Hot Cracking Susceptibility

of Austenitic Stainless Steels

Hot-cracking resistance is improved by (a) reducing

phosphorus and sulfur to < 0.002%, (b) decreasing silicon,
(c) increasing carbon when niobium is high, and (d)
increasing nitrogen when niobium is absent

BY T. OGAWA AND E. TSUNETOMI

ABSTRACT. In fully austenitic weld metal, Introduction application of the results to the produc-
the hot-cracking susceptibility increases tion of materials for actual welded struc-
remarkably for phosphorus and sulfur The control of impurities and alloying
tures. Hot-cracking susceptibility was
contents in excess of 0.015% and elements in steels has been facilitated by
investigated by the Varestraint test
0.010%, respectively. Varestraint tests the recent progress of steel refining tech-
method developed by W. F. Savage and
revealed that an extra-low-P and -S (both niques. This has led to intense activity in
C. D. Lundin (Ref. 1), and the effects of C,
less than 0.002%) austenitic stainless steel, research and development of stainless
Si, P, S, Nb, N and <5 ferrite on hot cracking
refined by the Metal-bearing Solution steels. Austenitic stainless steels play an
of austenitic stainless steels were investi-
Refining (MSR) process exhibits superb important role as structural materials in
gated. Segregation of these elements at
hot-cracking resistance comparable to nuclear reactor and general chemical
crack surfaces was investigated by auger
that of Type 304, even in fully austenitic plant applications. To enhance the quality
electron spectroscopy (beam diameter
weld metal. Niobium has a detrimental of such structural materials, this progress
0.2-1.0 nm).
effect on fully austenitic weld metal, since of steel refining techniques should be
accompanied by a similar advancement From these investigations, minimization
the hot-cracking susceptibility increases
in the field of welding techniques. of phosphorus, sulfur and other impuri-
considerably when the niobium content
ties was found to lower hot-cracking
exceeds 0.30%. The same tendency is Hot-cracking susceptibility of austenitic
susceptibility of fully austenitic stainless
also observed in the heat-affected zone stainless steels depends upon metallurgi-
steel to the level of that of Type 304.
of welds. cal phenomena related to the chemical
With Nb-stabilized austenitic stainless
When the niobium content is high composition, and constitutes one of the
steel (Type 347), which is known to have
(0.78%), increasing carbon decreases the most important factors affecting weld-
a high hot-cracking susceptibility, strict
hot-cracking susceptibility. With carbon ability. Many studies have been made in
control of niobium, impurities and 8 fer-
contents higher than 0.10%, the cracking this respect (Ref. 5,6), but they have not
rite proved to impart high resistance to
of high-Nb steels is remarkably reduced. succeeded in thoroughly clarifying the
hot cracking.
O n the other hand, when the niobium mechanism of hot cracking and the
effects of chemical composition. From These findings were applied to pro-
content is low (0.26% or less), the cracks duction of 25 and 40 mm (1 and 1.6 in.)
produced are small and carbon shows the view-point of steelmaking, there are
still some difficulties in quantitatively thick plates of modified Type 347 austen-
hardly any effect. itic stainless steel as a new material for the
determining the effect of each alloy ele-
Phosphorus, silicon and niobium have piping system in boiling water nuclear
ment on the hot-cracking susceptibility of
considerable effects on the duplex reactors (BWR). Joints of these thick
stainless steel.
(7 + a) welds. That of niobium is espe- plates welded by gas tungsten arc weld-
cially strong; when present at the 0.30- A study was made of the effect of ing (GTAW), shielded metal arc welding
0.40% level, it increases the hot-cracking chemical composition of austenitic stain- (SMAW), and submerged arc welding
susceptibility of the weld metal and heat- less steels on weldability. Emphasis was (SAW) exhibited excellent hot-cracking
affected zone greatly, even when 5% 5 placed on quantitative measurement hot- resistance (Ref. 8). Weldability of these
ferrite is present. Auger electron spectro- cracking resistance with a view to the plates will be reported separately,
graphic (AES) analyses revealed pro- together with their stress-corrosion-
nounced segregation of phosphorus, sul- cracking resistance and mechanical prop-
fur, silicon and niobium on hot-cracking Paper presented at the 10th Golden Gate
erties (Ref. 21).
Metals and Welding Conference held lanuary
fracture surfaces of Type 310S, and in
3 l-February 2, 1979 in San Francisco, Califor-
both high-Si and Nb-containing stainless nia.
steels. Hot-crack fracture surfaces of Experimental Procedure
MSR-treated extra-low-P and -S steel T. OGAWA is a Senior Researcher and E.
TSUNETOMI is Manager of Welding Laborato- Tables 1-3 show the chemical compo-
showed much less phosphorus and sulfur
ry I, Products Research & Development Labo- sitions of the test materials used. The 10,
segregation than did an ordinary com- ratories. Nippon Steel Corporation, Fuchinobe 20 and 50 kg (22, 44 and 110 Ib) ingots
mercial grade. Sagamihara kanagawa-ken, lapan. were forged, rolled and solid-solution-

8 2 - s | M A R C H 1982
heat-treated at 1050°C (1922°F) in the electron spectroscopy (AES). The amount Hot-cracking susceptibility increases
laboratory. of 8 ferrite in the weld metal was mea- markedly as the phosphorus content
The hot-cracking susceptibility was sured by a Ferrite Scope (manufactured exceeds 0.015%. Figure 3 shows the
evaluated either by the Varestraint (longi- by Helmut Fischer G.m.b.H., West Ger- effect of sulfur; here phosphorus content
tudinal type) test (Ref. 1) or by TIGAMA- many). was maintained at approximately 0.002%
IIG (Spot Varestraint) test (Ref. 2,27,28). to eliminate its effect. Note that the
The welding conditions for GTAW were Results hot-cracking susceptibility increases rap-
as follows: Varestraint test at 70 amperes idly for sulfur content above 0.010%.
(A) and 16 volts (V) with a travel speed 75 A comprehensive study of the hot- Therefore, even fully austenitic weld
mm/min. (2.95 ipm) on 5.0 mm (0.2 in.) cracking susceptibility of austenitic stain- metals such as 25% Cr-20% Ni (Type
thick plates, and TIGAMAJIG test at 40 A less steels should include a study of fully 310S), can be improved substantially in
and 10 V arc-spot with 5 second (s) arc austenitic weld metals as well as of hot-cracking resistance by lowering both
time on 2.0 mm (0.08 in.) thick plates. duplex weld metals consisting of a mix- phosphorus and sulfur contents to
Hot cracking occurred both in the ture of austenite and ferrite. This is be- approximately 0.002%. It should be
weld metal and the heat-affected zone. cause of the well-known fact that 8 ferrite noted that both affect the hot-cracking
In the evaluation of hot cracking suscep- has such a great effect on the hot- susceptibility, as will be discussed later.
tibility by the TIGAMAJIG test, the total cracking susceptibility that the effect of Figure 4 shows the results of Vare-
crack length in the weld metal was nor- chemical components cannot be deter- straint tests of fully austenitic stainless
mined readily when ferrite is present.
malized on the basis of the molten pool steels refined by the Metal-bearing Solu-
Accordingly, the effect of phosphorus,
diameter. TIGAMAJIG tests of dissimilar tion Refining (MSR) process which is one
sulfur, silicon and niobium were studied in
metal joints were performed by applying of the most advanced steel refining tech-
both fully austenitic and duplex weld
the arc-spot to the parts of specimens niques (Ref. 12). They are compared with
metals. *^
containing the dissimilar weld metal. Fig- the test results for a commercial steel and
ure 1 shows the Varestraint and TIGA- a low-P, low-S steel made by the induc-
MAJIG test methods in schematic form. Effects of P, S, Si, Nb, C and N on Hot tion melting process. The MSR-treated
The crack and fracture surfaces in the Cracking Susceptibility of Fully Austenitic extra-low-P and -S steels have extremely
Alloys high hot-cracking resistance compared to
welds were studied by dye penetrant
inspection, optical microscopy, electron Phosphorus and sulfur both have great that of the commercial Type 304 as
probe microanalysis (EPMA), scanning influence on hot-cracking susceptibility. discussed later, and the low-P, low-S air
electron microscopy (SEM) and auger Figure 2 shows the effect of phosphorus. melted steel had intermediate resistance

Table 1.—Chemical Composit on (wl %) of Laboratory Heats (10kg Melt) and Commercials- - C : X 1 0 ~ \ P,S: X 1 0 ' 3

C Si Mn P S Cr Ni Nb N ,5F<a> Note

S1 204 0.11 1.62 2 3 24.67 19.43 Nil f

S2 400 0.24 1.52 2 3 24.88 19.70 Nil ^
S3 385 0.23 1.53 2 5 24.79 19.58 Nil 0
S4 438 0.24 1,51 2 12 25.02 19.69 Nil g u00
S5 550 0.24 1.49 2 14 22.14 19.98 Nil 1 -
S6 160 0.25 1.51 2 15 21.80 19.87 Nil .«
S7 422 0.24 1.50 2 18 24.80 19.83 Nil \
PI 287 0.18 1.54 2 3 24.79 19.75 Nil -_
P2 272 0.21 1.48 12 3 23.87 18.45 Nil c N
P3 211 0.19 1.51 15 4 24.92 19.62 Nil ..§ o. *
P4 200 0.24 1.54 18 5 21.66 20.14 Nil -g =
P5 211 0. 18 1.54 21 4 25.07 19.71 'TO Nil fI
c
Sil 470 0.16 1.11 14 8 25.0 18.5 Nil
Si2 520 0.18 1.11 14 9 25.0 18.0 — Nil
Si3 470 0.75 1.12 15 7 25.3 18.6 2. Nil
Si4 670 1.00 1.54 10 9 26.2 18.7 Nil
Si5 470 1.35 1.11 14 8 25.1 18.7 Nil
Si6 460 1.85 1.07 14 6 25.2 17.8 Nil I5
c5
Si7 600 2.00 1.15 16 13 24.8 18.5 Nil
Si8 630 3.01 0.39 14 6 18.4 13.4 0.3 i- .9Pu
Si9 670 3.30 1.62 18 9 26.6 18.5 3.0 .] i -
Si 10 590 3.55 0.39 13 5 19.2 13.5 17 -i
sin 204 1.22 1.47 2 4 23.50 20.43 Nil
Si12 180 1.24 1.52 2 2 22.74 21 9 8 Nil
Si13 130 1.65 1.40 3 2 22.64 20.93 Nil
Si14 234 2.04 1.50 2 3 22.94 20.54 Nil
J
Si15 640 0.24 1.50 2 3 25.64 19.80 Nil
304 500 0.75 0.94 26 7 18.3 9.4 160 5.0 A
316 700 0.66 1.01 20 6 16.7 12.4 M o 2.38 - 1.0
316L 200 0.76 1.04 20 8 17.5 14.8 M o 2.33 - 0.2
347 440 0.56 1.68 22 7 17.2 11.1 0.72 - Nil -l
310S 530 0.65 1.00 26 4 25.1 19.9 — 170 Nil a
STEEL-A 900 2.13 1.54 25 5 24.2 13.4 Mo. 0.77 2510 2.0 |
STEEL-B 300 0.76 1.47 23 2 24.5 13.1 M o 0.78 3500 2.0 c
STEEL-C 660 3.32 0.25 18 4 19.4 12.8 - - 3.0
Incoloy 800 400 0.34 0.86 6 4 22.2 31.1 Ti 0.50 200 Nil

•a>i5F —Sferrite in autogenous weld metal.

WELDING RESEARCH SUPPLEMENT | 83-s

Table 2.—Chemical Composition (wt-%) of Laboratory Heats (10kg Melt) with Variable Niobium Content—C: X10" 4 , S,P: X 10"

Mn Cr Ni Nb <5F Note
NB1 120 0.19 1.35 2 12 23.31 21.93 - Nil I
NB2 130 0.18 1.47 2 13 24.43 23.04 0.15 — Nil Z
NB3 150 0.19 1.46 3 13 24.34 22.89 0.24 03 Nil "O ^
NB4 170 0.19 1.46 2 13 24.53 22.93 0.31 Nil O DO
NB5 130 0.22 1.49 3 13 23.65 22.51 0.41 Nil 'Z-z L L .
• %

NB6 120 0.20 1.40 2 13 22.27 22.41 0.52 2. Nil

NB7 150 0.21 1.50 3 13 24.35 22.99 0.73 Nil >
NB8 160 0.19 1.43 3 12 24.15 22.69 0.94 Nil
I
NB13 140 0.17 1.51 2 14 24.41 22.86 0.26 353 Nil
NB14 390 0.19 1.51 3 12 24.01 22.66 0.25 416 Nil
NB15 460 0.19 1.49 3 13 24.13 22.68 0.27 446 Nil .2 2 •
NB16 990 0.18 1.47 3 12 24.13 22.79 0.27 393 Nil ~ S ul
NB17 1280 0.17 1.47 3 12 24.38 22.95 0.27 376 Nil > §

NB18 180 0.17 1.48 3 12 24.16 22.72 0.75 397 Nil

NB19 330 0.17 1.44 3 12 23.90 22.45 0.76 390 Nil c 5
NB20 480 0.21 1.49 3 12 24.51 22.96 0.80 340 Nil
NB21 880 0.19 1.50 2 12 24.16 23.05 0.78 319 Nil
NB22 1120 0.20 1.46 3 12 24.17 22.66 0.77 406 Nil
|1
1 u
T
NB23 150 0.19 1.45 5 4 21.05 22.51 0.25 247 Nil z
NB24 150 0.20 1.47 5 4 20.81 22.33 0.25 379 Nil 0
NB25 160 0.19 1.48 5 4 20.99 22.26 0.24 506 Nil c °^
NB26 190 0.19 1.54 5 4 21.13 22.40 0.24 722 Nil m u-
NB27 200 0.19 1.47 5 4 20.93 22.52 0.24 983 Nil T3

NB28 250 0.19 1.44 5 4 20.97 22.46 0.25 1320 Nil 4.

7B6 54 0.18 1.48 5 3 18.77 9.69 0.26 237 4.1 J

7B7 49 0.18 1.47 4 3 18.61 9.72 0.53 218 4.6 ±£ rg
5 %
7B8 58 0.19 1.49 4 2 18.67 9.76 0.78 221 4.7 _Q £
7B9 54 0.18 1.43 5 3 18.57 9.63 1.18 243 4.0 Z -o r-j
Kl 150 0.34 1.69 13 7 17.7 11.0 0.12 40 2.5 _ > oo
K2 140 0.34 1.72 14 7 17.6 10.9 0.15 36 2.6 OS*
K3 150 0.36 1.60 14 7 17.7 11.2 0.18 35 2.9 3 +
K4 160 0.36 1.62 14 7 17.6 11.1 0.23 35 3.1 $£
K5 150 0.36 1.68 15 7 17.7 11.0 0.30 34 3.1
K6 220 0.37 1.70 15 7 17.7 11.2 0.45 37 2.5
i
r
7B1 200 0.20 1.56 5 3 18.46 10.00 0.23 223 2.7 -5-
Oj O ffl
7B2 180 0.20 1.47 4 3 18.68 10.02 0.25 315 3.0
7B3 240 0.20 1.48 4 3 18.81 9.35 0.25 461 3.2 5 'TJ M
7B4 220 0.20 1.52 5 3 18.88 9.31 0.25 932 2.3 4 - fc "•
7B5 290 0.20 1.46 5 3 20.22 9.28 0.25 1330 3.4

Table 3—Chemical Composition wt-°») of Laboratory Heats (20 - 50kg)—c: X10" 4 , P.S:X10" 3

Mn Cr Ni Nb nF Note

K11 150 0.35 1.61 8 4 18.0 9.9 0.29 220 4.5 D_

K12 150 0.34 1.61 4 4 17.6 10.0 0.29 490 3.5
o *
K13 140 0.33 1.62 3 4 17.9 9.8 0.29 480 3.0
K14 140 0.32 1.72 3 4 17.8 9.4 0.29 500 3.3 I 03 DC
K15 160 0.34 1.76 24 8 17.9 9.9 0.28 450 3.2 z^
K16 140 0.35 1.67 27 7 17.7 9.7 0.28 470 3.0
K17 140 0.33 1.73 26 7 17.9 9.5 0.30 460 4.0
K7 0.38 1.74 18 8 17.5 11.4 0.30 300 _0)
150 1.0
K8 160 0.36 1.69 25 7 18.0 11.1 0.29 430 0.2 m a—

K9 160 0.35 1.69 25 7 17.7 11.4 0.30 620 Nil > op

KK) 180 0.35 1.70 25 7 18.0 11.2 0.30 800 Nil z
No.94 140 0.01 1.48 2 3 24.8 21.0 - Nil 1
No.95 130 0.01 1.48 2 3 24.8 21.2 - Nil -9 c/,
I - ^r

B 423 0.29 2.79 4 < 2 16.57 12.64

Mo.
0.94 290 Nil L .9P
Extr<

MSR
low S

No.942 2.50 0.20 1.50 < 2 <2 25.00 21.01 Nil

Mo.
No. 104 240 0.18 1.80 < 2 < 2 25.15 29.57 0.27 376 Nil
<

84-s | MARCH 1982

 normalized total crack length does not ing tests for the Nb-containing (0.25%)
fall below about 2.5 mm (0.1 in.) even stainless steel, the nitrogen content of
when its silicon content is lowered to which was varied by adjusting the nitro-
0.15%. This value is much higher than the gen addition during the melting and refin-
value (under 1.0 mm) in the vicinity of ing processes. Figure 9 also shows the
3.5% Si content where 5 ferrite is pres- results for this steel after having been
ent. heat-treated at 860°C (1580°F) for 17 h
Figure 6 shows the effect of niobium to precipitate niobium carbo-nitride.
on the hot-cracking susceptibility of a Within the nitrogen range from 0.02 to
fully austenitic weld metal. In the weld 0.13%, the hot-cracking susceptibility of
metal, the hot-cracking susceptibility the low Nb (0.25%) fully austenitic weld
increases rapidly as the niobium content metal did not vary significantly with the
exceeds 0.30%, and is equivalent to or nitrogen content whether or not heat
rather higher than that of commercial treatment was applied prior to welding.
Type 310S at a niobium content of
0.40%. The heat-affected zone exhibits a Effects of 6 Ferrite on Hot-Cracking
similar tendency; larger cracks occur as Susceptibility
the niobium content exceeds 0.40%.
Figure 10 shows the effect of 8 ferrite
Figure 7 shows the effect of increasing
on hot-cracking susceptibility, as investi-
carbon content on the hot-cracking sus-
gated by Varestraint test. The nitrogen
ceptibility of 0.26% and 0.78% Nb-con-
content of the weld metal was varied by
taining steels. The effect of carbon is
adding nitrogen to the argon shielding
remarkable with 0.787X, Nb-containing
gas. In pure argon shielding gas, 3 to 4%
steels. Note that the hot-cracking suscep-
Herrite exists in the weld metal of Type
tibility decreases substantially with
Fig. 1 — Varestraint test: A — longitudinal-type increase in carbon content. On the other 304 and steel A and steel B. With a 5
test; B- TIGAMAIIG (spot-Varestraint test) vol-% nitrogen addition, the 8 ferrite is
hand, the hot-cracking susceptibility of
eliminated in steel A and steel B, but a
0.26% Nb steel is low and relatively
little remains in Type 304. The crack
to hot-cracking. independent of carbon content below
length increases as the amount of
Silicon has an effect similar to that of about 0.10% C.
5-ferrite decreases, and decreases to 6 to
phosphorus and sulfur on the hot-crack- Nitrogen has a differing effect on hot-
8 mm (0.24 to 0.31 in.) in steel A, to 3 to 5
ing susceptibility of fully austenitic weld cracking susceptibility, depending on
mm (0.12 to 0.2 in.) in steel B, and to
metal. Figure 5 shows the effect of silicon whether or not niobium is present. Figure
approximately 1 mm (0.04 in.) in Type
on hot-cracking susceptibility of 0.01% P, 8 shows the effect of nitrogen on the
304.
0.008% S steel and another steel whose hot-cracking susceptibility of a commer-
Figure 11 shows the results of hot-
phosphorus and sulfur contents were cial 25% Cr-20% Ni Type 310S, which is a
cracking tests of laboratory-melted, nio-
limited to below 0.003% in order to niobium-free steel. In this case, nitrogen
bium-containing materials with varying
minimize their effects. was added to the weld metal by adding
nitrogen contents. As with the results
The effect of Si is remarkable in both nitrogen gas to the argon shielding gas
shown in Fig. 10, considerably larger
steels. Hot-cracking susceptibility in- used in performing the Varestraint Test.
cracks were observed in the weld metal
creases linearly with silicon content until The strain rate was varied from 1.0% per
without ferrite. The cracks were some-
approximately 1.5%, drops sharply be- second to 10.0% per second. At all strain
what larger than in commercial Type
yond 2.0%, at which point 8 ferrite begins rates, the crack lengths were shorter with
310S —Fig. 4.
to form, and settles to the level of Type (Ar + 5%N) shielding than with pure Ar
The addition of nitrogen has a great
304 in the vicinity of 3.5% Si. In the steel shielding by a substantially constant
effect on duplex weld metal as it renders
with higher phosphorus and sulfur con- amount.
8 ferrite unstable. Because of the strong
tent, also shown in Fig. 5, the value of Figure 9 shows the results of hot crack-
influence of nitrogen as an austenite

Specimen :0.02/0.03C-1.5Mn-0.003S-25Cr-20Ni Specimen :0.02/0.05C-1.5Mn-0.002P-25Cr-20Ni

fully 7- fully 7
Aug. Strain: 1.0%
Thickness.:2.0mm

Aug. Strain: 1.0%

1 Thickness, :2.0mm

0.010 0.020 0.030 0.010 0.020 0.030

P(%) S(%)
Fig. 2 — Effect of P on hot-cracking susceptibility in fully austenitic weld Fig. 3-- Effect of S on hot-cracking susceptibility in fully austenitic weld
metal metal

WELDING RESEARCH SUPPLEMENT | 85-s

 MSR:Metal Bearing Solution Refining Fig. 4 — Effect of P and S on hot-cracking susceptibility
of fully austenitic stainless steels
CN
•
N • n— several percent of 5 ferrite in weld metal.
< - a However, even in weld metal containing
A ^ - . ^ - - ~ * o
0- A A 0
_ ~"
8 ferrite, other elements have significant
O o • • — m ^'"^ O effects on the hot-cracking susceptibili-
u
a g— "ggg •§• °A T ty.
00
Figure 12 shows the effect of niobium
fully 7 ^*<^ on the hot-cracking susceptibility of the
\ A^
X ,_ t low S P ^ D welds in an extra-low C steel with 4-5% 5
o r~ • 25Cr-30Ni-3.7Mo-0.27Nblf' *^ y\ \ \
ferrite. Both in the weld metal and the
cc E MSR \ \.^
J&*^~^^^
J^A
< e ° 25Cr-20Ni J v^\ \ heat-affected zone, the hot-cracking sus-
LU D
25Cr-20Ni. low S.P v«\ <c^\^Jf^^
(A ceptibility increases rapidly when the nio-
Ul
CC c
CO - A 25Cr-20Ni, Commercial >^" Q O ^ \^0r bium content exceeds about 0.4%.
_l Figures 13 and 14 show how the hot-

o
S<CA^\^y ^ S cracking susceptibility of Nb-containing
CU
steel changes wth varying nitrogen and
a. "ra phosphorus contents. The effect of
O
o
_ i
a
nitrogen on the hot-cracking susceptibili-
"ra
LJ o ty is not very significant within the range
>
LU
1-
-a-<
TJ
CD

#
yCx
s ' y
A
/ •
/' <<\
of 0.015 to 0.12%. However, both the
weld metal and the heat-affected zone

£
CO
Q CO
\ i_
-i-a
exhibit higher hot-cracking susceptibility
X U) with 0.025% phosphorus than with
O 0
tr 1—
CN 0.005%, as is shown in Fig. 14.
< ra
ui >
<s> Microscopic Observation of Weld Crack
ui
cr
z
,_
a^PS**
/</^t*^\
°9J < ^ \ \ \ V ^ extra low S.P
Surfaces
Silicon is one of the elements that
LJ o i increases the hot-cracking tendency con-
1 nr> ' ' '
o_ w
5 siderably. The following are some obser-
o 0 1° 2 3 4
vations on the surfaces of cracks caused
Augumented Strain (%)
by high silicon in welded joints in high
> silicon material. Cracks of this type are
LU
Q often found in dissimilar weld joints
former, the weld metal is made fully A marked increase in the crack length
I where the steels, such as high-Si austenit-
O austenitic by nitrogen pick-up and its was observed to occur when sufficient
CC
ic stainless steels (e.g., Type 302B and
hot-cracking susceptibility increases con- nitrogen was added to eliminate the 8
< 3.5% Si-19% Cr-13% Ni, steel C) and fully
siderably. The increase in hot-cracking ferrite. In the case of Type 304, which has
LU austenitic steels (e.g., Type 310S and
susceptibility is particularly severe, both in sufficient 8 ferrite in the weld metal, the
Incoloy 800), are welded together.
steel A, with 2.13% Si, and the laborato- hot-cracking susceptibility is lower even
ry-melted materials in Fig. 11, with 0.3% when nitrogen is added. Hence, it is an The weld bead cracks shown in Fig. 15
z Nb. accepted welding procedure to provide are examples.
LLI
Figure 16 summarizes the number and
m length of cracks in an actual dissimilar
a. metal joint and the results of a TIGAMA-
o
_i JIG test applied to such dissimilar weld
UJ
beads. Figure 17 is an example of the
> crack surface in such dissimilar weld
LLI

X metal, and Fig. 18 shows the results of

o AES analyses made on the crack surface
shown in Fig. 17 at a point near the center
or of the figure. Argon sputtering to remove
< the absorbed gas layer from the crack
UJ
surface increased the auger electron
intensity from each element. Further
sputtering reduced the amount of segre-
a. gation until the concentration of the seg-
o
_ i
regated elements ultimately fell to their
UJ level in the matrix. This segregation is
> particularly pronounced with silicon,
phosphorus and sulfur. As discussed later,
X the segregation of phosphorus and sulfur
O is sharply reduced in the metal produced
< by special refining processes, such as
LLI MSR. This reduced segregation in the
(/) weld metal significantly decreased the
UJ
cc

Fig. 5 —Effect of Si on hot-cracking susceptibility in

austenitic weld metal

86-s | MARCH 1982

 Specimen :0.013C-0.2Si-1.4Mn-0.003P-0.011S-24Cr-23Ni Specimen :0.20Si-1.5Mn-0.003P-0.011S-24Cr-23Ni-0.04N
" a^ „
IS ° e •
CJ -C
E
M
*ir~~h~~t°}°
i ^§cf •! "
O
\ - fully 7

\
o o -
.c
0.78Nb—V* \\ Nb
*
0.78% 0.26%
°
on \
cID \\
a »
_J LO #
\ \
.*o \ \
CO
O
O -
\^ • \
aE \ * \
,° E
F-v5 o \ \
ro
^c
J5 \ \ [Aug. S t r a i n 1.2%
co (1) o \ \ | Cond. 70A-7.5cpm
iS o \ \f | Thickness, 5.0mm
in CN 0.26Nb \ • *
C=
1U "m
TJ
>s °o o ^
q o
o \ *o •
Aug. Strain 1.2%
70A-7.5cpm o
Thickness 5.0mm " ^ o
o . I o i i i l

o
m o " m
o *— *-
o
Nb (%) c(%)°
Fig. 7 —Effect of carbon on hot-cracking susceptibility of fully austenitic
Fig. 6—Effect of Nb on hot-cracking susceptibility of fully austenitic
Nb-containing stainless steel
stainless steel

h o t - c r a c k i n g susceptibility. yses m a d e o n a s m o o t h dendritic cell w a l l c a r b o n is principally a result o f the c o n -

Figure 19 s h o w s the results of EPMA at the crack surface in t h e 0.75% n i o b i u m tamination of the hot-crack surface b y
analysis o f a crack in t h e fusion z o n e o f steel s h o w n in Fig. 20F. R e p e a t e d anal- e x p o s u r e t o air. T h e crack surfaces o f
the a b o v e dissimilar w e l d and indicates yses w e r e m a d e by p r o j e c t i n g e l e c t r o n 0.75% N b - c o n t a i n i n g steel exhibit either
extensive segregation of silicon at the beams at o n e p o i n t o n crack surfaces. s m o o t h dendritic cell boundaries or
boundaries of the cellular dendrites. T h e A r g o n sputtering for 220 s t o r e m o v e s m o o t h grain boundaries s t u d d e d w i t h
cracks in t h e w e l d b e a d , s h o w n in Fig. material f r o m t h e surface decreased the r o u g h , island-like residues and precipi-
15A, occur at boundaries o f cellular d e n - n i o b i u m and n i t r o g e n peaks, indicating tates. Examples of this m o r p h o l o g y can
drites, p r e d o m i n a n t l y at austenite grain that segregation of these elements be seen in the w e l d metal crack surface
boundaries. existed in the 0.75% N b w e l d metal. of Fig. 20F ( s h o w n schematically as point
Like silicon, n i o b i u m increases h o t - H o w e v e r , the segregation w a s v e r y slight F in Fig. 23), a n d o n the H A Z crack surface
cracking susceptibility. Fig. 2 0 s h o w s the o n the crack surface of the 0.24% n i o b i - of Fig. 20E ( s h o w n schematically as point
fracture surfaces of hot cracks in a V a r e - u m s p e c i m e n , as s h o w n in Fig. 22, w h i c h E in Fig. 23).
straint test s p e c i m e n of t w o l o w - C , fully s h o w s the results of AES analysis m a d e o n Even w h e n the n i o b i u m at the s m o o t h
austenitic N b - c o n t a i n i n g stainless steels. a cell in t h e central part of Fig. 20C. cell a n d grain boundaries is n o longer
Figure 21 s h o w s the results of AES anal- T h e high auger e l e c t r o n intensity f o r d e t e c t a b l e (as a result of a m p l e a r g o n

Specimen :0.20Si-t.5Mn-0.005P-0.004S-21Cr -22Ni-0.25Nb

fully r- CN

Commercial 310S
(25Cr-20Ni) N
1" < -
GTAW Shield Gas I
O Ar JC o 0
• (Ar + 5%N)
uo
c
5
o
rf :S sr &• 2 1
u y
(0 ( A u g . S t r a i n 1.2% i o as welded
70A-7.5cpm
1 p l a t e t h i c k n e s s 5.0mm
2 E • 8 6 0 C X l 7 h r W.Q.
Heat T r e a t nent
"o
h 5
c 2 CM
«J S 0
i5
tn
9
• 1
BJ 0) — Aug. Strain 1.2% co • ° 8 0 0
70A-7.5cm min 0 • o
Thickness. 5.0mm >
o • •
o CM ^r to co O CN T
2 3 4 5 6 7 o
o
o
o
o
o
o
o O O O
Augmented Strain Rate (%/sec). N {%)

Fig. 8 — Effect on N on hot-cracking susceptibility in fully austenitic stainless steel Fig. 9 —Effect of N on hot cracking susceptibility in fully
weld metal without Nb austenitic Nb-containing weld metal

WELDING RESEARCH SUPPLEMENT 187-s

 Fig. 10 — Influence of 8 ferrite on hot-cracking suscep-
tibility in austenitic stainless steel weld metal

impurities at grain boundaries is reduced

in materials extra-low in both phosphorus
and sulfur, such as those prepared by the
MSR process. Figure 28 shows the hot
cracks in both MSR-treated material and
commercial type 310S produced by the
TIGAMAJIG test. Note that despite its
fully austenitic microstructure, the MSR-
treated material shows only small, fine
cracks.
As can be seen from the above and
Fig. 4, it can be said that hot-cracking
resistance as good as that of Type 304
(Ref. 8) can be obtained in fully-austenitic
steels by decreasing the level of phos-
phorus and sulfur to very low values.
However, commercial production of
stainless steels that are extra low in both
phosphorus and sulfur is not yet possible
because of a lack of availability of suitable
raw materials, together with the com-
plexity of the refining techniques and the
high production costs. Therefore, the
effect of 5 ferrite in reducing hot-cracking
susceptibility continues to be widely uti-
lized. The discovery of this effect dates
2 5 10 back to early times, and it has been
N added in shield gas A r (Vol.%)—- s described in many reports (Ref. 14,15,
24). Because these impurities are more
— D^-
-o- soluble in 8 ferrite than in austenite, the <5
ferrite is thought to serve as a kind of
impurity absorber.
D STEEL-A
Silicon was found to impair hot-crack-
A STEEL-B ing resistance considerably in fully austen-
O type 3 0 4 itic weld metal-Figs. 5 and 10 (Ref. 16).
However, when present at levels which
cause some 8 ferrite to form, silicon
GTAW 70A-7.5cm/min, Augmented S t r a i n 1.2%, improves the hot-cracking resistance
thickness 5.0mm remarkably. Fully austenitic weld metal
always shows silicon enrichment at the
grain boundaries —Fig. 19. The binary
sputtering), the studded residue at point F Nb(CN) as it shows high electron peaks Ni-Si phase diagram has several low-
of Fig. 23 still exhibits high auger electron for niobium, carbon and nitrogen. AES melting intermediate phases, such as
intensities for both niobium and nitrogen. results on the matrix near the eutectic Ni5Si2, Ni2Si, and NiSi. There are also
Therefore, this residue is considered to constituent indicate no auger electron typical low-melting eutectics, such as NiSi-
be a compound containing considerable peaks for niobium, and only minor ones NiSi2 (996°C or 1825°F), NiSi-Ni3Si2
quantities of niobium and niobium car- of carbon and nitrogen. (964°C or 1767°F) and Ni-Ni3Si (1152°C
bonitride Nb(CN). On the crack surfaces or 2106°F) (Ref. 6,9).
of the 0.24% Nb-steel, such residues and The segregation of silicon at grain
precipitates are not observed. However, Discussion
boundaries appears to cause the forma-
several inclusions of manganese-chromi- It is well-known that sulfur and phos- tion of such intermediate phases and
um oxides located at points B and C, and phorus combine with iron to form such eutectics. Therefore, elements or phases
small pebble-like sulfides including some low-melting compounds as FeS (1190°C which inhibit the enrichment of silicon at
niobium compounds at points A and D or 2174°F) and Fe3P (1166°C or 2131°F), grain boundaries can decrease hot-crack-
(refer to the sketch of Fig. 20B contained which in turn could form such low- ing susceptibility. If S-ferrite is the primary
in Fig. 24). melting eutectics as FeS-Fe (988°C or phase during solidification, silicon, being a
Figure 25 shows the microstructure in 1810°F) and Fe3P-Fe (1050°C or 1922°F). ferrite former, can be readily absorbed in
the vicinity of hot-cracks in the 0.75% Nb There also exist other low-melting eutec- the 8 ferrite and the segregation to auste-
Varestraint test specimen. Eutectic struc- tics, such as Ni3S2-Ni (637°C or 1179°F) nite grain boundaries thus minimized.
tures can be seen both at the tip of the and Ni3P-Ni (875°C or 1607°F). These The addition of nitrogen lowered the
crack in the weld metal in Fig. 24A and at constituents, which increase hot-cracking hot-cracking susceptibility of a fully aus-
the grain boundaries in the HAZ in Fig. susceptibility, are considered to form tenitic weld metal (Commercial Type
24B. Figure 26 shows an SEM image of when sulfur and phosphorus segregate at 310S) as was shown in Fig. 8. Nitrogen is
the eutectic constituent, the results of the austenitic grain boundaries during believed to inhibit the enrichment of
AES analysis of the eutectic constituent at solidification (Ref. 5,6,9,10,11,29). silicon at grain boundaries somewhat.
point G and an AES scan of the nearby The results of AES analyses of crack However, at this point additional studies
matrix at H. The eutectic constituent con- surfaces, summarized in Fig. 27, have of the precipitates at grain boundaries
sists mainly of niobium carbonitride confirmed that the segregation of such should be made in order to understand

88-s | M A R C H 1982
 Specimen :0.015C-0.35Si-1.65Mn-17.6Cr-11.3Ni-0.015P-0.30Nb-0.01S Specimen :0.005C-0.20Si-1.5Mn-0.004P-0.003P-l8.6Cr-9.7Ni-0.025N

Jr
CN
.—
N j ^ r
J, 1
•a-. ID
<
X _
1-^1
CO c
cu o
I N
c IN o
(1)
d
-—(r+a) -
—
r
o
CO s—<.
L_

O £ GTAW cond. 70A-7.5cm/min L.

[Aug. Strain : 1.2% E CO
. thickness : 5.0mm thickness 5.0mm ^ ^ ^
|70A-7.5cm/min .c
•*-*

o | / \ A u g . Strain 5.0%
h c
111 tn
_l
JC
o
CO

O CO

CO CD

1-
:>
1 3 <"
CD

c %
CO
CM
l/> /
ID
CN CD CO <0 / Aug. Strain 2.5%
o

r r
O
o o >
o o o o *"
N {%) r
IJ n-
Fig. 11 — (Above) Influence of N and 5 ferrite on hot-cracking suscepti- o
bilty in low G, Nb-containing stainless steel welds /* CN NT CD CO O CN

° Nb°(%) °
Fig. 12 — (Right) Influence of Nb on hot-cracking susceptibility of low-C,
Nb-containing stainless steel with duplex (y+a) weld metal

the role o f n i t r o g e n m o r e fully. structures should b e c o n s i d e r e d (Ref. 7,17,23). In general, t h e hot-cracking sus-
In d e t e r m i n i n g t h e a p p r o p r i a t e a m o u n t 17,18). ceptibility o f w e l d s increases greatly w i t h
o f 8 ferrite, b o t h e m b r i t t l e m e n t a n d D u r i n g t h e p e r i o d w h e n T y p e 347 w a s increase in n i o b i u m c o n t e n t , in b o t h fully
d e t e r i o r a t i o n o f c o r r o s i o n resistance in used w i d e l y in t h e U n i t e d States, N b - austenitic (Fig. 6) and duplex w e l d metals
a n d a r o u n d c5 ferrite as w e l l as o t h e r containing austenitic stainless steels f r e - (Fig. 12). T h e hot-cracking susceptibility
p r o b l e m s e n c o u n t e r e d in w e l d i n g actual q u e n t l y exhibited w e l d cracks (Ref. 3,4, increases considerably at a n d a b o v e

Specimen :0.02C-0.20Si-1.5Mn-0.005P-0.003S-18.8Cr-9.3/10Ni-0.25Nb Specimen :Q.015C-0.35Si-1.65Mn-17.6Cr-10/9.4Ni-0.3Nb-0.007S

(NI •—

c
CD
_l
r'-i
Jen-
-(r+a)- 5.0mmt
ru r£i flu
70A-7.5cpm
O
3 m
Aug. S t r a i n 1.2%
t7
o
To
GTAW : 70A-7.5cm/min
,0 thickness 5.0mm

Aug. Strain 5.0%

-lil-r4*_ril_ & lil

low P 0.005 — H - — P 0.025 "IS C/>;° (/)</)
Dw D r~-> D O
- N —f- N added 0.04 —-I f ) C/> Ifl p,

0.02 Heats Commercials

Fig. 13 —(Left) Influence of N on hot-cracking susceptibility of low-C,

Nb-containing stainless steel with duplex (y + a) weld metal

Fig. 14 — (Above) Influence of P and N on hot-cracking susceptibility in

low-C, Nb-containing stainless steel welds

WELDING RESEARCH SUPPLEMENT I 89-s

 (3.5Si-0.025P-19Cr-13Ni/0.8Si-0.025P-25Cr-20r\li)

® Dissimilar Weld Joint by GTAW

Thickness of plate :2.0mm. 110A-l2V-30cpm. Ar shielded ; 15*/min

TIGAMAJIG Crack Test Result Crack Number in B u t t Weld

thickness 2.0mm. GTAW :40A-5sec. Aug. S t r a i n : 1.0% Joint (350mm bead length)

Hot Crack Susceptibility ^Ferrite Face bead Root bead

/cr.ck \ /cr,ck \ Total
0 1 2 3 4 5 6 Uength mm' Vlength mm/

' "*-*V . *
9
II
0 10 19
. ® (1.0/1.5™) (1.0/l.5mm)

- r*^\\^. with filler*

^h 1.1 0 1
(1.5mm)
1

* -• >_. .** ' :, * :0.05C-3.7Si-1.9Mn-0.008P-0.005S-19Cr-10Ni

Fig. 16 —Hot cracking susceptibility on TIGAMAIIG crack test and butt joint weld
metal
30//,
-- / '
F/g. 15 —Hot crack in high-Si fully austenitic
weld metal: A — macrostructure; B-micro-
structure
/rov

— • — S i oxide

0.30% Nb in fully austenitic and above

0.40% Nb in duplex weld metals. Niobium
thus is more detrimental in fully austenitic
welds. Note that in the Varestraint test,
Uft
the influence of niobium on the total
crack length in fully austenitic steels at
only 1.2% strain (Fig. 6) is roughly com-
parable to that produced with 5% strain
in duplex weld metal (Fig. 12). Note also
that at 2.5% strain only slight cracking was
evident even at the 1.2% Nb level. Fig. 17 —Fracture surface of hot cracking in 50 100 500 1000
high-Si fully austenitic weld metal AES analysis Ar s p u t t e r i n g time (sec)
In weld metal deposited by SMAW, was made on a cell in the central part 13 5Si-19Crl3Ni-0 025P)steel
carbon improves the hot-cracking resis-
tance of Nb-containing steels even with (0 8Si-25Cr-20Ni-0 025PI Steel
niobium contents as high as 0.78% (Ref.
Fig. 18 — Segregation of Si on dendritic cell
3). The Varestraint test results shown in reported. Such thermal faceting is
fracture surfaces of hot-crack lr. high-Si fully
Fig. 7 indicate a considerable improve- observed not only on crack surfaces at austenitic weld metals as revealed by AFS
ment in hot-cracking resistance with a grain boundaries in the heat-affected
carbon content of 0.10% in a 0.78% Nb zone, but also on the cellular dendrites at
steel. However, in low-Nb steels, this crack surfaces in weld metal when
tendency is less pronounced. viewed at higher magnification —Fig. of eutectics which result from the reac-
Nitrogen improves the hot-cracking 20F. tion between the constituents which
resistance moderately for Nb-free fully Liquid films at grain boundaries are form the isolated, rough features seen in
austenitic steels —Fig. 8. However, the considered to be formed by liquification Figs. 20E and F, and the matrix in the
nitrogen causes a moderate decrease in
hot-cracking resistance when niobium is
present in duplex weld metal —Fig. 13 Specimen Current ..
(Ref. 25). The AES analyses of crack
surfaces in both the weld metal and the ^location of crack
heat-affected zone in high Nb steels at
locations where there were no precipi-
tates indicated that grain boundary segre- te a^jfe-
gation of not only niobium but also O.lOPAc ftiy^-y-
nitrogen and carbon were present.
(3.5Si-19Cr-13Ni) '
The morphology of the crack surfaces /'(20O-30N,) ^0050p
in the heat-affected zone of Nb-contain-
ing steels exhibits a predominantly 2.0mm thick plate -»FeB G*T
smooth appearance with some step-and bun joint melt run by GTAW — — : r Specimen Current

wave-like features — Figs. 20A, B, D and E. 110A-30cm min \d^l7

This morphology suggests that a liquid
film plays a role in the cracking mecha-
nism, as pointed out by Honeycombe,
Gooch, and Lundin (Ref. 19,26). The step-
and wave-like features are thought to be
thermal faceting (Ref. 6) which results
from the exposure of free surfaces to \^%^AJ-
high temperature during solidification as Fig. 19 — Segregation of Si and P on grain boundaries in dissimilar high-Si weld metal by
Hemsworth, Boniszewski, et al. have EPMA

90-s I M A R C H 1982
 Fig. 20 —Fracture surface of hot-cracking pro-
duced in a Varestraint test specimen oi a fully
austenitic low-C, Nb-containing stainless steel
by Varestraint Test

higher Nb steels. Similar reactions proba-

bly occur between the small pebble-like
sulfides shown in Fig. 20B and matrix in
lower Nb-containing steels.
The eutectic reaction (~1315°C or
2399°F) between niobium carbonitride
and 7 iron has been thought to be partly
responsible for high hot-cracking suscep-
tibility of high-Nb austenitic stainless
steels (Ref. 6,20). However, in the case of
the low-Nb (0.25%) steel, little change
was observed in the hot-cracking suscep-
tibility of specimens even when sub-
jected to a precipitation heat treatment at
860°C (1580°F) for 17 h — Fig. 9.

Conclusions

1. Fully austenitic stainless steels can

attain a highly improved hot-cracking
resistance, comparable to that of Type
304, if the segregation of phosphorus
and sulfur at grain boundaries is re-
Fig. 21 —(Below left) Segregation of Nb as
revealed by AES on a solidification crack sur-
face in high-Nb stainless steel weld metal—
center part of Fig. 20F
0 021C-0 1 9Si-1 58Mn-0 004P-0 004S 0.01 BC-0.1 9Si-1 .46Mn-0.003P-0.01 3S\ Fig. 22 — (Below right) Slight segregation of Nb
22.37Cr-21.06Ni 0.73Nb 0.095N ) ( 24 34Cr-22.89Ni-0.24Nb-0.03N / as revealed by AES on a solidification crack
high Nb low Nb surface in low-Nb stainless steel weld metal—
Center part of Fig. 20C

Specimen 0 01 5C-0 19S.-1 46Mn-0.003P-0 01 3S-24 3Cr-22 9N|- 0.24Nb-0.03N

Specimen:0.021 C-0.1 9Si- 158Mn-0 004P -0.004S-22.4Cr -21.1 Ni-0.73Nb-0.095N

Ar bomb.: 220sec Ar bomb.: 220sec

Ar bomb.: 60sec

Ar bomb.: 40sec

Ar bomb: 40sec

Ar bomb.: 10sec

Ar bomb.: 1 Osec

Ar bomb.: no

Ar bomb : non

Electron Volt(eV)

WELDING RESEARCH SUPPLEMENT 191-s

 Precipitates or eutectics on liquation
crack surface in heat affected zone
Sketch of Fig.20(e)

A "

Compounds on solidification crack

surface in weld metal
Sketch of Fig. 20(f)

100 200 300 400 0 '00 800 ' J 6^—J _

IM 200 300
Fig. 23 —AES analysis on hot-crack surface of high-Nb stainless steel
Fig. 24-AES analysis on crack surface of low-Nb stainless steel

stricted. This can be achieved by reduc-

ing both phosphorus and sulfur content
ZOv Kin
to less than 0.002%.
J.-V"' 2. Increasing silicon increases hot-
cracking susceptibility of the fully auste-
nitic weld metal in an almost linear
fashion.
fe 4&X 3. Niobium increases hot-cracking sus-
ceptibility by segregating at grain bound-
aries during welding. This segregation
® was confirmed by AES analysis of the
surfaces of both weld metal and heat-
affected zone cracks. The grain bounda-
ries of Nb-containing steels are studded
Fig. 25 —Eutecticonstituent
c observed in weld metal and HAZ of fully austenitic high-Nb stainless with island-like reaction products that
steel: A—at tip of crack in weld metal; B — at the grain boundaries and matrix of HAZ contain considerable quantities of nio-

SpecimenO.023C-OI9Si-l.53Mn-0.005P-0.004S-22.3Cr-2t.2Ni-0.73Nb-0.072N

After 360sec. c l e a n i n g t h r o u g h Ar bomb.

92-s | M A R C H 1982
 metals of fully austenitic stainless steels (rep. II).
Trans. IWRI 6 (1): 105-116.
12. Nakamura, Y., and Abe, S. 1977. The
manufacture and characteristics of ultra-high
purity austenitic stainless steels by the MSR
process. Nippon Steel Technical Report Over-
seas 10(11): 114-119.
13. Ogawa, T.; Aoki, S.; and Ito, T. 1976
(Aug.). The influence of S, P. C, O and other
impurity contents on the properties of autenit-
ic stainless steel weld metal. 65th Committee
of Welding Metallurgy, lapan Welding Society
(in lapanese).
14. Hull, F. C. 1967. Effect of delta ferrite on
the hot cracking of austenitic chromium-nickel
stainless steel. Welding journal 46 (9): 399-s to
409-s.
15. Lundin, C. D.; DeLong, W. T.; and
Ar sputtering (sec) Spond, D. F. 1975. Ferrite fissuring relationship
Fig. 27 — Segregation of S and P on the fracture in austenitic stainless steel weld metals. Weld-
surface of a solidification crack as revealed by ing lournal 54 (8): 241-s to 246-s.
AES 16. Polygary, S. 1969. (Feb.). The influence
of silicon content on cracking in austenitic
stainless steel weld metal with particular refer-
ence to 18- Cr-8 Ni steel. Metal Constr. and
Brit. Weld, j 93-97.
b i u m carbo-nitride. 17. Moore, Thomas, ). 1959. Elevated-tem-
4. Increasing the carbon content perature properties of modified Type 347
weld metals. Welding lournal 38 (12): 457-s to
i m p r o v e s the hot-cracking resistance
474-s.
considerably in steels w i t h a high n i o b i u m
c o n t e n t (0.78%), b u t this effect is less .. s {- 18. Thomas, R. G. 1978. The effect of
o-ferrite on the creep rupture properties of
p r o n o u n c e d w h e n n i o b i u m c o n t e n t is austenitic weld metals. Welding Journal 57 (3):
l o w (0.26%). 81-s to 86-s.
5. Increasing t h e n i t r o g e n content 19. Honeycombe, M. )., and Gooch, T. G.
i m p r o v e d the hot-cracking resistance of a 1970 (Sept.). Microcracking in fully austenitic
fully austenitic stainless steel (25%Cr-20% stainless steel weld metal. Metal Constr. and
Ni) w h i c h c o n t a i n e d n o n i o b i u m . Brit. Weld, j: 373-380.
20. Tamura, H., and Watanabe, T. 1973.
6. N i t r o g e n c o n t e n t s a b o v e 0.02% in
Mechanism of liquidation cracking in weld heat
n i o b i u m - c o n t a i n i n g steels d e c r e a s e d the
affect zone of austenitic stainless steels.
hot-cracking resistance significantly, and /. lapan Weld. Society 42 (5): 489 (in lapa-
seems t o increase h o t - c r a c k i n g suscepti- nese).
bility slightly in d u p l e x w e l d s containing 21. In preparation.
niobium. 22. U. S. NRC regulatory 1.31 control of
7. T h e 8 ferrite has a r e m a r k a b l e effect ferrite content in stainless steel weldmetal April
in decreasing the hot-cracking susceptibil- Fig. 28 —Macro- and microstructure of TIGA- 1978, acceptance criteria for bend tests, article
ity of austenitic stainless steel welds. 1, QW-163, section IX of ASME boiler and
MAJIG solidification crack: A - B-Q position;
pressure vessel code edited 1977, and class 1
H o w e v e r , t h e e x t e n t of t h e e f f e c t varies B - MSR melt B; C - commercial 310S
vessel for nuclear power plant, article 36,
considerably w i t h chemical c o m p o s i t i o n .
schedule 1, chapter 3, Ministrial Order of MITI
Even w i t h 8 ferrite present, the w e l d (Ministry of International Trade of Japan) issued
metal is not f r e e f r o m the effect of silicon, The welding of Type 347 steels. Welding September 3, 1970.
niobium, phosphorus, nitrogen and other lournal ib (10): 442-s to 448-s. 23. Sandowski, E. P. 1974. Modification of
elements, w h i c h t e n d t o increase h o t - 4. Christoffel, R. |. 1962. Cracking in Type cast 25Cr-20Ni for improved crack resistance.
cracking susceptibility. 347 heat affected zone during stress relaxa- Welding journal 53 (2): 49-s to 58-s.
8. A f t e r e x p o s u r e t o a p r e c i p i t a t i o n tion. Welding /ournal 41 (6): 251-s to 256-s. 24. Astrom, H.; Loberg, B.; Bengtsson, B.;
heat t r e a t m e n t at 8 6 0 ° C (1580°F) for 17 5. Borland, |. C. 1962. Suggested explana- and Easterling, K. E. 1976. Hot cracking and
tion of hot cracking in mild and low alloy steel micro-segregation in 18-10 stainless steel
h, l o w - N b (0.25%) fully austenitic stainless
weld. Welding Research Abroad VIII (2): 73- welds. Metal Science (7): 225-234.
steel s h o w e d n o significant change in its
87. 25. Pascarel, L.; Hubert, M.; and Messager,
hot-cracking susceptibility.
6. Hemsworth, B.; Boniszewski, T.; and C. 1971 (Dec). Aspects particuliers de la fissu-
Eator, N. F. 1969. Classification and definition ration a chaud par soudaga d'aciers austeniti-
Acknowledgments of high temperature welding cracks in alloys. ques; influence de I'azote. Revue de Metallur-
Metal Constr. and Brit. Weld. j 2: 5S-16S. gie: 809-828.
T h e authors express their g r a t i t u d e f o r
7. Poole, L. K. 1953. The incidence of crack- 26. Lundin, C. D., and Spond, D. F. 1976.
the helpful suggestions a n d discussions ing in welding Type 347 steels. Welding Jour- The nature and morphology of fissures in
f r o m Drs. Scoichi Nakanishi (Director) and nal 32 (8): 403-s to 412-s. austenitic stainless steel weld metals. Welding
Harumasa Kakamura of t h e W e l d i n g 8. Abe, S., and Ogawa, T. 1979. A modified Journal 55 (11): 356-s to 367-s.
Research C e n t e r , Products R & D Labora- Type 347 stainless steel for nuclear power 27. Brook, |. A. 1974. Effect of alloy modifi-
tories, N i p p o n Steel C o r p o r a t i o n . applications. Metal Progress 116 (4): 61-65. cations on HAZ cracking of A-286 stainless
9. Hansen, R. P.; Elliott, R. P.; and Shunk, F. steel. Welding lournal 53 (11): 517-s to
References A. 1958, 1965 and 1969. Constitution of binary 523-s.
alloys. 1st and 2nd eds. McGraw-Hill. 28. Groeschel, Richard C. 1968 (Nov.).
1. Savage, W . F., and Lundin, C. D. 1965. 10. Brooks, |. A., and Lambert, F. |. |r. 1978. Kokomo Laboratory rep. no. 7620. Union
The Varestraint test. Welding journal 44 (10): The effect of phosphorus, sulfur and ferrite Carbide Corp.
433-s to 442-s. content on weld cracking of Type 309 stainless 29. Savage, W . F.; Nippes, E. F.; and Miller,
2. Goodwin, G. M. 1968 (June). Ph. D. steel. Welding journal 57 (5): 139-s to 143-s. T. W . 1976. Microsegregation in partially
thesis. Rensselaer Polytechnic Institute. 11. Arata, Y.; Matsuda, F.; Katayama, S. melted regions of 70Cu-30Ni weldments.
3. Hoerl, A., and Moore, Thomas ). 1957. 1977. Solidification crack susceptibility in weld Welding lournal 55 (6): 165-s to 173-s.

WELDING RESEARCH SUPPLEMENT 193-s