Minimizing
W o o d Shrinkage and Swelling The hygroscopicity and subsequent swelling and shrinking of dry wood are decreased by heating in various gases above thermal decomposition ternperatures. Greater reductions in hygroscopicity are obtained in an oxidizing than in a reducing ,atmosphere for the same heating conditions : but, by increasing the temperature, equal reductions in hygroscopicity can be obtained in reducing atmospheres. The darkening of the wood on heating appears to vary directly with the resulting reduction in hygroscopicity, regardless of heating conditions. Soaking in water after heating has but a slight tendency to restore the original hygroscopicity. Heating wood in water-saturated atmospheres has no permanent effect upon the swelling and shrinking.
Effect of Heating in Various Gases' ALFRED J. STAMM AND L. A. HANSEN Forest Products Laboratory, Madison, Wis.
in the toughness of 50 to 75 per cent. Data of Greenhill (3) on the relation between maximum strength of beech in tension perpendicular to the grain and the moisture content at different temperatures indicate that the loss of strength on heating of the wood decreases with a decrease in moisture content to a negligible value for very dry wood. Although the temperature range covered by these data is below the temperatures required to obtain appreciable antishrink efficiencies, they indicate the possibility that less loss in strength will occur if the wood is adequately seasoned a t normal kiln temperatures before being subjected to the higher temperatures required to reduce its subsequent swelling and shrinking.
Heating Experiments The purpose of this preliminary research was to confirm this loss of hygroscopicity of wood which occurs on heating, and to determine its permanence and how i t is affected by the
F
OR a long time it has been recognized that excessive
heating of wood reduces its hygroscopicity. Tiemann ( I O ) found that heating air-dry wood in superheated steam to about 150' C. for 4 hours reduced the subsequent moisture absorption by 10 to 25 per cent with but relatively small reductions of the strength, except for red oak which showed a reduction in crushing strength and modulus of rupture of about 60 per cent. An unpublished Forest Products Laboratory report ( I ) shows that heating black gum in dry air at 205" C. for 6 hours reduces the subsequent hygroscopicity to almost half of its original value with only a s li g h t accompanying decrease in its strength. Koehler and Pillow (5) and Pillow (6) heated a i r - d r y S i t k a s p r u c e and ash to 138' C. for 1 to 8 days and obtained reductions in the equilibrium moisture c o n t e n t a t s e v e r a l different relative humidities for the longer time of heating of 30 to 40 per cent with accompanying reductions in the crushing strength of 15 to 25 per cent and reductions 1 Other papers in this series appeared in April, 1935, page 401,i n ' December, 1935, page 1480, and in October, 1936, page 1164.
medium in which heating occurs. Sections of white pine 9 cm. long in the tangential direction (the direction of maximum swelling), 2 cm. radially, and 0.6 om. in the fiber direction, were used. The short dimension in the fiber direction ensured rapid attainment of moisture equilibrium and made possible the cutting of a number of adjacent sections from the wood with a minimum variation of structure from section to section, Four specimens were suspended in a small steel bomb that was heated on the outside by an electrical resistance coil and lagged with asbestos insulation. The temperature was determined by a t h e r m o couple inserted into a well in the wall of the bomb. The temperature was controlled m a n u a l l y to about 2' C. W h e n t h e specimens were heated in gases other than air, the bomb was evacuated and the gas admitted s e v e r a l times to ensure the elimination of air. The m o i s t u r e content of wood used in the tests was about 6 per cent. When the heating was carried out in presence of water vapor, a large excess of water over that necessary t o saturate CHARACTERISTIC SHRINKAGE AND DISTORTION OF FLATS, the specimens was placed in SQUARES, AND ROUNDS AS AFFECTED BY THE DIRECTION OF THE the bottom of the bomb. ANNUALGROWTH RINGS 831
832
INDUSTRIAL AND ENGINEERIXG CHEMISTRY
Effect of Gases Table I gives the effect of heating the white pine sections in differentdry atmospheres upon the subsequent swelling and shrinking. Measurements were made of both the tangential dimension change and the weight change occurring when the specimens were alternately brought to equilibrium with 30 and 90 per cent relative humidity in humidity rooms held at 26.7" C. The specimens were exposed 2 weeks at each relative humidity, which proved adequate for the attainment of equilibrium. The antishrink efficiencies are calculated on the basis of the reduction in dimensional change or weight change between 90 and 30 per cent.relative humidity. These changes
VOL. 29, NO. 'I
still more so after the time was increased to 6 hours. I n each of these cases the darkening obtained by heating in the different gases increased in the following order: hydrogen, illuminating gas, air, and oxygen. The specimens heated to 260" C . in hydrogen, however, were as dark as those heated for 6 hours at 205" C. in oxygen. The darkest specimens were about the color of unfinished walnut. The antishrink efficiency appears to parallel the darkening of the wood, irrespective of the temperature and the gas used. Although part of the antishrink efficiency may have been due to oxidation in the cases where the wood was heated in air and oxygen, it is hard to imagine that this was a major factor since equal efficiencies can be obtained by heating in hydrogen at a slightly elevated temperature. The phenomenon can best be On the OF HEATING DRYWoo0 TABLEI. EFFECTUPON ANTISHRINKEFFICIENCY basis of thermal decomposition. Loss of water IN VAR~OUS GASESAND SUBSEQUENT SOAK IN^ IN WATERFOR 5 DAYS --Antishiink Effioienoy-of constitution is the first thermal reaction. If -Before Soakingal---Aftei Sosking*this loss were due to the formation of an ether Time Tangential Tansential of dimension Weight dimension Weight linkage between two adjacent cellulore chains GSS Temp. Hesting basis basis basis basis through adjacent hydroxyl groups, the loss in " c. HI. % % % :8 hygroscopicity conld be readily explained. Not Hydrogeo 165 0.25 5.9 8.3 2.8 20s 2.00 18.0 17.0 11.5 11.4 only would the hygroscopicity he reduced be260 2.w 32.0 32.0 31.8 31.2 4 , 8 cause of the substitution of the less hygroscopic Illurniosting gss ins 0.25 8.2 8.5 6.3 205 2.00 18.0 19.0 14.0 13.2 ether group for the more hygroscopic hydroxyl 205 6.00 20.6 19.0 19.8 17.8 4,4 groups, but also because of the parallel bonding Air 185 0.25 8.3 8.4 4.9 20s 2.00 17.5 19.0 14.0 12.3 of the cellulose chains. Staudinger (9) showed 23.2 ".' ".* 21.0 that. the formation of such bridges between the oxygen ins 0.25 10.0 12.0 7.0 n.1 206 2.w 20.7 21.0 15.4 13.8 chains in polystyrene resins with p-divinylben205 6.W 28.0 30.0 28.7 27.2 cuts down the tremendously even * In term of retardation of the dimension and weight ohangee, for the average of four g-. unit change of the u n t + e d controls when sltemstely brought t o enuilibrlum when only enough pdivinylbenzene is used to with 30 ad 90 per cent relatlve h + m h t y . form a Single bridge for s e v e r a l t h o u s a n d bna~edupon the second humidity ohsoge oyok m the first ia appreciably affected by byatereek (11). molecule8 of m o n o m e r i c styrene. Just an occasional cross link evidently cuts down a p preeiahly the tendency for water to he taken up between structural chains. The formation of ether linkagea are referred to the corresponding change for the control. The between the hygroscopic hydroxyl group not only explains the sections heated in dry atmospheres were soaked in water for decreased hygroscopicity of wood heated in dry atmospheres 5 days after the humidity cycles were complete and then subbut also the fact that heating in the presence of a large excess jected again to the humidity change cycles. Only the second of water vapor causesnochangein hygroscopicity. The pressubsequent humidity cycle was used in the calculations, since ence of an excess of water vapor would depress the thermal the first would involve a higher desorption curve due to the reaction in which water is evolved according to the principle soaking and thus give results affectedby thesorption hystereof LeChatelier and thus markedly reduce the tendency to sis (If). The same is true for sections heated in the presence form the ether bridges. If the change in hsKroscoDicits were of water. The first cycle gave appreciable negative efficiencies merely a physical change, such as &at po&Iatez by-Urqubecause of this hyst&esis-effec
