343 3khb
343 3khb
343 3khb
by
Roger A. Scheid
Stub Column Test and Residual Stress Measurement (A572, Grade 65)
by
Roger A. Scheid
May, 1968
343.3
TABLE OF CONTENTS
No.
Page
ABSTRACT
ii
1.
INTRODUCTION .
2.
SHAPE CHARACTERISTICS
3.
4.
RESIDUAL STRESS
5.
COLUMN STRENGTH
6.
LOCAL BUCKLING
10
7.
SUMMARY
13
8.
ACKNOWLEDGEMENTS
16
9.
FIGURES
17
REFERENCES
31
10.
343.3
-ii
ABSTRACT
This study forms part of Project No. 343 - Plastic Design "in A572
(Grade 65) Steel, sponsored by the American Institute of Steel Construction.
This portiqn of the project was aided also by a National Science Foundation
determi~e
carbon steel can be extended to the high strength steel, A572 in particular.
Thi's report is concerned with a stub column test, a set of residual stress
measurements, and theoretical predictions based upon them.
The maximum compressive residual stress measured in a l6WF7l section
was found to be 16.8 ksi, which is higher than that found in A7 or A36
steels~
343.3
-1
1.
INTRODUCTION
It is the
purpose of this report to present the residual stress pattern due to both
cooling and rotarizing, and to give the compressive properties of a
16WF7l shape of ASTM A572 (Grade 65) steel.
a~alysis
of column strength
to other tests.
The study reported in this paper is related to a larger program,
Project No. 343,
whos~
pl~stic
design criteria developed for lower grade steels can be applied also to
the higher strength material.
343.3
-2
2.
SHAPE CHARACTERISTICS
A l6WF7l section was used in both the stub column test and the
residual stress measurements.
mechanical'properties of
~he
Both the flange thickness and the web thickness are average values
determined from several sets of micrometer measurements.
The cross
Table 1:
Measured
Handbook
Flange width, b
Flange thickness,
8.54 in.
8.58 in.
795 in.
10.75
.486 in.
33.25
bit
Maximum value
for E =500ksi
st
.80 in
10.72
.50 in.
32.50
21.35 sq.in.
Web thickness, w
d/w
Cross sectional area, A 20.8.6 sq,. in.
10.8
30.6
It is seen that the measured dimensions of the shape ,agree with handbook values within about 2%.
ca leu"la tions
h
W1t
st
3
averages at 12.0 x 10- in/in,
h
F19. I . (1)
autograp h1C recor d
S 18 sown
1n
-3
343.3
The average strain-hardening modulus Est for the two f+ange specimens
is 679.
Test
Specimen
Location
Web
Web
4.l3.2W
Flange
4.l3.3F
Flange
4.l3.4F
Stub column
Weighted average
of flange and web
4.13.1W
Mill
Test
Yield
Point,
in ksi
72.9
72.9
Static
Yield
a
k~i
,in
62.6
61.0
62.2
63.7
63.5
62.7
StrainStrainhardening
Strain, Est' hardening
Modulus,
in
in/in x 10"':3 Est, in ksi
18.0
590
12.0
11.9
688
670
245
343.3
-4
3.
A stub column 34 inches long was tested following the standard stub
column test procedure. (2)
not to disturb the residual stresses, and the second to prevent the member
from buckling as
with a minimum
column.
le~gth
2
nEst
(1)
{L/r )2
y
equal to
With r
dial gages, placed at the corners of the specimen, were used for alignment.
Four SR-4 strain gages mounted on the outside flange tips at midheight, and
two 1/10,000 inch dial gages mounted on two box frames with a gage length
of ten inches were used for measuring the deformation.
.
(2)
accordance w1th the standard stub column test procedure.
343.3
-5
29.6 x 10
53.0 ksi
63.5 ksi
E,
Young's modulus
IT,
Proportional limit
a P , Static yield
ys
ksi
st~ain
of
The
proportional limit of 53.0 ksi and a static yield of 63.5 ksi results.in an
apparent compressive residual stress of 10.5 ksi, which does not correspond
very well with the 16.8 ksi from the measured residual stress.
proportion~l.limit
The
whe~
long and not the required 42.25 inches which is required not to disturb
the residual stresses, the reduced value of 10.5 ksi is possible.
However,
mov~ment
of the cross-heads.
The static yield level of the stub column was then determined as the
average value in that portion of the plastic range between load no. 42 and
when buckling was first observed (Load No. 48).
would be obtained if the full yield range had been used in the average.
The static yield of 63.5 ksi of the whole cross-sectional area agrees
fairly well with the tension test results which gave an average web value
343.3
-6
of 61.8 ksi, a flange value of 63.0 ksi and a weighted value of 62.7 ksi.
This is within about 1% of the stub column value, as would be expected,
considering that the tension specimens were taken from the same heat,
ingot, and, piece as the stub column.
In Fig. 5 and 6 (both taken at Load No. 38) the flanges have completely
yielded, and the web has begun to yield.
F~gs.
completely yielded', and the section has moved from an elastic-plastic state
to a totally plastic state.
to wrinkle slightly as sQown in Figs. 8 and 9 (Load No. 56), and continued
to do so until the test was stopped at Load No. 65 (Figs. 10 and, 11).
65 corresponds to an average strain of 55 x 10-
Load
in/in.
Fig. 4 shows that the stub column apparently began to strain harden
at a strain of 13.7 x 10-
The value of E
st
245 ksi, a value considerably less than the value obtained from the tension
specimens.
-7
343.3
4.
RESIDUAL STRESS
Residual stresses are stresses that are formed in the steel due to
def
'
(3)
p 1 ast1c
ormat1~n.
readings, made with a Whittemore strain gage, were made on both sides of
the web and flanges, following the procedure set forth in Ref. 3.
results of these measurements are shown in Fig. 12.
The
and of the measured residual stress pattern, it appears that the shape
had been rotarized.
The maximum compressive_residual stress for this shape was 18.8 ksi
on the outside of the lower flange (using the orientation of Fig. 12).
The inside stress at the same location was
inside of the flanges.
location.
decreases and finally enters the positive region, as would be expected from
the straightening process.
343.3
-8
The
upper and lower flange patterns are quite different and are very
unsymmetrical.
bent more than the top, and especially the right side of the bottom flange.
The similarities between the stress pattern of this rotarized shape
and the cold-bent A7 steel in Fig. 13 can be seen, with the cold bent
Even though the shape had been rotarized, the maximum and
The residual
maX1mum
reS1dUB 1 stress. (5)
343.3
-9
5.
COLUMN STRENGTH
Using this
a,pproximation and the curved symmetrical pattern show very good correlation.
Except for a small region at L/r about 30, Fig. 18 shows that the
higher strength material has relatively less sensitivity to residual stress
effects than is the case for A36 steel.
Neither calculation take~into account the lack of symmetry in the
residual stress distribution shown in Fig. 12.
str~ss-strain
curve measured in
~he
343.3
-10
6.
LOCAL BUCKLING
Local flange buckling was measured with a 1/10,000 inch dial gage
between the flange tips on one side of the stub column (see inset, Fig. 2).
These measurements were made every 3 inches the entire length of the specimen
except for the top 7 inches and the bottom 6 inches.,
symmetrical.
c~lumn VB.
section of the tangents drawn along the straight section of the curve. (6)
in/in strain, which is just before the strain when flange buckling
consistent with the degree to which the geometry of the shape met the
flange and web requirements (Table 2).
The critical bit ratio at which it is possible for a fully yielded
section to start to buckle is given by (7)
-11
343.3
fG t
1
1
cry
bIt
+ 0.381
(w)2
st
(2)
1+\>+
where
(3)
~
st
\>
.3
Young's Modulus
Est
'=
Web thickness
= Flange thickness
C!y
= Static yield
A
F
= Area of flange
Area of web
Using the values obtained for the A572 (Grade 65) material and shape,
the required values are bit
30.6.
the 16WF7l were very close to these requirements (see Table 2).
According to the theory if the actual bit value is less than this
critical value, local buckling will not
begi~
value in the l6WF7l shape, local buckling "should follow the theory.(i.e.
buckling should not occur prior to est).
has been observed that local
buc~ling
(8) st
in/in.
-12
343.3
show that the local buckling did not begin at the strain-hardening strain,
but began at approximately 7.5 x 10value.
At present there
approximately 1350 kips, the load had dropped only to 1100 kips (a loss of
-3
19%), at 55.0 x 10. in/in strain, a value equal to 20 times the yield strain
and 7 times the strain when local buckling was first observed.
This is
st
190.
At the onset
of strain-hardening the load had dropped to 61.6 ksi, due to local buckling,
but at a strain of 21.6 x 10-
After this the load began to drop at a slow rate, giving considerable
additional postbuckling deformation (Fig. 4).
343.3
-13
7.
SUMMARY
2.
(61.8 ksi) was 16% lower than that obtained in the mill report
(72.9 ksi) which is fairly consistent with prior observations.
3~
The compressive static yield level of the stub column was 63.5 ksi,
compared to the weighted average of 62.7 ksi from the tension tests
(slightly over 1% difference).
4.
~s~ls
The maximum
co~pressive
steel.
5.
6.
343.3
7.
-M
st
8.
in/in compared to
tion capacity.
It is the
Altho~gh
-15
343.3
The same thing may not be true when the shape is used as a beam.
Local buckling, in that case, could trigger lateral failure.
tests are being
p~rformed
to give
infor~tion
Other
-16
343.3
8.
ACKNOWLEDGEMENTS
This study was made during the summer of 1967 at the Fritz Engineering
Laboratory, Lehigh University, Bethlehem, Pennsylvania, with the aid of a
National Science Foundation Undergraduate Research Participation grant in
the Civil Engineering Department for the period.
~.
~p
the
Geoffrey Kroll,
Mr. David Miller and Mr. Suresh Desai for their aid in carrying out the
tests.
!
1
1
1
1
1
1
-17
1
1
1
343.3
1
1
1
1
1
1
1
1
70
~straln
1
1
Hardening (Est)
1
1
60
1
1
1
1
50
1
1
40
STRESS
(KSI)
30
20
10
3
10
12
14
16
18
20 x10-
STRAIN (in.llnJ
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
343.3
-18
FIG. 2
343.3
-19
70
60
fr
50
STRESS
(KSI)
40
Predicted
Flange
Buckling
30
20
10
:3
STRAIN
FIG. 3,
lOx 10-3
On. lin.)
70
Load 45 rLoad 56
60
50
STRESS 40
(KS!)
30
20
10
10
15
20
25
30
35
40
45
FIG. 4
50x 10-3
343.3
-20
FIG. 5
YIELDED FLANGE
FIG. 7
FIG. 6
YIELDED WEB
343.3
FIG. 8
FIG. 10
-21
FLANGE WRINKLING
(WEST SIDE)
FIG. 9
FIG. 11
FIANGE WRINKLING
(EAST SIDE)
-22
343.3
-12.0
-10
KSI
10
Mill Flake
(+)
KSI
10
Outside
x Inside
o
-16.8
I
2"
-10
4"
-23
343.3
-5
KSI
5
I
KSI
5
-5
343.3
-24
SHAPE
FLANGE
1========114VF43
12 W=50
FIG. 14
WEB
-25
343.3
-11.9 ksf
-10
a
I
KSI
-6 ksi
10
/'
/
I
22.1 ksi
l
1
I
I
I
\
\..
'-""-.
.............
~
I
KSI
I
a
FIG. 15
I
11
11
10
-10
343.3
-26
70
Strong Axis
----- ------,
\
\
60
\
\
\
50
40
STRESS
(KSI)
30
20
10
o
FIG. 16
20
40
60
80
343.3
-27
~~n
r~
-10.6 ksi
-IOksi
21.4 ksi
...............
...............
KSI
2 11
4"
10
-10
SCALE
FIG. 17
343.3
-28
70
Strong Axis
-------------,
\
\
\
60
50
40
Weak Axis
Approximate Column
Curve for Rolled Shape (A 36)
STRESS
(KSI)
30
20
10
o
FIG. 18
20
40
60
80
100
343.3
-29
Stub Column
Stress - Strain Curve
70
------...---,
60
\
\
50
40
STRESS
(KSI)
30
20
10
o
FIG. 19
20
40
60
80
100
343.3
-30
.30
.25
.20
.15
MAXIMUM
DEFLECTION
(INCH)
.10
I
/
I
_k~
I
.05
FIG. 20
Critical StrQin
4
6
STRAIN (in./in.)
343.3
-31
10.
1.
REFERENCES
Desai, S.
MECHANI~AL
Engin~ering
2.
Johnston, B. G.
COLUMN RESEARCH COUNCIL GUIDE TO DESIGN CRITERIA FOR METAL
COMPRESSION MEMBERS, Second- Edition, John Wiley & Sons,
Inc. New York 1966
3.
4.
Beedl~,
5.
Marquez, E. S.
RESIDUAL STRESS MEASUREMENTS ON 12B19 SHAPE (A572), Fritz
Engineering Laboratory Report No. 343~1 (In preparation)
6.-
7.
L. S. and Tall, L.
BASIC COLUMN STRENGTH, Proc. ASeE, 86(ST7), p. 139, July 1960
~CE,
Lay, M. G.
FLANGE LOCAL BUCKLING IN WIDE FLANGE SHAPES, Froe. ASeE,