/j/R. / JJ
w 7_ 11/ 2-- 7._ ~ / 6J;-t
File D 00.12/ 122/No. 3296
Auburn University Libra ries
llllllllll~llllllllllllllllllllllll~llllllllllllllllllllllllllllllllll
N 3 1706 025 85187 9
AIR CORPS INFORMATl CIRCULAR
PUBLISHED BY THE CHIEF OF THE AIR CORPS, WASHINGTON, D. C.
Vol. VII December 1, 1930 No. 656
TEST OF ARMSTRONG-WHITWORTH STEEL SPARS
UNDER COMBINED AXIAL AND TRANSVERSE LOADING
(AIRPLANE BRANCH REPORT)
UNITED STATES
GOVERNMENT PRINTING OFFICE
WASHINGTON: 1930
Ralph Brown Draughon
LIBRARY
JUN 18 Z013
Non·Oepoitory
Auburn University
--
.... ~ ,.,., ~
~ Lit'>~"
TEST OF ARMSTRONG-WHITWORTH STEEL SPARS UNDER
COMBINED AXIAL AND TRANSVERSE LOADING
Prepared by C. G. Brown, Materiel Division, Air Corps, Wright Field, Dayton, Ohio, August 4, 1930
OBJECT
The object of this test was to obtain data for comparison
of the Armstrong-Whitworth steel beams with
a se ries of 6X-inch beams previously tested by the
Materiel Division .
RESULTS
To compare these beams with those tested by the
division is difficult because of the different test conditions
employed.
In Air Corps Information Circular No. 622, Tests
on 6).{-inch Metal Spars, the test length was 96 inches,
and the distance between the applied side loads was
51 inches. The English beams were about 60 inches
in length with the side loads applied at the one-third
points of the span. Further, the English beams were
shallower than those of Air Corps Information Circular
No. 622, being but 5~ inches in height. These beams
were tested with varying ratios of side to end load,
whereas the method of Air Corps Information Circular
No. 622 was to employ a constant ratio of side to end
load of 20 per cent.
From Table 1 it may be seen that the rigidity per
unit weight of the English beams is less than any of
those taken from Air Corps Information Circular No.
622. However, the fiber stress developed is considerably
higher. If the comparison were based only on
fiber stress, the beams would be excellent. If the comparison
were based only on rigidity, they would be
poor. Just which quality is· to be the more highly
stressed is a matter of the qualities of the airplane in
which they are to be used. For extremely high speeds
their lack of rigidity might be regarded with suspicion,
espec ially if the wing tip overhang were great. If the
magnitude of the spar or wing tip deflections 'Yere of
secondary importance, then cost would enter as another
detrimental factor from an experimental point of view.
The sections used in these beams must be drawn to
shape through a series of dies or rolls, hand fo rming
being out of the question. Hand riveting, too, in
such thin sections, appears impractical. Experimental
development of this type of spar would call for considerable
special equipment that would not ord"inarily
be used for production purposes.
TABLE !.-Comparison of beams
Armstrong-Whitworth
Riveted strip steel construction
(a) w=lb./ft. 1.6 _______ ______ ___ _______________ _
(b) ~~r;~·rii·~.000---_~~:: :::: :: :: :::: :::::::: l
(c) EwI- -24,400,000_ ______ ___ ___ _____________ ____ _
(d) Max. Comp.:
Stress, 158,800 lb./sq. in __________ ___ ___ <e> Mas.t~~~P·'-46,650 ___ _____ ___ ___________ _
"\Vt. of bar l" x 1" x 12"
21928-30
No. 20A
Hexagonal dural flanges,
corrugated dural web,
corrugations horizontal
2.09
83, 700, 000
40, 000, 000
31, 330
25, 800
(1)
No. 23A
Dural corrugated
web
l. 85
73, 000, 000
39, 400, 000
40, 360
33, 250
No. 20
~
'2d
Wood box beam
1. 93
78, 330, 000
40, 600, 000
No. 32A
Steel tube Warren truss
2. 04
76, 000,000
37, 000, 000
80, 390
23, 600
SPAR RIGIDITY
In order to determine the effectiYe EI of the beam,
the theoretical deflections from t he expression
1[ WJ sin J J YL=p - L- -W.
2 cos 2J
were plotted against J , as in Figure 6.
W = Side load at each point of application in
terms of P
Actual a=21 inches}
Actual L = 62 inches between ball centers
P =End load
J= ~EJ
Then, for a given deflection, it was possible to determine
P from the deflection curves of Figure 7, and
J from Figure 6. EI was then computed from the
expression EI = J 2P.
Though two beams were tested, computations of
rigidity were made on but one, 300- 2, as an inspection
of the deflection curves of t he two, Figures 7 and 8,
showed this beam to be the more rigid of the two.
.l!Uimu.m /i'ber slress
JJeam 300-Z
p i/-=~~~2~"1 =-~~~~§-~,
FIGURE 1
2
The beam carried 7,800 pounds end load, and fai led
at~about 7,850 pounds.
Moment of inertia, 1.159 in.3
Distance to extreme fiber, 2.733 in.
Section modulus, 0.4194 in.2
Area, 0.259 in.2
Deflection at center for load of 7,800 pounds=
0.619 in .
Moment due to side load = 0.30 X 7,800 X 21 = 49,100
in. lb.
Moment due to end load = 7,800 X 0.619 = 4,830 in. lb.
Total moment, 49,100 +4,830=53,930 in. lb.
Mc 53930
I = 0_4194=128,700 lb. per sq. in.
p 7800
11 = 0_259= 30, 100 lb. per sq. in.
Total fi ber stress before failure, 128,700+30,100=
158, 00 pounds per square inch.
In Table 1, line (e) a strength weight ratio based
on fiber stress, and weight of material has been included.
It is the ultimate compressive stress divided by the
weight of a bar of t he same material 1 inch sq uare
and 1 foot long.
The weight of a bar of dural 1 by 1 by 12-inch=
1.213 pounds.
The weight of a bar of steel 1 by 1 by 12-inch=3.4
pounds.
DESCRIPTION OF BEAMS
The Armstrong-Whitworth beams are of built-up
steel construction. The general details of the beams
may be seen from Figures 9 to 13. Hollow steel rivets
are used throughout.
The construction of these beams is such that the
distribution of flange material may approximate the
stress distribution. The compression and tension
flanges are the most highly stressed portion of the beam,
occupying approximately the outermost 120 per cent
of the total flange, and are of the heaviest gauge used
in the beams. The cornices, joining the flanges and
webs, being less highly stressed, are made of lighter
gauge. The web is the lightest of the t hree major portions
of the beam. The pyysical properties of the
material used were taken from the Armstrong-Whitworth
report accompanying these beams, and are
presented in the following table:
T ABLE 1- A.-Physical properties of beam material
BEAM 300-2
Com- Com- I 'fen~ pres- pres- Web sion sion sion flange flange cornices
----------!·--------
Thickness (in .) ______________ 0. 0198
Proportional limit (#/0') ____ 51, 000
Proof stress (#/0') _____ __ ____ 145, 000
mtimate strength (#/0') _____ 203, 700
Brinell hardness__ _______ ____ 429
Modnlus of elasticity, mil-lions
of (fl/O')____ __________ 28
0. 0120
51, 000
174, 000
190, 000
473
28
-------- ~----~
BEAM 300-1
0. 0102 1 0. 0200
67, 250 67, 250
164, 000 159, 500
197, 000 203, 500
437 493
28 j 28
-
Ten-siou
cornice
0.0120
78, 500
159, 500
195, 700
437
28
and Com- .
Tension I
com- pression Tension Web
pression cornice cornice
____________, _ fl_an_ge_s ---------
Thickness _________ ___ ___ ----- ____ _
Proportional limit (#/O'l---------Proof
stress (fl/O' l---------- -- ----Ultimate
strength (#/O' l---------- Brinell hardness _________________ _
Modnlus of elasticity, millions of
(#/O ' l--------- ----- ------- ----- -
0. 020
67, 250
158,000
203, 700
446
28
0. 0125 0. 0123
56, 000 78, 500
150, 000 165, 000
186, 000 198, 000 4:: I 4:
0. 011
56,000
149, 000
186, 500
413
28
The above sections are heat treated a fter forming.
The joining strips are hardened and tempered in the
flat, and are not heat treated after rolling to section.
The proof stress mentioned in Table 1-A is a figure
corresponding in a way to a yield point. It is the stress
at which the departure from a straight stress strain
line becomes 0.1 per cent. This measure is resorted to,
it is believed, because the actual stress strain curves
of the test specimens were a very smooth curve clear
to the ultimate, in most cases, with no break that would
indicate a definite yield point.
3
The following are the section properties of the beams:
Beam 300-2
Area of compression boom
Area of tension boom
Area of web
= 0.1132 sq . in.
= 0.1137 sq. in.
= 0.0319 sq. in.
= 0.2588 sq. in .
= 0.005 in.
= l.159 in.•
Total area
Shift of C. G.
Moment of inertia
Distance to outermost fibcr = 2.763 in.
Section modulus = 0.4194 in.a
B eam 300- 1
Arca of compression boom
Arca of tension boom
Area of web
= 0.1157 sq. in.
= 0.1149 sq. in.
= 0.0344 sq. in.
= 0.2650 sq. in.
= 0.005 in.
= 1.166 in.4
Total area
Shift of C. G.
l\fomcnt of inertia
Distance to outermost fibcr = 2:727 in.
Section modulus = 0.4278 in.
APPARATUS
The apparatus used in making the tests is shown iu
Figw·es 2 and 3.
p
A
-----
f '
/ li ___
B'
T"~~ -
F ,
-----
A I
p a b~
FIGURE 2
D
c
Member AB was a
steel bar ! ~ inche s square
and 17 inches long. The
remaining members were
1 ~ -inch by ){-inch stock.
A flat plate with a ball
scat was placed on each
end of the beam, and a
round steel ball, E, used
to give approximately a
pin ended condition. A
knife edge was placed at
point F and was moved
between A and D to vary
the ratio of side to end
loads.
The action of the apparatus
is as follows:
The end load P is divided
between A and B
such that the load at A is
This load causes tension in member BD, the component
of which produces the side load on the beam. This
load is
Deflections were measured by means of two Wisler
dials attached approximately at the center of the
beam. While it was intended that the steel balls used
at the ends of the beam should give a pin ended condition,
a comparison of the deflections obtained, and
those computed from the "Precise" formula for this
type of loading showed that there was an end moment
acting on the beam.
Substituting the measured value of "a" and " L "
in the equation on page 2 gave the solid curves of
Figure 6. When the effective EI was computed from
these latter curves, it exceeded the geometric EI of the
section.
TESTS
The testing was performed by D. M. Warner and
H . Sedam in the physical testing laboratory of the
Materiels Branch. For a more complete description of
the apparatus and method of testing, reference is
given to a report of these tests to be prepared by the
Materiels Branch .
TABLE 2
20 30 Sin~ Cos '!Q J J J J
------
40 0.500 0. 750 0. 47943 0. 73169
50 . 400 .600 . 38942 2534
60 . 333 . 500 . 32729 . 87758
I
80 . 250 . 375 . 24740 . 93(}!9
JOO . 200 . 300 . 19867 . 95534
120 .166 . 250 . 16589 • 96891
TABLE 3
1 [nr J Sin J Deflections from Y l:_ = p L
2 Cos ZJ
I
40
50
60
80
100
120
J
---
40
50
60
80
100
120
o. 62 0. 92 1. 235 ]. 55
. 36 . 53 . 71 .90
. 24 . 36 .48 . 59
.12 .18 . 24 . 31
. 0 5 . 13 .15 . 20
. 055 .08 .11 . l4
I
TABLE 4
21 31 8. 21
J J ll17
----
0. 525 0. 775 0. 49904
. 420 . 620 . 40776
. 350 . 517 . 34290
. 2625 . 3975 . 25949
. 210 . 310 . 20846
.175 . 258 . 17411
TABLE 5
1.86
1. OS
. 71
. 36
. 23
1. 55
Cos~
I
J
0. 71441 I
. 81388
. 86946
I
. 92202
. 95233
. 96689
[
W J Sin.<:£ J Deflections from Y .!:. =pl: L J TV a a= 21
L =62 2 Cos - -
J
1--
40
50
60
80
100
120
- 2J
1;~0.10 0.15 I 0.20 II 0.25 0.30 I
_ _ , ___ I
0. 70 0. 98 ' l. 39 ' 1. 75 1 2. 50
. 405 . 61 . 81 1. 01 ]. 22
. 295 . 40 . 54 . 67 . 81
.15 .23 .30 .38 .46
.09 .14 .18 .23 .28 .oo .09 .12 .15 .rn
•
·~
~
4.
TABLE 6.- Delermination of E I a=20 L =60
I
20 per cent 30 per cent 40 per cent 50 per cent 60 per cent
Y ~ l_J_ _P_ 1.c!~ooo ___P_ 1.i!~ooo _J_ _P_ 1.c!~ooo _____ P_ 1,c!~ooo 1
__ J_ l_ _P _1. c!~oo0
0.10 88. 0 3,900
.15 72. 5 .5, 715
. 20 64. 0 7, 335
. 2.~ 38. 5
. 30 -----r====:
.\lean EI=30,200,000.
Total
trans-verse
load
% P
I
30. 2 112 2, 567 32. 2 126. 0 1, 965 31. 2 -------- --- --- -- -- - - ----- -
30. 1 90 3, 500 28. 35 100. 0 2,900 29.0 116. 5 2, 350 31.8
30. 0 77 4, 933 29. 2 87. 0 3, 833 29. 0 100. 5 3, 150 31. 8
---------------- ----- -- ------- -- - 78. 5 4, 700 29.0 88. 5 3, 925 31. 4
--- --- --- - ------- ----- ---------- 73. 0 5, 650 30. I ---- ---- - - --- --- ----------
TABLE 7._:Combined column and transverse loading
IlEAM 300-1
I
20 30 40 I 50
-
-------- ------ --
123. 0 2,000
106. 0 2, 667
95. 5 3,300
87. 5 3, 950
60
I -
I 'l"'ransverse deflec-
I
Transrnrse defiec- J Transverse dellec- Transverse defiec- 'rransverse deficc-tion
inches tion inches I tion inches tion inches tion inches
l.ood P
Left Right I Left Right Left Hight Left Hight Left Hight
-------- ------------ ---- -------- -- ·--- ----
500 0. 0144 o. 0132 I 0. 0306 o. 0200 0. 0275 o. 0275 0. 0332 0. 0335 o. 0370 o. 0395
1, 000 • Q?..SO . 0276 . 0424 .0400 . 0557 . 0532 .0696 . 0685 . 0744 . 0758
I
1, 500 . 0430 . 040
I
. 0632 . 0606 . 0843 . 0810 . 1032 .1027 . 1145 . 1168
2.000 .0580 . 05.52 . 0854 .0816 . 1131 . 1095 I . 1386 .1383 .1550 . 1570
2, 500 . 0734 . 0700 . 1074 . 1026 . 1415 . 1372 . 1770 . 1738 . 1970 . 1975
3,000 .0884 . 0840 .1298 .1240 . 1725 .1670 I . 2140 . 2100 . 2408 . 2410
I 3, 500 . 1036 . 0992
I
. 1530 . 1466 . 2037 . 1975 . 2520 . 2460 . 2850 . 2845
4. 000 .1200 .1148 . 1770 . 1694 . 2349 . 2272 . 2910 . 2842 . 3296 .3278
4, 500 . 1362 . 1304 . 2010 .1920 . 2665 . 2577 . 3268 .32'27 . 3770 . 3710
5,000 . 1532 . 1470 . 2250 . 2144 . 2997 .2886 . 3720 . 3610 . 4224 . 4173
5,500 .1694 .1620 . 2490 . 2378 . 3321 . 3190 . 4132 . 4005 . 4670 . 4625
6, 000 .1864 . 1784 I . 2740 . 2620 . 3663 . 3510 . 4542 . 4410 . 5118 . 5108
6,500 . 2020 . 1940 . 2990 . 2864 . 3999 . 3835 . 4950 • 4818 Stopped
7,000 . 2190 . 2104
I
. 3230 . 3106 . 4341 . 4180 . 5354 . 5260
7, 500
--- --- ----
30. 2
30. 0
30. l
30. 2
I
I
I
Beam was not injured so far as could be noted. De.Oection readings are for left and right dials. J,eft dial attached 1 iuch above center. Itight
dial attached at center. Left dial on compression flange side. Right dial on tension flange side.
'l'otal
transverse
load
3 P -1
Load P
500
1, 000
1. 500
2, 000
2, 500
3,000
3. :i00
4. 000
4, f>OO
.\, 000
5, 500
6,000
fi,500
7,000
7,500
7,800
7,850
TABLE 8.- Combined column and transverse loading
IlEAM 300-2
20
Transverse deflection
inches
Left Right
0.0130 0. 0121
. 0265 .0240
. 0392 . 0356
. 0520 . 0482
. 0646 . 0598
. 0780 . 072fl
. 0914 . 0866
.1054 . 099fl
. 1l90 .1128
. 1336 .1276
.1476 . 1412
. IG24 .1556
. 1770 .1700
.1924 .1846
---------- ----------
---A.i)i)roxi'11iiic- ---
30 40 60
Transverse deflee- l-T-ran_s_v-ers_e_d_e_.O_e_c--
1
-'1-'-ra_n_sv-ers_e_d_e_fl_ec--- l-,-1-'r_a_n-sv_e_r-se-d--ef-lec-tion
inches tion inches tion inches tion inches
Left Right Left Right Left Right Left Hight
0. 0202 0. 0190 0. 0263 o. 0255 0. 0324 0. 0310 o. 0376 o. 037fl
. 0404 .0380 . 0523 . 0483 . 0648 . 0622 . 0736 . 0728
. 0600 . 0570 . 0773 . 0717 .0994 . 0940 .1114 . \IOI
.0802 . 0760 .1037 .0971 . 1308 .1236 . 1492 . 1481
.1004 . 0948 .1303 . 1241 .1628 . 1550 . l8f>6 . 18!\ll
.1210 .1150 .1577 . 1515 .1944 .1862 . 2256 • 22!"}1)
. 1424 .1356 . 1853 .1795 . 2290 . 2196 . 2650 . 2fi4fi
. Jf>34 .1558 . 2129 . 2061 . 2614 . 2516 . 305fl . 30fii;
. 1844 .1772 . 2423 . 2349 . 2924 . 2850 . 3468 . 3496
. 2074 .1996 . 2715 . 2(i39 . 3244 . 3174 . 3890 . 3908
. 228G .2202 . 2993 . 2915 . 3578 . 3524 . 4316 . 4336
. 2520 . 2426 . 3287 . 3217 . 3958 . 3914 . 4740 . 4786
. 2750 . 2656 . 3579 . 3519 . 4374 . 4318 . 5180 . 5231
.W86 . 2880 . 3843 . 3795 . 4794 • 4714 . 5726 . 57ll
---------- -- -- ------ --------- -- --------- --- ------- ---- ------ . 6132 . 6251
---------- ---------- -------- -- ---------- ---- ------ :::::::::r -----¥ai1ed . 6656
----- ---------------
Bcalll 300-2 fai led in compression mcmhcr about 1 inch helow center. ·Failuro was ahrupt. Jt was due to buckling rnther th:111 1o y icldi11~ tir
11taLcrial. .f\ II loading wa.s by hand operation or the Olsen 20,000-pound mnch inc. Iic rt dial on co rn press ion Hauge sillc. Hight cl inl on lcrisio11 rl;:i ngc
side.
-
5
-
FIGURE 3.-Combined axial and transverse loading jig
F!G"O'R! 4,- Loading jig set-up in 20,000-pound Olsen testing machine
(6)
:FIGURE 5.- Failure or beam 300- 2
r"
~
1.. ~
~
~
~
ilj
~
t
~ ·
o.s ii)
0
- zo
8
y:, . 1 [ WJ s in. !f wa
z P cos},
,A
F IGURE 6
,,
r;,
a- zz • L c 6Z"
a - zo· L - 60'
00 h o
j
J
J-t
H ~To z a l side load in% end load
1-1-
- 1-H -++
±;:fc 1 t' 1 ifu iT :i +i:o.1 m++ =to.2 q: ± a3 ::ti:±rlo.4j
H.Def t ec;,u on wch_es
.. FIGURE 7
i. 9
8~;41'1 soo-z
Cornbi n•a' lood/n~ lt!..sf
7000 . ,
~· -+.._~ :Fi, .. ~ +i
:( ~I 1 4 ~ ij
6000
' '
t>,
sooo] '
~
400 1g
<:)
'ii; '
~o, o~ '
'
ZOi-~ R-r<tfco of (o a{ .ride {OIJfi
lb end lo11d
1000
0 .De/f~ctio.n - l";zdzes
l::t!. 0 i:u: q.1 1:1aZt:t:t::i::t:UO.J t::t::tt:t:t:l/l4t 5~
FIGURE 8
l
When used as ii cover plale cul off lip,s dS
0
I.35 .18
1.35
10
.18
A
.03 R.
. r3R ~
'o3R. ~
Web
:u·wun.E t2
2.7
.l/R..
/
.04
11
.24
I~
i--'--- . 2 _ __ __,,_. Actual,
.s1.:ze
FIGURE 10.- J oin ting strip
f. OF RfVETS -._
--------
20°46
1.0
N OTE :
.04 R . WHEN USED A S A
COVER PLA TE, CUT
<i. OF RIVETS OFF L /PS AS SHOWN
-~·-- 1.0
Car.ntce
FIGURE II
-
LIAS Riwded - tf ho!! ow
Rivets af J"pitc.Jz ,,
--- ao --------- ----; ..... 1--------z o --------~ ... ---------20------------l..,
h5Rowsf dta . .hollow rivets
~! \ 6 Rivds per TOW ,!-
View of e11d _pla fe L, 2. 7 -
£nd oJ ,;pecimeJZ fo .be Rat
ar.d pcrpendfcalar to axz's.
21928-30. (Follows p. 11.)
1.3750.D-.ECS.V.CiT.S T'l!BE
(Cham end.s a tu.be I
-~DIA. POP R. "ET8
j!J Ela. Jzo!lo.w rt~rd,s
.rz'de plates . 02.5 (ft[dc:
IJ. TD. 5_,,_t, Lz 1t:ie.ned.
'
~.--------_,,_-----}
I;.:--;:.------------=- =----:..::.~--~~;-~~--_ ,
~----3--~--
Weh n'v1Zled j- lzo
Air Service Information Circulars
