean TE, this
standard deviation indicates a sample size of six observations
for each test series.  Table 23 provides the basis for this
estimation.

     A test matrix was designed to provide six runs for each of
12 configurations.  The 12 configurations are described in
Table 24.  Each run has both the vertical cylinder and flat
panel targets, for a total of eight foils per run (one observa-
tion is the average of four foils).

TEST SEQUENCE (1)

The actual test sequence was:

March 18 - Set up equipment, check out scales

March 21 - Began TE test runs, completed HSB-67-FPVC 1-3

March 22 - Ran AAE-67-FPVC 1-6

           (AAE-67-FPVC-6 used the original VC
           wrapping technique)

March 23 - Ran AAE-51-FPVC 1-8

March 24 - Ran HSB-51-FPVC 1-6

           Ran HSB-67-FPVC 3-6
(1) See Table 7 for nomenclature.


                                70

-------
   TABLE 23.  ESTIMATION OF REQUIRED TE TEST SAMPLE SIZE
              FOR FUTURE LABORATORY TESTS
Estimated Required
Allowable
Error, +/-(!)
0.5
1.0
1.5
2.0
2.5
3.0

0.5
4
1
1
1
1
1
for
1.0
16
4
2
1
1
1
Sample
Standard Deviation
1.5
36
9
4
3
2
1
2.0
62
16
7
4
3
2
Size
of
2.5
96
36
11
6
4
3


3.0
138
62
16
9
6
4
Basis:

   o  95% level of confidence

   o  Sample standard deviation = population standard deviation

   o  Estimated Sample Size (n) = (1.96 x s/e)2
                      where  s  = standard deviation
                        and  e  = allowable error
(1)  The allowable error is the size of the difference between
     the means (measured mean TE and population mean TE) that
     is considered acceptable.
                            71

-------
         TABLE  24.   TEST MATRIX FOR PHASE  II LABORATORY TESTS
                    OF TE STANDARD  TEST METHOD
Spray Coating
Equipment Wt. Solids
51
Air Atomized
Electrostatic
67
51
Air Atomized
Conventional
67
51
High Speed
Bell
67
Target
Configuration
Vert. Cyl.
Flat Panel

Vert. Cyl.
Flat Panel
Vert. Cyl.
Flat Panel

Vert. Cyl.
Flat Panel
Vert. Cyl.
Flat Panel
Vert. Cyl.
Flat Panel
Test Series No. of
Nomenclature Observations (1)
AAE-51-VC
AAE-51-FP

AAE-67-VC
AAE-67-FP
AAC-51-VC
AAC-51-FP

AAC-67-VC
AAC-67-FP
HSB-51-VC
HSB-51-FP
HSB-67-VC
HSB-67-FP
6
6

6
6
6
6

6
6
6
6
6
6
(1)  One observation is  the  average of  four target  foils.
                                72

-------
March 25 - Ran AAC-67-FPVC 1-6

           Ran AAC-67-FPVC 1A-4A using original
           VC wrapping technique

March 28 - Ran AAC-51-FPVC 1-6

           Clean up


FOIL HANDLING PROCEDURES

     The original vertical cylinder  (VC) wrapping technique was
cumbersome during the PPG test.  For the second test, a candy-
cane wrapping technique for VC's was proposed.  The  foil was  to
be secured at each end by 0-rings  (Figure 9).

     The recommended candy-cane wrapping technique was tried  for
several preliminary runs.  This method also was cumbersome and
it allowed underspray through  the  seams onto the VC  post.  It
was abandoned in favor of Ransburg's own technique.

     Ransburg cuts foil to the length of the target, then hand
wraps the foil vertically around the length of each  cylinder
(Figure 5).  The foil is crimped into place by hand.  No tape or
0-rings are required.

     This VC wrapping technique was  used throughout  the Ransburg
tests.  Four runs (16 cylinders) were made using the original
wrapping technique for comparison.

QA/QC PROCEDURES

     All quality assurance/quality control procedures in the
approved Test Plan were followed (See Section 4).  In addition  to
the recommended practices, it  is recommended that all associated
pressure gages be calibrated before  testing.  Calibration
devices are usually standard equipment for testing laboratories
but may be obtained at a reasonable  cost from suppliers.

     Other recommendations are detailed in Section 9.

TEST RESULTS

General

     Transfer efficiency measurements were made for  a total of
12 test series.  These series  encompassed a test matrix of three
types of spray equipment, two  coatings, and two target configura-
tions.  Six runs of eight foils (four on vertical cylinder, four
on flat panel)  were coated to  determine transfer efficiency,
constituting each test series.  The  rationale for selecting a


                                73

-------
              VERTICAL CYLINDER TARGET
 oaring to  secure foil
       3.2cm  (1-1/4 inch)
       diameter aluminum pipe,
   wrapped candy-cane style
   with 6 in.s 1»5 mil, foil
Top View of Cylinder

secure foil with double-
sided tape until 0 ring
is installed
      double sided tape to secure
      foil before 0 ring is placed
    bottom O ring
                  FLAT PANEL  TARGET
15.24cm(6 in.)-*!
foil
\

Flat Panel Target


X


^
s"



h-


s
s

^ douh


\s"


)le -sided tape


Figure  9.   Foil Attachment  Techniques for Vertical
            Cylinder (VC) and Flat Panel  (FP)  Targets
            Proposed for Ransburg Test
                            74

-------
sample size of six observations for each test series was dis-
cussed earlier.  Forty-eight measurements per test series were
obtained by coating eight foils per target assembly for each of
the six runs.  Table 25 summarizes the results of the tests
based on evaluating the data under the following two scenarios:

     •  Each test series is composed of 24 observations from
        six groups of four foils each.

     •  Each test series is composed of six observations; each
        observation is the average of a group of four foils.

     These two different scenarios for evaluating the data are
shown to reconfirm the original assessment of whether or not a
group of four foils should be treated as four separate observa-
tions or as one observation that is the average of four foils.
As outlined in Section 5, treating the test data for each test
series as two observations composed of the averages of four
foils each is a conservative approach.  Realistically, the data
obtained in the tests are equivalent to a sample size somewhere
between 6 and 24.  The more conservative approach (data treated
as six observations of four foils each) is the preferred
technique for estimating the overall precision of the test
method.

     In both scenarios, the measured mean TE values are equal.
However, the standard deviations and coefficients of variation
are different.  For example, the measured mean TE values ranged
from a low of 14.7 percent for the air atomized conventional gun
spraying 67-percent solids paint at a vertical cylinder target,
to a high of over 100 percent (this is explained later) for a
high speed rotating bell delivering 67-percent solids paint to a
vertical cylinder target.

     For the standard deviations, however, the two different
data handling scenarios produced noticeably different results as
expected. In the first case (data treated as 24 individual
observations), the standard deviations ranged from 0.8 to 5.2.
The standard deviations in the second case (data treated as six
observations of four foils each), ranged from 0.2 to 2.4.
Similarly, the coefficients of variation, which are a measure of
the relative precision among test series, ranged from 1.8 to 9.5
percent for the first case and 1.0 to 3.4 percent for the second
case.

     Having calculated the standard deviations for each test
series, the analyses proceeded to an evaluation of the pool-
ability of the variances to determine if a statement could be
made concerning the overall precision of the test method.
Bartlett's Test was used to examine the homogeneity of the
variances of the test series.
                                75

-------
     From the Bartlett's Test it was found that the variances of
all 12 test series could not be considered homogeneous for
either of the cases in Table 25.  The analyses then proceeded to
an examination of the following various groups of data:

     •  Compare poolability of variances for TE's obtained from
        the same spray equipment type (4 TE's = 2 targets x 2
        coatings); 3 gun groups to be evaluated.

     •  Compare poolability of variances for TE's obtained for
        the same target configuration (6 TE's = 3 spray equip-
        ment x 2 coatings); 2 target groups to be evaluated.

     •  Compare poolability of variances for TE's obtained for
        the same coating (6 TE's = 3 spray equipment x 2 tar-
        gets); 2 coatings groups to be evaluated.

     •  Several combinations of the first group.

     These analyses were performed for only the second case,
i.e., treating the data for each test series as six samples of
four foils each.  The results of these analyses are shown in
Table 26. The pooled standard deviations ranged from 1.5 to
2.0; however, the pooled standard deviation for the maximum
sample size of 48 was 1.8.  This value is lower than the 2,5
estimated from the Phase I laboratory test.

     The results on Table 26 do not include HSB-67-FPVC-2,
which was eliminated through an outlier analysis (ASTM E180).
Confirmation runs using the original VC wrapping technique are
also not included.

Air Atomized Electrostatic Test Results

     Table 27 presents the results of six test runs using air
atomized electrostatic paint spray equipment.  Transfer effi-
ciencies for flat panel targets ranged from 87.4 (AAE-51-FP-6)
to 93.4 (AAE-51-FP-2) percent.  The standard deviation was 1.7
to 2.4.  The relative precision of these runs as expressed by
coefficient of variation was 1.9 to 2.6 percent.

     Vertical cylinder TE's ranged from 50.4 (AAE-67-VC-6) to
63.5 (AAE-51-VC-5) percent.  The standard deviation was 1.8 to
2.1.  The coefficient of variation was 3.4 to 3.5 percent.

     A summary of equipment operating conditions for the air
atomized electrostatic test series is shown in Table 28.
                                76

-------
                        TABLE 25.   SUMMARY OF  TE TEST RESULTS
                                   RANSBURG TEST

Test Series
AAE-51-VC
51-FP
67-VC
67-FP
AAO51-VC
51-FP
67-VC
67-FP
HSB-51-VC
51-FP
67-VC
67-FP
RANGE

TEM, %
61.3
90.7
51.5
90.3
15.8
76.5
14.7
75.5
103.0
.101.6
101.0
102.1

CASE 1:
24 indiv
S
2.3
3.7
1.9
2.6
0.8
3.3
1.4
3.2
5.2
1.8
4.7
2.0
0.8-5.2
Eata Treated as
. observations
COV,%
3.8
4.1
3.7
2.9
5.1
4.3
9.5
4.2
5.0
1.8
4.6
2.0
1.8-9.5
CASE 2: C&ta
observations
S
2.1
2.4
1.8
1.7
0.2
0.8
0.5
2.1
1.5
1.4
1.5
1.6
0.2-2.4
treated as 6
of 4 foils ea.
COV,%
3.4
2.6
3.5
1.9
1.3
1.0
3.4
2.8
1.5
1.4
1.5
1.6
1.0-3.4
NOTES:

TEM = Mean Transfer Efficiency

  S = Standard deviation = [TEi-TEM)2 + (TE2-TEM)2 + ...]°*5

COV = Coefficient of Variation (S/TEM) * 100%
                                      77

-------
    TABLE  26.  RESULTS OF BARTLETTS'S TEST FOR EVALUATION OF POOLABILITY
               OF VARIANCES AT 95% LEVEL OF CONFIDENCE            	
                                 PHASE  II  (RANSBURG)
   Test Series
     Evaluated
                                Pooled
   No. of       No. of         Standard
Test Series  Observations(1)  Deviation,%
  I.  All Test  series

 II.  Groups of test series

     A.  Same  spray equipment type
       12
72
Not poolable
1. Air atomized electrostatic
2. Air atomized conventional
3. High speed bell
B. Same target configuration
1. Vertical cylinders
2. Flat panels
C. Sane coating
1. Paint 2 (51% wt)
2. Paint 1 (67% wt)
III. Combinations of groups
A. Group Al & A2
B. Group Al & A3
C. Group A2 & A3
4
4
4

6
6

6
6

8
8
8
24
24
24

36
36

36
36

48
48
48
2.0
Not poolable
1.5

Not poolable
1.7

Not poolable
1.6

Not poolable
1.8
Not poolable
(1)  Data  treated as six observations of four  foils each per  test series.
                                    78

-------
TABLE 27.  TRANSFER EFFICIENCY  DATA,  RANSBURG LABORATORY TEST,
           MARCH 21-29,  1983  (1)

Run 1
Faint
Ut »
Solids
AAE
51
51
67
67
AAC
51
51
-J 67
VO
67
HSB
51
51
67
67
foil No.
Target
vc
FP
vc
FP
vc
FP
vc
FP
VC
FP
VC
FP
1
63.3
88.6
50.9
87.6
16.5
75. B
15.1
70.4
106.6
102.2
103.2
102.5
2
60.6
92.6
52.0
92.0
14.7
76.7
13.7
7J.4
103.1
101.4
96.8
99.3
3
62.3
94. U
51.6
91.6
15.2
72.2
15.1
72.2
99.3
101.7
98.6
98.9
4
63.2
92.1
54.0
90.1
16.4
75.0
13.9
76.0
110.9
103.8
103.5
100.5
TE1 (2) 1
62.4
92.0
52.1
90. J
15.7
74.9
14.5
73.0
105.0
102.3
100.5
100.3
63.2
85.9
52.7
90.3
17.4
68.5
16.9
71.5
104.1
100.4


Run 2
Foil No.
2
61.2
92.4
54.3
94.2
14.9
78.0
14.1
78.7
97.2
98.0


3
61.0
98.5
54.1
95.9
15.7
16.1
14.1
78.9
96.1
98.6
OUTLIER

4
64. U
96.8
56.0
93.0
15.7
82.8
15.5
82.3
108.2
101.8


TE2 (2)
62.6
93.4
54.3
93.3
15.9
76.9
15.2
77.9
101.4
99.7


1
61.2
94.3
52.1
87.2
17.2
79.5
17.7
75.0
105.7
100.6
101.1
103.3
Run 3
Foil No.
2
60.9
94. J
51.4
91.5
15.7
80.0
15.2
77.9
93.4
98.9
102.5
101.0
3
61.2
92.0
50.9
92.3
15.1
74.9
13.4
72.1
96.3
98.5
95.3
99.4
4
63.7
88.8
53.1
90.1
16.1
74.6
15.4
78.0
106.4
101.5
105.1
103.5
TE3 (2)
61.8
92.4
51.9
90.3
16.0
77.3
15.4
75.8
101.7
99.9
101.0
101.8
1
60.4
86.8
48.7
85.6
16.3
73.8
16.3
73.8
106.6
102.9
104.3
102.9
Run 4
Foil No.
2
57.5
90.5
4U.2
89.9
15.8
77.7
14.4
76.1
99.5
102.1
95.1
102.0
3
59.4
93.5
48.8
91.1
14.5
76.1
13.6
73.9
97.3
100.7
94.4
101.7
4
61.2
91.6
50.4
88.6
16.1
78.1
15.8
77.5
109.1
103.5
106.4
104.2
TC4 (2)
59.6
90.6
49.0
88.8
15.7
76.4
15.0
75.3
103.1
102.3
100.0
102.7

-------
                                                          TABLE  27  (continued)
CO
O

Ha i nt
Wt %
solids
AAE
51
51
67
67
AAC
51
51
67
67
MSB
51
51
67
67
Run 5
Foil No.

Target

VC
FP
VC
FP

VC
FP
VC
FP

VC
FP
VC
FP

1

63.5
84.9
49.5
84.7

16.0
70.8
12.7
78.0

108.2
103.1
103.0
102.0

2

61.4
89.0
51.7
90.2

15.7
78.7
17.3
80.9

101.1
101.4
95.6
99.9

3

63.5
91.0
51.3
90.6

15.4
77.7
13.4
77.2

111.3
101.2
94.1
100.4

4

65.7
88.6
51.6
88.3

14.8
81.0
13.4
74.6

97.6
103.2
105.8
101.3

TC5 (2)

63.5
8U.4
51.0
88.5

15.5
77.0
14.2
77.7

104.6
102.2
99.6
100.9

1

57.5
83.6
50.0
88.2

17.3
72.9
15.5
75.1

103.8
104.6
106.6
105.3
Run
Foil

2

57.6
87.5
50.4
90.2

14.9
77.8
13.1
75.1

97.5
102.1
100.3
102.7
6
No.

3

58.6
90.3
50.3
93.2

15.4
74.9
13.6
70.6

97.0
104.0
99.5
104.2


4

58.8
88.3
51.0
89.9

15.9
79.9
14.7
71.9

110.2
101.3
109.2
107.0


TE6 (2) 1

58.1
87.4
50.4(6)
90.4

15.9
76.4
14.2
73.2

102.1
103.0
103.9 103.4
104.8 103.6
Run 7
Foil No.

2 3 4 TE7 (2) TEm (3) S (4) COV (5)

61.3 2.1 3.4
90.7 2.4 2.6
51.5 1.8 3.5
90.3 1.7 1.9

15.8 0.2 1.3
76.5 0.8 1.0
14.7 0.5 3.4
75.5 2.1 2.8

103.0 1.5 1.5
101.6 1.4 1.4
94.9 97.0 108.1 101.0 101.0 1.5 1.5
99.5 101.2 103.3 101.9 102.1 1.6 1.6
        (1) TE calc'd to 0.1%, statistics calc'd to 0.1%.    (4) S=  standard deviation =  [(TEl-TUn)2 + (TE2^rE,)2 + ...j
        (2) TEn= arithjietic avg o£ foils (l-4)n.            (5) COV= coefficient of variation = S/TEm X 100.
        (3) Tthi= arithmetic avy of all TEn.                (6) This run using Tape Wrap Tech.

-------
             TABLE 28.  AAE PAINT SPRAY AND  PERIPHERAL
                        EQUIPMENT SPECIFICATIONS    	
A. Paint Supply Tank

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, gallons
   Pressure Pot
      Binks
    83-5508
B. Paint Spray Equipment

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cc/min
   6. Air Cap
   7. Fluid Tip
   8. Needle
Electrostatic Hand Gun
    Nordson	
     AN-8
     987
     228
C. Paint Spray Booth
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity,  cfm
    Open
D. Conveyor
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   Over-Head
  Richards-Wilcox Mfg,
   Forced Draft Oven
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.

   Paint Heaters
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   Electric
  Young Brothers
                               81

-------
                        TABLE 28 (continued)
                 AAE EQUIPMENT OPERATING CONDITIONS
A.  Paint Spray Equipment

   1. Paint Pressure at Paint Pot, kPa (psig)   5.32  (22)
   2. Paint Pressre at Spray Gun, kPa (psig)       N/A
   3. Atomizing Air Pressure at Spray Gun,
        kPa (psig)
   4. Operating Voltage, kV                        63
   5. Disk or Bell Speed, rpm                   _
      a. With Paint Applied                     ______
      b. Without Paint Applied              •    _____
   6. Shaping Air for Bell, psig                _____
   7. Paint Temperature at Paint Pot, °F
   8. Gun to Target Distance, cm (in)            30.5  (12)

B. Paint Spray Booth

   1. Ambient Temperature, °C (°F)                 24  (76)
   2. Relative Humidity, %                        19
   3. Air Flow Velocity, ra/s (fpm)               0.5  (100)
   4. Air Flow Direction                    Normal to  target

C. Target Parameters

   1. Average Wet Film Thickness, cm (mils)   5xlO~3  (2.0)
   2. Vertical Paint Coverage	
   3. Target Height, cm (in)                      .
   4. % Vertical Coverage                         45 Jest)
   5. Resistance to Ground, Ohm                   >10b
D. Forced Draft Oven (1)

   1. Cure Time, minutes
      a. Foil Dish (sample)   67% 600 s @ 177°C  (10 min at  350°)
      b. Target Foil         51% 1800 s @ 121°C  (30 min at  250°)

   2. Cure Temperature
      a. Foil Dish (sample)                      See above	
      b. Target Foil                           ^	
   Paint Heaters

   1. Temperature In
   2. Temperature Out
F. Conveyor Speed Setpoint, cm/s    (fps)        15.8  (0.52)
(1) The same cure schedules were used for foils and dishes.
                                 82

-------
Air Atomized Conventional Test Results

     Table 27 summarizes the results of all six test  series  for
air atomized conventional paint spray equipment.   Flat  panel
transfer efficiencies ranged from 73.0  (AAC-67-FP-1)  to  77.9
(AAC-67-FP-2).  Vertical cylinder transfer efficiencies  ranged
from 14.2 (AAC-67-VC-5 and 6) to 16.0 (AAC-51-VC-3).  The
standard deviations were 0.2 to 2.1, indicating good  precision.

     In addition to the six scheduled runs, four additional  runs
were made using the original vertical cylinder wrapping  tech-
nique (Figure 7) with conventional flat panel targets accompany-
ing as controls.  These runs are compared with the  new  (Ransburg)
wrapping method in Section 7.

     A summary of operating conditions  for the air  atomized
conventional equipment is presented in  Table 29.

High Speed Bell Test Results

     Table 27 summarizes the results from seven test  runs on
the high speed bell paint spray equipment.  Test conditions  are
summarized on Table 30.  One set of runs  (HSB-67-FP-VC-2)
was eliminated by an outlier analysis,  leaving six  complete
runs.  Transfer efficiencies for flat panel targets range from
99.7 (HSB-51-FP-2) to 104.8  (HSB-67-FP-6), with a  standard
deviation of 1.4 to 1.6.  The coefficient of variation  is low,
from 1.4 to 1.6 percent.

     Transfer efficiencies for vertical cylinder targets range
from 99.6 (HSB-51-VC-1) to 105.0 (HSB-51-VC-1), with  a standard
deviation of 1.5.  The coefficient of variation is  a  very low
1.5 percent.

     It is obvious that transfer efficiency cannot  exceed 100
percent.  The HSB data points to a consistent system  or method
error in determining TE.  When TE's began appearing at  100+
percent, every possible physical testing error was  examined.
Several more weight percent solids determinations were made,
scales were recalibrated, and new runs were made.   No equipment
error was found, but an explanation for the high TE's was
discovered .

     The original test method called for 400 g of  paint  to be
sprayed before passing the targets in front of the  gun.  In  the
Ransburg test about 100 g was sprayed before the target was
passed through. (Paint was in short supply.)   When  the  100 g
spray was initiated, the targets were about 5 feet  down  the
conveyor in the open booth, moving towards the gun.   Some of
this paint had the capability of remaining suspended  long enough
to eventually coat the targets considering the electrostatic
                               83

-------
             TABLE 29.  AAC  PAINT  SPRAY AND PERIPHERAL
             	EQUIPMENT  SPECIFICATIONS
A. Paint Supply Tank

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, gallons
   Pressure Pot
      Binks
    83-5508
B. Paint Spray Equipment

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cc/min
   6. Air Cap
   7. Fluid Tip
   8. Needle
Conventional Air Spray
    Binks	
     21	
    36113	

     63 PB
     63C
C. Paint Spray Booth
   1. Type
   2. Manufacturer
   3, Model No.
   4. Serial No.
   5. Rated Capacity, cfm
    Open
D. Conveyor
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   Over-Head
  Richards-Wilcox Mfg.
E. Forced Draft Oven
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.

F. Paint Heaters
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   Electric
  Young Brothers
                               84

-------
                        TABLE 29 (continued)
                 AAC EQUIPMENT OPERATING CONDITIONS
                                                 3.73 (11)
                                                  453 (51)
                                                   0
                                                 30.5 (12)
A. Paint Spray Equipment

   1. Paint Pressure at Paint Pot, kPa (psig)
   2. Paint Pressre at Spray Gun, kPa (psig)
   3. Atomizing Air Pressure at Spray Gun,
        kPa (psig)
   4. Operating Voltage, kV
   5. Disk or Bell Speed, rpm
      a. With Paint Applied
      b. Without Paint Applied
   6. Shaping Air for Bell, psig
   7. Paint Temperature at Paint Pot, °F
   8. Gun to Target Distance, cm (in)

B. Paint Spray Booth

   1. Ambient Temperature, °C (°F)
   2. Relative Humidity, %
   3. Air Flow Velocity, m/s (fpm)
   4. Air Flow Direction

C. Target Parameters

   1. Average Wet Film Thickness, cm (mils)
   2. Vertical Paint Coverage, cm (in)
   3. Target Height, cm (in)
   4. % Vertical Coverage
   5. Resistance to Ground, Ohm

D. Forced Draft Oven (1)
                                                 23 (73)
                                                  25
                                                 0.5 (100)
                                                 Normal
                                              7.6xlO-3 (3.0)
                                                 (16)
                                                 varied
                                                  45
   1.  Cure Time,  minutes
      a.  Foil Dish (sample)
      b.  Target Foil
                             67%-60Q s @ 177°C (10 min at 350°F)
                            51%-1800 s § 121°C (30 min at 250°F)
   2. Cure Temperature
      a. Foil Dish (sample)
      b. Target Foil

E. Paint Heaters

   1. Temperature In
   2. Temperature Out

F. Conveyor Speed Setpoint, cm/s (fpm)
                                                See above
                                               16.0 (0.52)
(1)  The same cure schedule was used for foils and dishes.
                                85

-------
              TABLE 30.   HSB PAINT  SPRAY AND PERIPHERAL
                         EQUIPMENT  SPECIFICATIONS
A. Paint  Supply  Tank

   1. Type                                      DC Pump
   2. Manufacturer                             Ransburg
   3. Model  No.                                  9966-01
   4. Serial No.                               ___________
   5. Rated  Capacity, gallons                 _____________
B. Paint  Spray  Equipment

   1. Type                                     High Speed Bell
   2. Manufacturer                              Ransburg	
    3. Model  No.                                  20  865-04
    4. Serial No.	
    5. Rated  Capacity, cc/min	
    6. Air  Cap                                 _______________
    7. Fluid  Tip                                _____________
    8. Needle
    Paint  Spray Booth
    1. Type                                       Open
    2. Manufacturer                            ________
    3. Model  No.                                _,______,
    4. Serial No.                               ________
    5. Rated  Capacity, cfm  (ft3/min)           _______
 D. Conveyor
   1.  Type                                      Over-Head
    2.  Manufacturer                            Richards-Wilcox Mfg,
    3.  Model  No.                                _______________________
    4.  Serial No.
    Forced  Draft Oven
    1.  Type                                     Electric
    2.  Manufacturer                            Young  Brothers
    3.  Model No.	.
    4.  Serial No.
    Paint Heaters
    1.  Type
    2.  Manufacturer
    3.  Model No.
    4.  Serial No,
                                86

-------
                        TABLE  30  (continued)
                 HSB  EQUIPMENT OPERATING  CONDITIONS
A.
B.
C.
D.
Paint Spray Equipment
1. Paint Pressure at Paint Pot, kPa (psig
2. Paint Pressre at Spray Gun, kPa (psig)
3. Atomizing Air Pressure at Spray Gun,
kPa (psig)
4. Operating Voltage, kV
5. Disk or Bell Speed, rps (rpm)
a. With Paint Applied
b. Without Paint Applied
6. Shaping Air for Bell, kPa (psig)
7. Paint Temperature at Paint Pot, °C ( 8F
8. Gun to Target Distance, cm (in)
9. Pump Setting
Paint Spray Booth
1. Ambient Temperature, °F
2. Relative Humidity, %
3. Air Flow Velocity, m/s (rpm)
4. Air Flow Direction
Target Parameters
1. Average Wet Film Thickness, cm (mils)
2. Average Dry Film Thickness
2. Vertical Paint Coverage, cm (in)
3. Target Height, cm (in)
4. % Vertical Coverage
5. Resistance to Ground, Ohm
Forced Draft Oven (1)
1. Cure Time, minutes
a. Foil Dish (sample) 67%-600 s @ 177
b. Target .Foil 51%-1800 s § 121
2. Cure Temperature, °F
a. Foil Dish (sample)
b. Target Foil
)
-
205 (15)
80

167 (10000)
333 (20000)
308 (30)
)
28 (11)
1.45
75°
18
0.5 (100)
Normal to target
S.lxlO-3 (2.0)
1.2
81 (32)
152 (60)
53
<1 million
°C (10 min at 350°F)
°C (30 min at 250°F)
See above

E. Paint Heaters
   1. Temperature In, °F
   2. Temperature Out, °F

F. Conveyor Speed Setpoint, cm/s (fpm)
                                                16  (0.52)
(1)  Same  cure  schedule  for  foils  and  dishes.
                               87

-------
nature of the process.   In the original test, the targets were
outside the closed booth while 400 g were sprayed.  Thus, the
targets in the Ransburg test were probably exposed to more paint
than was measured.

     Table 31 represents an approximation of the TE's when cor-
rected for the paint sprayed 2 feet before the first timing
mark.  This approximation is based on observations during the
tests.  As can be seen, the correction brings HSB TE's into a
range more comparable to Phase I.

     Based on the correction, HSB flat panel TE's were 88,7 to
93.2 percent with standard deviation of 1.2 'to 1.5.  The co-
efficient of variation was 1.3 to 1.7 percent.  Vertical
cylinder TE's were 88.9 to 93.3 percent with standard deviation
of 1.4.  The VC coefficient of variation was 1.5 to 1.6 percent.
                               88

-------
                                 TABLE  31.    TRANSFER EFFICIENCY  DATA




CO
vo

Faint
wt %
Sol ids
51
51
67
67

Target
VC
FP
VC
FP


1
96.2
91.7
91.6
90.7
Run 5
Foil No.
234
89.8 98.9 87.8
90.1 89.9 91.7
85.0 83.7 94.0
88.8 89.2 90.0
Ransburg Laboratory Test March 21-29, 1983
Corrected HSU Transfer Efficiency I&ta (1)
Run 6 Run 7
Foil No.
TE5 (2)
93.2
90.9
88.6
89.7
1
92.2
93.0
94.8
93.6
2
86.6
90.8
89.2
91.3
3
86.2
92.5
88.5
92.7
4 TE6 (2) 1
97.9 90.7
90.0 91.6
97.1 92.4 92.0
95.1 93.2 92.1
Foil No.
2 3 4 TE7 (2) TEm (3)
- - - 91.6
- 90.3
84.3 86.2 96.1 89.7 89.8
88.5 90.0 91.8 90.6 90.8
S (4) COV (5)
1.4 1.5
1.2 1.3
1.4 1.6
1.5 1.7
(1)  These calculations assume that  the paint sprayed 2 ft prior to target arrival has the capability of staying dispersed long enough to attach
    to the targets.  Constant flow  rate and conveyor speed are assumed tor  the additional 2 ft.  (See Section 7 for further discussion.)
    TE calc'd to 0.1», statistics calc'd to 0.1%.
(2) TBn= arithmetic  avg of foils (l-4)n.

(3) TEm= arithmetic  avg of all TEn.
(4) S= standard deviation « l(TEl-TEm)2 + (TE2-TEm)2 + ...)0.5

(5) COV= coefficient of variation = S/TEm X 100.

-------
                                           TABLE   31   (continued)


Paint
hit $
Solids Target 1
51 VC 96.2
51 t'P 91.7
67 VC 91.6
67 fP 90.7
Ransbury Laboratory Test March 21-29, 1983
Corrected HSH Transfer Efficiency L&ta (1)
Run 5 Run 6 Run 7
Foil No. Foil No. Foil No.
234 TE5 (2) 1 2 3 4 TE6 {2) 1 2 3 4 TE7 (2> TCm (3) a (4) CUV (5)
89.8 98.9 87.8 93.2 92.2 86.6 86.2 97.9 90.7 ----- 91.6 1.4 1.5
90.1 89.9 91.7 90.9 93.0 90.8 92.5 90.0 91.6 ----- 90.3 j.2 1.3
85.0 83. 7 94.0 88.6 94.8 89.2 88.5 97.1 92.4 92.0 84.3 86.2 96.1 89.7 89.8 1.4 1.6
a«.8 89.2 90.0 89.7 93.6 91.3 92.7 95.1 93.2 92.1 88.5 90.0 91.8 90.6 90.8 1.5 1.7
(1)  These calculations assume that the paint sprayed 2 ft prior to target arrival  has the capability of staying dispersed long enough to attach
    to the targets.  Constant flow rate and conveyor speed are assumed for the additional 2 ft.   (See Section 7 for  further discussion.)
    TB calc'd to 0.1», statistics calc'd to 0.1».

(2)  TEn= arithjiEtic  avy of foils (l-4)n.     (4) S= standard deviation = ((TEl-TEm)2 *  (TE2-TEro)2 + ...)0-5

(3)  Ttln= arithmetic  avg of all TEn.          (5) COV= coefficient  of variation » S/fEm X 100.

-------
                            SECTION 7

                         TEST COMPARISON


EFFECT OF FOIL WRAP TECHNIQUE - VERTICAL CYLINDER

     The method used to attach the foil to the vertical cylinder
targets during Phase I testing proved highly unsatisfactory.  As
discussed in Section 5, the foil could be removed from the
double-sided tape without losing paint only with great diffi-
culty.  Therefore, it was decided that the "tape wrap" method
would not be used in subsequent tests.  Instead the "crimp"
method described in Section 6 was used in Phase II.

     It was decided to conduct a series of tests to compare
transfer efficiency results obtained with each method of foil
attachment. The objective of these tests was to provide data
necessary for the determination of the extent to which the foil
attachment method influenced the resulting transfer efficiency,
if at all.  The t-test of the difference between two means was
then applied to the resulting data as follows.

     The air atomized conventional (AAC) spray gun was selected
for this purpose in order to minimize the number of observations
required to detect a small difference in transfer efficiency.
The standard deviation of the transfer efficiency results ob-
tained during Phase I testing using the AAC spray gun (taking
each cylinder as an individual observation) had been found to be
the least of any spray apparatus.

     Thus, using the AAC configuration, four runs of four
cylinders each were performed as a special test using the tape
wrap method of foil attachment.  Each of the four runs was made
under the same conditions using the air atomized conventional
gun with the high solids paint.  Resulting transfer efficiencies
are given in Table 32.
                                91

-------
 TABLE 32.  TESTS TO DETERMINE EFFECT OF FOIL ATTACHMENT METHOD
                  AAC-67-VC:  TAPE WRAP METHOD

                       Transfer Efficiency

                         Foil No.
Run No.
TE
1A
2A
3A
4A
14.0
15.1
16.8
15.1
12.5
13.9
14.3
13.7
13.1
13.8
11.8
12.2
13.6
13.8
15.3
13.0
13.4
14.2
14.6
13.5
     These results may be compared with the following values,
reproduced here from Table 27, showing transfer efficiencies
resulting from tests carried out under identical conditions
using the crimp method of foil attachment.
 TABLE 33.  TESTS TO DETERMINE EFFECT OF FOIL ATTACHMENT METHOD
                  AAC-67-VC;  CRIMP METHOD(l)

                       Transfer Efficiency

                         Foil No.
Run No.
TE

1
2
3
4
5
6

15.1
16.9
17.7
16.3
12.7
15o5

13.7
14.1
15.2
14.4
17.3
13.1

15.1
14.1
13.4
13.6
13.4
13 06

13.9
15.5
15e4
15.8
13.4
14 c 7
	 -n
14.5
15.2
15.4
15oO
14.2
14,2
(1)  From Table 27.

     The standard deviation of the 16 observations using the
tape wrap method (Table 32) is 1.25.  The standard deviation of
the  24 observations using the crimp method is 1.37.  The ratio
of variances (F-statistic)  then is 1.19.  Since the critical
value of F at the 0.1 level, for 23 and 15 degrees of freedom,
is 1.9, these variances are homogenous and therefore poolable.
The  pooled standard deviation may be shown to be 1.32.
                               92

-------
     Given a standard deviation of 1.32, the number of observa-
tions required for the t-test to detect a difference of 2 while
controlling both the  a risk and the /5  risk to the 0.05 level
is 13.   Eighteen observations are required to reduce the
risk to 0.01.  With 16 observations of the tape wrap method and
24 of the crimp method, the  /3 risk may be said to be controlled
at a level below 0.05 when the t-test is performed at the 0.05
level.  The  oi risk is the risk of deciding that two samples
were drawn from two different populations when in reality they
came from the same population.  The  ft risk, perhaps more
crucial for the present consideration, is the risk of deciding
that the samples are drawn from the same population when in
reality there is a real difference of at least the magnitude
specified.

     The value of the t-statistic is given by:

         TE  (crimp) - TE  (wrap)
     t = —	 = 1.99
               \l ±- + -L-
               \f 24 + 16
with 38 degrees of freedom.


     The critical value of t at the 0.05 level with 38 degrees
of freedom is 2.03.  Thus, the value of t for this test is not
significant at the specified level.  It is therefore concluded
that the transfer efficiency results from the two methods of
foil attachment do not differ significantly, with the risk of
their actually differing by 2 or more equal to less than 0.05.

     This result is further verified by the single run (Run 6)
performed using the tape wrap technique in the AAE-HS-VC series
(see Table 27, footnote 6).  That run resulted in a transfer
efficiency of 50.4, compared to the mean transfer efficiency for
all 6 runs taken together of 51.5.

COMPARISON OF INTERLABORATORY VARIANCES

     The results from each of the two participating laboratories
may be compared to determine if a significant difference in test
precision exists between laboratories.  Table 34 shows the
standard deviations at each laboratory for each test series,
treating each target as a separate observation.
 Davies, 0. L., The Design and Analysis of Industrial Experi-
  ments, 2nd Edition, Table E-l, pp 609-611.  Imperial Chemical
  Industries Ltd., London.
                               93

-------
     The F-statistic is equal to the ratio of the variances, the
squares of the standard deviations.   Degrees of freedom associ-
ated with numerator and denominator are 23 and 7, or 7 and 23,
depending on whether Phase I results or Phase II results ex-
hibited the greater variance.  The test for significance is a
two-tailed test, because the question being tested is whether
the precision at either laboratory is significantly different
(greater or less) than the precision at the other.

TABLE 34.  COMPARISON OF VARIANCES FOR TESTS AT EACH LABORATORY (1!

AAE-51-VC
* 51-FP
67-VC
67-FP
AAC-51-VC
* 51-FP
* 67-VC
67-FP
HSB-51-VC
51-FP
67-VC
67-FP
ST(8)
1.7
1.3
1.6
1.9
0.7
1.2
0 = 6
1.6
2,8
1.9
5.2
1 = 1
SIl(24)
2.3
3.7
1.9
2.6
0.8
3o3
1.4
3,2
5.2
1.8
4.7
2.0
F
1.83
8.10
1.41
1.87
1.31
7.56
5.44
4.00
3 = 45
1.11
1.22
3.31
F.05
4.4
4.4
4.4
4.4
4.4
4.4
4.4
4 = 4
4.4
2.9
2,9
4.4
(1) Each Target Considered a Separate Observation.

* Significant at 0.05 level


     Table 34 shows that variances for 3 test series are found
to differ significantly between laboratories.  In each of these
three cases, the precision exhibited in the Phase II testing at
Ransburg is poorer.  No significant difference at the 0.05 level
is found for the other nine test series.

     The precision of the test at each laboratory was found to
be 2.5 standard deviation for Phase I and 1.8 standard deviation
for Phase II based on the largest number of poolable observa-
tions in each case.  (For both laboratories it was found that
all air atomized electrostatic and all high speed bell results
could be pooled.)  This pooling was performed using standard
deviations based on treating the mean of four targets as a sin-
gle observation.  In contrast^ Table 34 is based on treating
each target as a separate observation, because the single degree
of freedom associated with Phase I results when treating four
targets as a single observation is not sufficient to provide for
a meaningful test.
                               94

-------
     The pooled estimates of precision for each laboratory may
be tested to determine if-they differ significantly.  The ratio
of variances is (2.5/1.8)  = 1.93, with 15 and 47 degrees of
freedom.  The critical value of F for a double-tailed test at
the 0.05 level is 2.12.  Thus, the overall test precision
(repeatability) at each of the two laboratories is not signifi-
cantly different.   Pooling the two estimates of standard devia-
tion gives an overall test precision of 2.0.

ANALYSIS OF VARIANCE FOR PAINT, LABORATORY

     This section presents a two-way analysis of variance (ANOVA)
for the transfer efficiency results obtained using each combina-
tion of spray gun and target.  The objective of the analysis is
the determination of the relative influence of laboratory and
paint type on transfer efficiency, and the estimation of intra-
laboratory precision (repeatability) and the interlaboratory
precision (reproducibility).

     Table 35 presents values of transfer efficiency obtained
under each test condition.  Considering a single observation to
be the mean of four targets, two observations were obtained for
each treatment during Phase I, and six observations for Phase
II, which constitutes an unbalanced experimental design.

       TABLE 35.  ANALYSIS OF VARIANCE - TRANSFER EFFICIENCY
           Laboratory
                       II
  Laboratory
I            II
   51
Paint

   67
    51
Paint

    67
    51
Paint

    67
29.7
32.0
26.0
25.4
a.
10.9
10.7
10.2
10.2
c.
93.0
89.6
102.4
93.9
62.4
62.6
61.8
52.1
54.3
51.9
AAE-VC
15.7
15.9
16.0
14.5
15.2
15.4
AAC-VC
105.0
101.4
101.7
100.5
101.0
100.0
59.6
63.5
58.1
49.0
51.0
50.4

15.7
15.5
15.9
15.0
14.2
14.2

103.1
104.6
102.1
99.6
103.9
101.0
81.4
81.1
80.2
82.7
b.
61.4
62.1
60.8
60.8
d.
91.0
91.7
93.2
92.3
92.0
93.4
92.4
90.3
93.3
90.3
AAE-FP
74.9
76.9
77.3
73.0
77.9
75.8
AAC-FP
102.3
99.7
99.9
100.3
101.8
102.7
90.6
88.4
87.4
88.8
88.5
90.4

76.4
77.0
76.4
75.3
77.7
73.2

102.3
102.2
103.0
100.9
104.8
101.9
            e.  HSB-VC
 f.  HSB-FP
                                95

-------
     Table 36 presents the ANOVA results in standard format.
Those effects significant at the 0.01 level are noted by an
asterisk.  The column effect (laboratory)  is significant for
every configuration.  The row effect (paint) is significant^for
two of the six configurations,  but considerably less significant
than the laboratory effect even in those cases.  Interaction is
essentially absent, although it is just significant at the  .01
level for the single instance of the high speed bell - vertical
cylinder case.  Although there  is no effect of paint for this
case, the fact that interaction is present means that there is a
real effect, but the effect is  in opposite direction at each
laboratory.  Having noted this, it is interesting that the paint
exhibits a real effect on transfer efficiency in all three
vertical cylinder target configurations and in none of the flat
panel target configurations.

     Although the paints used for each phase of testing were not
split from identical 67 percent solids and 51 percent solids
batches, as would have been ideal, they were formulated to be
similar as delivered to each laboratory.  Nevertheless, because
the physical properties of the  two higher solids and the two
lower solids paints differ to some extent between laboratories,
there is concern that such a difference may be the major cause
of the difference observed in transfer efficiency results at
each laboratory.  The analysis  of variance presented here shows
this to be possible but unlikely.  The effect of paint on
transfer efficiency, even when  the paint weight percent solids
is purposely varied, has been shown to be either absent or to be
much less significant than the  observed effect of the laboratory.
Thus the reasons for the difference in results between labora-
tories must be sought elsewhere, as is done in the next section.

     The error mean square (sum of squares divided by degrees of
freedom) is an estimate of the  variance associated with repeated
measurements under identical conditions.  Although these vari-
ance estimates are not all poolable by Bartlett's test at the
Oe05 level, the variances associated with five of the six test
configurations  (all except the  air atomized conventional spray
with the vertical cylinder target) can be pooled.  This pooled
variance estimate is 3.33.  The precision estimate associated
with this variance is 1.82 (standard deviation), the test
repeatability.

     Reproducibility, the precision of interlaboratory testing,
for these results is clearly not acceptable.  The inter-
laboratory variance estimates (columns) shown in Table 36 are
large and are poolable, with a  pooled variance of 630.6, corre-
sponding to a reproducibility estimate of 25 (standard devia-
tion) .  The next section analyzes the causes of this poor test
reproducibility.
                                96

-------
TABLE 36.  ANOVA RESULTS FOR PAINT AND LABORATORY


Rows
Cols
Int
Error
Total

Row
Col
Int
Error
Total

Row
Col
Int
Error
Total

Row
Col
Int
Error
Total

Row
Col
Int
Error
Total

Row
Col
Int
Error
Total
*Significant at

SS
302.
2371.
16.
40.
2731.

0.
250.
0.
45.
296.

3.
649.
0.
26.
679.

3.
68.
0.
1.
73.

0.
158.
58.
65.
282.

2.
286.
0.
22.
311.
.01 level


8
6
8
0
2

30
25
31
72
58

90
01
00"
14
05

42
16
15
54
27

20
40
54
15
29

10
16
61
90
77
•
AAE-VC

1
1
1
12
15
AAE-FP
1
1
1
12
15
AAC-FP
1
1
1
12
15
AAC-VC
1
1
1
12
15
HSB-VC
1
1
1
12
15
HSB-FP
1
1
1
12
15


MS
302.
2371.
16.
3.


0.
250.
0.
3.


3.
649.
0.
2.


3.
68.
0.
0.


0.
158.
58.
5.


2.
286.
0.
1.




8
6
8
3


30
25
31
81


90
01
00
18


42
16
15
13


20
40
54
43


10
16
61
91



F
91.
718.
5.



0.
65.
0.



1.
247.
0.



26.
524.
1.



0.
29.
10.



1.
149.
0.







75*
66
09



08
*





68*
08



79
90
00



30
30
15



04
17
78



10





*




*
*





*
*




82*
32







                         97

-------
TEST REPRODUCIBILITY

     As presented in Sections 5 and 6, high precision was ob-
served for the results from each laboratory test.  When inter-
laboratory results were compared, however, the reproducibility
was very poor.  The combination of high individual precision with
low interlaboratory reproducibility pointed to systematic
differences between the two laboratory tests.  The test proce-
dure, materials, and equipment for the two sets of tests were
methodically compared to define any systematic differences.
Three major differences between tests were documented.  These
differences qualitatively account for the observed discrepancies
between laboratory results.

     The first major difference between Phase I and Phase II
testing was the paint spray booth air flow rate.  During Phase I
testing, the linear air velocity was measured up to 1.0 m/s
(200 fpm)  in the plane of the target.  The air flow in the plane
of the target in Phase II was undetectable.  Directly in front
of the fan it was only 0.5 m/s (100 fpm).  This is a large dif-
ference between Phase I and Phase II conditions.

     At higher booth air velocity the paint may flow more
quickly past the targets towards the exhaust*  Thus the atomized
paint has less opportunity to coat the targets at higher booth
air flow rates.  This effect is observed in the significantly
higher TE's for all Phase II test configurations compared to
Phase I results  (see Table 37).

            TABLE 37-  TRANSFER EFFICIENCY COMPARISON

vc
AAE-51-VC
AAE-67-VC
AAC-51-VC
AAC-67-VC
HSB-51-VC
HSB-67-VC
Average
FP
AAE-51-FP
AAE-67-FP
AAC-51-FP
AAC-67-FP
HSB-51-FP
HSB-67-FP
Average
Phase I
TEM

30.8
25,7
10.8
10.2
91.3
93.9
43.8

81.2
81.5
61.7
60e8
91.3
92.8
78.2
Phase II
TEM

61.3
51.5
15.8
14.7
103,0
101.0
57.9

90.7
90.3
76.5
75.5
101.6
102.1
89»5
Net Difference
%

99*0
94e2
46.3
44.1
12.8
7.6
50.7

11.7
10.8
24.0
24 . 2
11.3
10.0
15.3
                               98

-------
     As confirmation of the nature of this phenomenon, the VC
target results would be expected to be more affected than FP
target results.  The VC targets have open air space between them,
which allow paint to be drawn past each foil; the FP foils have
no interstice to allow easy air (and atomized paint) flow past
the foils.  The interlaboratory data support this expectation,
with the VC results being 50.7 percent different and the FP
results only 15.3 percent different between laboratories (Table
37).

     Another logical outcome of low booth air rates in Phase II
would be more of an effect for runs with electrostatic equip-
ment.  The electrostatic attraction would help attract slow
moving atomized paint to the targets.  This effect is seen for
AAC and AAE runs, but is obscured by other differences in the
HSB runs.

     Interlaboratory test results for air atomized electrostatic
equipment varied more than any other test series (refer to
Table 37).  Vertical cylinder TE's were dramatically more
affected than the flat panel TE's, pointing to a difference in
attractive forces (operating voltage).  Although the same model
power supply was used in both laboratories, the operating volt-
age was about 47 kV in Phase I and 62 kV in Phase II.  Higher
TE's (especially for VC targets) are the reasonable consequence
of higher voltages in the Phase II tests.

     Finally, the test procedure was slightly altered for Phase
II testing.  The test procedure calls for 0.4 kg of paint to be
sprayed before initiating the conveyor to pass the targets
through the gun.  (This requirement was made to minimize the
error of having a +_ 5-g paint scale accuracy for mass flow
determination.)  There was not enough paint left from Phase I to
complete Phase II and meet the 0.4-kg requirement, so the de-
cision was made to use 0.1 kg.  It was expected that any in-
accuracy introduced here would be offset by more test repeti-
tions, as was shown.

     However, the 0.4 kg in Phase I was discharged with the
targets out of the booth.  In Phase II, the 0.1 kg was dis-
charged as the targets moved towards the gun, inside the
booth.  There was no way to determine mass flow with the targets
on the conveyor  (but outside the booth) in Phase II because of
the booth configuration.

     Although this procedural change should not have affected
the AAC runs, the electrostatic equipment was affected.  The
paint sprayed as the target moved towards the gun had the
capability of staying airborne long enough to be attracted to
the targets as they passed through.  The effect on AAE runs is
probably small because the paint is sprayed directly towards the
targets and exhaust system.  The horizontal mass velocity of the

                                99

-------
paint tends to carry it quickly towards the exhaust.  The effect
on the HSB runs is larger because the HSB produces a fog of
paint with very low horizontal mass velocity.  Thus the paint
sprayed by the HSB is capable of staying suspended in the
vicinity of the targets long enough to coat the targets.  The
adherence of extra paint to the HSB targets threw the TE ' s over
100 percent, a clear physical impossibility.

     In an effort to salvage the HSB data, engineering judgment
was used to correct the data for the additional paint the tar-
gets were exposed to.  Based on observations during the Ransburg
test, the paint was sprayed 5 feet in front of the target with
the target in the booth.  At PPG, the targets entered the booth
3 feet from the spray gun.  The difference of 2 feet was con-
verted to mass and subtracted from the HSB data.  Corrected HSB
data were presented in Section 6, Table 31.  The correction,
though based solely on engineering judgment, aligns Phase I re-
sults more closely than Phase II.  This alignment is considered
corroborative of the described effect.

     Four AAE runs (AAE-67-1-4)  were made during Phase II timing
the flow of paint using marks at the beginning of the first
scavenger and the end of the last scavenger.  These runs were
made to determine the effect of reduced accuracy in the paint
weighing operation on the transfer efficiency determination.,
The standard deviation of the four runs AAE-67-FP-1 through 4
(treating the mean of 4 targets as a single observation) was
2.18.  In the case of the four runs AAE-67-VC-1 through 4, the
standard deviation was 1.89.  A test was performed to see if
either of these represents a different level of precision than
that of the overall transfer efficiency test.

     The pooled standard deviation, S", for Phase II testing was
1.8 based on 48 observations.  While Bartlett's test is appropri-
ate for comparison of more than 2 variances, the F-statistic is
used to determine if a pair of sample variances differ signifi-
cantly.

     The following are the calculations.

                            2                        —c^ /•?'2
                                                     ~S /S
                                                 3,47
AAE-67-FP(l-4)    2.18    4.73      3.24            1.46
AAE-67-VC(l-4)    1.89    3.56      3.24            1.10

With 3 and 47 degrees of freedom, the critical value of F for a
two-sided test at the 0.1 significance level is 2.84.  Therefore
the null hypothesis cannot be rejected, leading to the conclu-
sion that the variance associated with the altered test method
is poolable with that of the standard method.
                               100

-------
It is therefore recommended that the test procedure be revised
to set the timing and mass flow rate marks at the scavengers.
This recommendation does not have a significant effect on the
test precision, while it avoids the problem of premature paint
flow initiation biasing the transfer efficiency results.
                               101

-------
                            SECTION 8

                      THIRD LABORATORY TEST
FACILITIES

     The third laboratory test was performed at Nordson using
the original test plan with modifications to accommodate varying
several parameters for each successive run.  The test was run in
a 3 m (10 ft)  open-faced booth contained in a 4.5 x 6.1 m (150 x
20 ft)  closed  room.  Ventilation was provided by a multispeed
fan drawing outside air once through the booth.  Test equipment
specifications matched or exceeded the requirements of the test
plan.

DESCRIPTION OF PAINT

     A 70-percent solids enamel paint with polyester resin was
obtained from  PPG for this test (see Table 38).  The paint
differs slightly from the 67-percent solids paints used in
earlier tests.  The weight percent solids was higher, and a
small additive present in the original test paints had been
deleted by the manufacturer.

     A SANES Model 877 was used to measure paint resistance.
ASTM D-1200-70 was used for viscosity determination.  Paint
weight percent solids was determined by ASTM D-2369-81 using the
manufacturer's recommended cure schedule.

QA/QC PROCEDURES

     All quality assurance/quality control procedures in the
approved Test  Plan were followed.  As in Phase I and Phase II,
a Master Data  book was prepared prior to testing.  All test
documentation  and data were recorded (and checked) in the Master
Data book.  A log book was also maintained for special notations
and observations.

TEST DESIGN

     According to ASTM 691-79, "The first requirement is the ex-
istence of a valid, well written test method that has been
developed in one or more competent laboratories and has been
                               102

-------
               TABLE 38.   PAINT SPECIFICATIONS
   1.  Paint  Type

   2.  Resin  Type

   3.  Manufacturer

   4.  Manufacturer's Paint ID No.

   5.  Lot  No.

   6.  Color

   7.  Recommended Cure Schedule

   8.  Viscosity (uncut)  (1)

   9.  Reducing Solvent

  10.  Vol. of  Solvent Put into
      Vol. Paint



  11.  Viscosity - Spray (cut) (1)

  12.  wt/gal - Spray

  13.  wt.  Solids - Spray

  14.  Resistance
     Enamel
     Polyester
     PPG
     W45458
     Tan
     10 min  @ 350 °F
  — Sec.tt — Ford Cup @  °F

     Solvesso
 _ 700 cc (vol) solvent in
    1 1/2   (vol) paint

   15 1/2 s   Zahn 3 _

28.5 s    #4 Ford Cup @ 75°F

 _ ™ _ Ibs/gal _
   1.7X108
(1)  Use ASTM D-1200-70,  "Viscosity of Paints, Varnishes, and
    Lacquers by Ford Viscosity Cup."  Viscosity may also be
    determined by ASTM D-3794, Part 6 (Zahn Cup method)
    in addition to the Ford Cup measurements.
                               103

-------
subjected to a screening procedure. . .".  The third laboratory
test was developed to provide the required screening procedure
to determine the extent to which previously uncontrolled factors
influence the resulting transfer efficiency for the three equip-
ment types previously tested.  Variables were selected primarily
through an analysis of the differences between Phase I and Phase
II test conditions.  Unknown or uncontrolled factors were
selected for further testing.  In addition to these variables,
the Steering Committee suggested other parameters for screening
testing.  The following parameters were selected for screening:

          AAE                                  AAC

   Booth air rate, fpm            •  Booth air rate, fpm
   Paint mass flow, g/s           •  Paint mass flow, g/s
   Lag discharge distance, ft     •  Lag discharge distance, ft
   Shaping air                    •  Shaping air
   Atomizing air, psig            •  Atomizing air, psig
   Tip voltage, kV

                             HSB

                   Booth air rate, fpm
                   Paint mass flow, g/s
                   Log discharge distance, ft
                   Shaping air
                   Bell rpms
                   Tip voltage, kV

     When the results of two or more factors are to be studied,
a factorial design is usually the most efficient method to use-
The basic idea of factorial design is to alter several aspects
of a procedure at a time, but in such a way that the effects of
individual alterations may be determined.  Fractional factorial
designs sacrifice the ability to test for some or all inter-
actions but are able to test for a number of main effects very
efficiently.

For TE screening testing, the effect of five or six parameters
needed to be evaluated.  The Association of Official Analytical
Chemists  (AOAC) design (the 1/16 fraction of a 2  factorial)
for examining the effects of seven variables in eight trials was
selected for the third TE test, and is presented in Table 39.
One or two "dummy" variables were added to complete the AOAC
matrix for TE testing.  (Dummy variables introduce a meaningless
action, such as checking your watch prior to a test run, into
the matrix.  If this action is later demonstrated to be signifi-
cant, it points to other difficulties with the procedure that
must be addressed.)
                               104

-------
               TABLE 39.   AOAC(1^ SCREENING  TEST DESIGN
 Eight combinations of  seven factors used  to test the  ruggedness
 of an analytical method

                                       Combination or Detn  No.
Factor
A
B
C
D
E
F
G
or
or
or
or
or
or
or
Value
a
b
c
d
e
f
g
Observed result .
i
A
B
C
D
E
F
G
s
2
A
B
c
D
e
f
g
t
3
A
b
C
d
E
f
g
u
4
A
b
c
d
e
F
G
V
5
a
B
C
d
e
F
g
w
6
a
B
c
d
E
f
G
X
7
a
b
C
D
e
f
G
y
8
a
b
c
D
E
F
g
z
A, B, C,  D,  E,  F, G = nominal values  for  7 different  factors
                       that  might influence the result  if  their
                       nominal values  are  slightly changed.

a, b, c,  d,  e,  f, g = the alternative value of A, B,  C, D,  E,
                       F, G


    Youden,  W.  J. and E. H.  Steiner,  "Statistical Manual  of the
    Association of Official  Analytical  Chemists," AAOC, Arlington,
    VA,  1975.

 Note:  This test design does not  address  the interactions of the main  effects
       to each other.  For example, paint flow,  although not significant at
       the 90and 95 percent confidence level is  confounded with the inter-
       actions of voltage and atomizing air or lag distance is confounded
       with the interactions of voltage and booth air and others.
                                 105

-------
     From Table 39  it  is  evident  that  each  factor  can be evalu-
ated by some combination  of  the eight  determinations.  More is
presented on the data  analysis later  in  this  section.

     Following the  AOAC design, a  test matrix  was  developed for
each equipment type.   These  matrices  are  presented in Tables 40,
41, and 42.  The performance order  for each set  of eight runs
(for each equipment type) was randomized.   This  randomized set
was then assigned sequential run  numbers  to identify the pre-
scribed order of testing.  Tables  40  through  42  show the actual
values of each factor  attained during  the testing, as well as
the resulting values of transfer  efficiency for  each equipment
configuration.

TEST PARAMETERS

     Each equipment type, HSB, AAC, and  AAE,  was tested accord-
ing to the AOAC test design.  The  equipment specifications (with
the exception of intentionally varied  factors) are presented in
Tables 43 through 49.

TEST SEQUENCE(1)

The actual test sequence  was?

Monday (6/27/83)
     o  Obtained platform scale
     o  Set up HSB
     o  Assembled targets
     o  Cut and weighed foils for  HSB  runs
     o  Set conveyor speed
     o  Fabricated  scavengers

Tuesday (6/28/83)
     o  Completed HSB  set up
     o  Cut paint and  documented
     o  Conducted trial mass flow and  spray pattern  runs
     o  Conducted dummy run
     o  Conducted HSB  5-8

Wednesday (6/29/83)
     o  Performed HSB  1-4
     o  Performed AAE  1-8

Thursday (6/30/83)
     o  Performed AAC  1-8
     o  Performed blank/solvent run
     o  Completed all  weighing
	o  Clean up

    See Table 7 for  nomenclature.
                               106

-------
        TABLE 40.  AIR ATOMIZED ELECTROSTATIC TEST MATRIX
                   AND RESULTS (NORDSON)
FLAT PANEL
Combination No.
Factor Value
A
B
C
D
E
F
G

= Voltage at tip, kV
= Atomizing air, psig
= Booth air at target, fpm
= Paint mass flow, g/s
= Lag discharge distance, ft
= Shaping air, high=l, Lcw=0
= Dummy one, Yes=l, No=0
TE
1
60
50
65
3.8
3
1
1
83.9
2
60
50
45
3.7
0
0
0
91.8
3
60
30
65
3.0
3
0
0
89.0
4
60
30
45
2.9
0
1
1
90.4
5
45
50
65
2.9
0
1
0
80.5
6
45
50
45
2.4
3
0
1
87.5
7
45
30
65
3.7
0
0
1
89.2
8
45
30
45
3.8
3
1
0
88.3
VERTICAL CYLINDER
Combination No
Factor Value
A
B
C
D
E
F
G
= Voltage at tip, kV
= Atomizing air, psig
= Booth air at target, fpm
= Paint mass flow, g/s
= Lag discharge distance, ft
= Shaping air, high=l, Low=0
= Dummy one, Yes=l, No=0
1
60
50
65
3.8
3
1
1
2
60
50
45
3.7
0
0
0
3
60
30
65
3.0
3
0
0
4
60
30
45
2.9
0
1
1
5
45
50
65
2.9
0
1
0
*
6
45
50
45
2.4
3
0
1

7
45
30
65
3.7
0
0
1

8
45
30
45
3.8
3
1
0
TE
21.4  23.9  28.3  36.7  21.1  17.8  26.8  30.6
                                 107

-------
   TABLE  41.  HIGH SPEED BELL TEST MATRIX AND RESULTS  (NORDSON)
FIAT PANEL
Combination No.
Factor Value
A =
B =
C =
D -
E =
F =
G =

Shaping air, psig at gun
Voltage at tip, kV
Paint mass flow, g/s
Lag discharge distance, ft
Booth air at target, fpm
Thousand rpms
Dummy one, Yes=l, No=0
TE 96
1
40
90
6.5
3
105
11
1
.4
2
40
90
3.8
3
80
9.3
0
104.4
3
40
72
6.5
0
80
9.2
0
91.1
4
40
72
3.6
0
50
11
1
97.6
5
30
90
6.4
0
65
11
0
101.1
6
30
90
3.4
0
105
9.3
1
101.3
7
30
72
7.0
3
65
9.3
1
107
8
30
72
3.6
3
67
11
0
.1 100.5
VERTICAL CYLINDER
Combination No.
Factor Value
A =
B =
C =
D =
E =
F =
G =
Shaping air, psig at gun
Voltage at tip, kV
Paint mass flow, g/s
Lag discharge distance, ft
Booth air at target, fpm
Thousand rpms
Dummy one, Yes=l, No=0
1
40
90
6.5
3
105
11
1
2
40
90
3.8
3
80
9.3
0
3
40
72
6.5
0
80
9.2
0
4
40
72
3.6
0
50
11
1
5
30
90
6.4
0
65
11
0
6
30
90
3.4
0
105
9.3
1
7
30
72
7.0
3
65
9.3
1
8
30
72
3.6
3
67
11
0
TE                       97.8  103.1  92.7  98.6  109.7  103.4  107  101.2
                                 108

-------
          TABLE 42.  AIR ATOMIZED CONVENTIONAL TEST MATRIX
                     AND RESULTS (NORDSON)
FLAT PANEL
Combination No.
Factor Value
A
B
C
D
E
F
G

= Atomizing Air, psig
= Paint mass flow, g/s
= Booth air at target, fpm
= Lag discharge distance, ft
= Shaping air, High=l, Lcw=0
= Dummy one, Yes=l, No=0
= Dummy two, Yes=l, No=0
TE
1
50
6.2
75
3
1
1
1
76.6
2
50
6.2
40
3
0
0
0
90.3
3
50
1.8
75
0
1
0
0
70.1
4
50
2.4
40
0
0
1
1
85.0
5
30
6.3
75
0
0
1
0
86.9
6
30
6.4
40
0
1
0
1
86.0
7
30
3.0
75
3
0
0
1
85.4
8
30
2.6
40
3
1
1
0
79.9
VERTICAL CYLINDER
Combination No.
Factor Value
A
B
C
D
E
F
G
= Atomizing Air, psig
= Paint mass flow, g/s
= Booth air at target, fpm
= Lag discharge distance, ft
= Shaping air, High=l, Lcw=0
= Dummy one, Yes=l, No=0
= Dummy two, Yes=l, No=0
1
50
6.2
75
3
1
1
1
2
50
6.2
40
3
0
0
0
3
50
1.8
75
0
1
0
0
4
50
2.4
40
0
0
1
1
5
30
6.3
75
0
0
1
0
6
30
6.4
40
0
1
0
1
7
30
3.0
75
3
0
0
1
8
30
2.6
40
3
1
1
0
TE                          14.7   17.4   13.3   15.3   18.3   17.2   15.9   14.8


                                  109

-------
       TABLE 43.   NORDSON TEST EQUIPMENT SPECIFICATIONS
A. We
1.
2.
3.
4.
ight Percent Solids Measurement
Laboratory Scales
a. Manufacturer
b. Model No.
c. Serial No.
d. Capacity, g
e. Rated accuracy, g
Foil Dishes
a« Type
b. Size
Syringe
a. Type
b. Capacity, ml
Solvent Type
Equipment
Mettler
H78AR
623434
160






2" dia. aluminum
2" dia.
Plastic
10
Solvesso



( paint )
B. Conveyor Speed Measurement Equipment
   1.  Rule
       a. Type
       b. Graduations
   2.  Electronic Timer
                                             NA
                                             NA
                                             Autotron Electronic
                                             Autotron
    a. Type
    b. Manufacturer
    c. Model No.  Countomatic (Digital Display)  A579SA
    d. Serial No.
    e. Rated accuracy, s
Mass Flow Measurement Equipment
                                                  364
                                                  0.01
       c.
       d.
   Target
   1.
   2.
   3.
   Wet
    Platform Scales
    a.  Manufacturer Metrodyne

    b.  Model No.       SS-100
    c.  Serial No.      100#
    d.  Capacity, kg
    e.  Rated Accuracy, g
    Stopwatch
    a.  Manufacturer
    b.  Model No.
       Serial No.
       Rated Accuracy, s
       Foil
    Type
    Nominal Thickness, mils
    Temper
                                 (Platform)  Filina Scales-Con-
    Film Measurement Equipment
    a. Manufacturer
    b. Model No.
Voltage Measuring Equipment Electrostatic
    a. Manufacturer
    b. Model No.
Linear Air Velocity Measurement
    a. Anemometer Mfg.
    b. Model No.
                                             solidated Controls
                                             UMC2000 AAAA
                                                5521    	
                                                45
                                              Marcel & Cie
1/5
1.5
Med
                                             System Analyzer
                                               Nordson	
                                               P160102B
                                               Alnor
                                               8500 +1
       f pm
                                 110

-------
          TABLE 44.  AAE PAINT SPRAY AND PERIPHERAL
                     EQUIPMENT SPECIFICATIONS  (NORDSON
A. Paint Supply Tank

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, gallons
  Pressure Pot
  Binks
  2 gal
B. Paint Spray Equipment

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cc/min
   6. Air Cap
   7. Fluid Tip
   8. Needle
    Conventional
Nordson Electrostatic
   Nordson	
   AN8A 2465028	
   A2K01472
   245-987
 228(59/1000
C. Paint Spray Booth
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cfm
 Open 10X17 ft
D. Conveyor
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   Overhead
   Unibuilt
   Forced Draft Oven
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.

   Paint Heaters
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
  2 door  8'XIO'
  Grieve Co.	
  SC550	
  49037
  NA
  NA
  NA
  NA
                               111

-------
   TABLE 45.  AAE EQUIPMENT OPERATING CONDITIONS  (NORDSON
A. Paint Spray Equipment
   1. Paint Pressure at Paint Pot, psig
   2. Paint Pressure at Spray Gun, psig
   3. Atomizing/Turbine Air Pressure at
       Spray Gun, psig
   4. Operating Voltage, kV
   5. Disk or Bell Speed, rpm
      a. With Paint Applied
      b. Without Paint Applied
   6. Horn Air for Bell, psig
   7. Paint Temperature at Paint Pot,  °F
   8. Gun to Target Distance, cm
   9. Pump Setting

B. Paint Spray Booth
   1. Ambient Temperature, °F
   2. Relative Humidity, %
   3. Air Flow Velocity, fpm,
       at plane of target
   4. Air Flow, fpm, measured 4 ft
      in front of fan
   5. Air Flow Direction

C. Target Parameters
   1. Average Wet Film Thickness, mils
   2. Average Dry Film Thickness
   3. Vertical Paint Coverage, cm (in)
   4. Target Height, cm (in)
   5. % Vertical Coverage
   6. Resistance to Ground, Ohm

D. Forced Draft Oven (1)
   1. Cure Time, minutes
      a. Foil Dish (sample)
      b. Target Foil

   2. Cure Temperature, °F
      a. Foil Dish (sample)
      b. Target Foil

E. Paint Heaters
   1. Temperature In, °F
   2. Temperature Out, °F

F. Conveyor Speed Setpoint, fpm  (cm/s)
 12
 50
 60
 74
 12"
 74
 61%
 30-60 fpm
Normal
100 microns
15"
30"
50%
24
  ^
10
350
350
 NA
 NA
 NA
(1) Same cure schedule as foils.
                                112

-------
   TABLE 46.  HSB EQUIPMENT OPERATING CONDITIONS  (NQRDSON
A. Paint Spray Equipment
   1. Paint Pressure at Paint Pot, psig
   2. Paint Pressure at Spray Gun, psig
   3. Atomizing/Turbine Air Pressure at
       Spray Gun, psig
   4. Operating Voltage, kV
   5. Disk or Bell Speed, rpm
      a. With Paint Applied
      b. Without Paint Applied
   6. Shaping Air for Bell, psig
   7. Paint Temperature at Paint Pot, °F
   8. Gun to Target Distance, cm
B
   9. Pump Setting
   Paint Spray Booth
   1. Ambient Temperature,  °F
   2. Relative Humidity, %
   3. Air Flow Velocity, fpm,
       at plane of target
   4. Air Flow, fpm, measured  4 ft
      in front of fan
   5. Air Flow Direction
C. Target Parameters
   1. Average Wet Film Thickness, mils
   2. Average Dry Film Thickness
      Vertical Paint Coverage, cm (in)
      Target Height, cm (in)
      % Vertical Coverage
   3,
   4.
   5,
   6. Resistance to Ground, Ohm

D. Forced Draft Oven  (1)
   1. Cure Time, minutes
      a. Foil Dish  (sample)
      b. Target Foil

   2. Cure Temperature, °F
      a. Foil Dish  (sample)
      b. Target Foil

E. Paint Heaters
   1. Temperature In,  °F
   2. Temperature Out, °F

F. Conveyor Speed Setpoint, fpm  (cm/s)
                                                    20.5
                                                    90
                                                    9300
                                                    12700
                                                    40
                                                    74
                                                    12"
                                                    .062
                                                   74
                                                   61%
                                                  Normal
                                                  20 microns
30
60
50%
P.452
                                                  10
                                                  10
                                                  350
                                                  350
                                                   NA
                                                   NA
                                                   NA
(1) Same cure schedule as foils.
                               113

-------
          TABLE 47.  HSB PAINT SPRAY AND  PERIPHERAL
                     EQUIPMENT SPECIFICATIONS  (NORDSON
A. Paint Supply Tank

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, gal
       2 gal STD
       Sinks
       2 gal
B. Paint Spray Equipment

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cc/min
   6. Air Cap
   7. Fluid Tip
   8. Needle
         Turbobell
        Ransburg
        3655 01
        20074- 12 head
C. Paint Spray Booth
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cfm

D. Conveyor
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
      Open Room 10X17 ft
50-80 fpm at plane of tgt.
      Overhead-square run
        Unibuilt
   Forced Draft Oven
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.

   Paint Heaters
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
       2 door  8BX10'
       Grieve Co.	
       SC55Q
       49037
       NA
       NA
       NA
       NA
                                114

-------
   TABLE 48.  AAC EQUIPMENT OPERATING CONDITIONS  (NORDSON;

A. Paint Spray Equipment
   1. Paint Pressure at Paint Pot, psig             9	
   2. Paint Pressure at Spray Gun, psig             —	
   3. Atomizing/Turbine Air Pressure at
       Spray Gun, psig                              50
   4. Operating Voltage, kV                         —
   5. Disk or Bell Speed, rpm                       —
      a. With Paint Applied                         —
      b. Without Paint Applied                      —
      Horn Air for Bell, psig
   7. Paint Temperature at Paint Pot, °F           85
   8. Gun to Target Distance, cm                   12"
   9. Pump Setting                                 —
B. Paint Spray Booth
   1. Ambient Temperature,  °F                      85
   2. Relative Humidity, %                         65%
      Air Flow Velocity, fpm,                      30-50
       at plane of target
   4. Air Flow, fpm, measured 4 ft                 —
      in front of fan
   5. Air Flow Direction                          Normal
C. Target Parameters
   1. Average Wet Film Thickness, mils            350 microns
   2. Average Dry Film Thickness                  —	
   3. Vertical Paint Coverage, cm (in)             8	
   4. Target Height, cm  (in)                      30
   5. % Vertical Coverage                         —	
   6. Resistance to Ground, Ohm
D. Forced Draft Oven  (1)
   1. Cure Time, minutes
      a. Foil Dish  (sample)                       10
      b. Target Foil                              10
   2. Cure Temperature, °F
      a. Foil Dish  (sample)                       350
      b. Target Foil                              350

E. Paint Heaters
   1. Temperature In,  °F                           NA
   2. Temperature Out, °F                          NA
F. Conveyor Speed Setpoint, fpm  (cm/s)             NA
(1) Same cure schedule as foils.


                               115

-------
              TABLE 49.  AAC PAINT  SPRAY  AND PERIPHERAL
                         EQUIPMENT  SPECIFICATIONS (NORDSON)
A. Paint Supply Tank

   1. Type
   2, Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, gal
  Pressurized Std
  Sinks
  2 gal
B. Paint Spray Equipment

   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   5. Rated Capacity, cc/min
   6. Air Cap
   7. Fluid Tip    (Had to change  fluid  tip)
   8,, Needle
Air Atom.Conventional
   Sinks	
   610	
   56789
   63P3 or 63PB
C. Paint Spray Booth
   1. Type
   20 Manufacturer
   3. Model No.
   4. Serial No.
   5c Rated Capacity, cfm
 Open 10X17 ft
D. Conveyor
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.
   Overhead
   Unibuilt
   Forced  Draft Oven
   1. Type
   2. Manufacturer
   3. Model No.
   4. Serial No.

   Paint Heaters
   1. Type
   2o Manufacturer
   3. Model No.
   4. Serial No.
 8'XIO1 Conventional
  Grieve Co.	
  SC550	
  49037
  NA
  NA
  NA
  NA
                                116

-------
Friday (7/1/83)
     o  Completed clean up
     o  Prepared test materials for shipment


SOLVENT-ONLY RUN

     A test run was conducted using Solvesso  (no paint) as the
spray in order to determine if the solvent or tape were con-
tributing any weight gain to the foils.  The foils averaged a
0.001-g weight gain.  This weight gain is undetectable in
reported TE's calculated to tenths.

DATA ANALYSIS

     During the data analysis an error was found in the record-
ed cured foil weights for AAE runs E and F, which happened to
have been run consecutively in the randomized order.  It was
discovered that the cured foil weights for runs E and F were
inadvertently recorded on the incorrect sheets.  The following
data analysis is based on the corrected weights.

     Table 50 presents the analysis of the experiment.  The
computer printout gives the contrasts and variances associated
with each factor.  The contrast is the difference between the
mean transfer efficiency values obtained when each factor is at
its high and low values.

     The variance associated with the dummy variable or vari-
ables is small in each case, as is to be expected.  At least
one other factor exhibits a variance similar to that exhibited
by the dummy variable(s) for each experiment.  Such variances
were pooled with the dummy variance(s) to get an estimate of
error for each experimental configuration.

     Table 51 presents the results of an analysis of variance
based on the variances from Table 50.  The degrees of freedom
associated with the error term is noted in each case.  Those
factors found to be significant at the 0.1 level are marked
with an asterisk.

     A regression equation may be written to show the manner
in which transfer efficiency varies with each factor over the
range tested.  Regressions for each experimental configuration
with the factors found to be significant as independent vari-
ables are presented in Table 52.  The coefficients are simply
the contrasts  (1) from Table 50.
 (1) Myers, Raymond H.,  Response  Surface Methodology pp  47-48,
   1976.

                                117

-------
                 TABLE 50=   SCREENING EXPERIMENT ANALYSIS
                            TABULATED VALUES  OF VARIANCE AND
                            CONTRASTS
        AAC  -VC
Factor  Variance
c
B
F
A
E
D
G

C
B
F
A
E
D
G


C
B
F
A
E
D
G

C
B
F
A
E
D
g
0.78125
8.61125
0.06125
3.78125
5.95125
0.21125
0.06125
Contrast
0.625
-2.075
-0.175
1.375
-1.725.
-0.325
0.175
AAC -FP
Variance
61.605
47.045
1.445
32.805
153.125
2.205
4.205
Contrast
5.55
-4.85
-0.85
4.05
-8.75
1.05
-1.45
  MSB  -VC
  Variance

  0.10125
 26.28125
  0.15125
105.85125
 67.86125
  2.76125
  0.00125

  Contrast

 -0.225
 -3.625
  0.275
  7.275
 -5.825
  1.175
 -0.025

  HSB  -FP
  Variance

  8.20125
  5.95125
  8.61125
 52.53125
 54.60125
 37.41125
  3.51125

  Contrast

  2.025
 -1.725
 -2.075
  5.125
 -5.225
  4.325
 -1.325
  AAE  -VC
  Variance

 16.245
182.405
 21.125
 24.5
 13.52
  0.13
  0.18

  Contrast

  2.85
  9.55
  3.25
 -3.5
 -2.6
 -0.3
  0.3

  AAE  -FP
  Variance

 29.645
 21.78
 25.92
 11.52
  1.28
  4.205
  0.245

  Contrast

  3.85
  3.3
 -3.6
 -2.4
 -0.9
  1.45
 -0.35
                                   118

-------
TABLE 51,
ANOVA RESULTS FOR SCREENING  TEST

Source
Regression
on mean
on A
on B
on C
on E
error
Total

Regression
on mean
on A
on B
on C
on E
error
Total

Regression
on mean
on A
on B
on D
on E
error
Total

Regression
on mean
on A
on D
on E
error
Total
AAC-VC
SS

2012.9512
3.78125
8.61125
.78125
5.95125
.33375
2032.41
AAC-FP

54483.005
32.805
47.045
61.605
153.125
7.855
54785.44
HSB-VC

82722.781
105.85125
26.28125
2.76125
67.86125
.25375
82925.789
HSB-FP

79900.031
52.53125
37.41125
54.60125
26.275
80070.849

DF

1
1
1
1
1
3
8


1
1
1
1
1
3
8


1
1
1
1
1
3
8


1
1
1
1
4
8

MS


3.78125
8.61125
.78125
5.75125
.11125




32.805
47.045
61.605
153.125
2.618




105.85125
26.28125
2.76125
67.86125
.085




52.53125
37.41125
54.60725
6.56875


F


33.99*
77.40*
7.02*
53.49*





12.53*
17.97*
23.53*
58.48*





3724.8*
924.8*
97.2*
2388.0*





8.00*
5.69*
8.31*


                         119

-------
                  TABLE 51.   (Continued)



Source
AAE-VC
SS

DF



MS



F
Regress ion
on
on
on
on
on
on
error
mean
A
B
C
E
F

Total


5335
24
182
16
13
21

5593

.445
.5
.405
.245
.52
.125
.36
.6
AAE-FP
1
1
1
1
1
1
2
8


24
182
16
13
21




.5
.405
.245
.52
.125
.18



136
1013
90
75
117




.1*
.4*
.3*
. 1*
.4*



Regression
on
on
on
on
on
on
error
mean
A
B
C
D
F

Total
61355
11
21
29
4
25
1
61449
.045
.52
.78
.645
.205
.92
.525
.64
1
1
1
1
1
1
2
8

11
21
29
4
25



.52
.78
.645
.205
.92
.7625


15
28
38
5
33



ol*
.6*
.9*
.5
.9*


*Significant at .10 level
                             120

-------
  TABLE 52.  REGRESSION MODELS DERIVED FROM  THE  SCREENING TESTS

  '                        AAC-VC                      '
     TE = 15.862 -       A +       B  - -    C  -
     TE = 82.525 -      A
A = Atomizing air          +1 =  50 psig    -1 = 30 psig
B = Paint mass flow        +1 =  6.275      -1 = 2.45
C = Booth air at target    +1 =  75 fpm     -1 = 40 fpm
E = Shaping air            +1 =  high       -1 = low

                          HSB-VC

     TE = 101.6875 - 14™ A + *l™ B + ii££ D - 1
         TE = 99.9375 -
                          HSB-FP

                                  4.325     5.225
         TE = 87.575 -I-
A = Shaping air, psig at gun   +1 =40         -1 = 30
B =. Voltage at tip, kV         +1 = 90         -1 = 72
D = Lag disch. dist., ft       +1=3         -1=0
E = Air rate at target, fpm    +1 = 89.25      -1 = 65

                          AAE-VC

     TE = 25.825 +!5A-iB_2c_26E +
A = Voltage at tip, kV         +1 =60        -1 = 45
B = Atomizing air, psig        +1 =50        -1 = 30
C = Booth air at target, fpm   +1 =65        -1 = 45
E = Lag disch. dist., ft       +1 =  3        -1=0
F = Shaping air                +1 = high      -1 = low
                               121

-------
     The differences observed between transfer efficiency
results at PPG and at Ransburg for the same test configurations
can now be seen to be explained by the uncontrolled variables.
Table 53 presents, for each configuration, two sets of test
conditions which, according to the appropriate regression, would
result in transfer efficiency values similar to those actually
observed at each laboratory.  The values of each variable
utilized in construction of Table 53 are the actual experimental
values for those variables for which data was recorded.

     The predicted transfer efficiency values shown in Table 53
are similar to those actually obtained during the Phase I and
Phase II testing.  In general, they are within one standard
deviation of the observed values, where the precision of the
test method has been specified as a standard deviation of 2.0.

     Appendix B presents the results of a multiple linear re-
gression computer analysis of the screening experiment, utiliz-
ing as independent variables for each regression the factors
found to be significant by ANOVA (see Table 51).  Actual experi-
mental values of each factor for each experiment were used in
Appendix B rather than the mean of each high and low setting as
was done here.
                               122

-------
       TABLE 53.   PREDICTED TRANSFER EFFICIENCY RESULTS
      	(PPG AND RANSBURG)	


                    Air Atomized Conventional

                                      PPG               Ransburg

Independent Variable

  A - atomizing air, psig             50                   51
  B - paint mass flow, g/s             3.04                 2.76
  C - booth air at  target, fpm       100                   50
  E - shaping air                      2                    0

Predicted TE

  Vertical cylinder                   11.97                14.37
  Flat panel                          63.33                79.45


                        High Speed Bell

Independent Variable

  A - shaping air,  psig               35                   45
 -B - voltage at tip, kV              80                   80
  D - lag distance, ft                 0                    5
  E - booth air at  target, fpm       100                   50

Predicted TE

  Vertical cylinder                   95.4                102.1
  Flat panel                          92.8                105.7


                    Air Atomized Electrostatic

Independent Variable

  A - voltage at tip, kV              47                   63
  B - atomizing air, psig             30                   10
  C - booth air at  target, fpm       100                   50
  E - lag distance, ft                 0                    5
  F - shaping air                      0                    4

Predicted TE

  Vertical cylinder                   24.2                 46.8
  Flat panel                          79.7                 88.0
                               123

-------
                            SECTION 9

                 CONCLUSIONS AND RECOMMENDATIONS


     The results of Phase I and Phase II testing demonstrated
the viability of the draft spray painting TE method.  While
these tests exhibited good precision, tney produced clearly
different TE results.  Subsequently, a third test was performed
to define the factors contributing to the differences between
Phase I and II results.

     The following factors were determined to have a significant
impact on TE for at least one type of spray equipment tested:

     o  Booth air rate at target, fpm
     o  Shaping air
     o  Atomizing air, psig
     o  Voltage at tip, kV (for electrostatic equipment)
     o  Paint discharge technique (lag discharge distance)
     o  Paint mass flow, g/s

     A complete TE method must specifically address control of
these factors (in addition to those already controlled)  to avoid
introduction of significant errors.  The draft test method pre-
sented in Section 4 incorporates these findings as summarized
below:

     o  The booth air rate at the targets should be controlled.,
        While a certain air rate may be specified in the TE
        method, most laboratories do not have the capability to
        adjust air rates.  A conservative alternative would be to
        set a minimum acceptable air rate for all tests.

     o  Shaping air pressure must be controlled.  Generally
        shaping air is adjusted to provide a properly shaped
        spray.  The adjustment is qualitative.

     o  Atomizing air pressure must be controlled.

     o  Tip voltage must be controlled; an adjustable power
        supply is strongly recommended.
                               124

-------
     o  The paint discharge technique must be standardized to
        provide identical opportunity for paint to adhere for
        each run.

     o  The paint mass flow rate must be controlled for air
        atomized equipment.

     The following changes are also recommended in the draft TE
procedure to provide a more precise, simpler test performance:

     o  The test panels and cylinders can be mounted in a less
        complex manner, provided dimensions and materials are
        consistent with the original targets.

     o  A single source of paint should be used for each series
        of tests.

     o  The same oven and timer should be used for all curing.

     o  Plastic syringes are acceptable for use in weight
        percent solids determination.

     o  All pressure gages should be calibrated before
        performing tests.

     It is essential to control all previously listed factors to
the same level for verification testing of the draft TE method.

     The draft test method arrived at through development and
screening tests as described in this report is now ready to be
subjected to a systematic program of interlaboratory testing.
Such a program will enable the determination of the reproduc-
ibility component of precision.  It is recommended that an
interlaboratory test program including from 6 to 10 laboratories
be undertaken, using as a basis the draft test method developed
here.
                               125

-------
                          BIBLIOGRAPHY
Air Pollution Control Board of the State of Indiana, VOC
     Regulations-General Provisions, Section 325 IAC 8-1.1.

Aldorfer, D.M.  A Report on Transfer Efficiency of Waterborne
     Enamel,  Preliminary Results.  Warren,  Michigan:  General
     Motors Technical Center,  Fisher Body Division, April 24,
     1979.

Antonelli, J. A.  "Liquid High Solids-Organic Coatings," Society
     of Manufacturing Engineers,  1974 Technical Paper FC74-654.

Brewer, G.E.F.  "Calculations of Painting Wasteloads Associated
     with Metal Finishing," for Industrial Environmental
     Research Laboratory, Cincinnati, EPA-600/2-80-144, June 1980,

Brewer, G.E.F.  "Mathematics of Emissions and Transfer Effici-
     ency," U. S. Environmental Protection Agency Seminar on
     Volatile Organic Compound Control in Surface Coating
     Industries.  Brighton, Michigan:  George E.F. Brewer
     Coating Consultants, 1979.

Bublick, Timothy  "Robots for Spray Finishing," Plating and
     Surface Finishing,  November 1980, V. 67,  No. 11, PSFMDH 67.

Chapman, G.R.  "High Solids Coatings," 1982 Products Finishing
     Directory, September 1981, V. 45, No.  12A.

DeVilbiss Company,  "The ABC's of Spray Equipment," Third
     Edition, Toledo, Ohio., 1954.

Drum, E.W.  Letter to Dave Salman, U. S. Environmental Protec-
     tion Agency, Research Triangle Park, N.C..  Indianapolis:
     Ransburg Electrostatic Equipment, A Division of Ransburg
     Corporation, October 31,  1978.
                               126

-------
Farrell, Ron   "Powder Coatings,"  1982 Products Finishing
     Directory, September 1981, V. 45, No.  12A.

General Motors Corporation.  GMAD Oklahoma  City, Oklahoma Trans-
     fer Efficiency Experiments.  Warren, Michigan:  General
     Motors Technical Center, Fisher Body Division, December 13,
     1979.

        Baseline Transfer Efficiency of Waterborne Enamel Top-
     coat, Warren, Michigan:  General Motors Technical Center,
     Fisher Body Division, October 1, 1979.

Godovoy. A.   "Deposition Efficiency in Electrostatic Spraying
     of Powder Coating."  Journal of Paint  Technology, Vol. 15
     No. 580. May 1973.

Gunsel, S.J.  Letter (SG-51) to Central Docket Section (A-130),
     U. S. Environmental Protection Agency, Washington, DC.
     Amherst, Ohio:  Nordson Corporation, February 19, 1981.

Harding, S.K.  Trip Report—April 19-21, 1982.  Ft. Lauderdale,
     Florida:  CENTEC Corporation.

Health and Human Services, U. S.  Department of.  An Evaluation
     of Engineering Control Technology for  Spray Painting, NIOSH
     Technical Report, -Publication No. 81-121. "Washington, DC:
     U. S. Government Printing Office, June 1981.

Hungerford, A.O.  Transfer Efficiencies of  Painting Method, A
     Case History.  Galesburg, Illinois:  Butler Manufacturing
     Company, March 1, 1982.

Industrial Finishing, October 1978, Part Two.

Industrial Finishing, August 1982, page 43.

Industrial Finishing, February 1982,  pages 9, 10, 16.

McCrodden, B.J.  Final Trip Report — Ransburg Corporation.
     Research Triangle Park, N.C.:  Research Triangle Institute,
     June 19, 1980.

Metal Finishing, January 1980, page 112.

Metal Finishing Annual Review Issue,  February 1981, February
     1982, and April 1982.

Miller, E.P-  "Powder Coating Compared with Other Techniques
     Available for Meeting OSHA and EPA Regulations," Society of
     Manufacturing Engineers, 1973 Technical Paper, FC73-553.
                               127

-------
Myers, Raymond H.  Response Surface Methodology,  1976.

Nickerson, R.S.  "Applying High Solids Coatings,"  Products
     Finishing, November 1981, pages 81-88.

Plating and Surface Finishing, "EPA Reveals  New  Policy  on
     Hydrocarbon Emissions," No date.

Powell, G. and Bussjaeger, S.  "Water-bourne Coatings,"  1982
     Products Finishing Directory, September 1981,  V. 45,  No.
     T2A^

Product Source Listings, 1982 Product Finishing  Directory,
     September 1981, V. 45, No. 12A.'

Roobol, N.R.  "Focus on Automotive Coatings," Metal  Finishing,
     September 1980, pages 31-35.

Scarbrough, D.  "Electrostatic Spray Coating .  .  .   Focus  on
     Profitability," Finishers' Management,  March  1981.

Sloan, E.M.  Trip Report -- GMAD  - Fremont,  California Transfer
     Efficiency Experiments.  Warren, Michigan:  General  Motors
     Technical Center,  Fisher Body Division,  April,  1980.
        Vehicle Paint Transfer Efficiency Test -- A New  Tool  to
     Compare Solvent Emissions (GMMD - 79-065), SME Paper  No.
     FC80-610.  Dearborn, Michigan:  Association for  Finishing
     Processes, Society of Manufacturing Engineers, 1980.

manable, Y.  Feasibility Study for Base Coat/Clear Coat  Painting
     System Using New High Solid Paints.  Marysville, Ohio:
     Honda of America Manufacturing, Inc., n.d.

U. S. Department of Health and Human Services, "An Evaluation of
     Engineering Control Technology for Spray Painting," June
     1981 .

U. S. Environmental Protection Agency,  "Industrial Surface
     Coating Appliances-Background Information for Proposed
     Standards," Office of Air Quality Planning and Standards,
     November 1930, EPA-450/3-80-037a (NTIS PB 82-152174).

U. S. Environmental Protection Agency, "Standards of  Performance
     for New Stationary Sources:  Industrial Surface  Coating:
     Appliances," Proposed Rule and Notice of Public  Hearing,
     Federal Register, V. 45, No. 249, December 24, 1980.
                               128

-------
U. S. Environmental Protection Agency, "Standards of Performance
     for New Stationary Sources:  Industrial Surface Coating:
     Appliances," Correction, Federal Register, V. 46, No. 18,
     January 28, 1981.

U. S. Environmental Protection Agency, "Standards of Performance
     for New Stationary Sources:  Industrial Surface Coating:
     Appliances, Correction," Federal Register, V. 46, No. 107,
     June 4, 1981.

U. S. Environmental Protection Agency, "Measurement of Volatile
     Organic Compounds," Office of Air Quality Planning and
     Standards 1979, EPA-450/2-78-041 (NTIS PB80-221674).

Walberg, A.C.  "Painting Efficiency of Automatic Electrostatic
     Systems," A paper presented at Chemical Coaters Association
     Meeting, June 12-14, 1979.

Welch, R.A.  "Cyrogenic Paint Stripping," Industrial Finishing,
     May 1984.

Youden, W.J. and E.H. Steiner-  "Statistical Manual of the
     Association of Official Analytical Chemists," AAOC,
     Arlington, VA, 1975.

Ziegeweid, J.E.  "Applying Organic Coatings, Tools of the
     Trade," Metal Finishing, October 1980, pages 57-61.

Ziegeweid, J.E.  "Applying Organic Coatings — Airless Comes
     of Age," Metal Finishing, June 1981, pages 37-41.
                               129

-------
                           APPENDIX A

            WORTH ASSESSMENT MODEL OF TE TEST METHODS


     Nine TE test methods were evaluated according to the
following criteria:

1.   How representative is the test method expected to be of
    actual line TE's?

         1.0  Identical
         0.8  Very close
         0*5  Defines range of TE's
         0.3  Poor
         0.0  Very poor

2.   How complex will the procedure be to perform?

         1.0  Easy
         008  Fairly easy
         0.5  Moderate
         0.3  Fairly difficult
         0.0  Difficult

3.   How precise is the method expected to be?

         1.0  Excellent
         0,8  Very good
         0,5  Fair
         0.3  Poor
         0.0  Very poor

4.   How expensive will the method be to perform?

         1.0  Little expense involved
         0.8  Operating expense only
         0.5  Little operating and capital expense
         0.3  Moderate capital and operating expense
         0.0  Substantial capital and operating expense
                               130

-------
5.   Does the method bias against certain equipment
    manufacturers?

         1.0  None
         0.8  Little
         0.5  Moderate
         0.3  Substantial
         0.0  Severe

6.   Does the method bias against certain target geometries?

         1.0  None
         0.8  Little
         0.5  Moderate
         0.3  Substantial
         0.0  Severe

7.   How well developed is the method?   (How much more work will
    be required for a final method?)

         1.0  Ready for use
         0.8  Ready for verification testing
         0.5  Needs some development
         0.3  Needs much development
         0.0  Develop from scratch

8.   How much process modification is required to use the method?

         1.0  None
         0.8  Little
         0.5  Moderate
         0.3  Substantial
         0.0  Severe

9.   Are there special limitations or restrictions to the method
    which must be considered?

         1.0  None
         0.8  Little
         0.5  Moderate
         0.3  Substantial
         0.0  Severe
                               131

-------
     Each criteria was weighed according to its relative
importance as shown:

         Criteria                 Weighting

     1 Representation               0.20
     2 Complexity                   0.09
     3 Precision                    0.15
     4 Expense                      0.08
     5 Mfg.  Bias                    0.12
     6 Tgt.  Bias                    0.11
     7 Level of Dev.                 0.07
     8 Process Mod.                 0.08
     9 Restrictions                 0.10
     Each TE test method is described by a three section
descriptor:

                     AAA/BBB/CCC

where

     AAA = plant method or laboratory method

     BBB = target type; actual workpiece (opr) target,
           standardized targets, or shim

     CCC = test parameters; actual operating (line) conditions,
           standardized spray conditions

The relative rankings of the methods are shown in Table A-l.

     The individual components of the assessment follow in
computer print-out form.
                               132

-------
                                  Table A-l

                         TEST METHOD ALTERNATIVES
UJ
U)
METHOD TYPE
PLANT
LAB
LAB
LAB
LAB
PLANT
LAB
LAB
PLANT
TARGET TYPE
WORKPIECE
WORKPIECE
STANDARD
STANDARD
WORKPIECE
STANDARD
SHIM
SHIM
SHIM
TEST PARAMETERS
ACTUAL OPERATING
ACTUAL OPERATING
ACTUAL OPERATING
STANDARD
STANDARD
ACTUAL OPERATING
ACTUAL OPERATING
STANDARD
ACTUAL OPERATING
SCORE
72.5
71.5
67.3
61.8
60.8
60.3
58.0
53.0
18.3

-------
8   PLANT/OPR  TGT/OPR COND     .725
3   LAB/OPR  TGT/OPR COND       .715
2   LAB/STD  TGT/OPR COND       .6725
1   LAB/STD  TGT/STD COND       .6175
5   LAB/OPR  TGT/STD COND       .6075
7   PLAN/STD  TGTS/OPR COND     .6025
4   LAB/SHIM/OPR COND           .58
6   LAB/SHIM/STD COND           .58
9   PLANT/SHIM/OPR COND         .4825
                134

-------
(1)    LAS/STD TCT/STD
 *  FACTOR  NAME

 1  REPRESENTATION
 2  COMF1. EX I TV
 3  PRECISION
 4  EXPENSE
 :  MFC BIAS
 6  TGT EIA3
 7  LEVEL OF DEV
 8  PROCESS MOD
 9  RESTRICTIONS

TOTAL
COND
 VALUE SELECTION  DESCRIPTION

  •3  3  POOR
  0  8  FAIRLY  EASY
  1  0  EXCELLENT
  0 . 3  CPR ONLY
  0  3  SUBSTANTIAL
  0.3  SUBSTANTIAL
  0  3  AEADY FOR  VEftIF
   1.0  NONE
  1.0  NONE
WE I CHT
Y / N
               ft VALUE
0
0
Q
0
0 ,
0
0
0
0
200
090
100
030
. 1 20
1 10
070
.080
.100
YES
YES
V E S
YES
YES
YES
YES
YES
YES
0
0
0
0
0
0
0
0
0
03000
06750
15000
06000
03000
02750
.05250
. 08000
.10000
                                                          0  61750
 (2)   LAB/STD  TCT/OPR COND
 * FACTOR NAME          VALUE
         SELECTION DESCRIPTION
WEIGHT   Y/N   ff  VALUE
 1 REPRESENTATION
 2 COMPLEXITY
 3 PRECISION
 4 EXPENSE
 5 MFC  BIAS
 6 TCT  BIAS
 7 LEVEL  OF DEV
 8 PROCESS MOD
 9 RESTRICTIONS
   0 3   DEFINES  RANGE
   0 3   FAIRLY  EASY
   1 0   EXCELLENT
   0 8   OPR ONLY
   C.5   MODERATE
   0.3   SUBSTANTIAL
   0.9   READY  FOR VERIF
   i.0   NOME
   0.8   LITTLE
0
0 ,
0 .
0 ,
0
0
0 .
0
0
200
090
150
, 080
120
1 10
, 070
.080
1 00
YES
YES
YES
YES
YES
YES
YES-
YES
YES
0
0
0
o
0
0
0
0
0
10000
06750
15000
, 06000
06000
02750
05250
08000
07500
 rUTAL
                                                           0  67250
 13)    LA3/OPR  TGT/OPR COIID
  « FACTOR NAME           VALUE
         SELECTION DESCRIPTION
 WEIGH:
 Y/N  *  VALUE
  1 REPRESENTATION
  2 COMPLEXITY
  3 PRECISION
  4 EXPENSE
  5 MFC  BIAS
  & TGT  BIAS
  7 LEVEL OF  DEV
  3 PROCESS MOD
  9 RESTRICTIONS

 rOTAL
   0.3   VERY  CLOSE
   0.8   FAIRLY EASY
   0 5   FAIR
   0.8   OPR  ONLY
   0.8   LITTLE
   0.8   LITTLE
   0.5   NEEDS DEV
   1.0   NONE
   0.8   LITTLE
0 .
0
0 ,
0
0
0
0
0
0
200
. 090
.150
080
. 120
.110
. 070
080
100
YES
YES
YES
YES
YES
YES
YES
YES
YES
0 .
0
0 .
0
C
0
0
0
0
15000
06750
07500
06000
C9000
08250
33500
.03000
07500
                                                           0 71500
                                            135

-------
(4)    LAE/SHIM/CrR
 # FACTOR NAME

                       VALUI  SELECTION DESCRIPTION
                                                             WEIGHT  Y/N   -  VALUE
1 REPRESENTATION
2 COMPLEXITY
3 PRECISION
4 EXPENSE
3 MFC BIAS
6 TGT SI AS
7 LEVEL OF LEV
8 PROCESS MOD
9 RESTRICTIONS
TOTAL
(5) LAS /OPR TGT/STD
4i FACTOR NAME
1 REPRESENTATION
2 CO?!?LEXITY
3 PRECISION
4 EXPENSE
3 MFC BIAS
6 TGT BIAS
7 LEVEL OF DEV
8 PROCESS MOD
9 RESTRICTIONS
TOTAL
(6) LAB/SHIM/STD CO
# FACTOR NAME
1 REPRESENTATION
2 COMPLEXITY
3 PRECISION
4 EXPENSE
5 MFC BIAS
6 TGT BIAS
7 LEVEL OF DEV
6 PROCESS MOD
9 RESTRICTIONS
0 3
0 . 8
o :
0 8
a . s
0 . 3
0 S
1 0
0 8

COND
VALUE
0 5
o a
0 S
0 8
0 3
0 . 3
0 3
: . o
0 8

NO
VALUE
0 3
0 . 8
0 S
0 8
0 3
0 3
0 5
1 0
0 3
DEFINES RANGE
FAIRLY EASY
FAIR
OPR ONLY
MODERATE
SUBSTANTIAL
NEEDS DEV
NONE
LITTLE


SELECTION DESCRIPTION
DEFINES RANGE
FAIRLY EASY
FAIR
OPR ONLY
MODERATE
MODERATE
NEEDS DEV
NONE
LITTLE


SELECTION DESCRIPTION
POOR
FAIRLY EASY
FAIR
OPR ONLY
MODERATE
SUBSTANTIAL
NEEDS DEV
NONE
LITTLE
0
c
0
0
0
a
0
0
0 ,
200
090
i SQ
080
1 20
1 10
070
080
ICO
YES
YES
YES
YES
YES
YES
YES
YES
YES
0
0
0
0
0
0
0
0
0
10000
06730
07500
04 000
06000
02750
03500
.08000
075 00
                                                                           0  S3000
                                                             WEIGHT  Y/N   S  VALUE
0
Cr
0
0
0
0
0
0
0

WE
0
0
0
0
0
0
0
0
0
200
090
.150
. 030
120
1 1 0
070
030
.100

IGHT
.200
090
1 SO
.080
. 120
.110
. 070
030
100
YES
YE 3
YES
YES
YES
YES
YES-
YES
YES

Y/N
YES
YES
YES
YES
YES
YES
YES
YES
YES
0
0
0
0
0
0
0-
0 .
0
0 .
ft
0
0
0
0
0
0
0
0
0
10000
06750
07500
06000
06000
OS500
OSS 00
08000
07^500
60750
VALUE
05000
06750
07500
06000
06000
02750
Q3500
08000
07500
TOTAL
                                                                           0 53000
                                       136

-------
<7)    PLANT/3TD TG'
 # FACTOR NAME

 1 REPRESENTATION
 2 COMPLEXITY
 3 PRECISION
 4 EXPENSE
 3 MFC BIAS
 6 TGT BIAS
 7 LEVEL OF DEV
 8 PROCESS MOD
 9 RESTRICTIONS

TOTAL
,/OFR COND
   VALUE SELECTION  DESCRIPTION
     0  5  DEFINES RANGE
     0  3  FAIRLY  EASY
     1  Q  EXCELLENT
     0.5  LITTLE  CAP&OPR
     0.5  MODERATE
     0  5  MODERATE
     0.5  NEEDS DEV
     0.3  SUBSTANTIAL
     0.S  LITTLE
WEIGHT  Y/N  » VALUE
0
0
0 .
0
0
0
0
0
0
200
090
1 50
080
120
1 1 0
070
. G30
100
YES
YES
YES
YES
YES
YES
YES
YES
YES
0
0
0
0
0
0
0
0
0
10000
04730
15000
04000
06000
05500
03500
02000
07500
                                                     0  60250
<8>   PLANT/OPR TGT/OFR  COND
 S FACTOR NAME         VALUE
          .ELECTION DESCRIPTION
WEIGHT  Y/N  * VALUE
 1 REPRESENTATION
 2 COMPLEXITY
 3 PRECISION
 4 EXPENSE
 5 MFC BIAS
 6 TCT BIAS
 7 LEVEL OF DEV
 8 PROCESS MOD
 9 RESTRICTIONS
     1  0   IDENTICAL
     0 . 8   FAIRLY EASY
     0.5   FAIR
     0  5   LITTLE CAP&OPR
     1.0   NONE
     1  0   HONE
     0  3   NEEDS MUCH DEV
     0.3   SUBSTANTIAL
     0.8   LITTLE
0
0
0
0
0
0
0
0
0
200
0 90
1 50
080
1 20
.110
.070
. 080
. 100
YES
YES
YES
YES
YES
YES
YES"
YES
YES
0
0
0
0
0
0
0
0
0
20000
06750
07500
04000
12000
11000
01750
02000
07500
TOTAL
                                                     0 72500
 (9)    PLANT/SHIM/OPR  COND
  #  FACTOR  NAME          VALUE

  1  REPRESENTATION       0 5
  2  COMPLEXITY           0 8
  3  PRECISION            0.5
  4  EXPENSE              0 5
  5  MFC BIAS              0 S
  6  TGT BIAS              0.3
  7  LEVEL OF  DEV         0.3
  8  PROCESS MOD           0.3
  9  RESTRICTIONS         0 8
          SELECTION DESCRIPTION

          DEFINES RANGE
          FAIRLY EASY
          FAIR
          LITTLE CAF&OFR
          MODERATE
          SUBSTANTIAL
          NEEDS MUCH DEV
          SUBSTANTIAL
          LITTLE
WEIGHT  Y/N   *  VALUE
  0
  0
  0
  0
  0 .
  0
  0
  0
  0
200
090
150
080
120
1 10
070
080
100
YES
YES
YES
YES
YES
YES
YES
YES
YES
0 10000
0.06750
0 07500
0 04000
0.06000
0 02750
0.01750
0 02000
0 07500
 TOTAL
                                                      0  48250
                                       137

-------
                            APPENDIX B

                        SCREENING PROCEDURE
        MULTIPLE LINEAR REGRESSION FOR TRANSFER EFFICIENCY
     The regressions presented in Section 8 of this report were
derived from the contrasts associated with the independent
variables.  Only those independent variables found to be signi-
ficant based on an analysis of variance were included.  The
resulting regression coefficients were based on normalized
values of the experimental factor levels, 1 = the high value and
-1 = the low value.

     This appendix presents nearly equivalent regression expres-
sions derived from a conventional multiple linear regression
program.  Again, only independent variables found to be signifi-
cant by ANOVA are included.  Whereas the Section 8 analysis
considered each factor to have taken on only two values (the
mean of the four actual high values and the mean of the four
actual low values), the analysis presented here used the actual
experimental data in every case.  Differences are slight
because, for the most part, the planned high and low values of
each factor were actually achieved.

     Results are presented in Table B-l.
                               138

-------
                                    Table B-l

                 SCREENING PROCEDURE--MULTIPLE LINEAR REGRESSION

                           Air Atomized Electrostatic
                   Means
     Standard Deviations
Correlation Coefficients
Vertical Cylinder

   52.5               xA
   40.0               xB
   55.                xC
    1.5               xE
    0.5               xF
    6.072831771       SO
      .5345224838      S5
    1.603567451       S4
   10.69044968        S3
   10.69044968        S2
    8.017837257       SI
      .3080652921     RO1
    -.8405781542     R02
    -.2508531664     RO3
    -.2288485027     RO4
      .2860606284     RO5
                                                            Flat Plate
           52
           40
           55
            0
            3

           10
           10,
            8,
                                                              -0
   676080988
   5345224889
   69044968
   69044968
   017837257
   3489732586
   4798382306
   559811269
                                                               -.5234598879
             xA
             xB
             •xC
             xF
             SO
             34
             S3
             S2
             SL
            ROL
            RO2
            RO3
            R04
Determinant of the Corre-
lation Matrix Inverse

       Beta Coefficients
 Regression Coefficients




               Intercept


             t Statistic
          Standard Error
                             0
                             1
                             2
                             3
                             4
                             5
    1.

      .3080652921
    -.8405781542
    -.2508531664
    -.2288485027
      .2860606284

      .2333333333
   -0.4775
   -0.1425
    -.8666666667
    3.25
   40.1875
   11.66666666
   -31.83333331
   -9.499999992
   -8.666666659
   10.33333332
   27.02921615
Bl
B2
B3
B4
B5

Bl
B2
B3
B4
B5
BO
Tl
T2
T3
T4
T5
TO
 1.

  .3489732586
 -.4798382306
-0. 559811269
 -.5234598879
 0
-0
-0
-3
     .9986054889      R2
     .4242640691      SE
    Transfer Efficiency      0
    A Voltage                1
    B Atomizing Air          2
    C Booth Air              3
    E Lag Distance           4
    F Shaping Air
         139
16
165
1925
6
98.1625
 2.455893648
-3.376853767
-3.939662728
-3.683840473
20.37178647
                                                                 -9394259739
Bl
B2
B3
34

Bl
B2
B3
B4
BO
             Tl
             T2
             T3
             T4
             TO
                R3
                                                                1.382027496     SE
           Transfer Efficiency
           A Voltage
           B Atomizing Air
           C Booth Air
           F Shaping Air

-------
                               Table B-l (continued)

                           Air Atomized Conventional
                             Vertical Cylinder
                   Means
     Standard Deviations
Correlation
Determinant of the Corre-
lation Matrix Inverse

       Beta Coefficients
 Regression Coefficients



               Intercept

             t Statistic
40.
 4.3625
 57.5
 0.5
 1

18
 2
10
          Standard Error
667279649
5345224838
70828693
071533249
69044968
4408189206
7291604789
2003722366
5530273731
  .9842287046

 -.3689709194
  .6551665119
 -.1876931776
-0.514990196

 -.0575446051
  .5273127008
-.0167271871
-1.606354642
17.62887312

-3.292647716
 5.835660816
 -1.684830484
-4.615799271
16.72299965

 0.962783067
  .4913230413
 xA
 xB
 xC
 xE
 SO
 34
 S3
 S2
 SI
RO1
RO2
RO3
R04
                Bl
                B2
                B3
                B4

                Bl
                B2
                B3
                B4
                BO

                Tl
                T2
                T3
                T4
                TO

                R3
                SE
                         Flat Plate

                           40.
                            4.3625
                            57.5
 0
 6

18
 2
10
                                                              -0
573051042
5345224838
70828693
071533249
69044968
3293472157
4894084014
451327666
7115526266
                              .9842287046
 xA
 xB
 xC
 xE
 SO
 S4
 S3
 S2
 SI
RO1
RO2
R03
R04
-.2844171654
.4097074919
-.4433988336
-.6877661293
-.1748746406
1.300016912
-0.155785678
-8.457496195
97.03508643
-3.833560108
5.511954203
-6.011676679
-9.310694592
35.26575728
.9836862467
1.282426085
Bl
B2
B3
B4
81
B2
B3
B4
BO
Tl
T2
T3
T4
TO
R3
SE
                                0 =   Transfer Efficiency
                                1 = A Atomizing Air
                                2 = B Paint Mass Flow
                                3 = C Booth Air
                                4 = E Shaping Air
                                       140

-------
                                 TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
 1. REPORT NO
  EPA-600/2-88-026a
2.
                            3. RECIPIENT'S ACCESSION-NO.
 4. TITLE AND SUBTITLE
 Development of Proposed Standard Test Method for
  Spray Painting Transfer Efficiency; Volume I.
  Laboratory Development	
                            5. REPORT DATE
                             April 1988
                            6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
                            8. PERFORMING ORGANIZATION REPORT NO.
 K. C. Kennedy
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Centec Corporation
 11260 Roger Bacon Drive
 Reston, Virginia 22090
                            10. PROGRAM ELEMENT NO.
                            11. CONTRACT/GRANT NO.

                             68-03-1721, Task 2
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Air and Energy Engineering Research Laboratory
 Research Triangle Park, NC 27711
                            13. TYPE OF REPORT AND PERIOD COVERED
                             Final:  1/82 - 1/87
                            14. SPONSORING AGENCY CODE
                             EPA/600/13
 16. SUPPLEMENTARY NOTES AEERL project officer is Charles H. Darvin,  Mail Drop 62b,
 919/541-7633.  Volume II describes the verification program.
 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 best
 characteristics of these methods 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.
17.
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                            b.lDENTIFIERS/OPEN ENDED TERMS
                                         c.  COSATI Field/Group
 Pollution
 Spray Painting
 Tests
                Pollution Control
                Stationary Sources
                Transfer Efficiency
13B
13H
14B
 8. DISTRIBUTION STATEMENT

 Release to Public
                19. SECURITY CLASS (ThisReport)
                Unclassified
21. NO. OF PAGES
     150
               20. SECURITY CLASS (Thispage)
                Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
             141

-------