\s)
FOREST PRODUCTS LIBRARY
FOREST RESEARCH LABORATOR
(r)
OREGON STATE UNIVERSITY
EFFECT OF INORGANIC SALTS
UPON TUE SWELLING AND
TUE SIHRINICING OF WOOD
Information Reviewed and Reaffirmed
March 1936
No. 1156
Minim Ain imp p
minim mill, lip
UNITED STATES DEPARTMENT OF AGRICULTURE
FOREST PRODUCTS LABORATORY
MADISON 5, WISCONSIN FOREST SERVICE
In Cooperation with the University of Wisconsin
CO c
EFFECT OF INORGANIC SALTS UPON THE SWELLING
AND THE SHRINKING OF WOOD
By
A. J. STAMM, Chemist
Forest Products Laboratory,? Forest Service
U. S. Department of Agriculture
Wood, like all other fibrous materials, possesses the property of shrinking
and swelling upon the loss or grain of moisture. The swelling, according to
the classification of Katz,3 is of the maximum swelling type. On drying
green wood no shrinkage occurs until the moisture content is reduced below
the fiber-saturation point (the moisture content below which the activity of
the water becomes less than one). The shrinkage as manifested by the percentage change in external dimensions increases with an increase in the
density of the wood. It is greatest in the tangential direction (tangent to
the annual ring), varying from 6.o to 12.0 percent. In the radial direction
(from the center of the tree to the periphery) it varies from 3.0 to 7.0 percent and in the longitudinal direction (parallel to the fibers) it is only a
small fraction of a percent. The practically negligible longitudinal swelling and shrinking is accounted for on the basis of the swelling taking place
only between the threadlike micelles which, in general, are oriented parallel
to the length of the fibers. The greater shrinkage in the tangential direction than in the radial is accounted for on the basis of ray cells which have
their longitudinal component in the radial direction of the wood, restraining
the dimension changes in that direction.
X-ray studies have shown that the swelling of a cellulosic fiber by water
does not affect its crystal lattice. 3- The water must go between the constituent units of the structure rather than within these units (intermicellar
swelling as contrasted to intramicellar swelling). Concentrated solutions of
certain salts such as lithium thiocyanate, as well as of alkalies, however,
change the crystal lattice,t showing that their action is intramicellar.
–Presented before the Colloid Division, American Chemical Society, Chicago,
Ill., Sept. 11-15, 1933. Published in their Journal May 1934.
2Maintained at Madison, Wis., in cooperation with the University of Wisconsin.
Katz. Trans. Faraday Soc. 29, 279 (1933).
-Katz and Derksen. Rec. Tray . Chim. 50, 149, 737, 746 (1931).
Rept. No. 1156
-1-
Agriculture-Madison
Unfortunately, it is extremely difficult to determine the total swelling of
fibrous materials from the dimension changes. With wood, where the external
dimension changes are easy to follow, it is extremely difficult to even
approximate the internal changes that take place in the lumen of the fibers.
In a homogeneous material free from stresses and strains the lumen would
change in size on swelling and shrinking of the material, the same as the external dimensions of a piece of the material which would just fill the lumen.
This, however, is not the case for wood or any other fibrous material.
Because of these difficulties, swelling and shrinking measurements have, in
general, been made in terms of weight change. This procedure may, however,
be very misleading as swelling can only be proportional to the weight change
over the adsorption range. Absorption by the grosser capillary structure cannot cause swelling. Data for the absorption of salt solutions by wood, such
as that of Schwalbe and Fischer are, therefore, hardly a measure of the
swelling and further give no information as to whether the salts remain in
solution in the water which becomes sorbed, or if they are merely dissolved
in the free capillary water. There also seems to be a lack of information on
the effect that these salts have upon the subsequent shrinkage of the fibrous
materials.
Swelling Measurements
To avoid the foregoing difficulties as far as possible, measurements in the
studies reported here were made on a combined weight change and external
dimension change basis on small wood sections of Sitka spruce and white pine,
approximately 4 by 4 by 0.2 cm., the short dimension being in the direction
of the fiber length. Practically all the fibers were cut across at least
once, thus avoiding difficulties in penetration and in removing of air from
the sections. For most of the measurements the sections were first oven
dried to determine the weight of the oven-dry wood (although this does
slightly affect the subsequent swelling, it was shown to have no effect upon
the nature of the results). They were then soaked in water with the application of alternate suction and pressure to remove the air. This procedure
took practically a day to replace all the air with water. The sections were
then rapidly wiped on the surface with a towel, weighed, and the dimensions
determined. The completeness of the removal of air was tested by a comparison of the weight of the water taken up with the theoretical amount that
could be taken up, calculated from the total void volume. The fractional
void volume is given by the equation
V = 1 - d
do
(1)
in which d is the overall density of the wood and d o the density of wood
Schwalbe and Fischer. Kolloid Z. 57, 338 (1931).
Rept. No. 1156
-2-
6 At the bottom of columns
substance determined in water. —
5 and 6 of Table 2
such a comparison is given. The agreement is satisfactory, considering the
difficulty in removing surface water without removing capillary water when
wiping the sections.
The water-soaked sections were then placed in saturated salt solutions with
an excess of salt present and allowed to remain for at least two weeks.
Swelling beyond the water-swollen dimensions was found to be complete for
most of the salts in a few days, but calcium and magnesium chlorides required
about a week to come to equilibrium. The swelling of Sitka spruce, which
is expressed throughout this section of the paper as the percentage swelling
beyond the water-swollen condition in terms of external dimension changes, is
given in Table 1 (columns 10, 11, and 12) for a series of chloride and potassium salts together with some of the physical properties of the salt solutions.
In no case was there a measurable longitudinal swelling. The radial swelling
was much smaller than the tangential swelling and where the tangential swelling was small the radial swelling was negligible. The order of decreasing
tangential and volumetric swelling for the different salts is practically the
same as the order of decreasing solubility of the salts in terms of mol equivalents per kilogram of solvent (see Table 1, column 3). The only reversal of
order occurs with magnesium and manganese chlorides. In this case the solubilities differ but slightly and as magnesium salts stand higher in the lyotropic series for swelling and dispersion of colloids in general, such a
reversal might be expected. The relative positions of these two salts with
respect to other physical properties given tends to further confirm this
reversal effect. Potassium bichromate is the only salt considerably out of
position. This salt seems to be reduced by the wood as shown by the greenish
color. The amount taken up by the wood is also abnormal, as shown in Table 2,
column 7.
The order of decreasing swelling is also the same as the order of increasing
relative vapor pressure over the salt solutions (see Table 1, column 5) in all
cases but the abnormal potassium bichromate. If the salts entered only the
gross capillary structure of the wood and not that within which swelling takes
place, the sections should all shrink rather than swell and those which gave
the greatest swelling should have shrunk the most for this would be equivalent to having small desiccators throughout the wood for controlling the
moisture content of the cell walls, according to the relative vapor pressure.
The fact that wood is swollen the most in salt solutions which show the
lowest relative vapor pressure is strong evidence that the salt solutions must
enter the swelling structure.
It is interesting to note that the order of decreasing swelling is also the
order of decreasing surface tension of the saturated salt solutions (see
Table 1, column 6) with the one reversal of position of potassium oidide and
chromate. The fractional volume of the salt solution occupied by salt
in solution (see Table 1, column 4)
-Stamm. J. Phys. Chem. 33, 398 (1929).
Rept. No. 1156
-3-
100 + g
ds
V
sol
=
-
100.1
(2)
100 + g
ds
where g is the solubility of the salt in grams per 100 grams of water and dB
is the density of the salt solution, follows the swelling order with but two
changes of position.
The effect of the hydrogen-ion concentration upon the swelling was determined.
The measurements were made with a glass electrode. Hydrochloric acid and
potassium hydroxide were used for varying the pH in a salt-free unbuffered
system. The swelling-pH relationship is shown in Figure 1. There is no
swelling on the acid side, but an appreciable swelling on the alkaline side
at higher pH values. The pH values of the saturated salt solutions (see
Table 1, column 7) were also determined. These show no apparent relationships with swelling. Only in the real alkaline salts could any of the swelling effect be attributed to the pH.
The order of the solvent power for cellulose of saturated salt solutions at
their boiling point (see Table 1, columns 8 and 9) as given by Herzog and
Beck,1 is the same, as far as the data go, as the order of swelling of wood in
the salt solutions at room temperature.
Measurements of the swelling of the wood sections were also made by placing
the oven-dry sections directly in the saturated-salt solutions. The same
equilibrium swelling is, in general, established as when the sections were
.first water swollen, but the process is much slower. The salts tested required at least 9 days to give swelling equilibrium. Magnesium chloride,
however, required 58 days to swell to the water-swollen volume. After
100 days the swelling had not yet reached the equilibrium value attained
in 7 days by previously swelling in water. This effect might be expected on
the following basis: Water alone enters the fine structure readily, causing
swelling. After the fine structure is opened the salts can diffuse therein
without any great difficulty. If the dry wood is placed directly in the
saturated salt solution, the salt solution, which in the case of calcium and
magnesium chlorides shows no tendency to wet wood as indicated by the
spherical shape assumed by small drops placed on the surface of dry wood, can
evidently not enter as such. Presumably the wood tries to take the water from
the salt, but such a displacement of the salt-water equilibrium requires that
additional energy be expended in the adsorbing process. Only a small amount
of swelling could thus take place before salt would have to diffuse into the
adsorbed water in order to make possible further swelling. This diffusion into an only partially swollen structure would be very slow.
7
–Herzog and Beck. Z. Physiol. Chem. 111, 287 (1920).
Rept. No. 1156
-4-
The swelling of wood in the saturated salt solutions is practically reversible
for salts which do not hydrolyze. When the salt is removed by repeatedly immersing in fresh water for a couple of weeks, the original water-swollen
dimensions are nearly resumed.
Measurements were made of the swelling of Sitka spruce in solutions of lithium
chloride and potassium thiocyanate of different concentrations. The sections
were oven dried, weighed, soaked in water and the air removed, reweighed,
dimensions determined, and then placed in a small measured volume of salt solution of known concentration. The containers were sealed and allowed to stand
for two weeks to insure swelling equilibrium. The dimensions of the sections
were then determined. The equilibrium concentration of the solutions was also
determined by evaporation and weighing the residue. The percentage volumetric
swelling beyond the water-swollen dimensions is plotted in Figure 2 against
the concentration of the salt solution in equilibrium with the wood. No consistent variations between the initial and equilibrium concentrations were
obtained, indicating that any selective adsorption is probably small. This is
in agreement with earlier measurements of the author which showed that the
selective adsorption of ions from dilute solution is very small.—
Shrinkage Upon Oven Drying
The data given in Table 2 show the absorption of salts by wood and their effect upon the shrinkage when oven dried. The actual weight of absorbed salt
solution (column 5) was obtained from the difference between the weight of a
salt-solution swollen section and the original weight of the oven-dry section.
The theoretical weight (column 6) was obtained from the void volume (calculated
with the aid of equation 1) and the density of the saturated-salt solution
(column 3). The agreement is as good as could be expected, considering the
difficulty in removing only surface solution and not capillary solution when
wiping the sections for weighing. The actual weight of absorbed salt (column
7) was obtained from the difference in the weight of the oven-dry wood with
and without the salt present. This again is in fair agreement with the theoretical weight of salt (column 8) obtained from the void volume and solubility
of the salt in grams per 100 cc. of solution (column 2). Both of these comparisons again indicate that the salts virtually attain the same concentrations
in the sorbed water as in the free water. It is true that these data, as well
as the analytical measurements just cited, do not constitute so conclusive a
proof of this generalization as might be desired due to the fact that the
sorbed water occupies only a small fraction of the total void volume (in this
case 0.155). Any change in the concentration of the salt in the sorbed water
would thus be 6.5 times the measured difference for these data and considerably
more in the analytical case cited earlier where a still greater excess of
liquid had to be present for analysis. Although the differences between the
actual and theoretical values given in Table 2 are practically within experimental error, there seems to be a slight tendency for the former to be greater
than the latter, thus indicating the possibility of a slight positive adsorption of salt. According to the following data this adsorption must be very
small, however.
Rept. No. 1156
-5-
The shrinkage of these sections from the salt-swollen condition to the ovendry condition (the sections having been dried at room temperature over P205
prior to oven drying at 115° C.) are given in Table 2, column 9. The measurements show a retardation of the shrinkage even after the moisture is virtually
all removed. The more soluble salts in general retard the shrinkage to the
greatest extent as would be expected if these salts were deposited in the
swelling structure. Assuming that the retardation of shrinkage is entirely
due to this effect and that the sorbed-water salt concentration is virtually
the same as the free-water concentration, it is possible to calculate the expected shrinkage from the fractional solution volume occupied by dry salt
(column 4) by means of the equation
V dry = gs
(3)
where gs is the solubility of the salt in grams per 100 cc. of solution and
qi is the density of the dry salt. The expected shrinkage is
S = (1 V) S o
(4)
where S o is the shrinkage from the completely swollen condition to the dimensions that would be obtained on drying if no salt were present. These theoretical values are given in column 10. In general, they agree quite well with
the actual measured values, especially where the swelling correction is very
small. This agreement furnishes the best evidence obtained that the sorbed
water attains a salt concentration virtually the same as that in the bulk
water. The quantity measured here is a function of the sorbed water and not
of the total water as in the other comparisons so that the diffgrences obtained
cannot be due to a larger change in a small part of the system. This virtually uniform concentration of the salts in the swollen wood structure and in
the bulk water is not in contradiction to Donnan membrane equilibrium. Neale,
has estimated the ionization constant of cellulose to be of the order of
2 x 10-14 . This extremely low ionization constant, together with high salt
concentrations, would give an unmeasurable concentration difference.
Shrinkage measurements were at first made on salt-treated sections that were
dried in the oven at 115° C. without first drying at roam temperature over
aln connection with some earlier electrical conductivity measurements by the
author (J. Phys. Chem. 36, 312, 1932) it was concluded that dilute salt
solutions do not enter the cell wall because of the reduced conductivity at
the fiber-saturation point. This explanation of the experimental facts is
undoubtedly wrong. The results can still be explained, however, on the
basis of the salt solution dispersion not being sufficiently continuous to
give the full conductivity effect. No change in the calculations or other
conclusions drawn in the paper need be made because of this correction.
Neale. J. Textile Inst. gc, T373 (1929).
Rept. No. 1156
-6-
P 2 0 5 . These sections gave smaller shrinkages than those given in ,Table 2 for
the same salts. This was especially noticeable for sections treated with
salts, the solubility of which increased greatly with an increase-An temperature. Apparently the concentration of the solution in the gross capillary
structure increased as water was evaporated and the salt diffused to a
greater extent into the swelling structure because of the increased solubility.
Shrinkage and Moisture Content at Different
Relative Vapor Pressures
Salt-treated sections were dried in humidity chambers to obtain the shrinkage
and moisture content at different relative vapor pressures. Air was drawn
through towers containing saturated salt solutions with an excess of salt
present. The towers contained about an inch of mercury on the bottom. The
air supply tubes dipped below the mercury surface to prevent the tubes from
becoming clogged with salt crystals. The air from the towers was then bubbled
through open dishes of the saturated-salt solutions placed in the bottom of
vacuum desiccators which served as the humidity chambers. Air was drawn
through the system under vacuum of about half an atmosphere at the approximate
rate of 10 liters per hour. Under these conditions the theoretical relative
vapor pressures in equilibrium with the saturated salt solutions were virtually attained. The temperature of the desiccators was not automatically
controlled as the change in relative vapor pressure with moderate changes in
temperature was small for the salts used. Measurements were made with barium
chloride, sodium chloride, manganese chloride, magnesium chloride, and lithium
chloride in the humidifiers. The equilibrium relative vapor pressures for each
of these salts are given in Table 1, column 5.
The dimensions and weights of the conditioning sections were determined at
varying intervals to determine the shrinkage and moisture content. The salttreated sections in general required longer to attain equilibrium than the
controls. Several weeks' time was required to insure complete equilibrium.
Difficulty was encountered in making the moisture determinations on sections
that had been treated with saturated salt solutions because of the flaking
off of surface-deposited salt while determining the dimensions. It was
necessary when the salt concentration was high to use two matched sections,
one for shrinkage data and the other, which did not require handling, for
moisture content data.
The percentage volumetric shrinkage is plotted for a number of different
salts and two woods in Figures 3, 4, 6, and 8 against the relative vapor
pressures and in Figures 3, 5, and 7 against the moisture content. In all
cases the shrinkage begins when the relative vapor pressure under which the
section is being dried is slightly less than the equilibrium relative vapor
pressure of water over the salt solution at the concentration attained when
evaporation has proceeded to the fiber-saturation point.
Rept. No. 1156
-7-
Figure 3 gives the data obtained by drying sections treated with different
concentrations of sodium chloride. The points for sections treated with 36,
25, and 15 grams per 100 grams of water all fall on the same curve. The
shrinkage is slightly greater for 5 grams per 100 grams of water and considerably greater for 2 grams per 100 grams of water. When the sections are
slowly dried the concentration of the unsaturated solutions increases in the
void structure and evidently diffuses into the swelling structure. In the
white pine the sorbed water occupies 0.137 of the total void volume so that
when the moisture content is reduced to the fiber-saturation point, the concentration is increased 7.3 times. This would cause saturation in all but
the most dilute solution, and in the section treated with a solution containing 5 grams per 100 grams of water the concentration would just slightly
exceed that required to give saturation. Under these conditions any slight
deposition of salt in the grosser structure would prevent complete saturation
of the swelling structure and account for the slightly greater shrinkage.
According to the relative vapor pressure at which shrinkage begins for this
specimen the effective concentration is 30 grams per 100 grams of water instead of 36.5 as calculated for complete diffusion into the swelling structure. The former concentration gives a calculated oven-dry shrinkage of
9.2 percent which is practically identical with the measured value. For the
most dilute solution the effective concentration, according to the relative
vapor pressure at which shrinkage begins, is 10 grams per 100 grams of
water instead of 14.6 as calculated for complete diffusion into the fine
structure. The former gives a calculated oven-dry shrinkage , of 10.05 percent. This again agrees nicely with the actual measured value. The maximum
antishrink effect can thus be obtained at a concentration slightly in excess
of that which will give saturation upon evaporation to the fiber-saturation
point.
Figures 4, 6, and 8 give the volumetric shrinkage-relative vapor pressure
curves for other salts. Figures 4 and 6 give data for sections of Sitka
spruce and white pine treated with quarter-saturated salt solutions and
Figure 8 gives data for sections of white pine treated with saturated salts
which also act as fire retardants. In the quarter-saturated solutions the
swelling is negligible, but upon drying to the fiber-saturation point swelling takes place as if the sections had been soaked in saturated solutions,
thus giving further evidence that the salts diffuse into the fine swelling
structure while drying. The shrinkages are calculated from the saturated
salt-swollen volume rather than the initial volume to make the results comparable with those given in Table 2. The shrinkages to the oven-dry condition shown in Figure 4 are nearly identical with those given in Table 2 for
sections treated with saturated solutions.
The volumetric shrinkage-moisture content data are plotted in Figures 3, 5,
and 7. In all cases a linear relationship exists between the shrinkage and
the moisture lost from the fiber-saturation point to oven dry. A similar
linear relationship exists between the tangential shrinkage and moisture content and the radial shrinkage and moisture content, thus showing that the
ratio of tangential to radial shrinkage is constant over any moisture content range. Further, the ratio of the tangential to the radial shrinkage
Rept. 1156
-8-
is the same for salt-treated specimens, for which the swelling beyond the
water-soaked dimensions was small, as it is for the controls. In sections
treated with swelling salts the ratio is somewhat affected as would be expected, due to the abnormal tangential swelling.
The linear relationship obtained on the external-shrinkage basis might be expected to hold on a total shrinkage basis, over the sorption range if the adsorption compression effect is negligible. The fact that it holds for the
shrinkage manifested externally indicates that internal shrinkage in the
grosser capillary structure must be either proportional to the external
change or practically negligible. For the small specimens thus far tested
the following relationship seems to hold:
S=dxf
where S is the fractional shrinkage from the fiber-saturation point to, the
oven-dry condition, d is the density of the wood on a dry weight-green volume
basis, and f is the fiber-saturation point expressed as a fraction. The
relationship requires that the volume of the grosser capillaries does not
change with changes in moisture content. This internal swelling is being
studied more critically at the present time by a different means. If it is
found to be practically negligible according to the foregoing indications, it
will account for the linear relationship on the external dimension shrinkage
basis.
With the preceding relationship it is possible to calculate the moisture content of the salt-solution swollen specimens at the fiber-saturation point.
For example, if a section of Sitka spruce could be swollen to the volume obtained in saturated lithium chloride solution in water alone, the fiber-saturation point would be raised from 29 to 33.9 percent. The fractional volume
of the solution occupied by dissolved salt is, however, 0.282 (see Table 1,
column 4) so that (1 - 0.282) x 33.9 24.4 gives the percentage of water
held at the fiber-saturation point. This is in good agreement with the value
obtained in Figure 5 by extrapolating to a zero shrinkage. The values calculated for the sections treated in the same manner with potassium iodide and
potassium thiocyanate are 23.6 and 15.7 percent, respectively, which are also
in good agreement with the values obtained from Figure 5. According to this
the salts can in part displace the sorbed water. The balance between the
forces of attraction of wood for water, salt for water, and the resistance
to swelling of the wood beyond the normal green dimensions results in the
wood taking up a greater volume of salt than the volume of sorbed water displaced, thus causing swelling. From this it follows naturally that swelling
should be greatest with the most soluble salts.
From Figures 4 and 6 it is obvious that no antishrink benefit from the standpoint of constancy of dimensions can be obtained with any of the salts if the
prevailing relative humidities are always below the relative vapor pressures
in equilibrium with the saturated salt solutions for the rate of change of
shrinkage with changes in relative humidity is greater for all of the salttreated sections than for the control. From the standpoint of seasoning,
Rept. No. 1156
however, where the extent of degrade in the form of checking and warping of
the wood is a direct function of the total shrinkage, the antishrink effect
of the salt treatment may be of considerable value.
From Figures 4 and 6 it also appears that treatment with a very soluble salt
such as lithium chloride would solve the shrinking problem completely for
relative humidities below 11 percent are very rarely encountered. Unfortunately, however, sections treated with a saturated solution of lithium
chloride are always wet and sticky at humidities just slightly above 11 percent. They tend to take on more moisture from the air, but as the capillary
structure is already full of solution, dripping at the surface will occur
with the loss of salt. This will continue until a concentration is reached
that has a relative vapor pressure the same as the prevailing relative
humidity. Treatment with quarter-saturated solutions will give the same
antishrink effect, but the sections will not drip, unless the relative humidity is above 78 percent, which is the relative vapor pressure for the concentration. Similarly, a section treated with quarter-saturated magnesium
chloride will not drip unless the relative humidity is above 82 percent.
Although the sections treated with quarter-saturated salts will not lose
salt as readily as those treated with saturated salt solutions, they will be
wet and sticky above the equilibrium relative humidities.
Summary
1. The increased swelling of water-swollen wood in saturated salt solutions
was measured.
(a) The order of increasing volumetric swelling was practically the
same as the order of increasing solubility of the salts, increasing fractional volume of the salt in solution, increasing surface tension of the salt
solutions, and decreasing relative vapor pressure over the salt solutions.
(b) The pH of the solution had practically a negligible effect upon
the swelling, although the wood swelled in alkalies and not in acids.
(c) Swelling equilibrium was attained more slowly when dry wood was
placed directly in the saturated salt solutions than when previously swollen
in water.
2. The shrinkage and moisture content obtained on drying salt-treated sections at different relative humidities were measured.
(a) Shrinkage started when the relative humidity under which the sections were dried was less than the relative vapor pressure over the treating
solution at the concentration attained when evaporation had proceeded to the
fiber-saturation point.
Rept. No. 1156
-10-
(b) The shrinkage occurring on oven drying of the salt-treated sections was less than that of the controls because of the taking up of salt by
the structure within which swelling took place. This shrinkage was also
calculated from the fractional solution volume occupied by dry salt by
assuming that the salt concentration was the same inside and outside of the
swelling structure. Other evidence was also obtained to show that the
sorbed water virtually attained the same salt concentration as the free
water.
(c) The shrinkage of all the sections was directly proportional to
the moisture lost. The salt was shown to partly displace the sorbed water
from the wood. Those salts which tended to displace the most sorbed water
caused the greatest swelling.
.3-22
Rept. No. 1156
-11-
0
0 •-4
a)
5, 4d,
M •4
0
0
a
•
o
O0
CM
O
(V Is
0 co
"4 I
.-I
r+ a)
0 4.
al
c0
se
vo.
04.
4.
0
0)
0
1-1
a)
R.
CS
O
O
H
del0*
shok
el.so
sW
Or
Po
sp•
Nri
.0
0 \ 0 kJ) te\000 0 0
ONCU0Jr400000
•
0
-p
0
H
b() 03
•r4
W4
1.-H6060 Cll CU 0-4-0frNISO aclac H HO
• • • • • • • • •
ONLC1I1\r4 MD CU CU MI
WC:1\1,-H hr\r4H r4r4
• • • • • • • • •
H 8-48 . 1
C\ I
1:11
H +3
00
0
•
•
•
•
•
•
•
$4
s.
v.
N0100000000
CVCV0000000
•
•
•
•
•
•
•
•
•
•
.
\DA" VO X0
0.10.4' 0
011\14-Ve\ •-11-1 0 Is-0
• • • • •
• • •
r-1
14\1"-H ONO al CU CM Pr\
CT\ rr\I-1
HH
• • • • • • • • •
8-I 8-1 r-1
E-•
.A
4.11
0*
• CO
CO M
41.1 H
P
,o
sd E.
>,.
+3 40
4-1 3-( 0 0 60 0 0
0
+-I 0 +3 0
-•41.r4 cd •r4
Q
/4 _0 M M
H
r4
H otT
03 0
• • • •
14'
0
0
C) TS
01 I>
0000.400
044 H
Ft 0 41 r0
-1 0 03 0
0 CO 0
CO CO
.14
F-4
oo
so
O. wo
ow es OW mm
o• wd
ss
kJ) \-i• 0 I's:*r-6010
me°
H rev)) \O now:,
• • . • • •
•
•
060..1- MM -1- 14-.4- 4-
1--
O
O
M_1-
cu
arN
4. 43
•
s•
oo
oo
oo
sh•
qle
ms
s.
..
r-reNt-0000 arNnCV
rev411/40 crscr.t.mo o CV
• • • • • • • • •
1 - 0160 11-1- ∎13 111.0- ‘.0
O
ao
A .
H
al 0
0
rsCV
o
•n••
4. al
• cd
1+
CU 0 F-I
0.7
O CD 4.
0
r-1
•
00
00
O
co
cd.)
'-.0
F.
0
ro 0
a) 0
0
1.1- 1
co
CM 04%* H 60 ON0190
• • • • • • • • •
CM 0 I's- Pr \ 0 CU 0 1:11f`-
no ma\ aye° 00 r--r--
N8-1 01600'-0 N\000•
• • • • • • • •
0 C"CONLL"\-.1- MM
000000 r- r--r- r- r- r-
\ CI 0
0 4. 0
.1. 0 I-1
4J,
00
•
N
.•
al
O - al
• Ctl
0) 4-, al
cd
H'
0
r-i
0
43
co
7=1
cd
0
4-3
al
M
•
14 C.)
M
1-1
14 0
0
ra•
0
+30
cd
co CV
C
4.
0
a,
O
cd m ri•
4-,
0 0) 417.0
O
-1-,
0 P. cd
cry -c-1
0 0
a
0
SC4
$4
0
0
4-3
0)
0
P 4-I cd
•cd 0 4-4
ca E-• P. 4+
cd 0
ca
H ••
0 arN
qd
E-od 0 0 •
O M al
H
Er4
• •
0
••••1
1.
so vs
A- 0
Kw° sso
1111-1 8-1 C\1121
WNW r4 8-4 H H CU HO
• • • • • • • • •
cu r-to Cr.* KN.* H kr/
11\0NT-rap CM 8-4 CM M N
LCNCU H H r4 r4 0 0 0
• • • • • • • • •
O
0 a.1 4-, 4..
.4
4.
CO H .H
H
•v-1
••-I
44 90 a)
0 0 0 CO
H
H o o 0.0
•
•r1
0
1 4.
04-1
H
0 ++
P
.0
to cu
_M-
0P
PO ow .0 s• ae der
a
44
•-4
44 •-1J
a) co cd
Y••
413 44 ++
a)
0
a) P. 0
•ci
O
10
0 8.-1 0
0'
a)
•
H
O
ISO mm
• • • •
00/10
CM N-*0\1014-\ Is-LC\
0 WA' 11160 8-1 (TAD.*
• • • • • • • • •
• • • • • • • • •
$'4\8-1
000re\rdl 8-1\010.4• MCU tO
CU
LC\ 8-1 1-1 8-1
a000000
c)0013cd
0 0 0 tzt)
ss H o
r4
4-4 sir c•--1 to
A
gH 3•0 b TS a)
0
CU
a 0) a) M
4-3-p
0:10000
N (moral al
Hx xxx
H H1/4.0 \O.*
1--4
14
d
H
d rl
O 00
O
° J
O
EI H
H cd a3 M cd
0 „DO
0 irt
9-1
0‘4.
00
0H
rn
co
H
a)
or)
al
4.) 0
O Pm
co 0
co a)
Cu a) +3
0
• •
co
a)
0.
P.
0 0.1 LC\O CVO 060 t•r\
• • • • • • • • •
1 - 01e0 •-* ONIINVO to 01
_I- n.,;) sro 00 'co cr. ON a's 01
Cd.
cl-t
0 el
cd
5-1
0 1*--reu-1rcve0 CU \O
• • • • • • • • •
LC1010N1-1
0 I-1 CV
14" \ MCC\ P.-1,-00 ON
CO
cd
O
-1-
0
CU.4- CU re\
0.1 0.1 CU r-1
CU
14'0 14 0
c (0
H
r4r4HHHHQ
0 /4 r4 0 CO C-11,4
IA
000C.100 - HO
0 1-4 CUM 0 M 0.1 CVO
0 44 03 M A aiM th GS
ha isel ha *I 13.7 ha hsl NI bd
N•40XIMIM=1:403
a
0
CU
M
11190
H v-4 H
00000
000 0 0
4a 0
1-1 4-9
a,
O
m0.4
-14
td
El
A-I
0 1-1 •cl
&I 0 0
11-1 X 0
co 0
bl)
41 O
0 +3 I
.400
r40
0
o
0) co b.
sd
0
O
4a
0
43
O
Xi 40 0 11140 111.4. 0 LC\
\.0.4.
• • N•Lf•\ • • • • •
LC\ 0\ 01 MON ON 01010
r4
O
0
••n
14
4)
P4
Er\
0
0
.•e
+a
al 0
000noornon
• • • • • • • • •
ko1300101010'06-40
0 3'
N
NI°
co El
•-n •-1 1-1
00 00
42
0
al
O
Vr4o040WN-r-0
• • • • • • • • •
O
Cr%
$4
41)
tnaNtommaNmmo
pg
•p 4)
r-1 .0 4a
1-1 1d C1041 Fd H
00.i00 41
0 A-1 0)
03 03
.0
E.
43
reveo Lc1Lr1P-re.-1- o
ovco
• • • • • • • • •
\OW) 0\ 0\ CA 0\ 0 0 H
6--1 r-I 6-4
pew\
• •
r-I
6--1 6-4
1-4
.4a
0
co
150
u-Nr•-•revz r-co
010140 0101
• • • • • • • • •
Lc\
•-1 r-I
gl
t•-0\1110.11^V400.1LC‘0.1
0••-4\0\00‘10141010.1
• ••••••••
CU KNr-I r-1
r4
0
2
4)
.04-,
03
r-I
b134-1
4. A-1 0 0 al
O 4) CO 0)
•4 it
1--1
Lb
r-i
+ ce
0)
cd
0
0
re --10100
r4 01014 N-400\400
x4001111001150010
• • • • • • • • • 00.1%.0\00NCAM,..001
•••••••••
LINHH1-10.1
•H Pr\KNHH
1
4.
0
H
a)
4.
0
1-4
.
00 00 00 00 00 0 0 00
4-I
0 •0 0
0 0
4) r4 4. AD 4a A4
O
4.
C.)
ca
14
.0 $46-4 4a
O 0 bD 0 Cd
4) A-1 A-4 CO CO
0.0
[-•
as
g cd
0-4
0
co
\0\0 Li1150.4. 12\OCU040 \1:1 \OON•0
CV t0
•\ Cr• \• \ 0• ‘.0• \ 0• Cu• • •
• • 1-1
• r4• •.0
• Inn)
• • • 6-4•
MM MM A- LA* KNKN44Nrr\ Pr‘
MVO
• •
c‘.1
to• 04.
01/401,-al
nDfs•
•
•
• • • • •
\.0.4A.4 4. MM MM
0%40
• •
Cu al
a• 00 00 Oa .0 00 0. 00 .0
0 •0 0
4) 0
.0
-la .4..
0
.0
▪ 1)0 0 r-o d
.4 m cd •-I
.400 co 0
la 0
CO
t-1
cd
0
O
LC\
C.5
•
.4•• H•
•
\01/40
CV \-0 el
• • • • •
c."\ rr\ cc\ Pr\ cc
.1 00 MO OM 00 00
d.
OD ap 410 • • • • 00
mdm 4. Oa OM
.0 00
4 11
0
0 >.•
•6-4 .0
9-4
0
r-I
•
cd +3
0 0 ,--10 43 o
0
A-1 4) 0 A-I 030
(\I
0 0 E..
r-1 4-4 0 $44.3
$406-1 0 •0 cd
cd 0
co
›-•
010
r-re\ re\ trLO
40 N-40 6-4 rim 01 trie0
trau
C\10.10,1
r-I
r-1
6-1
0•
•
•
•
•
•
•
•
•
0
0)
cd
ED
O
O
IA a)
co la
-4
4.1,1 CU
C5
inrenr•-tno 0 .04 a. P•010 0`t lflrl U1L11
kr:ma al Cu 1,-0
r4r440Wr-.*
tUr.4
N 4P-MMI-41-1000
Or4 cu
• • • • • • • • •
• • • • • • • • •
r1 0-4 6-i0-1 r-1 0-1 H Hr-41-41-4Hrir4r4H
4A
0 0
FA
0
4) .
43.
Cl
OMO1Cs..1-40COW6-1
• • • • • • • • •
0100100.10-10101M
1.1
al
•
r44.20
44 r-1
.04)o
m0
•-1
044
clb 0
0 0
0 co
.-c
• ..
4)
cd
43
r-1
IA
r-1
.
.0 0.
a!
g r40
00
O
:10
44
+3
744
F-I
0
0
04)
.q
.0
•••
441 CO
0
0 • 03
0)47
r-t H 4)
00,0
100 Mr11/40 0 Mr-4 1•10-1
0 al 0 01LnEI1.4•4.
nye\
•-I• • FA
0 0• 0•
•
•
•
•
•
O
1
97 r-I 40
q 0•
>.F
4.3 cd
H q 0
CO 0 +. 0 0
4. Alc•
a) 0 0 Cu
Q c0 0 0
.0
•
.0
t-1
• •
cd
ri
00 .0
1---Lm0.4-10encucmre.
r4
00
00
00 00 WO
0.
tO•Lflai
OJ
1,-0
00
•
•
•
•
•
•
•
•
UN* 0\40 0
ONO UNL.Ckal (V 0-1 "-I
0-1
.0 00 O. 00 00 WO 09
CV CU CV CV r4
CV
•-I 0-1 r-1 0-1r4r4 C..) 0-1
000000 4,-4Q
0.401:00 cOMO CO
N •-1 0 Al
14 = b4 rcl
.0
.0 00 00 P. PP 40
PP .0 .0
or-.4 CV 14N
res0 14 0
CO $4
m 0146-40cno6-•1 0
01-4081:10Z CV0.10 CV
b4 04 b4 04 b4 Erl NI W 04 M
-1-
0
0
r4 r4
Or-I 0
• cd
0
F300
0,40
f44•3$4
4409-4
0
4) 0 4)
43 +a
Ce00
•-;
O 0
O El 0
r40r4
Id N cd
044
44%*
NON
r-11.4 r4
000
000
2.5
11:co
o
to E.
2. 0
a,
c.o
0
•o „,
•
<
11,
o• coct
Po
o
1. 5
1. 0
o"
CD
o
CD OQ
.5
0 0
0
0
0
1
2
0
3
4
5
6
7
8
pH
9 10 11 12 13 14
Figure 1. --Swelling-pH relationship for white pine.
Z M 2 2 40 8 F
2.4
2.2
2.0
`')
0 c
0 co
rs
1.8
1.6
0
b0 0
• ,4
1.4
1:1
a.)
1.2
u
u
1.0
0
.8
TD"
.6
fd
.4
.2
0
Concentration of salt (gms. per cc. )
Figure 2. --ffect of concentration of salts upon
the swelling of Sitka spruce.
Z 22401 F
0 0
N N
x
an
04
0
m
0
0
P4
6
0
fa.
6
cd
Z z
•
•
0
CO
ss)
ul
n7 M
0
•
-1a)
PI
0
;.•
0 04
k
0
04
cr.I
•
0
>
0
o
0
r.4
r.,
0
cd 4-n
• r4
a)
r4
tm
00
o
4a'
1)
..-▪ '+4
N
(
00
U)
•8
'0 Ti
0r9
a;
a) c.)
k g
0 0
-1-J U
4-1
•••n••• •ri
(I)
• r-I
id
k
o -t
c.) a)
co
co
a)
bA
b.0
(i)
0
0
a) ▪4cvs-'
4-4
r
cc) 1:1
(I)
-
g -Im'
•••1 w
k (1)
X k
01 +4
I (1)
(I) 0
P-I *1.4
a.)
O
CI)
CO)
P14
fil
Q.)
4-)
cl) • n••
" cd
z Z
-
rd
En
Q
...-..
cn
?id
cn
O
0
o
g
c0c3
M
O
.14
4.4
ci
cd
GI 0
k
•
k 4
Cli
k )-4
0 0 0
la
t . •-k •,-•
s
01 5
F14 -8
4a) "..4
.,_,
(in
En
a)
110 -0
•
a)
tti 0
t7:4
4a)
N3 a)
O "'a'
o
U
a)
.
0
4..)
o u)
.1-1
0
, a)
.—I I-,
......
I a) c
1 )10
. cis 0
..-1
ce) ... Ti
0 co0
cl.) •,-1
k k
0 4 4•4
01) En 0
.•,4
r•Z4
Volumetric shrinkage (percent)
..'Z'
c.)
.0
in
01 ni
Volumetric shrinkage (percent)
a
•
No salt
4D LiC1
BaC1
e!) MnC1
0 Ki
n NaC1
•
10
2
MgC1 20 KCNS
15
20
25
Moisture content (percent)
Figure 5. --Moisture content-shrinkage relationship
for Sitka spruce treated with different quartersaturated salt solutions.
Z M 22409 F
N
Ir
•41
N
Volumetric shrinkage (percent)
In1
O No salt
•
MnC1
O BaC1
e
MgC12
2
3 LiC1
Q NaC1
10
15
2
20
Figure 7. --Moisture content-shrinkage relationship for white pine treated with different
quarter-saturated salt solutions.
Z M 22402 F
Volumetric shrinkage (percent)