Forest Products Laboratory resin-treated, laminated, compressed wood (Compreg)


Material Information

Forest Products Laboratory resin-treated, laminated, compressed wood (Compreg)
Physical Description:
Mixed Material
Stamm, Alfred J ( Alfred Joaquim ), b. 1897
Seborg, R. M
United States -- Forest Service
Forest Products Laboratory (U.S.)
University of Wisconsin
U.S. Dept. of Agriculture, Forest Service, Forest Products Laboratory ( Madison, Wis )
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 29359614
oclc - 78089997
System ID:

Full Text

IRvised July 1944


No. 1381

Madison, Wisconsin
I& Ceepmtlim with the University of VWiteea






A previous report in this series, (mimeograph 1380, entitled "Forest
Products Laboratory Resin-Treated Wood (Impreg)" described the most effec-
tive treatment thus far tested by the Forest Products Laboratory for per-
manently reducing the swelling and shrinking of wood, It consists of the
formation within the cell-wall structure of a phenol-formaldehyde resin
after the wood has been treated with an aqueous solution of a completely
water-soluble, virtually unpolymerized phenol-formaldehyde mix.

ihe treatment imparts a number of other important properties to the
wood. One of these, not discussed in the previous report, is the wood's
plasticity at polymerization temperatures, prior to the setting of the resin.
Because of this plasticizing action of the resin-forming constituents, the
wood can be compressed under considerably lower pressures than dry untreated

For example, treated spruce, cottonwood, and aspen veneer dried to a
moisture content of about 6 percent under the conditions described in mimeo-
graph 1380 but not cured, will compress when subjected to a pressure of only
250 pounds per square inch at 300 F. to about half the original thickness
and a specific gravity of about 1. A few preliminary tests indicate that
redwood will compress to about the same degree under a pressure of only 200
pounds per square inch. The dry untreated veneers in contrast will compress
only 5 to 10 percent under the same conditions. Resin-treated sweet, black,
and tupelo --um and yellow-poplar require somewhat higher pressures, 300 to
400 pounds per square inch, to be compressed to half the original thickness.
Woods of higher specific gravity, such as birch and maple, cannot be com-
pressed to half their original thickness under any pressure.

Under a pressure of 1,000 to 1,200 pounds per square inch most of the
treated veneers compress to specific gravities, between 1.3 and 1.4. In

!This mimeograph is one of a series of progress reports issued by the Forest
Products Laboratory to aid the Nation's war program. Results here reported
are often preliminary and may be revised as additional data become

.. ITo. 1381 -1-

doing so, woods like spruce, cottonwood, and aspen compress to about one-
third of the original thickness. Th. denser, harder woods naturally are
reduced less in thickness. With pressures of 1,000 to 1,2'70 pounds per
square inch the treated woods are compressed almost to the maximum extent;
that is, the void volume approaches zero. This is evident from the fact
that the specific r:,'vity of wood substance is about 1.46 and that of the
resin about 1.28. Dry, untreated wood will, in general, require pressures
of 2,000 to 5,000 pounds per square inch to be compressed to specific
gravities of 1.3 to 1.4.

The increased compressibility of veneer treated by the Forest Products
Laboratory method not only simplifies the manufacture of compreg, but it also
makes possible the simultaneous compression of resin-treated plies and their
assembly with either untreated veneer, or resin-treated veneer in which the
resin has been procured by the application of heat alone, without compression.
For example, resin-treated but uncured face plies of spruce can be compressed
to about half their original thickness and simultaneously assembled to a dry,
untreated spruce core at 250 pounds pressure per square inch with a resultant
compression of the core of only 5 to 10 percent. If the core plies were
similarly treated but the resin was set within their structure by the : tion of heat prior to assembly with the treated faces, the core would hardly
compress at all under 250 pounds per square inch pressure. If the core was
of untreated poplar, it would also be practically uncompressed under this
pressure. If the spruce has a dry weight-dry volume specific gravity of 0.4,
the poplar of 0.5, and the resin treatment increases the specific gravity of
the wood by 18 percent (30 percent increase in weight and 12 percent incrcasc
in volume), then the specific gravity of the core in the three cases will,
after assembly, vary from about 0.42 to 0.52 and the specific gravity of the
faces will be about 1.0. It is thus possible to cause a variation in the
specific gravity of the product of about twofold between the compressed and
the uncompressed plies. The product can be made with the higher specific
gravity plies either on the surface or in the interior, as desired. The only
restriction is that the structure be balanced to avoid warping when the
trLated compressed plies are combined with untreated plies.

Besides the possibility of varying the specific gravity of the product
in the thickness direction, it is possible to vary the specific gravity in
t'r.e length direction by pressing a wedcge-shaped pile of plies. Figure 1
shows one of a number of possible ways of laying un treated veneer to obtain
specimens with varying specific gravity from one end to the other, together
with the finish. d product. ;Wrhon pressing such unsymmetrically shaped piles
of plies, the pres-ure is concentrated almost entirely at the h^avy end. 'o
avoid amn:-lo to the press because of the uneve-r. londirng, several specimens
should be pressed at once with the heavy ends symmetrically arranged over the
plate. area. Specific gravities ranging from tha-t of the .uncompressed
treated wood to about 1.4 can be obtained.

Mimeo. Yo. 1381

Gluing and Pressing

Bonding Glues

When resin-treated veneer is highly compressed in making parallel-
laminated compreg, it is not necessary to use a bonding glue between the
plies, provided the resin content exceeds 30 percent on the basis of the
dry weight of the untreated wood.! When the resin content is below 30
percent, and when the veneer is cross banded or only partially compressed,
an additional bonding agent should be used to obtain optimum shearing
strengths. Hot-press spreading phenolic glue seems to be most satisfactory
for this purpose. Slightly less than normal spread is sufficient.

Compreg panels can be satisfactorily glued to each other or to
ordinary wood only after removing the surface glaze by sanding or machining.
If the panels are thick it is important to machine the surfaces very flat
to avoid locally thick glue lines.

Gluing can be done satisfactorily with a number of glues. Alkaline
catalyzed phenolic glues that set below the boiling point of water appear
most satisfactory and have been most extensively used3.

Most Favorable Moisture Content

Experience has shown that it is desirable, especially when making
thick specimens of highly compressed resin-treated wood, to dry the treated
plies under nonp6lymerizing conditions (see mimeograph 1380) to as low a
moisture content as is practical, that is, about 2 percent moisture content.
When hot-press glues are used, it is further desirable to redry the plies
that have been coated with glue for about an hour at 160 to 170 F. prior
to assembly of the plies. This procedure, it has been shown, practically
eliminates end checking of thick pressed products, which may be very
serious when the moisture content of the veneer is appreciable.

When resin-treated faces are being compressed and assembled with a
dry untreated core or treated procured oore in a single operation, it does
not seem to be necessary to have the faces at so low a moisture content
as 2 percent to avoid checking. In fact, it is undesirable to have them
so dry if loss in differential compression between the faces and core is
to be avoided. The moisture content of the faces at the time of assembly
can, however, be too high, resulting in washboarding of the surface or face
crazing when the product is taken hot from the press. The best compromise
2 *_
-See Forest Products Laboratory Mimeo. ITo. 1384, "Comparison of Commercial
Water-soluble Phenol-formaldehyde Resinoids for Wood Imprergnation," by
Horace T:. Burr, Assistant Chemist and Alfred J. Stammn, Principal Chemist.
-See Forest Products Laboratory Mimeo. Fo. 1346, "Gluing of Thin Compreg,"
by Herbert W. Eickner.

Mlimco. 14o. 1381


moisture content for the face plies prior to pressing has not been defi-
nitely determined as yet for the various species. A moisture content of 6
percent, however, seems to be satisfactory for at least some species, such
as cottonwood, aspen, _Lnd yellow-poplar.

Temperature and Time of Prtssing

Experience has shown that the higher the ,unt'erature of pressing, the
greater will be the tendency to check. This is due to embrittlcmcrnt of the
treating resin. The best results have been obtained by pressing at 285 to
300 F.

;,'I'r. heating from the press platens, the tiMe of heatinr will natu-
rally depend upon the thickness of plies between the. platens. If all the
heat came from the platens the time for pressing would vary as the square
of the thickness. In the case of the resin-forming mixes used, there is an
appreciable amount of exothermic heat resulting from the reaction within
the wood structure. As a result of this, the time required for setting the
resin in thick specimens is somewhat reduced because of the fact that the
internal temperature is built up more rapidly than by conduction alone.
Cases have, in fact, been recorded in which the center temperature, as
indicated by a thermocouple inserted at the center of thick assemblies
(2-1/2 inches in the compressed condition) of resin-treated plies, rose
800 E. above the platen temperature of 310 F. and actually caused a slight
charring. In making compreg blocks six inches thick, which were heated by
high frequency to temperatures above 250 F., the temperature rose suffi-
ciently becausekhe exothermic reaction to cause charring at the center of
the block. This is due to the fact that heat was evolved from the reaction
at the center of the wood more rapidly than it could be dissipated by
conduction. In making 2-1/2 inch thick compreg when the platens were held
at 2P5 F., the exothermic reaction was sufficiently slow that the generated
heat could be conducted away as rapidly as it was generated, thus avoiding
the undesirable building up of heat at the center. This is further reason
for avoiding temperatures appreciably above 285 F.

Because of the effect of the thickness of the material pressed upon
the pressing time when heating from the platens, the pressing time is pref-
erably expressed as the time that the center of the wood should be held at
the desired temperature. This can be estimated from the curinr: temperatu':c-
curing time-swelling curves of figure 2 for 17 parallel laminat.,d plies of
1/16-inch birch veneer. These had been treated with enough Bakelite
Resinoid XR5995 to give a potential resin content of 30 percent of the
weight of the dry untreated wood, then dried at 170 F. for 5 hours at a
relative humidity of 45 percent giving a moisture content of abmut 6 percent,
and pressed at 1,000 pounds per square inch. In each case it trok from 10
to 15 minutes to attain the desired temperature at the center and about 5
minutes to cool the panels to 200 F. at the center subsequent to curing.
The cooling was necessary to prevent immediate sprin,:back of panels pressed
under incomplete curing conditions. It further gives im-Tjroved surfaces on
all panels. The curves of figure 2 show, from the large swelling ,Ir.Id
springback occurring upon immersion in water, that the resin was not cured


Mimco. :[o. 1381

in any of the panels at 2350 F. At 260 F. the resin is partially cured only
under the longest curing time of 45 minutes. At 2856 F. it is completely
cured in about 20 minutes and at 3000 F. in about 10 minutes.

With the use of electrostatic heating equipment, the time required to
bring the wood to the polmerization temperature should be markedly reduced.
l'o data are as yet available on curing times by this method.

It has been found desirable to apply heat when possible before exert-
ing pressures great enough to compress the wood, as less stresses and
rupture of the structure seem to result under these conditions due to plastici-
zation of the wood. This procedure, however, cannot be followed in making
thick, comrrpressed material heated only by the platens. In suchacase it is
necessary to apply compressing pressures before the center of the wood is
plastic, to avoid setting the resin in the outer plies before they are com-
pressed. This difficulty can be largely avoided by preheating the plies or
by using high frequency heating.

Thick material should be cooled in the pres's until the center of the
wood is down to about 200 to 2200 F. before releasing the pres.iure, especially
when the moisture content of the wood is appreciably above 2 percent. This
procedure is necessary to avoid the formation of steam blisters, crazing
of the surface, and washboarding of the surface in woods with contrasty grain.
Surface crazing can, however, be avoided with most of the species tested by
using very dry treated veneer. It 'is advisable that the cooling step be
omitted only when making relatively thin panels, using quite dry treated
veneer of uniform textured woods, such as cottonwood, aspen, and yellow-
poplar or when the pressed surface is to be machined from thick blocks of
compreg made at moisture contents beliw 2 percent.


Moisture Absorption and Swelling

Compreg is far more resistant to moisture absorption from the liquid
phase than the corresponding uncompressed resin-treated wood. This is due
to the relative lack of mechanical voids and capillary structure in the com-
prezsed material. Thc final equilibrium adsorption of water from the vapor
phase, hcwevcr, is practically unaffected by compression, although the rate
of adsorption is considerably less for the compressed material. Similarly,
the rate of swelling of compreg is considerably less than that for impreg,
'.ut the former will swell to a greater degree, due to the fact that the
amount of fiber substance per unit dimension is increased.

Compreg made with less stabilizing resins will not only absorb con-
siderably more water and swell to a considerably greater extent but it will
permanently recover from its compressed state to an appreciable degree when



Small blocks of spruce comprcg- 7/8-inch long in the fiber direction
swelled only 3.6 nprcent in thickness after 50 days' immersion in water and
after 1 year they swelled but 5.4 percent. Speci:nmens of spruce compr .- 1
by 10 by 10 centimeters in size absorbed 0.5 percent of moisture by weight
in 1 day, 1.2 in 4 days, and 1.8 percent in 7 days of cc-lete
immersion. The German specifications allow a weight increase for laminated,
resin-treated, compress'ed wood prepared by their methods of 5.0, 7.0, and
8.0 percent, using specimens of the same size, and for the same ler.hs of
time. A group of highly compressed opectmeris of birchrc:,mpreg 7/8-inch long
in the fiber direction that were cured under different conditions absorb.; ,
on the average, 5.0 percent of water and swelled about 5 percent in thick-
ness after 30 dayst immersion in water (figure 2). Similar specimens
adsorbed about 4 percent of water vapor and swelled about 3.5 percent in
thickness when subjected to a relative humidity of 97 percent for 44 dO .
Similar specimens of maple compreg 7/8-inch lor- in the fiber direction
swelled 0.7 percent after 4 days' immersion in water, 1.51 percent after 11
days, 3.7 percent after 30 days, and 5.4 percent after 60 days.

Army Air Forces specification 1-5. 15065 of June 10, 1Iq2 for ccmpreg
allowed a maximum water absorption, after 24 hours' water immersion, of 6
perc, r.t by a specimen 3 inches cy 1 inch by 3/8 inch (1 inch in the fiber
direction). The now specificatie-, 17o. 15065-A, Irrch 15, 1944, aloxs but
2.5 percent water absorption. Compreg made from resin-treated veneer
according to the Forest Products Laboratorry procedure (Mimeo. 1380) wil
absorb less than 1 percent moisture under these conditions.

A few tests indicate that compreg serves as even a better moisture
barrier than the uncompressed rcsin-treated wood (sec minAeo 1380). T1-.
moisture transfusion through a panel under a relative humidity gradient is,
for most purposes, negligible.

Surface Finish

Compr,:r has a lustrous varnish-like finish when it is compressed
b,.tween .i. hlv polished platens. 't.c degree of luster diminishes with a
decrease in the polish of the mold, a decrease in th.e co:ipression of the
wood, and an increase in the amount of precurin;" of the plies prior to press-
ing. Cut arfaces of the compressed material in which t:e rcsWn within the
cell-wall structure was cured at th.c time of pressing can be sanded and
buffed to give fully as lustrous a finish as can bo obt incd with the
platens, T-.. wood is finished throughout the structure. Sanding and buffing
to give a smooth surface merely bri.-: out the finish. Articles manAfacturd
from resin-treated, compressed wood can thus be restored to original
finish when scratched or marred by merely sandi.-., and buffing.

Compreg made according to the Forest Products Laboratory method b -
tween polished metal platens has a surface hrdnoss and finish ac shwn
by tests made with a Sword surface hardnes-- tester, whic. easures a com-
bination of smoothness and hardr..'ss. The instrument, wh'ch is calibrat-- to
give an ermirical rair. of 100 for plate 1 sse, r vs lucs rnnwin from


-1 -

65 to 90 for different resin-treated, compressed wood specimens with dif-
ferent resin contents made under different degrees of compression. Ordinary
smooth spruce gave a value of 6. The latter with a good varnish finish
gave a value of 18.

Only a few tests have thus far been made on painting compreg with a
yellow lacquer and a yellow enamel used by the Army for painting insignia on
metal airplanes. One coat was sprayed on half of the surface of panels with
resin-treated, compressed faces of spruce and on untreated, uncompressed
spruce controls. The one coat gave a smooth finish on the treated panels,
but on the controls showed an obvious need for building up the finish.
Southern exposure out-of-door weathering tests after three years showed no
deleterious weathering of the film in any case. Some face checking of the
untreated controls occurred through the paint film, starting largely at the
exposed unpainted parts of the panels. As far as the tests go, it appears
that this type of lacquer or enamel will stand up satisfactorily on resin-
treated, compressed wood.

Strength Properties

The strength properties of compreg are, in general, appreciably
greater than those of normal untreated wood. The specific strength proper-
ties (strength per unit specific gravity), are, however, in all cases but
the compressive strength, less for compreg than for normal wood. The in-
creased strength is. primarily due to the compression of the wood. The
resin seems to be effective only in increasing the compressive-strength
properties and the shear. It further causes an appreciable decrease in the

Table 1 gives the strength values obtained on a panel consisting
of 16 parallel laminations of 1/16-inch rotary cut spruce veneer containing
about 35 percent of resin on the basis of the weight of the dry untreated
wood that had been compressed to 0.35 of the original thickness and an
average specific gravity of 1.32. No glue was used between the plies. The
data show that laminated, resin-treated, compressed wood has very high
strength properties. Because of the limited number of tests, these values
can be considered only as approximate properties.

Inasmuch as the data on properties are related to test methods, a
brief description of the tests is pertinent. Because of the small size of
the samples, and their thinness, the standard test methods are not appli-
cable without some modification.

Tension parallel to grain.--The specimen was approximately 14 inches
long with an end cross section 1 inch by approximately 0.35 inch (the
thickness of the panel) and a central cross section 3/8 by 3/16 inch. The
center 2-1/2 inches of the length of the specimen was of constant cross
section and the transition from the central cross section to the end cross
section was effected with a curve of 30-inch radius. A constant rate of
motion of the movable head of the testing machine of 0.025 inch per minute
Was used, and strain measurements over a 1-inch gage length were taken



during the early portion of th., test. This specimen is substandard, in.
that the tension teit recently developed calls for a s'.ci-mnn 2C-1/2 incns
long to provide a -,ore favorable filler radius.

C-rrc-sion parallel tc -rain.--A specimen 1-3/5 inc[.-s long (4 tiles
least ii:..ension) by 1 inch wide by approximately 0.35 inch (the thickness of
the panel) was u .d. A constant rate of movL-.:r.t of the movable head of t;:o
testing. machine of 0.004 inch per minute was used, and strain measurece.nts
over a 1/2-inch *".-ie length were taken iuring- the early portion of the test.

Static bcrviinr7.--The sc; .cimen used was 2 inches wide and sufficiently
lrrw to provide a ratio of span to depth of 14. Center loadr'.- was used,
with a rate of descent of the movable head of 0.018 inch per minute.

Shear pr.r'ill.i to :'rain.--The specimen used was 2-1/2 inches log., -
2 inches wide by the thicknes, of the panel, with a notch 1/2 by 3/4 inch
in one corner. The rate of descent of the movable I. a. was 0.015 inch per
minute. Thi, lucimen was an adaptation of the Forest Products Laboratory
st-indard shear test specimeL, differing fro:, the standard onil,, in that the
thickness or width was approximately 0.35 inch instead of 2 inches.

The Johnson shear tests- were -%'ide on specimens 1 inch. wide .l
1/2 inch deep cut fro'-, another sfi.cimen 1 inch thick.

In conjunction with curing tests on resin-treated, parallel-
l-,mir.nated, compressed birch bonded with phenolic film and made under the
conditions given on page 4, measurements were made of the modulus of rupture
Sand the modulus of elasticity in static bendir.-, and the shear parallel to
the grain and across the plies, by the Forest Products Laboratory method.
F",r these tests, 45 different specimens cured under different conditions were
used. '-.e average, the maximum, and the minimum values for the modulus of
rupture were 40,300, 47,f)0, and 36,500 pounds per square inch. The corre-
sponding modulus of elasticity values were 3.64, 3.92, and 3..J million
pounds per square inch, and the ccm-rarable maximum shearing str-.r. :th. values
were 4,000, 4,500, and 3,7C0 pounds per square inch.

Shear values are highly dependent u'-.r. the method. Values taken
from different sources hence should not be c':- ,ared unless the method and
the size of the specimens are identical. For example, the standard Forcst
Products Laboratory shear-test method for testing t. joints of glue blocks
gave about double the values on compreg that were obtained by the Laboratory's
shear-test method for normal wood when shearing in the plane of the plies.

The shear stren:rth of,- par llel to the grain is considerably
greater in the dir.-ction in which the surface of failure is parallel to the
direction of rncompression than it is when the surface of failure is at ri :-.t
ar.-les to the direction of compression, even when the bond ',.tween the plies
is sufficiently strong to give 100 percent wood failure. It has been shown
from tests on solid blocks of wood compressed in either the radial or

--escribed in Johnson's Materials of Construction, by WJithey and Aston, 7th
ed., John Wiley & Sons, 1930, p. 61.


tangential structural directions of the wood that the structural direction
of the wood plays only a minor part in the difference. The difference secms
to be due primarily to variations in the structure in the direction of com-
pression and at right angles to the direction caused by the compression.
The average shear strength parallel to the grain and in the direction in
which the compression was applied for 24 plies of laminated, resin-treated,
compressed sweetgum specimens that were bonded with phenolic film, was as
determined by the Forest Products Laboratory method, 3,200 pounds per square
inch. Th-.: corresponding value for the shear at right angles to the direction
of compression was 1,200 pounds per square inch. Similar averages for six
specimens of cottonwood were 2,600 pounds per square inch in the direction of
compression and 1,500 pounds at right angles to the direction of compression.
Only the shear strength, in the direction in which the plane of rupture is
parallel to the direction of compression, is appreciably increased over that
for the normal wood.

The shear strength in the plane of the plies is highly dependent upon
the nature of the bonding material, especially when the dense, stronger woods
like birchi that cannot be greatly compressed are used. In the case of
parallel laminated highly compressible woods like spruce, 100 percent wood
failure is obtained with the treating resin exuding from the plies serves
entirely as the bonding medium. In the case of sweetgum and birch, glue failure
sometimes occurs, indicating that the bond may not be as strong as the wood.
When phenolic film is used as the bonding medium for resin-treated, parallel-
laminated sweetgum, 100 percent wood failure is obtained in the shear tests.
In the case of birch the failure is largely glue failure, indicating that the
wood is stronger than the paper lamina containing the resin glue. When hot-
press phenolic glues of the spreading type are used, 100 percent wood failure
is obtained even with birch. For example, birch compreg that was bonded with
phenolic film gave shear strength values in the plane of the plies in a
series of 45 tests of only 900 to 1,600 pounds per square inch. In all cases
glue failure predominated. When a hot-press phenolic spreading glue was
used, 10 specimens gave shear strengths averaging 2,000 pounds per square
inch, and complete wood failure resulted.

ir, mrc:dulus of rupture in bonding and the modulus of elasticity are
practically unaffected by curing conditions above a minimum threshold value-.
T., tou-.nss, however, is greatly affected by the condition of cure, over-
cure causing a decrease in toughness.

.-rre extensive strength data for birch compreg are given in the forth-
coming A17C-18 bulletin, "Design of Wood Aircraft Structures" to appear in
print soon.

5Sce Forest Products Laboratory Mimeo. :'o. 1383, "Effect of Resin Treatment
and Compression Upon the Properties of wood," by R. M. Seborg and
Alfred J. St- nm.-.
--See Forest Products Laboratory Mimeo. No. 1386, "Influence of manufacturingg
Variables on the Impact Resistance of Resin-treated Wood," by M. A. Millett,
1. M. Seborg. and A. J. Stamrm.



r Cr~ nomt uo t
It is more difficult to cut and ach than normal wood, 'ut
less difficult than metals. Special hardened saws and tools should be lower tool s-.':-ds than for nor-w,.! wood are desirable.

Because comnreg is more difficult to machine than treated uncom-
pressed and uncured word, the Forest Products Laboratory a method
for moldin-: prc-carved blanks rath. r than carving the final compr,-- blanks.
-.c process is now in use for moldln-- "club" motor test propellers and
aerial masts. Resin-tr ated veneer is glued uo into blanks with a phenolic
glue, such as Resinous Products PR14, or Bakelite cold-setting resin XCJ.'-31
with less than normal amount of catalyst XE2997, under conditions such that
the bonding glue sets only partially and the treatiL: rosin is unaffected
(temperatures below the point of water). The. blanks are then carved
to the desired width and shape but with a taickness 1-1/2 to 2-1/2 timos
that of the finis'.-d product. The carved blanks are then rE ated and com-
pressed in a split mold to the final desired specific gravity. A sli-ht
flash that can be readily machined off normally occurs at the parting line.

S-rie tL:p,.rimental work has been done i: cooperation with a Madison,
Wisconsin, co;:..:n; in molding airplane landi'.- wheels. in this case, the
plies have c,.:: .rr::.-ed with the grain of each ply at 45 degrees to that of
th.c next, rather than employing parallel laminations as is done with pro-
peller blanks. Blanks were glued up with a cold-setting phenolic glue with an
insufficient amount of catalyst present to set the resin completely at room
temperature, just as were the blanks for the molded airplane propellers.
-:.ose were turned to share, but with a greater thic-kness than that of the
f nishei product, and then pressed in a mold, consisting of a series of rings
and steps, to the final dimensions. Another -',ssible procedure would be to
punch discs of the proper sizes from the treated veneer, apply a hot-press
glue, and aseemb'le in the mold followed by pressir...


Co,'preg is finding its most extensive war use in airplane prri-llers,
motor testing "club" propellers, aerial masts, bearin.- plates, in: various
cr.nnectors that can be improved by the use of a material with ,-rater tensile,
c:-mpressive, and shear strer.;-ths than the nrorr.l word; also for various
tooling jigs. It likewise appears suitable for pulleys, silent crs, and
water-lubricated beari-.-s.

Pti-els with resin-treated compressed faces on an untreat-i or treated
and preciuri uncompressed core show promise for skin coveri--s of ai-r3irnes.
small boats, pontoons, and similar products. Preliminary tests shNow that
thin 3-ply plywood made from l/4Z-inch birch veneer ca, be steam bont almost
as easily as untreted plywood when the outer face is resin treated and com-
pressed. When both faces are treated and the outer face is com'"-essed and



the inner procured, it is more difficult to steam bend the material because
of the increased compressive strength of the inner ply, but it can be
accomplished with a radius of curvature of about 2 inches. Such panels show
promise for postwar use in house paneling, siding, and flooring.


The cost of resin-treated, compressed wood will be primarily dependent
upon the cost of the treating rein, the veneer, and the treating process,
all of which have been discussed in mimeograph No. 1380. The cost per unit
volume will be increased primarily because both the wood and the resin are
made to occupy a smaller volume under compression. The cost of compressing
and assembling in the form of flat panels will be but slightly more than
the cost of assembly of the resin-treated, uncompressed material. The cost
of production of resin-treated, compressed wood per pound will vary from
about 15 to 30 cents, depending upon the species of wood used and the thick-
ness of the panels.


Compreg is now manufactured by the following companies:

Camfield Hanufacturing Co., Grand Haven, Mich.
Farley-Loetscher Manufacturing Co., Dubuque, Iowa.
Formica Insulation Co., Cincinnati, Ohio.
Panelyte Division, St. Regis Paper Co., Trenton, N. J.
Par]kvood Corp., '-:'-efield, Mass.
Pluswood, Inc., Oshkosh, Wis.
The Rudolph Wurlitzer Co., DeXalb, Ill.

i- material is at present available only for war use.




1. .;-.o Antishrink Treatment of Wood with Synthetic Resin-fcr- ...
materials and Its Arrlication in V:, :i..; Superior Plywood, by
A. J. Stamm and R. i'. Seborg. Forest Products Laboratory
Mimeograph -Ri'13.

2. Resin-treated, Laminated, Compressed Wood, by A. J. Stamm and
R. 1. Seborg. Forest Products Laborat-r;,. ,1i.eograp.P El12Y(.

3. The Compression of W-od, by R. M. Seborg and A. J. Stamum. Forest
Products Laboratory Iimeo,:-rm:-,h 1258.

4. Effect of Resin Treatment and Ccrmrression upon the Fr-verties of
Wood, by R. M1. Seborg and A. J. Stamm. P.eport given at the
A,-rican Society of Mechanical Engineers meeting; in Louisville,
ritucky, October, 1941, Forest Products Laboratory iimeogrprh 1.7-3.

5. Forest Products Laboratory Resin-treated Wood (Impr.-;-), by
A. J. Stamm and R. M. Seborg. Forest Products Laboratory
Mimeograph 13.0.


TLbli i.--Stren;th values for resin-trected, parrllcl-laAinated

spruce co.ipresscd to a specific rpvity of 1,32

iTu.. 1-tr
of tests


Ger:.,an .
r--nC ci f -- t- :i ~.

Avorpje Lb. /

Tension parallel to the rain:
-i: ;.. tensile strcn th .......
[c'-ulus of elasticity ..........

Crops io p parallel to -;rsin:
a: i..u.: crushing strength......
oduli -i of ulasticity ..........

oc. ulus of rupture....... .....
Iolus of el sticity. .........

She ri 4 5prallel to the ;rain --
I n.u shuarin,_ stron,:tL
p erp.'uia r to plies:
o. ific F.P.L. method .......
Joson sin, Ile shear Liethod..
Johi-r'in double she, r ,.ethod..








-itr npt vlues taken fro.A the Ger,.an specifications for artificial
-:';in-treatod co..pressed wood (Kuaststoffe 30:58-62 j.1940]) are
:ir? ( for co:..oarrison.


M kdy35 r

Z M 39U5 F

Fig. 1. Means of laying up veneer to obtain
product with varying specific gravity from one
end to the other.

11111111 1 I 1 I II1
3 1262 08926 9376



% 40


L12 _____2 _____


220 240 260 280 300 3ZO 340 360

Fig. 2. Relation of the temperature at three
periods of cure to the combined swelling and re-
covery from compression of laminated, resin-treated,
compressed birch in the direction of compression
when immersed in water. Times are in terms of the
period that the center of the wood is held at the
designated temperature.