Document jg4QJZZ07X9vnrYvo2VR9gNRZ

COHFIBiEOTXAL DUPONT MARSHALL LABORATORY LIBRARY mE, 1, PORT BE NESSOBRS & CONPANT FABRICS & FINISHES DEPARTMENT PRODUCTION DIVISION BHLABELPHIA PROCESS EHSXMEERBI9 III ffil CUCIMTOi APPEARANCE CHARACTERIZATION & CONTROL OF STRXP-COATIN FINISHES - PART II BVs R* H. TOINQ 6J Bat issued: Member 1961 Projects P-60, Appesranee Characterization & control of Strip-Coating Finishes ABSTRACT Problems Inherent In eolor mmvtmmntB of flat and iai*gloss products are discussed and. instrumental methods jpermittlng such measurements are detailed. ?-60 APPEARANCE CHARACTERI2ATI0M M CONTROL OF STHXP-CQATIHS FINISHES Faarfe II Application and Alignment of Colormaster for Reflectance Measurements of Plat and $emi~0loss Finishes Introduction Part I of this report has reviewed the basic conditions that must prevail if color is to be shaded and controlled instrumentallj. If these prerequisites are recognized and accepted, then many of the problems hindering the successful application of instrumentation will be eliminated. In the interval, formulations not adhering to the basic concepts, particularly with respect to pigmentation, can be .successfully handled, but only by individuals capable of exercising a high degree of Judgement* The .belief that the'instrumental approach to color results in a mechanical system is highly erroneous, Judge ment is m integral part of the work* Experience has shown that panels measured at gloss levels . equal to standard, have not posed problems but color measurements of two panels at different gloss levels have been a problem* Part XX of this report covers the problems inherent in color measurement and control of flat and semi-gloss products versus standards of different gloss levels. The weakness of all tristimulus colorimeters where surface irregularities of the finishes being measured can result in light scattering, hence partial measure ment of specular gloss along with light emerging from the pigmented film, have been recognized and guarded against in the past through the design of the instruments themselves. In the measurement of flat and semi-gloss products on is faced with, by the vary nature of the products, color measurements on irregular surfaces where controlled scattering of specular gloss is an integral property of the product. It has long been accepted that color adjustments and batch turnover time on flat and semi-gloss products are much greater on these products than on high gloss products. The eye has difficulty in distinguishing between color difference that are solely func tions of gloss and real differences indicating hue variations, hence the need for instrumentation. The tri-stimulus colorimeter, properly standardized and used by skilled personnel can accurately measure color differences on these products. By analysis, the operator can then determine what portion of the measured color difference is due sole ly to gloss variations, and what portion is a result of hue variations, Techniques permitting accurate and reproducible reading using the Colcrmaster on flat and semi-gloss products are examined. Results It is apparent from the work completed to date that gloss variations of equal magnitude have an increasingly greater effect as the reflectance of the color decreases? l.e., dark colors are more greatly affected by gloss variations than are light colors * In all colors a reduction in gloss results In an apparent increase in the lightness of the color. This reflectance change is a result of the scattering of the gloss component of light reflected from DUP030002248 -2 - the film became of the irregularity of the fl gurfaen caused by the flattening agent. previous w o j ?te has shown that this gloss component has essentially the same color composition as the light striking the film. To neutral grays, the essentially equal G, H and B reflectance values are approximately equally affected by the gloss component, hence the effect of gloss variations is primarily in the lightness direction with little or no effect on the hue or saturation of the color (note Exhibit A). , In saturated colors, the unequal G, H and B reflectance values are not evenly affected by the gloss component' resulting in a three-fold effect, in that iostrumentally a pronounced change (loss) in saturation as well as in hue occurs, plus the Increase in lightness previously noted for neutral grays (note Exhibit B), Visually the eye accepts these colors as being equal to their respective standards and does not detect the hue and saturation changes measured instrurontally, The eye does-detect minor variations in lightness levels in the dark colors, but these are not considered to be objectionable based on visual inspection-by trained color personnel. During the course of this investigation it was noted that different Colormasters varied in their ability to measure panels of different gloss levels relative to a standard, . This fact was confirmed by a "round robin" test on panels of varying gloss incre ments prepared at Process Engineering and divisionally read, The data showing the variation between Colormasters within the Division have been compiled and distributed by B, A. Miller. These data are included as Exhibit C in this report. Conclusions 1) Techniques' that permit accurate and reproducible reflectance measurements on flat and semi-gloss panels have been developed and are detailed in this report. 2) lack of visual and instrumental correlation between panels differing in gloss can now be explained, since it has been established that tri-stimulus colorimeters detect color differences not discern ible by the eye. This condition exists In nonmetameric matches, 3) Alignment of Divisional "Colormasters" to make meaningful Division color tolerances can be readily established using techniques detailed in this report. Discussion The results of this investigation indicate that the response of the "Colormaster" Differential Colorimeter to gloss variations differs with each unit. These differences are' believed to be caused by very minor variations with respect to positioning of the DUP030002249 ___ 3 optical 090tea and phot tubes la each instrument, Other tristimulus instruments would show the same variations. la addition to the variation noted between clean units, it Is possible- to obtain considerable variation within an Instrument because of variations in the cleanliness of the filters, optics, mirrors and gloss plates above the photo tubes. Hals Is not surprising since the presence of grime and dust particles will tend to dis perse the light m much the saw say as the flattening component of a film does, this will produce an additive effect when coupled to film gloss of a panel. In order to remain instruaenbally consistent on a given Colermaster it Is essential that each instrument be rigorously cleaned prlorto initial use and that daily cleaning of the glass plate over the photo tube .and the mirror wing a camel hair brush be strictly enforced. Dust particles accumulating on these sur faces can result in major mmws'&mnt errors due to their ability to scatter light, To assist further it is essential that each plant site number their Individual colorimeters, and all readings pertaining to a batch be made on the same colorimeter. Cleaning of all optics following P,T,S.?, #18, page 9, section P-2, should be established on a cycle based on conditions existing at each location. The following procedure is recommended for obtaining instrumental color readings on batches which are to be instrumentally controlled, where gloss differences exist between the batch and standard due to the use of a high gloss color standard. 1) Select an approved batch that is visually considered to fee mUM equal to the dry platSHBOffinfal. 2} Calibrate the dry plate standard on a properly cleaned Colormaster versus the white ceramic For example Values determined for dry plate standard versus white ceramic plaque. 0 78,10 R - 79.60 B * 79.02 3) Read the visually approved and accepted batch panel versus the calibrated dry plate standard. Record 0, R and B values for the batch along with the Plant colorimeter number, t) Repeat on each additional Coloraastr, first standardising the Colortsaster using the 0, R and B values for the dry plate standard (determined in point 2). Read the selected batch versus the dry plate standard, recording 0, R and 3 values along with the Instrument number. Designate values as ' On, Rn and Bn. values of visually accepted batch DUP030002250 4 (point 1) versus dry plate standard.. #1 Oolpnasaster #2 Oolormaster 78.10 R 79.60 B 7902 1 78.41 Rl 80.50 B1 * 74.10 2 - 78.90 R2 ** 79.00 Bg * 76.10 , #3 golormaster To control production batohes, using a clean and properIf serviced goXormaster, the following procedure should he used: 5) Standardise the Oolornaster to the values determined for the dry plate standard in step 2. 6) Determine , R and B values for the production batch versus the dry plate standard using semiabsolute colorimetry. Reduction batch versus D.P.8. read on Oolongster #3s D.P.S. Batch - 78.10 R 79.60 B * 79.02 :U;8 -73.96 R B 7) Calculate delta values using for standard values, those values determined in step 3, not the values shown for the calibration of the dry plate standard. Care must he taken to select those values assigned for the particular Colorsaaster used. An example of the calculation involved is as follows:; determination of &l da, 4b values for Colprwster 3. Standard Batch RS3 - z7s9.40 B 76.14 I: ?f:H B - 68.84 AL - -2.40 A -1.04 Ah +1,52 A summary of the steps involved are shorn below: (A) Values determined for dry plate standard versus white ceramic plaque - * 78.10 R * 79.60 B * 79.02 DUP030002251 5 (B) Values of visually accepted batch (point 1) versus dry plat standard - P.?,S. #1 #2 #3 Colomaaster Colorsasster Colermaster 78.10 h * 79.60 b - 7%mt Ql 78.41 h i 80.50 Bi 74,10 2 78.90 HR 7900 Bg 76.10 3 79.50 H3 79.40 B3 76.14 (0) STeduction batch versus B. P.S. standard read on Colomaster 3? P.P.S. Batch S - 78.10 R - 79*60 79*2 0 73.9 1 Ill84 CP) Determination of 41* A&a A* for Colormaster 3s Standard Batch 3 79.50 S: H:l? G 73.96 R B 8:1 4L - -2.40 A sa -1,04 6h +1.52 Divisional Alignment of Colormaster When two or more location as? Manufacturing the am fpdue% code and close alignment is mandatory, it i essential that the CoiowMter at each site ha aligned with respect to the product. This is critical where gloss differences wist between product and approval standards, if alignment 1 not Bade, then passing tolerance with respect to a standard can he farcical because of varying sensitivity of each unit where gloss is a factor in measurements. Alignment of coXonsasters from plant to plant is relatively simple and straightforward. This can be accomplished as follows 1) The "parent" plant must end to d&ch manufacturing site a panel of the visually approved ;tateh, along with a dry plat standard, assuming this standard is not already available. 2) tFpoa receipt of these panels, the plant should standardise the Colcnwstar to the values shown on the dry plate standard. The visually approved batch panel should then be read gainst the dry plate standard and the values for G,?R\and B duly recorded along with the instrument number. DUP03Q002252 6 These values are then used for calculation purposes, as shown in the preceding example,, The passing batch panel should then be returned to the parent plant,* or alternately, if a calibrated dry plate standard is available at the plant then only the approved batch panel need be forwarded. tFpon receipt of this panel, the panel should be carefully read versus the dry plate standard vis serai-absolute colorimetry and a, R and B values duly recorded showing the instrument number the values apply to. These values should then be used for calculation purposes. The approved batch panel should then be returned to the parent plant. DUP030002253 neutral gray 0 Flatting Figment Lightness * 85 Gloss ' 86,0 0 E !I:i B 66,20 &h m m 1.5* 20 30 3o50 40 50 43,5 3:8 66.28 4* .01 - ...03 4* ,05 30.50 66.40 40.03 0.00 - .14 22,80 68.30 68,92 66,62 4* .20 * .08 - .04 12.50 6699,,0626 67,28 4* .56 - .07 4* .01 6.30 66.96 I?;! ?X:S + ,**5 5,30 70,19 S:3 4-1.13 4- .02 4* .42 4.30 70,12 70.80 68.06 4*1,10 - .15 4* .24 Lightness JiJ Gloss 85.50 G 21,16 R 21,54 B 20.64 Ah A &b 46.50 ' 34,50 21.32 21,44 21.73 21,83 20.80 20,93 +0.16 4* .28 +0.12 40,11 40.00 -0.02 23.70 21.52 21.88 21.04 40.36 40,02 -0.10 9,80 22.71 23,03 22.31 4*1.60 -0,17 -0.22 8,10 23,10 23,52 22.84 4*2.06 40.02 ~ .35 5.50 23.91 24.24 23.58 42.78 -0.12 -0.33 2.00 24.78 25.11 24.45 4*3.6 -0.17 -0.38 Mishtness ^ 15 Gloss 0 E B 87,0 2.755 2,697 2,939 45,00 1:$ 3,220 36.00 3.078 3.086 3,290 21.3 3.534 3.468 3.780 10.40 4,985 4.858 5.244 8.00 5.130 5.012 5.391 5.30 5,857 5.700 6.124 2.90 6,720 6.550 6.942 AS* 4 .84 4*1.08 42.46 7,57 4*7.97 49,85 4*11,88 Aa Afe 4 .46 - .18 4* .39 - .18 40.54 -0.32 -4*..2130 - .18 4* ,12 4* oil 4- .50 .50 lOTEs Glees readings were made at 60 and are reported to show order of magnitude of gloss change resulting from addition of flatting plgwnt, It is recognised that certain readings are outside the normal 60 limits, but are presented to avoid complexities resulting from seel changes. DUP030002254 EXHIBIT B 885-0557 - MgHtness % 85 % Flatting Pigment 1.50 a * 70,44 70,24 70.29 H 62.10 61,88 61.93 B 87,28 87,20 87.25 AL - .10 - .08 &a - .02 4* .01 At - oil - .11 2.00 70.45 62.10 87.41 0.00 *8- .02 - .08 3.00 70.88 62,44 87.94 + ,22 - ,02 - .09 3-5S 88.03 -AO + .01 4,00 71.19 62.72 88.30 + .36 - ,04 - ,15 5.00 71.82 mi + ,6? - .16 0.00 L V 50 0 20,66 H B 10. Jg 57,78 M* m Ab 20.92 10.81 57.92 + .28 + .39 + .31 21,13 11.08 58.10 + ,56 + ,62 + ,52 21,69 11.70 58.53 +1.10 + ,69 +1.12 22.31 12.36 59.10 +1.74 + .92 +1*66 22.68 12.76 59,41 +2.12 + .95 +2.01 22.92 13.01 59.63 +42..9367 +2,21 23,53 13.66 60,13 +2.98 +I.03 42,77 L A 15 0 2.400 R .940 B 15.22 AL m Ab 2.713 1.185 15,23 + .99 -1.39 +1.65 2.918 1.397 15.43 +1.88 -2.16 +2.64 3.139 1.610 15.71. 42.74 -2.92 +3.53 3.732 2.206 16.36 +4.90 -4.69 +5.57 4,042 2.457 I6.63 +5.94 -6.12 +6.99 4,378 2.831 17.13 46.98 -6.18 +7.68 5.228 3.696 18,09 +9.42 -7.50 +9.78 DU P030002255 EXHIBIT B (ContM) 885-036S - lightness 85 $> Platting Base & 1,5$6 Q R ** B 72,01 71-95 71.96 79o2? 65,94 P65.'8i4f 79.31 65.82 At - .02 - .02 M 00 *8*1.12 0 4 .05 + .11 2# 71.92 79.37 65.73 - .04 +2.24 + .13 3% 72.28 + .13 +2.10 + .03 3.5$S 72.41 I.83 66.26 + ,20 +2.18 + .05 73.68 81.11 67.48 + .81 +2.02 + .05 % 73.68 81.20 67.44 + .81 +2.15 + .09 h V 50 0 20.14 20.17 20.27 20.48 R 31.12 31.13 31.26 31.44 B 11915 11.22 11.34 11.68 At 8- .04 4*. .15 + .38 m - .06 - .03 - .22 Ah - .14 - .27 - .77 {*) Visually appears si, blue 21.04 31.98 12.31 + .98 - .56 -1,10 21.19 32.16 12.49 +1.14 - .55 -1.59 21.70 32,66 13.09 +1.68 - .89 -2.14 21089 32,83 13.42 +1.88 -2ill V* 3 0 74.89* 74.99* R 15.02 15.01 B 9.089* 10,59* (*) 10K multiplier 75.70* 15.10 11.06* 76.29* 15.16 11.93* 78.06* 15.38 13.61* 79.01* 15.43 15.50* 80.20* 15.58 16.54* 86.96* 16,25 24.88 i:84L 4 + ,03 + .03 t:U + .30 -$ + .85 -1.67 +1.09 -1,78 +2,41 -2.07 40 - .83 -1.20 -3.42 -5.17 -5.85* -10.99* {*) 'Visually appears slightly bluer than standard (**) Bluer-than standard, but not to extent shown instrumentally DUP030002256 KK3SIB1T B (Cont'd) 885-0630 - lightness 4 85 >1 Flatting Pigsaent urn $. 2 @ es 71,45 H es 80.18 B es 43.50 71.37 80.09 43.49 71.38 80.12 43.48 71.54 SO. 29 43.68 AL - .04 - .04 + .04 mA + .05 + .03 - .05 - .07 3$ m 44.59 .32 + .08 - .70 3.59* 78.67 81.20 45.32 + .60 XS V 72.53 81,12 45,08 * .54 - .02 - .90 %5 m 081 - .60 -1,08 55L * - Cl es *R B ss AL Aa Ah 25.14 35.17 31.35* 25.08 35.02 31.65 - ,o6 : :ll 2^.04 34.99 32.%* .1 - .05 - .90 25.19 35.12 34,76 + .05 - . 09 -2.10 25.52 35.42 39.32* .36 - .06 -4.19 25.74 35.60 42,01 4* .57 - ,31 -5,30 26.08 + .89 -r'M 26.67 36.52 54.63 1.44 - ,40 9.83 () . XQX aultlplltr DUP030002257 ill QQS 0, Eo VOUK . - MARSHALL LAB* J, 0, DOHERTf - VBSLho N, Ho EDVARDS Philadelphia Plant September 25, 1961 IROSESS EN0XNEERXNQ MEMO AN 2WESTIAT10I OF MEASUfflQia TWO PANELS OF THE SAME COLOR HIT DIFFERENT QLQgg LEVELS OH THREE CQMMASTRRS' A previous study indicated non-ruppoduoltls results between fete eolormeetere when measuring two panels of the mm color but different gloss levels ** one a High gloss and am & flat* This study was extended to cover all of Us colomasters on the Phila delphia Plant to determine degree of variance that might be expected in the division0 Location of Colomeeter Sale Dcveiojwent Pit, X,CA* Process Eng, Process Eng* Lab, Control lab. B roup Panel 85 6?,60 67.60 67.60 67*6 67,6 67*60 Panel 9-85 + 6s8 Syloid "jBil 9*93 69*35 69,87 69.98 70*14 70*10 2*33 2,27 2*32 2*0 2,5 A similar cheek will be made of tbe eolmostare in the division0 Panels to be airmailed to the plants September 22, I98I* The flat panel was oiled and then read on the Sales Develop ment and Plant X,C,A, ooloemaeters* X.C.A, Reading 1 1 Sales Develop, 1 2 P,E, loss @<-85 67*60 67*60 6f*6 67..60 67*60 67*60 67*6 67*60 67,60 Flat 0-85 66,91 66*0 66*89 66,82 66,80 66,0 68,1% (Vg-120) DUP030002258 EXHIBIT C (Cant o 1) The oil film thickness does affect the @ reading tut the ropiness of tb oil film seems to have some effect* It was found the ost reproducible results were obtained by oiling tto panel freely then wiping off with a. dry cloth. It was reeonaiiM to 0, Volk that some* mians be developed to rals the gloss of the flat panels and new values be established in order to avoid the variance between ooloraastara due to gloss. PROCESS EMXNSERBI& EAMsXIo E, A, Miller DUP030002259 EXHIBIT C CCS 0. R, VOIK - MARSHAL!. LAB. Wo H,, TOOIE - "* W0 Ho EDWARDS Philadelphia Plant October 16, 1961. COL0RMTER-CHART COLOR ADJUSTMENT S3TEfS (P-57) USE 08? SEMI-OLQSS STANDARDS FOR CONTROL OF FLAT PANELS _ DIVISIONAL When we were unable to reproduce tbs 40 value of a gloss and a flat panel using two eoloswsters in the Philadelphia plant, a survey was made to include all eolormasters in the division. Tbs following data, show the 40 difference as measured on color* masters in the plants and the same panels measured in Process Engira* eerlngg ... P.B. AO 1,80 " 1,42 " 1.59, " 1,21 l l-P ! 1.49 1 1.49 ; 2,33 * 2.33 Panels not returned. AO 1.07 ! 1.07 :?g # .70 1.30 1.20 0.96 i. 0,64 1,15 1.05 2,12 1.91 1,96 2.32 1,49 1.15 .60 :8 .57 Chicago Tucker Phlla. Newburgh 99 FlSiInt k$m Totlsedo 85 So .SF. Parlin 81 I XI I II X II I II III I II I II III P.Eo minus plant , 40 values 0.50 + 0.22 + 0.59 * 0.42 + 0.43 + 0.34 + 0.44 + 0,21 + 0.42 * 0,00 + 0.47 - 0,05 f 0.62 + 0.13 All of the panels sent to the plants were made from the same material and on the same day using the automatic spray gun. The parIs when read on the Process Engineering eolormaster should have had the same 40 values, but note the variance from O.fO to 2.33. This in dicates a weakness in reproducibility of the test. The panels were made in "puco'"* similar to the method used to make the present gloss versus color study. (*) Du Pont Registered Trademark DUP030002260 EXHIBIT C (Conft'd) Parlin reported that they could get smaller differences in 40 between colonsasters if they used the spinner. The large aperture was also used to evaluate panels with the following results on two eolormseterss FoEo Marshall lab. do 1.70 1.38 * O.32 (Ajax Panels) .81 0.50 + 0.31 ( fella, Panels) The same panels measured on the Marshall Laboratory colormaster on the regular port were as follows % 40 2.1177 (Ajax Panels AO 1.3399 (Phils o Kane Th large aperture decreases the AO values but did not eliminate differences between eolomoastera. panel. The only solution to date is to raise tbe gloss of the flat BAMtXL PROCESS ENOIHEERXHQ E. A. Miller DUP030002261