...

Campbell-TekkenTest-WJ_1976_05_s135.pdf

by user

on
Category: Documents
1

views

Report

Comments

Transcript

Campbell-TekkenTest-WJ_1976_05_s135.pdf
Experiences with HAZ Cold Cracking
Tests on a C-Mn Structural Steel
Significant details of the controlled thermal severity
(CTS) test are compared with those of the Tekken Ygroove test
BY W. P. CAMPBELL
ABSTRACT. The heat-affected zone
cold cracking tendency of a C-Mn
structural steel was evaluated using
controlled thermal severity (CTS) and
Tekken Y-groove specimens. This
permitted a comparison of the performance of the single-pass fillet weld
test and the single-pass butt weld test
and the development of improved u n derstanding of the performance of the
latter.
No cracking occurred in the CTS
tests but severe weld or heat-affected zone cracking was encountered in the Y-groove tests under c o n ditions which had been expected to
be less likely to produce cracking
than the conditions employed in the
CTS tests.
With the thickest plate tested, i.e.,
1-1/2 in. (38 mm), severe weld metal
cracking occurred when the restraint
level in the Y-groove test was at a
m a x i m u m , r e g a r d l e s s of the root
opening in the range of 0.04-0.08 in.
(1-2 mm). When the restraint level
was reduced by relief slots, weld
metal cracking was replaced by severe heat-affected zone cracking.
With 1 in. (25 mm) plate, only weld
metal cracking occurred when the
root opening was 0.02 in. (0.5 mm)
and only heat-affected zone cracking
occurred when the root opening was
0.08 in. (2 mm). Thus certain optimum
conditions were indicated in order to
u t i l i z e the test for s t u d y i n g t h e
W. P. CAMPBELL is Research Scientist,
Welding Section, in the Physical Metallurgy Research Laboratories of the Department of Energy. Mines and Resources.
Ottawa, Ontario
K1A-OG1
Canada.
p r o p e n s i t y to heat-affected zone
cracking.
It was shown that special procedures are necessary in order to control the critical root opening in the
Y-groove specimen. One method is to
use e l e c t r o - d i s c h a r g e
machining
(EDM) to make the root opening in a
one piece specimen. However,this has
the disadvantage that metallurgical
effects of the EDM process upon the
cut surface of the o p e n i n g may
possibly influence crack initiation. A
better method is to employ a two
piece specimen which has wide root
faces in the restraint portions. When
these faces are in contact, the correct root opening is maintained in the
test weld portion after welding the
restraint portions.
Introduction
The tendency of steels to heat-affected zone cold cracking may be assessed by the performance of the
steels in single-pass, fillet welded
specimens as employed in the C o n trolled Thermal Severity (CTS) test.
This has been used extensively in
many countries. The cracking t e n dency may also be evaluated in
single-pass butt welded specimens as
employed in the Y-groove Tekken
test. This has been used primarily in
Japan, where it was developed, but its
use appears to be increasing in other
countries.
In order to gain experience with
and to evaluate the Y-groove test, it
was decided to undertake comparative studies of the cracking tendency
of single-pass welds in a commonly
used structural steel with both CTS
WELDING
and the Y-groove specimens. It was
hoped that this study would contribute to more accurate and realistic
evaluation of w e l d a b i l i t y behavior
based upon small scale welding tests.
E x p e r i m e n t a l Details
Materials
All steel used in the study was from
the same heat and rolled to a thickness of either 1 in. (25 mm) or 1-1/2
in. (38 mm). The plate compositions
were determined to be as given in
Table 1.
Electrodes
All welding was performed using
5/32 in. (4 mm) E7018 classification
electrodes from one manufacturer
and lot. The moisture content of the
covering was monitored at intervals
during the program using the proced u r e specified in CSA S t a n d a r d
W48.1-1969. All test welds were deposited with e l e c t r o d e s having a
c o v e r i n g m o i s t u r e c o n t e n t in the
range 0.4 to 0.6%.
Test Procedure and Results
CTS Tests on 1 in. (25 mm) Plate
Single-pass fillet weld tests C1-C4
inclusive were made using the CTS
specimen shown in Fig. 1. Both top
and b o t t o m m e m b e r s were m a chined from the 1 in. (25 mm) plate.
The anchor welds were deposited first
at a higher level of energy input than
that of 30-36 k J / i n . (1.2-1.4 k J / m m )
which was used for depositing the two
test welds. Each test weld was dep o s i t e d w i t h the s p e c i m e n at a
RESEARCH SUPPLEMENT!
135-8
Table 1 — Composition of Steel Tested, Wt. %
1 in. (25 m m ) plate
11/2 in. (38 mm) plate
C
Mn
Si
S
P
Total
Al
N
0.22
0.19
1.32
1.27
0.20
0.20
0.021
0.022
0.006
0.006
0.025
0.016
0.005
0.005
CE(a,
0.44
0.40
(a) Carbon-equivalent = %C + % Mn/6
temperature of about 70 F (21 C), and
the second weld was made 2 h after
the first. Tests CC1 and CC2 were
similar except that the specimens
were 8 in. (203 mm) rather than 4 in.
(102 mm) wide, i.e., the test welds
were 8 in. (203 m m ) l o n g . Each
specimen had two anchor bolts. Tests
C5 and C6 were similar to C1-C4 inclusive, except that the specimens
were stiffened by the addition of a
1 X 4 X 1 2 in. (25X102X305 mm) plate,
positioned at right angles to the bottom plate at a central longitudinal
location, and d o u b l e - f i l l e t w e l d e d
along the c o m m o n 12 in. (305 mm)
dimension to f o r m a T-shaped bottom assembly. These modifications to
the "standard" CTS specimen were
made in attempting to increase the
cracking severity of the CTS test.
At least 48 h from completion of
welding, each test weld was sectioned transversely, using a liquid
cooled abrasive cut-off wheel, at three
equally spaced locations along the
length. All six surfaces thus exposed
in each test weld were inspected for
cracking under ultraviolet lighting
after magnetization while flowing a
liquid containing fluorescent magnetic powder over the surfaces. No
cracking was found in any of the test
welds.
Y-Groove Butt Weld Tests on
1 1/2 in. (38 mm) Plate
Single-pass butt weld tests (Table
2) were made in the 1-1/2 in. (38 mm)
plate using Y-groove specimens intended to conform to either Fig. 2, 3 or
5. The root opening* in specimens to
Fig. 3 were made using electro-discharge machining (EDM).
While welding the restraint welds in
the two piece specimen (Fig. 2), a
steel shim plate, 0.08 in. (2 m m ) thick,
was inserted in the test weld portion
of the specimen and was removed before depositing the test weld.
Somewhat later in the study, an
Australian paper (Ref. 1) was noted in
which attention was drawn to the c r i tical nature of the width of the root
opening. According to this study, the
root opening should be held in the
range 0.079 ±0.004 in. (2.0 ± 0 . 1 mm).
It was reported that m a x i m u m crack"Referred to as gap width in Figs. 2, 3 and
4.
136-8 I M A Y
1976
Cm
TEST WELD-TRITHERMALHEAT FLOW
BOTTOM PLATE
r|in. DRILL FOR £in. DIA BOLT
(14mm)
(13mm)
h*.i±*
4in.STEEL SHIM PLATE
(2 x 89x 102 mm)
CONTROLLED THERMAL SEVERITY (CTS) SPECIMEN
Fig. 1 — Controlled thermal severity (CTS) weldability specimens, used for 1 in. f25 mm)
plate
ing occurred when the opening was
0.08 in. (2 mm) wide and that cracking was reduced almost to zero when
the opening was either reduced or increased by approximately 0.02 in. (0.6
mm). It was further noted in this paper
that a Japanese Industrial Standard
on the Tekken test had specified the
opening to be 0.079 ± 0.008 in. (2 ±
0.2 mm). However, in the Japanese
publications which had been examined prior to commencing our project,
there had been no reference to root
opening
tolerances
nor
to
the
Japanese standard.
In view of the apparent importance
of the root opening dimension, the
transverse sections from the
specimens which were intended to
conform to either of Fig. 2 or 3 were
subsequently re-examined, not for
cracking as in the original examination, but to determine the root o p e n ing. The measurements for the root
opening, representative of all tests ex-
cept Y15 and Y16, are given in Table
2.
Measurements for Y15 and Y16
could not be obtained because these
specimens had been prepared for examination at least 48 h after completion of welding, first by heating them
in an oven for 1 h at 750 F (399 C) to
oxidize any cracks and then by breaking the welds open, after cutting off
the restraint p o r t i o n s , to p e r m i t
observation of the extent of oxidized,
cracked areas.
In all of the other tests, after at least
48 h from completion of welding, the
specimens were cut transversely at
intervals of approximately 1/2 in. (13
mm), using a liquid cooled abrasive
cut-off wheel, and all ten surfaces
from each weld were examined by the
magnetic particle process as for the
CTS specimens. As shown by Table 2,
the root openings had not been maintained within the r e c o m m e n d e d range
when using the two piece specimens,
intended to conform to Fig. 2. Typically, the root openings were of the order
of 0.04 in. (1 mm). In contrast, the root
o p e n i n g s for the one piece
specimens, intended to conform to
Fig. 3, were within the recommended
range.
Near the completion of the study,
tests YA and YB were made using a
two piece modified specimen. In Fig.
2, the joint for the restraint welds has
a chisel edge, i.e., no root face. In
specimens YA and YB, the joints for
the restraint welds were machined to
have a root face equal to 1/3 of the
plate thickness. Additionally, the joint
for the test weld portion was modified
so that when the specimens were
clamped together with the root faces
of the restraint weld portions in contact, the 0.08 in. (2 mm) opening was
automatically provided in the test portion. Similar joint preparations were
shown in a Japanese publication (Ref.
3) but no explanation was given for
their use.
Experiments by the author with this
modified specimen indicated that,
even with the relatively heavy root
faces in the restraint weld portions,
the root opening at the test weld portion was reduced by about 0.004 in.
(0.1 mm) after the anchor welds had
cooled to room temperature. Preheating the specimen to about 300 F (149
C) caused expansion of the root
opening by about 0.004 in. (0.1 mm).
Consequently the workshop drawings were modified so that the opening, when the specimens were
clamped together in preparation for
making the anchor welds, was 0.080
+ 0.003-0.001 in. (2+0.08-0.03 mm).
Thus the opening just before making
the test weld should always be in the
range 0.079 ±0.004 (2.0 ±0.1 mm)
with or without preheating up to at
least 3*00 F (249 C). This was the tolerance recommended by the Australian researchers (Ref. 1). The
modified workshop drawing is shown
in Fig. 5.
In tests Y5, Y6, Y15, Y16, Y24 and
Y25 (Table 2), a series of parallel slots
5.9in
5.9 in.
(I50mm)
(I50mm)
—
t— SAW CUT DEPTHS 1
I..
A
-4
c\;
,
'
o
UJ
2
—
z>
E 6
t-
" ll
i
I
in
3h
i_JdaL_
ENLARGED SECTION B - 8
* DEPTH OF RESTRAINT RELIEVING SAW CUTS = 0
FOR FULL RESTRAINT, OR SPECIFIED SEPARATELY
WHEN REDUCED RESTRAINT IS EMPLOYED
Fig. 2 — Two piece Y-groove weldability
specimen used for 1-1/2 in. (38 mm) plate
Fig. 2
Fig. 2
Full
Full
70(21)
70(21)
20-60
20-70
20-80
(Fig. 6)
None
30-50
Y4
Fig. 3
Full
70(21)
0.08 (2)
Y15
Fig. 3
0.79-in.
(20-mm)
slots
70(21)
Not measured < c >
but more probably
0.08 (2)
60-100
(Fig. 8)
50-100
Y16
Fig. 3
Y5
Fig. 3
Y6
Fig. 3
Y24
Y25
YA
YB
Fig.
Fig.
Fig.
Fig.
3
3
5
5
70(21)
1.6-in.
(40-mm)
slots
"
"
Full
Full
i
1
l
i ;!
(a)
0.04(1)
0.04(1)
0.04-0.07
(1-1.8)
0.04(1)
0.06 (1.5)
Y7
Y8
i
Fig. 3 — One piece Y-groove weldability
specimen used for 1-1/2 in. (38 mm) plate
70(21)
70(21)
70(21)
Full
.^v--Js_.
j i ;
CD
* DEPTH OF RESTRAINT RELIEVING SAW CUTS = O
FOR FULL RESTRAINT, OR SPECIFIED SEPARATELY
WHEN REDUCED RESTRAINT IS EMPLOYED
Full
Full
Fig. 2
Fig. 2
Fig. 2
—
..GAP WIDTH
. 0 8 in. (2 mm)
ENLARGED SECTION A - A
ENLARGED SECTION A -A
Root crack
length — % of
weld depth
Y1
Y2
Y3
Tiff
-RESTRAINT WELD
Root'"'
opening
in. (mm)
Restraint
condition
A
A
Initial
Temp.
F(C)
Test
specimen
T-
UJ
in
-4
Table 2 — Y-Groove Tests on 1-1/2 in. (38 mm) Plate
Test
no.
Q UJ
•]s»
p
B
H
p—SAV CUT DEPTHS
25-40
70(21)
0.08 (2)
None
70(21)
0.08(2)
None
200 (93)
200(93)
300(149)
300(149)
0.08(2)
0.08 (2)
0.08(2)
0.08 (2)
None
None
None
None
Other type
cracks — % of
weld depth
None
None
Toe cracks — see Fig. 6
HAZ cracks — see Fig. 9
HAZ cracks extending
30% from root and
30 to 8 0 % f r o m toe
None
HAZ cracks extending
50-100% from root
HAZ cracks extending
25-40% from root
HAZ cracks extending
20-90% from root —
see Fig. 7
HAZ cracks extending
20-50% from root
None
None
None
None
(a) Energy i n p u t w a s held c o n s t a n t at 4 0 - 4 3 k J / i n . (1.6-1.7 k J / m m )
(b) M e a s u r e m e n t s m a d e o n t r a n s v e r s e s e c t i o n s
(c) C r a c k s w e r e o x i d i z e d by h e a t i n g s p e c i m e n s in a f u r n a c e a n d w e l d s w e r e b r o k e n a p a r t f o r visual e x a m i n a t i o n .
W E L D I N G R E S E A R C H S U P P L E M E N T ! 137-8
either 0.79 in. (20 mm) or 1.6 in. (40
mm) in length were cut from the 8 in.
(203 mm) edges of the specimens, in
the manner illustrated in Fig. 2 or 3.
These slots provided a reduction in
t h e r e s t r a i n t as c o m p a r e d t o
s p e c i m e n s c o n t a i n i n g no s l o t s .
Reduction of restraint was greatest
with the specimens having a slot
depth of 1.6 in. (40 mm).
Preheating, used in tests Y24, Y25,
YA and YB, was achieved by heating
the specimens for at least one hour
per inch of plate thickness in an oven
set at the desired temperature. Test
welds were c o m m e n c e d within a few
seconds after the specimens were removed from the oven.
a
Table 3 — Y-Groove Tests on 1 in (25 mm) Plate< >
Test
no.
Restraint
condition
Y9
Full
Y10
Y11
Full
0.79-in.
(20-mm)
slots
Energy
input
kJ/in. (kJ/mm)
Root < b |
opening,
in. (mm)
Root crack
length — % of
welddepth,c)
41-42
(1.6-1.7)
0.02(0.5)
30 max
0.02(0.5)
0.02-0.04
(0.5-1)
30 max
None
"
Y12
Y13
Y14
Y17
Y18
20 max
None
0.02-0.03
(0.5-0.8)
1.6-in.
(40-mm)
slots
"
"
None
30 max
30 max
"
"
33(1.3)
(a) Initial plate temperature 70 F (21 C)
(b) Measurements made on transverse sections
(c) No other type of cracking occurred
5.9 ii>.
(150 mm)
5 . 9 in.
(150 mm)
i
i
1
1
1 ~
c
r<S
E
$
|
%
A
L
c E
a* O
K O
A
4
B
CO
r<i
^.
£
l.9in.
cn E
00
CO
tj
"
sy
*-E
>mm)
',, "
E
E
CO
fc
B
B
i- 4
i
A>_
f
'
,62in.
D I A - 2 HOLES
(16mm)
DIA-2 HOLES
60°,
V"7
<
IL
GAP WIDTH
,08in. (2mm)
ENLARGED SECTION A - A
GAP WIDTH
.08in. (2mm)
ENLARGED SECTION B - B
ANGLED - GAP SPECIMEN
STRAIGHT-GAP SPECIMEN
Fig. 4 — One piece Y-groove weldability specimens used tor 1 in. (25 mm) plate
138-8 |
MAY
1976
Y-Groove Butt Weld Tests on
1 in. (25 mm) Plate
Before it was realized that the root
opening in the two piece specimen
was not being maintained in the reco m m e n d e d range, several tests
(Table 3) were made, with specimens
intended to conform to Fig. 2, on 1 in.
(25 mm) plate using the steel shim
plate, 0.08 in. (2 mm) thick, to set the
root opening in the test weld portions. Other details of the test procedure were as employed in the tests on
1-1/2 in. (38 mm) plate and as shown
in Table 3. All specimens were examined for cracking by magnetic particle testing of transverse sections as
was used for all of the CTS specimens
and most of the Y-groove specimens
in 1-1/2 in. (38 mm) plate.
After data were found concerning
the r e c o m m e n d e d tolerance on the
root opening, measurements of the
opening were made (Table 3) on the
transverse sections as had been done
for the specimens in the 1-1/2 in. (38
mm) plate.
Three specimens in 1 in. (25 mm)
plate were also prepared similar to
Fig. 2 except for additional machining in the test weld location on the
plate having the single bevel preparation. An additional 0.08 in. (2 mm)
thickness of metal was removed from
the bevel face so that when the two
chisel edged "noses" of the anchor
weld joint preparation were in contact, a root opening of 0.08 in. (2 mm)
was obtained in the test region. However, when the anchor welds were
completed, it was found that the
opening at the test weld had been reduced well below the desired tolerance. It was not until later in the study
that a further modification of this idea
was employed as exemplified by tests
YA and YB (Table 2).
Experience with the Y-groove test
to this point had shown that it was difficult in the two piece assembly to
control the root opening with the accuracy indicated by the Australian
study (Ref. 1) to be necessary. In contrast, no difficulty was experienced in
maintaining the 0.08 in. (2 mm) opening when it was made by the EDM
process in a one piece specimen,
such as that shown in Fig. 3, and in
the 1-1/2 in. (38 mm) plate. Also at this
time, it was noted that other research
(Ref. 2) had indicated that more
severe test conditions could result if a
straight rather than an angled root
o p e n i n g w a s e m p l o y e d in t h e
specimen.
Consequently, another series of
tests (Table 4) on 1 in. (25 mm) plate
was u n d e r t a k e n w i t h o n e p i e c e
specimens having either straight or
angled root openings produced by
the EDM process (Fig. 4). Prior to
welding, the width of the root open-
ing was checked with a feeler gage
and in all specimens was found to
conform to the limits 0.079 ±0.004 in.
(2.0 ±0.1 mm). No preheating was
employed and all specimens p r o vided full restraint. Examination for
cracking was performed by the magnetic particle method as in previous
tests. Measurements (Table 4) were
also made on the transverse sections
to recheck on the root opening width.
All other tests summarized in Table
2 were made using specimens which
permitted control of the root opening
within the r e c o m m e n d e d range of
0.079 ±0.004 in. (2 ± 0 . 1 mm). Only
severe weld metal cracking occurred
in test Y4 under full restraint conditions. Cracking appeared to be some-
that shown in this Figure. In these
specimens,
weld
cracking
predominated, with some toe cracking occurring in tests Y3 and Y8 and
some heat-affected zone cracking initiating from the root in test Y8. Only
the latter form of cracking occurred in
test Y7.
Table 4 — Y-Groove Tests on 1 in. (25 mm) Plate with Root Opening
Made by Electro-Discharge Machining (EDM)
Auxiliary Tests on
Transverse Sections
Occasionally, it was not possible to
determine from the magnetic particle
examination whether cracking was in
the weld or in the heat-affected zone.
In these cases, metallographic examination was undertaken in order to
ascertain the location of the cracks.
Macrophotographs were prepared
(Figs. 6-10) to illustrate the types of
cracking observed. Figure 11 shows a
typical heat-affected zone crack at
higher magnification.
Metallographic examination was
made on a section from test YY1
(Table 4) in order to assess the metallurgical effect of the EDM process
upon the steel adjacent to the weld
root. Very small, metallurgically
altered areas were noted intermittently along the cut edge of the
root opening preparation (Fig. 12).
Where these areas were located in the
heat-affected zone of the test weld,
there was evidence of coarse martensite formation (Fig. 13).
Microhardness tests were made on
representative sections of various test
welds. The results, summarized in
Table 6, were obtained using a Tukon
testing m a c h i n e with a d i a m o n d
pyramid indenter and a 10 kg load.
Test
no.
Root
opening
(Fig.4)(b)
Energy
Initial
input
temp..
k J / i n . ( k J / m m ; F(C)
YY1
YY2
YY3
YY4
YY11
YY12
YY5
YY6
YY7
YY8
YY9
YY10
Angled
Angled
Straight
Straight
Straight
Straight
Straight
Straight
Angled
Angled
Angled
Angled
35(1.4)
32(1.3)
32(1.3)
32(1.3)
30(1.2)
30(1.2)
32(1.3)
32(1.3)
31 (1.2)
31 (1.2)
40(1.6)
43(1.7)
HAZ crack
length-% of
weld d e p t h l c )
70(21)
70(21)
70(21)
70(21)
70(21)
70(21)
200(93)
200(93)
200(93)
200(93)
70(21)
70(21)
25-50 (Fig. 9)
40
50-75
To 100 (Fig. 10)
50-90
Up to 100
None
None
None
None
25-40
25-40
(a) All specimens were under full restraint.
{bl All root openings were 0.08 in. (2.0 mm) measured on transverse section.
(c) Cracks initiated at the root of weld, propagated upwards in the HAZ and in some cases changed direction to terminate in the weld No other weld cracking occurred.
-8" rnvrsit"""'
Discussion
In the CTS tests, each test weld
provided "trithermal" heat flow conditions and thus the " c o m b i n e d plate
thickness" at the test welds was always 3 in. (76 mm). No cracking of
any kind was f o u n d in any of the
specimens despite attempts to increase test severity by either doubling
the sample width (tests CC1 and CC2)
or by increasing the rigidity of the bottom plate (tests C5 and C6). The use
of an opening at the root of the test
weld, in all of the specimens, is known
to increase the cracking tendency.
Consequently the absence of any
cracking in this series of tests suggests that little difficulty would be expected in practical fillet welded joints.
It is evident from tests Y 1 , Y2, Y3
and Y8 (Table 2) that severe cracking
can occur in fully restrained Y-groove
weldability specimens intended to
conform with Fig. 2 but, in fact, having a root opening only about one half
T
I )
DEPTH Of RESTRAINT RELIEVING SAW CUTS - 0
FOA FULL RESTRAINT, Oft SPECIFIED S E M t M T E L Y
WHEN REDUCED RESTRAINT tS EMPLOYED-.
ja
i
i
*Z.7
190
i 23.4
MM
09«
2.90
l
1
C
•M
hi
0
mm
rw
192 4 . M
3 . 0 M 71.11
0 7 * 2,01
7.34
3.17 1 80.34
K)2 2 3 9
123 3 . 1 2
. 171 4 . 3 4
209
.242 LIB
.3*9 9.79
3.242 • 2 . 3 5
9.§3
STt I4.W
3 317 M . 0 3
. 4 0 * II. M
• 7 4 17.12
3.4W
• 7JK
.196 4 7 2
,7TO 19.3*
3931
89.69
20* 3 2 3
4
4
3ft. 1
z
90. • . 5 3 1
4*4
•
A
ta.
MM
M7
13.49
169 4 . i »
Fig. 5 — Modified two piece Y-groove weldability specimen showing dimensions required
for machining each piece
WELDING
RESEARCH SUPPLEMENT!
139-8
what more servere than in tests
Y1, Y2, Y3 and Y8. Because severe
cracking, apparently with weld cracking predominating, also occurred in
tests Y15 and Y16, it is evident that the
reduction in restraint due to the 0.79
in. (20 mm) slots had no major influence.
However, with still further reduction in restraint due to the 1.6 in. (40
mm) slots, a marked change was observed in the nature of cracking in
Table 5 — Summary of Microhardness Test Data a)
Sections
from
test no.
Reference
Table no.
Combined plate
thickness
in. (mm)
Energy input
kJ/in.
(kJ/mm)
CC1.CC2
See text
2 (51)
Y5.Y6
2
3 (76)
Y24, Y25
2
3 (76)
Y11, Y12
3
2 (51)
Y17, Y18
3
2 (51)
30-36
(1.2-1.4)
40-43
(1.6-1.7)
40-43
(1.6-1.7)
41 - 4 2
(1.6-1.7)
33 (1.3)
YY11, YY12
4
2 (51)
30 (1.2)
YY9, YY10
4
2 (51)
40-43
(1.6-1.7)
Vickers , b )
(10 kg load)
DPN
211-250
(387-442)
222-228
(345-390)
196-218
(312-382)
209-213
(294-363)
190-211
(330-390)
232-237
(315-374)
205-232
(307-372)
(a) All specimens were welded at room temperature, i.e.. 70 F (21 C).
(b) Areas tested were the weld metal and the HAZ very close to the fusion line. Numbers without parentheses are for weld
metal; numbers in parentheses are for HAZ values.
Fig. 6 — HAZ toe and weld metal cracks (Test Y3 — Table 2). X8,
reduced 17%
Fig. 7 — HAZ crack from root deflecting
Table 2). X8, reduced 18%.
140-8 I MAY 1976
tests Y5 and Y6. In these tests, only
continuous, severe heat-affected
zone cracking was found, initiating
from the root of the joint. The results
indicate that, under the higher conditions of restraint, as in test Y4, weld
metal cracking usually occurs and
thus precludes or inhibits heat-affected zone cracking by reducing the
stresses imposed upon the heat-affected zone. With much lower restraints, as in tests Y5 and Y6, the
weld cools without cracking through
the temperature range which favors
weld cracking but the stresses imposed upon the heat-affected zone in
the cold cracking temperature range
are still sufficiently high to contribute
to this form of cracking.
As shown by tests Y24 and Y25,
preheating to 200 F (93 C) is sufficient to prevent such cracking. In
tests YA and YB, preheating to 300 F
(149 C) prevented cracking in
specimens providing the maximum
restraint. Judging from tests Y1, Y2,
Y3, Y4, Y7 and Y8, this level of
preheating has prevented, in tests YA
and YB, all forms of cracking such as
were encountered in the former tests.
Fig. 8 — Weld metal crack (Test Y4 — Table 2). X8, reduced
16%
into weld (Test Y5
HAZ crack (Test YY1 — 7ab/e 4). X8, reduced
16%
Fig. 12 — Remelted areas on surface of root opening made by the
EDM procedure in Y-groove specimen, remote from test weld
(Test YY1 — Table 4). X600, reduced 26%
Fig. 10 — HAZ crack (Test YY4 — Table 4). X8, reduced 20%
Fig. 13 — Remelted area on surface of root opening made by the
EDM procedure in Y-groove specimen, very close to test weld
(Test YY1 — Table 4). X600, reduced 37%
Fig. 11 — Microstructure and crack in HAZ ot Y-groove specimen
(Test YY1 — Table 4). X200, reduced 23%
Possibly some temperature intermediate between 200 F (93 C) and 300
F (149 C) would also have been as
effective.
Certain factors should be considered when comparing the extensive cracking in the Y-groove tests on
1-1/2 in. (38 mm) plate with the absence of cracking in the CTS tests on
1 in. (25 mm) plate. In both tests, the
"combined plate thickness" was the
same. However, the energy input level
was lower in the CTS tests than in the
Y-groove tests and the carbonequivalent of the steel in the CTS tests
was higher than that of the thicker
plate in the Y-groove tests. These two
factors would be expected to result in
a greater tendency for cracking in the
CTS tests because of the development of more crack-susceptible
microstructures in the heat-affected
zone. The hardness data in Table 5
support this view. Considerably
higher hardness values were found in
the heat-affected zone very close to
the weld in the CTS specimens, as
compared with hardness values at
corresponding locations in butt weld
specimens. Because of the differ-
ence in cracking behavior in the two
types of specimens, it must be concluded that the level of stress, in the
root area, developed by the specimens is a chief reason for the difference. Even with the deepest slots
employed, i.e., 1.6 in. (40 mm),
the stress level in this area of the 1-1/2
in. (38 mm) Y-groove specimen must
be greater than that of the 1 in. (25
mm) CTS specimen.
In the Y-groove tests on 1 in. (25
mm) plate, Table 3 shows that weld
metal cracking, extending from the
root, was the only form of cracking
encountered. The root opening in all
of these tests was significantly less
than that which had been intended.
With an energy input of 41-42 kJ/in.
(1.6-1.7 kJ/mm), moderate weld
m e t a l c r a c k i n g o c c u r r e d in
specimens providing full restraint
(tests Y9 and Y10) and there was evidence of reduction in cracking in tests
Y11 and Y12 which provided some
reduction in restraint. It is noted also
that cracking was considerably more
severe in similar tests (i.e., Y15 and
Y16, Table 2) on the 1-1/2 in. (38 mm)
plate. This trend would be expected
as restraint increases with the plate
thickness.
In tests Y13 and Y14, the further
lessening of restraint by still deeper
slots, caused complete elimination of
weld metal cracking. Thus it seemed
probable that, with the elimination of
weld metal cracking, some heat-affected zone cracking would then occur as was the case with tests (Y5, Y6,
Table 2) on the 1-1/2 in. (36 mm)
plate. A probable, partial explanation
for the absence of heat-affected zone
cracking in tests Y11, Y13 and Y14 is
that the root opening was too narrow
and thus, based on the Australian research (Ref. 1), inhibited crack initiation in the heat-affected zone.
Another factor is that the greater
cooling rates resulting from the
greater "combined plate thickness" in
tests Y5 and Y6 could have resulted in
more crack sensitive, heat-affected
zone microstructures than in tests
Y11, Y13 and Y14 despite the differences in plate carbon-equivalents.
The development of more crack sensitive, heat-affected zone microstructures in tests Y5 and Y6 is indicated
by the higher hardness values shown
W E L D I N G R E S E A R C H S U P P L E M E N T ! 141-s
in Table 5 for corresponding specimens.
Weld cracking occurred in tests Y17
and Y18 under the same restraint
conditions as for tests Y13 and Y14.
This is attributed to the higher stress
level imposed on the weld due to the
smaller weld size in tests Y17 and Y18
resulting from the lower energy input,
relative to tests Y13 and Y14. Hardness data indicate that, despite the
differences in energy input, the weld
metal microstructures were similar
and hence should have a similar
cracking tendency.
With the correct root opening provided by EDM in 1 in. (25 mm) plate,
all cracking (Table 4) was in the heataffected zone except in some cases
where the upper portion of the crack
deflected into the weld (as in Fig. 7).
With the angled or oblique root opening (Fig. 4), cracking was " m e d i u m " in
severity. With the vertical or straight
root opening (Fig. 4), cracking severity
was
increased
significantly.
Preheating to 200 F (93 C) eliminated
cracking with both types of root opening orientations. It has been suggested (Ref. 2) that the reason for the
greater severity of the specimen having a straight root opening, as c o m pared to the specimen having an
angled root opening, is the greater
stress concentration in the vicinity of
the root of weld and the base metal in
t h e f o r m e r s p e c i m e n . T h i s was
thought to be related to the difference in the angle f o r m e d between
the weld metal and the base metal at
the weld root; the angle being smallest in t h e s p e c i m e n h a v i n g t h e
straight root opening.
A comparison of tests Y9 and Y10
(Table 3) with tests YY9 and YY10
(Table 4) indicates that, in the 1 in. (25
mm) Y-groove specimens, a root
opening much less than the recommended value of 0.08 in. (2 mm) tends
to produce only weld metal cracking,
w h e r e a s w i t h the r e c o m m e n d e d
o p e n i n g only heat-affected zone
cracking occurs.
The configuration of the weld metal
appears to have a significant effect
upon cracking in the Y-groove test.
Examination of many sections from
the current study showed that, with an
angled root opening of about 0.08 in.
(2 mm), the weld root surface was
usually close to being perpendicular
to the thickness dimension of the
specimen as in Fig. 7, although Fig. 9
shows an e x c e p t i o n to this. The
angles formed by the root surface and
the root faces were nominally usually
about 60 and 120 degrees respectively, with the smallest angle being at
the side of the joint having the singlebevel p r e p a r a t i o n . However, the
sharpness of the notch was variable
even with the same nominal angular
formation, due to variations in the
142-s | M A Y
1976
shape of the junction of the root surface with the root face. For example,
at the junction of the root surface with
the side of the joint having the singlebevel preparation, Fig. 7 shows a
rather gentle notch effect, but Fig. 8
shows a sharper effect. In specimens
having a straight root opening of
about 0.08 in. (2 mm), the sharpest
notch was at the junction with the side
having the 90 degree preparation.
The weld root surface formed typically an angle of about 60 degrees
(Fig. 10), and sometimes as low as 30
degrees, with this side of the root
face.
In the angled root openings that
had widths of about 0.02-0.04 in. (0.51 mm), the weld root surface was
usually not perpendicular to the thickness dimension of the specimen and
an angle of about 30 degrees was
formed by the root surface and the
root face on the side of the joint having the single-bevel preparation. This
usually resulted in the presence of a
particularly sharp notch at the junction of the weld and root face. In all
such specimens, when cracking occ u r r e d , it was initiated at these
notches and propagated into the weld
metal in a direction approximately
parallel to the c o l u m n a r g r o w t h
pattern. The origin and path of cracking were similar to that shown in Fig.
6, although in this case the notch
effect was less severe than in most of
the specimens having root openings
of 0.02-0.04 in. (0.5-1 m m ) . It is
thought that with root openings in this
range, weld cracking was usually
favored by the combination of the
sharper notch and the orientation of
the columnar growth pattern of the
weld metal. The narrower root
openings appeared to promote a
growth pattern (Fig. 6) in the region of
the notch, adjacent to the side of the
joint having the single-bevel preparation, that was more nearly transv e r s e to t h e d i r e c t i o n of s t r e s s
developed by the weld shrinkage acting against the restraint of the specimen, than in the case of specimens
(Fig. 7) having a root opening of the
order of 0.08 in. (2 mm).
With root openings of about 0.08 in.
(2 mm), cracking was normally initiated at the intersection of the root
surface and the side of the joint having the double-bevel preparation,
(Figs. 7, 8 and 9). Except for test Y4
(Fig. 8) in 1-1/2 in. (38 mm) plate
under full restraint conditions, cracking was always initiated as heat-affected zone cracking. Figure 7, which
is representative of most of the sections from specimens having a root
opening of 0.08 in. (2 mm), shows a
"wineglass" configuration in which
the lower portion of the fusion line
on the side of the joint having the
double-bevel preparation is more
nearly transverse to the stress d i rection than that on the other side of
the j o i n t . This w o u l d o r i e n t t h e
hardened heat-affected zone into a
more favorable position for cracking
on the one side of the joint. This may
also be a factor in the greater tendency for cracking of the Y-groove specimen having a straight rather than an
angled root opening (Fig. 10).
The occasional occurrence of toe
cracks (Fig. 6) is related, most p r o b ably, to a particularly sharp junction
of the surface of the weld and the
steel. Such junctions could also contribute to the form of weld cracking
shown in Fig. 8 where it is probable
that cracking traveled upward from
the root and downward from the toe.
Root openings prepared by EDM
machining in one piece Y-groove
specimens were found to be consistently within the recommended tolerence for the opening. However, a possibly adverse factor relative to using
EDM for root opening preparation
was the observation that the cutting
process resulted in the presence of
small areas of remelted metal located
intermittently along the cut surfaces
and when such areas were austenitized by the test weld, a coarse, only
slightly tempered martensite resulted.
It is expected that such zones would
occur at some points along the test
portion just at the intersection of the
root face and the test weld. Because
of the microstructure in these zones,
cracking would be more apt to be init i a t e d t h a n in t h e n o r m a l h e a t affected zone, i.e., resulting only from
the test weld thermal cycle.
The following conclusions are particularly valid for the steel plate and
electrodes as employed in this study
but should be at least partially valid
for other steels and electrodes.
Conclusions
1. The CTS fillet weld specimen
provides less severe conditions for
the d e v e l o p m e n t of heat-affected
zone cracking than does the Y-groove
butt weld specimen. Cracking may
not occur in CTS tests under even
m o r e severe c o n d i t i o n s , s u c h as
lower energy input levels, higher plate
carbon-equivalent values and c o m bined plate thicknesses, which result
in severe heat-affected zone cracking in Y-groove tests. Thus, much
caution should be employed in basing
conclusions concerning cracking t e n dencies at the root pass of butt joints
upon the results of fillet weld cracking studies.
2. In employing the Y-groove specimen on heavier plate, such as 1 1/2 in. (38 mm) thick plate, severe
weld metal cracking is likely to occur
preferentially at high restraint levels,
thus inhibiting the assessment of the
propensity of the steel to heat-
affected zone cold cracking. At a
much lower level of restraint in 1-1/2
in. (38 mm) thick plate, weld metal
cracking is eliminated and severe
heat-affected zone cracking then occurs. Such cracking is eliminated by
preheating at a temperature of 300 F
(149 C).
3. Weld metal cracking is severe
even with root openings of the order
of 0.04 in. (1 mm), i.e. much less
than the recommended value of 0.08
in. (2 mm), when the 1-1/2 in. (38 mm)
Y-groove specimen is unslotted and
thus provides high restraint.
4. Moderate to severe heat-affected zone cracking, initiating at weld
toes which cause a particularly severe
notch effect, is likely to occur in Ygroove specimens in 1-1/2 in. (38 mm)
plate under conditions of high restraint.
5. In 1 in. (25 mm) Y-groove specimens, weld metal cracking rather
than heat-affected zone cracking occurs when the root opening is much
less than the recommended value of
0.08 in. (2 mm). When the latter value
is provided by the EDM process, only
heat-affected zone cracking occurs
under otherwise similar test conditions.
6. In Y-groove specimens, cracking which initiates from the root of
weld appears to be influenced significantly by the configuration of the
weld metal and this configuration is
affected by the width of the root opening.
7. In 1 in. (25 mm) Y-groove specimens, a significantly greater extent of
heat-affected zone cracking tends to
occur when the root opening orientation is straight rather than angled as is
normally employed. All cracking is
eliminated by preheating to 200 F
(93 C). The greater severity of specimens having the straight root opening may be due, in part, to orientation
of the heat-affected zone to a more
favorable position for crack development, relative to the stress direction,
than in specimens having the angled
root opening.
8. Although the EDM process permits good control of the root opening,
it produces small zones along the cut
edge which, when subjected to the
thermal cycle of the test weld may become more sensitive to crack initiation than the steel which is being
examined for heat-affected zone
cracking susceptibility. Consequently, it is concluded that the two piece
specimen, specially machined to provide automatic control of the root
opening is a more reliable solution to
the problem.
References
1. Hensler, J. H., Graham, J . W., and
Cullen, G. V., "Cold Cracking Tests for
Determination of Weldability-Assessment
of the Tekken Test," Australian Welding
Research, 1 (9), p 1, 1970.
2. Sasaki, H., Watanabe, K.. Kirihara,
S., and Sejima, I., "Effects of Restraint
Stress and Intensity of Restraint on Del a y e d C r a c k s in t h e W e l d s of 80
k g / m m 2 / H i g h - S t r e n g t h Thick-Plate Steel"
— IIW Document IX-784-72 (X-654-72),
July 1972.
3. Goda S., Sato, M., Nakai, T.. Ando,
A., and Minehisa, S., "Weld Bond Fracture and a New 80 k g / m m 2 High-Strength
Steel for Submerged Arc Welding" — IIW
Document
IX-632-69 (X-527-69), July
1969.
WRC Bulletin
No. 184
June 1973
"Submerged Arc Weld Hardness and
Cracking in W e t Sulfide Service"
by D. J. Kotecki and D. G. Howden
This study was undertaken to determine:
(1) The causes of higher-than-normal hardness in submerged-arc welds in plaincarbon steels
(2) The levels of strength or hardness which will not be susceptible to sulfidecorrosion cracking
(3) Welding procedures which will assure that nonsusceptible welds will be
produced.
Concentration is primarily on weld metal, though some consideration to the
weld heat-affected zone is given. The study covered a two-year period. The first
year was concerned with a macroscopic view of the weldments. In that first-year
study, some inhomogeneities were observed in weldments which are not obvious
in a macroscopic view of the weldment. It appeared likely that these
inhomogeneities could affect the behavior of the weldment in aqueous hydrogensulfide service. Accordingly, their presence and effects were investigated during
the second year.
The price of WRC Bulletin 184 is $3.50 per copy. Orders should be sent to the
Welding Research Council, 345 East 47th Street, New York, N.Y. 10017.
W E L D I N G R E S E A R C H SU P P L E M E N T I 143-s
1975 STRUCTURAL WELDING CODE
The new edition of the Structural Welding Code, AWS
Dl.1-75 is here — completely revised with important requirements that you need to know.
Automatic Revision Service
The price of both the looseleaf and the soft-cover
bound versions of the 1975 Structural Welding Code includes
the 1976 and 1977 Revisions, automatically sent to you as
they are published.
Sturdy 3-Ring Binder
The heavy-duty vinyl binder will hold the Structural
Welding Code, the soon-to-be published Code Commentary,
and the Reinforcing Steel Welding Code, AWS D12.1-75 (the
latter two documents to be sold separately). The binder includes dividers for the main sections of Dl.1-75 as well as the
other two publications mentioned.
A limited number of soft cover bound copies is available.
Structural Welding Code
1. Looseleaf text, Dl.1-75, in binder with dividers plus
1976 and 1977 revisions
2. Soft cover bound text, Dl.1-75, plus 1976 and 1977 revisions
$25.00
$24.00
Reinforcing Steel Welding Code, AWS D12.1.75 soft-cover, bound
$5.00
Code Commentary, soft-cover bound
Price not set
Discounts: 25 percent to A and B members; 20 percent to bookstores,
public libraries and schools; 15 percent to C and D members. Add 4 percent sales tax in Florida. Send orders to: American Welding Society, 2501
N.W. 7th St., Miami, FL 33125.
144-s I MAY 1976
Fly UP