I form ApprovedREPORT DOCUMENTATION PAGE OM ft. 07 4188

A~~ D -8 5r3 it-At1 t-3 Avtk3 I htOy' 0-!f -WWo. Ifludleg tht t"YMq 44of 19qWV%9, IrterIMO0S. 69orMN, 0mmW i data "w(TA D -A 2 5 2 3 3 4!?5tIn1. rie-,rt" oo.It

* a,"d -r' I~'*,WQ " ... I t ' ffneM.lttI.,I lipfl c~mmentS r. r*elnnq t!h,$ vV4t 91,lIM*llt 0. any 0th., fl,1111! Moti 111110 11111sm I/illts 11111 1"11Wl!I I!1141 toP

l *e.

! J ,"-)f- a,'wdJ Sudge.t. PFX~rrwvgr R11du~lon ftr¢|"I (004-018$, Wl~ltor,, 0t JOS41.

•REPORT OATE . . 3. REPORT TYPE AND OATES COVERED

April 17, 1992 Final - 3/1/91 to 3/4/92"S. FUNDING NUMBERS

"Research Women's Shirt Production" DLA90--87-D-0017 DO 0023 (C)

6. AUTHOR(S)

Nancy J. StaplesI EkCTE II

. PERFORMING ORGANIZATION NAME(S) AND AD .JN 1 2.FRMINGORANIATIO5) JUN1 2 1992 REPORT NUMBERClemson Apparel Research S J0

500 Lebanon Road CPendleton, SC 29670

9. SPONSOING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING

P. fense Logistics Agency AGENCY REPORT NUMBER

,ameron StationAlexandria, VA 22304-6100

11. -UPPLMENTARY NOTES

-isa OISTRIBUIIONIAVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

I1J. ADS IRACT (Mairimunt 100wordis)

The purpose of this project was to manufacture women's shirts for fit and performance testing by the U.S.Army. Patterns were received from the Navy, Air Force, and Coast Guard by way of US Army Natick andarrived In the form of ozalids and computer files. Two shirt styles were to be produced, #1 with AirForce styling and Air Force Collar, #2 with Navy styling, Air Force collar, and Coast Guard pocket. Dueto multiple inconsistency and inaccuracy problems with the patterns, it was resolved that the Navy bodywouid be used as the foundation for the two styles. Accurate, consistency graded patterns were createdfor the two styles, and, following the approval of additional prototypes, 48 shirts were produced.Detailed garment dimension specification ;charts were provided for future quality control. Recommendationswere made for modifications to enhance producibility. Following a fit test, medifications to thepatterns were made by Natick and 880 shirts for wear test were produced. During the course of theproject, problems related to the inappropriate use of CAR technology, typical of the industry at large,surfaced and were addressed.

14. SUai CRMS IS. NUMHE OR PAGES25

Apparel manufacturing, women's shirts C RO'

17. SECURITY CLAWSIFICATIoN 15. SEC.W ITY CLASSIFICATION 19. SECURITY CLA$i ICATiON 0. UMATATIO OF ASTRACT

OF REPORT j OF TIllS PAGE OF ABSTRArf

Un lass] fled I Unclassified I Unclassified UL

NSN 7540-01-80-500 ''tndard Form 295 (Rev. 23-9)09t ,. W 11A0. 139-16Molot

Research Women's Shirt Production

for

Clemson Apparel ResearchDLA900-87-D-0017 DO 0023

by

Dr. Nancy J. StaplesVisiting Professor

Clemson Apparel ResearchClemson University

Clemson, South Carolina

Aceas51 For

NTIS QII&I

* kaouveo Ua ]

Statement A per telecon Availability Codes-Helen Kenlin DLA/PRM Cameron Station Aai--a~d/orAlexandria, VA 22304-6100 D vst penil

NWW 6/11/92

92-15187

92 6 09 083

PRODUCTION FOR FIT TEST

Receipt of Patterns

The project to manufacture sample women's shirts for fit andperformance testing by the US Army was begun in early March 1991 withthe arrival of patterns at CAR from Natick. Shirt #1 arrived as ozalids ofthe US Navy women's shirts in sizes 32 through 42 with neck sizes 12through 16 and sleeve lengths 29, 30, 32, and 34. The graded Coast Guardpocket pattern and a dimensionally adapted version for the Army were alsoprovided. Shirt #2, the Air Force women's shirt in 17 odd and even sizesfrom 6 to 22 with regular and plus bust sizes (spec difference 3/4 inch) andshort, regular, and long sleeve lengths for each size, with the necklineadapted to receive the Air Force women's banded collar, was received asMicrodynamics computer files of the sample size 10 and an accompanyinggrade rule table.

Analysis of Patterns

An analysis of the patterns revealed that there were multipleproblems. Shirt #1 was erroneously graded such that the cross shoulderwidth grew with each increasing neck size for a given bust size (SeeAppendix A). This resulted in a difference of 1-3/4 inches of cross shoulderwidth for the range of a size 12 neck to a size 15 neck with a 34 bust. Due tothe difficulties associated with trying to adapt this grading to the Air Forceneckline, it was decided by Martha McCagg to leave the Navy neckline andattach an Air Force collar of compatible neck length for the Shirt #1prototypes. The Coast Guard pocket grading was eliminated and one sizepocket was used on all sizes. The pocket pattern provided had to beincreased in width by one inch in order to achieve the required finisheddimensions. Three shirts were thus produced in sizes 32/12/32, 36/14/30,and 40/16/34 (bust/neck/sleeve).

Shirt #2 was riddled with problems. The patterns had been input asgraded nests. This technique, if not "cleaned up" after input, results inexcessive numbers of grade rules with inconsistent grade increments. Itallows the computer to determine grade increments between sizes base(; onthe input in every size of each grade point on the perimeter of the pattern.Since the rules are written in ten-thousandths of an inch, even minutedifferences are picked up by the computer as separate rules. The fact thatthe regular and plus bust sizes had been input with the necks in oppositedirections only complicated matters further. Pattern pieces which shouldhave been symmetrical were not. Necklines which should have been thessme shape were not. Sleeve lengths were digitized separately instead of asgrade rule variations to the same sleeve. The side front was graded in such

a way that the princess seam over the bust had the identical curve for all 17sizes. Pieces which should have had seams conforming to the x or y axiswere skewed. Upon examination it was discovered that the bust of thecenter back regular size was lar than that of the plus size (see AppendixB) and that the cuff was too small for the sleeve. As a result of theseproblems, Barbara Quinn provided a revised Shirt #2 computer file whichapplied the Army grade rules to the Air Force shirt. Authorization wasgiven by Martha McCagg to improve the precision of the pattern pieces suchthat parts fit each other and that the plus sizes were in fact 3/4 inch largerin the bust than the regular sizes. Three shirts were produced in sizes 6short, 14 regular, and 22+ long.

COTR Review and Resulting Modifications

On March 28 and 29, Martha McCagg and Captain Kathleen Battenmet with the PI at CAR to inspect the pre-production shirt prototypes. Theproblems associated with each shirt were discussed, as well as the potentialvolume of uncontrollable variables prone to error in assessing fit based onthese two shirts. It was resolved that the shirts for fit test would be limited Yto the use of the Navy shirt body. Anthropologists Ken Parham and HelenBelaources wou!d provide biacromial breadths, intercye breadths, and neckto apex measurements from which the erroneous shoulder grade could becorrected for both shirts and the Shirt #2 could be remade using the Navybody with Air Force styling. For each bust size, all shirts would use the Air,Force neckline and collar in two sizes (small and large) plus two of fouravailable sleeve lengths (A=29, B=30, C=31, D=32), each with double buttonsfor adjustable fit. Following the receipt of the necessary data in early April,six additional prototype shirts were produced, Shirt #1 with Navy stylingand Coast Guard pockets in sizes 32SB, 368C, and 40SD and Shirt #2 withAir Force styling in sizes 32LB, 36LC, and 40LD.

It should be noted here that the anthropologists were unable toprovide the measurements necessary to confirm a cross shoulder widththat would conform to current anthropometric data. The nearestmeasurement available was an Intercye II, the distance between the top ofthe axilla fold on each side of the body across the back. The biacromialbreadth was not among those recorded by the anthropologists, in spite of thefact that the cross shoulder measurement is critical for patternmaking andquality control in the manufacture of tops. As a result, the most logical-seeming cross shoulder width of the four in the Navy size 34 sample sizewas selected as the foundation for the re-styled pattern.

Manually-traced patterns on the US Navy ozalid were evaluated forprecision by nesting individual pattern pieces in the full range of sizes. Bythis means, Lhe inLended grade rules were determined and inconsistenciesinherent in manual transfer were identified. To avoid the possibility of anygraded size being too small, since the bust grade, for example, was notconsistent between sizes, the smallest size (32) was used as the sample size

2

to be input into the computer and to which grade rules were applied. Afterthe pattern pieces had been digitized, the seam allowances were removedand all pieces were "cleaned up" to insure a precise fit in garmentconstruction. The line lengths and shapes of all seams on all pattern piecesthat would be sewn together in manufacturing were checked to -- insureprecise fit (the computer quickly and easily displays line lengths andpattern pieces can be manipulated to lock together on seams of the sameshape or to "walk" automatically to check the relationship of curves to eachother). The Air Force neckline was tranferred to the Navy body to createsmall and large neck sizes for each bust size. Grade rules were written toaccomplish the specified neck size changes. Graded nests were checked forline lengths in all sizes on coordinating pattern pieces and grade ruleswere adjusted where necessary. Only after all pattern pieces in all sizeshad been determined to be correct, and measurements had been recorded,were the seam allowances added back, the markers prepared, and theshirts cut and assembled.

The specified shirts, with an accompanying detailed garmentdimension specifications chart (see Appendix C), were delivered to theCOTR on April 26, 1991. The specifications chart included half bust, halfsweep, cross shoulder, oversleeve, undersleeve, half bicep, half cuff(buttoned) center back length, and collar circumference, representingactual pattern dimensions measured by the computer and recorded in threedecimal places. The prototypes as constructed were approved by the COTR.

Production and Distribution of Fit Test Garments

Authorization was obtained on May 28 to produce one of each styleshirt in each of six bust sizes, two neck sizes, and two sleeve lengths for atotal of 48 shirts. Production was begun May 29. The shirts were measuredand approved at a COTR inspection on June 7 prior to fit test June 10through July 3.

Following suggestions made by the plant operators, patternmodifications were made to improve the sewability of the shirts. Theseincluded making the seam allowance of all seams to be safety stitched 3/8:inch (to facilitate finishing precisely at 1/4 inch), increasing the seamallowance on the pockets to 1/2 inch (to facilitate creasing), and slightlyadjusting the radius of the curve on the cuff (to be compatible withautomatic cuff running, turning, and topstitching equipment).

The principal investigator assisted at the fit test of shirts, skirts, andpants in Alexandria, Virginia on June 17 and 18. Observations were madeand suggestions given for modifying some construction features to makethe uniform components inherently more alterable. This included a ratherunconventional sequence of operations for a princess-to-the-shoulder, three-,,piece-back shirt. It was suggested that if the side panels were joined at theshoulder, and the center panels joined at the shoulder, followed by the

3

princess seam being joined from back hem to front hem, a shouldernarrowing alteration could easily be made to this safety stitched seamwithout having to rip out any stitches. A similar suggestion was made forthe attachment of the skirt waistband at the side seams.

PRODUCTION FOR WEAR TEST

Receipt of Patterns and First Article Approval

Following the fit test in June, during the months of July throughOctober the project was on hold, awaiting the beginning of Phase 2,manufacturing shirts for the wear test. Meanwhile, during the month ofSeptember a sample necktab was received from Natick, digitized, andreturned in preparation for possible production of this revised necktab forwear test.

During the month of November the project continued with the receiptof revised patterns, sample garments, and instructions for the production ofClass 1 and 2 tuck-in first articles. The two shirts for the tuck-in firstarticle were produced. During the month of December the projectcontinued with the receipt of patterns, sample garments, and instructionsfor the production of Class 1 and 2 overblouse first articles. The two shirtsfor the overblouse first article were produced and on December 10 theprincipal investigator went to Natick for the approval of all four firstarticles by the COTR. A list of the numbers of garments in each of the fifty--:" 4

six sizes to be produced was received, and authorization to begin production ;was given (pending the installation of a new numerically controlled cutterat CAR). An error in the coordination of sizes between non-proportionaltuck-ins and proportional overblouses was corrected by the COTR. Thegarment dimensions chart provided for quality control included half bust,half sweep, center back length, sleeve length, and collar circumference.

Observations of Methods and Procedures at Natick

4

During the visit to Natick, the procedure for producing samples andfirst patterns and the work flow through digitizing, grading, and patterncutting were discussed. The principal investigator sought to determine-thesource of inaccuracy in the patterns coming out of Natick. Fit test patternssent to Natick in Spring/Summer 1991 had been checked, without seamallowances, to insure that all pattern pieces fit together. In addition, gradednests had been measured to insure that grade rules had been correctly;..written such that all pattern pieces would fit together in all sizes. '.The:modified patterns received from Natick in November and December 1991, did;'not reflect this precision.

Because it had been assumed that the patterns received were asprecise as those sent earlier in the year, in preparation for the assembly of.the first article shirts, the principal investigator had only adjusted the seam Iallowance depth (to permit more accurate safety stitched seams), then-made markers, had the pieces cut, and authorized the assembly of the partswithout checking pattern precision. The principal investigator was firstalerted to potential problems when a sewing machine operator pointed out Aher difficulty in attaching the collar band to the collar of the first articletuck-in. The collar band was too long for the collar and the notches did notmatch.

Upon investigating the problem (by calling up the pattern pieces onthe computer, stripping the seam allowances, and walking the patternpieces) it was determined that the collar band was 3/8" longer than thecollar edge to which it was to be attached and the notches were out ofalignment by nearly 1/4" (see Appendix. D). Further investigationdetermined that the grade rules applied to corresponding points on thecollar band and collar resulted in collars that would could not meet thespecs provided and, even if the sample size collar were adjusted to fit thecollar band, the collar and collar band would not fit each other in all sizes(see Appendix E). In addition, there were duplicate rules in each rule table,rule tables already inflated in size due to the assumed inability of Natick'scardboard pattern cutting device to read mirrored pieces and opposite rules(Microdynamics later confirmed that the cardboard cutting device receivesall data as x,y coordinates and therefore CAN read mirrored pieces and,opposite rules).

During the visit to Natick it was discovered that the procedure usedencouraged the inaccuracy encountered in this project. AlthoughMicrodynamics computer equipment is employed for digitizing andgrading of pattern pieces, first patterns are still made manually andtransferred by digitizing the cut lines. The resulting patterns are neveracleaned up" to insure that all pieces will fit precisely. Because seamallowances are not removed, graded pattern pieces cannot be determined tofit properly in all sizes. The erroneously assumed constraints of the patterncutting device causes excessive numbers of grade rules. It is unfortunatethat the method of operation at Natick does not permit using the investmentin Microdynamics CAD equipment to a greater level of its capability.

5

Production and Distribution of Wear Test Garments

Following the visit to Natick, corrections involving improvingprecision, but not modifying garment design, were authorized by the COTR.The patterns were "cleaned up" and during the month of January theproduction of the shirts for wear test was begun. As an aid to keeping allthe sizes straight, each ply of cloth cut included one of each shirt in thesame size (four shirts per marker). A chart developed for all the cuts (seeAppendix F) aided the marker maker, cutter, and plant manager tocoordinate the work. The first delivery, the long-sleeve tuck-in shirts, wassent on February 18. This included 220 each of a shirt with pockets and ashirt without pockets (see Appendix G). The overblouses were shippedMarch 4. This included 220 each of a long-sleeved and a short-sleevedoverblouse without pockets (see Appendix I). The principal investigatorwent to Alexandria, VA on February 27 to observe the distribution ofuniforms for wear test.

FINAL OBSERVATIONS

Although the intent of this project was originally simply tomanufacture sample women's shirts for fit and performance testing by theUS Army, by necessity it additionally evolved into a comprehensive analysisof military patterns and patternmaking procedures. The PI believed thatthere was no sense in producing, for fit or wear test, garments which wereinherently imprecise due to pattern problems. The resulting discoveriespointed out numerous ways in which the available technology could be usedto produce more precise and more consistent uniforms by insuring that thepatterns distributed from Natick are the best that they can be.

Lest it be assumed that the personnel at Natick are at fault for thequality of product they provide, in spite of their access to technology, itshould be pointed out that they are not alone in the problems this projectidentified. The apparel industry as a whole has either: 1) not embracedcurrently available technology to help it do a better job, or 2) not followedcapital investment in technology with a change in methods and proceduresto exploit fully the benefits of the technology. When CAD equipment ispurchased, the CAD vendor provides instruction on the use of the 2,equipment, how to use the vendor's software to perform specific tasks. The 'trainer does not, however, have time, nor does the vendor necessarily havethe responsibility during initial training, to provide additional instruction

6S,-- 'm

in how to change from a manual mode of organizing work to a computer-aided mode, including allowing the computer to perform tasks that wereeither impractical, or not even possible manually. During the initialtraining the focus is on learning to use the tool, and rightly so.Unfortunately no one has yet formally addressed the question of when, bywhom, and at whose expense training to adjust methods of operationemploying CAD technology should occur. Since it is the mission of theAAMTD's to encourage the proper application of advanced technology, it isonly fitting that the raising of these issues should emanate from ClemsonApparel Research.

7

Appendix A

I Si #e e g

4"l

/ " 1

I Shirt #1 was erroneously gradedsuch that the cross shoulder width..grew with each increasing neck sizefor a given bust size. This resulted ina difference of 1-3/4 inches of cross 1shoulder width for the range of a size12 neck to a size 15 neck with a 34bust. "

II 1

I I ,I

s I

I'

t / ~;', /i '~

N -

, .. , ,d.',..

.._-_,,.__,.__ _ j

J . |

Appendix B

,N.~ I -~

I. 4

I 1..1 II 1.I I .1I I I.I II. 1.

I I..1 I /I.

N. .1 (N. I I

I II II 1

'I I. I*1* .1 1

I I.I "KJ I1 .1*1 I.I I. II I.I I. 1~*1 I -- ~.----.. -~

g Upon examination it was discoveredI. that the bust of the center backI regular size was hrgm than that. ofI.

the plus size (the plus sizel isindicated in red).

.1

I -. ~-----*1 I..1

I..1I. A

* . .~ *.t4*

Appendix C

C- 0 C#) (D ccccoc o 00 U') IqN fin I* C4 l U) Nr I Ln A LO C6N C Li

.4 14 i r- i (6 r- in o 0 )CM N Mm N 04CI N~

o (0 U - 14 101101 in to on . Q -D

cm 0 co cm N cm N N N- i ',N co m .m co In C6 C) ) u

co w ~ tn it m1(N CD cD m- 0 -DLO

CM 0 V > 0 o (D((( CD o 00 CY - CM N10 v~ m in 000 C) CD c

z N Nl N T- N CMJC CJ CMo6

LL(D (0 I) I~ t C) C) C) 0 3) 0) c)) . ~cm 0 V) C~t NmNNiN (D C) cc

o an N V- cococcc 00C C ). Lj

a.- C- (60 i 14'J~( In C6 U)NO0 N N~ V- N cmN N j

v w-t c oc o c V- m T - -n L

to 10 N- P- LO ( c ~ )~ In inL an m 4 u

N 0 cm) cm 100m 1 1m cm N c

Noo U Int CO Nr CI 101 V01 IIl 0- a) 0

0q 0 '- m0 cn' ( mC m m N '- Ui

Nm Nm Nm .m m N

o~~~~ -n -- - -~'c ' -) -) . - -N 0 C) cc NNNN cc N a

a. 0 0- 10 CO'J1( U) t

N W LL

U) LL c c

iCj W) Wo cc 0Jw ccl

>)_0__' 0 0

o C,4 C-

Appendix D

Upon investigating the problem withthe collar and the band (by calling upthe pattern pieces on the computer,stripping the seam allowances, and

walking the pattern pieces), it was

determined that the collar band was

3/8" longer than the collar edge to

which it was to be attached and thenotches were out of alignment by

nearly 1/4".

Appendix E

1 4

A

I,

4I * ~

0))

0o0*~ ~ 1.4 0L C

pc4 c4 -4 0 -q -- -q -q 04

c.co

Appendix F

00 0-1 01 -4 (r-4r-4 r-4 r4

a) a a)a) CO

E-1E 4 1- E 4 E-4 E-4

140 10 40 40 140 140Iol -Io

~CD 0Aa 00 00r- -

'CID

10I DC 0000 000 2 r414 1- -

ca co ) U

cd)I c a)$ ) a)CdcoC a) c),0 40 10 4 40 40 40 40 P4 0

~co r-1

.1 IU $w

0-

E -4 E -4 E-4 E -40 .41 0 1 -40

CO) CO) co Cl) CO)

640 140

A.o

~?-- co''40 40 ~~0 '0 '00

4

210C) l C"O, Cl)l

0 4101 140 140 '10 o 4

r-q cq: - qL

ClCD

00.

*t !

d - ~ -4El

$ cI2$'.1 *0

U)C~ VR 1-4

AE-

CeCL

"Z t ..

. ..... . 4 ...... . . .. . ........ . . .... .

Appendix G

"all.

ly ,

Appendix H

fi