Document 5LKj2yYqbqmpj2G6YrbKdXYde

INDEXED" Uh/.i Mr. A. 1. Burkt-Wlliu Br A. Jo Hefisear3vHilla<> Ms?. Wo K0 Xaffett-Fliat Mr. E,, T,, Haekg-Parlin MPo Jo Ho towsli-Paxlln Hr. L. I. Ba&@r=F&2*Xin Mr Ho A. BWall-ParIia. L&t>= Fll?arllB PABLXH MB0RATOBY' MEMORAIIUJM BEPORY HO. 50 IT1110 MARSHALL UDQBATOBY LfflflJlBY. .fBtKHQE COM net t o r oRciaATioii SUBJECTS A STUB! OF THE CEACKXHG- CMBAOTmESTXCS. OF 233=35097 PRIMm-SUBFAOlS BATE XOSOTEBs MY 8$, l$k$ k o t h b o o i r e f s h s h o s i mm h o 22*2, p a g e s i t h e o o t 25 PREPARED BYs ARPHOTHP Bis N 29339 PARLIN LABORATORY MMORANPUM REPORT HO. 50 SUBJECT ? A STUDY OF THE CRACKING CHARACTERISTICS OF 233- 35097 PRIMEEUSURFACER INTRODUCTION During the latter part of 19^7 numerous complaints were received at Parlln on Philadelphia (233-35097) primersurfaeer production. These complaints registered the tendency of a thin film of this material to crack or separate from the surface to which it was applied. Actual evaluation of these complaints or the assign ment of responsibility in their investigation was* at th time. Impossible Thr@ was no test by which tMrs oharaoteristic could b accurately determined and, consequently, there was no means by which the identification of the contributing fac tors could b established either for the assignment of re sponsibility or for corrective recommendation in manufacture. The purpose of the work covered in this report was, consequently, of a two-phase nature? Cl) to develop a test which would accurately measure this crashing condition and (2) with such a test available to evaluate and determine which factors in formulation and/or processing were responsible for the cracking characteristics of this material. In addition, it was felt that the results of this work would make available valuable information which could be advantageously used in correcting existing formulation deficiencies and at the same time be applied profitably in future development work on primersurfaeer materials. A comparison of the active Parlln formula, on which there had been no cracking complaints, with the Philadelphia formula revealed several significant differences in formula tion. There was a marked variation in solvent composition along with differences in the amount and type of carbon black used. The difference in solvent composition was the result of Philadelphia having aligned against a Flint formula instead of a Parlln formula. It should further be explained that the difference in type of carbon black used was due to a shortage of the preferred pigment (-2^0 Peerless Black). As a result Philad@lphia8s manufacture was based-on Lamp Black (-205). The higher carbon black content employed by Philadelphia was a result of their attempt to maintain satisfactory color. It is worthy of not that -205 had also bees Used in Parlln pro duction at various times during this same period without apparent difficulty. A comparison of processing conditions showed that Philadelphia was using larger diameter mills and higher steel ball ratios in the grind portion of th formulas which, when compared to Parlln processing, suggested more severe grind ing of the Philadelphia-produced material. p DUP030000880 On or possibly a combination of those variables was suspected of being responsible for or contributing to the cracks ing characteristics under study. On this basis* a program ms established to evaluate both the effect of each variable sepa rately and the cumulative effect of pyramiding these variables. SUMMARY AND OOHChgSIOHS An accurate test to measure the cracking character istics of primer-surfacers was developed* fh test consists of pouring a sample of th primer-surfaesr* which has been adjusted by thinning to a given viscosity* on a specially pre pared panel and visually noting the cracking characteristics of the film of primer-surfacer on drying. In addition to show* ing the cracking characteristics of the film* the extent to which the cracking progresses measures epantltatively* to a reasonable degree* the deficiency f the material under test, fhe test has shown satisfactory reproducibility in extensive use and has been Issued as f.M. #103 for Division distribution use. Results of the evaluation work in a comparison of the Philadelphia and Parlln formulations have led to the following conclusions s (1) Under all conditions the Philadelphia vehicle is con siderably inferior, to the Parlin vehicle. In fact, there was not a single experiment involving Phila delphia vehicle that did not produce some degree of cracking. (2) Lamp Black (-2055 is apparently inferior in cracking characteristics to Peerless Black Cw~240) in that it is considerably more susceptible to the influence of over-pigmentation and extended or harsh grinding con ditions. (3) % effects of long or harsh grinding conditions can not be over emphasized in that any formulation weak ness* either as a result of vehicle composition, type or amount of pigmentation become more marked under these conditions of grind. The one exception* worthy of note* however* is that ^2^ pigment is not nearly so sensitive to the effect of over-grinding as the W-205 pigment. SKPERIMrom X. Development of 3?est to Measure Primer-Burfacer --,--,--,-------- ~~--------- Until the time of this work the only, known method of determining the cracking characteristics of primer-surfaeer materials consisted of a test wherein a bar spot was sanded DUP030000881 on a HDucoH panel which had previously been exposed (outside) for six months, feathering the edges of the sanded spot and spot-spraying with a sample of the primer-surfaer material under teat*, In most oases inferior material would erack at the feather-edge upon drying, Th test, however, was not de pendable in that van with the best known technique results were not always reproducible. Factors suoh as wet or dry spray, the proximity of the spray gun to the panel while spraying, the thickness of the spray coat, and rate of dry all influenced the results of this test, These Inconsistencies in conjunction with the fact that a six month old *,E>ueo'! panel was a prime requisite made the test undesirable and impractical for control or evaluation use. #ith these factors in mind the course of attack in developing an improved test for this characteristic consisted of determining what film of material on a panel would act in th same manner as th six month old "hueo* panel in promoting cracking and which could be prepared with little or no aging period. Secondly, it was necessary to devise a means of placing a film of the material under test at a reproducible thickness and consistency on such a panel and eliminate sal the variable conditions evident in the -spot spray method, it the characteristics of an aged Duoo* film which cause it to promote cracking were difficult to define? hone the development of a suitable replacement which could be pre pared without long aging posed a considerable problem* Extern$!ve trials and testing finally segregated 233-9607 White Pyre-Prep material as the most satisfactory film in this re spect. In addition, the overall time required to prepare and age a panel of this material for use in this test was only one-half hour. The method of applying a thin film of the primersurfaeer under test to th panel under reproducible conditions of film thickness, '`wetness*, etc. resolved into simply pour ing the material at a controlled viscosity. The results by this technique were reproducible and to a reasonable degree measure the aggregate deficiencies of th sample under test by the extent to which cracking progresses as the film dries. The actual test method developed and issued as a result of this work is contained as Table VI in th latter part of this report, XX. Evaluation of Prtmer-Surfacer Formulations , ______ ....................................... . - Exhibit Mo. 1 illustrates th general inferiority of the eomplainsd-of primer-surfaeer materials. The exhibited panels represent samples of these materials taken at random DUP030000882 from the Philadelphia production and compared to Parlin stands ard production In respect to cracking characteristics, _ By serial interpretation* the first four panels of Philadelphia material in this, exhibit represent November 1946 production, while serial P-49446 was produced in January 1947* If this one panel is representative of the quiP.it/ of produc tion at this later date, some apparent improvement is indi cated over the previously mad material which, if true, would probably indicate some change in processing in the Interven ing period* W are, however, at this time unable it either confirm such a statement as to the relative quality as repre sented by this one sample or to assign a definite meaning to any such comparison. As a basis on which to start the evaluation work, the Parlin formula in use at the time and on which there had been no cracking complaints was compared to the Philadelphia formula used in the manufacture of the complained-f materials. (Table II and Table III.) It was noted that in addition to variations in vehicle composition and type and amounts of pigment used, there were also variations in grinding conditions (Table V), Although no efforts were mad to simulate the Philadelphia conditions of higher steel ball ratios la the grind, both long and short grinding cycle runs were made in the valua tion of all the variable conditions under study. In an effort to study the effects of these known variables in the reapactive Parlin and Philadelphia formula tions, the following laboratory program was outlined; The variables under evaluation were; (1) Parlin versus Philadelphia vehicle. (2) Peerless Black (W-24o) versus Lamp Black (W-205i* <3> Over-pigmentation with Carbon Black (Parlin 3 5$ versus Philadelphia 5G*> (see Table ! ). (4) Extended grinding cycles. These variables were evaluated by modifying the formulas and grinding conditions as noted belowi 1. Parlin vehi^_e_^narlffljsatal IfflanflaSfcffii. -.Bfr%gES8gf *~24o pi^ (b) Extra amount W240 pigment. (c) Regular amount W-205 pigment. (d) Extra amount W-205 pigment,. DUP030000883 (a) Extra amount 5^240 pigment, (to) Regular amount W-205 pigment, (c) Extra amount W-205 pigment, B.Formulatlons ItjQgMfik&Sfa .Formula This formula (Table 12) is apparently a derivative of tii Flint formula dated August 1946 and secured toy Phila delphia from Flint on or about September 1946 for use in the productlea of 233*"3597 FrimrSrfae@r. This formula varies in several respects from the Pariin foimO.a in that it employs a solvent composition com posed of high amounts of HB-b, B-35, and H-47 and is,, there fore, somewhat richer than the solvent composition of the Parlin formula In which the main ingredients are HA-65, H-51, and B-29o In addition, the Philadelphia formula as derived from a representative hatch slip contains a relatively higher amount and a different type of cartoon black than the Parlin formula. In the course of this work the Philadelphia formula has been shown to toe weak in the following characteristics in respect to cracking, 'a) Weak in vehicle composition (Exhibit 2 through 5^ to) 0@es an inferior pigment W-205 (Eshibtt 4), c) Over-pigmented. (Exhibits 2-3 and 5), (d) Apparently overground (Exhibit 3 and 5), The Parlin formula (Table III) used in this study was the one in active use at Parlin at the time this study was begun. No fundamental weaknesses were observed in this formula. However, it was shown in this work that this formu lation would not permit over-pigmentation with the inferior W-205 pigment without racking on extended grinding (Exhibit 3), 3. Vehicle Evaluation The superiority of the Parlin vehicle over the Philadelphia vehicle under all conditions of variation in formulation is confirmed throughout (Exhibits 2-3~4 and 5) This difference, however, is most prominent in Exhibits 2 and 3 where series of Parlin versus Philadelphia vehicles show the aggregate effects of variations in type and amount of pig ment la each vehicle on both short and long grinding cycles. DUP030000884 It is interesting to note that in only one instance has the Parlin vehicle failed to protea satisfactory quality material,, file conditions fresponsible for this failure were a combination of overgrlnding,, ovsr-pigaent&tlon, and the us if the inferior pigment (W-20)= She Philadelphia vehicle on the other hand failed to produce entirely satisfactory material under any conditions of type or amount of pigmentation or grinding- (It should be noted that W-g4o in the regular amount was not tested in this vehicle- It is possible that this com-* bination would have been satisfactory in oracking-) A summary of the individual and aggregate effects of the variables in volved has been tabulated as fable I for reference in this comparison.. *b.. In Exhibit 2 it is apparent that there Is a appreei^ able difference between $-205 and Wg40 pigments in short grinds- However, Exhibit 3 clearly illustrates th@ inferiority of W-205 to $-240 pigment whm the grind is extended., 5 fh Effect of Varying the Amount of $-205 .-- -- In making a comparison of the effects of grinding on regular and extra amounts of W-205 pigment, Exhibit 4 clear ly illustrates that in oases of the higher pigmentation the effects resulting fro long or harsh grinding conditions are decidedly more marked than in those oases where lesser or regu lar amounts of pigment are used, 6,, The Effect of Grinding Cycle on .......<kt___, A comparison of the over-pigmented formulations in volving the two types of carbon black under varied grind con ditions, (Exhibit 5) again points out the inferiority of W-205 to $-240 pigment- This Inferiority Is not apparent in the short grinds but 1 Very pronmaused In the long grinds- It is thereby evident that $-240 Is not nearly so sensitive to overgrinding as $-205- 1* . Piryit^t1. Exhibit $ Is self-explanatory in that the Individual variables known to produce cracking have been pyramided to show the simulative effect of the differences between the Parlin and Philadelphia formulations- It is Implied in this progression that Philadelphia material has received longer and harsher grinding than the Parlin material- Reference to fable P, wherein a description of grinding conditions, as ob tained fro the respective Plant mentioned is noted, will confirm this contention- His of mills, steel ball ratio, ete- are both greater at Philadelphia then at Parlin and it is reasonable, therefore, to assume that grinding conditions at Philadelphia are more harsh than at Parlin- DUP030000885 7In this connection it has "been noted through a r law of the Philadelphia weekly reports (re Sept,, 9-13, 1946 May 19~2?s 1947, and Dee. &~l4, 1947) that immediately prior to manufacturing 233=35097 materials difficulty was being experienced in maintaining normal grinding teiroeratures in the hall mills and temperatures as high as 1203*\ had been encountered. References to these difficulties are again made in May 1947 and December 1947. Any such conditions would un doubtedly contribute further to over-grinding and b very detrimental to this product as illustrated in the various ex hibits. It will be noted in a review of this exhibit (5) that step by step the intensity of the cracking idthin the product progresses as the individual variables are introduced and combined until the ultimate change from the most to the least desirable conditions of formulation and processing is obtained. f dDRsBJfg 5/26/44 DUP030000886 TaJ>l He. I XX IX-A 111 III-A If f fl Mias IHPgX Summary of Variables Affecting Smok ing Character!sties in Prls@rSurf&eer Formulation*. Philadelphia 233-35097 formula* Philadelphia 233-35097 formula breakdown by ingredient. Parlia 233-35097 formula. Parlin 233-35097 formula breakdown by ingredient*. i El*renee notes on formulations 1 Comparison of Plant grinding conditions* Pesoription of cracking test procedure. DUP030000887 f&BLB 1 SIMMS' OF TARIABLFS AFFBGTXHG- GMGICIM GE&MCfsaXSTXGS IN PRIMlR^SCRFACmS CBAOKXH STATUS Bating Aslant ' Begalar Affiomt W-205 st*ft Affiomt R ; *15 Extra Amount W=2*iO 0 0 0 O 1 X* 2 .H^fira AaQttnt s Bgai&r Ajaomt W20'5 fetm Amount * *-|05 Extra Amount &~240 0 o 0 3 1 2 Bating 0 X 2 3 fc. 2ih &bM2II Ho Cracking Slight Cracking Pofinit Cracking Bad .Cracking ry Bad Grafting DUP030000888 TABLE XI PHILADELPHIA FOMULA This formula was obtained through a breakdown of Philadelphia 233=35097* F-51536, batch slip and includes in gredient adjustments resulting through 233-025 additions for olor controls Ester Gum (Gw60S) Butyl Acetate HBd) Methyl Ethyl Ketone Toluol m Methyl Isobutyl Ketone h -4&) Isobutyl Acetate H-62) Bibutyl Phthalat HE-111) Soya Bean Lacquer Oil h -,194) Butyl Alcohol H~12) Titanium Bioxide W-2) Elan Fix W-30) Talc H.D* W132> J5 = s, ;<f-g>)= Ester Gum Methyl Ethyl Ketone SoXual-- _ ______ (G-60S) JB-35) (See Breakdown Formula) (=363) (See Breakdown Formula)(B-501) (See Breakdown Formula) (B-w) Phosphor! Aid (E5J&) (G-697) ik 9$ S,i 3.0 M Col Mill Grind 0.3 3.0 iloo 4o0 -So :^fjT ~ ": Add to pull (Se Motel DUP030000889 TM&LE X1~A PHILADELPHIA FORMULA IHORlDISSIg BREAKDOWN Eater Gum Butyl Acetate Methyl Ethyl Ketone toluol Methyl Xsobutyl Ketone Isobutyl Asetate Stbutyl Phthalate Soya Beau Lacquer Oil Butyl Alcohol Methyl Isobutyl Canblnyl Acetate Isopropyl Alcohol .Anhy&roua Naphtha 95=l4oeC0 Methyl Isobutyl Carbluol Petroleum Xylol Bell Amyl A.tat# [EM Phosphoric Asia &5% [0-691 Titanium Dlca&a [*-26) Blanc Fix ;w~3o} Tale H, Oo B. To Lamp Blech \zm\ PX Hltroeotton Alcohol of Dehydration U 1,1% t*-71 II. p 2.4 0.3 3.0 tl u 1,9 0.6 1.3 0.4 p.3 4.0 Il6J..6S 5.0 WAw a7 /.*Ht (e> oA %A BA ai>'%& $A hS l.b %% OA 1nA ai 3.5 *7.1 3A DUP030000890 TABLE IXX FABLXH fmsmiA gates* Gum (G*6oB) Petroleum naphtha (ES or <BS9) 131*0) Bibutyl Pbthaiat (HjMLll) Blown Soya 3asi Oil Belle Amyl Asetate im(See Breakdown Formula) naphtha 95-140*0.. Titanium Bioxide Blanc Fixe Tale H,,0 Ii*4IS=Bia#s s = s = s } 2.3 3*0 2.0 9-2 1.5 6.5 5.2 Uo9 1?. 4 .1*5 Mill Grind Petroleum Maphtlsa (S$ or <h -*29) 131*0.) i%8_B^^kdo^FomlaI 1^131 (See Breakdown Formula) (->363' (See Breakdown Formula) (B~01 Phosphoric Acid {G-697 Ester Gum (GMSos! 3.0 Add to pull ~ %k, 75.5T 0.3 100.6^ (See. Mot# 1) DUP030000891 amjgfe* PABLIH F0BMUL4 iHSBEDiegT BREMBOWM Eater Chra ;g -6o 8) Petroleum Naphtha 88 or 1316. Dlfcutyl Phthal&t Blown Soya Beau 03.1 Bell Aayi Acetate [Sill] Naphtha 95l4oGt, Methyl fsohutyl GarMnol (8J Isopropyl A-ioohol Anhydrous 1&6S) Methyl Ethyl Ketone Methyl Ieobatyl Ketone Toluol Butyl Astate Methyl Isohutyl CarMnyl Acetate Petroleum Xylol Phosphoric Aeld Titanium Bioxlde Blanc Fix Tale HoOo Peerieea Blaeifc PX Nitrocotton Aleohol of Dehydration 7-7* 5.3 3.P 2.0 9*.22 1.8 2,1 0,9 1.2 0.9 1.6 0.3 5.2 1117..9^ DUP030000892 HOT&0 OH FOmmATIOm Nte.is For the sake of standardisation In th@ experlmental woA the amount of Carbon Black assign natd in th@ Paulin formulation was taken as the sBgalar Momt and that amount designated in the Philadelphia formtlatloa was taken as the wExtra Amount* rs Begniar Amount Carbon Black Extra Amount Carbon Biaek 3*5# 5.0# DUP030000893 TABLE V 003 HI ia . 3 *1 U*> m0 19 EHCsfdOv-hsr#>t^-' I A? >fi** 91 m o MILL S M B DATA 13 I d 2H * % t> ra o d 3 I tS & g S3 4 ffl r*\ g J* fl 6* 33 0 DUP030000894 TABLE VI GRA0SIM TEST FOE PRIMER^SUBFACEES iBmM. To determine the tendency of a thin film of prlmersurfacer to crack or separate from a surface to which it has been applied# Measurement of this cracking tendenoy by the method described herein consists of haring get op as standard, a ma terial which has proven to be satisfactory in cracking charac teristics nd comparing questionable or newly produced materials to it in this respect# Status This method has been recently developed in the .labora tory for use at Parlin to measure the cracking characteristics of *Dark Gray Primer-Surfacer#. Time Hequiremeat Man Hours...........,30 minutes Overall Test Time.... 2 hours Equipment Heeded 4x6* Steel Panels 233-9607 mt@ Pym-Prep g.^l. -2.QS.dl tlonin^, A 4 x 6# steel panel Is sanded with 280A Trl~Mtt paper wet with H-29 to remove grease and grit and is wiped clean. The panel is then washed with solvent and wiped dry. A portion of 233-9607 is cut with an equal volume of 36000 Thinner and throughly mixed. A single coat of the thinned material is sprayed to a thickness of 0.4 to 0.8 mil on a pre viously prepared panel and fore dried for 15-20 minutes at about 65. and allowed to eool to room temperature. Film thickness may be measured by us of the G !L Thickness Gauge or other suit able instrument. It is important that freshly prepared panels be used in this test. Each batch of 233-9607 baa coat should b pre-tested with primer-surfacer materials of known cracking characteristics as a precautionary measure before us to assure that the 2339607 is normal In all respects and that both negative and posi tive cracking characteristics may be duplicated with its use. Th formulation of 233-9607 dated February 13, 1948 on which this cracking test was developed together with manu facturing directions for the same Is noted below. This will DUP030000895 TABLE VI (continued} ~2 serve as a permanent record in case of drift in tost character istics as a result of future formula revisions* 233^9607 WHITE SEALER ^T|tXojltion HHHH~--4S23-275-lll B-563 B-501 wx-ooas - 697 %9% Wt./gal. 8 6.24 lbs. S 25*0. load with mixer Vise. 1 60-60* 10 cup @ 25Gt, reaming. Us buggy or barrel with Solids special head and 11.6# Ir valve for base,. Base to be added HC-12007(solid) HE-111 151-.7h in stream the H--22 '1.4 10.1 14..62 thickness of a phoeunrcsi.l. Mix 2 Pi & Anal. 0v35.0 n-,3 Add and mix 2 hours Developed for Refinish Sales as J-233-X-6545 (233-34120)s reissued to provide simpler cod. Revised to lower PX content. Issued by P&rlin 2/13/46 Superseding formula dated 12/5/47 JWOsE'ill Portions of the standard sample and the batch under going test are thinned to pouring consistency with 36006 Thin ner using the ratio of 4-5 parts sample to 55.parti of 36000 Thinner. When the samples have been thinned and thoroughly mixed pours of the standard and test sample are mad side by side on a freshly prepared test panel. Pours are allowed to drain and dry at room temperature 1*11 the panel remains in a vertical position. Comparison of the teat sample is then made visually to the standard. Any batch comparable to or better than standard in racking characteristics is to be considered satisfactory while any batch exhibiting a greater tendency to crack than the standard sample Is considered unsatisfactory. Parlin batch 233-35097* Serial 3*6356, has been sel ected as the standard sample for use in this test. In order to guarantee uniform comparison over a long period of time, 5-1 gallon cans of this material have been obtained and reserved for us as standard. I Using the 2|3"350#, Serial 366356, as a basis for comparison respectivefstandArds should bs developed on each type of primer-surfacetr material subjected to this test. Proper preparation of test panels in regards to film thickness and the use of freshly prepared panels is very important. DUP030000896 t a bl e ; rt (continued) -3" Best results are obtained when the samples under test are arouna 17-lg* viseosity <10 cup) which an normally b obtainit by diluting 0.1. viscosity material in the ratio to 36OOO Thinner described above* presence of grit on the test panel will cans the poured surf&oer samples to split by flowing around such an obstruction. This should not be confused with *r&ofclng* and if present may easily be identified as such. However* this condition my easily b controlled by eliminating grit from the thinned 232-9607 through filtration before spraying the test panel. DUP030000897 INDEX TO EXHIBITS Ho. 1 No. 2 No. 3 No, * No. 5 No. 6 Philadelphia, production versus Panin standard. Pigment variables with constant 10 hour grind ing cycle Ins Parlln Vehicle Series A Philadelphia Vehicle Series B Pigment Tar.tables with constant $6 hour grind ing cycle in Parlln Vehicle Series 0 Philadelphia Vehicle Series B She effect of varying the amount of ?-20;5 pig ment at various grinding cycles Parlln Vehicle Series M Philadelphia Vehicle ` Series P The effect of grinding cycle on ower~plgmented formulations Parlln Vehicle Series S Philadelphia Vehicle * Series H Pyramid of variable factors influencing cracking Notes Photostats of the above exhibits are shorn as negatives in order to avoid the loss in detail In making positive prints. The vertical breaks noticeable on some of the exhibited panels should not be confused with cracking. These latitudinal breaks are a result of film imperfections and should not be Interpreted as the longitudinal crack ing. CBRsBMZ 5/26/4$ DUP030000898 DUP030000899 DUP030000900 DUP030000901 DUP030000903 DUP030000904