.axr.
5.". .wx.
a v..
Lurm k
J ’ .9 no ...
DAM: «up. I “‘0‘ w“. 0.3
. . a - . . “3
A . .

’4‘»:
. I

“‘
«a.
' x
\q‘
5'

l
a
o
. .
a
.
'

\ ‘. "a_"\
V1 \‘3
0‘" 1 :
\‘J- i. I \-
vrz: '5
. '3
‘4' .-
I» a
"'4'- I z z
. 283%. “3.4!. "

R .
.fl 3...
p)».

in

K
1
O
I
3
83
1
{a
I

‘0
a

"fi
:’
a
‘K'.

“I
\
v
.0
' '3'.
‘ .‘o...’ r“
1'

‘.’

2.
Y.

.3

0' an n .- rm“.
1%, «7.... 1.. x 3 c

\ ‘ D‘..' a. I
i “\J OHL “UV. 9 or. c
.w .3 «3 .5 . a:
Iv. it n o... .. .3 5
2a 23 a... a. .
. a ax. 3 C G
\v v“ xuWQI ‘ n /r
“u “7.x. a . . . ..

\J. x
.. A Lu
I. cw".- h

2‘

..\c"‘:-. .‘."

E:12:3,;_,.:_,E.__:,_25.32:;

 

 

‘‘‘‘‘

 

J,“

NAIL-GLUED WOOD-PLYWOOD I-BEAMS

FOR RESIDENTIAL CONSTRUCTION

By
Donald Frederick Luebs

A Thesis
Submitted to the Faculty
of
Michigan State University
of

Agriculture and Applied Science

In Partial Fulfillment of the
Requirements for the Degree
of

MASTER OF SCIENCE

Department of Forest Products

1959

ACKNOWLEDGEMENTS

The writer is deeply indebted to Professor B. M. Radcliffe
for his direction, assistance and continued interest in the pre-
paration of this paper.

Sincere thanks are also due Doctor A. W. Sliker for his

helpful suggestions and valuable advice.

ii

LIST OF FIGURES . . .
ABSTRACT. . . . . . .
INTRODUCTION. . . . .
PREVIOUS WORK . . . .

HiStory O O 0 0
Design Data. . .

MATERIALS AND METHODS

Fabrication. . .
Method of Test .

TEST RESULTS. . . . .

TABLE OF

Loads at Allowable Deflection.

Stiffness Factor . . . . .
Moment of Inertia. . . . .
Modulus of Elasticity . .
Shear Deflection . . . . .
Fiber Stress in Flanges. .
Fiber Stress in Splices. .
Horizontal Shear Stress. .
Rolling Shear Stress . . .
Beams with Panels Removed..

O O
0 O
O O
O O
O O
O O
O O
O O
O 0

DISCUSSION OF RESULTS . . . . . . .

CONCLUSIONS AND RECOMMENDATIONS . .

LEEMURE CITED. 0 O O O O O O O O

APPEIQDE A O O O O O O O O O O O O

APPmmm B O O O O O O O O O O O 0

iii

CONTENTS

0 O O O O O O O O O

Page
iv

LIST OF FIGURES

Page

8
ll
12
13
15
16

18

19
20

23

27

29

29

Figure
1. Construction Details and Nomenclature of
Nail-Glmd I-BeamS. o o o o o o o o o o o
2. method of Testing Beams . . . . . . . . . .
3. Deflection Measuring Device . . . . . . . .
4. Beams Being Loaded . . . . . . . . . . . .
5. Smaller Beams after Failure. . . . . . . .
6. Deeper Beams under Full Load. . . . . . . .
7. Load-Deflection Curves for Beams of Various
Section Depth . . . . . . . . . . . . . .
8. Shear and Bending Moment Diagrams for Uniform
wading O O C O C O O O O O O O O O O O O
9. Experimental Stiffness Factors of Beams . .
10. Chart Showing Lack of Correlation between Beam
Depth and MOde or ElaStiCity o o o o o o 0
ll. Load-Reduction Factors for Beams with Panels
Removed after Fabrication . . . . . . . . .
12. Beam Failure at Tension Flange Splice in
14 jun. Beam 0 O O O O O O O O O O O O I O O O
13. Beam Failure at Tension Flange Splice in
16 in. Beam 0 O C O I O O O O O O O O O O O O
14. Loaded Beams Forming a Smooth Curve of Deflection.31

iv

ABSTRACT

There are many places in the structure of a house where clear-
span wood-and-plywood girders can be used to eliminate supporting
walls and columns and in general allow more design flexibility and
facilitate construction. Appreciable savings in material and weight
are possible by using a built-up construction which places material
where it may resist stresses most efficiently. It has been demon-
strated that costs may be reduced in many cases by replacing other
structural members with such beams.

The basic purpose of this investigation was to evaluate the
practical application of certain plywood beams of simple construction
for residential use. Such a beam must be easy to construct and must
employ locally available materials to be of real utility. In addi-
tion, it is desirable that no special equipment he required in
assembly.

Full scale tests were run on eighteen beams of various section
depth; the test method simulated actual use conditions. Load-deflection
data was taken, performance was observed and thenature and location of
failures was noted.

It was determined that nail-gluing with casein glue was an effi-
cient and practical method of joining members which seemed ideally suited
to this particular application. Shop fabrication where conditions are
more easily controlled is recommended. The beams should be used in
protected ventilated locations and should not be subjected continuously

to high humidity conditions.

It was concluded that beams of this description should be
designed on the basis of allowable deflection at midspan. No
failures occurred in the web due to horizontal shear stress even
though calculations indicated values much in excess of the arbi-
trarily accepted allowable stresses. A well-balanced beam would
not result if such values were used in the design. Compressive
stress at supports should be checked on heavily loaded beams to
eliminate the possibility of crushing of the wood.

Sections of the web may be cut out subsequent to fabrication
if the proper load-reduction factor is employed. The removal of
such panels might be necessary in construction to accommodate heat
ducts, plumbing, etc.

These nail—glued plywood girders should prove more economical than
steel beams or wood-steel flitch plate beams in many applications in
residential construction. They will certainly be much easier to
handle and position as well as fasten in place and apply finishing

materials to.

vi

NAIL-GLUED WOOD-PLYWOOD I-BEAMS

FOR RESIDENTIAL CONSTRUCTION

INTRODUCTION

There are many places in residential construction where built-up
wood-plywood girders may be advantageously used to replace more expen-
sive structural members. Such installations would include headers
above garage door and other wide Openings, main floor support girders
and wall and ridge beams for post and beam construction. The use of such
beams for clear-span floor and roof systems is also a promising appli-
cation.

Because the design of these girders places material where it may
be stressed to the best advantage, considerable savings in wood as well
as weight are possible. This combined with the high natural strength-
weight ratio of wood is an important factor; handling ease is promoted
while dead weight is reduced. For example, a wood-plywood girder for
a 16 ft. garage door header weighs less than half its steel I-beam
counterpart of similar strength and stiffness.

The fastening problem encountered with steel members is greatly
simplified since the plywood girder may be nailed in place. Finishing
materials may also be nailed directly to the beams. This is a great
contribution to flexibility of construction.

Working stresses for any grade of lumber within a species are
dependent upon a number of possible defects, particularly in the cen-
ter third of the span. The built-up construction of the plywood beam

promotes an increase in strength as the randomly occurring defects in

the members glued together reinforce one another. Shrinkage and
twisting tendencies found in solid wood members are also resisted
by the built-up construction.

One of the biggest problems with any built-up construction is
adequate joining of the members so that stresses may be distributed
directly to the members which are designed to resist them. Often
unnecessary costs are incurred by oversizing members merely to acco-
mmodate fasteners. A good glue joint of sufficient area is the most
efficient device for joining these members, but the use of presses
or clamps is often a hinderance, especially with larger assemblies.
It has been demonstrated that the use of nails to apply pressure to
glued joints is a simple yet effective method.10 Such nail-gluing
seems to be ideally suited to built-up plywood beams.

In the series of tests reported in this paper it was the object
to evaluate the practical application of certain selected nail-glued,
wood-plywood I-beams for residential construction. The beams were to
be simple in design and constructed of stock-size materials with no
special equipment. Full-scale tests which simulated actual use con-
ditions were to be run and data was taken to evaluate the performance.

The nature and location of the failures were to be observed.

PREVIOUS WORK

History

There has been much develOpment in structural uses of plywood
in the past 25 years. [Previous to 1935 little such develOpment had
taken place, partially due to a limited supply of plywood and partially
due to the lack of suitable adhesives. Actually, development was tak-
ing place only slowly previous to World War II.

COne of the earliest intensive uses of plywood beams in this country
was in a 125,000 sq. ft. warehouse built in 1942 for the RCA Manufactur-
ing 00. at Camden, New Jersey.19 A total of 198 identical plywood
girders, 36 ft. long, were job-fabricated using webs that were nailed
only to the lumber flanges with 8d cement-coated nails. After ten years
of service the warehouse was taken over by the government. The beams
were found to be in excellent condition at that time; they had not
sagged and had required no maintenance.

With the advent of World War II a search for structural members

other than steel took place. The necessity of conserving materials was

a guiding policy in this search.

Design Data

 

Extensive work was done in the years 1943-1944 in the design of box
beams by the Forest Products Laboratory under the supervision of the
Aeronautical Board of the U. S. government. They applied conventional
engineering calculations using allowable design values of wood and made

recommendations concerning the webs, flanges, and stiffeners of box beams.

4

They were largely concerned with buckling of the web14’15 and hori—

12 The Forest Products Laboratory found that

zontal shear stresses.
it made little difference whether the face grain of the plywood webs
was horizontal or vertical. Plywood webs oriented at 45°, however,
were found to be substantially more efficient.15 They also found
that for thin beam webs significant increases in web shear resis-
tance could be secured by reducing stiffener spacing.13

David Countryman in full-scale tests of plywood beams in 1944,
found that nail-gluing was an effective method of fabrication.10
Butt-joining plywood web Splices was also determined to be adequate
to develop the full beam strength in both bending and shear. Country-
man's tests showed no buckling in the webs nor were any beam failures
caused by horizontal shear faults, even though this was the limiting
design stress in many of the beams. He concluded that a better bal-
anced beam might have resulted had the allowable horizontal shear
stresses for the plywood been higher.

The Forest Products Laboratory has published some approximate me-
thods of calculating the strength of plywood based on the basic allow-
able stresses for Douglas-fir.l6 They allow a 25% increase in recommended
working values for use in dry locations (M.C. 16% or less) with a reduc-
tion depending on the grade of plywood used. These allowable working
stresses are widely applied today and are recommended in other coun-
tries1 as well as in later publications in the United States.6’9

The Douglas-fir Plywood Association has conducted additional tests
and developed simplified design methods for plywood box and I—beams.9

The fellowing procedure is recommended. The type of beam and method

of manufacture is decided upon, then a preliminary cross-section is

selected and checked for flexure, horizontal shear, shear between the
web and flange and deflection. In addition it is recommended that the
upper flanges be considered as columns to compute the degree of lateral
stability.

David Countryman and Vernon D. Haskell have presented this same
information for D.F.P.A.11 It is in a more generally usable form with
tables of design factors presented to further simplify it.

The Wood Handbook contains a unit on the design of wood-plywood
beams which is a collection of data from several sources.7 It is in a
very usable form simplified to the point where beams of constant cross-
section are demonstrated. It is suggested that the sizes of beams may
be limited by: (1) Clearance or other headroom limitations (2) Width
of plywood economically available, (3) Ratio of height to width and span
for lateral stability. These factors should be evaluated previous to
considering beam design.

most of the beams originally tested were constructed to resist
stresses in the most traditionally efficient engineering manner. Many
of them had complicated cross-sections which varied along the length
so that flange area increased toward the center to resist bending
stresses while web area increased toward the ends to resist shear.

This would require non-stock specially-milled lumber in addition to _
complicated fabrication procedure.

Many of these beams did not fail as the engineering data had pre-
dicted. This indicated that either the data did not accurately picture

the stresses present or that the usual arbitrarily selected allowable

stresses for the materials were at fault; very possibly both conditions

were present. Far more realistic design is possible by basing criteria
on actual performance.

While knowledge of built-up plywood beams is fairly widespread and
recognition of their attributes is given, it appears that relatively
little work has actually been done on them. Most design data and re-

commendations originate directly from one or two sources.

NMTERIALS AND METHODS

Fabrication

 

Test beams were constructed in 24 ft. lengths and depths ranging
from 14 to 22 inches. Referring to Fig. l, a constant, balanced I-
section was used with nominal 2 x 4 members glued to either side of the
3/8" plywood webs to form the top as well as the bottom flange. This
construction gave a beam of 3-5/8 in. thickness which would fit con-
veniently into a conventional stud wall.

The flanges consisted of central 2 x 4 - 16's with 2 x 4 - 4's
spliced to each end to form the 24 ft. length. Splices were directly
opposite each other.

The flange members for the test beams were of mixed Douglas-fir and
White Fir of No. 1 grade. No attempt was made to select lumber except
that bright lumber was specified. Low-grade material was considered
suitable for the 2 x 4 stiffeners. Moisture content of the lumber
ranged from 14% to 21% with an average of 17%. Medulus of Elasticity
values averaged .94 x 106 in ten full-sized 2 x 4 samples tested which
were obtained from the same source (See Appendix A).

Stiffeners were placed at two-foot intervals along the length of
the beam; thus, a stiffener occurred at the flange splices and at
splices in the plywood web. Stiffeners were fitted accurately between
t0p and bottom flange members to restrain their movement toward each
other as the beam was stressed. Their main purpose, however, was to re-

sist buckling of the web.

 

it. it'll} 9':l.h.|..l .II ‘lcl!l I.
. I... - ‘II.II-'.I"-" I | r15

a
_ O

-- _‘.-“.

 

 

 

 

 

rum » UCOHQW. nu 1.-
. >1 2. .H .1

.u
r C

 

 

 

 

. - .- -.._..m-J
p

 

 

 

 

 

 

 

 

 

 

 

 

 

o w
u . . -
. V . Q < ul’ 1. .
.- ;$--------!.--:-.._-:a - . . _ .1 .. , 3-- .1
‘ . ‘11:" I! |tl‘n‘."ll"' "ITI' C t
_ 5 o o 4 -ol 1.... 1 -. r v- . J
. . . . I «1 .‘l‘. .Il'lll. {i
w
I s- t . O O .m . ‘ V- .x. o& t
V! -
11-1 .5 I. III:- _ Q o l . .1. t v 1
. $.-a-sy-I-II---.1. i-i- --i;.-1 21...!- 1. q o L \ . . M l. w}... .- H» .x.
\L _ 4... .
_ T- - 1 --i - . . I
.117 - A1,:- 7 ..\ .s i 1 II “.4. L‘
(- \ . “ 1o. , v
.. x v I t *- _ .L
‘1. u!) -.
Jwv...“ \ .
. ./;
‘1 ‘\ n
- -- o - L
\‘I.’ i " U . (
.41" II. ‘1 Q‘JIK II- ‘1‘ l - -v|-.'.‘l'l..|. .
i4 til-n14 -1 -!-|uul.lu 1. - I- it .1- I - I»
II no. 1 Ir oil-«ll 11:! I. ullr-i.
14 J :11}. {13:11 '0‘."-
JI .u-II .- || 1 -l.-o. -1; .- I! .3 t. 1.9-
“ . .Jj _..1 I a 4.16!!! .1!- 1 fl w nil-v. 1-: . .- . L.— 4 w
A. w . -4- 1.... - 4.. an .- -.||I-I.I|ao. 1!- .2.-. - l
_ ‘_ w :“ .v _ g p~ .% fl~ .1. -iinI-fi.1 -1.
. _ .
n - _ _
_ _ . ‘ _ . . . -. ,
||7 ‘I n‘,‘ n! -| . . . _ . m . .
- - 1.1.1- -1--- -- -.-r1_!-u---1..1 . « I x L _ . l _ I _ _
--i---- . 5--- .r- , 1 - - ------..--_--.-..-,.-illi- . _. . . . . _
‘6' 1' w - I >71! :01... 'l tl-’lu>i-.l|.ll -14. P ._ I
I“: . A . ./ I IOIII ll {Iii-a 0‘ .fl ‘1
”A ‘b ' . ’ ‘ \ ..) J L r t \ « I-D. 'ylll 1:1,!!! ’1‘] ‘ ‘P L
.- o \ - P -./ ... )4 _ no...\.- 4 «1) 2/ _ . 1/ u \ N
. . x .. ._ . fl... | ~.0 A .C.
o .2 /’ O ’ z\ t . . l !\ \
_ L.
‘
x . .C.
. H J
[3k let's) lIla'l I \«lnt. “1 (x— (M .IEAI-
, o . - -
- I.
II. I l v ||v.|..1 1|- ltl I. II i-hM’I’LU
. 1...: -I I --,i.. -. nub- u:l.n..u.. .1 I. . . -. H .-|| :1 I a
-- - - - J- :- ..ri. _- .- ~ -1 a.---.--|..1-.1. 1m

 

Casein glue meeting U. S. specification MMM-A-l25 was used as
the adhesive. A relatively heavy mixture (2:l%, water to glue by
weight) was used to increase gap-filling qualities. A single spread
of 110 lbs./1000 sq. ft. was used. The ambient temperature was 75°F.
during fabrication. Sufficient glue squeeze-out evidenced adequate
contact.

In the assembly of the girders, the flanges and stiffeners for
one side only were first laid out and tacked together. Glue was then
spread on these members. The three plywood web sections, each 8 ft.
long, were subsequently laid on this assembly and tacked in place.
These consisted of 3/8 in. Douglas-fir interior grade Plyscord. The
flange members and stiffeners for the remaining side of the beam were
then spread with glue and laid in place.

Pressure was applied by 16d common nails driven down through the
entire assembly from the one side only. Nails were spaced at 8 in. and
staggered; they were driven hard to promote good contact at the glue
line. Three nails were driven into each pair of stiffeners. The nails
are to bring pressure to bear on the glue line; they do not contribute
to the strength ofthe beam except to support it in moving immediately
after assembly.

The last step in fabrication was to spread the splice plates with
glue and secure them by nailing to both sides of the beam. The plates
were 20 in. long with 10 in. over each of the butted flange 2 x 4'3.
Six 6d common nails were used on each side of the butt-joint. Splice
plates were of 1 x 4 soft pine to resist splitting; 3/8 in. plywood

was used in several beams to observe its performance.

10
Beams were stock-piled immediately after fabrication. Care was
used in stacking to avoid deformity which would remain after the glue

had set. A minimum curing time of 24 hours was allowed before testing.

Method of Test

 

The girders were erected in pairs 20 in. 0.0., simply supported
at the ends on 2 x 4 "plates" resting on concrete blocks; cross-bracing
was used at the ends only. A 1/2" plywood roof was tacked across the
pair to form a loading platform. A typical test set-up is shown in
Figure-2.

Nails were driven into the stiffeners at either end of each beam
just above the supports at the neutral axis. A nylon line was then
stretched along the span of each beam by tying to the nail at one end
and suspending a weight over the other. A scale with 1/10 in. incre-
ments was fastened to the stiffener at the mid-span of each beam so
that deflection readings could be taken by referring to the nylon
string. The string was positioned as close to the scale as possible
without touching the beam. Figure 3 shows this deflection-measuring
device.

A "zero" deflection reading was taken previous to loading the beams.
The load was then applied in the form of 8 in. concrete blocks which
weighed 38 lbs. each as determined from a representative sample.

Blocks were placed two abreast along the length of the beams. A
deflection reading was taken after each layer of 74 blocks (1400 lbs.
per beam) had been applied. Figure 4 shows a pair of test beams

being loaded.

\-

”a--—-——‘
I

 

 

 

A..Lo

I‘-
1“

11

I - - _
. .L.---.,.__-— a. “.»~_L—....- “-4‘4—“_..—.

v . -
1
~15
.-_
‘9
y.
1“

w

I .

' i

\,/

i

1......— .4-

O

.-

I
-.-- 5.-.--

O

I

1* ' .
——.- ._.L-.L.---, __ L. 4. -.}

r»»;--o—“ - -h‘ —~
r—.—..V--Y_.—
‘
I
—.
‘-
\

 

 

‘0.

n.

\L

f

w.
w.
w.
1
m
l
a
I.-

 

[ 4... --

IQIOI. lvllld' ' K

 

_ --_ - -.)~-

.- -- —---. ~ _. ~~-.—. .-- v.-— - -

-.-~---.—.— .4-

 

—..—-- - *o——..- ._._ - v

;.‘~I‘—L.J... ~-— ~-

3“"

1
a
\

 

 

 

 

 

 

 

 

12

.-—~—.

 

 

 

‘.&

 

3“!!-

o. 1.00.

 

 

 

 

 

 

. .
A
«Sillllm .
Lilli-0".--
.
I
l. . Q.
.
- I ‘9‘ h ‘ v'i. —.
plat--1vn|u||¢'llllftrvt .5.-
\ Q -1 Ar L
I. u .
w . . .
H .
L I r 1 I» > Ann» v” r ‘
‘11 .—
..\ . I
.sh. .. p Q -
.. \- f.

 

 

Figure 3

Deflection Measuring Device

”5&3; U" “THEM"? 15'”-
Iii-"1allflig’wlruunl-—- ’1... 1.. '
E.*:113-155ti1.-: 1.3..- .

mamas-31-1 11122.n-

 

Figure 4

Beams Being Loaded

 

13

14

Test to failure was conducted on beams of 14 in. and 16 in. section
depths. Figure 5 shows a typical test pair after failure. For beams of
deeper sections six layers of blocks were applied. This was sufficient
to cause several times the allowable deflection without collapse. Figure
6 shows such a typical test pair under full load.

A subsequent series of tests was conducted to determine the effect
of removing panels of the plywood web between adjacent stiffeners. In
actual usage such Openings might be required to provide for heat ducts,
plumbing or other utilities.

Panels were removed from various positions in the four 18 in. beams
for these tests. The beams had been previously loaded for deflection data
as described above,but stresses had not been allowed to approach propor-

tional limits.

 

 

 

 

Figure 5
Smaller Beams After Failure

15

 

 

 

”“3?
. 1; g “WW-11?;
'- “Him-1111mm .. g
‘ "“ "r"_Vb'-'l*.-..-_

' " '4" "’ ’b x
,, pt-

 

 

Figure 6

Deeper Beams under Full LOad

 

16

17
TEST RESULTS
Loads at Allowable Deflection

A load versus mid-span deflection curve was drawn for each beam
tested. The plots are shown in Figure 7. The loads at which a deflec—
tion of 1/360 of the span occurred (.8 in. for a 24 ft. span) are noted
in Table l of Appendix B.

Sgiffness Factor

For each beam a deflection at corresponding loads was used to com-
pute a stiffness factor according to the equation for a simply supported

beam under a uniformly distributed load.

EI . at.
3846
where:
E - modulus of elasticity, p.s.i.
I = moment of inertia, in.4
W = load, lbs. per lin. in.
1 2 span, in.
A : deflection, in.

These experimental stiffness factors are given in Table 2, Appendix B.

They are also compared with expected values in Figure 9.

---

g .

.-
‘\—-
r- - -_
1 \
, ,._1
- --.-
1’ --"1
/
g.-.............
/
..t .... --
r-‘V“‘
k __
'_ _- -.-

 

  

I\ . A 1. / _ .L . . n
I IV \ r
8 I 11.1 111: / oil-.11. .
l .\.J . n ._ __ I .1114 \1/ .. y / . \l \IJ .. .1}, . _ 3 \1 In \.I
1. .. .51.. I , l . _-. -. ...-.1;
(U I v 1“ . :1... A/ w K a :L s _. z u < w ’ // I. / .\ /\.. A
I .
z; .4
A... o a s m .m .m
x ._ _ / ..xl; .- - . 1. \4
>,\ (art-173k spa-C
...
“u..— .t..~
_ -- -:11|11-1-1.1- 1111111
0
‘\-\~
\\\s..\\-.
\\ \~,.
\- -.-\
\\\. \
\-. ...
\ .x...-
\\ \ .o. - \ ...-
\\\ \\ \\ \ .. xx...-
\\ -.\ \\ . \ .xx.
\ \nx \- . \ l.)-
--\ ..\ .... x \ ...-q
.\ s. \ x-x - ... . \
\« .\ .\ . \.\. a. \.\._.~
\. . -. f. .\ ...
\ \\ . x u. \ . ... .s -_1 \
\ \ . \,\- x . ......

\\. \ \‘ _
. I .\
.1. \ . -, . .
r\ l|_ \\ . \ ..x.\ \ ... . x. l. a
\ U . \ \\. ,. 2 _\ I ‘
. \\ . . . . .-
. \ .. \ . . ... . I
\ x \- ~ ~ .I
\. x x . . -_ x \ ..
. kx . I k
\ v x s
.\ x . \\. x . x c .
\ . . -
\ .- \ . . .. .
\ x . \. s . \ a .
\. . s - \ \ I x .. a
-\ \l x .. I . .
. a x -. .
.\ ‘ II 1. \pl \ . x \ e .\
1 1 . \ I
c g \ a _ a

II .. I t ‘
.. fl. (\ t I\ r.\ v\ . ‘1 .\\ x ‘ . .H o C x \\ .e
.1 \- xvu . \I a I. m \..
. o s. .- - lib . a. L. H
« l“. . ‘ \ . I ‘1 . m 1|; .-
. 1111 l I Q ' y y 1 \.~ . . * t .\ I. ~ . ~ c \1 \ j! c p
C .
\ ' l n.
x . .1! m 1 rs ~ . a n .
\ -\ H. a. ... ... m.
‘ \
. \ x \ . . ..
.. . x \ .-
\ e \ \ .. a «-

.4— U. a I.
..t ... ‘— .-\..\

 

2h

 

 

, .39:

19

 

 

 

 

 

I” / . COT
. | j ' I I , I . ,
I 1 1 I 1 1 I I I I I I I L I ‘ ‘ I 1 1 1 L ‘. ‘. j
I I
I f‘ 4 l .I
In ~—- «~7 z I-3 . u—l [>,_"" ,3-
I .
g I i + ‘ \f/ I‘I 5.: I] In
I ‘ I . * ______.__...
. l |
; I I I g A # - "f - -
'—
* ’//,..—-«—-—--~~\\
‘//"‘ . \\l\..
,/ ‘\.
’// ‘-\.\
///' \\
/ + .
// x
, A . - . ."'\
/ , . . \\

/ f ; . 1 I \
A - . i ; ...m - A - __ 4' 3
EDIGII\\ I’war
F i g u r e -8-

p \
L‘ . ~— ‘\
\‘l‘ I . 3 . I ‘ ’I I [I \ I \J ?\/4 fr." I\ AI“-
' F11" ..4. \ X’ \K . I It.) I V I I I I 7 ' 1~ 0
VI .- .‘ .1)th II] II JI.E.._II H‘IV

I"~ \

/
II};

J’ ~ I’C’
UAILJ II)

«OFF
‘wr--‘

$

‘N‘A
’0
0

5t» -.

20

I 1‘

a

s 1
I
l
I
|
.
I

 

i .
i ; z
21 '2’
i
l_ C
‘6 a
i
a O
'0
9" "
h" z
i
’ .Q
,’ K.
1‘ Kb *T »
NJ .
' 4/l ,.
/’.
L1
i" '1
1' -
‘ I ‘1) fl
30 / ,
1’0“
, 1:1: .
I "l‘ ’/ I r * I. n
"n , — “ *‘ {waned v1.3.1?“ 9.: my
f. -, .1 .
/ I . r" l.'\ ’
A -... . 1 ~. ‘.
t. “MES 017 “(LG \Qu
D ‘ / /\
”'2‘? ,~
5 ‘ ‘ k .7“! X 1'3 7
L ' ' 31.: . .4
-~--«-- ---‘-.nrx.r.M-.~-znz expeflw
I 1" l '\
V1 ,1 5- 0:: w‘JL . 1“}
N
\ / L ','\C
of 'a.;‘ '. Us K is)

 

 

.4‘ u, as 2: :1
0‘ t .\
V; 2.2mm- Der-f r: (In;

Figure -9-

" f‘
h F‘l '\ I. " ;”T"‘\ I \.."""‘!°"“ ‘§ r f" C r‘ f "7"“ ’\ f
XPL M""-.1‘{ ‘ "' W"   ‘x‘
.. \ {M Luci ; .L‘ ‘\_/' :l \L‘vv’ I‘L. 1'..' /\.z

I:
H \l w r‘ 'h .' (i
5* LS k /.\ w ‘

'i1iiilll l'l’l‘l. v ‘I‘
. r

21
Moment of Inertia

The theoretical moment of inertia was computed for each beam.

 

b(h3- h/3)
12
where:
I = moment of inertia, in.4
b 3 thickness of flanges, in.
h = depth of beam, in.

h/- distance between flanges, in.

The web area was neglected in these computations; only the 2 x 4 flange
members were considered. The I-values for the beams of various section
depth are given in Table 3, Appendix B.

Modulus of Elasticity

 

The effective modulus of elasticity was then determined for each
beam by the deflection formula as used for computing the stiffness
factor (see Table 4, Appendix B). The term "effective" is used since
the approximate equation for deflection does not take into account de-
flection due to shear. Inasmuch as the approximate deflection equation
will be used for design along with performance data, based on the same
assumption, the inclusion of the shear deflection component is not

necessary.

22

Shear Deflection

As a matter of interest the component of the allowable deflection

(.8 in.) attributable to shear was computed by a theoretical method.

= Pl
A8 G

K 1120
I
where:
A s :: deflection due to shear, in.
P = total load on beam, lbs.
1 = span, in.

section constant

5‘ :21
a

depth of the beam, in.
C = uniform load coefficient

G = shearing modulus of the webs, p.s.i.
4

I moment of inertia, in.

This theoretical deflection due to shear is noted in Table 5, Appendix
B. However, Figure 10 demonstrates that in this investigation there
is no definite relationship between apparent stiffness of the material
and beam depth as the shear deflection formula would indicate. This

is due to the variable performance of non-uniform materials.

2‘

R

(lbs/WK :0“

hasty
H

[loo

.5
Of

("3

:moi Mode In:

Exbemmr

1.!

LO

 

23

 

Q
. E
© t
r.
5’3.
\x’ /\
f.
a ‘. I”
7' ‘ 531‘:
Q.
'7) A
”'3 k: kt)
L')
l4 l6 l8 20 22
n
exacilon [>8pr (m)
F i g u r e -10-

Cw:

Ram/[LN BEAN N’NN

CNo \N: LN: N ,. 15m

N

O

ICN

Fiber Stress in Flanges
The extreme fiber stress in the flanges at mid-span corresponding
to the loading which caused a deflection equal to 1/360 of the span

was calculated by the conventional flexure formula.

_Mc
G "T

where:
63>: extreme fiber stress, p.s.i.
M ' bending moments, in.-lbs.
c = distance from neutral axis to extreme edge, in.
I = moment of inertia, in.4

Computed values may be found in Table 6 of Appendix B.
A form factor may be included in the flexure formula to increase

the estimate of fiber stress.9

.14

The form factor for an I-section may be considered as the ratio of
the fiber stress (at the proportional limit in its application to
working loads) to the similar prOperty of a rectangular beam 2 in. x 2
in. which allowable loading is based upon. The inclusion of a form
factor was deemed unnecessary, however, because the neglected web area
in the previous computation of the moment of inertia has served the
same function of increasing apparent flange stress. In addition, re-
duction of 9/10 is already included in allowable working stresses for
wood to compensate for the depth of rectangular beams commonly used

in construction. Further penalizing of the strength of the wood

would seem unrealistic.

25

Fiber Stre§§_inF§plices

The extreme fiber stress in the l x 4 splice plates was calculated in
the same manner as for the flanges. The plates were considered as form—
ing an I-section with the plywood web; web area (face plies only) was
included in the calculation of moment of inertia for the computation.

Values are given in Table 7, Appendix B.
Horizontal Shear Stress

The maximum horizontal shear stress in the plywood web was com-

puted for the test beams for loads at allowable deflection.

041%

where:
I : horizontal shear stress, p.s.i.
V : vertical shear, lbs.
Q ' statical moment about neutral axis, in.3
I = moment of inertia, in.4
b = web thickness, in.

These values may be fbund in.Table 8, Appendix B.

Rolling_§hear Stress

 

Where the flanges and webs are joined by glue, the joint is

parallel with the plane of the plies. The rolling shear stress in

26
plywood core developed under these circumstances was checked for loads

at maximum deflection.

where:

:0
ll

flange-web shear, p.s.i.

vertical shear, lbs.
3

' statical moment of a flange member, in.

moment of inertia, in.4

QHD,<:
I

depth of the flange member, in.

Beams with Panels Removed

 

The effective modulus of elasticity for each beam with a panel
removed was determined as described previously. The position of re-
moved panels was designated as shown in Figure 11. In each case a
ratio of the modulus of elasticity of the beam to that of the same
beam intact was computed. A plot of this ratio versus the relative
position of the removed panel was made. The curve, whose values are
suggested as a reduction factor for allowable loading, is also given

27

\r‘

. Ava
fu

_ v >.\I
l _ , . I-.-
m w

. . ~.~—-..-‘-

i”.
\‘L/
...

,

I

I

i

'

C

I

Y

1

\

ll

,4
.11-

a
I
o
l

I
‘-
r.

‘\_~
‘ I
I
p

i
5‘;

l
‘1 / '
‘, I
I
I

I
’3
9—4

1
l
A
>2
L;

L

Figure

I. .
.... . . .. -\ .I
p .. . .. . . .
H x r
l . In ,5. -
l .x.... f.
, . . l *1 I1
, l .
m . _ .
it e . . _ 1-119 v a

u a.
J . . M

aw . c r-..r 6 ’ l- .W

a, . _ H . ...:
-h . . - s

x / l

.
_ l
._ . C .. a K’ilf
- l . _ r, _
%

 

 

“ . la
a .. l \I - - w
_ . -
I! l i x A.-.»

III lilliil|tl||WlI {it A II .niiil Ax" ll'-'.|‘.n\ 'III‘X IIUI

.
kl. .-- .
_ , ...) a w , \.. .. .
a j 7. I. u. ”L — a. \. 4
l . . K v

I I‘D

. .
. _
.
, H
‘ Ir.l|||l¢ ,witllilnll, .

28

DISCUSSION OF RESULTS

The most significant observation drawn from the tests was that
deflection at mid-span should be the design criterion for these
girders. Also in the relatively short, deep beams capable of carry-
ing greater loads there must be sufficient bearing to prevent compres-
sion at the supports.

All beams tested to failure failed at the tension flange splices
at a load of about three times that at allowable deflection (see
Figures 12 and 13). No other factor appeared to contribute to failure.

No buckling of the web was observed in the course of testing,
nor did any shear failures occur although the usual computation of
horizontal shear stress indicated values much in excess of the accepted
allowable. At and near the web stiffeners the shear stress distribution
would be quite complex. The usual equation should apply between stiff-
eners, however.

It is suggested in some publications that loads within a distance
from both supports equal to the height of the beam should be neglected
in vertical shear calculations which are used in the computation of
horizontal shear.7 This would yield somewhat lower calculated values
but still would not explain the great difference between theory and
performance.

Inasmuch as no failures occurred in such a significantly large

number of tests, it was concluded that horizontal shear was not the

controlling factor for beams within the range of sizes tested. This

-.- __--_-A

 

 

 

rig-I012
”humour-nun
nag-391mm: “in.”

 

 

 

30

may be attributed to the fact that very large safety factors are used
in determining an allowable shear strength for Douglas-fir plywood.

All other calculated stresses in the test beams were within ac-
cepted allowable values at a deflection of 1/360 of the span. The
overall aspect was a rather well-balanced beam of a very practical
nature.

The deflection of the beams appeared to be regular so as to
form a smooth curve; no points of concentrated bending were apparent
at splices or other locations. This feature is demonstrated in
Figure 14.

The reduction in stiffness of the beams with panels removed was
as expected. The reduction factor increased with the distance from
mid-span, becoming a maximum at the last panel over the support. Thus,
the increase in deflection is primarily the result of horizontal shear

stress.

 

31

 

 

Figure 14 - Loaded Beams Forming a

° Smooth Curve of Deflection

32

CONCLUSIONS AND RECOI‘JEFENDAT IONS

On the basis of the observations made and the results found in

the investigation, the fellowing conclusions have been drawn:

1.

4.

5.

Nail-gluing is an effective as well as practical means

of joining members. A slightly heavier-than-recommended
casein glue mixture promoted good gap-filling character-
istics while the 8 in. staggered nail spacing provided
adequate pre ssure .

Beams of this description should be designed on the basis
of allowable deflection at midspan. The usual methods of
structural analysis would seem to apply.

Butt-joining of the 2 x 4 flanges using 1 x 4 lumber was
of sufficient strength to support loads of 2%-to 3%-times
those at allowable deflection. Splices should be located
as close to the ends of the beam as possible, preferably
not within the central two-thirds.

No failures occurred in the web due to horizontal shear
stress as conventional calculations and design load would
indicate. A well-balanced beam would not result if design
was based upon such unrealistic values.

Compression at the supports may occur in the case of heavily
loaded beams. Adequate bearing must be provided to limit
this crushing stress.

Panels in the web may be cut out to accommodate heat ducts,

etc. if the proper load-reduction factor is employed.

33

7. Beams of this description should be used in protected,
ventilated locations. They should not be subjected
continuously to high humidity conditions. Q

8. Shop fabrication where conditions may be controlled is
the recommended_procedure. Labor specialization and a
permanent set-up are other inherent advantages which may
be gained.

9. These nail-glued plywood girders should prove more
economical than steel beams or wood-steel flitch plate

beams in many applications in residential construction.

LITERATURE

CITED

34

1.

4.

5.

9.

10.

12.

13.

14.

35
LITERATURE CITED

Pearson, R. C., Kloot, N. H., Boyd, J. D., Timber Engineerigg
Desigp Handbook, Cambridge University Press, New York, 1958.

 

Perry, T. D., Modern Plywood, Pitman Publishing Corporation,
New York, 1948.

Timoshenko, 8., Strength 9; Naterialg, Part II, D. VanNostrand
Company, Inc., New York, 1941.

Laurson, P. C., Cox, 3. J., Mechanics 2; Materials, John Wiley &
Sons, Inc., New York, 1947.

Parker, H., Gay, C. M., NacGuire, J. W., Materials ard Methods 9;
Architectpral Construction, John Wiley & Sons, Inc., New York, 1958.

Timber Engineering Company, Timber Design and Construction Handbook,
F. W. Dodge Corporation, New York, 1950.

 

 

___, Wood Handbook, U. S. D.A. Handbook No. 72.

 

Frees, A.D., Selbo, M.L., Fabrication gng Design pf Glued Laminated
Wood Structural Members, U. S.D. A. Technical Bulletin No. 1069,

1954.

 

 

 

 

igphpiggl Data 9n Plytggg, Douglas Fir Plywood Association,
Tacoma,Washington, 1948.

Countryman, D., Th2 Fabrication Egg Testing gf'Full Scale Plywood
Beams, Report No. 30, Douglas Fir Plywood Association, Tacoma,
Washington, 1944.

 

 

Countryman, D., Haskell, V. D., Plywood Beam Design Factors,
Report No. 58, Douglas Fir Plywood Association, Tacoma,
Washington, 1952.

Design 2; Plywood Webs in Box Beams,

 

 

...—...,
F. P. L. Report No. 1318, 1943.

Lewis, N. C., Dawley, E.R., Stiffeners 13 B93 Beams gag Details
2f Design, F.P.L. Report No. IBIS-A, 1943.

, Ieebink, T. B., Cottingham, N. S.,

 

15.

16.

17.

18.

19.

36

, Heebink, T. B., Cottingham, W. 8., Buckling 229
Ultimate Strengths 9f Shear Webs gf‘ggx Beams Having Plywood

Face grain Direction Parallel 23 Perpendicular tg thg Axig gf‘thg
Beams, F. P. L. Report No. 1318-D, 1944.

 

 

Markwardt, L. J., Frees, A.D., Approximate Methods 9: Calculating
Egg Strength 9; Plywood, F. P.L. Report No. R1630, 1950.

 

 

Radcliffe, B. M., Granum, H., Design 9; Nail-Glued Plywood
Gusset Plates Eng Solid Wood Splice Plates £93 Softwoods,
Purdue University Agricultural Experiment Station Bulletin
N . 613, 1954.

 

Angleton, H.D., £93 £9 Nail-Glue Hardwoods Egg Softwoods,
Purdue University Agricultural Extension Service Bulletin
NO. 441, 1957.

Anonymous, "Plywood Girders Roof a Warehouse", Engineering
News—Record, 128:625-6, April 23, 1942.

 

APPF
...NDI
X

37

38
APPENDIX A
Moisture Content and Modulus of Elasticity
Value of Ten Representative 2 x 4'3

Obtained from the Same Source as Members
Used in Beams

 

 

Sample M.C. E Species
1 17% 1.37x 106 p.s.i. D.F.
2 16 .98 Wt.F.
3 15 9. 1.01 n
4 21 1.36 ' D.F.
5 14 .98 Wt.F.
6 19 .54 D.F.
7 16 .93 "

8 18 .69 "

9 18 .60 "

10 16 .91 "
Ave. 17% .937x106p.s.i.

Testing Conditions -—
3
Center Loading -- E : p 1
48 y I
Span '- 48 in.

Rate of Loading -- 1/8 in./min.

 

 

mmwm B
TEST Beam Depth (in.)
Beam 14, 16 18 2o 22

 

1-1* 94 130 166 237 272
1-2 87 134 162 240 267
2-1 100 158 154 220 ---

2-2 115 136 178 223 ---

 

Table 1
Load at 1/360 Span Deflection (lbs./lin. ft.)

 

 

 

 

Test Bean Depth (in.)
peen 14 16 18 20 22
1-1 884x106 1222 1560 2228 2557
1-2 818 1260 1523 2256 2510
2-1 940 1485 1448 2068 ----
2-2 1081 1278 1673 2096 ----
Table 2 2
Experimental Stiffness Factors (lbs.-in. )

*First digit refers to test pair, the second to individual beam.

39

 

Section . I
Depth (iml Values
14 : 661 in.4
16 : 929
18 : 1245
20 : 1608
22 : 2018
Table 3

Moments of Inertia

 

 

 

 

Test Beam Depthfin.)

Beam 14 16 18 20 22
1-1 1.34::106 1.31 1.25 1.39 1.27
1-2 1.24 1.35 1.22 1.40 1.24
2-1 1.42 1.59 1.16 1.29 ----
2-2 1.63 1.38 1.34 1.30 ----

Table 4

Effective Modulus of Elasticity (1bs./in.2)

 

 

 

 

Test Beam Depth (in)

Beam 14 16 18 2O 22
1-1 .119 .150 .190 .256 .277
1-2 .111 .162 .185 .259 .272
2’1 .127 .191 .176 .238 ---
2-2 .146 .165 .203 .241 ---

Table 5

Shear Deflection Component of Total Allow-

 

 

 

 

able Deflection ( in.)

Test Beam Depth (in?

Beam 14 16 18 20 22
1-1 860 967 1937 1273 1281
1-2 796 997 1012 1287 1257
2-1 915 1175 962 1181 ----
2-2 1052 1012 1112 1197 ---—

Table 6

Extreme Fiber Stress in Flanges
at Allowable Deflection (p.s.i.)

 

 

 

 

Test Beam Depth (in.)

Beam 14 16 18 20 22

1-1 839 936 989 1194 1202

1-2 777 965 966 1210 1180

2-1 893 1138 918 1109 ----

2-2 1027 979 1061 1124 '---
Table 7

Fiber Stress in 25/32" Wood
Splice Plates at Allowable Deflection

(p.s.i.)

 

 

 

 

Test Beam Beam Depth (in.)
14 16 18 20 22
l-l 279 328 363 457 469
1-2 259 338 354 463 460
2-1 297 398 337 424 ---
2-2 342 343 389 430 ---
Table 8

Horizontal Shear Stress

at Allowable Deflection (p.s.i.)

43

 

Beam Depth (in.)

 

 

 

Test

Beam 14 16 18 20 22

1-1 14.44 17.00 18.72 23.61 24.15

1-2 13.36 17.53 18.27 23.90 23.71

2-1 15.36 20.67 17.37 21.91 -----

2-2 17.66 17.79 20.08 22.21 ---—-
Table 9

Rolling Shear at Joint Between

Flange and Web at Allowable Deflection (p.s.i.)

"11111111! 131111111155