Lehigh University

Lehigh Preserve
Fritz Laboratory Reports Civil and Environmental Engineering
1-1-1965
Residual stresses in welded built-up T-1 shapes,
October 1965, Reprint No. 320 (67-3)
E. Odar
F. Nishino
Follow this and additional works at: htp://preserve.lehigh.edu/engr-civil-environmental-fritz-lab-
reports
Tis Technical Report is brought to you for free and open access by the Civil and Environmental Engineering at Lehigh Preserve. It has been accepted
for inclusion in Fritz Laboratory Reports by an authorized administrator of Lehigh Preserve. For more information, please contact
preserve@lehigh.edu.
Recommended Citation
Odar, E. and Nishino, F., "Residual stresses in welded built-up T-1 shapes, October 1965, Reprint No. 320 (67-3)" (1965). Fritz
Laboratory Reports. Paper 177.
htp://preserve.lehigh.edu/engr-civil-environmental-fritz-lab-reports/177
by
Enver dar
Fumio ishino
Lambert: Tall
elded and RolledT-1 olumns
El
Fritz Engineering laboratory Report N00 29 08
Welded Built-Up and Rolled Heat-Treated T-I Golumns
RESIDUAL STRESSES IN WELDED T-l SHAPES
by
Enver Odar
Fumio Nishino
Lambert Tall
This work has been, carried out as part of an
investigation sponsored by the United States Steel
Corporation.
Fritz Laboratory
Department of Civil Engineering.
Lehigh University
Bethlehem, Pennsylvania
October 1965
Fritz Engineering Laboratory Report No. 290.8
2.
3.
5.
6.
8.
9.
TABLE OF CONTENTS
ABSTRACT
INTRODUCTION
DESCRIPTION OF TESTS
Residual Stress Measurements
Mechanical Property Tests
Stub Column Tests
TEST RESULTS AND DISCUSSIONS
Residual Stress Measurements
Mechanical Property Tests
Stub Column Tests
CO:MPARISON WITH PLATE RESULTS AND RESULTS OF
OTHER STEELS
SuMMARY AND CONCLUSIONS
ACKNOWLEDGEMENTS
DEFINITION OF TERMS
TABLES AND FIGURES
REFERENCES
1
2
2
2
3
4
4
6
6
7
10
11
12
14
35
290.8
ABSTRACT
)
\
This report presents the obtained from an experimental
investigation of residual stresses in welded built-up columns made of
liT-I" steel. fiT-I" steel meets the requirements of ASTM Designation
AS140
The column shapes were made by welding T-l steel plates to-
gether. Three welded built-up H shapes and Box shapes were
included in the investigation.
The investigation included residual stress measurements, ten-
si1e coupon tests and stub column tests. Comparison was made of the
residual stresses in shapes and the residual stresses in their compon-
ent plates, and the residual stresses were shown to be similar. The
knowledge of residual stress in the cross section makes it possible to
predict the strength of a member made of steel.
1. INTRODUCTION
"Residual Stresses in Welded Built-Up Shapes Made of T-l
Steel" is one phase of the research project on "Welded Built-Up and
"k
Rolled Heat Treated T-l Columns". An extensive study was conducted
on Residual Stresses in T-1 Steel Plates, and the results are presented
in Reference 1. This report gives results obtained from measuring
residual stresses in shapes made from welding T-l steel plates, and
the results are compared with those obtained from measuring welded
plates, and also with results from other steels.
Studies on the carrying capacity of columns show that the
strength of axially loaded steel columns is greatly affected by the
(2' 3)
residual stresses within the cross section > This effect is more
pronounced for welded sections (4) The theoretical prediction of
strength reduction in sections containing residual stresses is avqil-
able in m?ny sources such as in References 2, 3, 5, and 6.
"T-I" steel is a high strength structural alloy steel produced by the
United States Steel Corporation; it meets the requirements of ASTM
A514 and /o.r A5l7 steel.
-1-
290.8
2. DESCRIPTION OF TESTS
Five shapes were included in the study. Three of the shapes
were welded H shapes, and two shapes were welded Box shapes. They
were made of flame-cut T-l plates, except for one 6 x 7 H shape where
sheared edge plates were used. Table 1 shows th,e shapes tested and
their dimensions. The welding was by the automatic submerged arc
welding procedure utilizing Lincoln L70 wire and L840 flux.
Residual Stress Measurements
The residual stresses were obtained by the method of section-
ing(3). The holes were laid off on the sections in more than one place
along the length to results. Figures I and 2 show the typi-
cal layouts for marking and sectioning for welded H and Box shapes
respectively. The hole dimensions were spaced in such a way as to
give more readings at the regions where a transition from tensile to
compressive residual stress was expected. The readings were taken on
both faces of the H shapes and on one face of the Box shape, since the
plates were sufficiently to, have uniform residual
stresses throughout the thickness(7). Figure 3 shows the residual
stress distribution in welded H and Box shapes and in their component
plates.
Mechanical Property Tests
A number of tensile coupons to ascertain the
mechanical properties. The tension coupons were standard ASTM(8)_flat
-2-
290.8 -3-
specimens. Table 2 shows the coupons included in the study and Fig. 4
shows the locations from which the coupons were taken.
Stub Column Tests
Five stub columns (2) , one f ~ o m each shape, were t ~ s t e d to
obtain the average load-strain relationship for the section as a
whole, and also to verify the measured residual stresses. The testing
set-up, procedure and the required lengths were as specified in Reference
9. Figure 5 shows a stub column ready for testing.
290.8
3. TEST RESULTS AND DISCUSSION
The patterns of residual stress distribution obtained from
measurements on H and Box shapes are present'ed in Figs. 6 through 10,.
The results for welded H-shapes are the readings of both top and bottom
faces, whereas the readings shown for Box shapes are the outside face
readings only. Two sets of measurements taken from different sections
are presented for each shape; these show that there is little varia-
tion in residual stress along the length of the shape. Table 3 gives
a summary of results. The results are average valuee of top and
bottom readings for H shapes and one face reading for Box shapes.
The resulting curves and the notation used in the Table are shown in
Fig. 3.
Figures 6 through 8 show the resiqual stresses for welded H
shapes. The pattern of residual stress distribution had tensile
residual stresses at the junction of flange and web and at the flange
tips, and compressi ve stresses over the rest of the shape, except for
the shape made of plates with sheared edges, Fig. 6. The 6" x 7
11
H-shape made of plates with sheared edges and the shape made of plates
with flame-cut edges showed a similar stress pattern except for the
flange tips where the residual stress pattern of one face is totally
different from the other. This is a characteristic pattern resulting'
from the shearing of the edges (1) There were tensile residual
stresses at the junction of flange and web and at the web ends; this
-4-
290.8 -5-
was up to 80 ksi at the web ends in magnitude with an average of 25 ksi
for the flanges. Compressive residual stresses covered the rest of the
&hape with a rather constant distribution and values not exceeding
25 ksi, the average being 20 ksi. In shapes made of flame-cut plates,
Fig. 7, the tensile residual stress at the flange tips less than
20 ksi, and reached 80 ksi at the web ends on the weld metal. The
bottom face reading was not taken at the junction of web and flange,
but the closest measured points suggest that it would have been over
80 ksi on the weld. The compressive residual stresses were distributed
rather uniformly over most of the flange and web, with an average of
20 to 22 ksi for both flange and web. The measured results for the
9" welded H shape are shown in Fig. 8. The tensile stresses
were about 10-12 ksi at the flange tips and less than 40 ksi at the
web tips. The compressive residual stress was spread over most of
the flange width and web. The distribution is fairly constant with a
value of 16 to 19 ksi in the flanges and 6 to 8 ksi in the web. In all
cases the difference between two sets of readings is insignificant,
suggesting that the variation of residual stress along the length of
the shape is rather small.
The results of measurements on welded Box shapes are shown
in Figs. 9 and 10. The 6 x 6 Box shape had tensile residual stresses
at the welded corners as high as 80 ksi with an average of 65 ksi.
Compressive residual stresses existed over most of the area, varying
from 24 to 28 ksi, with a rather uniform distribution. In the larger
shape, the 10 x 10 Box, the pattern was similar with tensile
290.8 -6-
stresses of 80 ksi at the four corners and with a fairly constant com-
pressive residual stress distribution throughout the of the section.
The average compressive residual stress was about 10 - 16 ksi. In all
Box shapes, the component plates had compressive residual stress
uniformly distributed over 80% of the total width.
Mechanical Property Tests
The results obtained from some of the tension coupon test p are
given in Table 4 and a typical load strain relationship is shown in
Fig. 11. The static yield stress* was around 110 ksi for the coupons
tested.
Stub Column Tests
A typical load-strain relationship obtained from testing a
6 x 7 made of plates with flame-cut edges is shown in Fig. 12.
The effect of the residual stresses pre&ent in the cross section on the
load strain relationship is seen from the existence of a rounded knee
in this figure(2). The proportional limit was reduced below that
obtained from the tension coupon test, (Fig. 11,) and the yielding
started at 74% of the total yield load of the cross section.
This reduction confirmed the magnitude of the measured residual stresses,
which were relatively high, 20 - 22 ksi) for this particular shape.
1<
The "static yield stress" is defined in Section 7, Definitions.
290.8
4. COMPARISON WITH PLATE RESULTS AND RESULTS OF OTHER STEELS
Welded shapes are formed from a number of plates, welded to-
gether; the plates may be center-welded, both edges welded or single
plates. An extensive study(l) was conducted on T-l steel
plates welded and unwelded, and it is of interest to determine whether
an estimate of residual stress existing in welded shapes can be made
from a knowledge of the residual stress in its component plates. The
factors affecting the possible difference between residual stress in
plates and shapes may be as follows:
1. The general rise in temperature after welding may be differ-
ent to additional plates for heat flow in shapes and,
2. Each plate in a shape has some restraining effect on the
other. It may be expected that the effect depends greatly on the rela-
tive sizes of the plate elements; namely the effect of restraining will
be much more pronounced if a small plate is welded to a large plate,
d 1 d
-f h - f 1 we1ded - -1 (10,11):
an ess t e 0 pates are ar .k
In this study the plates were of almost the same size and it
may be assumed that the effect of restraint is very small. Figure 13
shows the residual stress distribution measured in a 6" x 7
11
H-shape
with plates having flame.cut edges, and the residual stress distribu-
tioD measured from a number of center-welded and edge-welded plates of
similar geometry. The results are similar; the compreSSive
residual stresses in the web and flanges of a built-up section can be

290.8 -8-
represented by the distribution from measuring its component
plates. The tensile stress on the top face reading at the junction of
flange and web is less for the shape probably due to the fact that weld
metal did not penetrate far into the flange plate, wher,eas the separate
plate was grooved and the beam laid in the groove. However, the main
purpose of this study was to accurately determine the compressive resid4al
stresses, rather than tensile residual stresses, since it is the com-
pressive residu'al stress which affects the strength of compression members.
The comparison of residual stresses in a Box shape and its component
plate is shown in Fig. 14. The patterns shown are the top face readings
of the 6" x 6" Box and the top face reading of a 6" x plate welded on
both edges. Again, the similarity exists. Thus, it is clear that com-
ponent plates have a residual stress pattern and magnitude similar to
that in the welded built-up shape made of these plates.
The results of a study on the residual stresses in welded A7
steel shapes, and applicable to A36 steel shapes, are presented in
Ref. 10. It is of interest to compare these results with those of this
study since the are of identical geometry. There is no possibility
of direct comparison for H-shapes) since the A7 steel shapes were made
from universal mill plates; however, for Box shapes the effect of flame-
cutting is almost eliminated by subsequent welding of the edges. Figure
15 shows the residual stresses measured in 6" x 6" box shapes of A.7 and
T-l steels. The general pattern of the residual stress is similar for
both steels. However, 'the magnitude of compressive residual stress is
slightly less for A7 steel.
290.8 -9-
Although for T-l steel the value of the compressive residual
stress is slightly higher and is over a wider portion of the section,
nevertheless, it is a fraction of the yield strength. The
effect of residual stress on the capacity of welded
T-I steel columns will be smaller than the on the shapes made of
A7 and A36
Finally, in comparing the re:;;idual stresses in welded 6 x 7
H-shapes with residual stresses in a rolled heat-treated shape,
both of T-l steel, the magnitude of stress in the rolled
heat-treated shape is much lower probably as a result of the ,tempering
Figure 16 shows this comparison of residual stresses in
welded built-up and rolled heat-treated ahapea(12).
290.8
5. SUMMARY AND CONCLUSIONS
A summary of the results obtained from the experimental in-
vestigation into the residual stresses in weLded shapes is
presented here.
The study, together with previous results obtained for welded
of T-l steel, a basic knowledge needed for estimating
the strength of medium-size welded built-up columns of T-l steel.
Three H and two Box ahapes were included in the study. The
shapes were built.up by welding T-,l steel plates together. The com-
ponent plates had edges either flame-cut or sheared and were welded
automatically by the submerged-arc welding process utilizing Lincoln
L70 wire and L840 flux.
The following conclusions can be drawn from this study.
1. Residual stresses in component plates give a close re-
presentation of the residual stresses in the shapes themselves. (Sec.
3, Figs 13, 14.)
2. Welded built-up T-I steel shapes have residual stresses
considerably higher than rolled heat-treated T-l steel shapes. (Sec.
3, Fig. 16.)
3. Residual stresses in welded steel sections are
slightly higher in magnitude than those in welded built-up shapes
A7 steel. Since they constitute a smaller fraction of the yield strength
for T-l steel, it is expected that the strength of T-l steel compression
should be affected less by their presence. (Sec. 3, Fig. 15.)
-10-
290.8
6. ACKNOWLEDGEMENTS
This study has been carried out as, a part of the research
project on "Welded Built-Up and Rolled Heat-Treated T-I Steel Columns"
being at Engineering Laboratory, Department of Civil
Engineering, Lehigh University in Bethlehem, Pennsylvania.
The project ls sponsored by the United States Steel'Corporation.
Charles G. Schilling of that corporation made many valuable Gomments,
both in the course of the program, and in the preparation of this re-
port. Task Group 1 of the Column Research Council of' the Engineering
Foundation acted in an advisory capacity under the chairmanship 6f
John A. Gilligan. T.he assistance of Ching-KuQ Yu during the study is
acknowledged.
-11-
290.8
7. DEFINITION OF TERMS
Residual Stresses
Residual stresses are the stre'sses formed' as a result of
permanent deformations of certain fibers during the fp.b.rication or
cooling processes.
Yield Point
The yield point is the first stress in a material less than
the maximum attainable stress, at which an increase in strain occurs
without an increase in stress(8).
Yield Strength
The yield strength is the stress at which a material exhibits
a specified limiting deviation from the proportionality of stress to
strain. The deviation is expressed in terms of strain(8).
Yield Stress Level
The yield stress level is the average stress during actual
yielding in the plastic range. It remains fairly constant for structural
steel, provided the strain rate remains constant.
Static Yield Stress Level
The yield stress level is the average stress during
actual yielding in the plastic with zero strain rate.
Stub Column
A stub column is a short compression test specimen, sufficient-
ly long for use in measuring the stress-strain relationship for the
-12-
'."i.'
-13-
complete cross section, but short enough to avoid buckling as a column
in the elastic and plastic ranges.
290.8
8 TABLES AND FIGURES
-14.,.
290.8
NO.
TABLE 1
SHAPE
SHAPES TESTED
EDGE
PREPARATION
PLATE SIZE (IN)
-15-
{ 6 x
1/2 flange
W-A 6 x 7 - H Sheared
6 x 3/8 web
r
6 x 1/2 flangE?
W-B 6 x 7 - H Flame-Cut
l
6 x 3/8 web
9 x'3/4 flange
W.. C
9 x l O ~ - H Flame-Cut
{
9 x 1/2 web
6 x 1/4
W-D 6 x 6 Box Flame-Cut
{
5 ~ x 1/4
10 x 1/2
W-E 10 x 10 Box Flame-Cut
(
9 x 1/2
TABLE 2 TENSION COUPONS TESTED
-16-
SHAPE COUPON NO. PLACE TAKEN
W-A-l Flange
6 x 7 H
W-A-2 Flange
(Sheared)
W-A-3 Web
W-B-l Flange
6 x 7 H
W-B-2 Flange
(Flame Cut)
W-B-3 Web
W-C-l Flange
9 x l O ~ H W-C-2 Web
W - C ~ 3 Flange
W-D-l
*
W-D-2
6 'x 6 Box
W-D-3
W-D-4
W-E-l
'1<
W-E-2
10 x 10 Box
W-E-3
W-E-4
*
One tension coupon from each side was taken for box sections.
Readings are average of top and bott?m faces for H shapes.
I
I--'-
..........
I'
290.8
TABLE 4 RESULTS OF TENSION COUPON TESTS
-18-
Cut
I \
1_ 19" ~ I 11" 1_
24" -I
-19-
gil X 10 y ~ 1 Welded
H Shape
FIG. 1
gil
TYPICAL LAYOUT FOR SECTIONING OF WELDED H-SHAPES
Cut
I \
---+
-20-
@
I ~ II II
4 12 =2
6x6
Welded Box
FIG. 2 TYPICAL LAYOUT FOR SECTIONING OF WELDED BOX SHAPES
Compression
+ Tension
0'
4
- 2 1 ~
FIG. 3
,
r
PATTERNS AND NOTATIONS FOR RESIDUAL STRESSES
IN WELDED BUILT-UP T-l SHAPES
1M
I
FIG. 4
-22-
I
Two coupons from the
flanges
One cou pon from the
web
H- Shape
One coupon from
each component plate
Box Shape
LOCATION OF TENSION COUPONS
-23-
...
..
. , .
FIG. 5 STUB COLUMN TEST SET-UP
-24- .
40
ksi 0
,
~
,
-40
\
b
\
----o-----<r--
h
Top Surface
2 in.
---l::r--6---
Bottom Surface
ksi
6
I
I
40 0 -40
Ii
I
I
d
"
FIG. 6 RESIDUAL STRESSES IN 6" x 7" H-SHAPE (SHEARED EDGES)
2 in
ksi
I I
o -40
FIG. 7 RESIDUAL STRESSES IN 6" x 7" H-SHAPE (FLA'ME-CUT)
FIG. 8
2in
ksi
I I I
40 0 -40
RESIDUAL STRESSES IN 9" x 1 0 ~ " H-SHAPE (FLAME-CDT)
-26-
Dimension
Stress
Scale
Scale
I
2in
40 80
ksi
-27-
FIG. 9 RESIDUAL STRESSES IN 6" x 6" BOX SHAPE (FLAME-CUT)
Dimension Stress
Scale Scale
I I I
2in 0 40 80
ksi
-28-
FIG. 10
RESIDUAL STRESSES IN 10" x 10" BOX SHAPE
LOAD (kips)
80
60
40
20
o 0.005
Area =0.776 in
2
oy =106.2 ksi
~ =119.3 ksi
0.010 0.015
STRAI N (micro in/in)
0.020
FIG. 11 LOAD-STRAIN RELATIONSHIP OF A TYPICAL TENSION COUPON
I
N
\.0
I
LOAD (kips)
1000
o 0 ~ 0 0 5 0.010
STRAIN (micro in/in)
800
600
400
200
--L
/
/
/
/
A
6
11
X 7" H - Shape
0.015
FIG. 12 LOAD-STRAIN RELATIONSHIP FROM A STUB COLUMN TEST I
lJ,.)
o
I
6 x7H- Shape
(Flame Cut)
t r - ~
I \
I \
I \
J,-yoQ \
IT''''"' \
I
\ l)-"\J I
\ I
\ I
'rr.li
A
ksi
I I I
40 0 -40
1 I
2in
I ..
6 X Y2 Plate
( ~ ~ I A Weld)
r . . f r ~
lro-\\
I
40
ksi J... 0
-40
FIG. 13 CO:MPARISON OF WELDED H-SHAPE WITH ITS CO:MPONENT PLATES
I
lJ,j
f--A
I
6
11
x 6" Box Shape
ksi
I I J
40 0 -40
I I
2in
6
11
X yd
l
Plate
Fig. 14 CO:MPARISON OF WEIDED BOX SHAPE WITH ITS COlYIPONENT PLATES
I
W
N
I
FIG. 15
-33-
Residual Stress in 6
11
)( 6" Box
A-7 Steel
Stress
Scole
, I I
o 20 40
ksi
Dimension Scale
2in
Residual Stress in SIiX 6
11
Box
T-I Steel
COMPARISON OF RESIDUAL STRESS IN A-7 AND T-l BOX SHAPES
aW:-31
Stress
Scale
I I I
o 40 80
ksi
6x7 H
Flame Cut
FIG. 16 COMPARISON OF RESIDUAL STRESS IN WELDED BUILT-UP SHAPE
AND ROLLED HEAT-TREATED SECTION
I
W
+'
I
290.8
REFERENCES
Ie Odar, E., Nishino, F., and Tall, L.
RESIDUAL STRESSES IN T-I PLATES, Fritz Laboratory Report
290.4, Lehigh University, (To be published in
Welding Journal).
2. Beedle, L. S., and Tall, L.
BASIC COLUMN STRENGTH, ASeE Prac. Paper 2555, Vol. 86,
ST7, July 1960.
3. Huber, A. W., and Beedle, L. S.
RESIDUAL STRESS AND THE COMPRESSIVE STRENGTH OF STEELS,
The Welding Journal, 33 (12), p. 589-8, December 1954.
4. Estuar, F., and Tall, L.
EXPERIMENTAL INVESTIGATION OF WELDED BUILT-UP COLUMNS,
The Welding Journal, p . 1963.
5. Yang, C. H., Beedle, L. S., and Johnston, B. G.
RESIDUAL STRESS AND THE YIELD STRENGTH OF STEEL BEAMS,
The Welding Journal, Research Supplement, Vol. 31,
April 1952.
6. Osgood, W. R.
THE EFFECT OF RESIDUAL STRESSES ON COLUMN STRENGTH,
Prac. First National Congress Applied Mechanics, 1951.
7. NagarajaRao, N. R., and Tall, L.
RESIDUAL STRESSES IN WELDED PLATES, The Welding Journal,
Vol. 40, October 1961.
ASTM Specifications No. A370-557.
9. Tall, L.
STUB COLUMN PROCEDURE, Fritz Laboratory Report,
220 A-36, February 1961. Also, IrW Document
Oslo, July 1962.
10. NagarajaRao, N. R., Estuar, and Tall, L.
RESIDUAL STRESSES IN WELDED SHAPES, The Welding Journal,
Vol. 43, July 1964.
11. Tall, L.
RESIDUAL STRESSES IN WELDED PLATES - A THEORETICAL STUDY,
The Welding Journal, Vol. 43, January. 1964.

290.8 -36-
12. Odar, E., Nishino, F., and Tall, L.
RESIDUAL STRESSES IN ROLLED HEAT-TREATED T-l SHAPES,
Fritz Laboratory Report 29005, Lehigh University,
1965. (To be published in Welding Journal.)