UNIT STRESSES IN TIMBE R
1927
:4
FOREST RESEARCH LABORATOR
Y
LIBRAR Y
UNITED STATES DEPARTMENT OF AGRICULTUR E
FOREST SERVIC E
FOREST PRODUCTS LABORATOR Y
Madison, Wisconsi n
In Cooperation with the University of Wisconsin
MOOR FO EST FORUM IPORATOR Y
LISRA Ry .
... .
1
.
UNIT STRESSES IN TIIVIBERBY
,,' :,!. -
J-. A .. 1EWLIN, Principal Enginee r
The fact that there are several hundred species of arbOrescevt plant s
in the U'll4t?“te.take-samd .a greater mAtiPlioit,ofsue4 u:eOpl iDiAhe traipUe.s
has led nlaRy.leRg4neexe .4.o consider the subject9fmot and. J*,piVi:es.,
too indRfIrritle-aWoP14sato4 t o 4pltafY PvPE-g w4ptieTfi4a7-'6t%d5;V MPch ,
less a close anaIyets .
The speaker t s .purp .
is to correct se.le 7 m1sconceutiO48' . PotfiWDe
wood . Under his supervisiem about 700P0OO ' seie4gth tests of iic.e%bltO
performed, many of which were on structural tubers . Many others wert'Ade
during the World War in contaction with aeroplaz,e ionstrvwtiph .
I
People see pieces of balsa wood, O r read abgi t .,>
0 alY . aye
sixth as heavy as water l, and.0out other speffq:bOkl$ Ivri-Ati% ;ishan w Vkol,
They hear that the Army and . pp.vy are using for 4-fatia 0*esS in extren
fiber in bending of , 9 x 400 W43?er . sq . "in &P , airi'CW0tA,itoplanes have._ , _
been examined and evi4ence's,' fOv.d that the tim*'iOlems: Abeen subje(OW
to a stress of 11,000- to 12;000 lb .'without fal,Waid:then) ,as
climax, they see a table of stresses'!recotmendeiPlioresil P pag45P :
Laboratory, in which a "Select" grade of spruce- it dated a=s 6:6 d
lb . per sq . in. Usually, engineers and contractors stop here and say, . "Well,
since the factor of safety for spruce is evidently about 12, the ?4ad- PaP
cated by a stress of 1,140 M .,- way be multiplied by 4 and 'frtil n bJeAtt'h:1. a good margin of safety ; and with this wide mOgin..alpoet any wdO ies can
be substituted for another in the same sizes, ."- ' TheedG0,hotl'ob* 40ed tiler=
; e - wa:FF1.1eW. - Onc
04
.oughly,rtewdncoivfthsmargn
does fail . Most of the failures occur in concrete forms seBI-clAll- wRot& I
The optimistic contractor thinks he has this, large factor of safety, h e
quadruples the load p and down goes the structure .
resented at the special meeting of the Structural Division, America n
Society of Civil Engineers, Oct . 2S, 1925 and published in Transaction s
of American Society of Civil Engineers, Vol . 91, p . 387 (December 1927) .
R1221
does io t ` ieed1 .te-bicame an ex •rt in the
s, or does he have to be familiar with the proper order to use wood intelligently . It ins: heto4sually
stinguish between species that are gori, t. e* . ike itt
under these conditions their strtgths would,
]
far as stren th is concerned, for instance, it '
. istinguish etween t ;. .- fifty-sev different }
e divided
lair *era
impossi in f ac
1 L -~ 4 044
e
lumb e
Pariou
classed as Sbuthern'e . If
eaf' pine
iec ?l 40
loblo l
has the growth and
perties o f
stance of't Id a
safi t
%3af ; or a
in weight
e 6f true 1gleaf Mir be soft and
al
of "Aipine than th e average of short-.
leaf pine
Arkansas oft pie" is selected .
,Af
The extremes of the species,
use for structural timbers ,
n'e
'when considered from a strength int, are Eastern hemlock and whit e
fir, the weakest, and southern yellow pine and Douglas fir, the strongest .
The latter have about one and one-half times the strength and stiffness o f
the•$e "ock-' and. white fir . Southern pine and Douglas fir furnish not onl y
'most o America's structural timbers, but also more than one-half of all it s
1140e r
11kh
A
A
The chi
ics of- the species, their inherent properties, th e
ce of certain defects and their relative likelihood to develop othe r
rvice, and the influence of duration of stress are all con, vg working stresses . Different limitations imposed o n
Mkared in ass =
factA;-g j,
~tnmon to all species explain the difference in recommende d
taking stresses for wood in aeroplanes and in house construction . Thes e
faetars will be illustrated in the inverse order of their importance, a s
follows : (1)
variability of the strength of the clear wood ; (2) th e
tress ; and (4) defects ,
rte.,
;
~- , ;r r
. .t!
r~;~
4
Str
th of clear ' Rio „• r ; ~r
~. r
~
1
;-,,';d
1
.14 17.41,
reen condition . Si
's tak e
able on this speci :
gracti c
y {~ ry . .r
ce in airs
4;
' .a .carv e
,►
God . It w-'
.
s (6
iit of 1,3010 giv
.•e -*ariation , lying . ` va
•orres Mthnito the hig a
.w the
age . it density curve f•
T
.anal .
ity
ve .
-
g
• . :•
ge, whi~f
e a
point on - he curv e
y given species i
_-fi t
properties of a specie s
tional but greater tha n
the type shown in Figure 1 .
ti
In assigning working stresses, primary consideration is given to th e
one-fourth of'• he pieces lowest in strength . For aeroplanes, a lower limi t
for specific, gravity is set which eliminates most of the lower one-tenth o f
the material but does not change the most probable value . In southern
yellow pine and Douglas fir it is possible to eliminate a little of thi s
low material (along with a little of the higher) by the "ring" rule or "i rat e
of growth" rule, A more satisfactory rule for selecting material of thes e
species is that known as the density rule, by which density is determine d
from an estimate of the proportion of summerwood ,
Variation of strength with moistur e
In Figure 2 is shown the influence of moisture on the strength o f
spruce wood, These are old curves . They represent work that was done quit e
a number of years ago on-relatively few pieces of small size . More recen t
data show the average strength,of the green material to be slightly highe r
than these early tests indicate . The horizontal line at the right of eac h
curve represents the strength of the green material ; when the fiber saturation point is reached, ,that is, when all the free water is out of the cell ,
the cell wall begins to lose its water and the strength of the material
goes up . quite rapidly . Since in any ordinary drying some of the cells wil l
contain free water while others are below the fiber saturation point, i n
practice the curve } instead of being straight to the fiber saturation point ,
and then having a sudden break upward, is a transition curve of varyin g
radius, depending on numerous conditions . As between this ."green" strengt h
and the strength at , 15% moisture, the most recent curve for modulus o f
rupture shows practically an increase .of 50 percent . The increase in th e
shear strength is not sufficient to be of much importance in structura l
material .
Duration of stres s
The diagram, Figure 3, shows the influence of duration of stress o n
the strength of timber . A logarithmic scale is used for the time co-ordinate because it illustrates the law better than a direct plotting . Th e
upper curve is for fiber stress at the elastic limit and the lower one fo r
modulus of rupture . In each curve the value from the Forest Product s
Laboratory's standard static-bending tests, which require on the averag e
about 5 min ., is taken as 100% (point marked a, Fig . 3)„ The point farthes t
to the right (b) represents the average strength of pieces which broke i n
from six months to a year under dead load . No rupture strength data ar e
available for periods of t,me much less than 1 sec . The point in the extreme left-hand corner (c) = represents an elastic limit stress in impact *
In developing the curves little weight was given this point (even thoug h
it represents thousands of tests), since there are several factors whic h
cannot be eliminated and tend to raise the point slightly .
2
From Bulletin 556, U .S . Dept . of Agriculture .
R1221
-3-
-
-r'jlf-76' •41-011.--di
ane flight i is impossible to maintain the condition s
wh
stress for more than a few seconds, in which case ther e
It'4actor
. almost V-wo as compared with the load that the same materia l
wi
long time--as in a library floor, for instance . This show s
y
L
r st
ures should be designed for static loads and the impac t
$&ve log ' neglec$ea unless the impact stress exceeds the stati c
I
Ole to the live ]oad .
a►'t?
.1
-t"
t'
leer impor t
ing the
i-lo-i
fects . O
- f the mos t erious de . :_ .- s decay ,
be allowed in any important structural member . Timber tha
air-dry and constantly kept in that condition will not de c
in contact with other material that is decaying . A mill f l
phi al Pa ., recently examined, revealed that wet pine had bee n
a sub -. floor, with paper and wet concrete below and a tight to p
thoroughly oiled, making the conditions ideal for decay . The sub-floo r
rotted out completely, and the fungi passed through the top floor and bas e
moulding into the longleaf pine columns, rotting them out in two years .
These columns in the ground, unprotected, ought to have stood for seven o r
eight years . Decay must be prevented in timber buildings .
r 4
Figure L . shows a knot, another serious type of defect in timbe r
which often reduces the strength of a piece almost to zero . A knot is a
part of the limb which,, as the tree grew, became a part of the trunk . Th e
fibers run from the bottom out into the limb . On the top they produce a
kind of crinkle around the knot . Some of them seem to end there, but a
number run through into the limb, The strength of the wood along the grai n
ern& across the grain is tremendously different . The variation is fiftee n
to fortyfold, Therefore, it would be better to have a hole the size of th e
knot proper than to have the knot with its attendant cross-grain .
The influence that a knot on the tension face at the point of maximum
moment . exerts on the strength of a green beam is approximately measured b y
the ratio of the diameter of the knot to the width of the face . The diameter of the knot is measured between lines parallel to the edges of th e
face .- .That is, a knot one-fourth the width of the face g ill -reduce th e
Ift0elp4ilt,ettilti 25 percent . The same knot on the compression side would hav e
It ;lone-half this influence .' Large knots have a somewhat greate-r influewe on the strength than-is indicated by this rule, owing-to the rdlhtivel y
larger distortion of the grain around them . This effect is' 'taken care o f
in the grading rules- recommended by the T . S . Forest Products-Laborator y
tintvOmstu Select" grade will insure 75% of the s't'rength of the olear' good.,'
Large timbers never dry to . as low a moisture content as small stocky
ate$ they develop more checks, especially around knots . The result is that
although the average strength is, increased by dryingi_ about- 25'g of th e
~~~'f3.ii dhow no increase, and, therefore, no increase in working stress o n
R1221
account of seasoning
less in thickness-, a
plane stock the full
stress of 11. sec . may
is allowed . In the higher grades of joists, 4 in.. an d
slight increase in s-tr .eAgth- s Fecagn :zed . In aerostrength of clear weod:at 15` meisturt- and duration- off '
be developed .
How working stresses
are determir-ie d
.
In arriving at a safe stress for a ''Select" grade of spruce structura l
timber, the starting point would be 5,760 lb . . per sq . in ., which is th e
average modulus of rupture for small, clear stock tested green at standad . . ,
speed . Reduce this by one-fourth to take care of variability of the clea r
wood ; make another reduction of one-fourth to take care of the maximum defec t
permitted in=the grade ; then take off about seven-sixteenths for the-dea d
load effect over a, number of years, This give s
5,760 x - x - x i6 y,
or 1,•820 lb . per sq ._ in .
as the stress under long-time leading at which an-occasional timber- . ' possibly
1 in 25, would be expected to fail . Thus, the recommended s'tTess of . 1,10 0
lb . per sq . in . gives a factor of safety of 1-2/3 for-this bad timber an d
this worst condition of loading . The average timber has a fa-. or of safet y
of about 2. 1/4 for long-time loading, and for a few minutes will• skoW a ;e-or eEf
safety of 4 . With average timbers under impact there would be a factor o f
about 6 . When the influence of duration of stress on the Strength of timbe r
is considered, it is apparent why it is recommended that timber be designed '
for dead load and for a long period-of ' time negl:ectirg any impact stresse s
unless they exceed the stresses due to the ' live load which produces thos e
stresses . Wh:Wiever tests of structural timbers are available they ar e
found to check these factors quite closely .
For-make years the Forest Products Laboratory' refrained fro m . assign- ,
ing any stress to-timber and it always has refused to recommend ' s.tresse s
where the grading rule does not limit quite closely the weakest timber tha t
could be accepted . The stresses were finally recommended at the urgent re quest of engineers, and although the lumber interests were given attention :
they have had no vote in fixing these final values . '
-
1
y
MI= IN=
•
IEEE
ENE=
Kil,„,
NEI ME
°
'‘
AEI■■■ ummmin ■■■ =4
IIEEWI
•
111111111NM
UM*
o
t
SPRUC E
12 1 000_
11,00010,000
C)
\a
H
t,
,
o0
5
50
15
25
35 40 45
20
30
MOISTURE-PERCENT OF DRY +lE I GHT
Fig . 2 .-Comparison of the various strength values of spruce with variatio n
in moisture . (The two upper curves are from bending tests an d
the lowest one from compression at right angles to grain . )
81221
10
e
r
•
Figure 3 .--Relation between mechanical property o f
Sitka spruce and duration of stress .
81221
h
ti
(la
to
t
t..1
--amo
W
h
ti
O JW
h
o
R rry
W
SAd0 0/
o
dQ
2 t
.
0
ve; /
51100H 0/
Q
5,Y/70H 5
ao
0
-I
k
h
Z
w
Y
ti
o
V
'K
p~.
=
p O
g
is
-I Z.
h
o
p p p ti
Q
a
Q
p
h~u
Z Q
h
C}
O
.z
y
k,
p
4i
J
v
ti
h
¢ h
4s
Q
ti
W ?
W h
Z
o
°
2
h
tZ,
ca Q
h o W
4.)
vt
aFo2
y q
u h
¢
FIB o '2c k `"
L~
QOp
44
p h
k
o ip 44
W
o o4
av h
c 1 oq W :t
m
a -e
a•
0
=
(cC
Y00H /
N/A/ o f
N//4/ o r
1
-
-N/A/ S
'/v/A./ /
•
.735 O r
O
1
0
031 0 /
_T
035 S
•
o
a
_
••
J •
1
D5 /'0
1
1
1
1
+
+
J
f
0
i
o
v
v,
e
~v
•°
o
m
h
(1N3O i/3d) 31//N/W 2/1d 71AfrWl d0 HON/ 50/0 iv 9N/0N39
0/1GLS O61 VONd1S N/ 031031 /l/3d0Nd 2t/7/12//S 01 t1113d02'd 7d,7/NVH.73W 10 0/1VE/
x.
m
0
0
0
(la
Fig . 14 .--Knot showing characteristic differ ence in behavior of fibers on lower an d
upper sides .