TECHNICAL BULLETIN NO.

205 OCTOBER, 1930

GLUING WOOD IN AIRCRAFT

MANUFACTURE

BY

T. R. TRUAX
Senior Wood Technologist, Forest Products Laboratory
Branch of Research, Forest Service

UNITED STATES DEPARTMENT OF AGRICULTURE, WASHINGTON, D. C.

 TECHNICAL BULLETIN NO. 205 OCTOBER, 1930

UNITED STATES DEPARTMENT OF AGRICULTURE

WASHINGTON, D. C.

GLUING WOOD IN AIRCRAFT

MANUFACTURE
By T. R. TEUAX,* Senior Wood Technoloffist, Forest Products Laboratory^^
Brcmch of Research, Forest Service

CONTENTS

Page Page
Introduction _ 1 Gluing different species and surfaces of wood, 34
Glues for aircraft 2 Species glued.__ 34
Properties of aircraft glues 3 Recommendations for gluing side-grain
Other moisture-resistant adhesives 4 surfaces of various species 35
Durability of aircraft glues.. 6 Treating the wood before gluing 38
Increasing the durability of glued joints. 9 Gluing end-grain surfaces 38
Testing glues for quality 10 Drying and conditioning glued stock 39
Tests for casein glue 10 Recommended practice for the principal air-
Tests for blood-albumin glues 14 craft gluing operations 40
Tests for animal glue 14 Propellers.._ 41
Preparation of glues for use in aircraft 15 Plywood . 42
Preparing casein glue . 15 Laminated spars, spar flanges, bow ends,
Preparing animal glue 16 pontoon ribs, and reinforcing blocks... 43
Preparing blood-albumin glues 17
Preparation of wood for gluing 17 Scarf-type joints . 45
Moisture content 17 Wing ribs .46
Lumber . 19 Box beams 46
Scratched and tooth-planed j oints 20 General assembly 48
Shaped joints contrasted with, plain Principles of cross-banded and laminated-
jomts 20 wood construction 49
Veneer 20 Cross-banded construction 49
Gluing technic as related to the quality of Balanced construction 50
glue joints 22 Laminated-wood construction 51
Satisfactory and unsatisfactory joints 22 Glue formulas for use in aircraft 52
Spreading the glue 22 Casein glue formula No. 11 52
Single and double spreading 23 Casein glue formula No. 4B 54
Quantity of glue spread 23 Blood-albumin glue—hot press formula.. 54
Consistency of glue mixture 24 Paraformaldehyde-blood-albumin glue
Effect of assembly time and tempera- formula. 54
ture 24 Water-resistant animal glue form ula 55
Pressing and clamping 27 Literature cited 66

INTRODUCTION

The modern airplane requires from about 250 to 2,000 board feet
or more of lumber and considerable quantities of veneer and plywood
in its construction. It is estimated that 8,000 to 10,000 airplanes,
chiefly of commercial type, were built in 1929, of which approxi-
mately 90 per cent had wing structures of wood. Even in the fast
1 Acknowledgment is made to various members of the Forest Products Laboratory and
of the War and Navy Departments for assistance in the preparation of the manuscript
and to several aircraft manufacturers for contributed pliotographs and information.
a Maintained by the U. S. Department of Agriculture at Madison, Wis., in cooperation
with the University of Wisconsin.
119542°—30 1 1
2 TECHNICAL BULLETIIST 2 0 5, U. S. DEPT. OF AGRICULTUEE •

combat type of military plane wooden wing structures are used. In

?ome planes the entire wing framework and covering is of glued-
wood construction. (Pis. 1 and 2.) In propellers, control surfaces,
fuselages, and the pontoons and hulls of seaplanes and flying boats
wood is also used (i^, 28),^ (Pis. 2 and 3.) A partial list of more
detail parts, made wholly or partly of wood, includes : Wing beams
(solid and IDOX types), ribs, leading and trailing edges, bow ends,
braces, reinforcing or shear blocks, and light compartments; aileron
ribs, braces, and covering; rudder, stabilizer, and elevator parts;
tank compartments ; propellers ; center covers, step boards and walk-
ways; turtle decks; engine bearers; bulkheads; instrument boards;
floors; seats; cabin sides; fairings; doors; strut streamlining; and
keels, ribs, partitions, braces, and covering of pontoons and hulls.
(Pis. 4, 5, 6, and 7.)
Glue is used not only in laminating and building up large and
irregular wooden parts and in the making of plywood but it also
affords the principal means of fastening the various wooden parts
together into the finished structure. Joints, together with fasten-
ings, are, however, generally regarded as the weakest part of the
built-up construction. Consequently they control the design in very
large part, despite the fact that it is known that the technic of gluing
as practiced in aircraft is capable of considerable improvement

Since nearly every article of glued-wood construction represents

an economy in the use of timber resources the Forest Products Lab-
oratory has conducted studies upon glues and gluing,* a large part
of which has pertained to the use of glue in aircraft.^ The purpose
of the present bulletin is to give specific information about the gluing
of wood that is directly applicable to aircraft.
GLUES FOR AIRCRAFT
Most uses of glues in aircraft require adhesives that retain a large
proportion of their strength under moist conditions, even to the
extent of remaining safe in service after exposure to free water. Of
the various adhesives commonly used in woodworking, blood-albu-
min and casein glues most nearly meet these requirements.^ This
principle excludes at once the vegetable glues made from starch,
as well as fish glues and untreated animal glues, which, although
admirably suited for interior use, lose their strength rapidly and
almost completely when exposed for even short periods'to a mois-
ture-saturated atmosphere or to free water. Only where the glued
members are thoroughly protected against moisture changes by suit-
able coatings are any of the less moisture-resistant glues useful in
aircraft. In wooden propellers, high-grade animal glue has been
used successfully. Any increase in moisture content large enough
3 Italic numbers in parentheses refer to Literature Cited, p. 56.
*A general discussion of the use of glue in woodworking industries is given in The
Gluing of Wood i2Jf).
^ Research conducted in cooperation with the War and Navy Departments and the
National Advisory Committee for Aeronautics.
^This bulletin is confined to a discussion of glues that set and make strong joints in
wood. Marine glues, which are used in aircraft to make constructions water-tight, are
not included among the so-called woodworking glues since they do not set.
 GLUIKG WOOD IK ATECRAFT MANUFACTUBE ó

to weaken the glue seriously makes the propellers unserviceable on

account of changes in shape; hence highly moisture-resistant coat-
ings are used that protect the whole construction.
PROPEÈTIES OF AIRCRAFT GLUES

The principal adhesives of proved value for making wood joints

in aircraft are casein glues, blood-albumin glues, and animal glues
when properly protected against large moisture changes. In Table
1 the general properties and characteristics of casein, blood-albumin,
and animal glues used in aircraft are listed. On most of the points
of comparison there is a lack of definite and specific knowledge,
and as a result only general terminology can be used in describing
them. Furthermore, there is a wide variation among the glues of
the three classes. Only the strongest and most durable glues of each
class are described.
TABLE 1.—Properties cmd oharacteristios of Mfferent classes of woodworking
glues iised in aircraft^

Property or characteristic Casein glue Blood-albumin glue Animal glue (untreated)

Strength (dry) 2 _ Very high to medium... Medium to low -_._- __ Very high.
Strength (wet after soak- About 25 to 60 per cent About 60 to nearly 100 Very low.
ing in water 48 hours). of dry strength; varies per cent of dry
with glue. strength.
Durability in 100 per cent Deteriorates eventually; Deteriorates slowly but Deteriorates quickly.
relative humidity or rate varies with glue. usually completely in
prolonged soaking in time.
water.
Rate of setting... Rapid Very rapid with heat Rapid.
Working life _. Few to several hours 3__ Several hours to a few Few hours to several
days. days. 3
Consistency Medium to thick Variable, thin to thick, Variable from thin to
depending on formula. very thick with tem-
perature changes.
Temperature require- Used at ordinary room Heat required to set Control important for
ments. temperatures. glues; cold-press for- glue, wood, and room.
mula an exception.
Mixing and application... Mixed with cold water; Mixed with cold water; Soaked in water and
applied cold by hand applied cold by hand melted; applied warpi
or mechanical spread- or mechanical spread- by hand or mechanical
ers. ers. spreaders.
Tendency to foam... Slight to medium ._ Slight to pronounced Usually slight.
Tendency to stain wood.. Marked with some None, except that dark None to very slight.
woods. glue may show
through thin veneer.
Dulling effect on tools * -. Moderate to pronounced. Slight Moderate.
Spreading capacity: «
Extremes reported * _ 30 to 80 30 to 100 - 20 to 55.
Common range * 35 to 55 25 to 35.

1 Grades and quality only of glues that pass U. S. Government aircraft specifications.
. 2 Based chiefly on joint strength tests.
3 Casein and animal glues are likely to deteriorate seriously if kept liquid more than 1 day.
* Based on reports from commercial operators.
« Expressed in square feet of single glue line per pound of dry glue for veneer work.

Apart from joint tests there is comparatively little specific infor-

mation on the mechanical and physical properties of glues as mate-
rials of construction. Some data, however, are available on the ten-
sile strength of animal glues (^, 9j 10^ 16^ 17) and on the relation
between moisture content and relative humidity for animal and casein
glues. In Figure 1 is shown the tensile strength—moisture-content
relation for an animal glue that meets the specifications for aircraft
work. Small specimens of the glue at different moisture content were
 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGEICULTTJEE

tested in tension. The decrease in strength with increase in moisture

is striking. Curves shown in Figure 2 indicate the relation of rela-
tive humidity to moisture content of the same animal glue and of
one type of a water-resistant casein glue. The casein glue absorbed
an even larger percent-
age of moisture than
the animal glue, but
from other tests was
found to retain a
greater proportion of
its dry strength.
Water-resistant glues,
particularly casein and
blood albumin, set on
wood at almost any
moisture content, but
animal glue will not set
on wood of high mois-
ture content, 15 per
cent being about the
maximum for satisfac-
tory results.
Animal glue treated
with formaldehyde-
yielding materials gives
joints of about the same
initial resistance in cold
water as the water-
resistant casein glues.
However, tests indicate
that it is less durable
under prolonged expo-
sure. Blood - albumin
glues are less affected
16 Z4 3Z 40 46
MOISTURE CONTENT (PER CE//f)
by soaking in water
than other water-resist-
FIGURE 1.—The effect of moisture content on the ten-
sile strength of an animal glue that meets aircraft ant glues.
epecifications Formulas for the
preparation of water-
resistant casein, blood-albumin, and animal glues, which rate well
in water resistance among the glues of their classes and which are
suitable for use in aircraft, are given on p. 52.
OTHER MOISTURE-RESISTANT ADHESIVES

Besides the glues that have been discussed there are other adhesives
that compare favorably in many respects with those now in use. An
improvement in one or more respects may make their use in aircraft
desirable. Some of these adhesives are discussed in the next three
paragraphs.
 G^UIKG WOOD IN AIBCRAFT MAISTUFACTUEE

30 40 50 , 60 JO too
RñLATIV£ HUMIPITY (PER CENT)
FiGUEE 2.—Effect of relative humidity on moisture content of animal and casein
glues suitable for aircraft. Tests made at 80° F.
 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OE AGBICtJLTURE

VEGETABLE-PROTEIN GLUES

Soybean and peanut meal serve as bases for adhesives^ which in

general properties resemble casein glues. Sorrig of them have a good
degree of water resistance and are relatively cKeaj), but in their pres-
ent stage of development they do not show so high a dry strength
as the better quality casein glues.
PHENOL-ALDEHYDE CONDENSATION PRODUCTS

By heating a phenol with an aldehyde it is possible to produce

hard, strong adhesive substances, which when set by hot pressing
appear to be unaffected by moisture in any form. Nonaqueous solu-
tions of these materials have been prepared that can be applied to
wood. Teöts of plywood glued with these products indicate very
satisfactory joint strengths under all moisture conditions tried.
Apparently phenol-aldehyde adhesives are more durable under
extremely unfavorable conditions than the glues now used for
woodworking, but their high cost has thus far practically prohibited
their use. The necessity for hot pressing also restricts their appli-
cation.
MISCELLANEOUS MATERIALS

Other materials, including cellulose cements, asphalts, resins,

gums, and rubber have been tried as adhesives for wood. Cellulose
cements have set very slowly in wood joints in tests at the Forest
Products Laboratory and are otherwise difficult to use to obtain
satisfactory adhesion. Likewise asphalts, resins, and gums have
been tried out alone or in combination with other materials, but the
results have not been entirely satisfactory. In their present forms
these materials are not suitable as glues for aircraft joints, although
they appear to have possibilities of development. Plywood has been
glued by the hot-press process with an adhesive having rubber as its
basic material. Joint tests of such plywood in both wet and dry con-
dition indicate satisfactory strength. High cost is a present practi-
cal limitation affecting its use.
DURABILITY OF AIRCRAFT GLUES

None of the glues in practical use at present can be expected to

form permanent joints in aircraft parts that are subject to pro-
longed saturation with water, such as seaplane hulls and floats,
unless the glues or joints are especially treated. Even the most
water-resistant blood-albumin or casein-glued joints, which show a
strength when first saturated with water of 25 to nearly 100 per
cent of their dry strength, fail completely when exposed without
protection for a long time to free water or to extremely high atmos-
pheric humidities. Failure in such cases is apparently caused by
chemical decomposition of the glue or by its deterioration from the
'^ The following patents relate to soybean and peanut-meal adhesives : JOHNSON, O.
ADHESIVE. (U. S. Patent No. 1,460,757.) U. S. Patent Office, Off. Gaz. 312: 132.
1923; (Re 16,422). U. S. Patent Office, Off. Gaz. 350: 289. 1926. LAUCKS, I. F., and
DAVIDSON, G. VEGETABLE GLUE AND PEOCESS OF MAKING SAME. (U. S. Patents Nos.
1,680,732 and 1,691,661.) U. S. Patent Office, Off. Gaz. 375:1136. 1928, and 376:468.
1928. OsGOOD, G. H. GLUE. (U. S. Patents Nos. 1,601,506 and 1,601,507.) U. S. Patent
Office, Off. Gaz. 350 :861. 1926. OSGOOD, G. H. WATERPROOFING VEGETABLE PROTEIN—
BASE GLUE. (U. S. Patent No. 1,706,674.) Ü. S. Patent Office, Off. Gaz. 380:875. 1929.
 GLUING WOOD I]Sr AIRCKAFT MANUFACTUKE

action of fungi and bacteria or pçrhaps both. Furthermore, unpro-

tected water-resistant joints, which are known to withstand a limited
number of soaking periods of several days each followed by drying
out without seriously affecting their strength, fail eventually if sub-
jected to a long series of large moisture changes by alternate wetting
and drying. Under such cyclic conditions mechanical failure may
be a factor in the breakdown of the joints in addition to chemical
decomposition or the action of fungi.
Casein and blood-albumin glues, or even high-grade animal glues,
do not deteriorate under conditions that produce about 12 per cent
moisture content in wood, which is close to the average found in a
survey of aircraft stations. (Table 2.) At such a moisture con-
25r 1

30 1
/
/
J f
I" / f/
/

1 20 / /
/ f/ /
/ /
r • /
y /
/5
y ' • ^ •>
y^
7 iS^ ^^ ^^ _y /
W ^
^ ^ "inl^i :
^ *«
^ • .-** ^.^¿>
,^^ ^^, .-- • —' "^ ^
^i ii"' --
10 ZO 30 40 50 60 70 . 60 90 100
RELATIVE HUM!PITY IN ATMOSPHERE (PER CENT)
FIGURE 3.—Relation of the equilibrium moisture content of wood to the relative
humidity of the surrounding atmosphere at three temperatures

tent, which is considered fairly representative, well-made joints of

properly constructed members may reasonably be expected to remain
permanent with only ordinary protection. Glue failures, however,
are reported to occur in parts of aircraft under unusually adverse
service conditions, where the moisture content of the wood un-
doubtedly exceeds the 12 per cent average considerably. Adequate
data are not available for determining the exact critical moisture
content at which casein or blood-albumin glues retain their strengths
permanently and above which deterioration sets in. The available
data indicate, however, that the critical range of moisture content
is within 15 to 20 per cent for wood. Wood subjected to relative
humidities of approximately 75 to 85 per cent and to a temperature
of 70° F. takes on an equilibrium moisture content within this
range. (Fig. 3.) Under conditions where the wood retains 20 per
cent or more moisture there is no positive assurance of the perman-
ence of glues without special treatment.
8 TECHIíICAL BULLETIN- 2 05, U. S. DEPT. OF AGBICTJLTURE

6 6 10 12 14 Ib 16 20 22 24 26
TIME OF 'EXPOSURE TO 95 PER CENT RELATIVE HUM¡PIT Y (MONTHS)
FIGURE 4.—Effect of treatment of casein glues and plywood joints on durability:
A, Untreated ; B, plywood specimens dipped in asphalt paint ; C, plywood speci-
mens treated with coal-tar creosote ; D, plywood specimens treated with ß naphthol
in linseed oil ; E, creosote added to glue before making joints ; F, ß naphthol added
to glue before making joints ; G, creosote added to glue and plywood specimens
coated with aluminum powder in spar varnisih ; H, ß naphthol added to glue and
plywood specinnens coated with aluminum powder in spar varnisih. Joint strength
tests made every month on 30 specimens for each treatment
 Tech. Bul. 205. U. S. Dept. of Agriculture PLATE I

SOME TYPES OF COMMERCIAL AIRCRAFT EMPLOYING WOOD

A, 12-passenger monoplane In which the entire wing stnictare, including the covering, is made
of wood. Wood is also used in part in the cabin construction; B, 6-passenger monoplane with
framework and leading edge of wing and various parts of the fuselage made of wood; 0, 2-pas-
seuger seaplane with wing structure» main float and wing tip floats constructed of wood.
Tech. Bul. 205, U. S. Dept. of Agriculture PLATE 2

TYPICAU WOODWORKING DEPARTMENTS IN COMMERCIAU AIRCRAFT

FACTORIES
A.Building wings entirely of wood for a large transport plane. The frame of the wing is covered
with plywood; B, using wood for turtle decks and fairings in fuselages.
 Tech. Bul. 205. U. S. Dept. of Agriculture PLATE 3

PltZiSIit-Hj

FUYING BOAT AND PROPELLER CONSTRUCTION

A, Constructing flying boat hulls of wood; B, manuiacturing propellers of wood. Laying out and
baud sawing laminations preparatory to gluing.
 Tech. Bul. 205, U. S. Dept. of Agriculture PLATE 4

WING BUILDING OPERATIONS

A, Making wooden wing straetures for both biplanes and monoplanes; B, inspecting a wing struc-
ture before covering; C, covering wings with cloth.
 GLUING WOOD IN AmCRAFT MANUFACTURE

TABLE 2.—Moisture content fov/nd in wooden airplane parts under service

conditions

Moisture content-
Stations
where Kinds of Speci- Maxi- Mini-
Kind of construction Service samples woods mens Average for mum for
were tested tested for speci- mum
anyone anyone
taken 1 mens station station
tested 2 and wood and wood

Number Number Number Per cent Per cent Per cent

Solid and laminated wood s.. - Navy 10 10 419 12.7 15.3 8.8
Plywood ...do - 6 4 35 13.8 17.8 9.1
Solid and laminated wood ^-. Army 10 4 371 11.6 14.0 9.3
Propellers --do - 9 76 10.3 11.3 8.4
Plywood - —do 7 39 13.9 16.7 11.7

1 Army and Navy stations are considered separately, although they are frequently located close to each
other.
* Grand average for all stations where determinations were made; station averages were prorated on
number of specimens tested.
3 Exclusive of propellers.

INCREASING THE DURABILITY OP GLUED JOINTS

As a result of research work at the Forest Products Laboratory *

special treatments of glues and joints have been discovered that
materially increase the durability of glued joints under severe expo-
sure. In Figure 4 are shown the results of a still incompleted series
of prolonged exposure tests on casein-glued plywood. The results
shown are based on two casein glues—one made in accordance with
Forest Products Laboratory formula 4B (p. 54) and the other a
special commercial aircraft glue. Because of the similarity of the
results obtained with the two glues the data were averaged. The
special treatments were of three general kinds; namely, (1) addi-
tion of preservative materials to the glues before making the joints ;
(2) preservative treatment of plywood that had been glued with
untreated glues; and (3) treatment of glues before making joints,
followed by the application to the glued joints of aluminum
powder in spar varnish, which is an -effective moisture-excluding
coating.
In similar tests, blood-albumin glue was treated with sodium
chromate and made into plywood. Ten per cent of sodium chromate
proved to be effective in increasing the durability of the blood-
albumin glue. Plywood, which had been made with untreated
blood-albumin glue, was also treated either with 6 naphthol in
linseed oil or with creosote. The treatment of the plywood speci-
mens both with the 6 naphthol in linseed oil and with the creosote
increased the durability of the joints. In one series of tests,
plywood which had been glued with blood-albumin glue and treated
with creosote after gluing, was exposed for four and one-half years
in a moisture-saturated atmosphere without any appreciable loss in
strength.
In applying special treatments to aircraft joints it is preferable
to treat, as far as possible, complete assemblies, such as wing sections
8 Tests were conducted in cooperation with the National Advisory Committee for
Aeronautics.
10 TECHNICAL BULLETIN 2 05, U. S. DEPT. OF AGRICULTURE

and floats (pis. 5 and 7, A) before covering. It is recommended

that the assembled parts be dipped in a hot bath of boiled linseed
oil containing 25 per cent of 6 naphthoP or in coal-tar creosote.
Better protection will be afforded by dipping the member first in
a hot bath, about 200° F., and then in a cold bath, allowing the
member to remain about one-half hour in each. The linseed oil and
6 naphthol mixture is mushy at ordinary room temperatures and
the " cold bath " must be kept at about 150° to obtain proper
absorption and to prevent the formation of a thick layer on the
surface of the wood. The weight of the member will be increased
somewhat more with the increased temperature and with the time it
remains in the baths. The weight added in a treating process will
depend also upon the size and shape of the members. Small
pieces having large surface areas in proportion to their volume,
such as the specimens used in the tests on which Figure 4 are based,
will have more weight added per unit volume than large, compact
parts.
The foregoing treatments are distinctly superior to the present
methods of dipping, spraying, or brushing with varnish or asphaltic
materials. Applying the preservative to the glue alone will increase
the weight of the member only slightly and, although not so effective
as treating the whole member, it may give sufficient protection for
certain aircraft parts exposed under average conditions of use.
TESTING GLUES FOR QUALITY
Due vigilance should be exercised to see that glues for use in air-
craft conform to standard specifications or requirements. Specifica-
tions for glues and glue materials are valuable both as a guaranty of
quality and as a means of obtaining more uniform results in use.
No attempt is made in this publication to describe acceptance tests
in detail, since the published literature pertaining to the subject is
ample and readily available ^^ (^5, ii, 6^ 27).
TESTS FOR CASEIN GLUE

Joint tests constitute at present the principal means for judging

the value ofl casein glues for use in aircraft. Considerable time is
required to make joint tests. Moreover, they are difficult to perform
accurately, and at best yield only indirect comparisons of glue proper-
ties. On the other hand, joint tests, when properlv carried out, afford
highly important information about the value oi the glue in service
and its adaptability to manufacturing conditions.
Block shear and plywood tests for joints have been adopted as
standard by the Army and Navy. Except for certain changes as to
the form of specimen and shearing tool used in the block shear test
and the method of measuring water resistance in the plywood test,
^The treatment here recommended is the same as used on the plywood specimens indi-
cated in Figure 4. The minimum amount of 6 naphthol, which will properly protect
the glued members is not known.
^^ Detailed information on U. S. Army and Navy acceptance tests are contained in
United States Army, Glue, Casein, Specif. No. 98-14020-D, 5 p., illus, 1926 ; United
States Army, Glue, Hide, Specif. No. 3_140, 3 p., 1927 ; United States Navy, Glue, Casein.
Dept. Specif. No. 52G8, 5 p., illus., 1924; and United States Navy, Glue, Hide, Dept.
Specif. No. 52G4C, 2 p., 1925. j-. . » t'
 GLUING WOOD IN" AmCEAFT MANUFACTUItE 11^
the details of the two tests are substantially the same as those origi-
nally worked out by the Forest Products Laboratory (i). The form
of test specimens, shearing tool, and grips used at present at the
laboratory are shown in Figures 5 and 6. Water resistance is now
measured at the laboratory and by both the Army and the Navy by
testing plywood specimens wet after soaking in cold water for 48
hours, whereas originally the only requirement was that plyivood
show no separation of the plies after soaking in hot or cold water.

SECTION A-A

FIGURE 5.—Specinuen and shearing tool used at Forest Products Laboratory for
block-shear joint test. Experience has shown that it is necessary to conform
strictly to the details shown in order to obtain comparable results

Experience has shown that the following points are important in

making joint tests :
BLOCK SHEAR JOINT TEST

(1) Select a wood (hard maple is satisfactory), which glues well, is high in
density, of straight grain, and free from defects. Condition to a uniform
moisture content of about 7 per cent.
(2) Glue the joints large enough that four or more specimens (fig. 5) can
be cut from each joint ; also test two or more joints for each glue.
(3) Surface the wood pieces smoothly and to a uniform thickness* imme-
diately before gluing.
(4) Follow directions carefully in preparing glue for use.
(5) Spread approximately 1^ ounces of wet glue evenly to each square foot
of joint a-nd apply pressure uniformly to the surfaces when the glue is at the
proper consistency. (P. 24 for details.)
(6) Allow the glue to set under pressure for 15 hours or more and then
condition joints to a uniform moisture content of approximately the same as
12 s. DEPT, OF AGRICULTURE

before gluing. (Seven days' conditioning time at room temperature is usually

sufläcient.)
(7) Prepare specimens of the type shown in Figure 5, and test on a testing
machine equipped with a shearing tool as shown also in Figure 5. Apply the
load to the specimen at a rate not greater than 0.0157 inch per minute.
(8) Record for each specimen both the breaking load and the approximate
percentage of wood failure occurring over the glue-line area.
(9) Whenever a specimen fails with a load less than the required average"
and the failure occurs 50 per cent or more in the wood, the specimen should be
rejected in computing the average. If the variation among individual speci-
mens is 10 per cent or more and the breaking strength of one or more specimens
is equal to or higher than the
required average strength, the
test should be repeated.
PLYWOOD JOINT TEST
(1) Use a strong wood (yel-
\y
low birch is quite satisfactory).
(2) Select veneer that is
straight grained, comparatively,
free from checks, smooth, uni-
4 1? 0 ®f form in thickness, and free from

'CM other visible defects. Results

t rkV ej
obtained from specimens with
cross grain through the face
SPECIMEN fi
\ plies and with badly checked
y) cores do not represent the full
strength of the joint and should
be disregarded in testing the
'"in- quality of glues.
(3) Use, where possible, a
standard construction. Three
plies, each one-sixteenth of an
inch thick, glued with the grain
of the core at right angles to
the faces, are generally used.
(4) Condition all veneer to
approximately the same mois-
SPECIMEN Ö ture content before gluing. It
is recommended that the mois-
FIGURE 6.—Specimens and grips used at Forest ture content be not higher than
Products Laboratory for plywood joint test. Ex-
perience has shown that it is necessary to con- 12 nor lower than 7 per cent.
form strictly to the details shown in order to (5) Prepare glues for use in
obtain comparable results accordance with manufacturer's
directions.
(6) Glue the plywood under carefully controlled conditions. (P. 35 for
details.)
(7) Leave the panels under pressure for at least 15 hours and then condition
them to approximately the same moisture content that the veneer had before
gluing. Three days' conditioning is usually sufficient.
(8) Cut specimens a^ shown in Figure 6. Where the face plies do not with-
stand the loads under test with specimen A, the shearing area should be reduced
to one-half square inch as shown in specimen B.
(9) Use care in placing specimen in machine and adjusting the grips. (Fig.
6.) Apply the load to the specimen at a rate of 600 to 1,000 pounds per minute.
(10) Record both breaking load and nature of failure.
(11) Test an equal number of specimens dry and wet from each panel and
preferably four or more panels for each glue.
" The average breaking strength required by the TJ. S. Navy Dept. Specif. No. 52G8
(1924) and the U. S. Army Specif. No. 98-14020-D (1925) is 2,400 pounds per square
inch. If the test is carefully performed several available high-grade casein glues give
joint strengths considerably in excess of 2,400. At the Forest Products Laboratory a
number have exceeded 2,800 pounds per square inch.
 GLUING WOOD IN AIECEAFT MANTJFACTUBB 13
(12) In case a specimen fails with a load per square inch less than the re-
quired average strength" and the failure occurs 50 per cent or more in the
wood, the specimen should be disregarded in computing the average.
Plywood made of different thicknesses of veneer will give different
values when tested in the manner shown in Figure 6. This is because
of the form of specimen, its tendency to bend, and the eccentric
loading which occurs under test. The test values vary with the
thickness of both the face and core plies, and the order of magnitude

•?

I
H
i
- T£3T£P PRY
-TESTED WÍT AFTER SOAKING IN WATER 4BHOURS
ALL FACES k, ¡HCH; CORE VARIABLE IN THICKNESS

^O.U V
^ 0.70

-J I I L- _I L_J L_
240 280 320 360 400 440 460
PL YWOOP STRENGTH (pOUNPS PER SQUARE INCH)

FIGURE 7.—Effect of ply thickness on strength values obtained in plywood test

of these variations is shown in Figure 7. It is important, therefore,

that a standard construction be used in making tests on glues.
In addition to joint tests casein glues are sometimes tested to
determine their working life. Highly water-resistant casein glues,
when mixed ready for use, thicken after a time, usually within a few
hours, to a jelly and become unusable. The working life is con-
sidered as the period in which the glue mixture remains in a satis-
12 U. S. Army Specif. No. 98-14020-D (1926) and U. S. Navy Dept. Specif. No. 52G8
(1924) require an average of 250 pounds per square inch tested dry and 125 pounds per
square inch tested wet after soaking in water at room temperature for 48 hours.
14 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGRICULTURE

factory consistency for spreading on wood. A requirement of four

to five hours' working life is reasonable. It must be remembered,
however, that some thickening will occur during this period. A glue
that does not thicken and set eventually to a firm jelly is almost
certain to be low in water resistance, even though the mixture is
used while fresh (p. 3). Hence the working-lie test is valuable
not only as a measure of the time during which the glue remains in
a satisfactory condition for use but as a rough check on its water
resistance. It seems possible that consistency and jelly strength tests
in addition to determining the working life of casein glues may
eventually be used to measure certain of their mechanical properties
(4).
TESTS FOR BLOOD-ALBUMIN GLUES

The blood-albumin glues used in the manufacture of aircraft

plywood are mixed by combining the various ingredients at the time
of use. The aircraft manufacturer is not concerned with the testing
of blood-albumin glues directly since this class of adhesives is used
chiefly in gluing plywood, most of which is made in plywood manu-
facturing plants. Joint tests are the principal means by which
blood-albumin glues are evaluated. The plywood test described for
casein glues (p. 12) can also be used for blood-albumin glues.
TESTS FOR ANIMAL GLUE

The viscosity and jellj strength of specified concentrations of

solutions, tested at definite temperatures, are at present the prin-
cipal criteria of quality in animal glues. For proper execution these
tests require careful control of conditions, accurate measuring ap-
paratus, and a trained observer. In general, the higher the vis-
cosity and jelly strength, considered jointly, the better the strength
and general serviceability of the clue. Either one of the two tests
without the other may yield a fictitious indication for a given glue.
The adoption of standard methods and equipment will greatly
simplify the testing of animal glues and largely overcome the con-
fusion which has existed in the past. The National Association of
Glue Manufacturers has worked out and adopted a uniform system
of testing and grading glues based on the viscosity and jelly strength,
which promises to bring about a very desirable simplification (5).
The ]elly point of an animal glue is the temperature at which a
solution of the glue changes from a liquid to a jelly. It deter-
mines in large part the necessity for warming the wood and is there-
fore of considerable importance in the shop. As an acceptance test,
however, it is doubtful whether the jelly point reveals any service
quality of the glue not indicated by viscosity and jelly strength.
No standard method for determining the jelly point has yet been
agreed upon, but an approximation can be easily obtained by cooling
a portion of glue solution and noting the temperature at which it
ceases to flow.
A number of other properties of animal glues are of practical im-
portance. Acidity or alkalinity, tendency to foam, odor and keep-
ing quality, and amount of grease and other extraneous materials
present all determine to a greater or less extent the fitness of a glue
for high-grade joint work as required in aircraft. The first three
 GLUING WOOD IN AmCBAFT MANUFACTURE 15

of these properties can be determined comparatively easily within

the limits required of a glue for joint work and should be made a
consideration in selection. The presence of grease and other for-
eign materials in small amounts is not particularly objectionable,
and the exact percentages can be determined only by quantitative
analyses, which would hardly be justified under all conditions. The
amount of moisture present should be taken into consideration in the
preparation of solutions for the viscosity and jelly strength tests.
PREPARATION OF GLUES FOR USE IN AIRCRAFT
Clean, cold water should be used with the glues that are under
consideration in this bulletin, and the quantity of dry glue and
water should be determined by weight rather than measure or guess.
The aim in mixing glues should be to produce a solution of the
proper consistency and free from air bubbles, foam, and lumps of
undissolved material. Machine mixing normally produces a more
thoroughly mixed glue than stirring by hand.
PREPARING CASEIN GLUE

Casein glues suitable for use in aircraft are of two kinds, those
prepared in dry form ready to be added to water and those in which
the various materials are combined at the time of mixing with
water. The first are generally referred to as " prepared " glues
and the second as " wet-mix " glues. Both kinds are mixed without
heating and in the same type of equipment. Various types of
mixers have been used successfully, but the dough type (pi. 8, A
and B), equipped with a mechanism for turning the paddle in a
double rotary motion at two or three different speeds, has been quite
generally used with excellent results. The chief requisites of a
mixer for casein glues are (1) thorough agitation, preferably with
different speeds of the paddle, and (2) a bowl that can be readily
removed from the machine for cleaning, made of some metal that
will not corrode rapidly from the action of alkali.
In aircraft repair shops in which only small quantities of glue are
needed the mixing may be done by hand, but this is recommended
only where machine stirring is not feasible. A small mixer of the
type shown in Plate 8, A and B may be procured, equipped with two
sizes of mixing bowls in which quantities of wet glue varying from
2 to 20 pounds may be mixed successfully. Where hand stirring is
necessary a paddle or stiff spatula should be used. Egg beaters may
be used but unless operated at low speed they have a tendency to stir
in air and thus produce a foamy mixture. The conditions and pro-
cedure recommended in succeeding paragraphs for mechanical mixers
should be approximated as nearly as possible in hand mixing.
A survey of aircraft-gluing operations (22) has shown that various
types of mixers are in use with varying results. Drill presses, to
which are attached a small, homemade paddle or stirrer are used in
a number of plants. These are usually run at too high a speed and
cause a foamy mixture. Furthermore, the stirrer does not agitate
the mixture uniformly, and lumps are apt to form while the dry
glue powder is being added. Other 1-speed mixers now in use fre-
quently permit lumps to form or whip air into the mixture. For
these reasons the use of a 2-speed mixer, equipped with a stirrer large
16 TECHNICAL BULLETIN âÔ5, tí. S. DUPT. OF AGKlCÜLTtfRE

enough to agitate the whole mixture uniformly at speeds of 100 to

120 and 50 to 60 revolutions per minute, respectively, will improve
the mixing in many aircraft factories.
The correct proportion of glue and water varies with the particular
brand of glue and somewhat with the tvpe of joint. Manufacturers
usually recommend the proportions of glue and water, and these
should be followed unless other proportions are known to give better
results. For most prepared casein glues a ratio of 1 part of glue to
2 parts of water (by weight) gives a proper consistency for side-
grain joints. For wet-mix glues the proportions of water and other
ingredients are given in the formulas in the appendix. For gluing
end-grain joints some variation in the glue-water ratio is necessary
as described on page 45.
With prepared casein glues the dry powder is simply mixed
thoroughly with the water and stirred until it has dissolved. The
water should first be placed in the bowl of the mixer and the glue
sprinkled or sifted in slowly with the paddle in rapid motion. Care
should be used that large lumps do not form. After the dry glue
has all been added the motion of the paddle should be slowed down,
but the stirring should be continued (usually 20 to 30 minutes) until
a smooth mixture of even consistency results. Variations in pro-
cedure are advisable for certain prepared casein glues. In such glues
the glue-water proportions and other details are usually recom-
mended by the manufacturer.
Wet-mix casein glues are prepared according to formulas by the
addition of the separate ingredients at the time of mixing. Detailed
directions for mixing casein glues suitable for aircraft work are
given on pages 52 and 54.
The prepared glues are convenient and easy to mix, while the wet-
mix glues require somewhat more attention and skill in mixing.
With either glue, however, a procedure can easily be acquired for
mixing the various ingredients which the average man can follow
without difficulty.
Most water-resistant casein glues become noticeably thicker on
standing at room temperatures, but in such cases the quality of joints
apparently is unaffected so long as the' glues can be applied satis-
factorily. On the other hand, casein glues that do not set to a jelly
until a long time after mixing frequently become thinner; such
thinning action usually indicates a deterioration of the glue, which
may render its use unsafe. In general, mixtures that remain liquid
for a long time do not produce joints of high water resistance, such
as required in aircraft, even though the mixture is used while fresh.
Mixers, spreaders, and other equipment used with casein glues
should be thoroughly cleaned at regular intervals to prevent deteri-
oration of the glue and the inclusion of pieces of dry glue in the
freshly prepared mixture. A thorough cleaning every working day,
before the glue hardens, is highly desirable.
PREPARING ANIMAL GLUE

An animal glue which meets aircraft specifications ^^ requires for

each pound of dry glue about 2% pounds of water for standard side-
13 Specification No. 3-140 (1927) of the U. S. Army and No. 52G4C (1925) of the Ü. S.
Navy cover animal glue for use in aircraft.
 c:

r
>
H
m

THE STRUCTURE OF A HALF MONOPLANE WING

This part of the wing, except the aileron, is made of wood. Its construction reqtiires the maliing of more than 1,000 glued joints. Note the many small pieces
of spruce and the use of plywood for the sides of the gas compartment, the sides of the bos beams, the leading edge, the webs of the ribs, and the gussets.
 Tech. Bul. 205, U. S. Dept. of Agriculture PLATE 6

WING STRUCTURE MADE LARGELY OF WOOD

A, (o) Gas tank compartment, (b) sides of box beams, (c) compression rib webs, and (d) numerous
gussets made of plywood, (e) cap strips, <J) rib diagonals, (g) spar flanges, and (*) braces made of
solid spruce; B, plywood use din (a) ailerons, ((>) spars, and (c) ribs. (3) Bow end laminated from
several pieces of spruce.
 Tech. Bul. 205, U. S. Dept. of Agriculture PLATE 7

WOOD IN PONTOON, FUSEUAGE, AND AMPHIBIAN CONSTRUCTION

A, Pontoon frames made largely of wood. The curved ribs are of thin laminated construction; B,
wood is iised on the frame of fuselages as a base to which is attached the covering and various acces-
sories. Seats, floors, instrument boards, bulk heads, and other parts are also made largely of
plywood; O, wood and plywood used in the construction of an amphibian hull.
Tech. Bul. 205. U. s. Dept. of Agriculture PLATE 8

SATISFACTORY MIXING EQUIPMENT FOR CASEIN AND ANIMAL GLUES

A, 2-speed, 6-quart size, electric, casein glue miier in use in an aircraft fectory. The dry glue
and water are weighed on the scale and the mixed glue is distributed to the workmen in the
small paraiDned paper cups; B, S-speed dough-type, electric mixer, equipped with 3 and 8
quart bowls and two sizes of paddles for mixing casein glue; C, two types of electric glue pots
with automatic temperature control for animal glue.
 GLUING WOOD IÎT AIRCEAFT MANUFACTURE 17

grain gluing. This proportion may be varied to 1 part glue to 2

parts water where a mixture of a heavier consistency is required, as
on end-grain surfaces or in quick pressing operations and on warm
wood. On the other hand, a mixture of 1 to 3 may be used for sizing
purposes.
The following recommendations are made for the preparation and
handling of animal glue for aircraft work: (1) Soak glue in clean,
cold water until all particles are softened.^* Mix thoroughly the
glue and water when combining so as to insure uniform absorption;
(2) melt glue in a heater from which the solution mav be removed
as soon as melted or in which the temperature can be accurately
controlled (pi. 8, C) ; (3) keep the melted glue at a temperature
between 140° and 150° F.; (4) use the glue as quickly as possible
after melting. Glue that has been heated for a total time of four
hours should be discarded ;^^ (5) clean thoroughly all glue
receptacles at least once during each working day.
PREPARING BLOOD-ALBUMIN GLUES

Blood-albumin glues are prepared by combining the several in-

gredients at the time of mixing. Dried soluble blood albumin is
generally used. The proportions of water and other materials added
to the dried blood albumin vary greatly with different formulas.
Two blood-glue formulas developed at the Forest Products Labo-
ratory, with detailed directions for mixing, are given on page 54.
In general blood glues, used commercially, are prepared in ac-
cordance with secret formulas and processes. Most formulas are
reported to have originated in European and Asiatic countries where
blood glues have been used for a longer time and more extensively
than in this country. The mixtures of most value for aircraft are
of a viscous, or noncoagulated type, and have a comparatively low
water content.
The working life of blood glues is extremely variable, depending
chiefly upon the proportions of alkali and water added to the dried
albumin. Glues can easily be mixed that have a working life of
several hours to a day or more. As with other glujes made from
animal material, precautions should be taken to prevent deteriora-
tion from bacteria and other organisms of decay.
PREPARATION OF WOOD FOR GLUING
MOISTURE CONTENT

Wood for gluing should have the proper moisture* content, uni-
formly distributed throughout, and should be free of casehardening
and other stresses {21),
1* The time required to soften the glue thoroughly may vary from one to several hours.
One hour is sufficient time for a ground glue that will pass an 8-mesh screen. Longer
periods are required for coarsely ground and flake glues. Glue which is soaked for several
hours should be placed in a refrigerator at a temperature of about 40" to 50° F. to prevent
deterioration.
15 Batches of mixed animal glue may be kept from one day to the next at a temperature
of 40° to 50° F. without any appreciable deterioration and remelted and used later with
safety. The batch should be discarded, however, when the combined heating periods total
four hours. Dry glue will keep for an indefinite period at ordinary room temperatures if
safeguarded from dampness.
119542°—30 2
18 TECHNICAL BULLETIN" 2 0 5, U. S. DEPT. OF AGRICULTURE

Any considerable change of moisture content in wood after glu-

ing usually develops stresses in the glue and reduces the load that
the member will withstand in service. The ideal moisture relation
to be aimed at in gluing for maximum joint strength, therefore, is
that the moisture content of the wood at the time of gluing plus the
moisture added by the glue should be equal to the average moisture
content that the member will have in service. In Table 2 are shown
the average maximum and minimum moisture-content values of wood
in service in airplane parts, as determined at various aircraft stations
throughout the United States. Table 3 shows the approximate per-
centages of moisture added to wood by the glue in certain types of
construction.
The effect of a different moisture content of the wood, before and
after gluing, on the joint strength of casein-glue plywood is illus-
trated in Figure 8. The plywood for the wet tests was soaked in
cold water for 48 hours and tested in the saturated condition. The
dry tests were made at a moisture content of 10 to 12 per cent.
TABLE 3.—Percentaffes of moisture added to wood in gluing ^

C3
Species and thickness of— 'S
nöc»
Construc- es o
tion
Faces
§1
Crossbands Core Laminations
5 1
No. In. P. Ct.
3 }4 8-inch, yellow 348-inch, yellow He 48.8
birch. birch.
3 ,}42-inch, yellow 3^2-inch, yellow %2 32.5
birch. birch.
3 do _. H 6-inch, yellow M 29.6
poplar.
3 î^o-inch, Hondu- H 0-inch, Hondu- Ma 19.6
Pljrwood.-- ran mahogany. ran mahogany.
6 ^8-inch, yellow î^o-inch, yellow H 0-inch, yellow %2 36 1
birch. poplar. poplar.
5 ^0-inch, yellow H 30.9
birch.
6 He-inch, yellow H2-inch, yellow H 2-inch, yellow % 21 2
birch. poplar. poplar.
7, —--do do.3 H 23.1
15 do.* do.3 1 23.4
7 ">í-íñch,"'s'itka' 17.3
Laminated- spruce.
10 %-inch, white oak 1.7
2 1-inch, white ash. 2 .8

1 Percentages calculated from average weights (oven dry based on volume when air dry) of various species
as given in Table 3 of Report No. 84 of the National Advisory Committee for Aeronautics (7) or Table 2
of U. S. Department of Agriculture Bulletin 556 (19). In the calculations it is assumed that all the sur-
plus moisture added by the glue is absorbed by the wood. This assumption is known to be somewhat
in error, but it nevertheless affords a fairly satisfactory basis for comparison.
2 A glue mixture of 1 part dry glue and 2 parts water and a spread of 40 square feet of single glue line
per pound of dry glue is assumed in this calculation. ^
8 Core consists of 3 plies.
4 Outer 12 plies all of same species and thickness.,

From these data it might seem that the proper moisture content
of wood at the time of gluing is an' extremely variable quantity,
depending upon the high or low moisture content of the finished
article in service and the exact thickness and other characteristics
of stock being glued. In general practice, however, adjustments can
not be made for all such widely varying factors. Experience has
shown that a moisture content of 5 to 10 per cent is satisfactory for
 GLIJIlSrG WOOD lî^ AIROKAFT MANUFACTXJIiE 19
veneer and thin laminations that are to be glued into plywood and
laminated construction, respectively, provided all plies or lamina-
tions are at approximately the same moisture content within this
range. For laminations one-half inch thick and thicker, a moisture
content of 10 to 12 per cent is recommended, except for propellers.
Propellers are commonly coated with an effective moisture-resistant
finish, so that a moisture content of about 8 per cent before gluing
is recommended. Wood for aircraft seldom must be conditioned
to a moisture content as low as 5 per cent or higher than 12 per cent.
Under average aircraft-manufacturing conditions a range of mois-
ture content from 7 to 12 per cent is the most easily obtained and
maintained and will ordinarily prove satisfactory for most veneer
and lumber.
The moisture content of wood that has been properly dried for
gluing may change in storage or in the factory during manufacture,

TESTEP WET

_I I I u -J I l__J L_
'""i? 5 10 15
MOISTURE CONTENT Of VENEER BEFORE GLUINQ (FER CENÍ)

FIGURE 8,—Effect of moisture content of veneer upon the

strengtli of plywood panels

depending upon the relative humidity of the surrounding atmos-

phere. In the manufacture of certain aircraft parts it may be nec-
essary to control the relative humidity in the storage room or factory
to obtain the best results. The relation of relative humidity to the
moisture content of wood is shown in Figure 3.
LUMBER

CONDITIONING

Lumber that has been dried to the approximate average moisture

content desired may still show differences between various boards
and between the center and the outside of the same piece. A varia-
tion of 1 per cent in moisture content may be expected between
boards of the same species even when they are of uniform dimen-
sions and are exposed continuously under the same conditions of
temperature and humidity.- Larger differences may be adjusted by
allowing the lumber to condition under suitable atmospheric tem-
perature and humidity until the moisture content of each board
comes to equilibrium with the moisture in the surrounding air.
20 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGRICULTURE

(Fig. 3.) The thickness of boards, kind of wood, and moisture dif-
ferences present all affect the time required for the lumber to come
to equilibrium.
MACHINING

Machining lumber before it is to be glued should accomplish three

important objects: (1) Elminátion of defects, (2) reduction to
the proper dimensions, and (3) preparation of the contact surfaces
for the glue. The stock should be conditioned to the desired mois-
ture content prior to the machining. It is best to machine the stock
just before gluing, so that the surfaces will not become distorted
from atmospheric moisture changes. The surfaces for gluing
should be smooth and true. Plies for cross-banded or laminated
construction (fig. 9, A to G) should be of uniform thickness. A
small variation in thickness in each piece may cause a difference of
serious proportions, as, for example, when a number of similar pieces
are piled in the same order as they come from a planer. Planer
marks, chipped or loosened grain, and other surface irregularities
are objectionable.
SCRATCHED AND TOOTH-PLANED JOINTS

Scratching, tooth planing, or sanding the surfaces to be glued is

still practiced by many operators, who either believe that roughen-
ing and tearing up the surface fibers gives the glue a better chance
to adhere to the wood or that the irregularities are cut away and
flatter pieces, more uniform in thickness, are obtained. Investiga-
tions into the penetration of glue into wood and tests of the strength
of well-made joints show no benefit from such practices (^4).
SHAPED JOINTS CONTRASTED WITH PLAIN JOINTS

The fact that the tongue-and-groove and other shaped joints

present a larger gluing surface than the plain joints is often con-
sidered a theoretical advantage. In practice, however, this advan-
tage is often lost entirely or m part. Shaped joints are more diffi-
cult to machine than plain straight edges so as to obtain a perfect
fit of the parts. Lack of contact may make the effective holding
area smaller in the shaped joint than on a flat surface, and this
may actually reduce the strength. Furthermore, if proper condi-
tions are maintained, plane tangential or radial surfaces (fig. 9, A
to G) of most woods can be glued so as to develop the full strength
of the wood, and the extra contact surface therefore becomes super-
fluous so far as strength is concerned.
In edge-joint gluing, however, the tongue and groove is an ad-
vantage in that the pieces are held in alignment during the assem-
bling process and there is less slipping of the parts when pressure
is applied. This makes possible more rapid clamping in many op-
erations. When end-grain surfaces have to be joined, as illustrated
in Figure 9, H and I, cutting to sharp angles and the use of reinforc-
ing members are usually necessary.
VENEER

Veneer is cut by three different methods—sawing, slicing, and

rotary cutting. By far the largest part of all veneer is cut by the
rotary process because it is cheaper and involves less waste in manu-
 ÔLUIKÔ WOOD m AmCËAÏ'î MANUFACTURE 21
facture. Both rotary-cut and sliced veneers are checked somewhat
on the side of the sheet next to the bolt or flitch (pi. 9, B) during
the manufacturing process whereas sawed veneer is more uniform

\////////////////^^^^^^

////////////////////////////////////////////

v/////M////////^////<>//^¿/ww//w//m

FIGURE 9.—Types of joints : A, Plywood (all veneer) ; B, plywood (lumber core) ; C,

plywood (lumber or veneer core) ; D, bent plywood (all veneer) ; E, curved lami-
nated construction ; F, laminated propeller hub ; G, section of laminated beam ;
H, serrate or finger joint; I, scarf joint

in properties on the two sides of the sheet. The amount of checking

varies with (1) the kind of wood, (2) the thickness of the veneer
cut, (3) the preparation of the bolt or flitch for cutting, and (4)
the efficiency of the machine-cutting operation. Woods with small
pores, such as birch, gum, and poplar, are easier to cut without
22
serious checking than woods with large cells, such as oak and ash.
Thick stock is more liable to check in cutting than are thin sheets.
Proper heat treatment of the bolt and flitches is important for best
results in both rotary cutting and slicing.
There appears to be no reason why plywood for aircraft should
not be made from well-cut veneer made by any of the three methods.
Comparative strength and warping tests, which were made on ply-
wood manufactured from well-cut veneer by the three processes and
from the same logs, showed no consistent differences in properties.
On the other hand, badly checked veneer when used as cores in 3-ply
panels (pi. 9, C) is known to reduce the joint strength in the plywood
and should therefore be rejected for aircraft work.
The following characteristics are important in selecting veneer
for use in aircraft: (1) Uniformity of thickness in the same piece;
(2) smoothness and flatness; (3) freedom from large checks or other
defects (pi. 9, B) ; and (4) straight grain and absence of decay
(pl.9,D).
GLUING TECHNIC AS RELATED TO THE QUALITY OF GLUE JOINTS
SATISFACTORY AND UNSATISFACTORY JOINTS

The conditions involved in the gluing operation must be controlled

so as to produce a continuous film of solid glue in the joint with
adequate adhesion to both pieces of wood. These conditions involve
a sufficient spread of the glue on the wood surfaces and a correct
balance between the consistency of the glue at the time of pressing
and the amount of pressure used. A satisfactory joint and two com-
mon types of unsatisfactory joints, resulting from different gluing
conditions, are shown in Plate 10. The three joints were all made
from the same kind and quality of glue and wood, prepared in the
same way. The "starved" joint (pi. 10, B) illustrates the result
of one extreme, where the glue is too thin, and the " chilled " or
"dried" joint (pi. 10, C) the opposite, where the glue is too thick,
for the amounts of pressure used. A correct balance of the con-
sistency of the glue and the amount of pressure used gave the good
joint. (PI. 10, A.) The starved and chilled joints are most com-
monly produced with animal glue, although the former may occur
with any thin glue mixture. Dried joints may occur with all kinds
of glues.
SPREADING THE GLUE

To make a satisfactory joint it is necessary to spread evenly the

amount of glue needed, and for certain classes of work this should be
done within as short a time as possible. These requirements can
often be most easily met by machine spreading, but in the construc-
tion of aircraft from many small and irregular-shaped pieces, it is
frequently necessary to spread by hand. Low-viscosity glues, such
as warm animal glues, and most blood-albumin glues, spread easily
by either method, but thick casein glues are difficult to spread by
hand, and it is therefore best to use a machine spreader for them
whenever possible.
 GLUING WOOD IN AIKCEAFT MANUFACTURE 23

SINGLE AND DOUBLE SPREADING

In most commercial gluing operations the glue is spread on but

one of the two contact faces of the pieces being glued (single spread-
ing). At times, however, both contact faces are coated if the great-
est precaution is being taken to insure good joints. In order to
determine the comparative value of the two methods a series of tests
was made on plywood, a part of which was single spread with casein
glue and the other double spread. The results of the tests, given in
Table 4, indicate that the increase in strength by double spreading
is slight.
TABLE 4.—Comparison of restUtSi of tests on single and double spreading in
ffluing

Average! joint
strength obtained
when—
Method of spreading
Tested Tested
dry wet

Lbs. per Lbs. per

sq. in. sq, in.
Single - 369 225
Double 377 235

1 Each test value is an average of at least 75 specimens.

A more extensive test was made on thick laminations under a

wider range of gluing conditions. The results again showed that
good joints can be made by either method and that no great dif-
ference in strength of joints ordinarily occurs, whether one or both
contact faces are coated, if other conditions are satisfactory. How-
ever, under adverse conditions of gluing, as when the glue became
very thick before pressing, double spreading was found to be more
reliable.
QUANTITY OF GLUE SPREAD

Within certain limits the stren^h of the glue joint increases with
the quantity of glue spread. This is illustrated in Figure 10. The
data are from plywood glued with casein glue under good gluing
conditions. In other tests it was found that with less favorable
gluing conditions the quantity of glue spread affected the joint
strength to a still greater extent. The same general relationship
holds for other glues, although the optimum amoimt of spread and
rate of change in strength vary somewhat.
A spread of 1^ ounces of wet glue per square foot of single glue
line (about 38 square feet per pound of dry glue mixed 1 part glue
to 2 parts water) is satisfactory for most conditions and kinds of
gluing. A larger or smaller quantity may be used advantageously
under certain conditions. With casein and blood-albumin glues a
spread of more than 1^ ounces apparently neither lowers nor in-
creases the joint strength materially, whereas a smaller quantity nor-
mally produces inferior joints. Tests have shown that animal glue
has a wider range of satisfactory spread for maximum joint strength,
the amount to use being determined by the gluing conditions; good
24 TECHITICAL BULLETIN 2 05, U. S. DEPT. OF AGEICULTUBE

joints may be obtained with spreads of 1 to 1% ounces per square

foot if other conditions are adjusted accordingly. A larger quantity
of glue is required when both contact surfaces are coated and when
conditions are such that the glue dries considerably on the wood
before pressing. Warm wood, a long assembly time, and exposure
to the air are factors that bring about quick drying of the glue.
CONSISTENCY OF GLUE MIXTURE

The consistency of the glue at the time of pressing is perhaps the

most important factor in making satisfactory joints. However, the
consistency of glue after being spread on the wood is extremely
variable, depending upon such factors as the kind of glue, glue-water
proportion of the mixture, quantity of glue spread, moisture content
of the wood, temperature of the glue, room, and wood, the time elaps-
500 ing between spreading
and pressing, and the
Î^ 400 T£ST£Û DRY
extent to which the
glue - coated surfaces
• "^9 are exposed to the air.
The kind of glue and
the moisture content
TESTED WET ^ .^
of the wood should be
predetermined by con-
ditions of service. The
$ IOC
other factors may be
varied somewhat to
1 I— L 1 .. i
0(9 0.') 1.0 U U 1.5 l.f Í.5 1.6 1.7 suit the particular
QUANTITY OF GLUE SPREAD (OUNCES PER SQUARE FOOT OF GL(J£ LINE S PRE A ^
gluing operation.
FIGURE 10.—Relation between quantity of glue spread and It is very difficult
plywood joint strength. Tests made on 3-ply panels
with i^s-inch birch faces and ^-inch red gum cores to measure the viscos-
glued with casein glue. Joining pressure, 200 pounds ity of a glue mixture
per square inch ; assembly time, 3 to 12 minutes
after it is spread on
wood surfaces, but it is possible to judge of the consistency in an
empirical way by touching the glue layer with the fingers. An
animal glue, for example, should be thick enough to form short,
thick strings, but not too thick to take an imprint or depression
readily. Between the two conditions thus defined good results are
produced with a moderate amount of pressure. At the time of press-
ing a glue of the proper consistency will flow sufficiently to show a
distinct line of glue at the joint edge. (PL 10, A.) The absence
of a line of glue at the joint indicates a chilled or dried joint.
(PL 10, C.) If, on the other hand, the glue flows out and down over
the edges of the pieces excessively it mdicates the likelihood of a
starved joint. (PL 10, B.)
EFFECT OF ASSEMBLY TIME AND TEMPERATURE

Where pieces of wood are coated and exposed freely to the air
a much more rapid change in consistency of the glue occurs than
where the pieces are laid together as soon as the spreading (single
or double) has been done. The condition of free exposure is con-
veniently referred to as " open assembly," and the other as " closed
assembly."
Tech. Bul. 205, U. s. Dept. of Agriculture PLATE 9

QUALITY OF VENEER AFFECTS STRENGTH OF PLYWOOD

A, strong joints glued from straight-grained well-cut veneer; B, veneer badly cheeked in cutting,
not suitable for aircraft plywood; C, plywood test specimens, with core of checised veneer shown
in B, failed by breaking from check to check and gave low strength values; D, cross-grained
veneer, which should be rejected for aircraft use; E, broken plywood specimens with a cross-
grained face ply similar to D. The cross-grained breaks resulted in low strength values.
Tech. Bul. 205. U. s. Dept. of Agriculture PLATE 10

STRONG AND WEAK JOINTS RESULTING FROM DIFFERENT GLUING CON-

DITIONS
A, Well-glued joint, made with a proper relation between pressure and consistency of glue; B,
starved joint, resmted Irom the application of pressure while the glue was too thin. Occurs fre-
quently with animal and other thin glue mixtures on certain woods; C, chilled joint, glue chilled
excessively and amount of pressure was insufficient to bring about complete contact. Glue
that dries excessively before pressing gives the same result.
 GLUIK6 WOOD IN AIRCRAPT MANUPACTUEE 25
The effect of the time element on the strengfc of casein-glue
joints under closed-assembly conditions is illustrated in Figure 11.
The results shown are for eight groups of plywood panels spread
with glue in succession and pressed together in the same press,
fiving different assembly times. The assembly time was shortest
or the last panels spread. The consistency of the glue at the
shortest assembly periods was too thin and at the longest periods
had become too thick to give the best joint strength at the pressure
used. In other tests the use of somewhat lighter pressure at early
periods and a heavier pressure at later periods was found to give
more uniform joint strength, but such variation is, of course, im-
possible where all the panels are pressed at one time.
Of the glues used for woodworking, animal glue normally shows
the largest change in consistency during the gluing operation. This
is due primarily to
the striking effect of ^^^
cooling in increasing
the viscosity of an
animal - glue solution,
illustrated in Figure
12. The change of vis-
cosity of other glues
from the effect of
temperature is small TE3TBP WET
in comparison, even if
they are subjected to
the same temperature
changes, which is sel-
dom. However, with
all glues a rather indi-
rect effect of tempera- I >) I 1 I I I I I I
IZ le 20
ture is the thickening ASSEMBLY TIME (MiNUTäS)
of the glue solution
by drying. High tem- FIGURE 11.—Effect of assembly time on the joint strength
of plywood glued with casein glue. Tests were made
peratures, especially of on 3-ply panda with -¿"s-inch. birch faoes and i"s-inch
red gum cores. The joining pressure was 200 pounds
the wood, dry the glue per square inch
on the wood surfaces
more rapidly than do lower temperatures. For animal glue on wood
at high temperatures (120° to 140° F.) this thickening from drying
is the only means of bringing the glue to a proper consistency for
pressing, and the process is slow. The cooling of the glue is rapid on
wood at relatively low temperatures (70° F.) and becomes the con-
trolling factor in producing a proper consistency, the drying effect
becoming relatively unimportant. With animal glue, therefore, a
definite control of temperature is necessary or, if changes in temper-
ature occur, other conditions must be varied to prevent or compensate
for differences in consistency (p. 27).
Satisfactory joints are obtained with both casein and blood-
albumin glues when applied at room temperatures ranging from
about 70° to 90° F., and variations within this range do not require
any important change in other conditions, such as assembly time or
amount of pressure.
26 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGRICULTURE

The approxim^e ranges in time at which average mixtures of

aircraft glues reach a satisfactory consistency for pressing under

60 90 100 no IZO m i4o 160 160 170

TEMPERATURE {PEGREES ff)

FiGUBB 12.—Relation between temperature and viscosity of an animal glue that

meets aircraft specifications. The tests were made on a solution mixed 1 part
of dry glue to 2% parts of water

open and closed assembly and different temperatures are shown in

Table 5.
 GLTJIKG WOOD IIT AIRCRAFT MANUFACTURE 27

TABLE 5.—Ramage in assembly time for différent glues *

Temperature of— Approxi-

Kind of glue Manner of assembly , mate as-
sembling
Boom Wood time

op Minvtes
Casein -- Closed 70 70 0 -20
Do . - Open 70 70 0-8
Blood-albumin Closed 70 70 6 -20
Animal 2 .. do 70 70
Do
Do
-
-
do
do
80
90
80
90
,,^i
10 -20
Do . _. Oüen . 70 70
Do. „1 do 80 80 i^:k ■

1 Conditions other than those given in the table were assumed as follows: Customaiy mixtures of the
glues, spread 1}4. ounces per square foot and pressed with 200 pounds pressure per square inch of joint area^
a Animal glue of the minimum properties required in U. S. Army Specif. No. 3-140 (1927) and U. S. Navy
Dept. Specif. No. 52G4C (1925).

The foregoing combinations of assemblies and temperatures are

given as a guide and may be varied, provided other conditions are
changed. For example, mixing a glue solution thicker than custom-
ary would lower the time limits between which the joints can safely
be pressed. A pressure of more than 200 pounds per square inch
would extend somewhat the maximum time limit, while a pressure
less than 200 pounds per square inch would lower the minimum time
limit. In general, longer assembly times are advisable with thin glue
mixtures, heavy spreads, and high pressures, and shorter assembly
times may be used successfully with thicker mixtures of glue and
lighter spreads and pressures.
In gluing small parts at room temperatures it is ordinarily not
difficult to apply the pressure when the glue has a proper consistency
for moderate pressure. In- larger operations, as in gluing propellers
or plywood, the lapse of time varies considerably, and consequently
the consistency of the glue spread first and that spread last in the
same construction or stack of panels is not the same. In such cases
a set of conditions must be selected that gives a sufficient time range
(p. 41).
Temperature, aside from its effect upon consistency and the result-
ing strength 01 the joint, also influences the rate at which the glue
sets and at which the joint gains strength under pressure (p. 32).
PRESSING AND CLAMPING

Pressure on the joint during the early setting of the glue is required
for best results in practically all types and forms of gluing. The
functions of pressure are to sprea(i the glue out into a continuous
film between the wood layers, to force air from the joint, to bring
the wood surfaces into intimate contact with the glue, and to hold
them in this position during the setting of the glue.
METHODS OF APPLYING PRESSUBE

The methods employed in applying pressure to glue joints in air-

craft manufacture range from the insertion of brads and screws to
the use of powerful hydraulic and electric power presses. Figure 13
illustrates some of the means used for applying pressure by hand.
28 TECHNICAL BULLETIN" 2 0 5, U. S. DEPT. OF AGEICULTURE

In a large number of joints in the wing structure the insertion of

brads and small nails is the method commonly used. (Fig. 13, A.)
The spring clamp (fig. 13, B) is used some in gluing small, narrow

FIGURE 13.—Hand devices used in applying pressure to glue joints: A, Brads or

nails ; B, spring clamp ; C, eccentric clamp ; D, wood clamp ; E, C clamp ; F, jack-
screw; G, bar clamp

joints, and the eccentric clamp (fig. 13, @) in applying pressure to

the flanges of box beams when gluing the reinforcing blocks in place.
 GLUING.WOOD IN AIRCRAFT MANUFACTURE 29
Wood clamps and C clamps (fig. 13, D and E) are extensively used
in gluing spars, spar ñanges, reinforcing blocks, and bow ends.
(Pis. 11 and 12.) The larger laminated members, such as spars and
propellers, are usually pressed under jackscrews (fig. 13, F)
mounted on frames (pi. 12, A) where several are used on the same
joints. The bar clamp (fig. 13, G) is useful when gluing pieces edge
to edge.
The amount of pressure that can be applied with different types
of equipment varies enormously. The amount of pressure applied
by a single brad may be only a matter of pounds whereas several
tons may be applied with the jackscrews. Tests made on some de-
vices of the types illustrated in Figure 13 have provided the data
shown in Table 6, which may serve as a guide in the use of similar
pressure equipment. Where the screws of clamps and presses differ
from those illustrated in Figure 13 and Table 6 the approximate
loads may be calculated from the formula given in footnote 2 of
Table 6. In gluing plywood, power presses, either of the hydraulic
or electric types (pi. 13), are commonly used in which loads up to
several hundred thousand pounds can be applied. Such presses are
usually equipped with pressure gauges.
TABLE 6.—Pressure data on screw cldmps and other devices used in gluing

Force Length Pitch of Diam- Total Coeffi-

Equipment tested applied of lever screw eter load 1 cient of
arm of screw friction

Pounds Inches Inches Pounds

Jackscrew 1170. 0 37 IMe 33,850 2 0.1978
Do 1170.0 18 IMe 16,350 2.1997
Do 1170. 0 31 2^6 16,720 2.2000
Do 1140. 0 7 7,500 2.2498
C clamp
Do
2 70.0
2 69.6
m
3
1,585
2,700
3.20
3.20
Do 2 70.2 2 1,370 3.20
Bar clamp 2 79.9 2H 2,810 3.20
Wood clamp <. 2 71.3 720 3.20

1 Measured in test.
2 Calculated from the formula FIf=WRm i Dj^^fj^ ) t which is applicable to square-thread screws.
Where i?'=force applied to lever arm in pounds. •
i=length of lever arm in inches.
W=load in pounds.
127»=mean radius of screw in inches.
/■=coefficient of friction.
^=3.1416.
i)77i=mean diameter of screw in inches.
íL'= pitch of screw (single, thread) or lead (multiple thread).
8 Assumed from results of previous tests.
* Screws with V-type threads; hence calculation from formula is only approximate.

To secure the best results in gluing, the pressure should be dis-

tributed uniformly over the joint area. High pressure at certain
points and low pressure at others results in weak spots or areas in
the joint. Blocks and cauls are frequently used between the clamp
or press and the layers being glued to distribute the load from the
point of contact to other parts not directly under the load. This is
particularly necessary where thin plies are glued. Obviously such
members must be true and uniform in dimension or they do not
fulfill their purpose. Other principal causes of unequal pressure
on joints are: (1) Irregular surfaces of the wood pieces being
30 TECHNICAL BULLETIN" 2 05, XT. S. DEPT. OF AGKICULTUBE

glued; (2) unequal thicknesses or width of stock; (3) improper

spacing of the pressure-bearing members; and (4) deflection and
other imperfections in press, clamps, or other pressing equipment.
Where heavy loads are applied, as in gluing propellers and ply-
wood, large deflections of the press members may, and frequently do,
occur. This is most often true with heavy screws of the type shown
as Figure 13, F, and where I beams and retaining clamps are used
with bundles of veneered panels for retention of pressure after re-
moval from the press. (PI. 13, A.) The presses should be designed
rigid enough and the retaining I beams should be stiff enough and
spaced sufficiently close so that undue deflection will not occur.
Likewise, the platens and other large» bearing surfaces of large
presses should be smooth, true, and parallel.
The load originally applied to glue joints tends to drop as the
glue squeezes out from between the wood layers and distributes itself
in the joint. Care should be used that the correct amount of pres-

ROOM AND WOOD AT %'F. ROOM AND WOOD AT W'F

AVERAGE VALUES

O
MINIMUM VALUÍ5
MINIMUM VALUES

_J I I L_
0 200 400 600 ZOO
JOINING PRESSURE (POUNDS PER SQUARE INCH)

FIGURE 14.—Relation between pressure and joint strength. Tests were made on
yellow birch and animal glue.. Shaded portion of average points indicates per-
centage of wood failure developed in testing joints. (A large percentage of wood
failure indicates that the joints are well made)

sure is maintained. It is therefore necessary to continue adjusting

the clamps or other devices for a short time after the original appli-
cation of the load.
AMOUNT OF PBESSUBE

A light pressure should be used with a thin glue, a heavy pres-

sure with a thick glue, and corresponding variations in pressure
should be made with glues of intermediate consistencies. In Figure
14 are shown thç results of joint tests made on 1-inch stock with
animal glue, which illustrates the relation between the amount of
pressure and the resulting average joint strength. From these and
other data the curve of Figure 15 was derived, showing pressures
recommended by the Forest Products Laboratory for different con-
sistencies of glue.
While it is possible to make strong joints with pressures of less
than 50 pounds per' square inch, such pressures are generally not
 GLUING WOOD IN AIKCRAFT MANUFACTURE 31
feasible under commercial conditions. The successful use of light
pressures presupposes that the wood surfaces are free from warping
and other irregularities—a condition that seldom prevails. Pres-

soir MEPIUM HIGH VERY HIGH

VISCOSITY OF CLUE AT TIME OF PRESSING
FIGURE 15.—Joining pressures recommended for glues of varying consistencies. Low
viscosity, represented by thinly mixed, warm animal glues and thinly mixed blood-
albumin glues ; medium viscosity, represented by thickly mixed blood-albumin
glues, slightly cooled animal glues, and thin casein glues; high viscosity, repre-
sented by thick casein glues, semijellied animal glues, and blood-albumin glues ;
very high viscosity, represented by firmly jellied animal glues

sures in excess of 200 pounds per square inch may crush certain
woods, and pressures as high as 400 pounds per square inch are
applicable only to the strongest species of wood. In general, ex-
32 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGRICULTUEE

perimental and practical work has justified the use of a pressure of

about 200 pounds per square inch and a medium-high viscosity of
glue where maximum strength of joint is required.
The occurrence of an occasional weak joint or part of a joint,
which may cause failure in service, is of great significance to the
builder of aircraft. The selection and control of gluing conditions
so as to avoid weak spots in the joints is of even greater importance
than the average strength of the joints. Figure 14 shows the effect
of the amount of pressure on minimum values in joint tests made
with animal glue.
Figure 16 shows the gluing conditions under which low strength
joints are most likely to occur when gluing with casein glue.^^

-® 0- <$>-

-&) &} ■0

-0——é> -0
k>——4
'o-(§)^

/l5S£MßCr (At/f/UT£S)

FIGURE 16.—Occurrence of low joint strength in relation to gluing conditions with

casein glue. Figures in circles represent the number of low-strength joints occur-
ring with each combination and pressure and a&sembly time.

Twenty-four species of wood were glued in these experiments, tinder

16 combinations of gluing conditions. Ten test specimens were
made from each species under each gluing condition which made a
total of 160 specimens for each species and 3,840 for the 24 different
woods. The lowest 10 per cent of the strength values of each species
were separated and arranged under the conditions producing them,
excluding, however, the values where failure was completely in the
wood. The largest number of weak joints occurred with long
assemblies and low pressures, and the smallest number with the short
assemblies and high pressures.
DURATION OF PRESSURE

Joints should be retained under pressure at least until they have a

sufficient strength to withstand the internal stresses tending to sepa-
16 The glue mixture contained about 10 per cent less water than shown in formula
No. 4B (p. 54) and was slightly thicker than an average casein glue.
 î
S-

TJ
r

A Box SPAR GLUING OPERATION

A, Laminated flanges being glued; S, conditioned; C, worked to size, and B, assembled preparatory to gluing on the plywood sides. Pressure is applied to the spar flanges in gluing
by means oí C clamps.
Tech. Bul. 205, U. S. Dept. of Agriculture PLATE 12

SPAR CONSTRUCTION
A, An aircraft factory press loaded with laminated si)ars of solid construction. Two pieces of spruce
are glued together to form one spar. Pressure per square inch of joint area is from 125 to ISO pounds;
B, gluing reinforcing blocks to the flanges of a box spar. Pressure, appUed by means of wood
elami)s, is insufficient and not properly distributed for making the best joints: C, a box spar show-
ing laminated flanges, .5-ply reinforcing block, and 3-ply veneer sides. The plywood sides were
pressed to the framework of the spar by the use of brads only.
Tech Bul. 205, U. S. Dept. of Agriculture PLATE 13

POWER PRESSES USED IN GLUING PLYWOOD

A, Hydraulic, cold press showing stack of panels with I teams and retaining clamps in place;
B, hydraulic, hot plate press used in the manufacture of blood-albumin glued plywood.
 GLUING WOOD IN AIKCRAFT MANUFACTURE 33
fate the wood pieces. It is safe to assume that under favorable
gluing conditions this stage will be reached in from two to seven
hours, according to the thickness and absorptive power of the wood.
A pressing period beyond the minimum is advisable as a precau-
tionary measure where operating conditions permit.
The rate at which the joints gain strength is the principal factor
determining the length of time in the press. Joints increase in
strength mainly as a result of drying of the glue layer, and drying is
in turn aflFected by several factors. The quickest release of pressure
is possible when a fast-setting glue, a thin spread, and warm, dry,
thick layers of wood are used in a warm room.
Z600

-
^Z400
'^
s
%2000 • •
- ^^^J¿^— O

-
I
^ /200
•/
^^^""^^ ^

Xo SITKA SPRUCr^
O

Ï ^^"-^ ^

I 600
°Á /^
*

I -
- /7
'/
LEGENP
• -fiN/MAL GLUE
0-CA5EIN GLUE

Ö/
1 1 1 1 f 1 1 1 1 1 1 1 t . „1. 1 1 1
'O .1 Z 3 4 5 e 7 6 9
JOINT ME (HOUße)

FIGURE 17.—Strength-time relation for animal and casein glue joints. Tests were
made on %-inch hard maple and Sitka spruce as soon as possible after joints were
taken from press. Each value is an average of at least 24 test specimens from
6 or more joints

Figure 17, showing the increase of strength of joints with time of

pressing, is based on tests of animal and casein glues. The gluing
was done under conditions favorable to the production of strong
joints. The rates of increase in strength of the two glues are ap-
proximately equal when the glues are used in a room and on wood at
a temperature of about 70° F. Animal-glue joints made in a
warmer room and on heated wood showed a more rapid increase in
strength, and in all probability the same would be true for casein
glue. The higher temperatures apparently increase the drying rate,
but they are also known to be a serious cause of lower final strength
of joints. In the tests represented in Figure 17 the glue was as
strong in shear after 6 hours of pressing as the maple wood ad-
jacent the glue line. In the tests with spruce, the glue was as strong
as the wood in 4 hours. The strength shown for sugar-maple
joints at the end of about 8 hours of pressing by no means repre-
119542^—30 3
34 TECHÎiriCAL BULLETIN 2 0 5, U. S. DEPT. OF AGRICULTURE

sents the final strength of the joints. A further conditioning, ex-

tending in all over a 7-day period, permitted an equalization of
moisture which resulted in an average joint strength on sugar maple
of approximately 3,000 pounds per square inch. In other tests of
the same series, joints, which were pressed for 2 hours and then
allowed to season for 22 hours, proved as strong as those that had
been pressed for 24 hours. Sugar-maple joints pressed for only
one-half hour and seasoned for 24 hours, although of fair average
strength, showed greater variation and would not be so dependable
where uniform maximum strength is important.
In gluing veneer the early setting of the glue is about as quick as
in gluing thick joints, but beyond this the veneer joint appears to
increase in strength more slowly. The thin layers of wood do not
continue to absorb the moisture as rapidly from the glue as the
thicker stock, and a longer pressing period is therefore required.
Plywood is usually left under pressure for 18 to 24 hours, but this is
determined more by routine of operation than by the rate of setting
of the glue. ^ Pressing plywood beyond 5 to 7 hours appears to be
unnecessary if it has been properly glued, although the customary
longer periods are not harmful and insure an additional margin of
safety.
USE OF HEAT IN PRESSING

With most blood-albumin glues hot-plate presses are required in

making joints. (PI. 13, B.) The glue must be brought to a mini-
mum temperature of about 160° F. to render the joint water-
resistant. Heat and pressure are applied.at the same time by means
of steam-heated platens. The temperature of the platens usually
ranges between 212° and 300°. With certain formulas temperatures
of about 300° give the strongest and most highly water-resistant
joints, but at the same time they are liable to develop blisters or
steam pockets in the panels during the pressing process. The time
required in pressing varies with the temperature of the platens and
thickness and kind of the material being pressed. The variation in
time in actual practice is from 2 or 3 minutes to as much as 30
minutes.
Casein glues for use in aircraft are usually pressed cold. Ex-
periments have shown, however, that the quality and water resist-
ance of casein-glue plywood can be substantially improved, by the
application of heat. Hot pressing after the joints had set but
before they had dried resulted in a somewhat greater improvement
than hot pressing at the time of gluing. With some brands of
casein glues the wet strength of the resulting joint was increased
by 35 to 40 per cent.
GLUING DIFFERENT SPECIES AND SURFACES OF WOOD
SPECIES GLUED

A very large part of the wood used in aircraft is Sitka spruce

which, fortunately, is comparatively easy ,to glue. Ash, basswood,
black walnut, mahogany, hard maple, white oak, and yellow poplar
are also used and glued to some extent. Spruce does not require so
 GLUING WOOD IN AIRCRAFT MANUFACTURE 35

careful control of gluing conditions as the other species in order

to make joints of maximum strength. Ash, black walnut, birch,
hard maple, and white oak will require very careful control of the
gluing operation to obtain the best results.
Gluing tests have been made at the Forest Products Laboratory
on about 40 different species, including those used most in aircraft
construction. The woods were glued with animal, casein, and vege-
table glues under a wide range of pressures and assembly times ^^
and the joints were tested in shear in the manner illustrated in Fig-
ure 5. The joints which showed no marked or general vari-
ation in strength were taken as a basis for comparing the gluing
characteristics of the different species of wood. From these tests
definite recommendations have been made for the gluing of each
species.^^ The results obtained with animal and casein glues are
applicable to aircraft gluing.
In Figure 18 the average joint strengths of 31 species, which have
been glued in aircraft or which from their properties and abundance
may warrant consideration for aircraft purposes, are plotted against
the specific gavity of the wood tested. The strength of wood is
known to vary in a general way with its density, and from Figure
18 the same relation is^seen to exist for glued joints, even where
complete wood failure did not occur in testing.^® The amount of
wood failure occurring during test over the glue-line area was gener-
ally in reverse order to the strength of the points and the specific
gravity of the woods. For example, Sitka spruce joints showed
practically 100 per cent average wood failure (0 per cent of glue-line
failure), mahogany about 80 per cent, and ash and birch each about
50 per cent.
RECOMMENDATIONS FOR GLUING SIDE-GRAIN SURFACES OF VARIOUS SPECIES

WITH CASEIN GLUÏ3

For gluing side-grain surfaces with casein glue the various species
listed in Figure 18 are divided into two groups according to the
conditions applicable, as follows :
GUdnst conêitîonê recovnmendeë
GBOTJP 1
Cedar, western red
Fir, white Glue of medium consistency, spread 1% to 1^
Hemlock, western ounces per square foot, pressure from 100 to
Pine, northern white- 150 pounds per square inch, and assembly
Pine, western yellow. time 0 to 20 minutes closed, or 0 to 8 minutes
Redwood open.
Spruce, Sitka —

1'^ Pressures of 100 to 400 pounds per squaie inch and assembly times of 1 to 25 minutes
were used. Since 400 pounds pressure would have crushed some of the weaker woods, the
highest pressure used was the maximum that the wood would stand without injury.
IS For a more complete discussion of results and recommendations for the gluing of the
40 species see The Gluing of Wood (24)-
1» In the type of specimen used in testing the glued joints (fig. 5) the unit stress in the
glue is higher than in the wood because the shearing load is borne by a larger area of
wood than of glue line. Furthermore, the stress may be very close to the ultimate
strength of the wood, and yet failure may occur chiefly or entirely in the glue line. For
these reasons a high percentage of glue failure (low percentage wood failure) may occur
In test and still the strength of the joint be as high as that of the wood itself^
36 TECHNICAL BULLETIN 2 0 5, IT. S. DEPT. OF AGRICULTURE

Species Olumff conditions recommended

GROUP 2
Ash, white
Basswood
Beech
Birch, yellow
Cedar, Alaska
Cherry, black
Cottonwood
Elm, American
Elm, rock
Douglas fir Glue of thick consistency, spread 1% to li^
Gum, black ounces per square foot, pressure 150 to 200
Gum, red (heart). pounds per square inch, and assembly time
Gum, red (sap)__. 0 to 12 minutes closed, or 0 to 6 minutes
Gum, túpelo open.
Hickory
Mahogany..
Magnolia
Maple, soft
Maple, hard
Oak, red
Oak, white
Poplar, yellow
Sycamore
Walnut, black
Occasional weak joints in red gum (heartwood), black cherry, and
hickory are obtained even when glued under the conditions recom-
mended. This trouble can be largely avoided by treating the wood
before gluing in the manner described on page 38. Sitka spruce,
which is used extensively in aircraft, is very satisfactorily glued
under the conditions recommended for Group 2 woods as well as
those for Group 1, in which it is listed. On the other hand, white
ash, birch, or hard maple is less satisfactorily glued under some of
the conditions recommended for Group 1 woods and should, there-
fore, be glued under the more restricted conditions of Group 2.
WITH ANIMAL GLUE

The principal species of woods glued with animal glue in aircraft

construction may be placed in a single group in so far as recom-
mendations for gluing practices are concerned. The following
species include the woods used for propellers and a few others
requiring similar gluing technic: Ash, yellow birch, black cherry,
red gum,, magnolia, mahogany,^^ soft maple, hard maple, red oak:,
white oak, and black walnut. Gluing conditions recommended for
these species of wood are shown in Table 7.
2» The gluing conditions required for mahogany are less rigid than for the other species
listed, but it glues satisfactorily under the conditions recommended.
 GLUING WOOD IN AIRCRAFT MANUFACTURE 37
TABLE 7.—Gluing conôitions reoommendeá} with animal glue for wooäs u^ed in
aircraft

Approxi- Temper- Closed

Glue-water ^ proportions, by weight mate glue ature of Pressure assembly 2
spread the wood time

Ozs. per sq. Lbs. per sq.

in. MirnUes
lto2^ -- ^ 70 150-200
Do -- 80 150-200 3-5
Do - 85 150-200 8 -15
Do - 90 150-200 10 -18

1 A glue which meets the requirements of Ü. S. Army Specification No. 3-140, and Navy Department
Specification No. 5204C.
2 Wood pieces are laid together as soon as spread with glue.
3 Weight of wet glue mixture applied per square foot of glue line.

LEGEND

• 'ANIMAL CLU£
O-CASEIN CLUE

I-ASH, WHITE
&
Z-BASS WOO P
3-BEECH
4-BIRCH, YELLOW
5-CEPAR, ALASKA
6 -CEPAR, WESTERN REP
7-CHERRY, BLACK
3-COTTONWOOD
9'POUGLAS FIR
lO-ELM, AMERICAN
II-ELM, ROCK <f
12-FIR, WHITE
13-GUM, BLACK o"
14-GUM, REP (HEARTWOO0
15-GUM, REP J^fiPWOOP)
é à*
'^ lb-CUM, TUPELO
17-HEMLOCK, WESTERN
lb-HICKORY
19-MAGNOLIA
20-MAHOGANY
é'
21-MAPLE, SOFT
22'MAPLE, HARD é'
23-OAK, REP
24-OAK, WHITE
25-PINE, NORTHERN WHITE
Zb-PINE, WESTERN YELLOW
21-POPLAR, YELLOW
28-REPWOOP
J^O
I
2')-SPRUCE, SITKA
30-SYCAMORE
31-WALNUT, BLACK A5 • m!^

O?
^
é"
d>

0.45 0.50 0.55

SPECIFIC GRAVITY

FIGURE 18.—Relation between specific gravity of wood and shear strength of joints
of various species glued with animal and casein glues
38 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGEICULTURE

Satisfactory gluing may be performed at other temperatures than

those indicated in Table 7, provided an adjustment is made in the
assembly time or other conditions. Heating the wood above 90° F. is
ordinarily not advisable. Where it is impossible to apply the full
amount of pressure recommended, the assembly time should be short-
ened or the temperature somewhat increased.
Because of the variability of results with red gum (heartwood) and
black cherry, these species should not be used where maximum
strength of joint is of primary importance unless treated in the
manner described in Table 8.
TREATING THE WOOD BEFORE GLUING

With woods that are naturally difficult to glue or where conditions

are not the most favorable to the production of strong joints it is pos-
sible to bring about stronger joints by treating the wood before
gluing.
With animal glue, caustic soda proved to be one of the most effiec-
tive of several treating materials tried out experimentally. The wood
surfaces to be joined were brushed with a 10 per cent solution of caus-
tic soda, wiped after* a few minutes to remove any dissolved material
or excess of the solvent, allowed to dry, and then glued under stand-
ard gluing practice. The data in Table 8, derived from joint tests of
yellow birch (heartwood) glued with animal glue, indicate the im-
provement resulting from the caustic-soda treatment.
TABLE 8.^-Result» of caustic soda treatment on joints of yellow 'birch (heart-
wood) Œtid am/mal fflue^

Shear Wood Specific

Condition of si)ecimen strength failure gravity
of joint of wood 2

Lbs. per
sq. in. Per cent
Wood untreated (good Eint conditions) 2,700 0.66
Wood untreated (starved joint conditions) 1,910 .66
Wood treated with 10 per cent caustic-soda solution (starved joint condi-
tions) ___ _ 2,770 .62

1 Each test value shown represents at least 10 specimens.

3 Specific gravity based on oven-dry weight and volume at 7 per cent moisture content.

Other species, which showed an improvement in joint strength

when treated before gluing with animal glue were: Black cherry, red
gum (heartwood), red gum (sapwood), hard maple, red oak, and
white oak.
With casein glue, hydrated lime (10 grams added to 90 grams of
water) gave slightly better results than caustic soda when used as a
surfacie treatment for black cherry, hickory, and red gum. Hickory,
which is difficult to glue with casein glue, produced appreciably
better joints after treatment with either lime water or caustic soda.
GLUING END-GRAIN SURFACES

The methods, practices, results of tests on joints, and recommenda-

tions which have thus far been presented relate more specifically to
the gluing of side-grain surfaces of wood. Such surfaces are in-
 GLUIÎ^G WOOD m AmCüAFT MÁÍ^tTPÁCTtfBE 39

volved exclusively in plywood and laminated constructions. (Fig.

9, A to G.) Joints between side-grain surfaces in most species can
be made as strong in shear parallel to the grain, tension across the
grain, or cleavage as the wood itself. The highest stresses developed
in these joints do not exceed 3,000 or at the most 4,000^ pounds per
square inch.
The gluing of end^grain surfaces, on the other hand, is not accom-
plished with the same degree of success. Straight end-grain butt
joints are rarely attempted in any type of construction, and where
wood is subjected to tension stresses parallel to the grain, joints of
this type can not be depended upon to develop more than a small
part of the strength of the wood.
Most North American species of wood are capable of withstand-
ing 6,000 to 20,000 pounds per square inch in tension parallel to
the grain. Tests made in gluing straight end-grain surfaces have
shown that such joints are erratic and rarely exceed about 3,000
or 4,000 pounds per square inch in strength. Their strength is
limited by several factors including: (1) Structure of the wood, (2)
penetration of the glue, (3) air bubbles in openings of the wood,
(4) quality of glue, (5) consistency of glue, (6) application of glue,
(7) amount and duration of pressure, and (8) shrinkage of glue.
Since factors 1, 3, and 8 are largely beyond operative control, any
improvement in strength must be brought about primarily through
the other factors. High-grade glue, insuring maximum tensile
strength (fig. 1), should be chosen. In general, a thick consistency
of glue solution gives better results than a thin one, so that the pro-
portion of water in the mixture should be somewhat less than for
side-grain gluing.
With even the most careful gluing of straight butt joints not
more than about 25 per cent of the tensile strength of the wood
parallel to the grain has been obtained in tests. It is evident, there-
fore, that in order to obtain a tensile strength of the various species
that is greater than 25 per cent of the tensile strength of the wood,
a scarf, serrated (fig. 9, H and I), or other form of joint must be
used instead of plain end gluing. Where it is necessary to elongate
members, such as longerons or beams, some such form of joint is
recommended. The plain scarf (fig. 9, I) is perhaps the easiest to
glue and involves fewer machining difficulties than the many angle
forms (fig. 9, H).
DRYING AND CONDITIONING GLUED STOCK
The gluing operation adds moisture to the wQod. Glue that has
set in joints contains only a part of the water added at the time of
mixing, the remainder being absorbed by the wood or removed by
evaporation. The absorbed moisture must be allowed to dry out or
to distribute itself through the wood in order to insure the full
strength of the joint and to reduce the tendency of the glued member
to warp.
Table 3 (p. 18) shows the percentages of moisture added to differ-
ent constructions in gluing, calculated for a specified glue mixture
and sprçad. No allowances have been made for surface evaporation,
but the computed percentages are checked closely by actual deter-
minations.
40 TECHNICAL BTTLLETUsT â05, tT. S. ÍMPT. OF AORICÜLÍ^ÜBE

In propellers (fig. 9, F) and similar constructions, glued from

thick laminations, the moisture from the glue need not be eliminated
but may simply be allowed to distribute itself throughout the con-
struction. Even complete equalization would require a very long
time. For birch and white oak in laminations about three-iourths
of an inch thick and under average room temperatures, a condition-
ing period of 7 to 10 days is sufficient. For quicker-drying woods,
such as spruce, which permit a more rapid distribution, a 2-day
period should suffice under most conditions. However, where con-
structions are glued from laminations one-fourth of an inch or less
in thickness they will normally contain too much moisture and
should be dried for one to three weeks or longer depending upon
their thickness and width. A spruce bow end, for example (fig.
9, E), will usually dry sufficiently at room conditions during 7 to
10 days, but a thick laminated spar flange would require two or
three times as long.
In plywood it is advisable to dry out by artificial methods a part,
at least, of the moisture added in gluing, until the construction has
reached the average service moisture content. Drying plywood to
an excessively low moisture content materially increases warping,
opening of joints, and other defects. It is therefore recommended
that for the general requirements of aircraft, plywood be dried to
an average moisture content of 10 to 12 per cent. This is somewhat
lower than the average found in service (Table 2) but is higher than
the moisture content in most fabricating shops in winter.
Tests have shown that it is possible to dry panels three-sixteenths
of an inch or less in thickness very easily in 8 to 16 hours under rela-
tively mild conditions. Satisfactory results may be obtained with a
rather wide range of drying temperatures, sucîfi as 70° to 140° F.,
îf proper relative humidities are used.
Table 9 shows several combinations of temperatures and relative
humidities, which will produce a moisture content of 8 to 12 per
cent in plywood panels within a reasonable drying pe^od. If ex-
posed for an indefinite period to the conditions outlined in Table
9, however, the plywood will come to a moisture content somewhat
lower than those indicated. (Fig, 3.)
TABLE 9.—Combmations of temperatures and relative hu/miMties^ suitable for
producing a moisture content of 8^ 10, or 12 per cent in plywood panels

Relative humidities for use with different temperatures

Moisture content desired, per
cent
70° F. 80° F. 90° F. 100° F. 110° F. 120° F. 140° F,

Per cent Per cent Per cent Per cent Per cent Per cent Per cent
8__ _ 30 31 32 33 35 37 41
10_ _ 43 44 45 46 48 50 53
12_ 55 56 57 58 69 61 65

RECOMMENDED PRACTICE FOR THE PRINCIPAL AIRCRAFT

GLUING OPERATIONS
The following pages give a summary of information and, specific
recommendations on important details in the principal operations in
the gluing of: (1) Propellers; (2) plywood; (3) laminated spars,
 GLUING WOOD IN- AIRCRAFT MAKUFACTURE , 41

spar flanges, bow ends, pontoon ribs, and reinforcing blocks; (4)
scarf-type joints; (5) wing ribs; (6) box beams; and (7) general
assembly.
Necessary prerequisites in all gluing operations are properly dried
and machined stock, and equipment that spreads the glue and applies
pressure uniformly over the joint. These requirements must be met
or weak joints will result despite the most careful control of other
conditions.
PROPELLERS

Propellers (pi. 3, B) are usually made in special propeller plants.

Their gluing and shaping requires equipment not usually available
in aircraft factories.
Most propellers are glued with animal glue. Casein glues are
used only to a small extent, since their water resistance for this
purpose is of little value. Birch, white oak, black walnut, and
mahogany are the principal species of wood used.
A propeller-glumg operation requires several minutes for com-
pletion which, combined with the difficulty of gluing the species
used, necessitates careful control of the gluing conditions. Since
animal glue is largely used in propellers, most of the following
details are applicable to animal glue :
GLUING BOOM

The gluing room should preferably be separated from the rest of

the factory. A room temperature of 70° to 75° F. is recommended,
but gluing may be successfully done at room temperatures of 85° to
90°. However, such high temperatures are uncomfortable for the
workmen and are unnecessary.
GLUE

Glues conforming to the requirements given in United States

Army and Navy Specifications Nos. 3-140 and 52G4C, respectively,
are recommended.
GLUE-WATER PEOPORTION

One part dry glue to 21/4 parts water (by weight) should be used
for a glue tl|^t meets the minimum requirements of the foregoing
specifications. (See preceding paragraph and p. 16.)
MOISTURE CONTENT OF WOOD

A moisture content of the wood at the time of gluing of approxi-

mately 8 per cent (p. 19) is recommended.
TEMPERATURE OF THE WOOD

Wood that is at approximately room temperature should be

warmed at 85° to 90° F. for about one hour before applying the
glue. Higher temperatures and shorter heating periods can also be
used, but higher temperatures are more apt to lead to trouble.
GLUE SPREAD

One and one-fourth ounces of glue mixture should be spread uni-

formly per square foot of glue line. (If both sides of the joint are
42 TECHKICAL BULLETIIT 2 0 5, U. S. DEPT. OF AGEICULTURE

spread, 25 per cent more glue should be used.) A more uniform

spread of glue is obtained by the use of mechanical spreaders than
by hand spreading. The glue-coated wood should be laid together
as soon as spread (closed assembly). (P. 24.)
ASSEMBLY TIME

A closed assembly time of 8 to 15 minutes or 10 to 18 minutes

with wood warmed at 85° or 90° F., respectiv^ely, is recommended.
Pressure should be applied not sooner than 15 or 18 minutes after the
first lamination is coated with glue, but not later than 8 or 10 minutes
after the last lamination is spread. (Table 5,)
PRES SUEE

A pressure of 150 to-200 pounds per square inch of single-joint

area should be applied for a minimum time of 5 hours. Pressure-
distributing blocks, cut from trimmings from the laminations being
glued, and cauls should be used to equalize the load over the joint
areas. (Pp. 27 to 29.)
CONDITIONING

The glued blocks should be conditioned at room temperatures for

at least two days before rough carving and five additional days
before finish carving (p. 39).
PLYWOOD

Aircraft plywood is usually manufactured in plywood plants

where large glue spreaders and power presses (pi. 13, A and B),
not usually found in aircraft factories, are available.
Either blood-albumin or casein glues are successfully used in the
manufacture of aircraft plywood. While a number of species of
wood are permitted in aircraft-plywood specifications ^^ the list of
woods actually used in the past has consisted almost entirely of
basswood, birch, mahogany, Sitka spruce, and yellow poplar.
Methods and practices that produce high-grade, water-resistant
plywood have been developed (^, W). The hot-press method is
used with blood-albumin glues and either the hot-press or the cold-
press method with casein glues (p. 34). The following is a sum-
, mary of recommended conditions for gluing aircraft plywood :
GLUING ROOM

Ordinary room temperatures and shop conditions are satisfactory.

GLUE

Highly water-resistant types of blood-albumin or casein glues

(p. 2) are recommended.
21 The species permitted in plywood in the U, S. Army and Navy specifications are driven
in U. S. Army, Plywood, Aircraft, Specif. No. 82-6-A, 7 p., illus., 1928, and in Ü. S. Navy,
Plywood (Aircraft Use), Dept. Specif. No. 39P13, 6 p., illus, 1929.
 GLUING WOOD I]Sr AIBCRAFT MANUFACTURE 43

GLUÎ3-WATER PROPORTION

A glue mixture of comparatively thick consistency (1 part by

weight of dry glue to 1.8 to 2 parts of water) is recommended for
most prepared casein glues. Spruce glues well with a somewhat
thinner mixture than required for hardwoods (p. 35).
QUALITY OF VENEER

Smooth, tightly cut veneer that is free from bad cross grain and
rot (p. 20) is recommended.
TEMPERATURE OF "Vy^OOD

Wood that is of ordinary room temperature is satisfactory.

GLUE SPREAD

One and one-fourth ounces of glue mixture per square foot of joint
area should be applied evenly with machine spreaders, single spread
(p. 23).
ASSEMBLY TIME

For blood-albumin glue an assembly time of 5 to 20 minutes (closed

assembly) is recommended. For casein glue an assembly time of 3
to 15 minutes (closed assembly) is recommended. On spruce wood
the assembly time may safely be extended to 18 or 20 minutes (p. 35).
PRESSURE

A pressure of 150 (for spruce and basswood) to 200 (for birch)

pounds per square inch of panel area should be applied in the cold-
press process for a minimum time of 5 hours. For the hot process,
the pressing time depends upon thickness and number of panels
pressed (p. 34).
DRYING

The glued panels should be dried to a moisture content of about 12

per cent (p. 39).
LAMINATED SPARS, SPAR FLANGES, BOW ENDS, PONTOON RIBS, AND
REINFORCING BLOCKS

In laminating spars, spar flanges, bow ends, pontoon ribs, and rein-
forcing blocks several minutes may be required from the time the
first glue is spread on the wood until pressure is applied. Further-
more, the large glue line areas of most of these members require large
presses (pi. 12, A) or the use of many hand clamps (pi. 12, B) in
order to obtain adequate pressure.
The gluing operations are somewhat simplified, however, in that
spruce is the principal and often the only wood used, and the gluing
is done with casein glues.
The following conditions are recommended for gluing Sitka spruce
into laminated members :
44 TECHIíIOAL BULLETIN- 2 0 5, U. S. DEPT. OF AGRICULTUEE

GLUING BOOM

Ordinary room temperatures and shop conditions are satisfactory.

GLUE

Glue conforming to United States Army and Navy Specifications

Nos. 98-14020-D and 52G8, respectively, is recommended.
GLUE-WATER PROPORTIONS

A glue mixture of medium consistency (by weight 1 part of dry

glue to 2 parts of water) is recommended for most prepared glues
(p. 15).
MOISTURE CONTENT OF WOOD

A moisture content of the wood at the time of gluing of 7 to 12

per cent, depending upon thickness of individual laminations and
other factors (p. 17 and Table 3), is recommended.
TEMPERATURE OF WOOD

Wood that is at ordinary room temperature is satisfactory.

GLUE SPREAD

One and one-fourth ounces of glue mixture should be spread uni-

formly per square foot of joint area. (If both sides of joint are
spread, 25 per cent more glue should be used.) Machine spreading
is preferred. The glue-coated pieces should be laid together as soon
as spread (p. 24).
ASSEMBLY TIME

Pressure should be applied within 20 minutes after glue is first

spread on the wood. Where it is impractical to lay the pieces to-
gether as soon as coated with glue, the maximum assembly time
should not exceed 8 minutes. (Table 5.)
PRESSURE

A pressure of 100 to 150 pounds per square inch of single joint

area should be applied for a minimum time of 4 hours (pp. 30 to 32).
CONDITIONING

The glued members should be conditioned for at least two days at

room temperature before machining. Where individual laminations
are one-fourth of an inch or less in thickness the conditioning period
should be one to three weeks, depending upon thickness and width
of laminate^ member (p. 39).
In laminating hardwoods with casein glue, a glue mixture of
thicker consistency (about 10 per cent less water than for spruce)
should be used with an assembly time of 12 or 6 minutes for closed
and open assemblies, respectively, and with a pressure of 150 to 200
pounds per square inch applied for at least 5 hours (p. 35).
 GLUING WOOD IN AIECEAFT MANtTPACTtTEE 45
SCARF-TYPE JOINTS

Scarf and finger, or serrate, joints (fig. 9, H and I) are fre-

quently made in building long pieces, such as in laminations for
beams and propellers. The gluing surfaces are partly end-grain
wood.
The following procedure is recommended for gluing scarf and
other énd-grain surfaces in aircraft :
The end-grain surfaces should be sized with a glue mixture some-
what thinner than that used for regular gluing. A sizing mixture
of 1 part glue to 3 parts water is recommended for animal and
prepared casein glues that meet aircraft requirements. The sizing
coat should be allowed to dry on the wood surfaces. The mixture
for the final gluing should be thicker than for side-grain gluing—
about 1 part by weight of glue to 1.8 and 2 parts water for casein
and animal glues, respectively. Both surfaces of the joint should
be coated with glue and approximately 200 pounds pressure per
square inch applied. For animal glue care should be used that other
conditions are maintained that will cause the glue to come to the
proper consistency for pressing (p. 24).
With the above-recommended procedure, the slopes shown in
Table 10 for plain scarf joints are considered necessary to produce
joints as strong as the wood.
TABLE 10.—Slopes of scarf Joints recommend^ed for several species of woods

Species Slope Species Slope

Red giirn^.. _. 1 in 8. Red oak 1 in 15.

Yellow poplar Do. White oak__ 1 in 15.
Sitka spruce. _ _ lin 10 (estimated).! Black walnut Do.
Honduran mahogany 1 in 10. White ash lin 15 (estimated).«
Yellow birch— lin 12 (
1 In tests, made on plain scarf wing-beam sections, it was found (8) that a slope of 1 in 10 was sufläcient
for Sitka spruce to develop practically 91 per cent of the full strength of the section in cross bending. The
plain scarf was considered the best of 4 difierent kinds of scarfs tested, and it was concluded that tooth-
planing did not increase the efficiency of the splice.
2 There are no test data showing the maximum slope of scarf for white ash that will give the full strength
of the wood. A comparison of white ash with other species tested as to strength and gluing properties
ndicates that the slope should not be greater than 1 in 15.

In finger or serrate joints the same slopes may be used as in the

plain scarf joint provided the parts of the joint fit accurately.
The sharply angular forms of the finger or serrate joints, however,
are more diiScult to machine to obtain a perfect fit of the parts and
are ordinarily more difficult to glue.
Glued scarf joints for aircraft have frequently been wrapped
with tape or reinforced with rivets with the intent of increasing
the factor of safety. The wrappings, however, are probably more
harmful than benencial. They are Imown to be relatively inefiicient,
and in all probability they reciuce rather than increase the durability
of the joint by favoring a high moisture content. On the other hand
mechanical fasteners do not reduce the strength and durability of
the glue and, if properly designed and applied, will preserve some
strength in the joint in case the glue temporarily softens or perma-
nently deteriorates in service.
119542°~30 á
46 TECHísTICAL BTJLLETIÜNT 2 0 5, U. S. DEPT. OF AGRICULTURE.

WING RIBS

Wing ribs are glued with casein glue. The principal species of
wood used are Sitka spruce, mahogany, yellow poplar, and basswood.
The gluing of the ribs involves the making of many joints of small
size and is almost entirely a hand operation. The ribs are assembled
in jigs or forms, the contact surfaces of the pieces are coated with
glue, and brads are commonly inserted at the joints. This procedure
is tedious, time consuming, and the gluing conditions are usually not
the best {22).
In rib-gluing operations now commonly used, three conditions are
apt to be unfavorable to the production of strong joints: (1) Ir-
regular glue spreads; (2) too long an assembly time on some joints;
and (3) inadequate pressure. The use of mecnanical glue spreaders
is recommended for obtaining more uniform glue spreads over the
joints. Pressing the joints with clamps or other devices, either with
or without the insertion of brads (pi. 14), is recommended for insur-
ing more complete contact and stronger glue joints. Small rollers,
revolving in a bath of glue, are sometimes used for spreading the
glue on cap strips. (Fig. 19, A and B.) Similar equipment may
be useful for coating stiffeners, diagonals, and other parts at the
contact points. Pressure forms are also used successfully for gluing
cap strips to the webs and gluing diagonals and stiffeners in place.
These methods will doubtless increase as models and types of aircraft
become more standardized and as the size of manufacturing opera-
tions increase. The use of such mechanical appliances for spread-
ing and pressing tends to reduce the assembly time and will result in
more economical production by gluing the ribs in multiples.
Most wood surfaces that are glued in rib construction are side grain,
therefore the gluing conditions recommended for gluing Sitka spruce
with casein glue into laminated construction (p. 43) are recommended
for the gluing of wing ribs. However, in gluing the cap strips to
the webs or ribs there may be involved the gluing of a certain amount
of end-grain wood. The use of plj^wood in the web results in the
end grains being exposed on approximately one-half the area of the
edges. This condition necessitates bringing a larger area of the cap
strip into contact with the plywood web, which can be done by set-
ting the web into a groove in the cap strip (fig. 19, A), or by dividing
the cap strip and gluing a piece to each side of the web (tig. 19, B).
Because of the difficulty involved in bringing about perfect contact
in the groove type of joint, the divided cap-strip type is easier to glue
to obtain a perfect joint.
BOX BEAMS

The construction of box beams consists of gluing diagonals, or

spacers, and reinforcing, or fitting blocks, between the flanges and
of gluing plywood sheets to the sides. (Pis. 11 and 12, B and C.)
Spruce is the prevailing wood for all parts, except the plywood,
which commonly has either spruce or mahogany face plies.
Where reinforcing blocks, such as in Figure 19, C, a^ and Plate
12, B, are glued into box beams at fitting points, the grain of the
blocks is approximately parallel to the grain of the flanges or it may
be at an angle of as much as 30°, Where the grain of the blocks is
 GLUING WOOD IN AIRCRAFT MANUFACTURE 47
at an angle of as much as 30° or more the procedure recommended for
gluing end-grain joints (p. 38) should be followed. Likewise,
diaphragms used to stiffen a beam or hollow member, as shown in
Figure 19, C, 6, and E, ú5, and in Plate 11 involve the gluing of end-

ên/arqeàf scchon
wind°^riJ>
L

FIGURE 19.—Special joints made In aircraft : A, Section of wing ribs ; B, wing

rib-beam assembly; C, details of box-beam construction ; D, cross section of
laminated I beams ; E, built-up laminated construction (airship) ; and I,
wrapped veneer member

grain to side-grain wood wholly or in part, and the gluing directions

for end-grain surfaces should be followed.
In gluing reinforcing blocks or spacers into box beams, clamps or
other pressure devices are ordinarily used. (PI. 12, B.) The clamps
should be of such size and so spaced as to apply the proper amount of
pressure per square inch of glue line area.
48 TECHNICAL BULLETIN 2 0 5, U. S. DEPT. OF AGRICULTURE

In gluing the plywood sides to the framework of box beams, both

surfaces are side grain. (Fig. 19, B and C and pi. 12, C.) Never-
theless the joints are subjected to relatively severe internal stresses
due to a difference in shrinkage with any moisture change of the
cross-banded faces and the solid wood flanges. Allowable working
stresses for such joints should be based on plywood rather than
laminated wood values. Maximum strength of such joints is im-
portant, and the recommendations for gluing laminated members
should be carefully followed (p. 43). In current practice pressure
is often applied by the insertion of brads and in some instances by
means of presses and clamps. Whether brads are used or not, it is
desirable that pressure be applied in a press or by means of suitable
hand clamps. Because of the size of the joint areas, however, it is
recommended that jackscrews (fig. 13, F) or power presses be used, ,
where practicable. If brads are used, the pressure may be applied
either before or after their insertion. If applied afterwards the time
elapsing between spreading the glue and pressing should not exceed
20 minutes. If pressed first, it is best to allow the glue to set for at
least 5 or 6 hours before inserting the brads, but the joint need not
be under pressure all that time. It is recommended that the pressure
be applied for at least one-half hour.
GENERAL ASSEMBLY

In wing assembly the ribs are fastened to the beams (pi. 15) and
other parts and in some cases all or a part of the wing structure is
covered with plywood. In most of the joints, side-grain faces are
in contact, and recommendations for gluing laminated constructions
should be followed (p. 43). The grain of adjacent pieces is fre-
quently at right angles, however, and severe internal stresses occur
on such joints with changes in moisture content of the wood, hence
the necessity for careful gluing.
In gluing the cap strips of the ribs to the beams, precautions must
be used to insure an adequate amount of glue uniformly applied to
the joint area. Pressure should also be applied to these joints by
means of clamps. This should be done even though brads or screws
are used.
In the gluing of shear blocks to the beam and wing rib (shown in
fig. 19, B, öü), the grain of the blocks is across or diagonal to the
face grain of both the rib and the beam, bringing into contact side-
grain faces in a similar way as in plywood. The pieces are usually
glued and nailed with brads, but the only pressure that is applied to
the joints in gluing results from the insertion of the brads. This
procedure is not likely to bring about complete contact with casein
glue, owing to insufficient and nonuniform pressure, and the strength
of the joints might therefore be expected to be erratic. A more
reliable method is to apply pressure with small clamps and allow the
glue to set. This may be done either immediately before or after
the insertion of brads. Where regular clamping is not feasible a
momentary application of the proper amount of pressure after insert-
ing the brads will be better than no clamping at all.
Reinforcing blocks of hardwoods, such as ash, birch, maple, or
oak, or of plywood are frequently glued to solid or laminated spruce
 Tech. Bul. 205, U. S. Dept. of Agriculture PLATE 14

WINGS GLUED UNDER PRESSURE

A, All wood wing structure assembled without the customary brads. Pressure was applied with
clamps or similar devices to all joints; B, wing rib glued in a jig under pressure. Pressure was
applied to more than 50 joints and brads were used only to hold the pieces in place imtil pressed;
C, the laminated cap strips of this wing rib were glued under pressure. The hundred odd joints
between gussets and other parts of the rib were glued under pressure applied by inserting brads
although such joints afford an opportunity for applying the pressure by power or clamp deyices.
Tech. Bul. 205. U. S. Dept. of Agriculture PLATE 15

WING ASSEMBLY
A, Gluing the ribs to solid spruce spars of a biplane wing. Pressure is applied only by means of
brads; B, starting to assemble a monoplane wmg. Joints which hold the ribs and beams together
are highly stressed and must be carefully glued.
 GLUII^G WOOD lî^ AIBCRAFT MANUFACTUEE 49

beams at the fitting points. These joints should be glued under

conditions recommended for gluing hardwoods with casein glue in
laminated construction (p. 43), and pressure should be applied by
means of clamps or presses.
In gluing plywood to the leading edge, to ailerons, or as a cover-
ing on wings or flying-boat hulls (pis. 2, A and 3, A) it is usually
not practicable to apply pressure with present clamping devices.^^
Inasmuch as brads must be relied upon to apply pressure to the
joints the glue should be of an easily flowing consistency and the
assembly time should be as short as possible.
In Figure 19, D and E are shown two special types of shaped
joints, involving the use of grooved construction to provide extra
gluing area. For a general discussion of plane, tongue-and-groove,
or other shaped joints, see page 20. When gluing refractory woods
under conditions not favorable to the production of strong joints,
the additional gluing surface provided by shaped joints may prove
to be of some value, but with such woods as Sitka spruce and with
good gluing conditions the additional gluing area is unnecessary, and
the plain flat joints are as strong as the woods themselves.
PRINCIPLES OF CROSS-BANDED AND LAMINATED WOOD
CONSTRUCTION
The making of satisfactory glued joints of any type depends not
only on good gluing but also on proper design. The principles or
rules that govern all glued-wood construction are based largely on
the avoidance of internal stresses or the balancing of these stresses
when necessarily present.
CROSS-BANDED CONSTRUCTION

Changes in the moisture content of wood induce or relieve internal

stresses. In cross-banded constructions (fig. 9) these stresses are
chiefly the result of the arrangement of the plies at different angles.
Adjacent plies tend to shrink or swell across the grain under moisture
changes; and since each ply, by reason of its small longitudinal
shrinking or swelling, restrains the adjoining ply or plies, large
internal stresses are set up in each layer. If the plywood is drying,
all plies tend to shrink in the direction across the grain. Thus all
plies are put in tension perpendicular to the grain and in compression
parallelto the grain. There is, in fact, a shearing force at work on
the glue joints. In plywood that is thus stressed from drying,
moistening the plies tends first to relieve the stresses and, if con-
tinued far enough, to introduce the exactly opposite forces. Moisten-
ing and drying, therefore, alternate in their action, each at first cor-
recting and finally reversing the effects of the other.
22 A unique veneer gluing operation is employed in a Pacific coast aircraft factory in
the construction of monoplanes. The fuselage, which is a monocoque type, is shaped and
glued in a concrete form using compressed air to apply pressure. Smgle-ply veneer is
first placed in the form, a coat of casein glue applied, and a second ply is then forced
into position in the form by a bag of compressed air. The grains of the two plies are
laid at right angles. This construction is reported to give great strength and positive
alignment, with extremely light weight (13). A similar type of construction has been
used in pontoons and flying-boat hulls ; the veneer, which is glued with marine glue and
fastened to the framework with brads or small nails, is applied as a covering.
50 TECHÍí^ICAL BULLETIN 2 05, U. S. DEPT. OF AGRICULTURE

BALANCED CONSTRUCTION

The stresses in cross-banded construction must be approximately-

balanced in the panel itself, or there will be a strong tendency toward
distortion and warping. The necessary balance of forces is provided
by using an odd number of plies and so arranging them that for
any given ply there is a similar, opposite, and parallel ply equally
distant from the core or central ply. As already explained, if two
plies are glued together with the grain at a right angle, each ply
tends to distort the other when any moisture change occurs, and cup-
ping is the result. By using a third ply with the grain parallel to
that of the other outside, or face ply, a balanced effect is produced.
The corresponding plies should not only be parallel to each other
and perpendicular to adjacent plies, but they should be of similar
properties (7).
Obtaining symmetrically placed similar plies involves a considera-
tion of: (1) Ply thickness, (2) kind of wood with particular refer-
ence to (a) shrinkage and (&) density, (3) moisture content of wood
at time of gluing, and (4) angle or relative direction of grain of
plies.
It is possible to get a balanced construction without having a
strictly parallel-and-perpendicular arrangement of plies. The wing-
beam section shown in Figure 19, C, illustrates a variation from the
90° arrangement of plies in which an approximate balance of in-
ternal stresses is still maintained in the member as a whole. The
plies should be symmetrically arranged about the core and for each
pair of opposite plies at any angle other than 0° or 90° with the
core there must be a corresponding pair of plies with the grain at a
right angle to that of the first pair. This principle applies only
to five or a larger odd number of plies. In some aircraft parts, how-
ever, freedom from warping tendencies is of less importance than
other mechanical characteristics, and more or less complete depar-
tures from the principle of balanced construction become necessary.
The amount of internal stress on the glue joints of cross-banded
construction is determined chiefly by the thickness of the plies and
the density and shrinkage of the woods involved. Thick layers and
woods of high density and high shrinkage properties develop the
largest stresses. Table 11 contains shrinkage and density values for
a number of woods which are now glued for aircraft together with
several woods whose properties or abundance indicate that they may
eventually be used. In plywood the face plies, being restrained on
only one side, exert a relatively large stress on the outer glue joints.
For this reason face plies should, in general, be thinner than the in-
terior plies. One-eighth inch is about the maximum thickness of
face plies that can be held securely in place under severe moisture
changes, and where considerable strength of joint is required thin-
ner plies must be used.
 GLUING WOOD IN AIECEAFT MANUFACTUBE 51

TABLE 11.—Shrinhage and density of woods fflued for aíroraft ^

Shrinkage 2 Shrinkage 2
Species Den- Species Den-
Tan- sity 3 Tan- Radial sity3
gential Radial gential

Mahogany (Swietenia spe- Per cent Per cent Per cent Per cent
cies) _ 4.7 3.4 0.46 Red oak 8.2 4.0 0.63
Northern white pine 6.0 2.3 .39 White ash (second growth). 8.7 6.3 .64
Western yellow pine 6.3 3.9 .41 Yellow birch... 9.2 7.2 .63
Yellow poplar 7.1 4.0 .41 White oak 9.0 5.3 .69
Black cherry 7.1 3.7 .61 C ottonwood, eastern 9.2 ,3.9 .43
Black walnut _ ___ 7.1 5.2 .57 Sugar maple. _ 9.5 4.9 .62
Sitka spruce ,.- _ 7.5 4.3 .38 Basswood _. 9.3 6.6 .38
Tujielo gnrn . .,, . 7.6 4.2 .62 American elm 9.5 4.2 .51
Sycamore 7.6 5.1 .50 Red gum ._ 9.9 6.2 .49
Douglas fir (west coast) 7.8 5.0 .61 Beech 10.6 4.8 .63
Red maple 8.2 4.0 .64

1 Data from U. S. Dept. Agr. Tech. Bui. 158 except for mahogany (18).
a Shrinkage from green to oven-dry conditions expressed in percentage of dimensions when green. Longi-
tudinal shrinkage of normal wood ranges from 0.1 to 0.3 per cent.
8 Density expressed as si)ecific gravity based on oven-dry weight and air-dry volume. They are the
specific gravities used in plotting Figure 18 and are on a different basis from that used in certain tables of
strength values of various woods for use in airplane design {26, 27).

In certain types of cross-ban(îed construction the magnitude of the

external stresses is of considerable importance. If, for example, the
shear block in Figure 19 B, a, were glued to a solid wing beam, the
glue would be subjected to very large shearing stresses during mois-
ture changes, and failure might occur from this cause alone. The
use of a plywood web on the beam greatly reduces the stresses on
the joint because the shrinkage of the plywood and that of the block
endwise are not greatly different.
LAMINATED-WOOD CONSTRUCTION

Propellers are an important example of laminated construction in

airplane work. In laminated construction (fig. 9, E, F, G) the
grain of plies is approximately parallel, and the shrinkage of all
plies combined is similar to that of solid wood.
However, no two pieces of wood are exactly alike in properties.
When glued together, even with the grain parallel, laminations with
different shrinkage properties develop stresses at the glued joints,
which weaken the member and cause it to change shape. In most
laminated construction, and especially in propellers, it is of the
highest importance that such stresses be avoided.
The construction of laminated-wood products free from internal
stresses involves the principle of combining pieces of similar proper-
ties, which, in turn, necessitates a consideration of (1) the kind of
wood, (2) the plane of cutting and direction of grain of pieces, and
(3) the moisture content of plies.
Pieces of wood of the same species are more apt to be alike than
pieces of different species. Variations in wood affecting its use in
laminated construction are mainly shrinkage (Table 11) and rate
of change in moisture content. A joint made of woods that have
about the same rate of moisture change will be freer from stresses
than one which includes both quick and slow changing woods. Con-
trary to the effect noted in cross-banded construction, differences in
density of adjacent laminations, when not accompanied by differ-
ences in shrinkage properties, do not induce serious internal stresses.
52 TECHI^ICAL BULLETIN 2 05, V. S. DEPT. OF AGRICULTUBE

Combining high and low density laminations of the same species,

provided they are otherwise alike, into the same laminated construc-
tion does not appear to cause internal stresses of large magnitude.
Laminated specimens under test have shown little weakening or
tendency to warp from this cause (H).
There is a difference in shrinkage between quartered and flat-sawn
material of the same species. Although this difference is greater in
some woods than in others, an average of 150 species of wood shows
a tangential shrinkage of about 8 per cent and a radial shrinkage
of about 4% per cent between the oven-dry and saturated conditions
(Í6*). Combining quartered and flat-sawn material is a disturbing
factor in certain classes of laminated construction, such as propellers,
but may be relatively unimportant in others.
The direction of the grain of adjacent laminations affects the
strength and shape of glued members. Cross grain in a lamination
results in a certain amount of longitudinal shrinkage of that par-
ticular lamination, whereas the longitudinal shrinkage of straight-
grained material is very small or negligible. Consequently a con-
struction made of cross-grained and straight-grained laminations is
very likely to warp.
Stresses in laminated members resulting from differences of mois-
ture content in the laminations have been found to disappear gradu-
ally and without danger of later recurrence if the glued block is kept
for a long time in the same atmospheric condition, because the mem-
ber checks or changes shape more or less to relieve the stressed con-
dition of the fibers. If, however, the original stresses are the result
of combining quartered and flat-sawn material they may die out
during a long conditioning, but as soon as the atmospheric conditions
change the stresses will be set up anew.
In general, then, it should be apparent that for laminated-wood
articles in which high strength and accurate shape are required all
plain-sawed or all quarter-sawed material of the same wood or simi-
lar woods should be used, and all pieces should be at a uniform
moisture content when glued. For the manufacture of articles where
slight changes in form are of no consequence, these precautions are
of the United States :
GLUE FORMULAS FOR USE IN AIRCRAFT
The following water-resistant glue formulas, with directions for
mixing, were developed at the Forest Products Laboratory by the
laboratory personnel and are available for the free use of the people
of the United States.
CASEIN GLUE FORMULA NO. 11 »a
Ingredients Parts by weight
Casein 100
Water 220-230
Hydrated lime 20- 30
Water 100
Silicate of soda (water glass) 70
Copper chloride 2-3
Water - 30- 50
23 This formula, developed at the Forest Products Laboratory, is covered by U. S. Patent
No. 1,456,842, which is available for the free use of the people of the United States.
BüTTERMAN, S., and COOPBRRIDER, C. K. PROCESS OF MANUFACTURING WATERPROOF ADHE-
SIVES. (U. S. Patent No. 1,456,842.) U. S. Patent Office, Off. Gaz. 310 ; 1129. 1923.
 GLUIKG WOOD IN AIECRAFT MANUPACTUEE 53

The 220 to 230 parts of water added to the casein is approximately

the right amount to use with Argentine (naturally soured) casein;
but if a different casein is used, tEe water requirement will lie some-
where between 150 and 250 parts by weight.
The formula presupposes that a high-calcium lime will be used. A
lime of lower grade may be used, but a proportionately larger
amount of it will have to be added unless one is willing to sacrifice
the water resistance of the glue. It is suggested that the user try
25 parts of lime to begin with. If this does not give good results,
the amount can be varied within the limits specified.
The density of the silicate of soda used should be about 40° Baume
with a silica-soda ratio of from 1 to 3.25. Copper sulphate can be
substituted for copper chloride.
The casein and water are placed in the bowl of the mixing ma-
chine, and the paddle is made to rotate, stirring the mixture until
all the water has been absorbed and all the casein moistened. By
allowing the casein to soak beforehand for a while it is more readily
dissolved in the mixing process. In a separate container the hyd-
rated lime is mixed with water. This mixture is stirred vigorously
at first, but just before it is added to the casein it should be stirred
with a gentle rotary motion, just enough to keep the finest particles of
lime in suspension. Pour this milk of lime quickly into the casein.
When casein and lime are first combined they form large lumps,
which are balls of casein coated with the partly dissolved casein.
These balls break up readily under the action of the mixer, becom-
ing smaller and smaller and finally disappearing. The solution, in
the meantime, is becoming thin and fluid. Within the first minute
after adding the lime it is well to stop the paddle and scrape the
sides and bottom of the can, and then stir again. If a deposit of
casein remains unacted on, it may cause more lumps later.
When about two minutes have elapsed since the lime and casein
were first put together, it will be noticed that the glue has begun to
thicken a little. The sodium silicate must be added now, or else the
glue will become too thick. Disregarding lumps in the casein, if
they are but few, pour in the sodium silicate. The glue will become
even thicker momentarily, but it will soon change to a thin, smooth,
and fluid consistency.
The stirring should continue until the glue is free from all lumps.
This should not take more than 15 or 20 minutes, counting from the
time the lime was added. If the glue is now a little too thick, a
small amount of water may/be added. If the glue is too thin, how-
ever, it will be necessary to discard the entire batch and start over
again, using a smaller proportion of water.
The copper salt may be added at different times during the mixing
process with apparently the same final results. If added as a pow-
der or a solution to the casein before soaking, it is likely to have a
corrosive action on the metal container, and for this reason it is
advisable to add it later during the mixing process. If added at
first, however, it should be thoroughly mixed with the casein before
the addition of the lime. It may be placed in solution and conven-
iently stirred into the moistened casein immediately before adding
the lime or after all the other ingredients have been combined and
the mixture rendered smooth and uniform in consistency. When
54 TECHNICAL BULLETIN 2 0 5, IT. S. DEPT. OF AGRICULTURE

added at the end of the mixing period, it should be poured into the
glue in a thin stream and the mixture stirred vigorously. Stirring
should be continued until any lumps that may have formed at first
by coagulation, when glue and copper solution mix, are broken up
and a smooth, violet-colored glue is obtained.
Glue prepared by formula No. 11 has proved to be exceptionally
strong and durable in aircraft construction, even under wet or damp
conditions.
CASEIN GLUE FORMULA NO. 4B ^
Formula No. 11, as above specified but without the copper-chloride
solution, represents an earlier stage of development, known as for-
mula 4B. The mixing is the same as for formula No. 11, except
for the omission of the copper-chloride solution. The glue has a
medium consistency, excellent working properties, and a good work-
ing life. It falls somewhat short of formula No. 11 in water-resist-
ing properties, however.
BLOOD-ALBUMIN GLUE—HOT-PRESS FORMULA î»
Ingredients Parts by weight
Blood albumin (90 per cent solubility) 100
Water , 170
Ammonium hydroxide (specific gravity 0.90) 4
Hydrated lime _ 3
Water , 10
Cold water should be poured over the blood albumin and the mix-
ture allowed to stand for an hour or two without stirring. The
soaked blood albumin can then be put into solution with a small
amount of stirring. After the blood albumin is in solution the am-
monia is added while the mixture is being stirred slowly. Slow
stirring is necessary to prevent foamy glue. The lime is then com-
bined with the smaller amount of water to form milk of lime. The
milk of lime is added and agitation should be continued for a few
minutes. Care should be exercised in the use of the lime, inasmuch
as a small excess will cause the mixture to thicken and become
a jellylike mass. The glue should be of medium consistency when
mixed and should remain suitable for use for several hours. The
exact proportions of blood albumin and water may be varied as
required to produce a glue of greater or less consistency or to suit
a blood albumin of different solubility from that specified.
PARAFORMALDEHYDE-BLOOD-ALBUMIN GLUE FORMULA 2«
Ingredients Parts by weight
Blood albumin (90 per cent solubility) 100
Water ^ 140-200
Ammonium hydroxide (si>eeific gravity 0.90) 5i/^
Paraformaldehyde . 15

24 This formula, developed at the Forest Products Laboratory, is covered by U. S. Patent

No. 1291396, which is available for the free use of the people of the United States.
BUTTERMAN^ S. PROGEISS OF MANUFACTURING WATERPROOF ADHBSIVES. (U. S. Patent No.
1291396.) U. S. Patent Office, Off. Gaz. 258:354. 1919.
25 This formula, developed at the Forest Products Laboratory, is covered by U. S. Patent
No. 1329599, which is available for the free use of the people of the United States.
HENNING, S. B. GLUE AND MANUFACTURING SAME. (U. S. Patent No. 1329599.) U. S.
Patent Office, Off. Gaz. 271 : 48. 1920.
28 This formula, developed at the Forest Products Laboratory, Is covered by U. S. Patent
No. 1459541, which is available for the free use of the people of the United States.
LiNDAUER, A. C. BLOOD-ALBUMIN GLUE. (U. S. Patent No. 1459541.) U. S. Patent
Office, Off. Gaz. 311: 669. 1923.
 GLUING WOOD IN AIBCEAFT MANUFACTURE 55

The blood albumin is covered with the water and the mixture is
allowed to stand for an hour or two, them stirred slowly. The am-
monium hydroxide is next added with more stirring. Then the para-
formaldehyde is sifted in, and the mixture is stirred constantly at
a fairly high speed. Paraformaldehyde should not be poured in so
rapidly as to form lumps nor so slowly that the mixture will thicken
and coagulate before the required amount has been added.
The mixture thickens considerably and usually reaches a consist-
ency where stirring is difficult or impossible. However, the thick-
ened mass will become fluid again in a short time at ordinary tem-
peratures and will return to a good working consistency in about an
hour. It will remain in this condition for 6 or 8 hours, but when the
liquid finally sets and dries, as in a glue joint, it forms a hard and
insoluble film.
This glue may be used in either hot or cold presses. When cold
pressed, however, it has only moderate strength, and for that reason
is not to be depended upon in aircraft construction where maximum
strength is required. If hot pressed, it is high in strength and very
water resistant.
WATER-RESISTANT ANIMAL GLUE FORMULA «^
Ingredients Parts by weight
Animal glue 100
Water . 225
Oxalic acid 5. 5
Paraformaldehyde 10
The glue is soaked in the water until the granules or flakes have
been softened. It is then melted at about 140° F. after which the
temperature is allowed to fall to between 105° and 115°. The oxalic
acid, in small crystals, and the paraformaldehyde, ground to a fine
white powder, are then mixed together and added to the glue. The
mixture is stirred until all of the oxalic acid has gone into solution,
after which it is ready for use. The paraformaldehyde does not
readily dissolve in the glue, and much of it remains as a finely
divided solid during the working life of the glue mixture. A certain
amount of agitation is, therefore, necessary to keep it evenly dis-
tributed throughout the mixture. The paraformaldehyde should be
fine enough to pass through a 50-mesh sieve.
The paraformaldehvde for use in this formula should be of the
slow-reacting type. A fast-reacting paraformaldehyde appreciably
shortens the working life of the glue mixture.
If kept at a temperature not exceeding 115° F., the glue will
remain in a fluid condition for 8 to 10 hours from the time of incor-
porating the paraformaldehyde and oxalic acid, after which it will
set to a tough, firm jelly which can not be melted again. It is
important to avoid heating the glue mixture much above 115° if a
long working life is required. Organic decomposition of the glue
will not seriously affect the quality of the glue at this temperature,
since the chemicals used in its preparation act as preservatives.
27 This formula, developed at the Forest Products Laboratory, is covered by U. S. Patent
No. 1712077, which is available for the free use of the people of the United States.
HRUBBSKY, C. E., and BROWNE. F. L. WATER-RESISTANT ANIMAL GLUE. (U. S. Patent
No. 1712077.) U. S. Patent Office, Off. Gaz. 382:201. 1929.
56 TECHNICAL BULLETIÎi* 2 0 5, U. S. DEPT. OF AGRICULTURE

Both oxalic acid and paraformaldehyde are poisonous materials

and should therefore be handled with care.
This glue combines the desirable characteristics of an animal glue
with a relatively high-water resistance. However, the water resist-
ance varies directly with the temperature of the water to which the
glued article is subjected. At room temperatures the glue is highly
water resistant, but at 140° F. it is very low in water resistance.
The joints should be conditioned for about two weeks at room tem-
perature to allow the glue to reach its maximum water resistance.
LITERATURE aTED
(1) ALLEN, S. W., and TRUAX, T. R.
1919. GLUES USED IN AIRPLANE PARTS. Nati. Advisory Com. Aeronaut.
Ann. Rpt. (1919) 5: 387-408, iUus. (Tech. Rpt. 66.)
(2) BATEMAN, E., and TOWN, G. G.
1923. THE HYGROSCOPICITT OP HIDE GLUES AND THE RELATION OP TENSILE
STRENGTH OP GLUE TO ITS MOISTURE CONTENT. Indus. and Engin.
Chem. 15: 371-375, illus.
(3) BROUSE, D.
1925. EFFECT OP GLUING CONDITIONS ON THE STRENGTH OF THE GLUE JOINT
IN PLYWOOD. 45 p., iUus. Madison, Wis. (Thesis, University of
Wisconsin.)
(4) BROWNE, F. L., and BROUSE, D.
1928. THE CONSISTENCY OP CASEIN GLUE). Weiser, H. B., Editor, Colloid
Symposium Monograph v. 5 p. 229-242, illus. New York.
(5) DEBBAUKELAER, F. L., POWELL, J. R., and BAHLMANN, E. F.
1930. STANDARD METHODS (REVISED) FOR DETERMINING VISCOSITY AND JELLY
STRENGTH OP GLUE. Indus. and Engin. Chem. Anal. Ed. 2:348-
351, inus.
(6) DOUGLAS, W. D., and PETITIFOR, C. B.
1929. THE TESTING OF ADHEsivES FOB TIMBER. JouF. Roy. Aeronaut. Soc.
33:92-128.
(7) ELMENDORP, A.
1921. DATA ON THE DESIGN OP PLYWOOD FOR AIRCRAFT. Nati. Advisory
Com. Aeronaut. Ann. Rpt. (1920) 6: 109-122, illus. (Tech.
Rpt. 84.)
(8) FRARY, H. D.
1921. MERITS OF DIFFERENT SPLICES FOR AIRPLANE WING BEAMS. JOUR. SOC.
Automotive Engin. 9:133-138, illus.
(9) GILL, A. H.
1915. A STUDY OP VARIOUS TESTS UPON GLUE, PARTICULARLY THE TENSILE
STRENGTH. ludus. and Engin. Chem. 7: 102-106, illus.
(10)
1925. A COMPARISON OP VARIOUS TESTS UPON GLUE, PARTICULARLY ITS TEN-
SILE STRENGTH. Indus. and Engin. Chem. 17: 297-298.
(11) [GREAT BRITAIN]. DEPARTMENT OF SCIENTIFIC AND INDUSTRIAL RESEARCH.
1926. MECHANICAL TESTS OF ADHESIVES POR TIMBER BY THE ROYAL AIRCRAFT
ESTABLISHMENT. [Gt. Brit.] Dept. Sei. and Indus. Research,
Adhesives Research Com. Rpt. 2: 90-121, illus.
(12) HARDECKER, J. F.
1928. WOOD AND VENEERS IN THE MANUFACTURE OF AIRPLANES. SO. LUM-
berman 133(1734) : 159-161, illus.
(13)
1929. VENEERS EMPLOYED IN MASS AIRPLANE CONSTRUCTION. VeneerS
23(3) : 18-20, illus.
(14) HEIM, A. L., KNAUSS, A. C, and SEUTTER, L.
1923. INTERNAL STRESSES IN LAMINATED CONSTRUCTION. Nati. Advisory
Com. Aeronaut. Ann. Rpt. (1922) 8: 327-382, illus. (Tech. Rpt.
145.)
(15) Hopp, G.
1920. DETTERMINATION OF THE TENSILE STRENGTH OP GLUE. JOUT. ludUS.
and Engin. Chem. 12: 356-358, illus.
 GLUIKÖ WOOD m AIBCRAFT MANUFACTURE 57

(16) KOÍ3HI.EB, A.
1924. THE PROPERTIES AND USES OF WOOD. 354 p., iUus. New York and
London.
(17) MCBAIN, J. W., and HOPKINS, D. G.
1925. ON ADHESivES AND ADHESIVE ACTION. Jour. Phys. Chem. 29:188-
204, illus.
(18) MARKWARDT, L. J.
1930. COMPARATIVE STRENGTH PROPERTIES OF WOODS GROWN IN THE TJNITED
STATES. U. S. Dept. Agr. Tech. Bui. 158, 39 p., illus.
(19) NEWLIN, J. A., and WILSON, T. R. C.
1917. MECHANICAL PROPERTIES OF WOODS GROWN IN THH UNITED STATES.
U. S. Dept. Agr. Bui. 556, 47 p., illus.
(20) SPONSLER, O. L.
1919. WATER PROOFING PANELS. FACTORS AFFECTING THE WATER RESISTANCE
OF PLYWOOD. Hardwood Rec. 47 (8) : 23-24, 26.
(21) THELEN, R.
1929. KILN DRYING HANDBOOK. U. S. Dept. Agr. Bul. 1136, 64 p., illus.
(Revised.)
(22) TRUAX, T. R.
1928. GLUING PRACTICE AT AIRCRAIT:' MANUFACTURING PLANTS. Nati. Ad-
visory Com. Aeronaut, Tech. Note 291, 11 p., illus.
(23) —
1928. GLUING WOOD IN AIRCRAFT WORK. Amer. Soc. Mech. Engin. Trans.
(Aer. 50-16) 49-50: 61-64, illus.
(24)
1929. THE GLUING OF WOOD. U. S. Dept. Agr. Bul. 1500, 78 p., illus.
(25) BROWNE, F. I/., and BROUSB, D.
1929. SIGNIFICANCE OF MECHANICAL WOOD-JOINT TESTS FOR THE SELECTION
OF WOODWORKING GLUES. Indus. and Engin. Chem. 21:74-79,
illus,
(26) UNITED STATES DEPARTMENT OF AGRICULTURE, FOREST SEIRVICE. FOREST
PRODUCTS LABORATORY.
1919. WOOD IN AIRCRAFT CONSTRUCTION. 149 p., iUus. Reprinted from
Navy Dept., Bur. Aeronaut. Aircraft Design Data No. 12.
(27)
1928. MANUAL FOR THE INSPECTION OF AIRCRAFT WOOD AND GLUE FOR THE
UNITED STATES NAVY. Navy Dept., Bur. Aeronaut. SD-31, 152 p.,
illus.
(28) UNITED STATES WAR DEPARTMENT, AIR CORPS.
1929. THE AIRPLANE. 54 p., illus. Washington, D. C.
ORGANIZATION OF THE UNITED STATES DEPARTMENT OF AGRICULTURE
WHEN THIS PUBLICATION WAS LAST PRINTED

ßearetary of Agriculture ARTHUB M. HYDE.

Assistant Secretary R. W, DUNLAP.
Director of Scientific Work A. F. WOODS.
Director of Regulatory Work WALTER G. CAMPEUX.
Director of Extension Work G. W. WABBURTON.
Director of Personnel and Business Adminis- W. W. STOCKBERGER.
tration.
Director of Information ^ M. S. EISENHOWEB.
Solicitor^ .: E. L. MARSHALL.
Weather Bureau CHARLES F. MARVIN, Chief,
Bureau of Animal Industry JOHN R. MOHLER, Chief.
Bureau of Dairy Industry O. E. REED, Chief.
Bureau of Plant Industry WTTJJAM A. TAYLOR, Chief.
Forest Service ^ R. Y. STTJABT, Chief.
Bureau of Chemistry and Soils -_ H. G. KNIGHT, Chief.
Bureau of Entomology , C. L. MARLATT, Chief.
Bureau of Biological Survey PAUL G. REDINGTON, Chief.
Bureau of Public Roads THOMAS H. MAODONALD, Chief.
Bureau of Agricultural Economics - NILS A. OLSEN, Chief.
Bureau of Home Economics- LOUISE STANLEY, Chief.
Plant Quarantine and Control Administration- LEE A. STRONG, Chief.
Grain Futures Administration J. W. T. DUVEL, Chief.
Food and Drug Administration WALTER G. CAMPBELL, Director of
Regulatory Work, in Charge.
Office of Experiment Stations , Chief.
Office of Cooperative Extension Work . C. B. SMITH, Chief.
Library , CLARIBEL R. BABNETT, Librarian.

This bulletin is a contribution from

Forest Service . R. Y. STUART, Chief.

Branch of Research . .. EARLE H. CLAPP, Assistant For*
ester, in Charge.
Forest Products Laboratory- CARLILE P. WINSLOW, Director
Section of Wood Preservation GEORGE M. HUNT, Senior Chemist,
in Charge.
58

U.S. GOVERNMENT PRINTING OFFICE: 1930