Document 4amkOjJ9x3a9jZnLrOj5BBaXj

FOREST PRODUCTS LABORATORY. LIBRARY DETERIORATION OF PAINT COATINGS DURING PROLONGED EXPOSURE INDOORS By F. L. BROWNE, Chemist LIMITED DISTRIBUTION COPY NOT FOR PUBLICATION nauma mvomm mxxm By F. L. Chemist Forest Products Laboratory.^1 Forest tiervico U.S. B&partiasmt of Agriculture 2 A previous publication ()"* reported tbs changes in vaight, volume, density, and other properties of coatings of linseed oil house paints ^hll drying in the laboratory for 10 days after spreading and during a succeeding period of 15 days of exposure to each of the following conditions: (1) laboratory air and light )yoca glased windows facing north, (2) laboratory air in darkness, (3) darkness in a roots at 97 percent relative humidity, (4) twice daily timeroIon for 3 hours in running \iater, (5) artificial sunlight in a wsatheroontar without spraying with i water, and (6) artificial sunlight in a Ueatheromster and spraying with water three times each hour. After the initial loss of volatile thinner during the first flaw hours after application, all of the coatings gained in might, reached a maximum in weight within 2 to 8 days, and than lost weight steadily throughout the rest of the 25-day period. Tba density of the coatings increased and the vol ume shrank continuously, even during the first few days while the might was increasing. Changes occurred in the other properties such as water absorption, swelling, and solubility. During the final 15-day exposure to different conditions, artificial sunlight caused the moat rapid changes in coatings, leaching with water stood next, high humidity next, and laboratory air, either in darkness or ' Maintained at Madison, Mis., In cooperation with the University of Wisconsin. g "Underlined nuuibers in parentheses refer to Literature Cited at the end of this report. in north light through window*, caused the slowest changes. Ssrerth&Less, all coatings were still losing weight and changing in properties at the end of the S-day period. Xt was of interest to sea whether a such longer period of exposure to the mild conditions of laboratory air and indoor light would bring the coatings to a stable state with no further change in weight. Accordingly, samples of the coatings that had been exposed only to laboratory air end indoor light were retained and are still being weighed at suitable intervals more than years after they were spread. Moreover, in the course of tests fen* other purposes, coatings of additional paints, including emulsion paints, were prepared and weighed at suitable intervals. The worn reported here consisted of extensions and supplements to teste made for other purposes. As a result, it was not planned In advance as carefully as could now be desired. Despite some shortcomings in technique, hewever, the data are top significant to be left uareportad, especially since the work could not be repeated by iagurevad methods in less than & years. jfeperfeaental Procedure The linseed oil paints tested were ground from their ingredients at the Forest Products laboratory. Th emulsion paints, emulsion-type primers, and oil-typo primers were commercial products among which all but brand C bore statements of composition on the label. The content of nonvolatile material and the weight per gallon of all ccsmrcial paints were determined experimentally and agreed with oooynatations from the.label formulas. Faints were spread by doctor blade and suction plate on clean, t&red tinplate, 4 by 6 inches in else. The weight of wet paint applied was determined by reueighing the tinplate as tniioMy as possible (within 1 alnuta) after tbs vat paint was spread* Since ail of the paints contained volatile thinners (usually raiaeral spirits oar water), there ms a slight loss in weight while the paints wars bains Spread and the specimens weighed. Ouch error did not exceed a few milligrams ih a total weight of nonvolatile that raagsd from 1 to 3 grans, lbs weight of nonvolatile applied, which was calculated from -the treasured weight of vet paint applied and the Known fraction of nonvolatile in tbs vet paint, was therefore underestimated by an amount that can be neglected safely. vIba coated tinplate specimens wore weighed frequently, usually daily, for the first 10 days after applying paint. After that, the intervals between weighings were lengths*nod more or less progressively as the coatings became older and the weight changed less rapidly. Tho change In weight of coating at any tim van calculated as follows: in which is the weight in grams of the specimen at time t, is the calcu lated weight in grans of tinplate pine paint nonvolatile originally applied, is the volume in cubic eentiaetors of paint nonvolatile originally applied, and is the change in weight at tisaa t in milligrams per cubic centlmater of nonvolatile originally applied. Typical Graphs of the Chamaas in height with Ttoo In figure 1, the changes in weight for each of five single -pisaant linseed oil paints are plotted against the logarithm of time in days for a period of more them 2 years. lbs paints were iwQ& with antimony oxlda (A), rutile titanium dioxide C), lead-free aeleular abac oxide (s), magnesium silicate (X), and basic carbonate white lead (L), receptively, 1a tbs following i^ropoartlcsas: <R5>3S>51i!SClSo3Cw^ Pigment idnased *>11 Mineral srplrita 0.261 609 -A3. 1.000 i3x linseed oil always cootaiaod tbs smut amount of lead-oimeanaBe napfcfctomate drier* Approaclmtaly 84 bouts after spreading the paints, there woa a mt mla in weight of aottmlatile coating ranging from 65 aOJUgrams per cubic eeatJUaeter for point L to 110 milligram per cubic aentlaster for paint A. Only a stroll fraction of the weight increase could be mineral eplrtto still retained in the coating* It would be overasawrous to assutas that, after 4 hours, the coating retained 5 percent of th 105 sailllgrwa of aimral spirits originally present, which would be only 5*5 aliligraae. Sven this waall error is offset by the error of raagptttyto due to lose of mineral spirits wills spreading acid weighing the wet paint. ilia palate reached maxima in weight at 123 milligrams per cubic centimeter after 2 days for the antimony oxide paint, 103 milligrams per cubic centimeter after 2 days for the titanium dioxide paint, *j milligrams per cubic centimeter after 6 days for sine oxide paint, 79 milligrams per cubic centimeter after 6 days for magnesium silicate paint, and 67 milligrams per cubic centimeter after 6 days tor vhite lead paint* Clearly the pi&aeuts exerted marked specific effects oa the mechanism of drying of the linseed oil in tta& paints, ouen chemically inert plguacuts as magnesium silicate, titanium dioxide, and antimony oxide seem to bo as effective as the reactive pigments white lead and zinc oxide. After passing the maximum, all of the coating fallowed a generally- down* vard course, subject to sane minor va'verlng, foe an long an the experimnte have continued. After 300 to 4C0 days, the coatings of paints L and X -weighed lose than the weight of the nonvolatile applied initially, before oxidation and drying. Paints A and T ww fallowing a roughly parallel course. Paint , how ever, lost weight lees rapidly than the other paints. All of the curves In figure 1 exhibited a distinct dip near the 10th day. Ouch dip seems to be characteristic of linseed oil paints because It appeared In all tests mads subsequently. Apparently there is a brief period following the first attainment of maximum weight during which gain in -weight from absorption of oxygen again exceeds the loss of volatile decotxypoaitioa products, tkxiw of the other fluctuations in the course of the curves were due to variations in moisture content of tint coatings vith cbongss in relative humidity of the labo ratory air as discussed farther an. Variability of Results Faint `fx&gQ, p/nv 0.3(1 was included in six series of tests, aod paint TxLj^, p/nv 0.30, in two series started at different times between August 19Sfl and July 1969* Since the plotted results present a confusion of overlapping curves, certain characteristics of the individual curves are given in table 1, naxssly tbs gain or loss in weight (above or below the weight of nonvolatile initially applied) after 1 day, at the maximum gain, at 30 days, at 300 days, and at the last day of observation, and the average rate of change in weight during the intervals between maximum gain and 30 days, 30 to 300 days, and 300 days to the last day of observation* Although the variability la results with a given point tooted at different tJUaes was substantial, It was net great enough to Obscure the largur differences in results obtained, with different paints# specially when comparisons were toads snaong points tasted octac^cm)iQ<3Usiy. `Iha variations imdoUbtedly arise chiefly trem the uncontrolled variations in teo^ratucrQ and relative humidity in the laboratory In 'which the work was done* it la wall known that the mobeaaXm of drying of oil paints is sensitive to t^jerature, relative tajcaidlty, and illumination. Aging after drying seems to be largely a continuation of the drying mechanism and pcmswaatoly is sensitive to the same oribient conditions. Previous tests (1^ mad with coatings and films whan 10 days oLd^ indicated that the weight of taaisrfcure absorbed Aron air at relative humidity below 6s per cent was less than the loss of volatile decomposition products during the time required to sake the measurements. When tbs present tests were started, there fore, it was thought that the variations In humidity and temperature in the laboratory could be neglected safely. Such proved to be nearly true for the first 30 days of aging, but was not the case later on whoa, as table 1 shows, the rate of loss of volatile products fell to a fraction of the rate during the first 30 days. It was soon noticed that beyond age 30 days, the plotted curves presented a eaw-tooth appearance with ups and downs related to the prevailing relative humidity* 3M* was most noticeable, of course, during the suom&r months, especially so during the unusually damp eusaasr and fall of i960. When the oil-type and emulsion-type points of scries k were 3& days old, the coatings were equilibrated in a desiccator over calcium chloride, weighed, then equilibrated and weighed successively in the 30, 65, 00, and 90 percent "O** relative humidity rooms, and finally equilibrated and weighed &0&iu in the desiccator. Tha raeulte are presented In tha stepped bar charts of figure 2, For each point, tbs weight of the coatlas 'dion first e<uii.ibratd in the desiccator is taken as tbs bass for calculating tha subsequent changes in weight. In each bar, tbs lengths of the steps in order from top to bottom represent the weight change attained in 30, 65, @0, and 90 percent relative humidity and when returned to the desiccator, respcctivoly* 3ha -weight gain from moisture absorbed in 90 percent relative humidity massed frets 10 milligrams per cubic centimeter for oil-type primer of brand C to 55.2 milligrams per cubic centimeter for a?aulsion-type primer of brand iS. '^hen equilibrated in the desiccator for the second tine, none of the paints returned exactly to its weight in the desiccator 9 days earlier) oil paint 'SXZ^q and emulsion paint brand retained less than 1 milligram per cubic centimeter of the weight gained in 90 percent relative humidity, whereas the other paints fall 0.8 to 5.5 milligrams pr cubic centimeter below their weight in the desiccator the first tins, fen* most paints, perhaps the exposure to high humidity, which softens and swells the coating, may facilitate the escape of accumulated volatile products. Of the two linseed oil paints, XxSgQ inrovad wore hygroscopic than IbcL^, in agreement with past experience* Oil-type primer brand A, which was TxJu^ , p/nv 0AS, made with bodied linseed and tung oils, in hygroseopielty fell between paints Ifx&gQ and fritter brand B, which was Vx, p/txv 0.24, made with linseed albyd resin, and priawr brand Sty&, which was p/nv 0.36, made with lung-linseed resin, were less hygroscopic than paint Primer brand C was outstandingly lew in hygroscoplcity, `Hares of the emuleicn-type paints, the two primers and brand B finish paint, were more hygroscopic than lineeed oil -7- paint at least in 90 percent relative humidity. AH esmilaion-type paints except brand C were taoro hygroscopic in 90 percent relative humidity Usui linseed oil paint 'ExlitQ* <Xfa difference between 30 and 65 percent relative humidity in the laboratory vhen the aoatinga vere veiled could anke a difference of 8 to 10 milligrams per cubic centimeter in the moisture content of the coatings and readily accounts for the saw-tooth course of the later portions of the curves of change in coating weight with tine* *%&n that tests of series 3 were weighed on the 13th day (May 20# I9&9)j the relative humidity in the laboratory was about 60 percent, ami the coatings were found heavier than they bad been on the 10th day when the humidity was lower* l"he coatings were than placed in a desiccator for 0 hours and weighed again, faint Tx&gQ lost 6.9 aailllgraais per cubic centimeter. Fbn )) other coatings in tbs series, which consisted of paint -with various additives# lost from 4,1 to 8*6 milHgraao per cubic centimeter Similarly, the paints of series 1 were found heavier an the 780th day (October i960) than they had boon an the 535th day (February i960)* P^piilibratlon in the desiccates* produced the following losses in milligrams per cubic centimeters paint Ta%0# 8.0* ICaie^gQ, 6.2j Xk ^q , 6.9i 1# 6.C* A, ?.0j and T, 9.4. Linseed Oil joints of Series 1 Series 1 consisted of five single-pigment paints and three iaixed~pi*mt paints# all made with the mm raw linseed oil vehicle at pi&aent volume 0.30. For the single-pigaent paints, the cosylete data for change with tlrns in the gain or loss in coating weight are plotted in figure 1. 'fable 2 records significant points on the curves for all eight paints together with the average rate of change in weight during three intervals of time. One day aftor applying tb palate, tJhe coatings were heavier thou the weight of aoHvolatll Applied by 6.S e&lli&nma per cable centimetor for paint L to 113 BillXlgra^aB per cubic centimeter for paint A, Assuming that all of the mineral spirits in the paint had evaporated, the gain in wight representa the difference between the wight of ossygou absorbed fma the air and the -weight of volatile docoEpoaiticn products given off. lose of volatile products was substantial from the earliest part of the drying process. Rhodes and V&n Wirt (b) measured the eacygsn absorbed and the volatile products given off by coatings of linseed oil containing lead llncdoate drier and from coatings of linseed oil points made with various pigmsts when the coatings -were ossposed to pure oxygen for various lengths of time. Sees of tboli resuits, ^calculated for comparison. with table 2, ora given in table 3, `Heir pigmented coatings wore much loser in piseat volume and dried much slower than the paints of table 2, and the sothodo of test differed greatly. Nevertheless, Rhodes and Van Wirt, in agrssvasnt with the findings of this report, observed a rise in mt gain in wight to a nradmea followed by decline in net gain and the net gains at 10 and 17 days were rmaaritahly close to the mt gains for the most nearly comparable paints at mxlmsa and 30 days, respectively, in table . In partiratar, however, the results of Rhodes and Van Wirt show ttiat volatile products are given off In a substantial amount even during the first day after paints are applied. Both the absorption of oxygen and -Use loss of volatiles increase continuously from the beginning; only their differences, the net gain in weight, passes through a maximum. Although the net gain in weight of white lead paint after the maximum has been passed Is always less than the comspoading net gain for sine oxide paint, both the oxygen absorbed and the volatile products given off are substantially greater in white lead paint than In sine oxide paint. As reported la table 2, pigmentation with antimony oxide and titanium dioxide lad to relatively high net gala la weight after 1 day asd at maximum weight, which was attained after only 2 daya. It la unlikely that these pigaante hastened the early absorption of oxygen* Presumably, therefore, they retarded the loss of volatile products* In the stages beyond 30 days, whan changes In weight vara due chiefly to loss of volatile products, the antimony oxide and titanium dioxide paints lost weight more rapidly than any of the paints that contained sine oxide* The characteristic points for the mixed-pigment paints fell throughout between those of the two or three aingle-pliant paints node with the coa^ponenta of the mixture* Whoa sine oxide was a component, however, the mixed-pigsent paint fell closer to purs sine oxide paint than to the titanium dioxide or the white lead paint* Previous studies of tha swelling of paint films in water likewise showed that sine oxide tends to dominate over other pigment* in mixedpigment paints (2). The three paints that contained sine oxide exhibited distinctly slower rates of loss in weight beyond the 30-day mark than any of the other paints. It is possible that the rate of less of volatile products from the xixu;-containing paints is greater than the weldings indicate by reason of an offsetting gain In weight from slow absorption of carbon dioxide from the air and conversion of sine oxide to sine carbonate. It was found previously (g) that stocks of sine oxide pigaont, when stored in an unhealed and occasionally dacqp shed, become seriously carbonated. Linseed Oil faints of Series 6 The linseed oil paints of series 6, reported in table 4, reveal the effect of variations in plgaent volume, which is the volumetric ratio of total pigment -10- In the total nonvolatile ingredients of a point. Among the titanium sine paints anti among the titanium lead paints, the rate of loss.in velvet usually decreased as the pigment voluma increased. Such tread might he expected because of the decrease in the amount of oil In the coating as the pigment volume increases* The paints of very high pigaent volume rcached their laaxlxaum wight sooner than similar paints made at lower pigment volume* Thss titanium sine paint of pigment values 0*47 and one of those of pigment volume 0*33, during the interval from 300 to M5 days, not only failed to lose wight hut actually became slightly heavier* There was x k > gain in moisture content because the wight after k6o leys was checked after equilibration in a desiccator. Sudh gain in wight offers further evidence that the sine oxide in paints become cmrtxsmted by absorption of carbon dioxide from the air. JJ The last four paints in table 4 contained the relatively new white pigment, barium mtaborete, vhleh has been proposed as a substitute fear sine oxide to provide resistance to mildew without ashing paint sensitive to moisture blistering, substitution of barium metobor&tc fear sine ootids consistently hastened the attainment of maximum wight gain and led to more rapid loss In wight during subsequent aging. avulsion paints of the acrylic-resin type dry presumably without oxidation by coalescence as ths water of the emulsion evaporates. Therefore no gain in weight over tbs wight of nonvolatile ingredients initially present should be expected during the process of drying and aging. Since the acrylic resins are said to be unusually stable to light and moisture, tbs wl&st of the dried 11 paint coating night be expected to regain constant during subsequent exposure to air in the laboratory* Seats of ccsoaartiial eimalaion'^ypa points, however, did not coufcma to the expectations* Sable 5 reports results with five brands of emlaion paints, 2 emulsiontype primers, and four oil-type primrs of brands ccarrasponding to the miaion paints. One ramUfcaturor supplied both an cdl-type and an emulsion-type prisxsr. Four of the emtlpioa paints, 1 omulaim primer, and 1 oil primer failed to reveal a period of Increase in weight following sraporatim of the volatile constituents, possibly because there were insufficient weighings during the first 24 hours after application. Hsv^rthelesa, all but one of these products yielded coatings that were decidedly heavier than the nonvolatile portion of the paint applied. All of the coatings, without exception, lost waigfat cqatlwuously during exposure )) and were still losing weight after 49S days:* "She weight gaim 1 day after application, except for emulsion paint of brand C, ranged from 48.0 to 98.V milligrams per cubic centimeter coopered with 72*0 and 88.0 s&lligratae per cubic cenfcioctcr for the two linseed oil paints tested ccsitssixraasously. Souleioa paint of brand B, which took 8 days to reach a xaaxltaua waigat gain, was made with an olltyd-vesin modified acrylic resin* Sara oxidation of the alhyd-rasin ccnponcsxt is possible* ftauisice primer of brand |2, which reached a tmxinsm weight gain after 4 days, was made with a linsed*lhyd-resin emulsion rather than an acrylic eoulsion* Pulsion vehicles contain numerous additives ouch as csulaifyiue agents, surfactants for pigaant dispersion, antilbming agents, depressants of the freezing point of water, and fhngleides* Stone of the additives listed in the numerous published formulas for acrylic-rasin finish coats, however, is a L) material that oxidises readily at ixsreaal bx^raturee. itawlBloa primers fir socratinme contain drying oil for batter pemtrailua end bending to chalky aurfaces, but no such additives to emulsion points tor finish coats appear la the literature. Tba larga weight galas above tha nonvolatile content of the liquid paint there fore suggest that the acrylic resins absorb oxygen as they dry, although the laochaniaia of such absorption is soot coherent. lba l4ay old coating of* esulslon paint brand C was lighter them the nonvolatile components of the liquid point by 41.9 milligram* per cubic ecntiraeter and subsequently* awttinued to loan vol^fet* lass infOruatlon about cot^oaltioa van given on the label tban <m those of any other commercial point tested. Hie initial content of nonvolatile ingredients, however, was deterained In the cuatoiaary vay be heating a sample at 106* C for 3 hours. It la unlikely that any Ingredient left after 3 hours at IDS* C. could evaporate to the extent of j)4l.9 mUligraas per cubic centiaetar In 24 hours at room teaorature even whan exposed in a coating only 2.6 alls thick, ihreaumaibly either the acrylic resin or an additive present In the paint in substantial proportion evolved volatile deccBVKraitina products during drying to on extent that exceeded the absorption of oxygon, if such absorption took plane. All of the waHlslna paints and the emulsion- or oil-type primers tor use with them decreased in wel#xt steadily after the first day or the aaaxhatm weight within 4 days* All were still losing weight after 495 days. l?ubllstoed formulas tar eoulslon paints contain such high boiling liquids as ethyl collosolv^, ethylene glycol, and polyglycol added to depress the freeaiag point. Bthyl cellosolve has a vapor pressure of 12 milllooters of mercury at 20* C., ethylene glycol only 0.12 millimeter of mroury. Such additives would evaporate from coatings very slowly, but It is doubtful that any would remain in coatings less 13- paints, the proportions of high boiling liquid additives range frcia O to 100 milligrams per cubic centimeter. Bam formulas for polyvinyl acetate emulsion paints afecotsaend as much as 220 milligrams per cubic centimeter. Thus tha paints tested might have contained, but did not necessarily' contain, hlgfc boiling liquids to account for the weight losses observed, provided that the emporation was alow enough to persist for more than 49c days, tteverthelosa, the initial weight gain by all but one of the emulsion painty together with the long continued subsequent loaa in weight, suggests that ostidation and volatilisation of decomposition products occur in emulsion paints to nearly as great cun extent as they do In oil paints. The paints of aeries 2, 3, and 5, reported in table 6, which la in two parts, were made primarily in a search for means of reducing the swelling in water of ainc-containing paints such as SxSgQ, p/nv 0*30. Xt vas of interest to learn also whether the additives affect the course of the changes in weight of coatings during aging in the laboratory. Many of the additives exerted little or no effect on the course of the weight changes during aging of the coatings. XM* group included additives that increased the swelling in water slightly, such as tetralsopropyl titanate, and additives that markedly decreased swelling, such as silazana 3X212. Other additives increased the net gain in weight after 1 day and the rate of loss in weight from the maximum to 30 days* This group included additives that increased the swelling in water greatly, such as seya secondary amine, and additives that Y*. I f decreased swelling substantially, such as octylene glycol titauato. ibus no relation was rvavmlad between tha effect on weight change and tbs effect on swelling in water. The effect of the additives on weight changes was exerted chiefly during the first jO days) after that tisaa, the further rate of lose la weight was usually similar to that of the point without additive. (1) After a linseed oil paint is spread as a coating exposed, to air aud its volatile thinner evaporates, its weight increases for a few days, reaches a usaximuta, ^ then declines. `fbx increase observed la the difference between the weight of easysn absorbed and the weight of volatile deccejpoaitioa products given off. At the asastuRsa, the two processes are in btklance) beyond the esutixauu, evolution of decomposition products exceeds any further absorption of oc<ygea. Since oxygen absorption and decatspoBitian are each affected indcpeadontly by envlreaecmtal conditions of temperature, relative humidity, and irradiation by light, quantitative measurements of a given paint started at different times vary slightly unless the environmental conditions ore ccmtrollad store closely than is the cose in our point laboratory, (2) ifce gain in net weight of a linseed oil point and the subsequent rate of decline from tbs tasatlmua vary widely according to the nature of the pigments in tbs point. Paints of high pigment volutes reach :aa*iiasa weight sooner and t-fry* usually decline in weight less rapidly than similar points of lower pi$mt volucaeu Points made at their critical pigment volute? may reach unusually high raygiMMfn weight, and sy decline rapidly fresa tbs raaxtnsuni during the next 30 days, there is also evidence that zinc oxide la paint my absorb carbon dioxide fpea the# air, which offsets part of the loss of decomposition products of the oil vehicle. -IS- (3) Although the rata of lose of decomposition products boocraaa slower as tlsss passes, it c<xrtinuos tor as long as tha observations have lasted, \jhlch Is laors than 2 years for tha points of series 1. Thus Oeccupoaitictt goes ou coutimously even under the very mild eixpootmj to the envlroumsatal ccnditioos of tha point laboratory. At tha end of ?6o days, uhite load paint was lighter than the wight of nonvolatile ingredients applied by 55.5 Milligrams per cubic contisieterj vhich was 10.2 percent of the linseed ail In the paint. (4) i&sisB additives to linseed oil point mrtodly offset the wight change during the first 30 days, They caay either increase or decrease tha net gain at mxlsain veight. Thaws 1 no relation, houover, between the effect of additives on wight changes and their offset an the duelling of paint coatings in water. (5) Surprisingly, umt acrylic-acmlsion paints, which do not depend on oxidation for their drying, exhibited a net gain in wight ccosporable to that of linseed oil paints. Hh* emlsiaa points, however, usually attained their iwrciHsua weight during the first day after application. Ho explanation has been found for the net gain in weitgat other than the assumption that the acrylic resins ore subject to oxidation by oaaa raachaniwa not yet isnown. Moreover, the eswlsioa points, on subsequent aging, lost weight steadily and were still losing weight after 495 days. Xt regains to bo determined whether the loss in weight is due to decos&Hbsltion of the acrylic resins or to slew evaporation of high boiling ll$*ld additives such as etfcylom glycol, polyglycol, and ethyl cellosolv, which are often used as freezing point depressants. (6) Coatings of all paints tested, whether of oil type or amLsion type, remain significantly hygroscopic even after 2 years or acre. Even the moderate changes in relative humidity in the paint laboratory produce corresponding changes in >i weight of the coatings. -l>- (1) Browns, y I , 1955. Owilling of Paint Fllais la Viator. XXI* Absorption and VolmaBtrlc Swelling of Bound and Fro Filnua Frees Air of different Illative ISumiditlefi. Forest throduets Journal Sj 9*2~6. Stalling of Paint Pilaus la -feier. XI. tlixed-Pieseat Paints in Linseed Oil. Forest Products Journal 7:248-52. (3) 1969* doderstand1n& the itocteaalaaa of Deterior&tlm of Ecus Palate. Forest Products Journal 9*. 417-27. ! (4) E&odee, ]?. a., and Van rfirt, A. E. 19^3* Effect of Various Pigasnts Upon tbs Bate of Osidatim of Linseed Oil. Industrial and i&glnsoring Ctoalstry IS: 1,135--j O. -IT asslia * 4 * - * O 4 S&88*8*8 o HI & Ml 3% o* <*. * ** VJ** 1 O onflow SiKK^fiS fli' % o o oooo l WOWOOU) to o* aU covo c*\w to `O 31 l PJ w O OOI OV &!&*$ rl o id to o o o * 10 O ! 2 (Ligglfei v lt ii Item Tim* after applying the coating; N **%** **!* ** ** * ; 1 day j 2 days t 10 days : 17 days * - Ljnaaad oil containing drier but no piaecanrfcfc3i ; Y^WS**' abaorbed per cc*; 168 _ /jointOn products lost ,..W0* per cc.: 65 '""Hart gain in vaiebt...is^. ppaarr ec.; 103 s t 233 t 102 : 131 Linseed oil paint L, p/av 0,08i Oxygen absorbed, .usg per cc.{ 136 196 per ec*t 34 t J& per ce.s 10c s HI Linseed oil paint p>mr 0*10: per cc.s per ec.i peer ecr.i 75*4 ; 134 8.4 $ J51 67.0 * 113 * i 326 J t 233 . t 93 3 i 320 s&l t T? 216 : Iff) j 92 3 1 * 1 3 331 &*< &} 337 22 6-4 23L Jgg. 84 ^ycfcale hove the following joinings*L, basic carbonate \Mte loadj Zt sine oxide* p/nsr, pigment vuLuuo or fraction of total pigasnt fcy volaraa in tba total nonvtxlatilo setter by v o Iu bbo , 'Uatols 5 --Tests o f oiltaroe and enilflico-^Eroe nrlaara aafl -paints o f eerlag 4. v.. <5 S *i*ll* IS #tmt M + #4 # 'O' 8<$ Ih*T pit if fi if*fi .** ** ** * ** * # gs W * ** *'4 ww 33 "` Ml *\CM tO d g to o id lO OVCO O o 3sfd<$ O * uw GJtfJ c- ill | * * *** fSt sH> 0 COCO ids *j 4* H t~0 HVO id is 4 ' * * * # t O Ml i o4H oW S *o*i**-*** n *> 3 o * i # IS l**s *-%*4*>* & tf M t CO ^ id mmm M o 04 9@tj2 * * t vO Oi-j O' dids* *4 *4 R M | ** oo oo |6 66 offtT a p T ww ^J(%Q% M*OU p iigi >1 %_m isiss m o * * I * Ml * * 44 ft* 4* 44 ft* *4 #Lft ft* 4* ft 9* ft* I %$l 3 f 8181 g88 $8 s $ d 1 5) Pi ** * 44 * ft* * ft* 4* ft* ft* ft* *4 ft* *ft ft* ftft ftft ftft ftft ftft ft* * 3 sj PR$3 ftp r| H * ** * *4 ft* ft* *4 * ft* * *4 ftft #* Ol6$4 IflH BmS sa&a <0X1. Wrj '** a$ Qs?s*??S8S k j j s O 0 32?^ AOinoo o o 33&5 ** ** ** * * #* ft# ft* ftft ftft ft* ftft .ft* ftft- o oo OOOO Wi> iut>) o o h wwwow t~o ' s gsidtj #sa$s as 44 w pps ftftft# ft* ftft * o *}0 w asris sj a ** * * ft* ft# ftft ftft ftft ft# 44 4* * wo w t- mtaooiA w gags s'4&8f 44 4ft *4 44 * 4* . wVO w*W moot *4 ft# 4* ft* ftft ftft 44 44 44 ftft *4 *'' <*"* <%***,**%.^ aeaa a d woo w mj ^oo<5 ttas* p | |js% *4 4% 4 * _ ____ ,, ,, eaaaa as wow w>u> 00 sg^&g s'g * * # *4 44 ** 44 44 4* ***' ** 44 4# 44 44 44 44 44 44 w ir-fr-t'* woo wowow ow M gSSti asgs &{= * 44 4* 4* ft* 44 44 44 * ** *4 ft* 4 4* 4* ftft ft* ftft 44 ftft 44 44 44 44 44 44 H 3833 3 3 3833 '""'S S" 4 *4 ft *4 4ft ft* 44 *4 44 ftft ftft 44 44 44 44 NO III tf tj * / f: K I1 f i- * tJ * * * *# ft t 41 It BUS 88M **# * ' * * * * t* I# I # t ** ft ft M It t * 1 ** t* ## I H $ 1 l* t *% H *ft t * M It ** O H 4* ft* ft I a 3 V * # 4* ** * #* n m itncrvo W* >4 # 4# t gPW5s^f^c0* % ** M oHn p j ms w o h U&ftijl sisi^^i8 * ** ** 44 * 44 4 if ff ft t* t ftt tt as SB &\0\ I* 9 ** 4* #4 *4 * MM0JB& 0| * ft- * * ft tt ft# ft i t t # 8S ssssHQHQHQtft oflrgtrotHow M> *H *t * 4* * It f # ft ft ft If tt ft *t tt Figure I.--Chaofijfta i wight with time of cnqposuro Indoors for cootinra of linseed oil points pl0a&ted * 0*30 ple&snt valutas with (A) antiiooay oxide, (T) rutile titaaiu* dioxide, (Z) nine cadde, (X) aaenasiua silicate, and (h) basic ** ** whlto lead, Chaugeo la weight os* In nUllgraas i>or cubic coatlmtor of coating or balcwr the weight of ncsivOl&tiU mterlal in tho liquid point whoa spread! tins is la days plotted on a lo^rithmlc scale. (MllSSSO) fr I 3 { 411 $ H llti 550 Figure 2.~ChBn0M la valgit duo to change iu moisture content of 384ay~old coating* of iwtoni paint* brought to equilibrium successively with tbs atmosphere (l) to desiccator over calcium chloride (o moisture content), (2) la 30 percent relative humidity roan, (3) la <>3 percent relative humidity roan, (4) la 00 percent relative tmutdity room, (3) in 90 percent relative humidity roam, anti (6) la the* desiccator again. 04 X10 931) m 118 EMULSION P A IN T EMULSION PRIMER -OIL TYPE PRIMER LINSEED