Effect of Hydrogen on the Intergranular

Stress Corrosion Cracking of Alloy 600

in High Temperature
Aqueous Environments*

N. TOTSUKA, * E. LUNARSKA, * * G. CRAGNOLINO*"*

and Z. SZKLARSKA-SMIALO WSKA * * * *

Abstract
Slow strain rate tests (SSRTs) on differently shaped tensile specimens machined from two
heats of Alloy 600 tubing were conducted at 350 C in deaerated aqueous solutions con-
taining smal) amounts of lithium hydroxide and boric acid at different partial pressures of
hydrogen, using a recirculating autoclave system. In addition, split half tube specimens cut
from the same two heats of Alloy 600 were subjected to reverse U-bend testing in pure water
and steam at 363 and 400 C, respectively, using two static autoclaves. In both types of tests,
intergranular stress corrosion cracking (IGSCC) was found to occur predominantly at sites in
the alloy where a complex state of stress and strain existed. During the SSRTs, the strained
portions of specimens absorbed much more hydrogen than the unstrained portions. Still
greater contents of hydrogen were found at sites close to the fracture zone. The IGSCC
susceptibility and the crack growth rate both increased with increasing partial pressure of
hydrogen; however, absorption of hydrogen by the alloy also occurred in absence of hydrogen
gas in the system. Results of this study suggest that hydrogen plays an essential role in the
IGSCC mechanism.

Introduction number of cracking incidents in pressurized water reactor

The intergranular stress cROsioni cracking (IGSCC) of Alloy
600 steam generator tubes in primary side environments of
pressurized water reactors has been a matter of increasing
concern in recent years. 1 Since Coriou, et al. reported in 1959
that Alloy 600 and other high nickel alloys were susceptible
to IGSCC in deaerated pure water at temperatures ranging
from 300 to 350 C at relatively high applied tensile stresses, 6
extensive studies in this field have been conducted. Most of
the early laboratory work was reviewed in 1975 by Van Rooyen 7
who took into consideration the mechanistic aspects of the
phenomena on the basis of the information availabie and the
existing stress corrosion cracking (SCC) theories. Recently
published reviews discuss the roles of mechanica) and envi-
ronmental variables at length on the basis of the large amount
of data generated in the last ten years, during which the
*Submitted for publication April 1986; revised September
1986.
plants started to increase substantially. Based on laboratory
work and observations of tubing failures in steam generator
service, some remedial measures have been proposed and
implemented. Especially, prolonged heating at —700 C was
found to have a double beneficial effect on the performance of
Alloy 600 tubes in primary water. This treatment resulted In
the release of residual stresses caused by bending and pro-
duced a coarse-grained microstructure with predominantly
intergranular carbide precipitation. However, many conflicting
resuits have also been obtained.
The main limitation confronted today is the Jack of under-
standing of the crack initiation and propagation mechanisms,
combined with an imprecise definition of the role of environ-
mental variables such as water chemistry, temperature, dis-
solved gases (especially hydrogen), and potential.
In recent years it has been found that the presence of
dissolved hydrogen in pure water increases the susceptibility
to IGSCC of Alloy 600 and accelerates the crack growth
rate. 10 11 However, no precise relationship between concentra-
'

*Corrosion Laboratory, Research Laboratories, Kawasaki tion or partial pressure of hydrogen and IGSCC has been
Steel Corp., I-Kawasaki-Cho, Chiba City, Japan. established to date.
""Institute of Physical Chemistry, Polish Academy of Sci- This study attempts to determine the effect of hydrogen
ences, Warsaw, Poland. on SCC of Alloy 600 in aqueous environments at high tem-
""'Brookhaven National Laboratory, Upton, New York 11973. peratures and to better define the part played in these detri-
****Ohio State University, Dept. of Metallurgical Engineer- mental processes by the localized stress fields and plastic
ing, Columbus, Ohio 43210. strains.
0010-9312/8710002651$3.0010
Vol. 43, No. 8, August 1987 © 1987, National Association of Corrosion Engineers 505
 TABLE 1 — The Chemical Composition (wt%) of Materials Used in This Study

Heat C Ni Cr Fe S P Si Mn Cu Co Al Ti B

1 0.052 72.24 15.37 9.62 0.001 0.010 0.25 0.25 0.25 0.050 0.27 0.23 0.004
II 0.038 75.00 15.81 8.04 0.002 — 0.32 0.26 0.01 0.022 — — —

FIGURE 1 - Micrographs of as•received Alloy 600 tube specimens: (a) and (b). Heat I; (c)
and (d) • electrolytic etch in phosphoric acid; (b) and (d) • electrolytic etch in nital.

Experimental
Materials
Two heats of Alloy 600 tubing, outside diameter (OD)-
22.2 mm, wall thickness (WT) 1.27 mm, were used in this study.
Table 1 gives the chemicai composition of both heats. Heat li
was processed (11 with an intermediate and final annealing at
~930 C and cooling at a rate of 256 Clmin. No information is
available on the thermal treatment of Heat I.
The microstructure of both heats of materials, as revealed
by electrolytic etching in nital and phosphoric acid, showed
( 1 Babcocx Wilcox Go., Alliance, OH.
the average grain size in Heat II to be less than 10µm, whereas
that of Heat 1 was significantly larger at 30 µm (Figure 1). From
the micrographs, the presence of carbide bands in Heat 1 is ap-
parent, accompanied by a noticeable carbide precipitation in
most of the grain boundaries. On the other hand, predominant
intragranular carbide precipitation occurs at random in Heat
II.
SSRTs
Test specimens. The mechanica) properties of both heats
of material as measured at room temperature (RT) are listed in
Table 2. Split tube tensile specimens shown in Figure 2 were
machined from both heats of Alloy 600 tubing. Type A-2

506 CORROSION—NACE
 TABLE 2 — Mechanica) Proporties as Measured The solution used in all SSRTs was 0.01 mol boric acid
at Room Temperature plus 0.001 mol lithium hydroxide per 1 kg water. This solution
is used to simulate the prevailing environment in the primary
side of pressured water reactors.
Heet 0.2% Proot Stress Ultimate Tensiie Strength Elongation Source
The tests were conducted under open circuit conditions
at varying pressures of hydrogen and strain rates. Two kinds
290 707 44.0 1(11 of purging gas were used for maintaining a deaerated environ-
327 706 34.6b 2
ment and obtaining a constant partial pressure of hydrogen.
389 737 37.0' One was a mixture of 95% nitrogen with 5% hydrogen, and the
390 728 37.1b other was pure hydrogen. To maintain a single liquid phase in
the loop over the temperature range from 300 to 350 C, the
total pressure was set at 11.4 to 20.7 MPa, by using an ap-
11) Legend: propriate back pressure regulator. The solution recirculation
1—as supplied by J. A. Jones Applied Research Co.
2—as measured on split tube tensile specimens at a strain rate of 0.1 mm/min. rate was —3.5 LIh.
3—as supplied by Babcock & Wilcox Co. After a tensile specimen was mounted inside the auto-
a—measured on 25.4 mm gage length specimens. clave, the solution contained in a storage tank was deaerated
b—measured on 20 mm gage length specimens.
by purging gas for ~24 h to obtain an oxygen concentration
less than 10 ppb. A high pressure PTFE diaphragm pump was
activated, the pressure was set via the back pressure
\6350
regulator, and the system was then heated to the desired tem-
6350 yR
peratu re.
Straining was started after steady temperature and
+ 20 — J 2° F- o pressure were obtained. During tests, both the potential and
load were recorded, except in the first two tests where type A-1
22000 specimens were used. A strain rate of 5.0 x 10 -7/s was used,
127 1 híckness
and an initial load of 440 N was applied before straining was
Type 4:I Specimen started. In one test, the specimen was first pulled at a rate of
8.0 x 10 -5/s until an elongation of — 15% was attained, and
A then straining was continued at a rate of 2.0 x 10 -7 1s.
6350+ 6.352
9R M
/ r.
-. .. _
`y^
-.-
Jn
. IS Constant Deformation Tests.
L7i F — '1' 1-_y Ii .•R Reverse U-bend test specimens have been used by many
-9.
researchers l ' 1016 because it is believed that the stress state
existing in these specimens better simulates that prevailing in
steam generator tubes than the common U-bend test with f lat-
tened specimens. It was expected that the use of this testing
b method, supplemented by the determination of hydrogen con-
Type A-2 Specimen tents in the alloy after various exposure times, could provide a
useful approach to the assessment of the role of hydrogen in
B the IGSCC mechanism.
C6.35m ^6.3.`^N i I n Test Specimens. Type B-1 specimens (B-1 Type) (Figure 3)
9R ^
^
were machined from the same tubings used for SSRT. Split
15 M
-10 sections of Alloy 600 tubing --10 cm long were bent using a
I'^ 20^i 20 ^3
`
p
2.54-cm diameter mandrel. No special precautions were used
to avoid stress relaxation before inserting the bolt and the nut
Type A-3 Specimen necessary to maintain a constant 2.54-cm separation between
the straight legs. The specimens were tested after careful
FIGURE 2 — Specimens used in SSRTs. Al] dimensions degreasing and cleaning.
are in mm. Type B-2 specimens were in the form of rings, 10 mm wide.
The specimens were tested after being flattened by 70% OD
reduction. The same type of specimen was used successfully
by Matsushima 17 to evaluate IGSCC of sensitized austenitic
specimens were made Erom Type A-1 specimens by pressing SS in polythionic acid at RT.
their middle part with a special dye. The elongation of all Type B-3 specimens were 3-mm wide strips cut parallel to
specimens was measured on a gage length of 20 mm. The the tube length. They were tested after mounting them (shown
gage length of Type A-2 specimens was marked before press- in Figure 4) into a holder made of Hastelloy C-276. 121 The inner
ing. For preliminary experiments, a double-notched Type A-3 side of the tube was under extension. The length of specimens
specimen, also shown in Figure 2, was used. was either 43 or 50 mm to obtain a different deflection (y) in the
Equipment. The specimens for the SSRTs were set in a 1 L °
center, and thus a different strain (E ) in the extreme fiber, ac-
AISI 316 stainless steel (SS) autoclave connected with a slow cording to the formula:
strain rate tensile machine. The specimens were isolated from
the grips and the pull rod by ceramic pieces of zirconia. The r2by
pull rod was electrically isolated from the autoclave by
locating a conax fitting with polytetrafluoroethylene (PTFE)
EO 2L2 (1)
seal well above the heat of the autoclave at the end of a seg-
where L is the length and b is the thickness of the specimen
ment of AISI 316 SS pipe. A coil made from copper tube and (1.2 mm). For L = 43 mm, y = 7 mm, e, = 2.2%, and for L = 50
brazed to the pipe was used to circulate water for cooling, °
mm, y = 13 mm, e = 3.0%. In both cases, the deformation in
avoiding the high temperature creep of the PTFE seal. A Pt foil the extreme fiber exceeded the elastic range limit of both
welded to a Pt wire served as reference electrode. An Alloy 600 materials, which was --1.7% at RT.
wire was welded to the specimen for electrical connection. For Equipment. The experimental set-up used for the above
the sake of electrical isolation, all connecting wires were
covered with alumina or zirconia tubes. 12 )Registered trade name.

Vol. 43, No. 8, August 1987 507

 ^— 30 —^-I 9.8R Pressure
Gauge f Ga s e re
Thermocouples 9

Q N N
1?
Distilled Valve VaNe Valve Valve
Water
Tank Feed
Pump
N ----- Vacuum
Furnoce -_ -- Furnoce Pump

i;. 1.27 Thickness Autoclave 1 Autoclave 2

Washer
k10 Nut
FIGURE 4 — Schematic representation of the apparatus
6.350 Bolt
used for constant deflection tests.
Type 8-! Specimen
Heat I Heat II
O • 400°C Steam
U) 100
N
-

--
N

+

h--- 33 ---1 127Thickness

^^ c
E
U
G)
Q
`

80 I ❑ n 363°C Water
() Number of Specimens
Tested
IOOhr. Exposure •(6)
Before Being Flattened After Being Flattened (1) 60 -

•4)
Type 8-2. Specimen
v
a) 40 -

L- n (4)
20
0 •(6) •(6)
98R y0

^-- 50 or 43 (6)@7(4)
O (6) (4)_ (4)(4)
1.27 Thickness 55 _ I
40
0 0.1 0.2 0.3
Type 8-3 Specimen and Ho/der PH2 (MPa)
FIGURE 5 — Resuits of reverse U-bend tests lasting 100
FIGURE 3 — Specimens used in constant deformation h (B-1 specimens).
tests. All dimensions are in mm.

tests is shown schematically in Figure 4. The specimens were
mounted in a SS holder, which was isolated with alumina
tubes. After setting specimens in the autoclave, the system
was evacuated by using a vacuum pump, and high purity water
(<0.3 µS/cm, Vol. 700 mL) was injected into the first autoclave.
In addition, hydrogen gas at various partial pressures could be
introduced into the autoclave system. The partial pressure of
hydrogen was varied from 0 to 0.27 MPa, corresponding to a
concentration of 0 to 43.0 cm 3 H 2 STP/kg H 2 O. The water in the
first autociave was heated to 363 C in equilibrium with steam
at a pressure of 19.3 MPa. Steam was then transferred to the
second autociave and overheated to 400 C at the same
pressure. By locating specimens in the liquid and steam phase
in the first autociave and in the steam phase in the second
autociave, comparative tests could be made. After each test
was completed, the specimens were removed from the auto-
claves and examined visually and by scanning electron micro-
scopy (SEM).
Hydrogen Determination
The total amount of hydrogen was determined in samples
cut from test specimens using a LECO( 3) Hydrogen Deter-
minator by heating the sample to ~ 1500 C. The specimens
were cut with a diamond saw into smaller samples to deter-
mine the hydrogen content in different parts of specimens. An
average time of 7 h elapsed between switching off the auto-
clave heating and performing hydrogen determination. Before
measurements, the samples were washed with acetone, using
an ultrasonic cleaner. To check if the result of hydrogen deter-
mination was affected by the presence of a film of corrosion
products, in some experiments, the film was removed from the
sample surface by polishing it with 600 grit SiC paper. Mean
values obtained in at least two measurements on samples
taken from adjacent parts of specimens were considered.
Fractography
Examination of the fracture surface and side surface of
each studied specimen was performed using a Hitachi S-510 (4 ►
scanning electron microscope. Before examination, the speci-
men was washed with acetone in an ultrasonic cleaner and
dried in warm air. In some cases, the corrosion products were
removed by pickling in dilute hydrochloric acid with 2 mg/L
hexamethylene tetraamine as corrosion inhibitor, followed by
ultrasonic cleaning.
Resuits
SSRTs
A summary of the SSRTs is given in Table 3, showing that
A-1 specimens revealed ductile fracture at both strain rates
used, and this was accompanied by a large elongation to frac-
ture (--30%), whereas A-2 specimens, specifically those tested
at a hydrogen partial pressure of 0.1 MPa, failed by IGSCC. A
notched A-3 specimen of Heat II, which was tested under the
same experimental conditions as A-2 specimens, did not
reveal IGSCC.
Constant Deformation Tests
Results of reverse U-bend tests (B-1) conducted in high
temperature water and steam are summarized in Figures 5 and
6, where the percentage of specimens failed in 100 and 250 h,
respectively, is related to the total number of specimens
tested at different partial pressures of hydrogen. Results in
Figure 5 indicate that the IGSCC susceptibility of Heat II is

(3) Registered trade name. (4)Hitachi, Tokyo, Japan.

508 CORROSION—NACE


Heat I Heat II
O • 400°C Steam
v0 o (6)CM(4)
r 363°C Water

E_ g 0 I () Numberof Specimens •(6) •(6)

U
Tested 0(4)
Q 250hr Exposure

(4)-(4)
_ 40
^ •(5)
a •(4)
li
20

0 (2)
O(6)
(3)
0.1 0.2 0.3
PH2 (MPa)
FIGURE 6 — Resuits of reverse U-bend tests lasting 250
h (B-1 specimens).

FIGURE 7 — Fracture surface of a type Al specimen

tested In air.

much higher than that of Heat I. Because of the scatter of ex-

perimental data obtained at longer times of exposure, resuits
shown in Figure 6 are less indicative than those in Figure 5,
but the greater resistance to IGSCC of Heat 1 over Heat II is
still apparent. After 400 h of exposure, ~ 50% of Heat II speci-
mens tested in the absence of hydrogen showed IGSCC, and
almost all specimens failed when some amount of H 2 was
present.
These results indicate that both heats of Alloy 600,
though in a different degree, are susceptible to IGSCC in the
absence of dissolved hydrogen in high temperature water, and
that increased contents of hydrogen cause IGSCC to increase
substantially. The plastically cold-deformed rings (B-2) did not
crack on exposure for 500 h to hot water or steam in the pres-
ence of up to 0.27 MPa hydrogen. In tests conducted under the
same experimental conditions described, B-3 specimens de-
formed to 2.2% in the outer fiber did not exhibit any traces of
cracking, while those deformed to 3.3% showed slight inter-
granular attack (IGA) on the most strained specimen zone.
Fractography
SSRT Specimens. Figure 7 shows the fracture surface of
a Type Al specimen of Heat 1 after tensile testing at RT in air.
A fully ductile fracture was observed with uniformly distrib-
FIGURE 8 (a) and (b) — Fracture surface of a Type A-2
specimen tested in air.
uted shear dimples visible at higher magnification. The same
type of fracture was found for Heat II specimens tested in air
and also for both the Heat 1 and Heat II specimens tested in
boric acid + lithium hydroxide solution at 350 C.
All A-2 specimens failed along the plane marked a-a' in
Figure 2, i.e., across the most deformed portion of the speci-
men. SEM fractographs shown in Figures 8(a) and (b), which
were taken for the fracture surface of a Heat II specimen after
testing in air at RT, exhibit a ductile pattern similar to that ob-
tained for Al specimens (Figure 7), but with dimples
elongated and coalesced along the specimen width, especially
in the center part of the specimen and at sites where the edge
of the dye was pressed during cold-deformation. In contrast,
A-2 specimens tested in H 3 B03 + LiOH + H 2 showed some
percentage of intergranular cracks distinguishabie on the frac-
ture surface. Table 3 indicates the percentage of intergranular
fracture found on the fracture surfaces of A-2 specimens of
Heat 1 and Heat li after exposure to the above solution at two
different partial pressures of hydrogen, showing that Heat II

Vol. 43, No. 8, August 1987 509

 FIGURE 9 — Fracture surface of a type A-2 Heat 1 speci-
men tested at 350 C in the presence of 0.1 MPa H 2 .

exhibited larger areas of intergranular fracture than Heat 1 and

that a higher partial pressure of hydrogen resulted in a higher
percentage of intergranular fracture.
Figure 9 depicts the fracture surface of a Type A-2 speci-
men after testing in the solution at 350 C, 0.1 MPa H 2 , showing FIGURE 10 — Intergranular fracture of type A-2 speci-
slight intergranular cracking (IGC). The fracture always mens of Heat 1 (a) and Heat 11(b) tested at 350 C in the
originated at site a' (Figure 2), where a state of highest cold- presence of 0.1 MPa H 2.

work existed. Some grains elongated normal to the fracture

surface, with numerous slip steps on the grain boundary inter-
faces, were observed in Heat 1 specimens (marked by arrows in
Figure 10(a), whereas Heat II specimens revealed finer and
nearly equiaxial grains with almost no slip steps apparent
[Figure 10(b)].
Figure 11 shows at higher magnification a region near the
tip of a crack (A). Intergranular voids (B) ahead of the crack tip
are present. From the shape of the intergranular crack (A) and
from marks on the grain boundary interface (C), which reveal
sites of the final rupture, it follows that the intergranular crack
has propagated by growth and coalescence of intergranular
voids.
Quite often, on the fracture surface of specimens, an in-
termediate portion of modified fracture (marked by arrows in
Figure 9) occurred. This fracture was characterized by deep
ditches situated along the specimen width (marked A in
Figures 12(a), with elongated and coalesced dimples on their
bottom, and with some portions of transgranular, cleavage-
like fracture [marked C in Figure 12(b)]. The extent of such a
type of intermediate fracture was greater on the fracture sur-
face of specimens tested at 0.005 MPa than at 0.1 MPa H 2 , and FIGURE 11 — Fracture surface of a type A-2 Heat 1
the same type of fracture also occurred in specimens that did specimen tested at 350 C in the presence of 0.1 MPa H 2:

not show IGC. A. intergranular crack, B- intergranular voids, C- marks

Typical intergranular cracked side surfaces of two A-2
specimens of Heat II tested in 350 C, 0.1 MPa H 2 , are shown in
Figure 13. Deep cracks parallel to the fracture surface [originat-
ing at a' in Figure 2(a)] are shown. Some IG cracks also oc-
curred naar the plane marked b'-b (these cracks originated at
point b marked in Figure 2). Numerous smal) intergranular
cracks shown in Figures 14(a) and (b) were revealed by suc-
on grain boundaries.
cessive polishing of the longitudinal section of a cracked
specimen. These cracks were situated far ahead of the main
cracks, and there was no evidence that interconnection be-
tween them existed.

510 CORROSION-NACE
 FIGURE 12 (a) and (b) — Modified type of fracture on a
type A-2 Heat 1 specimen fractured in the presence of
0.005 MPa H 2 . A. ditches, B - elongated dimples, C.
transgranular quasi-cleavage.

FIGURE 14 — Cracks on the longitudinal section of a

type A-2 Heat II specimen. Separate microcracks are
marked by arrows.

measured on the fracture surface and the time to failure of the

FIGURE 13 — Side surface of a type A-2 Heat II speci- given specimen. From these estimates, it follows that the
men tested in the presence of 0.1 MPa H 2 . crack propagation rate is greater in the material of Heat II than
in that of Heat 1 and increases as the partial pressure of hydro-
gen gas is increased.
Crack Growth Rate Estimates. The last column of Table 3 Constant Deformation Test Specimens. When tested in
contains data on crack growth rates estimated for A-2 speci- water or steam at high temperatures, all B-1 specimens ex-
mens exhibiting intergranular fracture. Assuming that the time hibited a fully intergranular fracture. Two typical examples are
necessary to initiate cracks was negligible, the crack growth given in Figures 15(a) and (b). These fracture surfaces were
rate could be calculated by considering the IG crack depth covered with corrosion products (white in photographs), the

Vol. 43, No. 8, August 1987 511


FIGURE 15 — Fracture surf ace of a type B-1 Heat II
specimen tested in steam (a) and in water (b).
amount of which was greater in the case of testing in steam
[Figure 15(a)] than in water (Figure 15(b)].
Note that many cracks were longitudinal, i.e., situated
along the axis of the tube, indicating that considerable hoop
stresses had developed during the split tube bending. The
cracks deviated from the longitudinal direction after some
growth and continued propagation at various angles, with
respect to the tube axis. However, in many cases the longiduti-
nal path was resumed after propagation over a short distance
in the transverse direction.
Hydrogen Content
The content of hydrogen was 1.8 and 2.2 ppm in the as-
received Heat 1 and Heat II material, respectively. These
residual amounts of hydrogen were subtracted from those
found in specimens after various SCC tests. Polishing with 600
grit SiC paper to remove corrosion products from the
specimen surface, did not affect the hydrogen determination
resuits, indicating that no or only negligible quantities of
hydroxides on the alloy surface were present.
SSRT Specimens. The contents of total hydrogen found in
samples taken from different parts of SSRT specimens
(marked 1 through 5 in Figure 16) are given in Table 3. Because
these results are based on single measurements, they are
highly scattered, nevertheless indicative. In the unstrained
portions of specimens (sites 4 and 5 in Figure 16), the
512 CORROSION—NACE
E
a
n
a)
0
U
c
a>
0
`+ J L
Part of the Specimen
FIGURE 16 — Hydrogen contents found in different
parts of A-2 Heat II specimens.
hydrogen content increases with time, achieving a steady-
state value of —12 ppm after 100 h of exposure for both heats
and partial pressures of environmental hydrogen. In the
strained portion, 2 the hydrogen content always is much higher
than in the unstrained one, and mostly higher at 0.1 than at
0.005 MPa of H 2 partial pressure. The highest hydrogen con-
tents occur close to the fracture. )
Constant Deformation Test Specimens. The average con-
tent of hydrogen found in B-1 specimens (reverse U-bend tests)
was approximately the same, 8 to 14 ppm, irrespective of the
material, temperature, H 2 pressure, and specimen portion from
which a sample for hydrogen determination was taken. How-
ever, as a rule, an exposure to hot water or steam for 100 h
resulted in a significantly greater hydrogen content, relative to
longer exposures.
In B-2 and B-3 specimens exposed to hot water or steam
for 240 and 540 h, the mean hydrogen content was —13 and 20
ppm, respectively. No discernible influence of the kind of
material or degree of cold-deformation on the hydrogen con-
tent was observed.
Discussion
Microstructure
As shown in Figure 1, the material of Heat II is fine-
grained and free of intergranular carbide precipitates, whereas
that of Heat 1 is coarse-grained, with numerous carbide par-
ticles in the grain boundaries. On the other hand, the above
resuits of SSRT and reverse U-bend tests indicate that Heat II
is much more susceptible to IGSCC than is Heat I. This agrees
with observations of other authors who found coarse-grained
Alloy 600 (<ASTM 6)( 5) Iess susceptible to IGSCC in high
temperature water when compared with fine-grained alloy
(>ASTM 8).(6 ) 5,13 It has also been reported 5 that significant
I 5),(6 IASTM,
Philadelphia, Pennsylvania, 4-6 Kandasurugadai
Chiyodaku Tokyo, Japan.
precipitation of chromium carbides in grain boundaries
characterizes materials more resistant to IGSCC in com-
parison with those containing relatively clean grain bound-
aries and predominantly intragranular carbide precipitates. 13
The effect of microstructure on IGSCC of Alloy 600 is
not well understood. Bruemmer and Henager assume that
grain boundary carbides act as Iow energy dislocation
sources, produce more homogeneous plastic deformation,
and modify the local stress state. 18 The presence of a semi-
continuous array of intergranular carbides may continuously
blunt the crack tip and reduce the SCC crack propagation.
Stress and Strain
From the results obtained, it follows that intergranular
cracks primarily initiate and develop at locations in the alloy
where a complex state of high stress and strain exists. This
conclusion is also in agreement with literature data. For exam-
ple, Norring reported that materials susceptible to IGSCC in
the reverse U-bend tests did not fail in SSRTs conducted in
high temperature water, even at very low tensile strain rates
(-1.0 x 10 -7 1s), and with H 2 present (2.5 to 3.0 ppm). 19
According to Norring, a complex state of stress (biaxial or
triaxial) is necessary to induce IGSCC of Alloy 600 in high
temperature water.
In type A-1 specimens used in the present study, no
IGSCC occurred at strain rates of 5.0 x 10 -7 /s or 2.0 x
10 -7 /s. In contrast, under the same experimental conditions,
type A-2 specimens suffered IGSCC in proximity of the hump,
where a complex state of stress and strain existed. However,
no IGSCC occurred on the notched A-3 specimen, in spite of
the development of triaxial stresses at the notch bottom dur-
ing straining. This suggests that local cold-work enhances
IGSCC, which also is consistent with previous findings. 11
In terms of the constant deformation tests, only B-1 speci-
mens (i.e., those subjected to reverse U-bend tests) suffered
severe IGSCC in high temperature water and steam. These
specimens were in a state of intense cold-deformation and
very high localized stresses. The B-2 specimens (ring speci-
mens), which were also predeformed, but with a lesser amount
of cold-work, did not show IGSCC. The B-3 specimens, which
were subjected to a lower strain (43 mm in length) did not suf-
fer IGSCC, while Jonger ones (50 mm in length) showed slight
IGA in the outermost fibers.
These results indicate that the susceptibility to IGSCC of
Alloy 600 increases with increasing intensity and complexi-
ty of stress fields, and especially with increasing amount of
localized cold-work.
Induction Time for IGSCC Initiation
The results of reverse U-bend tests [Figures 5(a) and (b)]
show that the induction time for crack initiation in water and
steam in the presence of gaseous hydrogen is less than 100 h
in Heat II and from 100 to 250 h in Heat 1. In the absence of
hydrogen in water, the induction time for crack initiation in
Heat II is longer than 250 h. These results indicate that the in-
duction time for IGSCC initiation depends on the alloy quality
(most probably its microstructure) and the partial pressure of
hydrogen gas in the water system. More accurate data con-
cerning induction time could be obtained on the basis of inter-
rupted SSRTs, but this type of experiment was not performed
in the present study.
Crack Growth Rates
The crack growth rates were only roughly estimated in
this work. Mean values obtained from SSRTs conducted in
H 3 B0 3 + LiOH with 0.1 MPa H 2 were 1.4 x 10 -9 and 3.8 x
10 -9 m/s for Heat 1 and Heat II, respectively (Table 3). At a
lower H 2 pressure, smaller values were observed. As men-
tioned previously, these estimates were made without con-
sidering the induction time for crack initiation, which certainly
was a rough simplification.
The average crack growth rates based on reversed U-bend
tests were 'estimated by considering only the length of
through-wall cracks and neglecting cracks propagating
lengthwise. The value obtained this way was 2.6 x 10 -9 mis,
hence in the same order of magnitude as in the SSRT.
Hydrogen Absorption
The hydrogen contents found in the unstrained portions
of A-1 and A-2 specimens (Table 3) and in the strained and
unstrained parts of B-1 specimens exposed for times Jonger
than 100 h were independent of the partial pressure of H 2 in all
the environments studied. These results suggest that: (1)
hydrogen is absorbed by the alloy not only from the water-
dissolved hydrogen gas, but from water molecules as well,
which probbaly occurs via their dissociative adsorption on the
alloy surface; 20 and (2) after exposure for 100 h, a steady state
is established between the H adsorbed on the surface and that
dissolved in the alloy.

TABLE 3 - Results of Slow Strain Rate Tests

at 350 C in H 3 B0 3 + LiOH + H 2 Solution

arl) b c d e f 9 h i 1 K r m

A-1 II 5.0 0.1 - - 343.0 29.0 54.1 13.1 6.8 0 -

A-1 II 2.0 0.1 - - 408.0 29.9 82.4 43.2 20.1 0 -

A-2 1 5.0 0.1 +14 +3 99.7 9.6 34.7 33.0 13.0 42.4 1.44
A-2 1 5.0 0.1 +3 +2 101.8 10.3 28.4 29.4 12.8 40.0 1.6
A-2 1 5.0 0.005 -14 + 150 182.4 17.8 27.1 23.1 12.2 1.0 -
A-2 1 5.0 0.005 -35 +50 132.3 12.3 65.8 17.2 9.9 5.6 0.5

A-2 II 5.0 0.1 -22 +3 63.5 6.3 27.1 22.4 10.5 60.0 2.9
A-2 II 5.0 0.1 -18 +3 60.5 3.5 65.8 29.0 12.8 67.8 3.5
A-2 II 5.0 0.1 -3 +3 60.5 5.7 83.0 29.2 9.1 66.3 3.3
A-2 II 5.0 0.005 -30 +114 79.3 7.6 62.4 17.8 8.0 20.8 1.05
A-2 II 5.0 0.005 +40 +250 125 9.8 21.1 16.2 11.3 0 -
A-2 II 5.0 0.005 0 -10 46.5 4.6 20.3 18.5 10.2 12.3 1.6

A-3 It 5.0 0.1 - - 80.5 34.6 10 - 0 -

" ) Legend: 9-time to fracture, hours

a-specimen type
b-Heat
c-straining rate x 1071s
d-partial pressure of H 2 , MPa
e-potential at start of test, mVp,
f-potential at end of test, mVP,
h-elongation at fracture. o/
i-hydrogen content in part 1 of specimens, ppm; (Figure 16)
j-hydrogen content in part 2 of specimens, ppm
k-hydrogen content in part 4 of specimens, ppm
I-IGSCC area on the fracture surf ace.
m-crack growth race. x 10 9 mis

Vol. 43, No. 8, August 1987 513

 In A-2 specimens, the content of hydrogen increases with
the amount of deformation (Figure 16), namely, in the following
order, respective of the different parts of the specimen:
holder<intermediate part between holder and gage length<
uniformly deformed part<hump region. It is welt-known that
plastic deformation resuits in the formation of voids in the
metal in which hydrogen can accumulate in molecular form. In
further stages of straining, it also can readsorb on the internal
surface of voids and diffuse in atomic form to energetically
more favorable locations in the piece of metal or alloy, for ex-
ample to sites were a state of triaxial stress becomes
operative, such as a crack tip.
Straining of A-1 and A-2 specimens during SSRTs at high
temperature caused the absorption of hydrogen by the
strained alloy to be much more pronounced at 0.1 MPa than at
0.005 MPa of H 2 in the environment. Despite the high concen-
tration of hydrogen in A-1 and A-3 specimens after SSRT, no
IGSCC occurred. Thus, it is obvious that a high hydrogen con-
tent does not suffice to cause cracking if no complex state of
stress or cold-worked zone in the specimen is present.
The reverse U-bend test specimens showed no remark-
able differences in the hydrogen content between the bent
(strained) and straight (unstrained) portions. This finding is
presently hard to explain. However, it should be noted that
because all measurements of hydrogen contents were made
after at least 7 h after fracture, the actual hydrogen concentra-
tion at the moment of the cracking event, especially in the
crack tip zone, could be much higher. On the other hand, the
hydrogen determination method used in this study (by melting
the alloy) measured the total amount of hydrogen, independ-
ent of its form, i.e., H, H 2 , CH 4 , H 2 S, etc.
Open Circuit Potential
Changes in the open circuit potential (OCP) of A-2 speci-
mens during SSRTs (Table 3) proved to be related to IGSCC.
For specimens suffering IGSCC, a relatively constant negative
potential was observed, whereas for those with ductile frac-
ture or with only a slight area of intergranular cracking, more
positive and increasing with time potential values were re-
corded. This suggests that the IGSCC of Alloy 600 is con-
trolled by a cathodic process. The effect of potential on IGSCC
deserves to be studied in more detail.
The results obtained in this study strongly suggest that,
in the system studied, hydrogen is responsible for IGSCC at
sites where complex states of stress and strain are present.
However, it would be premature to propose an exact mecha-
nism for this combined action.
Conciusions
1. IGSCC of Alloy 600 in high temperature water and
stearn is caused by a joint action of hydrogen and complex
states of intense stresses and strains.
2. The crack growth rate increases with increasing partial
pressure of H 2 in the aqueous environment.
3. The intergranular degradation of Alloy 600 is par-
ticularly severe in heavily cold-worked zones, where intense,
complex stresses exist.
Acknowledgment
This work was partly supported by Steam Generator
Owners Group and the Electric Power Research Institute. The
514 CORROSION—NACE
authors also wish to thank Dr. A. Mcllree for helpful discus-
sions.
References
1. EPRI Workshop Proc.: U-Bend Tube Cracking in Steam
Generators, C. E. Shoemaker, Ed., EPRI WS-80.136, Electric
Power Research Institute, Palo Alto, California, 1981.
2. S. J. Green, J. P. Paine, NuclearTechnology, Vol. 55, No. 10,
p. 10, 1981.
3. S. J. Green, J. P. Paine, Proc. Int. Symp. on Environmental
Degradation of Materials in Nuclear Power Systems-
Water Reactors, National Association of Corrosion Engi-
neers, Houston, Texas, p. 53, 1984.
4. EPRI Workshop Proc.: Primary Side SCC in Steam Gener-
ators, Electric Power Research Institute, Palo Alto, Cali-
fornia, March 1985.
5. J. A. Gorman, E. S. Hunt, A. R. Mcllree, Proc. 2nd Int. Symp.
on Environmental Degradation of Materials in Nuclear
Power Systems—Water Reactors, National Association of
Corrosion Engineers, Houston, Texas, p. 557, September
1985.
6. H. Coriou, L. Grall, Y. Le Gall, S. Vettier, 3-eme Colloque
de Metallurgie sur la Corrosion, Saclay, North Holland
Publ. Co., Amsterdam, Holland, p. 161, 1959.
7. D. Van Rooyen, Corrosion, Vol. 31, No. 9, p. 327, 1975.
8. E. Serra, Stress Corrosion Cracking of Alloy 600, EPRI
NP-2114 SR, Electric Power Research Institute, Palo Aito,
California, November 1981.
9. Y. S. Garud, T. L. Gerber, Intergranular SCC of Ni-Cr-Fe
Alloy 600 Tubes in PWR Primary Water—Review and As-
sessment for Model Development, EPRI NP-3057, Electric
Power Research Institute, Palo Alto, California, May 1983.
10. G. P. Airey, Optimization of Metallurgical Variables to
Improve Corrosion Resistance of Inconel Alloy 600, EPRI
NP-3051, Electric Power Research Institute, Palo Alto,
California, July 1983.
11. R. Bandy, D. Van Rooyen, Corrosion, Vol. 40, No. 8, p. 425,
1984.
12. D. Van Rooyen, EPRI Workshop Proc.: Primary Side SCC
in Steam Generators, Electric Power Research Institute,
Palo Alto, California, March 1985.
13. P. Saint Paul, EPRI Workshop Proc.: Primary Side SCC
in Steam Generators, Electric Power Research Institute,
Palo Alto, California, March 1985.
14. R. Bandy, D. Van Rooyen, Proc. 9th Int. Congress on Metal-
lic Corrosion, Ottawa, Canada, Vol. 2, p. 202, 1984.
15. T. S. Bulischek, D. Van Rooyen, Corrosion, Vol. 37, No. 10,
p. 597, 1981.
16. T. S. Bulischek, D. Van Rooyen, Nuclear Technology, Vol.
55, p. 383, 1981.
17. I. Matsushima, Corrosion Eng. (Japan), Vol. 22, p. 141,
1973.
18. S. M. Bruemmer, C. H. Henager, Jr., Proc. 2nd Int. Symp.
on Environmental Degradation of Materials in Nuclear
Power Systems—Water Reactors, National Association of
Corrosion Engineers, Houston, Texas, p. 293, September
1985.
19. K. Norring, Slow Strain Rate Testing of Steam Generator
Tube Materials in High Purity Water, STUDSVIK/EI.85119,
Studsvik, Energiteknik AB, January 1985, p. 12.
20. P. Marcus, J. Oudard, Hydrogen Degradation of Ferrous
Alloys, R. A. Oriani, Y. P. Hirth, M. Smialowski, Eds., Noyes
Publications, p. 36, 1985.