MECHANICAL PROPERTIES CI

LAMINATED MODIFIED WOOD

Information Reviewed and Reaffirmed 1965

!Revised 1952

No. 1039

1111111111

arroll11111iiiir r im'
1 14140114m1Waill
UNITED STATES DEPARTMENT OF AGRICULTURE
FOREST PRODUCT - ABORATORY 4
FOREST SERVICE
MADISON 5. WISCONSIN
erotical with the University of Whoeonain
 MECHANICAL PROPERTIES OF LAMINATED MODIFIED WOOD-

By

E. C. 0. ERICKSON, Engineer

Forest Products Laboratory, Forest Service

U. S. Department of Agriculture

Summary

This report summarizes the results of approximately 3,700 tests

made as a preliminary evaluation of some of the basic mechanical
properties of parallel-laminated and cross-laminated modified wood
prepared at the U. S. Forest Products Laboratory from yellow birch,
sweetgum, and Sitka spruce veneers. 'Test material consisted of
seventeen-ply panels of 1/16-inch rotary-cut veneer in the form of
normal laminated wood, impreg, semicompreg, compreg, and stay-
pak. Tension, compression, flexure, shear, toughness, impact,
water-absorption, and dimensional-stability tests were made on each
product.

The impreg, semicompreg, and compreg were comprised of veneers

impregnated with a phenolic resin, dried, and assembled under heat
and pressure. The resin treatment greatly improved the moisture
resistance and the compressive strengths of the woods, but also gave
them some of the brittle characteristics of the resin, as reflected
in reduced tensile, toughness, and impact values. In contrast, the
compressed heat-stabilized wood, staypak, containing no resin, had
increased strength and stiffness about in proportion to its increase
in density, and had about twice the toughness and impact strength

!Original report dated May 1947. Revised November 1952.

?Maintained at Madison, Wis. , in cooperation with the University of
Wisconsin.

Report No. 1639

of resin-treated compressed wood (compreg). However, since stay-
pak does not contain resin, the dimensional-stability of the wood was
not improved, but the normal tendency of wood to recover from com-
pression was practically eliminated. Average values resulting from
specified numbers of tests are presented in this report, together with
maximum and minimum values of each property. Data are also pre-
sented comparing the specific tensile, compressive, and flexural
strengths and the specific stiffnesses of the several types of material
in each of the three, species.

Comparative values of specific strength (based on strength-weight

ratios) indicated the marked advantage of the normal laminate over
the modified laminates in flexural strength and stiffness, and the
considerable advantage of the unimpregnated laminates over the resin-
treated materials in tension parallel to the grain. With respect to
compressive strength, resin impregnation provided a considerable
increase in the specific strength of parallel-laminated materials
parallel to the grain, and a most marked increase in compression
perpendicular to the grain (edgewise).

It is recognized that the resin adds considerably to the weight of the

materials, and that any quantity added beyond that required for dimen-
sional stability will obviously cause such impregnated materials to
show to much less advantage on a specific basis. Consequently, con-
clusions drawn from the limited number of comparative tests, and
particularly the comparisons on a specific basis, should be regarded
only as indicative because the total resin content of the treated materials
ranged from 37 to 58 percent, based on the weight of the dry untreated
wood, compared to 25 to 30 percent now used commercially and con-
sidered ample for optimum stability of treated products.

Introduction

Before the static tests covered by this report were begun early, in
1942, investigators at the Forest Products Laboratory had produced
several types of modified-wood and veneer combinations by impreg-
nating wood with the newly developed phenol-formaldehyde resins, by
compressing it, or by both impregnating and compressing it. Two of
these materials, a resin-treated uncompressed wood (impreg) and a
resin-treated compressed wood (compreg), showed appreciably greater
dimensional stability in the presence of moisture than did similar

Report 1639 -2-

commercial products then available. Several reports3—' 4 5— des-
cribe the treating method employed in obtaining the types with this
improved stability, as does a report on a modified form of highly
compressed wood containing no resin (staypak).7.—

Development panels had not been of sufficient size and quantity to

permit a critical comparison of the engineering properties of the
resin-treated laminated, compressed-wood types made according
to the improved Laboratory methods, to those made commercially
7
and reported in classified documents and engineering publications.—
Preliminary tests, however, had indicated that products made accor-
ding to Forest Products Laboratory methods from mechanically in-
ferior species would have mechanical properties equal to if not better
than those reported for commercial products made from maple and
beech, when compressed to the same specific gravity.

Consequently, the chief object of this investigation was to confirm

the results of the previous Laboratory development studies and to
provide at least a preliminary evaluation of the effect of such variables
as species, ply arrangement, treatment, and compression upon some
of the basic engineering properties of modified laminated wood. The
study was designed also to afford a comparison of the properties of
untreated wood to those of wood treated by the improved Laboratory
processes, as well as to those of carefully matched samples treated
by a commercial process under industrial operating conditions. Un-
foreseen resin restrictions prevented the carrying out of the commer-
cial process. However, three of four selected species were treated
by the Laboratory processes before the procurement of phenolic resins
became critical, and their treatment forms the basis of this report.

3
—"Forest Products Laboratory Resin-Treated Wood (Impreg)." Forest
Products Laboratory Report No. 1380.
4
—"Forest Products Laboratory Resin-Treated, Laminated, Compressed
Wood (Compreg)." Forest Products Laboratory Report No. 1381.
..52'Effect of Resin Treatment and Compression Upon the Properties of
Wood." Forest Products Laboratory Report No. 1383.
6"Heat-Stabilized Compression Wood (Staypak)." Forest Products
Laboratory Report No. 1580.
7
—"Superpressed Plywood," by R. K. Bernhard, T. D. Perry, and
E. G. Stern. Mechanical Engineering, Vol. 62, March 1940,
pp. 189-195.

Report No. 1639 -3-

A portion of this study, covering the properties of yellow birch compreg,
was first published in the restricted "Wood Aircraft Fabrication Manual"
in July 1942 by the Aeronautical Board. Later, the comparative prop-
erties of the several types of modified birch were published in the
revised ANC Bulletin No. 18, "Design of Wood Aircraft Structures,"
issued by the Army-Navy Civil Subcommittee on Wood Aircraft
Design Criteria. These earlier data are repeated here, together
with data on corresponding types of modified wood made from two
species of lesser density, sweetgum and,Sitka spruce.

The resin content of the treated materials reported herein ranged

from 37 to 58 percent, while only about 25 to 30 percent is now used
commercially and is considered ample for optimum stability of treated
products. A later (wartime) modification study! indicated that a re-
duction in the proportion of resin to that now used commercially will
provide a substantial increase in tensile and flexural strength, a modf,
erate loss in compressive strength, and essentially the same dimen-
sional stability.

Test Material

All material for test was obtained from logs received at the Laboratory
in a green condition.

Conversion

Two or more bolts, each approximately 52 inches in length, were cut

from each log, steamed, barked, and rotary-cut into 1/16-inch veneer.
The veneer was cut into 26-inch widths in such manner as to produce
sheets having clipped edges as nearly parallel to the grain as possible
throughout the 52-inch length. Each sheet was numbered consecutively
as cut. The veneer was then kiln dried to an average moisture content
of 10 percent, after which each sheet was cut to provide two 26- by
26-inch sheets, both bearing the same identification number.

Matching and Marking

Three sets of 26- by 26-inch test panels, each set comprising 18

matched panels of 17 plies each, were assembled from veneer of
each of the three species. Panels of a set were formed by assembling

!"Basic Properties of Yellow Birch Laminates Modified with Phenol and

Urea Resins." Forest Products Laboratory Report No. 1741.
Report No. 1639 -4-
consecutively plies of the 18 panels to correspond with consecutively
numbered sheets of veneer. As far as possible, the veneer in each
set was comprised of heartwood or sapwood taken from several bolts
of the same log. A few sets were made of both heartwood and sap-
wood from several logs. The sets were made up differently for each
species, as follows: birch, two sets of sapwood and one of heartwood;
sweetgum, two sets of sapwood and one of mixed heartwood and sap-
wood; and spruce, two sets of heartwood and one of mixed heartwood
and sapwood. Panels of a set were matched as previously described
in order to minimize the effect of variations of material within a bolt
or log, and to insure that, with respect to basic material, each panel
would be as nearly identical to every other panel in the same, set as
possible.

Following assembly, the plies of each panel were re-marked to denote

the location of each ply within the panel, the set number, the proposed
construction, the treatment, and the degree of compression. All identi-
fication marks were applied to the veneers on the face closest to the
bark side of the original bolt, and these faces were maintained uppermost
in laminating and in all subsequent flatwise tests.

Of the 54 panels of each species assembled in this way, 30, consisting

of 10 panels from each of the 3 sets, were selected for the tests covered
by this report. Of each 10 panels, 5 were designated for parallel-
laminated construction and 5 for cross-laminated (plywood) construction.
One panel from each group of 5 was processed into 1 of the following
types of material:

I. Normal laminate, unimpregnated and uncompressed.

II. Impreg, impregnated and uncompressed. (Some
unintended compression did occur.)
III. Semicompreg, impregnated and moderately compressed.
IV. Compreg, impregnated and highly compressed.
V. Staypak, unimpregnated and highly compressed.

There were thus three parallel-laminated and three cross-laminated

panels of each type for each species.

Impregnation

Veneers for the impregnated materials (impreg, semicompreg, and

compreg) were treated with a 50 percent aqueous solution of a water-
soluble phenolformaldehyde (resin) mixture by the cylinder treating

Report No. 1639 -5-

method. The treating time and pressure necessary for veneer approxi-
mately to double in weight by take-up of solution at room temperature
were 17 and 34 hours, respectively, for birch sapwood and heartwood
under a pressure of 50 pounds per square inch, 30 to 40 minutes for
sweetgum sapwood under 50 pounds per square inch, and 30 minutes
for spruce heartwood under 65 pounds per square inch. Following
treatment, the veneer was stored for 48 hours at room temperature,
under nondrying conditions, to allow the resin constituents in its
coarse capillary structure, to diffuse into its cell-wall structure.

Drying and Curing

All impregnated veneer was kiln dried at a temperature of 150° to

160° F. for about 3 hours at an initial relative humidity of 65 percent,
followed by 13 hours of dry heat to remove excess moisture. This
treatment resulted in a moisture content of about 2 percent and average
resin contents, on the basis of dry weight of treated veneer, of about
28 percent in the sweetgum and spruce and about 32 percent in the birch.
On the basis of the dry weight of the untreated veneer, the average resin
contents were 38, 39, and 47 percent for the sweetgum, spruce, and
birch, respectively.

Following kiln drying, the impregnated veneer that was to be made

into impreg was cured by heating for 1 hour at 295° F. in a hot press
at virtually zero pressure. This veneer was cured two sheets at a
time between three sheet-copper cauls, of which one was placed between
the sheets. After curing, this veneer was conditioned to approximate
equilibrium at 80° F. and 65 percent relative humidity prior to the
application of bonding resin.

All veneers designated for semicompreg and compreg were stored

after kiln drying at 80° F. and 30 percent relative humidity for a
short period prior to spreading with bonding resin.

Bonding Resin

Veneer for impreg, with the exception of the top ply of each panel,
was hand-spread over the upper side with a hot-press liquid phenolic-
resin glue, in amounts ranging from 9 grams of glue per square foot
on the spruce, and 11 grams on the sweetgum, to about 15 grams on
the birch. Thus, all impreg panels contained about 12 percent bonding

Report No. 1639 -6-

resin glue on the basis of the dry weight of untreated veneer. Hand-
spreading was used because the cured impreg veneer was found to be
too fragile to permit mechanical spreading without considerable
breakage.

Alternate plies of veneers for semicompreg and compreg were spread

on both sides with a hot-press liquid phenolic-resin glue in a mechanical
spreader. The average spread of glue per square foot of glue line was
about 7 grams. On the basis of dry weight of untreated veneer, this
amount of bonding resin was equivalent to a resin content of 8, 10,
and 13 percent in the birch, sweetgum, and spruce, respectively.

The normal laminates and staypak were assembled with a film type
of hot-press phenolic glue of about 5-1/2 grams per square foot of
glue line, which resulted in a 7, 7-1/2, and 10 percent bonding-resin
content in the birch, sweetgum, and spruce, respectively.

Conditioning

Following the application of bonding resin and subsequent drying at

room temperature, 'veneer for impreg was conditioned to 80° F. and
65 percent relative humidity, and veneers for semicompreg and com-
preg to 80° F. and 30 percent relative humidity, before pressing and
curing. All impregnated veneers were conditioned before pressing,
because conditioning of the final product would have taken much longer.
Different conditioning media were employed in order to bring the cured
and partially cured materials to approximately the same moisture
content. In this series, impreg, semicompreg, and compreg were
made up of impregnated veneer in approximate equilibrium with standard
12 percent moisture conditions (4 to 5 percent volatile content) in order
to obviate the adjustment of results of tests for differences in moisture
content. It has since been recognized that compreg should be made up
at moisture contents as low as 2 percent in order to produce material
that will not check.

Pressing

Panel assemblies consisting of 17 plies of 26- by 26-inch veneer were

pressed into flat panels in a 625-ton-capacity hydraulic press, equipped
with temperature and pressure controls. Conditions of pressure, platen
temperature,, and curing and cooling times, and the approximate amount
of compression of each type of material are shown in table 1.

Report No. 1639 -7-

It may be noted in the table that all compreg panels were subjected to
two separate curing periods. The original curing period of 35 to 38
minutes was concluded to have been insufficient to cure completely
the resin-forming constituents in these panels, because short-column
compression specimens taken from the birch compreg exhibited plastic
behavior and was readily deformed under load. Some specimens were
bent in a single curve, and some in an S-shaped curve with point of
inflection at midlength. Consequently, all compreg material was
reheated for a period of 1 hour, under the same conditions of tempera-
ture, pressure, and cooling time as employed originally, to assure
fulfillment of resin cure.

Preparation of Test Specimens

After pressing followed by cooling in accordance with the schedule

shown in table 1, each panel was trimmed to approximately 24 inches
square and was placed in a conditioning room, maintained at 75° F.
and 65 percent relative humidity, for lay-out and marking of test speci-
mens. Modulus-of-rigidity tests (nondamaging) were conducted on the
majority of trimmed panels before they were cut into specimens for the
other tests.

The location and direction of test specimens with respect to the trimmed
edges of the panels are indicated in the cutting diagrams, figure 1. In
general, four specimens of each type were taken from each panel to
represent each of the two principal grain directions, lengthwise and
crosswise (0° and 90°, or parallel and perpendicular, respectively,
to the grain of parallel-laminated panels, and to the face grain of the
cross-laminated panels). For tension tests, only two lengthwise and
three crosswise specimens were taken from each panel. Tensile
specimens were shaped as described in the appendix of this report.

In the absence of applicable specifications for the testing of modified

woods at the time this study was begun, the type and dimensions of
specimens used conformed to a modified version of American Society
for Testing Materials Specification (D143-52), "Standard Methods of
Testing Small Clear Specimens of Timber." Specimens were machined
with high-speed steel tools in such manner as to be virtually free from
tool marks or evidence of overheating, .and were not otherwise finished
prior to test.

Specimens for strength determinations were conditioned at 75° F. and

65 percent relative humidity for at least 1 week prior to test.

Report No. 1639 -8-

 Testing Procedure

Except as otherwise noted, tests were made on specimens with their

original or formed surfaces undisturbed. Physical tests were made
as described in the appendix of this report.

Results of Tests

Results of tests on seventeen-ply laminates of normal and modified

woods are summarized in tables 2 to 7, inclusive. Average absolute
values for a specified number of tests are reported, together with
maximum and minimum values. Properties of normal and modified
yellow birch, sweetgum, and Sitka spruce of parallel-laminated con-
struction are summarized separately in tables 2, 3, and 4, respectively;
and those of cross-laminated construction in tables 5, 6, and 7, re-
spectively.

In these tables, lengthwise and crosswise refer to the orientation of

the grain direction of the parallel laminates, and to that of the face
grain direction of the cross laminates with respect to the length of
the specimen. Edgewise refers to test load applied to the edges of
the laminations in a direction perpendicular to that of the laminating
pressure. Flatwise refers to test load applied to a surface of the
original material in the laminating pressure direction.

Specific-gravity data were obtained from measurements and weights

of the flexure specimens of each material taken immediately before
test. Specific-gravity values were based on weight and volume at
test, and were thus the density of the conditioned material in grams
per cubic centimeter. Moisture content at test is reported for the
normal laminates only, because the customary oven drying procedure
used in the determination of moisture content is not directly applicable
to modified materials containing resin volatiles. An approximation
of the moisture content of the type II impreg veneer was determined,
however, from the gain in weight during conditioning, following im-
pregnation and kiln drying. On such a basis the moisture content of
the impregnated materials ranged from 3.5 to 4.5 percent. Inasmuch
as all veneer was in approximate equilibrium at 80° F. and 65 percent
relative humidity before pressing, and since only the normal laminates
that were reconditioned before test lost any significant weight during
pressing, all materials were presumably in equilibrium with standard

Report No. 1639 -9 -

12 percent moisture conditions at time of test. Consequently, all
strength properties were determined and compared on the basis of
values at test, a basis representative of common use conditions.

Modulus-of-elasticity values in compression perpendicular to grain

(flatwise) are omitted in tables 2 to 7, inclusive, because in this
series of tests the limited specimen height made it necessary to
measure deformations between the heads of the testing machine, which
measurements were found to be erratic and subject to considerable
error. Comparative tests with deformation measurements made
between heads and by means of compressometers attached to speci-
mens of compreg in edgewise compression, permitted the following
approximations for birch and spruce compreg: The flatwise com-
pressive moduli for parallel-laminated and cross-laminated compreg
were, respectively, 80 and 40 percent of the corresponding edgewise
moduli perpendicular to grain.

Results of tests conducted on tension, compression, flexure, and

toughness specimens taken from the birch compreg panels prior to the
reheating period referred to in table 1, are also of interest. Modulus-
of-elasticity values for the undercured compreg were essentially the
same as the values obtained after further curing. However, the ul-
timate tensile strength and toughness were, respectively, 14 and
25 percent greater before reheating, whereas the modulus of rupture
in bending and the compressive strength parallel to grain were, re:-
spectively, 12 and 20 percent less before than after the longer curing
period. The data indicate that although the stiffness properties of
high-resin-content materials are practically independent of the degree
of cure, the strength properties are somewhat dependent on the degree
of cure at this temperature. Reheating the high-resin-content material
resulted in a lower toughness and tensile strength and a higher com-
pressive strength than would obtain at a lower degree of cure or from
fully cured materials of a lower resin content. While the chief object
of resin impregnation is to give the wood stability against swelling and
shrinking, other physical changes may result from resin treatment
and are discussed in previous Laboratory publications. 3, 4

In view of the marked difference in specific gravity between the modi-

fied materials, it is desirable to make comparisons both on the basis
of unit volume (absolute values) and on the basis of equal weight (specific
values). Consequently, strength weight ratios based on average absolute
values were also determined to provide specific strength and stiffness
values commonly used to compare the relative merits of structural

Report No. 1639 -10-

materials of different densities. Specific tensile, compressive, and
flexural properties of the parallel-laminated and of the cross-laminated
materials are presented in tables 8 and 9, respectively.

In these tables, values of specific ultimate strength and specific

modulus of elasticity in tension and compression represent the ratio
of average absolute values divided by the specific gravity (density).
In flexure, however, values of specific modulus of rupture are equiva-
lent to the average absolute strength divided. by density squared, and
those for specific modulus of elasticity are equal to Young' s modulus
divided by density cubed. Thus, the tables compare flexural proper-
ties from the point of view of equal weight, constant width, and variable
depth, and indicate that when materials of lower density are used,
beams can be made deeper for the same weight and the flexural rigidity
(EI) can be increased in greater proportion.

Some of the relationships between absolute properties and specific

gravity are also summarized graphically9 in figures 2 to 5, inclusive.
In each of these figures test data are shown plotted on logarithmic
paper, with the strength and stiffness as ordinates and specific gravity
as abscissas.

Figure 2 shows how the flexural strength and the stiffness of parallel-
laminated materials vary with specific gravity and modification.

Similar relationships between density and the tensile and compressive

properties of parallel-laminated materials are shown in figures 3 and
4, respectively. Figure 5 shows the relationship between density and
the tensile and compressive properties of cross-laminated material
as measured by the average of lengthwise and crosswise values.

The straight lines superimposed on the data at designated slopes

permit a ready comparison of the relative specific properties of the
several materials. In figures 3 to 5, inclusive, specific tensile
and compressive properties can be compared on the basis of first
power of density by comparing the intercepts on the Y axis (x = 1. 0)
of the plot of each property projected in a straight line whose slope
is 1:1 and whose intercept on the Y axis (sp. gr. = 1. 0) is the specific

2This method of presenting absolute values to facilitate comparisons of

specific values was developed by T. R. C. Wilson, former chief,
Division of Timber Mechanics, U. S. Forest Products Laboratory,
as reported by him in "Graphic Comparison of Specific Strengths,"
Civil Eng. , Feb. 1948, p. 47.

Report No. 1639 -11-

value of the property in question. Three slope lines are shown through
each group of flexural data in figure 2 to permit comparison of materials
for specific stiffness and flexural rigidity from several points of view, 10
as follows:

For a constant section width b and depth h adjusted to give constant

weight, EI = Ebh 3 /12. Then, if W is the weight per unit superficial
area and p is the density W = hp and h = Wip . Substituting value
of h in Eh 3 /12, EI = EW'/12p3 and criterion of merit in stiffness
becomes E/p 3 (corresponding to 3:1 slope). Similarly, flexural
strength as measured by the resisting moment Sbh 2 /6 becomes
SW 2 /6p 2 and criterion of merit becomes S/p 2 (slope, 2:1).

For section width b and depth h altered in same proportion to give

similar sections, let b = ab. Then EI = Enh 4 / 12, With weight per
foot W = nh 2 p , h = W 1 np , and El = EnW 2 /12n 2p 2 and criterion
of stiffness becomes E/p 2 (2:1 slope). Similarly, Sbh 2 /6 or Srxh3/6
becomes SnW 31 2/6n3"ip 3/2 and criterion of strength becomes
S/p 3 / 2 (slope 1.5 to 1).

For section width b variable and depth h constant, W = bp and b = W/p

Hence, EI = Ebh 3 /12 becomes Eh 3 W/12p and criterion of stiffness
equal to E/p (slope, 1: 1) . In strength, Sbh 2 /6p becomes Sh2W/6p
and criterion of merit becomes Sip (slope 1:1).

The inclined lines are shown in figure 2 to facilitate comparisons

on each of the three bases outlined previously. (Their slopes, but
not their positions, are significant.) Considering, for example,
stiffness with grain parallel to span (diagram at upper right in figure
2), it may be readily seen that lines at the 1:1 slope through the points
representing materials B1 and B5 intersect the unit vertical at approxi-
mately the same point. This shows that for beams of the same depth
and width adjusted to give the same weight, the two materials are
equal in specific stiffness. On the other hand, for lines at the 3:1
or the 2:1 slope, the intersections from BI are much higher on the
sheet than those from B5. This demonstrates that for beams with
width constant and depth adjusted to give the same weight (3:1 slope),
material 131 is some 3-3/4 times (6, 000 ÷ 1, 600) as high in specific
stiffness as B5. For beams of similar cross section and size adjusted
to give the same weight (2:1 slope), the ratio is nearly 2 (4, 400 ÷ 2,250).

.11"A Survey of Plastics from the Viewpoint of the Mechanical Engineer,"

by S. Livingston Smith. Institution of Mechanical Engineers Journal
and Proceedings, Vol. 152, pp. 29-43, June 1945.

Report No. 1639 -12-

In figures 2 to 4 the vertical scale is continuous and "grain parallel"
can be compared directly to "grain perpendicular." For example
(fig. 2), the modulus-of-rupture ratio for material SI is 12 4 . 0.78 =
15.4. Such comparisons as this obviously apply to any of the specific
strengths and to the absolute values as well.

Typical load-deflection curves for flatwise-flexure tests of parallel-

laminated modified birch materials up to maximum load at fracture
are shown in figure 6. Shown also are typical tensile and compressive
stress-strain diagrams of the same materials. All curves are based
on actual load-deformation data for individual specimens having
properties in close agreement with the average of the group.

Typical flexure and toughness failures of parallel-laminated modi-

fied yellow birch are pictured in figures 7 and 8. Typical Izod impact
failures of modified yellow birch and sweetgum are shown in figure 9.
The relatively brash failure surfaces of the impregnated materials,
as compared to the splintering type of failure of the unimpregnated
materials, are expressive of the embrittling effect of the resin.

Conclusions

Results of various tests to determine some of the basic properties of

modified yellow birch, sweetgum, and Sitka spruce laminates made
at the Forest Products Laboratory from seventeen-ply panels of 1/16-
inch rotary-cut veneer are presented. Certain properties of impreg,
semicompreg, compreg, and staypak, as determined by the tests, are
shown graphically for comparison with normal laminates on both abso-
lute and specific strength bases. Details of the preparation and lami-
nation of the modified wood are outlined, and in the appendix the testing
procedures used are described.

Water-absorption and swelling data show that the moisture resistance

of wood was greatly improved by resin impregnation, and other (impreg)
data presented show that the only substantial increase in strength
occurred in compression, notably across the grain. Resin treatment
improved the stiffness of all species (notably across the grain), and
to some extent improved the shear in the plane of the plies, but re-
duced the tensile, toughness, and impact strength of wood.

Report No. 1639 -13-

Resin impregnation plus compression produced a general increase in
all properties except toughness and impact strength. Compreg was
found to be much more brittle than normal wood, although less brittle
than impreg. All compressive properties were greatly improved,
whereas the tensile and flexural properties were increased about in
proportion to the degree of compression of the wood.

Compression to a high degree at elevated temperature without resin

impregnation (staypak) produced a general increase in most proper-
ties about in proportion to the increase in density. Wood modified
in this way had about the same toughness and impact strength as normal
wood. It was not, of course, moisture-resistant, a characteristic of
resin-treated wood, but in other respects it had the same general
characteristics as resin-treated compressed wood, plus the advantages
of increased toughness and about twice the impact strength. The stay-
pak exhibited some springback or recovery from compression, but
further research produced staypak that was practically free from
springback or the tendency to recover from compression when exposed
to swelling conditions.

Comparative results on the basis of equal weight (specific properties)

showed resin-impregnated wood to be superior to normal wood in
compression, but inferior to it in flexural strength and rigidity and in
tension parallel to grain. The superiority of impregnated wood in
specific compressive strength was most marked across the grain.
Moreover, when impregnated wood was also compressed, its com-
pressive strength across the grain was approximately twice that of
unimpregnated compressed wood of the same density. Further, it
should be emphasized that the materials containing more resin than
necessary for optimum stability showed to less advantage on a specific
basis because the excess resin was deposited in the coarse capillary
structure of the wood and only added weight without significant change
in volume. Unimpregnated, compressed wood was about comparable
to normal wood in specific tension and compression parallel to grain.

In view of the fact that specific properties of wood were, in general,

not improved by resin impregnation, except for compressive proper-
ties and stiffness in crosswise tension, it appears that such treatment
of wood is justified only where dimensional stability and other related
properties are sought, or when the absolute strength is of importance,
irrespective of an increase in weight.

Report No. 1639 -14-

 APPENDIX

Testing Procedure

Tension Parallel to Face Grain (1, tables 2 to 7, inclusive)

Tension specimens taken lengthwise were 1 inch wide by panel thickness

(t) by 24 inches long, shaped to have a central section 2-1/2 inches long
and 1/4 inch wide. The taper followed a 60-inch radius on each edge.

The end portions of all parallel-laminated lengthwise specimens of

normal Sitka spruce were reinforced With three inserts of 1/8-inch
birch veneer to prevent crushing by the gripping wedges during test.,
These inserts were glued into 3-inch deep saw kerfs cut in the plane
of the laminations• at the center and quarter points of the specimen
thickness with their grain at right angles to the grain direction of the
spruce.

Specimens were tested in a motor-driven, beam-and-poise type,

universal testing machine, operated at a no-load speed of 0.05 inch
per minute, using self-alining and self-tightening Templin grips.
Elongations were measured to the nearest 0.0001 inch, at equal
increments of load to failure with a 2-inch-gage-length, nonaveraging-
type (Tripolitis) extensometer. Specimens were gripped edgewise,
that is, grip pressure was applied to the edges of the laminations.
The extensometer was also applied to the edge surface.

Tension Perpendicular to Face Grain and Parallel to

Laminations (2, tables 2 to 7, inclusive)

Tension specimens taken crosswise were 1 inch wide by (t) by 16 inches

long, shaped to have a central section 2-1/2-inches long and 1/2 inch
wide for parallel-laminated panels and 1/4 inch wide for cross-laminated
panels, with radii of 30 and 20 inches, respectively. Specimens were
tested by the same method employed in conducting parallel-to-grain
tension tests.

Report No. 1639 -15-

Compression Parallel to Face Grain (3, tables 2 to 7, inclusive)
and Perpendicular to Face Grain and Parallel to Laminations
(4, tables 2 to 7, inclusive)

Specimens used for determination of stress at proportional limit and

modulus of elasticity were 1 inch by panel thickness (t) by 3-1/2 to
4 inches long for the normal laminates, impreg, and semicompreg
materials, and approximately 4t inches in length for maximum-strength
tests. Deformation data were obtained by means of a 2-inch-gage-length
Lamb' s roller compressometer attached to the edges of the specimens.

Compreg and staypak specimens 1 inch by (t) by 2 to 4 inches long

(approximately 6t) were used in tests to evaluate proportional-
limit and modulus properties, and specimens 1 by (t) by 1 to 2 inches
long (approximately 3t) for tests to determine maximum strength.
A Marten' s mirror compressometer of 1- and 2-inch gage length was
used to obtain deformation data. Load was applied to all compression
specimens tested through a spherical loading head in a beam-and-
poise-type machine, at a rate of motion of the movable head to pro-
portional limit calculated to produce a rate of fiber strain of 0.003
inch per inch of specimen length per minute. A rate of loading corres-
ponding to a rate of fiber strain of 0.006 inch per inch per minute
was used in obtaining maximum loads.

None of the compression specimens exceeded 4t inches in length for

maximum-strength tests or 6t inches in length for the determination
of elastic properties. Specimen lengths of 6t were chosen in the
latter case for all compressed materials in order to maintain a length
of approximately twice the gage length of the compressometer that
had 1-inch minimum gage length.

Compression (flatwi se) Perpendicular to Grain, and Plane of Lamina-

tions (5, tables 2 to 7, inclusive)

Specimens were 1 inch by 1 inch by panel thickness (t), except for

compreg and staypak, which consisted of two thicknesses of material,
each 1 inch square, placed one upon the other. Specimens were tested
in a special compression jig centered under a spherical loading head
in a beam-and-poise type of universal testing machine. Deformations
were measured between the fixed and movable heads of the testing
machine by means of two 0.0001-inch dial gages. The machine was
operated at no-load speeds ranging from 0.004 inch per minute while
taking load-deformation data, to 0.025 inch per minute for obtaining
maximum loads.

Report No. 16 39
-16-
Flexure flatwise), Face Grain Parallel to Span, and
Face Grain Perpendicular to Span (6 and 7, tables
2 to 7, inclusive)

Specimens 1 inch by depth (t) were tested as a simple beam with

center loading on spans ranging from 14t to 16t. The rate of
motion of the movable head of the testing machine was calculated
from the formula
2
ZL
N = where

N = speed of movable head of machine in inches per minute

Z = 0.0015, unit rate of fiber strain
L = span in inches
d = the depth of beam in inches.

Detailed procedure conformed closely to standards of the American

Society for Testing Materials Specification (D143-27). In general,
testing was not carried to complete failure, but was stopped when
the load had decreased to about one-half of the maximum.

Shear Parallel to Grain (8, tables 2, 3, and 4)

Notched specimens 2 inches by (t) by 2-1/2 inches, with shearing

surface 2 inches by (t), were tested in single shear across the lami-
nations as illustrated in figure 13 of American Society for Testing
Materials Specification (D143-52).

Specimens tested in a Johnson-type shear tool were 1 inch by (t) by

3 inches (two 1-inch by (t) shearing surfaces). Specimens were
broken parallel to grain by means of a compressive load applied to a
shear member and acting across the central portion of edgewise speci-
mens clamped at each end. Cylindrical double-shear specimens were
3/8 inch in diameter. Specimens were tested parallel to grain and
broken in planes parallel to the laminations in a three-plate jig by means
of tensile loading. Shear tests were conducted at rates of loading rang-
ing from 0.014 to 0.016 inch per minute, and in accordance with proce-
dures specified in Federal Specification (L-P-406b).

Report No. 1639 -17-

Modulus of Rigidity (9, tables 2, 3, 4, and 8,
tables 5, 6, and 7)

Tests were conducted on panels approximately 24 inches square by full

thickness of material, using the plate-shear method developed by the
Forest Products Laboratory for measuring the shearing moduli of
wood, as described in American Society for Testing Materials Specifi-
cation (D805-52). The normal laminated wood panels of each species
and the birch and spruce compreg panels were not tested in this manner
because these panels had already been cut into test specimens before
the plate-shear method had come into general use. Torsion tests,
however, were conducted on rectangular specimens of these materials
of width 3t by thickness (t) by 16 to 24 inches long, gripped flatwise
and with a detrusion-measuring device applied to their edges. Following
tests on these, with torque kept within the proportional limit, specimens
were cut to width of 2t and the test was repeated.

Toughness (10, tables 2, 3, and 4; and 9,

tables 5, 6, and 7)

Specimens 5/8 inch by (t) by 10 inches long, with face grain of material
parallel to length, were tested on the Forest Products Laboratory
toughness machine over an 8-inch span, with plane of laminations
parallel to direction of load.

act Izod t 11, tables and 4)

Specimens had the grain lengthwise and the notch in an original sur-
face. Some of the compreg and staypak specimens were less than
1/2 inch in thickness, but the dimensions from the base of the notch
to the opposite face were standard. Impact strengths were determined
as specified in American Society for Testing Materials Specification
(D256-56) Method A, in a 16-foot-pound-capacity, pendulum-type
impact machine.

Water Absorption (12, tables 2, 3, and 4)

Specimens were 1 inch by 3/8 inch by 3 inches with grain parallel

to the 1-inch dimension. One face was an original surface sanded,
and the other surfaces were machined. Specimens were heated for
24 hours at 122° F. , Cooled, weighed, and immersed in water at
room temperature for 24 hours, and the percentage increase in
weight during immersion was calculated.
Report No. 1639 - 18-
Dimensional Stability of Molded Thickness (t)
(13, tables Z, 3, and 4)

Equilibrium swelling and recovery from compression were determined

from specimens 1/8 inch by (t) by 2 inches long (grain parallel to the
1/8-inch dimension). Specimens were immersed in water at room
temperature until equilibrium moisture content was reached, and the
percentage increase in thickness (swelling plus recovery) was calcu-
lated. The specimens were then oven dried and measured, and percentage
recovery determined. The equilibrium swelling was then calculated.

1.-39
Report No. 1639 -19-
 :lb: I 11.4
14111
h
W.O.n
...N.. Sim.
NNW
RAS
Aar
t
tg I I
ty_ I
14 I R ag' 8.22
1– IFA1 1 1 1 1 III
oi
ill

g2g
.

• •I
Ca& CRC CC IV Hie

RRR CCP ERR

II SI 1 .11 1

C7.1"3.' ERR MM.

Lt..,
6.4 el .4

RRE Rio W WinW

44P
d
ha

ERE 2CR
O
gwit0
u-s0
NMN §,§,§,

7 000
wenn
worn, FIRE
FIRE WWW
NNW

I .
Vt.! 3 3g
g.tr-.=;.:1 iik I

2 41%
%Rr•
1-1!
,
1 1 :
In MOO 'I.GN
31! Iv, Ii
41 ;
I 1 . .11
0.Z. gas RE8
0 UN§
3 21: EH Z.%
rimil
P 24
n1
;
61 4 ; t
1
lim COO Er gm ono--tesrb0
nwo, Mt _65
I! E

.°1
= ;12.
.1• 1 -
1 •-

7.


i . 1•.-.,
1
. $,W W p. E. *6?. ..2, // 4i17IfiEli 2E/ R FF. F., •n
" -3e - 2
; 13
11
l / N ° 2 AV,4 " :.7x- r'n' zn to.-- 3

il 1 la L Z A .',..i3q US W; i€R /1 RUC; 10.1 f.KE ..: .,:s 1

01 i
431 -1 S' 1F:
I -iI a
SA4 WV:
i:A
ag E q m
,:4 .1
nr, u. ikam: EIV4 . EU 2 qi ! ;:2Z2&
k: .r.. d44""' W.r: '' W4.

illidlii . . » 5 2' 5 ... A !Irv. . . :3, '" ' i

II ! 7.- ' F, 6: 4, E. t XF!!/

...1
fli Y.5.
4 a,z /Eeli-,
PV:1-4 lfW LI
..;..n
" a:••• 2'°- ..,
Elfii 1
41^ 4
I
''
'Pi i 9i :i g: IIE
n'r.:4
.?..W
44
HS n',r; t/ ITM gE-70 ..•
q H
-14 " .71
nn nn
UN In
«11.'";-'
ill ''' ' '
Pi t Id i 8 .41/3!a.- EM .;..-.1
W. Et .a EE .i liWM g?,!, kk V.' :i V sl R 1
..,; '4k-''"
..r .–, 6 4
7: :."S'4
i,. . _ i 1
EN a! . ... 5 5 '"'"' 5 2, nn -- o' .:,., i
i ! .?.! g.' g. WA 3§..24! .1m Ell. El U6 .74; ifft
.. ---.; ..7. n;.1.: .14 1,11 '6 !Lq F.: ; R '1
" »8,,3 . .%: .;,,* .1
i f
•
0! 21 kt .f?..t4 TIR.- q UP: g/E 41
EA
4A
E.gr-1Z '15Vil
W .4A
Zif E '44 .: '" '7. '.: '4' 6.: NM UN
1.1 WI
I V

al
3 I 6 6 ri."; d.';';'.
13 1 ? ..z1 g A ag,P iR.5.1.2 1M HA fg R151 WA
a.- IA
• EIt
,.;, a ii t: :!::• 2 2 li
1.1
. L 5 5 2 5 ..m ., n. s .,-. ,,,
11 11E1

I
 kin! EAR !stag"' as t c
ag .7.„; VN.0
ga

IR,14 f21 2
•/. it :.:
2A:M
.nu-, • HE Ha §A2VA ' 2 .1"
aa

ARMS':
EN'^" 2A^
2R4ff
•
Or 84 gg " aa,a
UV": !UR gasag e4 § a
AAA
2 11E '4 '4
0.•

E§3A 29.2P/ 82 •
Eif aa
aa ag o$: NON •
"-‘

111 •
RR
if N1AA matip, kg r Z. 7i "11
ag
-••
22A8
••• 16,11 •
ARA% aifn 5R-ZA •.` 1
4d.rn
LO f2.1

§1E2. 2.9 Utv. W°. . . a

A.4
t
one-
dA aag

4 P_VA 3r?,-PL .282 g *7; skE2::1 2

mom earsom V, 10

---
RM 5C,J, aaa Y4 aprl, E ia
.a •

Hp Hp, 1§'"?`41
i•
2,14. 11A mma a 74 6: : 1 1!
I :A 14i
e gl gs VI 41]
alp R2.41R
a hi6q.-1; F.' T4 .21


1 1 Id ! .! 444! MA 44E las ING

: ... j. En !! !!!
••....••

3 44:i sal • -: 19!

4 1i . 1 —f
z1 .
4 ''' 64>. !r. !!! .. .A !! 1W ; IT ! !! mi- 1.1'0

4 1. i 1.-i t, 1., M q s E/ : tEE

...I A
6
Or:.5
'
tki H. ir.i...,,_ §IFA R: 5E4 R r ,,;P ; rri
.41 1 ! I'l l '7.n '.. n 9:1i.
MI..

Ei M i 1;1 10 . V• ‘A' ON V• 2 2Z2 ' Z

n7 OS
'11r3n'''i
i1 1 - ^^ alag2 dRE'f., ,..
; !!! riE !
1
11 1.1d -:,
: : I 1
.,
-:-.. ,. . .- WU
:l .. a..'
.,,,I

WI 1, Id , 2 RNA t.W Itif

4 41 i
1 1 - 3.4-. aA WI 1!"!
:7,1 6-,' !!
0 NA
kH....1
.......... ........... .................... .... 1
8 i1 f ..3:1 iAD!' g. IMS;, gt !! !!3/4 aalm FIE V v.: a A li 1 i
•.$:
11-! 1 1 ' .- .. r_ ... ........• ._ a
.„
E L. :61 - 4, - . ,
10 2.
CP 2 P1 i
..___..


31h7T:.: 535 H T," EE &,§

•

NEP 13;iF RE-3 22.

PrOM sap:
1E 3.9M Elt-q a
z2". otb d" 00

ERAI EEk 1.31V P.m

aaH
thd,"

REP 1:gA rRN

ESP; R1AR'q

MP!, lER,
an ti 46a 311 S5a,n
Mean
34P7.1: r a&

RS RAW? E EE EF1
P. 01

OA a
,S7 eV

Ea /2R1,171 §°-,§gs,g, WR a2
aid

33,12 ER §3
A.2 !!!17-.; 21

a04- EAE P;
E11.4 SS PANAF:
4ft"..n 46 a

2 1,01

.1 Its EA a§t HEIA ESP" 5

c.rw
I
E3F. kgpS a; Ellqg EEPA 2
3 al aa :42 v.
eo
1 2
ss
g
9.R=T mArn IT S
EEC-1
I ""
"
g 2 •
E.514 In/ aal EM Eng* R/M
.2.- a g ,1
g
ma RR2 9.24-qn
%21

I 1111 111! 111 111 11 ,— : 7.:::,_:

: 11 11
O 0 M M §§ 4.,
..' ..I §..
as . g.g. .:.: 3
1 i - -.4 4: I ..
z 7: ! 1 1 1 11 1- .4 1

i i g
14 . 1
3
1 1 I i 3
a H- - .
ti. 4- '- "..3-
3 0 ,
! 0
g V g211 11
it.A i-TA
:zzr: 141 b lid
11 4' ii? 11- 11 3 !!sdllil I
.] li;
N i r i CH Cl Nil! 1 411 1 il' E
r.3 .: . 4.

t p. t .4 11 41 ! 1 ii = t O l t 11E Ottx1 OttEi l.1

1 t :Ili zEis l 'El It429: 7iia22 Vb 1 'V ; "
I 2 ii llgit, ps i i0l ii12 IN i Offil tEllil I..; /..
il • 0
1 t
ws .W. A sa ..: a .;
4 A g .4
 °SW
go^ §§;' a Ate!

-1- .$, nA RIEIR a nn

"2'

IgV9: EbA SSA

" gFagil
AAA
PPS

. •

ag n kkwi Ate; RIM n 41 1

••

kkt
.e. n
aaak
,d,
fl fit" en

as

EfFi aim REW4 g

NtRt gf iR: SS"

"oA ak
aa
?, n
ad
§S2gg
Ma n
elate• t5 2ND asp.; A
7t4
aaa
AY,•J fit
aa
fF!
ad "

1154 Rif IfI PA

•••

1 Fa i; g2m UPC
°
! 1=1 ? RER tgffr': RE'.4 M.:42A
I 1- i 4 °
IP/

r-1
2RAZ; aig RE14;:

PR Mg' gaf gn MW,

.0..
Am Rtg as
aa
irk"

°
 Ealq EE PIE Rail; ElEgr. -rt
,T1d4" 144" RR

EIR ts
'Es
,

* 54' te: ••n;

•••%

3;11 asv!. Ali ma § 16f1;

,Aga
V
•••• ar:a

14
1.0* if IRPA guA; gp,

k ECU.' ilE EMNq i3A 7A 9

kt.g. 1E 31411 firA 9 2 4",

engem

01. 10

ILA EV EF-aVs: 81/A8 ei

.ca g.§ ak nz.6;

IF if 322 W4 EIRRq 6
L1-7,p; c4a A.;.;
g1
ES ff,' Zi• .a 8211.% I
ga .wga 4 a

•••1 1111,111

2 aaR ,71 VRt Mr: 1Wq

z VI
P.EB-4 UE 111 4:11. E
6 Aldipt

if
Eg5, 41
! f ;i
.1 :4
5. 2 2 1.11

113,Z g lava .a
tr•S

rm
• /El E,§SR ERE
!gal / 22
6A
.27 °
a
gg
U1' 1EgA EFATE: a
6 NI Pet "
&
1:1
2 2
1]..1
11-11 111i 111 111 11 1,111•11, 111.Ci 11 42
El A k kid Lri kdd aa I
n .;•,.1 ., -II


t.0 03,0 hr. N - '

I Fei ';1:44 40
...1rni to M .n••n Y'.
11/.

• Zsi:7 g . •.
-7

Id
N w
• ,
4 ▪ LO 91 g•-•
0.0
.11•• 0. g
le • CI, •
la I
t .7, 1 :2 •12, 142,' °:'013
t ' ' • 'Rtz N
F d
4 gA '9W Pg
1 1.71 a
I14 1
11 1 V, .47
.17 •NO• • n••-•
10.
1-1 N 47; CA •

I •1:
yF

1 '1.‘ 'I '2 2;

•
'=.2
KN 2
.1G

ytl

. 0;1
g
11 11' •-I •

i
! F. i m •••wo A 1,-. ...IC WZ, A
'° vi • Lr. va • nI's IV • ••
I I rt'e 11.0/ .401 -at A..•
1 ••
11.3 •2 1- as1.11 '1M RO 121
O NN n' :34 N' g
.
-
a . .
1
J 1 I: ,7 3 93C••0.--•• 'AI., R Ni. l a v.,,,
R...-. 4•
f I;
I 4 i
1 t' 1 I . I. N d
.I i E. 11-.; 1 .A, '9"'
9.:
23;
...1 •
".13411
& 01 r.:,
'13
..,„
,;;:-. ...g• 3. ti
.,
a .4 S 1 1 ' :tali:
a ... . 3.8 i
i i F i ° ...... . 0• .0. ,
.I.
KS 05 .-• n•••n I.,
1.-..... .1 0. Fs
1 v;
.1 I 0 7.i 0; ..I . '4 ; . Z ' r . .7
".' .-. ' ii,11.
• ll
.
1 P e: I
NW° .4,0 i
•e•- IS
.
'AV% ,0 1•••

.
.
ON, •0 :Z.? 10 v
. 1 !CAI
I
I
i
!
I
". N .4. . '4 NI 4N. • ." A
N 1

lil g I SI I In-
..1P I" .1..r, -....
• . IP
CO• 01 N• .°. * SE', ;0 I
•• ••
O N g 1 s' I 0 9,...1 ..1 v.
''''5, • N 00 •
P11
u ,
li I I
- - - - - - 3 1.1
A .
g '2:8 - ."8 '8`2 ft 1:.•
T
e.1.4 . . 7:11,
.4,1 SI
8.1 v. "c
.4,1 ..1,.
I NH NH NN NM 14/4 14 . 0 7,. 1n
i n
It: #
1 -;.' 1 i. I. 1 1 1 I 1 1 1 :. ' i 1
i1
;
1 7 I. 3 1 4 e
...
-2.5 •• g i...: ! I" 1
r• I i g . .... 1g 1 - -
I . - 1131
p. ?
.1
-.: 9h
- E.
. 3 1 i 'V. 4 T. ..1.11.4 ':: ;
-V. f' 1. ii ! i
i
V
i
6 r4 NJ 0 .N71
21.1 ,...• • ..5 .,4•••gt1d4a
.d .. ..n Ill.. rg.f.,

▪ I 1 *4 _ 5t1 iv: r•ril 514 .P.1 ali.1 1 1 t ;

F1 I i .i.: ff it V"; : .=t, :.", it; It; i 0 : ,
1 i L: 3 a: - -
t : i 4 i.'
41 1 1 6 6 .4114 .1 1 i 1:11 Ia l l V; Ili V" a 1 'i3
1 1 i1. 1 i52 : g 2 i g 2 1:: :. .12 .2° ..
., . g g IF _.1

TIJJ
51 I
I 1 "4 .." • • .
i i k A. .4 ..-. .. ..0 , I.:
 ••••• 212; N N TE4
41, N .N WN W.

n-
rs • WS FOD
l0 N•
'18 cM
l0 • N•
N .
O

n NW .M M m tO
M OW N
•
•
lO
00 .

00 1
O ,M 0, W, .10
MM WW •M WW • W
m.
W. 00 M.
0

g 0. o-1 •
NW
.s g c ki.21 1,1,0 M

• ..
• NW WM
N
w1 O
\0
m.
om ,S
. .0 M.0
.,
• M. W• '7.4- 1.1. ..
. . .. W . ..- .

U
O0
t,
•N
•W
0,N
N,a•
lO
• ON dP
s
1 :14 Nm •
pY

a NN NM SW NW
S 00.
10
OD.
N. N.
• 1,
00
Ol
1
•• • • • • • • • • • • • _ • • • • - - • - • • • - • • • • •••••-•-•••••••••••-
u.u1 1[10 0,
• S.
ti
O
.M
•
Dom
• MS
mm
UN •.7 0 tv •

a) 00 al Cr. ti MNW
ItleM • 0,
0 ,4 NS
I, •
N,

2 WM M, WN OW ,M
ow • WW •0
5 • 00
..M • 0. .-1

qm W •M •0 Om
01.-1 4 .
8

00S NN 0 O MS
01 •l0
so O 0;14 M. .M

•
11 K2 Old4 .4
M,

MM WM,M 0,N OO fi;

• a, • al
0 m4 m.

0; all m m m al
..
311
..
il irt ... .. 11.
.2 r"
H
r .1
.Li, t: "i„ .

i
'5 – 4
!,
IF 3 % 4. ,
2 2
it
t
C.
:1 «
1
ill . ' U
:1 2 -Iglit. . . . -'L.
.-.1J1 e'Ail 7.1.71 -;1 : :11:3 24111 .
:1, 1 vii l l E4.1] 10 trz
7: rE
.....; II; •
. P-4 :.:!.. pt
=11 vo .1. 3 iil 1;1
1i liZ4 F.A TIZ
3 ' iil
VI i5; 2-52 ,16" zjLI e437
1. ; .
4 4 4; .9 .: , ..-
 Type and Dimensions of Specimens Designated in

Cutting Diagrams, Figure 1

Specimen Type and Dimensions

1 Tension, 1 inch wide by panel thickness (t) by 24

inches long (before shaping).

2 Tension, 1 inch by (t) by 16 inches (before shaping).

3, 4 Compression (edgewise), 1 inch by (t) by 4t to 6t

for proportional limit stress and modulus of
elasticity determinations and 1 inch by (t) by
2t to 4t for maximum stress.

5 Compression (flatwise), 1 inch by (t) or (2t) by 1

inch.

6, 7 Flexure (flatwise), 1 inch by (t) by 14t or 16t,

plus 2 inches.

8a, 8b Shear, parallel to grain (edgewise) from parallel-

laminated panels only.

8a 2 inches by (t) by 2-1/2 inches (notched-type,

single shear).

8b 1 inch by (t) by 3 inches (Johnson-type, double

shear).

9b Torsion, 3t by (t) by 16 to 24 inches.

10 Toughness, 5/8 inch by (t) by 10 inches.

12 Water absorption, I inch by 3/8 inch by 3 inches.

Portion trimmed from ends of parallel-laminated

panels to eliminate light-colored areas of
questionable stability.
Izod, cylindrical shear, and dimensional-stability specimens were
taken from the unmarked portions of the panels. Compression speci-
mens of the normal laminates were taken from the undamaged ends
of tested flexural specimens.
Report No. 1639
• X 67004 I

Figure 8.--Flatwise views of typical edgewise toughness

failures of parallel-laminated modified wood made from
17 plies of 1/16-inch rotary-cut yellow birch veneer.
14 71964 F
 NORMAL LAMINATE

I MPREG

SEMI ® COMPREG

COMPREG

40N
STAY PA K

Figure 9.--Typical flatwise Izod impact failures of parallel-

laminated modified wood made from 1/16-inch rotary-cut
veneer. Yellow birch specimens are shown at the left and
sweetgum specimens at the right. Top views of each group
show notch in original surface and lengthwise grain.
z M 71965 F