Document j71G43dnOwoQD8OKpje9Gqjk

Twenty Fire Tears of Paint Testing" U. Van Loo The Shendn-Willisms Company, Chicago, Illinois ' The span of years marked by the twenty fifth anniversary of the Division / of Paint, Varnish, and Plastics Chemistry of the American Chemical Society has witnessed a tremendous advance in the wuality and general utility of organic finishes. This advance has been due in no small degree to a con siderable elaboration and refinement of testing technique. In many ways it is difficult to differentiate between cause and effect in the parallel improvement of both quality and test technique. The primary functions of paints, varnishes and lacouera are protection and/or decora tion. As an increasing number of usages for such coatings were developed, it is a natural corollary that methods of test had to be engineered to ascertain the capacity of the finish to meet the requirements imposed. At the same time, the wider scope of application and performance character istics which could be determined led to a greater appreciation of the versatility of the material. This in turn opened a broader vista of po tential 'usages. Thus the contribution of the test chemist and the tast ing engineer, both in control of quality and in expansion of usefulness of protective and decorating coatinga must be recognised. Progress in Testinp Technique The term "tasting" has three connotationsi a review of things that have already occurred; an inspection of the current situation; and on the basis of knowledge of all available faots, a prediction of the future. Testing as it applies to paint technology has several aapectai (1) In formulation, the determination of the characteristics of a finish which fix its particular niche In the protective and decorative field. (2) In production, the determination of the level and uniformity of qualify compared with a standard set up far that material. (3) In development, the analysis of service and durability records upon which may ba prsdicatsd Improvement of the product, (4) In research, the combination of findings from all such tests, giving an insight into ths complex structure of the paint film, whloh is esasntlal to all progress. Testing in ths paint industry certainly wes not s new development within the pest twenty five years. It is true that sines 1922 thsre has been a large increase in the number of chemists, chemical engineers and technic clans employed in the paint industry, and the importance of the test chemist and test engineer has increased proportionately. However, much still remains to be done, as was demonstrated in the 1946 report of the Dallas Faint and Varnish Production Club, covering a survey of methods used by paint manufacturers in testing paints and varnishes. (1) 0007-SWP-043928 / -2- (1) Official Digest, Federation Paint and Varnish Production Clubs 263, 653, 1946 Cut of approximately 290 laboratories reporting, sens IJf still did not test their production for such fundamental properties as fineness of grind, weight per gallon, consistency, brushing end flow oharaoteriotica, opacity, color, gloss, hardness, etc., or did so only when such teete were specified by the consigner. This meanB, however, that of the paint manufacturers run such testa as a matter of routine. Volum^wlaa, the proportion of paint which receives regular and careful inspection and testing ia un doubtedly much higher. Gardner's "Physical and Chemical Examinetion - Paints, Varnishes, Lacquers, Colors," now in its tenth edition, has for the past twenty five years been the outstanding compilation of teat methods employed by the paint industry. The first edition was published in 1922, ths second in 1925, with succes sive editions appearing at frequent intervals up to the tenth of 1946. Therefore, this outstanding work covers precisely ths period under consid eration. A review of the various editions as they appeared affords in itself a picture of the progress in technology and testing technique. This division of ths American Chemical Society has made numerous contribu tions to the present knowledge of paint testing. The American Society for Testing Materials, especially Caanittea D-l, has established methods of tests which are accepted standards for material suppliers, producers and consumers of coatings. Ths member clubs of the Federation of Paint and Varnish Production Clubs also have contributed much to the current know ledge of testing in the paint Industry. Federal and other governmental or military purchasing agencies and industrial consumers through compo sition or performance specifications present a wide range of pertinent tests. The literature constantly adds new methods of test or refinement of currant technique* The testing of industriel finishes was discussed in detail in a Symposium on Paint and Paint Materials, sponsored by the American Society for Test ing Materials in Philadelphia in February, 1947. Also, the Buxton Con ference of the Oil and Colour Chemists' Association held in England in May, 1947 had ae its general theme the assessment of paint performance, While same of the discussion applied to trade sales types of paint, coi>siderable emphasis was placed on the testing of industrial finishes. A complete review of all test procedures applying to the composition, development or formulation, and the performance of protective and decora tive coatings is neither necessary nor feasible within the scope of this paper. However, since the field of exterior house paints is one of the most widely explored in all paint technology, it will serve as an example of ths contribution of the test chemist and testing engineer. Progress in Paint Formulation - Ths improvement in performance and service qualities of exterior house paints over the past 25 years has been exten sive, See figure 1. The four photographs illustrate the comparative ap pearance and durability of two-coat systems of representative white ready mixed house paints of 1927, 1935> 1940 and 1943, all after four years' simultaneous exposure, Ths paints were all on one continuous board of white pine siding, exposed south vertical in suburban Chicago, Illinois. 0007-SWP-043929 0007-SWP-000127279  3- / / / / / Th 1927 paint la repreaentative of tba 60% lead, 30% also oxide, 10JK extenders paint of that ara. &uoh paints tandad to oollaot dirt which was than ahad unevenly, and tba film failad by checking and (tracking. By 1930 titanium pigments war# being used In oonjunation with lead and a sine pigments and extenders, with the tltanlvai naually introduced In tha barlua aulphata axtandad fana. By 1935* titaniw cod.de plua aalaotad aztandera waa replacing tha titanlua barlua plgnent to a con siderable extant. Appearance qualities bad improved, due to tha bright ness, opacity and obalking tandenoiee of tha titanium oxide, azri dura bility had been augmented. By 1940, specialised house paint undercoatare ware widely uaad, and tha pigmentation of tha top aoat had moved in tha direction of higher titanium and axtandar content with acmevhat reduoad lead pignenta. Up to 1943, pigment-volume-ooneentration usually waa bald at about 2& to 30%m War time ahertagaa necessitated conservation of drying oils and led to tha uaa of inereaalng propcrtlooa of bodied oils, with less total hinder, tba pigment-volume-ccncentratlon lnor eased to approxl-- mately 36%, The net result was a better house paint in such important qualitiaa aa flow, drying, and reduction in dirt collection, oheoking and cracking tendencies, Tha steady progress in protective and decorative properties of house paints Illustrated in Figure 1 waa made possible to a considerable de gree by the progress in methods of tasting, Tha backbona of this test ing waa aotual outdoor exposures - panels, housa testa and other fanu of field testing. There is probably no other type of paint coating in whiob durability over e span ef years plays a more important role. Since house paints have a Ufa expeotanoy of several years, testa whioh expose tha paint to the weather in much the same fashion as in regular service are necessarily laborious and time consuming. The improvement that has been made in bouae paints in tba past 2$ years has axtandad tba time requlrsd for exterior exposures to show significant failures. Hence paint durability teatlng continues to oooauma more and mors time and progress beoemea slower end slower, Hhan the test chemist facee such a situation he tries to devise special laboratory teats which will give him in a shirt time the same results which he would get on longer normal exposure. So far there baa been no suoh happy eolation to the problem of teatlng exterior bouae paints. Exterior bouse paints are exposed to air, sunlight, wind, various degrees of humidity, dew, fog, rain, aleet, lam, snow, various and varying tem peratures, dirt, soot, and in acme oesea industrial geese, mildew, salty almoephcra, ate. The aevwlty ef these conditions changes Aran day to day and year te year. All ef these states of weather influence both the initial drying and the aervioe performance of the paint. Other faotara exclusive of weather or composition include the wood or type of repaint surface on which the point la eppllad, film thickness, direction end angle of exposure, to mention only e few. 0007-SWP-04393 0 0007-SWP-000127280 > 'i Brora* boa pcIntad out that th* propartlaa of wood that affaot paint hold / ing characteristic* lnolud* density,(2) texture, content of raaina, oil* and / (2) Brown*, F, L., U, S, Dept, Agr., Misc. Pub, *0, 629, 1947 nolotors, and defects ouoh ae knots. Siding varies in annual ring atruotura, distribution of spring and auaaer wood, direction of grain, whathar the *urfaoa In bark alda or pith aid*, whathar high- or low-grada, *ta, Thua it becaaaa apparent that area In tha aalaotlon of th* aubstrat* th* paint test chaaiat faoaa a difficult problem to achieve an/ dagraa of uniformity, which la asBantial if valid comparisons aay ba wada batwaan Bombers of a barias or between series. Whan in addition ba attempts to aat up apeoial conditions to cover all tha poaaibilitiaa of variationa in application and atmoapherio exposure oonditiona, tba task becomes an extremely difficult ona. Labcraterr Teat* to Accelerate Weathering For maiy /aars paint tasters have used various labaratar/ dsvloss calculated to aocalarata tba weathering of paint*, Thar* bava bean two scboola of thought In tba development of such apparatus. On* school baa assembled tha various weathering factors in a tank, intensifying soma without being able to intensify all of tha faotora. Tba resultant accelerated weather ing waa of neeeaaity somewhat unbalanced, thus detracting true tba valid ity of tha test*. Tha aaoond school of thought attempted to eubjeot tba toat specimen to tba varioua weathering factara on* at a time and to ocmbias tha rssulta in soma suitable manner to Indioat* a measure of durabil ity or earvio# performance. While again tbia aaoond approach represents a departure from oonditiona of normal uaaga, It baa proven useful in tha testing af certain types of finish*a, especially in tba industrial and autoaotiva fields, Th* measure of success gained in evaluating such fin ishes should one our*go further Investigation of tba value of tha aaa* approaoh in house paints* Comparative data will have to ba accumulated to provide a background of axparlaooe, but obviously a start must b* mad* same time to gather tha naoasaary factual baa* material. Carbon Are True Brpoeure Apparatus - Devices of tbia general character represent developments of tha first school of thought and have bean in use for many years. (3) (4) (3) Balaam, H. A* end Schauta, F. C., Aa. Soo. Tasting Materials, Faroe,, 22, Part II, 458, 1922 (4) Belaon, H. A. and Schauta, F. C., Aa. Soo. Testing Materials, Proo., 24, Fart II, 920, 1924 0007-SWP-043931 0007-SWP-000127281 / --5 6 ^ tta types moat widely used, tha paint films arc radiated with ultra- i.l?- ^rUlmt light fra* aeroury Taper or carbon arcs, usually the latter, it T ^tarvala tha films ara subjected to thanal shock and aoiatura pene tration fro* a water spray, plus additional tbaraal shook and tha stresses / sat up by tha expansion on freezing of tha water retained in the fils if refrigeration is used* Expansion and contraction due to the alternate eoaking and drying out ooaur at various stages In the oycle* The inten sity, amount, and rage of radiation, ambient temperature and humidity, panel temperatures, thermal shook and degree of moisture penetration, reet periods and the relative time distribution of tost eonditlcns during the cycle are deterained by tha design and operation of the machine* Thus it appears that such devices introduce a large number of the test faotora present in natural meatharing and it might ba anticipated that they should permit prediction of performance in aotual service. Although they have bean In use for over 25 years, there baa been no unanimity in that conclusion, especially in the testing of long oil exterior paints* (5) (6) Thera is still no sgrssmsnt as (5) Am* Soc, Tasting Materials, Proc., I, 366, 1936 (6) Mat* Paint, Varnish Laoquer Assoc. Sci. Soot. Cire, 671, February, 1944* to tbs bast cycle to reproduce natural weathering, failures to oorrslate accelerated teat results with outdoor exposures, or tests run 00 identical materials by different oooperatara, or to classify various 0lassea of exterlv finishes la their aotual order of exterior durability or oolar re tention, sto. in some lnstanoes have been ascribed to failures to duplicate test conditions precisely* The Cleveland Paint and Varnish Production Olub has bean active for e nus*bar of years in an investigation of carbon are msoblne operation* (7) (3) (9) (7} Official Digest, Federation Paint and Varnish Production Clubs, 250, 454, 1945. (8) Official Digest, federation Paint and Varnish Production Clubs, 262, 516, 1946. (9) Am* Paint 32, *0. 7-B, 8, 1947 They found that a satisfaotcry correlation of auoh devices in ths Cleveland area and duplication of Miami, Florida test fence exposures did not extend to aro cabinets run by oooperstors nor# widely distributed geographically, who were not all confined to the Cleveland area, using the asms water supply* 0007-SWP-043932 0007-SWP-000127282 k4<-r 'U % - a T. v - jMy found aa apparent necessity far adjuahaent of water supply, duration .' of water exposure, maximum temperature and duration of nsarly maxima taeperature, to achieve unifora machine operation in various parts of the oountry, They Infer that after such adjustments are made further coopera tive work must be done before any atatementa can ba mada es to whether there la a correlation with Florida results, Thera are two prlnoiploe of eooeleretlon es far es laboratory weatnerlng tests ere ooncernod. First, it should bo posslbls to Intensify the weathering factors so that risible failure occurs more rapidly than under actual aerrioe conditions. This ia the scheme uaad moat widaly. Second, it should be posaible to diagnose the deterioration of the finish undertest at an earlier stage when risible failure is not apparent. The Clereland Club's efforts in the letter direction here been Interesting and hare net with sane success. In addition to the obserretien of tho usual loss of gloss, chalking, cheeking and oolor changes at orltaria of the effects of weathering, they checked additional properties such as distenslbility, adhesion, scrape hardness end rusting on the metal panels used in their tests. The iaerlcan Society for Testing Materials has Tentatire Specification D 622 46 f, "Tentatire Recommended Practice for Operating Light and Tatar atposura Apparatus (Carbon-Aro Type) for Testing Paint, Tarnish, Lacquer and Related Produota," (10) It specifies the types of carbons, measurement and control of - (10) Am, Soc, Tooting Materials, Standards, Part II, Honmetallle Materials Constructional, 1614, 1946, current, roltege, tasporatur# of panels, temperature end pressure of the water eprey, rotation of the teat speolments, and the operating cyole. The specifications for tho water other than temperature end pressure era quite general, Hollie (11) pointa out that it la dtslrabla that a light aeuroe be (11) Hollie, B, J. Oil Colour Chaaiete* Aaaoe. JO, 431 (1947). developed which (1) la constant in output, (2) gives a fair proportion of light In the wave bend from 2900 - 3300 but little in the extrmae ultra violet, and (3) doea not vary in intensity during use due to fouling of the glese glbbca or filters by degradation produota. Corex D glase globee how e relatively sharp cut-off of wave lengths below 2900 a, reasonably close to the law limits of solar radiation after passing through the earth's atmosphere, at about 2950 1. (12) (12) Coblenta, V, W. ami Stair, R,, B. of 3. J. of Res. 3, To. 5, 629,1929 0007-SWP--043933 0007-SWP-000127283 L--nMr >- / -7- 4,5.T.K. Specification D 922 -- 46 T (10) oallo for the replacmaent of the fleas filters (globes or windows) after 2000 hours of use, or when pronounced f diaoduration or milklneee develops. filters must be washed each day with / detergent and water. Thus It would appear that the radiation factors ar e susceptible to reasonable control. / k rather general orltioian of oarbon arc aaoblnas la that undue emphasis Is / placed on chalking tendeaciea at the expense of checking and cracking, especially in long oil paints. (6) (11) There is a growing realisation of the part aolature plays in causing dsatruotlan of the weathering film* (?) Wtrailing (13) found that plotting the nonthly naan taaperature and the percentage relative (13) Wirshing, R. J.t Tad, Eng. Chaa., 33, 23*, 1941* humidity at Kami, Florida mere nearly approached the failure curve of autoactive lacquers exposed at that point than did tha radiant energy curve* There ie a general trend te revise carbon aro machine oyoles to prolong the water spray period so that tha palnto booaaa thoroughly soaked with moisture* It is believed that tho sequence of swolling during the wet stage of tho cycle followed by oootraotion on drying out will accentuate rupture of tile file, and in conjunction with acme ultraviolet radiation will aInulate acre olosely the filxe failures which result fraa normal weathering. House Paints In Oarbon ire Unite. - 1 definite question in conducting oarbon are tests to aooelarate weathering of long oil air drying house paints is when to Initiate tho tost. (11) 711ns which roach maximum film strength quickly, such as those which dry by simple loos of solvent, or taken finishes, presumably may bo subjected to aooelerated teats rolativoly quickly. Howsvsr, slower air drying filas such as houso paints poss anotbar problsm. As soon as s paint is spread on a houso two motions are initiated. One la the drying of the paint in which oxidation and polymerisation and aol to gel transformation oocur which give tha film strength and railstaro# to tho elamenta. It io known that those processes continue few a long period eventually loading to unfavorable film qualities, snob as loss of elasticity* Also as soon as tha paint is spplisd, tho dostruotivo foreos of tbs slamonts start tholr attack. It is diffioult to fix tho point at wbioh tha balance changes from tha positlva to tho negative aide* D 922 - 46 T Specification (10) calls far a drying period of seven days for "imal drying paints** This would appear inadequate for house paints, but it is appreciated that if too much time elapses before the film may bs con sidered properly conditioned to bo placed under tost conditions, the main point -- that of accelerating the development of normal failure - ia lost to a degree, k ravia1tea of the early stages of tha exposure to reduce tha aeverlty of ultraviolet radiation and operation at lower temperature to avoid distortion of polymerisation processes is worthy of consideration. However, this is often lapraotioal in apparatus whioh is used simultaneous ly for various typeo of materials and groups which are started at different times. Cost is s consideration in addition of aufflelant Instrumentation* Also, tho proper proportionate.on of attacking factors over tho entire range of tha testing oyolo has not boon determined. 0007-SWP-043934 0007-SWP-000127284 ?V- g'0 tu thsre la no data available on tba results obtained on apparatus in / which all the maana of oontrol oa tha light source, water, and other fac- / tara aro employed, and tba options* aycla la utillsad* As raportad In tha Claveland Club investigation, (9) whan such instruments are available and proparly bandlad it should ba poaalbla to ocorralato results between dif ferent machines. Tha questions which atill would have to bo anawarad ara whathar ooata can ba hald within liaita and whathar thara would than ha a oorralatiou with natural waatharing raaulta* Outdoor Durability Taata Until fully dapandabla nathoda of acoalaratad waatharing taata ara devel opad, tha paint taohnologiata and teatara must roly on time consisting out door taata* Tha taohniqua of outdoor paint tasting has bean laprovad con siderably in tha paat twenty fiva yeara* Sarly taata wara often applied without too much regard far apraading rataa and film thickness, or type and uniformity of wood surface uaad, A rather common praotioa was to aitpose tha paints on flat boards at an angle of 45 faoing south, often at only ana geographical location, with Florida a rather popular selection* It has bean pointed out that exposures at 45 south at times do not simply aocalarata tbs normal waatharing of houaa paint exposed vertically and nay actually dlatort tha perfarmanoe in regular aarvloa* (14) (1$) (14) "Photographic Supplement to 'Leaves From a Paint Research Votebook, law Jaraay line Co., 1938* (15) Troutman, R B* and Tannoy, IT* O*, Ind. Big. Chi*,, 32, 232, 1940,, Today, moat house paints ara exposed on lengths of beveled clapboard aid ing, mounted in tha regular manner, which maana that tha paint surface is actually slightly off tha vartioal. Much greater difference* in dirt coL1notion and olaaning ara shown on tha slightly angular bevelad aiding titan an strictly vartioal boarda* (16) Aa a consequence, strictly vertical expo sures ara avoided (16) Vaonoy, V, c. Official Digest Federation Paint and Tarnish Production Clubs, 235, 177, 1944. except for special feypas auoh as trim paints, ahoss major usago would bo 00 such surfaces* forth vartioal exposures on beveled aiding are particu larly useful in studying dirt collection. Panel Bcposursa - There ara several good methods of panel exposures. A deaoription of all tba methods and an analysis of the effects ef tha variations la taohniqua ara beyond the limitations of tha paper. Therm-' fora, without prejudice to other nathoda tba discussion will bo liaitsd to thoos employed for tha evaluation of exterior performance of hwae paints in our own laboratory. Three-foot aaotlona of 6-inoh or 8-inch beveled siding are aalaotad for uniformity of wood characteristics within tha sarlea. Tba apeoiea used "` 0007-SWP-043935 0007-SWP-000127285 --9-- v* most are edge grain Western Red Ceaar for Group 1 wood, (2) White Pina for Group 2, especially In checking email differancaa is performance of high grade painta wbloh night not ba detaotad on wooda of auch good palntabll-- ity as cadar or radwood, and Pondaroaa Una for Group 3 wooda. Southern fellow Pina, aa a Oroup 4 wood, is now used only In apeolal atudias of painting systems for that wood, due to difficulties with warping whan bavaled siding of regular thickness vaa used. Tha painta ara applied by brushing at controlled apreadlng rates, preferably in an enclosed paint shop at the exposure fern during tho outdoor painting season, leather conditions, temperature and humidity at tfea time of paints ing are recorded, Tha wet panels ara then moved to tha adjacent drying room which ia roofed but open to tha weather on two cldea except for cheaaecloth to keep off insects. If the weather permits, the panels ara then mounted temporarily on southern exposure for the time remaining binfore recosting. The next ooat is usually applied two days after tha previous coat in good weather and preferably within four days under any cirouuatanoae. The top ooat la then allowed to dry until it hao hardened sufficiently to acoept stenciled identification. It is then ready to be pieced on the fence if it ia to be exposed locally, or a arias of panels may be boxed and shipped for exposure elsewhere. Such panels ae must be painted Indoora during winter montha ara given two hours of radiation in a special twin arc cabinet for each coat. This is done to simulate the beneficial drying effects of sunshine which a paint rsoelves outdoor*. Range of lasts - Since the wood characteristics of any one panel will not duplicate precisely those of any other panel, conclusions based on the re sults obtained on ajy one panel must always ba suspect. Elm (17) has pointed out that (17) Elm, A. C., Am. loo. Testing Materials, A5TU Bull,, 142, ?, 1?46. problems of this nature can be solved by e multiplication of testa on e suf ficiently extended seals and the application of etatiatloal methods to the exposure results. There has been a conaiderabla lnoreaas in tha application of statistical methods to paint testing in reoent years, and a further in crease ia to be anticipated. As mentioned earlier, neatber is not a fixed constant, - rather It varies from sseson to season and has wide fluctuations in temperatures, precipi tation, sunshine, eto., frae year to year. Tests run for relatively short periods of time will yield different results depending on the season of exposure, or they may not be representative of average behavior at that lo cation because of unusual weather oonditiona. Since house painta are usual ly kept on exposure far several years, deviations from the mean in weather conditions tend to cancel out. However, the effects of wide deviations will 0007-SVJP-043936 0007-SWP-000127286 / -10- wary according to tbe point et whioh they ocour in tbo tost period* There for*, it i* neoeaeary to run repeated exposures before tho normal perforaone* on bo dotorainod reliably. Recognising tbo possibility of erratio behavior of a point oo any single panel, whether duo to wood, opplloation or manner of exposure, a gnido esotion is appllod on every panel. A paint whoa* ncarnal behavior is thorough ly understood by- an adequate background of expoeur* oxperionoo Is applied cc 1/3 of eaeh panel* Tbs tost paint is applied on the regaining 2/3 of tbs panel and its perforaanoe is rated with regard to the guide on tbs sane panel* The broad experience with tbe guide also permits allowance to be aado in oaee unusual weather conditions are encountered* A multiplicity of panels is always desirable and duplioate or triplicate panels are used for oritloal tests* It is recognised that this does not giro s broad base for statistical evaluation* Horever, a working faro aust cover euoh a aase of problsos that there is a physical liait to tha nuobar of panels that can be prepared, mounted, and examined* For most testa tbe problem is bandied by the use of tbo guides and multiple expo sures* Usually this involves exposures south and north vertical In subur ban Chiaago, Illinois, south vertical at Coffeyville, Kansas, end south vertioal at Uiami, Florida* Tbe use of both edge grain eedtr and Ponderosa Fine or other combinations further multiplies the unite of testa, and in total cover a wide variety of panels, geographical distribution and direc tion of exposure* Operating in this manner, experlanoe has shown that tbe average perfcraanoe of a paint may be evaluated with sufficient reliability to allow national sale in largo volume wltfc safety* Tbe panele ere mounted on a vertioal fence as shown in Figure 2* They ere held against conventional studding by e fitted olaap, with an enclosed air spaas of 4" depth between the panels and tbe ehlplap backing. A double fenoe permits exposures both to tbe north end south* Readings are made and reearded at regular intervals and weather conditions are also recorded aa an essential part of the exposure history* Variations in fttpoaure Reaulte - Ihen Judgment ae dictates, the panels era called In from the eeveral exposure locations end jointly examined to get tbe ovwr-all picture of porformanoo* It should be pointed out that Florida exposures do not simply accelerate tbe incidence and development of film failures which appear in tbs acre northerly exposures, nor do tha Kansas exposures alaply lie Intermediate* la general, tbe Florida panele tend to accentuate checking and eraoking tendencies, with acme surface flaking end eroaicn a minor failure* Suburban Chicago exposures show relatively little checking and cracking in better grade bouse paints, but erosion tendencies are aggravated* Figure 3 shows oompenion exposures of five years' duration of tbe same paint, at tbe three locations, all south vertioal, Magnified areas Illustrate the type of failure at aaeh location* 4532 C, the Chisago panel, sheer* the re sults of active chalking loading into erosion, especially visible at the brush marks* There is no characterlatie ahaeking and vary little cracking. 0007-SWF-043 937 0007-SWP-000127287 L C,, * J *! / -- tbs very thin spots. 4532 COP, tha CoffeyvUla, Kansas panel,, 4ra very llttla ohalking and no aroaloo, but wall developed cracking lenc the brueh marks. The elllptloal eraeka are marks from hail^etonea, 4532 ft the Miami, llarlda panel, ahowe seas aurfaoa pitting from ohalfclug, sal orsoklng almost equal to the lamas exposure, Tbe faot that anob difference* exlat baa been proven by many exposures, using different paint aye teas, different apeoiea of wood, with panala painted and exposed at different tiaes of the year* Why auoh differences occur ia by no naans as clear. They would appear to Involve the total, anount of ultraviolet absorbed by the palate, the temperatures of the films, tbe anount of aolsture absorbed and bow long each moisture le re tained, the expansion end contraction due to temperature changes and alter nate soaking and drying, and similar factors. However, the precise reasons why tbs weather conditions at one point aoeeatuate chooking and at another, erosion, remain obaoure. Ibleago had approximately the same total hours of sunshine as Miami In both 1946 and 1947, for example. The angles of inoldenoe of ths sunra rays, of course, wars different. The average temperature in 1946 wea 51*8 P* in Chicago, 60 in Coffeyvills, Kansas, and 75*4 la Miami} in 1947, 49,3 in Chisago, 57 in CoffeyvUla, 76,1 in Miami, The extremes ia temperature in 1946 were -3 T and 166 In Chicago. 7 end 107 In CoffeyvUla, and 45 and 89s in Miami} in 1947, -2 and 101 in Ohieago, -7 and 105 in Coiffeyville, and 35 and 89 In the first eleven months In Miami, Total precipi tation in 1946 waa 31,88 inohea in Chicago, 33*15 inches in CoffeyviUe, and 75,46 inohea at the Miami exposure field} in 1947, 39,21 inohea in Chicago, 31*20 inohea in Coffeyvills, and 79,30 inohea at tbe Miami exposure field for ths first eleven aimths, Tbe oomparetlvely heavy precipitation at the Miami expoeura field and higher average temperatures would mean relatively higher average aolsturs 0ontent in the air than In CoffeyvUla cr Chicago, However, infcreation an hoe long the panels at the various locations remain wet after periods of precipitation is not available In this inetanoe. It Is apparent that the relation between weather and paint perfcraanoe needs nor# investigation. Tbe obvleua question alee arises whether aooalerated weathering devioee should be desired to simulate results of normal exposure In Florida *r at other points having weather representative of areas with larger paint consumption. Sines repaint mirk la proportionately auoh larger than application on new wood, a relatively large number of panels are on exposure constantly, utilising standard paint syeteaa, This afford* a backlog of surfaces for repaint studios, which otherwise are handled in the usual manner* House Teats - lhan speeific paints have shewn sufficient merit to deserve further evaluation they progress from tbe panel stage to tbe final step of field testing - practical house tests, in locations a elected over the entire country* Houses are ohoaen on the heels of representative construction and condition, with a knowledge of previous pointings if repaint work. Relative ly open exposure is {referred unless studies auoh ee milder, blistering or peeling, performance over knots and similar epealal oonditiooa are tbe ob jective, The house surface is usually divided into sections using control or comparison paints, Ths painters' oemments, weather conditions, spread ing rataa, drying, glees and finally service performance ere recorded. 0007-SWP-043938 0007-SWP-000127288 * Vi ?, --12-- Conclusion It 1* obvious that whan house paint* have eurvived snob rigorous tasting, ooesuasr protection is much graatar, The last chaalst and tasting engineer liar* not only contributed to paint technology bat to tha dsrelepmnt of battar paints which ere aada available to tha gsnaral public, However, it should also ba pointad oat that tha procedure is too laborious and tii-- ccnsmlng, and lqrevomant* whioh should hava basn aada In paints, to tha bsnafit of tha public, have baan unduly dalayad. Tharafar a, further study of aathoda of acoalaratad taatlng la not only desirable but mandatory. It is entirely poasibla that tha praaant instruaantation and taohniquo sill serve aa tha foundation upon whioh completely aatlafactory methods will ba built to evaluate all properties of tho paint, including untinats durability, Thara la still such to bo laarnsd about tha intarnal etruotura of tha film and tha obangaa whioh occur during Its Ufa cycle. Long (18) hao shown interesting (18) Long, J. S, Official Digest, federation Paint and Tarnish Production Clubs, 255, 150, 1946, possibilities In tbs study of absorption of light by olaar films, particu larly in tha untrariolefc range, as a measure of decomposition raaotions and tharafora a wassura of durability. It was planned to examine both ntrwally dried fHas and fils* whioh had boon woollen with water after they ware thoroughly dried. In this way tha Inflaanoe of satar as a factor in influencing ultraviolet absorption, and consequently reaction and decomposition, would ba measured, farther stadias of pigments, extendare, particle alia distribution, surfaoa areas, bindsr requirmeats, oritioal pigment volume concentrations, watting, dispersion, composition and tbs obsnistry of binders, distent 1bility, moisture permeability and many other footers ara lndioatod, Tho expansion in research and tooting will oentinus, and tho rooponsibilitiso and tho contribution of tho tost chemist and tooting engineer will increase aoedlngly What Is true of boose paints 1s squally true of all types of protective and dsocrative coatings. The Improvement in quality of finishes is always paralleled by a companion isprovmasnt in methods of evaluation and test ings There is no question but that tha progress aada during tbs past twenty five years will be dqsrfed by the progress whioh will be made in the next quarter century. jgbicwlsdgemant Tbs author is Indebted to A, C, Am, H. k Oardner and C C* Smrd for their helpful advloe and assistanoo In tha preparation of this paper. 0007-SWP-043939 0007-SWP-000127289