Document 6BomKxY2nj9E0ZOjrXwBQmEe6

/ # Ills 1 MlfiSilL L Indexed / R. B. Davis, Lai). 0. H. Bullitt, Exp. Sta. Patent Section, Lab. Marshall Lab. Library (5) Piles 183.0 Marshall Laboratory November 23, 1953. MEMORANDUM REPORT NO. M~6SS Prepared by w. P. Singleton m *> j.i| COMPARISON OF DOW AND DUPONT TESTS FOR STAB? REMOVAL AND SCRUB RESISTANCE OF LATEX FAXIT INTRODUCTION; The Dow Chemical Company examined samples of "Flow-Kote" on the composition current in the spring of 1953 -and compared these samples with 43 other commercial latex paints. Dow reports that 't.?low-Roten was poorer than average In .cleanability and scrubabilifcy. We have previously compared "Plow-Kobe" with several leading competitive paints and have found no deficiency in either of these properties. Since these results are in some contradiction, it is desirable to compare the methods used by Dow with our methods to make sure that our methods are adequate* SUMMARY;; Two representative wplow-Sotew colors and two corresponding colors of -Super Kern-Tone have been tested by the methods of both companies, Super Kern-Tone is considered an outstanding product .for stain removal end scrub resistance by the Dow Chemical Coiapaoy, In cur previous tests?, it has been easier to clean because It erodes faster than !5Flaw-*Kote,{. `'UXaw-Kote* 1b the current composition with Oelite as extender in the j white and ASP #400 Olay as extender in Berkeley Green. DOlTQLUSZQNSl 1. Both Dow and du-Pont methods agree in general, and they should lead to the sane practical Conclusions,, 2. The Dow test allows large variation between operators in -the N35485 number of strokes necessary to -clean os* erode a film. '`Flow-Kote'1 way be rated from very good to poor- depending on the operator. However, there is Qualitative agreement among operators in the order of exeellen.ee of a given set of films, .. 3. . .In scrub resistance, "Flow-Kobe*' and Super Kern-Tone White are close together. "Piow-Kcte" Berkeley Green may have less scrub resist- snee than the White. Super Kern-Tone Green is in a class by itself with extremely poor scrub resistance.. The showing of Super Kern-Tone White is relatively better in this test than in our previous tests, probably because the flits was aged two weeks instead of one week. 4. Super .ICeta-Tone cleans up e&ster. than "Flow-Koto". This is In accordance with previous results. The oleoreslnous modifier in Super Kern-Tone probably helps 1 stale removal? alkyd modification of "Flow-Hote* has been shown to improve the ease of cleaning, 5. The duPo.nt tests for stair removal and scrub resistance sight be made quicker and more reliable by using a weight on the .brush equivalent to the pressure of the band in scrubbing, and by other changes in detail. Since the duPont methods have teen standardised by agreement between the Mgmants Department and the Marshall Laboratory, no action is recommended s't^this tims. Further laboratory work would be needed in order to present definite recommendations EET1SLS; Type of Panel; The duMont panel Is wallboard with a coat of PrlneFTSoaTer followed by .a coat of Town and Country Flat wall. Paint tinted .gray, one coat of latex paint is applied ever this panel at a spreading rate of 420 sq.ft./gall The Dow panel is two coats of latex paint on tbs same type of waliboard, Dow does not specify the spreading rate; for this test, the first cost was. applied at 420 sq.ft./gal. and the second coat at 500 sq.ft,/gallon, (These spreading rates proved to be 0 little too high, 350 sq.ft./gal. and 4g'0 sq.ft,/gal, are a better choice fer .future work), prying Tiwe-s DuMont specifies one week drying for the final coat DW'speMTies 14 days drying. For the comparison of the two tests, all. panels were dried 13 or 14 days, Stains* DuPont procedure uses six stains applied to the panel about' oneUbur before the test "is run. The Dew test specifics three stEirvs3 of which one Is the same as in the dufent method, All eight ctKlns were -applied to each pan!, for comparison the methods 'The feteins are listed helows DUP030003618 DuPont Stains: Ink (Quink) rubbed on from a roll of saturated felt. Black Iron oxide pigment (W~215} rubbed on with the fingertip. #1 lead pencil. 866 Bb.erha.rd Faber red pencil. Lard containing l# black Iron oxide pigment smeared on the panel with the fingertips. Blalsdell .China Marking pencil# red (This stain is not Specified in the duPont method as written# but it has been used regularly in the Marshall Laboratory in addition to the five originally specified}, Bow Stains: #2 Lead pencil Blaisdell China marking pencil - red Blalsdell China marking pencil black Stain Removal Test* In the duFont stain removal test the panel is subjected-to three steps, eseh comprising fifty cycles with a cellulose sponge on the Gardner .straight line machine? Step Is Step -2? Step 3s Spang wet with %fi Ivory soap solution Sponge wet with 5% trieodium phosphate -solution Step 2 repeated with scouring powder sprinkled on the film, !%e -etains that remain are noted at the end of each step. An additional x ctteg is ?o#e at the end of the last step whether only' a trace of the sC-uin remains sr -whether the stain is still considerable. In in 'bow test for definability the panel is scrubbed by hand witha piece''of cloth towel that hae been folded to four thicknesses, lie towel is wet with a solution of Ivory Snow and wars water. ''Firm pressure is -exerted*, Scrubbing is stopped at forty -cycles. DUP030003619 /** i 4 Dow rates the panels from excellent to poor depending upon the number of cycles to remove all stains. Dow states, the reports on serubability and cleanability are comparative, with a group of samples being evaluated at the same time and results based on the average of two or more technicians The spread between operators was so great in our own laboratory using the Dow test that the number of strokes is not a significant way to rate the films. Test for Scrub Resistancei The Dow test for serubability is made at the same time as the test for cleanability. The number of cycles is noted when the film begins to erode or peel. Ratings are given according to the number of cycles. In the duPont "wet abrasion test", panels are scrubbed with a brush wet with 5/ trisodium phosphate solution in the Gardner straight line scrub machine* The test is stopped at 2000 cycles unless serious failure has oceured sooner. We notes (1) the number of cycles at which the first break in the film appears, (2) the number of cycles at which of the area of the film has been removed, and (3) stop the test when 2/3 of the area has been removed. The test is run in duplicate. Experimental Results in Stain Removal; There are eight systems in each test, comprising four paints each over two different primer systems. To summarise the results and to permit statistical analysis, the systems are ranked in each test ac cording to the number of cycles needed to remove the last stain. The Dow test as run by two operators is treated as two different tests because of the wide variation in number of strokes between the two oper ators. The degree of agreement between the tests can then be calculated according to a method of Kendall and Smith. In Table IA the eight paint systems are ranked according to the duPont and Dow methods, including only the stains that are specified for each method. The agreement between operators is good for the Dow method when ranking alone is considered, although the operators do not agree in the number of strokes. The agreement between the duPont and Dow methods Is not'satisfactory statistically, although inspection of the original data shows that there is a considerable degree of agreement in fact. The lack of agreement between the two methods occurs in the removal of the most difficult stains - red or black China Marking pencil-. In the duPont test China larking pencil always remains through the first two steps. The third step where the film is scrubbed with scouring powder DUP030003620 * -5' may remove the stain completely or may leave a tracer The variations at this point are largely accidental; therefore, differences between the films are minimised. In the Dow test, continued scrubbing with pressure gradually removes the China Marking stain and brings out small differences between the films * In Table IB the paint systems are ranked by omitting the China Marking pencil and including all the other six stains under both the duPont and the Dow scrubbing techniques. In this table, the agreement between the Dow and duPont techniques is highly significant in a statistical sense,, Experimental Results in Scrub Resistance; Data in the scrub tests are summarized in Table XX. All tests showed that the two systems with Super Kara-Tone Park Lane Green wear through very quickly. The other six systems are close together in scrub resistance, and the tests do not distinguish between them reliably. The data are not precise enough for analysis even by the method of ranking. Table XI illustrates the wide variation between operators in the Dow test. Although there is less variation in the'duPont scrub test, it is still subject to very large error, The confidence limits for the average of .1800 cycles shown for one system, for instance, at the 95# level of probability are 1000 cycles minimum and 3100 cycles maximum. REFERENCES; Notebook 5677 PP. 151, 16S through 174. Memorandum from Pigments Department? "Procedure for Wet Abrasion on Latex Emulsion Paints", received in Marshall Laboratory in September, 1952 Memorandum, "Stain Removal Tests - Latex Paints", R. A, Johnston, 9A9/52. "Evaluation of Latex Paints", Dow Chemical Co., page 18, revised 9/25/53,, M. G, Kendall and B, B, Smith, Annals of Mathematical Statistics, 10, pages 275-237 MARSHALL LABORATORY WPSjltffl 11/18 Attach. W, F, SINGLETON^ DUP030003621 A: Not s ig n ific a n t? F 2 .50 ( f o r 95# p r o b a b ility , 12&6 degrees o f fre e flo w , P=4.00) B? H ig h ly s ig n ific a n t? F=7.80 ( f o r 99# p r o b a b ilit y , 12&6 degrees o f freedom , F -7 .7 2 } HANK IN STAIN REMOVAL BY DOW AND DUPONT TESTS t- VO CO oo in cvj cm TO 6 vo t- c m sf in co VO tf\ in ini CVI CVI 00 f- Q in in 01 01 TABLE I : w | irv oo Sl- VO OJ S'- rl (I) in t- *-* VD CO CVJ CO 00 Jt c m cvi t"- n cu vo PH(0 < O fhll r-l & I CO 10 03 r-* tu 0) H (0 p t0EJ) rl U t-4 00 0 p rl <43 U 0 f(-04 <p r-4 03 rH u <0 ip h a fcH mA T> O 6 P Ua i <p r-t 0) to c 0) <8 a> H OJ H CO r-l (0 0E) P p co &p P-< (X< o V*4 r(-0< u Q) f-4 a 03 iH i < S<2S |0H ip ca C UH f<-40 P-ft, O OJ U O * <l 1 <u ofc j - <0 oo 0 p 4 I 03 aos- 03 OEI K 0a s 03 po r 1 so H -fc E1 0) w OaJ 0 03 DUP030003622 TABLE 21: CYCLES TO FAILURE IN DON AND DUPONT SCRUB TESTS rm -p ft, 1--t E-J m >p 5 O P* coo o vo CM CA .tft o m # VO <vr CM CM -KS- -Sf cn cn 88 8 CO CM TM! & r-t 13:, 'D & $ *ri r-1 U Q> <D i SO C-i 1 r-t (0 (D CO V* 4*i b Aj H S3 * Sc 4 O M ! $ : *l 1-5 &C6 4<0* r~i fe 'O & B r--1 h Q. j 50 $-s 1 i-, f-i Si <M p .a. <M sz & J3 0 Ps Bt B , T 14 U Sj & w 8 St .p CO fe `B *s 1 SS3i r-f rt % a to i1 <M u i--l m r! Si G t-U us p o* f <5 o )~J 'sr fe w H co 3s .p so 'Sc, TO g> s. B &> h j--i fc m 0. .Q> ! CO SM u ri G C 0 t: 4 Ss s fei 0. ;* a so o3OwU,'O raOM p ffl S3 <0 > Pg>foHcj OH ir* J co Pa s-* isfi ajs* s** p pTS*a a^c oi ou vi p P to <0 fep to <d oa pj U JC 8>P P *D P COM; <BH J? m H p S3 -rl B p <0 r-i yrJ CO ft! H I S3 o*$ pa 50 a .p53 *Co-li---I! 'HC Pto 03 A&O Pca a m rS 03 p fj Eu S3 3 8> '0 P P S'S ng Up B <*rMHM O M P .<U TOJJ . o J8 p 4* 3c hf`P CM CCS O SUS-i mo HC m so> 03 $4 03 W > r~* OS ;m 0 Ps " o O r-J P i~l o ro a, O @s j~i CM 44> o to &SP r-4 o ra <H C8 (l) u uo p 0 p VI Ua Vi o M o O& Pp 4 0 AC o jc M P ra to 0 ** M Oi ft9 M E~* O . tr* o: :hs a. DUP030003623