5.08

Irradiation Assisted Stress Corrosion Cracking

P. L. Andresen
GE Global Research Center, Schenectady, NY, USA

G. S. Was
University of Michigan, Ann Arbor, MI, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.08.1

Introduction

177

5.08.2
5.08.2.1
5.08.2.2
5.08.3
5.08.3.1
5.08.3.2
5.08.4
5.08.4.1
5.08.4.2
5.08.4.2.1
5.08.4.2.2
5.08.4.2.3
5.08.4.3

5.08.5
References

Irradiation Effects on SCC: Laboratory and Plant Data
Individual Effects of Radiation on IASCC
Service Experience
Irradiation Effects on Water Chemistry
Radiolysis and Its Effect on Corrosion Potential
Effects of Corrosion Potential on IASCC
Irradiation Effects on Microchemistry and Microstructure
Radiation-Induced Segregation
Microstructure, Radiation Hardening, and Deformation
Irradiated microstructure
Radiation hardening
Deformation mode
Radiation Creep and Stress Relaxation
Summary

180
180
183
187
187
189
190
190
194
194
196
198

201
202
202

Abbreviations
AES
AGR
BWR
CT
CW
DPA
FEGSTEM
FWHM
HWC
HWR
IASCC
IGSCC
NWC
PWR
RIS
RH
SCC
SFE
SGHWR
SS
SSRT
STEM

Auger electron spectroscopy
Advanced gas cooled-reactor

Boiling water reactor
Compact type (specimen)
Cold work
Displacements per atom
Field emission gun TEM
Full-width – half-max (for profiles)
Hydrogen water chemistry (in BWRs)
Heavy water reactor
Irradiation-assisted stress corrosion
cracking
Intergranular stress corrosion cracking
Normal water chemistry (in BWRs)
Pressurized water reactor
Radiation-induced segregation
Radiation hardening
Stress corrosion cracking
Stacking fault energy
Steam generating heavy water reactor
Stainless steel
Slow strain rate test
Scanning transmission electron
microscopy

TEM
TG

Transmission electron microscopy
Transgranular

5.08.1 Introduction

Nuclear power accounts for about 17% of the world’s
electricity production, and the rapid expansion in
nuclear power throughout the world will necessitate
that they operate with high reliability and safety.
Stress corrosion cracking (SCC) has occurred in all
water cooled reactors, including boiling-water reactors (BWRs) and pressurized-water reactors (PWRs),
with a greater incidence in unirradiated, out-of-core
components, especially between 1970 and 1990. As
these materials, component designs, and water chemistries have improved, an increasing percentage of
cracking problems has occurred in irradiated components. While irradiation-assisted stress corrosion
cracking (IASCC) has been observed since early
plant operation, increasing operating time and fluence has led to an increased incidence of cracking.
Setting aside zircaloy fuel cladding and pressure
vessel steels, most irradiated core components consist
of austenitic stainless steels and nickel-base alloys
177


178

Irradiation Assisted Stress Corrosion Cracking

exposed to environments that span oxygenated to
hydrogenated water at $270–340  C. The core of a
nuclear reactor is an extreme environment consisting
of high-temperature water, imposed stresses and
strains, and an intense radiation field that affects
the water chemistry, stress, and microstructure of
the core materials (Figure 1). For background, the
reader is referred to Chapter 1.03, RadiationInduced Effects on Microstructure; Chapter 1.05,

Radiation-Induced Effects on Material Properties
of Ceramics (Mechanical and Dimensional), and
Chapter 1.07, Radiation Damage Using Ion Beams
for more detailed treatments of radiation effects on
materials. Initially, the affected components were primarily small components (bolts, springs, etc.) or components designed for replacement (fuel rods, control
blades, or instrumentation tubes). However, in the
last $20 years, IASCC has been observed in structural
components (e.g., PWR baffle bolts and BWR core
shrouds and top guides).
Extensive literature exists for SCC under unirradiated conditions, and the basic factors and dependencies are well defined and reasonably well
modeled for austenitic stainless steel and nickel alloys
(e.g., Alloys 600 and its weld metals).1–9 A complete
consensus on the underlying mechanism of cracking
has not emerged although the well-behaved continuum in crack growth rate response versus material/
composition (including from stainless steels to nickel
Solution renewal
rate to crack-tip
Stress

Δf Anionic
transport

Oxide rupture
rate at
crack-tip

Environment

Microstructure


g-field
Crack tip f [A]–, pH
Passivation rate
at crack-tip
Grain boundary denudation

Hardening
Relaxation

N-fluence
Segregation

Figure 1 Schematic of the primary engineering
parameters that effect stress corrosion cracking – stress,
microstructure, and environment – and the underlying crack
tip processes that control stress corrosion cracking. The
primary ways in which radiation affects stress corrosion
cracking is also shown: segregation, hardening, relaxation,
and radiolysis. Radiolysis can increase the corrosion
potential, which in turn increases the potential gradient
(Á’) and the crack tip potential ’, anion concentration
[A], and pH.

alloys), water chemistry, temperature, and radiation
suggests that a common crack growth mechanism is
operative.8–13
Our understanding has evolved from the view that
SCC occurs under very specific and unique conditions
to the view that a continuum in response exists.8–13 With
steady improvement in laboratory and plant detection

of SCC, it is clear that SCC occurs under a wide range
of conditions and also at a wide range of growth rates.
Figures 2 and 3 show examples of the effect of environment (corrosion potential and water purity), material
condition (sensitized vs. cold-worked), and stress (stress
intensity factor) on SCC growth rates; the solid curves
are the predicted response.8,9,12,13 SCC occurs even at
low corrosion potential (Figure 2), and thus the behavior in BWRs and PWRs is linked, with the primary
differences being dissolved H2, temperature, and the
dissolved ion chemistry (B and Li are added to PWR
primary water; see Chapter 5.02, Water Chemistry
Control in LWRs).8–10,14 Of these, temperature has a
universal effect, variations in dissolved H2 are particularly important in nickel alloys, and B/Li has little or
no effect in deaerated water.9,14
Early plant (Figure 4) and laboratory (Figure 5)
observations showed that the same basic dependencies
existed for unirradiated and irradiated stainless steels,
and that increasing fluence produces a well-behaved
increase in SCC susceptibility (Figure 6). Figure 4
shows a strong effect of water purity for both unirradiated and irradiated BWR components, and Figure 5
shows a very similar response to corrosion potential to
that in Figure 2. Thus, it was proposed that radiation
enhances SCC primarily in four ways: segregation,
hardening, relaxation, and radiolysis (Figure 1). The
neutron fluence where these processes have an effect is
shown in Figure 7, along with the current end-of-life
fluence for various BWR and PWR components. The
primary radiation effects on materials operate in a
similar range of fluences, and thus their individual
contributions can be difficult to distinguish. An example of their interaction in altering SCC growth rate is
shown in the prediction of cracking of a weld in a BWR

core shroud (Figure 8) in which the individual effects
are plotted along with the resulting crack length versus
time. While many of the enhancements in SCC susceptibility from irradiation dose (neutron fluence)
have been well established, it remains possible that
additional factors will emerge at high fluences (e.g.,
>30 displacements per atom (dpa)).
Intergranular (IG) SCC is promoted in austenitic stainless steels above a ‘threshold’ fluence


Irradiation Assisted Stress Corrosion Cracking

179

1.0E–05
1

25 mm CT specimen
Furnace sensitized; 15 C cm–2
288 ЊC water
; 0.1–0.3 mS cm–1
Constant load
; 25 Ksi in1/2

6

Crack growth rate (mm s–1)

42.5 μin h–1

10–7


10

8

1.0E–06

11

Crack growth rate (mm s–1)

10–6

Sensitized 304 stainless steel
30 MPa m1/2, 288 ЊC water
0.06–0.4 μS cm–1, 0–25 ppb SO4
filled triangle = constant load
open squares = ‘gentle’ cyclic

5

14

14.2 μin h–1
Theoretical
curves

ααα
9


μS cm–1

0.3
0.2

0.1

10–8

2

200 ppb O2
500 ppb O2
2000 ppb O2

304 Stainless steel

3 7
4

Screened round robin data
- highest quality data
- corrected corr. potential
- growth rates corrected
to 30 MPa m1/2

42.5
28.3
14.2
–1

μin h

1.0E–07
GE pledge
predictions
30 MPa m1/2

0.5
2000 ppb O2
Ann. 304SS
200 ppb O2

0.25
1.0E–08
0.1

12

0.06 μS cm–1

Hydrogen water
chemistry
β

Normal water
chemistry
(ex-core)

–600


–400
–200
0
+200
Corrosion potential (mVshe)

+400

30 MPa m1/2
1.0E–09
–0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1
Corrosion potential (Vshe)

1.0E–05
Sensitized 304 stainless steel
30 MPa m1/2, 288 ЊC water
0.06–0.4 μS cm–1, 0–25 ppb SO4
SKI round robin data
filled triangle = constant load
open squares = ‘gentle’ cyclic

200 ppb O2
500 ppb O2
2000 ppb O2

β

10–9

–1


0.06 μS cm
Industry mean

0.2

0.3

0.4

4 dpa
304SS

Crack growth rate (mm s–1)

1.0E–06
316L (A14128, square)
304L (Grand gulf, circle)
nonsensitized SS
50% RA 140 C (black)
10% RA 140 C (gray)
1.0E–07

20% CW
A600
42.5
28.3
14.2
μin h–1


20% CW A600
GE pledge
predictions
30 MPa m1/2
Sens SS

0.5
2000 ppb O2
Ann. 304SS
200 ppb O2

0.25

1.0E–08

–1

0.1

0.1 μS cm
Means from analysis of
120 lit. sens SS data

0.06 μS cm–1

0.06 μS cm–1

GE pledge predictions for
Unsens. SS (upper curve for 20% CW)


1.0E–09
–0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1
Corrosion potential (Vshe)

0.2

0.3

0.4

Figure 2 Stress corrosion cracking growth rate versus corrosion potential for stainless steels tested in high-purity water
at 288  C containing 2000 ppb O2 and 95–3000 ppb H2. Dissolved O2 strongly influences corrosion potential, which in turn
affects crack chemistry and growth rate of sensitized stainless steels (two graphs at left) as well as cold-worked stainless
steels and Alloy 600 (large rectangular symbols on right graph) and irradiated stainless steel (large triangular symbols).
Cold-worked or irradiated materials have an elevated yield strength, which causes an increase in growth rate at both low
and high potential. RA, Reduction in area; CW, Cold work.


180

Irradiation Assisted Stress Corrosion Cracking

10–5

Stress intensity (ksi in1/2)
6 8 10
20 30 40 60 80

4


10–3
Sens. 304 stainless steel
288 ЊC water

10–4

10–7

NRC disposition
line

10–5

*
10–8

*

Theory
15 C cm–2, –50 mVshe
0.5 ms cm–1
Theory
–2
15 C cm , –50 mVshe
0.2 ms cm–1

10–9

10–6


*

Theory
15 C cm–2, –200 ® –500 mVshe
0.2 ms cm–1

10–10

4

6

Crack growth rate (in h–1)

Crack growth rate (mm s–1)

10–6

10–7

8 10
20 30 40 60 80
Stress intensity (MPa m1/2)

Figure 3 Effect of stress intensity factor on stress
corrosion cracking growth rate for sensitized stainless steel
exposed in various water chemistries at 288  C.

(Figures 5 and 6). This occurs in oxygenated (e.g.,
BWR) water above 2–5 Â 1020 n cm–2 (E > 1 MeV),

which corresponds to about 0.3–0.7 dpa, and depends
on the stress, water chemistry (especially, sulfate and
chloride), and other factors. Attempts to reproduce the
same level of IG cracking in inert environments have
been unsuccessful, confirming that it is an environmental cracking phenomenon, not simply a change in
the mechanical properties and response of the material
in an inert environment.
Cracking in hydrogenated water (i.e., BWR hydrogen water chemistry (HWC) or PWR water) is typically observed at roughly a 4Â higher fluence than in
oxidizing water, with IASCC enhanced at elevated
temperature (Figure 9). For both BWR and PWR
conditions, the same basic dependencies exist for
unirradiated and irradiated materials.
It is important to distinguish the results of different
kinds of SCC testing. Crack-growth testing typically
uses fracture mechanics specimens, commonly a compact type (CT) specimen. It has the significant

advantage of continuous, online monitoring of crack
length versus time, usually by employing an electric
current, potential drop technique. It can provide a
resolution of about 1 mm and can accurately characterize the inherent resistance to crack advance (i.e., beyond
the ill-defined initiation stage) as well as the dependencies on corrosion potential, stress intensity factor,
etc. Smooth specimen tests, whether by constant load,
constant displacement, or slow strain rate (SSR), are
simpler tests to perform, but represent some combination of initiation and growth. SSR tests impose failure
and can overstate or understate the SCC susceptibility.
Constant load or displacement tests generally require
periodic interruption for examination, and ‘initiation’
can be microscopic cracks or complete failure.
Work over the last 25 years has enabled many aspects
of IASCC phenomenology to be explained and predicted on the basis of the experience with intergranular

stress corrosion cracking (IGSCC) of nonirradiated
stainless steel in high-temperature water environments.
This continuum approach has successfully accounted
for radiation effects on water chemistry and its influence on electrochemical corrosion potential. However,
all radiation-induced microstructural and microchemical changes that promote IASCC are neither fully
known nor fully reproducible in similar materials.
Well-controlled data from well-characterized irradiated
materials remain insufficient due to the inherent experimental difficulties and financial limitations. Many of
the important metallurgical, mechanical, and environmental aspects that are believed to play a role in the
cracking process are illustrated in Figure 1. Only persistent material changes are required for IASCC to
occur, but in-core processes such as radiation creep
and radiolysis also have an important effect on IASCC.

5.08.2 Irradiation Effects on SCC:
Laboratory and Plant Data
5.08.2.1 Individual Effects of Radiation
on IASCC
IASCC can be categorized into radiation effects on the
water chemistry (radiolysis) and on the material/stress,
and the accepted definition of IASCC encompasses
cases where either factor is dominant (low-fluence
materials tested in water undergoing radiolysis, or preirradiated materials tested without an active radiation
flux). Radiation dose rate in rads/h is often used in
radiolysis, and in neutrons per square centimeter
(n cm–2) or displacements per atom in materials.
In light water reactors (LWRs), 1 dpa corresponds


Frequency of SCC initiation
increases dramatically with

increasing conductivity

0.2

Best fit

1.2
1.0
0.8
0.6

Threshold
conductivity for
SCC initiation
increases as level
of sensitization
decreases
Low carbon SS
High carbon SS

Nonsensitized
low carbon SS

0.2
0
0

0.2

(b)


1.0
0.9

Sensitized
high carbon SS

0.4

0.6

Plant average conductivity (ms cm–1)

0.4

0.6

Plant average conductivity (ms cm–1)

*

0.8

Frequency of IGSCC
initiation increases
with plant conductivity

0.7
0.6
0.5


*

0.4

* Experienced
substantial high
conductivity
excursions not
reflected in
average value

0.3
0.2

*

0.1
0

(c)

181

1.4
Upper bound

0.4

% with IGSCC/on-line months


(a)

1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0

% with IGSCC/on-line months

% with IGSCC/on-line month

Irradiation Assisted Stress Corrosion Cracking

0

0.1


0.2

0.3

0.4

0.5

0.6

0.7

Plant average conductivity (ms cm–1)

Figure 4 The effects of average plant water purity shown in field correlations of the core component cracking behavior for
(a) stainless steel intermediate and source range monitor dry tubes, (b) creviced stainless steel safe ends, and (c) creviced
Inconel 600 shroud head bolts, which also shows the predicted response versus conductivity. Adapted from Brown, K. S.;
Gordon, G. M. In Proceedings of 3rd Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors;
The American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): New York, NY, 1988; pp 243–248.

to $7 Â 1020 n cm–2 when counting neutrons with
E > 1 MeV, or $1.5 Â 1021 n cm–2 for E > 0.1 MeV.
The primary effects of radiation on materials8,15–22
include microcompositional effects (grain boundary
chemistry) and microstructural changes (formation of
dislocation loops, voids, precipitates, and the resulting changes hardening and deformation mode). In
terms of their effect on IASCC, the primary effects
of radiation are the following:
 Radiolysis of water, in which a variety of short- and

long-lived radicals and species are produced.
There is no evidence that the specific species
formed are important, and indeed their effect on
cracking appears to be fully captured by their overall effect on the corrosion potential of the material.
 Radiation-induced segregation (RIS), which produces an enrichment in some species (e.g., Ni and
Si) at grain boundaries and other defect sinks, and a
depletion in other species (e.g., Cr). Even though the
distance over which RIS occurs is very limited (a few

nanometers), studies of unirradiated materials have
shown that the narrow profiles can affect SCC.23,24
 Radiation hardening (RH), which results from radiation damage and the creation of vacancy and interstitial loops, which impede dislocation motion. Once
a few dislocations move along a given slip plane, they
clear the ‘channel’ of most of these barriers, and
subsequent dislocation occurs primarily in these
channels.
 Radiation creep relaxation, which reduces constant displacement stresses such as in bolts or associated with weld residual stress. During active
irradiation, radiation creep can promote dynamic
strain, and thereby SCC.
 Swelling, which occurs to a limited extent at temperatures above $300  C, but can be sufficient to
produce reloading of components such as PWR
baffle former bolts. Swelling occurs differently in
different materials, and is delayed in cold-worked
materials. Stresses due to swelling are balanced by


182

Irradiation Assisted Stress Corrosion Cracking


Data of Jacobs( ) and Kodama( )
Postirrad. SSRT 2–3 ϫ 10–7 s–1 288 ЊC
Comm. purity 304SS( ) and 316SS( )
42 ppm O2-sat’d vs.
0.02 ppm O2

100

80

»2

100

40

»42

ppm O2

60
%IGSCC

% IGSCC fracture

Data shifted right by Init. grain boundary Cr enrichment

»0.2
0.02
–0.05


fC, Vshe

20
–0.4

0
1019

–0.2

0

0.2

1020
1021
Neutron fluence (n cm–2 ) (E > 1 MeV)

1022

Percentage of spot welds inspected with IGSCC

Figure 5 Dependence of irradiation-assisted stress corrosion cracking on fast neutron fluence as measured in slow strain
rate tests at 3.7 Â 10À7 sÀ1 on preirradiated type 304 stainless steel in 288  C water. The effect of corrosion potential via
changes in dissolved oxygen is shown at a fluence of %2 Â 1021 n cm–2. Reproduced from Jacobs, A. J.; Hale, D. A.; Siegler, M.
Unpublished Data on SCC of Irradiated SS in 288  C Water and Inert Gas; GE Nuclear Energy: San Jose, CA, 1986.
SSRT, Slow strain rate test.

PWR control

BWR core BWR end rod failures
(IASCC)
component
of life
failures
(IASCC)

100

80

1020

BWR creviced control blade sheath
60
Threshold fluence for IGSCC
≈5 ϫ 1020 n cm–2

40

0.1

0
0

0.2

0.4

0.6


0.8

1

1.2

1.4

Neutron fluence (n cm–2 ϫ 1021) (E > 1 MeV)

Figure 6 Dependence of irradiation-assisted stress
corrosion cracking on fast neutron fluence for creviced control
blade sheath in high-conductivity boiling-water reactors.
Reproduced from Gordon, G. M.; Brown, K. S. In Proceedings
of 4th International Conference on Environmental Degradation
of Materials in Nuclear Power Systems – Water Reactors;
NACE: Houston, TX, 1990; pp 14-46–14-62.

radiation creep relaxation, but the resulting stress
can be sufficient to cause IASCC.
 Other microstructural changes, such as precipitation or dissolution of phases in materials. While
there is no clear evidence that such changes affect
IASCC response, this may only reflect the limited

PWR end
of life
PWR life
extension


1021
1022
1023
Neutron fluence (n cm–2) (E >1 MeV)
Irradiation dose (dpa)
1
10
Significant changes
in grain boundary
composition, alloy
strength, and ductility

20

PWR baffle
bolt failures
(IASCC)

100
Onset of significant
void swelling
and possible
embrittlement

Figure 7 Neutron fluence effects on irradiation-assisted
stress corrosion cracking susceptibility of type 304SS
in boiling-water reactor environments. Reproduced
from Bruemmer, S. M.; Simonen, E. P.; Scott, P. M.;
Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater.
1999, 274, 299–314.


characterization and IASCC studies that has been
performed on high fluence materials.
The individual (segregation, hardening, and creep,
Figure 10) and composite (SCC, Figures 5–9) effects
of radiation increase with dose in much the same
manner, which makes the isolation of, and attribution
to, individual contributions difficult. The dislocation
loop microstructure is closely tied to radiation


Irradiation Assisted Stress Corrosion Cracking

30

20

30
25

Effect of
rad segregation

20

Effect of rad hardening

15
10


Depth
10

Stress intensity (K)

5

Effect of stress relaxation
0
0

100

200
300
Time (month)

25
Cr
Cr depletion
20
Loop line
length
SCC

15

10

0

500

400

5

Figure 8 Predicted effect of radiation segregation,
radiation hardening, and radiation creep relaxation on a
boiling-water reactor core shroud, where the through-wall
weld residual stress profile is the primary source of
stress. Less aggressive water chemistry (corrosion
potential and water purity) would result in less crack
advance early in life, which would give a greater
opportunity for radiation creep relaxation. The leak depth
is the wall thickness of the shroud. While radiation
hardening continues to increase the yield strength, its effect
on crack growth is reduced (see Figure 21(a)). EPR,
Electrochemical potentiokinetic repassivation (a test for
sensitization).

Fraction of IG cracking area

Hardness

Arbitrary units

Crack depth (mm)

Leak depth


38.1-mm thick 304SS, two-sided weld
0.75 Vshe, 0.15 μS cm–1
EPR0 = 10.8 C cm–2 (0.050% C)
Flux = 3 ϫ 1019 n cm–2 year

K, segregation, hardening, or relaxation

40

30

183

CP-304SS
3.2 MeV protons
360 ЊC

0
0

1

2

3
Dose (dpa)

4

5


6

Figure 10 Schematic diagram showing the dose
dependence of key irradiated microstructure features
(radiation-induced segregation and dislocation
microstructure) and radiation hardening along with stress
corrosion cracking susceptibility. Reproduced from
Was, G. S. In Proceedings of 11th International
Conference on Environmental Degradation of Materials
in Nuclear Power Systems – Water Reactors; American
Nuclear Society (ANS): La Grange Park, IL, 2004;
pp 965–985.

100
80
60
40

In PWR primary water
CW316, 340 ЊC
CW316, 325 ЊC
CW316, 290 ЊC
Type 304, 325 ЊC

irradiation. Subsequent sections focus on the possible
mechanisms by which water chemistry, RIS, microstructure, hardening, deformation mode, and irradiation creep – individually or on concert – may affect
IASCC.

*


* Specimen broke at the pin hole

20
0

*

*
*
1.0E + 20

1.0E + 21
1.0E + 22
Fluence (n cm–2) (E > 0.1 MeV)

1.0E + 23

Figure 9 Percentage intergranular stress corrosion
cracking versus fluence for cold-worked type 316 stainless
steel tested at various temperatures. Despite the low
potential environment of pressurized-water reactor primary
water, at high fluence (especially at higher temperature)
there is significant susceptibility to stress corrosion
cracking.

hardening and both increase with dose until saturation occurs by $5 dpa. RIS also increases with dose
and tends to saturate by $5 dpa. Although dependent
on both metallurgical and environmental parameters,
IASCC generally occurs at doses between 0.5 dpa (for

BWRs) and 2–3 dpa (for PWRs), which encompasses
the steeply rising portion of the curves in Figure 4
that describe the changes in materials properties with

5.08.2.2

Service Experience

The IASCC service experience extends over 50 years,
and the early observations projected an accurate view
of the important characteristics and dependencies,
and pointed to the ‘assisted’ nature of radiation in
enhancing SCC. As with other forms of SCC, early
observations suggested that a growing incidence with
time and neutron fluence should be expected.
IASCC was first reported in the early
1960s7,8,15–22,25–36 and involved intergranular cracking of stainless steel fuel cladding, with the initiation of multiple cracks occurring from the water side.
By contrast, mostly ductile, transgranular cracking
was observed in postirradiation mechanical tests
performed in inert environments at various strain
rates and temperatures. Grain boundary carbide


184

Irradiation Assisted Stress Corrosion Cracking

precipitation was rarely observed although preexisting thermal sensitization was present in some cases.
A correlation between time-to-failure and stress level
was reported, with failure occurring first in thinwalled rods with small fuel-to-cladding gaps, where

fuel swelling strains were the largest. The highest
incidence of cracking occurred in peak heat flux
regions, corresponding to the highest fluence and
the greatest fuel–cladding interaction (highest stresses and strains.) Similar stainless steel cladding in
PWR service exhibited fewer instances of intergranular failure. At that time, off-chemistry conditions or
stress rupture were often considered to be the cause
of PWR failures.
In the last 30þ years, a growing number of other
stainless steel (and nickel alloy) core components
have exhibited IASCC, including neutron source
holders in 1976 and control rod absorber tubes in
1978. Instrument dry tubes and control blade handles
and sheaths, Figure 5, which are subject to very low
stresses, are also cracked, generally in creviced locations and at higher fluences.15,18,20,37,38 These initial
failures in the most susceptible components were

Table 1

followed by more numerous incidents of IASCC in
the past 20 years, perhaps most notably in BWR core
shrouds8,15,16,39 and PWR baffle bolts.19,40,41
A summary of reported failures of reactor internal
components is listed in Table 1 and demonstrates
that IASCC is not confined to a particular reactor
design, material, component, or water chemistry. For
example, stainless steel fuel cladding failures were
reported years ago in commercial PWRs and in PWR
test reactors.15,22,25,29–36,42 At the West Milton PWR
test loop, intergranular failure of vacuum-annealed
type 304 stainless steel fuel cladding was observed31

in 316  C ammoniated water (pH 10) when the cladding was stressed above yield. Similarly, IASCC was
observed in creviced stainless steel fuel element ferrules in the Winfrith steam generating heavy water
reactor (SGHWR),43,44 a 100 MWe plant in which
light water is boiled in pressure tubes, where the coolant chemistry is similar to other boiling-water reactor
designs. The 20% Cr–25% Ni–Nb stainless steel differs from type 304 primarily in Ni and Nb content, as
well as in its lower sulfur (%0.006%) and phosphorus
(%0.005%) contents. The ferrules were designed for

Irradiation-assisted stress corrosion cracking service experience

Component

Material

Reactor type

Possible sources of stress

Fuel cladding
Fuel cladding
Fuel claddinga
Fuel cladding ferrules
Neutron source holders
Instrument dry tubes
Control rod absorber tubes
Fuel bundle cap screws
Control rod follower rivets
Control blade handle
Control blade sheath
Control blades

Plate type control blade
Various boltsb
Steam separator dryer boltsb
Shroud head boltsb
Various bolts
Guide tube support pins
Jet pump beams
Various springs
Various springs
Baffle former bolts
Core shroud
Top guide

304SS
304SS
20% Cr–25% Ni–Nb
20% Cr–25% Ni–Nb
304SS
304SS
304/304L/316L SS
304SS
304SS
304SS
304SS
304SS
304SS
A-286
A-286
600
X-750

X-750
X-750
X-750
718
316SS cold work
304/316/347/L SS
304SS

BWR
PWR
AGR
SGHWR
BWR
BWR
BWR
BWR
BWR
BWR
BWR
PWR
BWR
PWR and BWR
BWR
BWR
BWR and PWR
PWR
BWR
BWR and PWR
PWR
PWR

BWR
BWR

Fuel swelling
Fuel swelling
Fuel swelling
Fabrication
Welding and Be swelling
Fabrication
B4C swelling
Fabrication
Fabrication
Low stress
Low stress
Low stress
Low stress
Service
Service
Service
Service
Service
Service
Service
Service
Torque, differential swelling
Weld residual stress
Low stress (bending)

a


Cracking in AGR fuel occurred during storage in spent fuel pond.
Cracking of core internal occurs away from high neutron and gamma fluxes.
AGR, Advanced gas-cooled reactor
b


Irradiation Assisted Stress Corrosion Cracking

a 5-year exposure during which the peak fast neutron
flux is 2–3 Â 1013 n cm–2 s (E > 1.5 MeV), yielding a
peak fluence over 5 years of 3–5 Â 1021 n cm–2.
The similarity of IASCC in BWRs and PWRs was
also noted in swelling tube tests performed in the
core44,45 on a variety of commercial and high-purity
heats of types 304, 316, and 348 stainless steel and
Alloys X-750, 718, and 625. Swelling was controlled
by varying the mix of Al2O3 and B4C within the
tubes; the latter swells as neutrons transmute B to
He. Nominally identical strings of specimens were
inserted into the core in place of fuel rods.
Historically, the oxidizing potential in a PWR
core is lower than in BWRs, but in the last decade,
most BWRs employ an electrocatalytic technology
called NobleChem™(46–48) to create a low corrosion
potential. Some early investigators attributed PWR
cracking to low ductility stress rupture (of course,
this mechanism would apply equally to BWRs).
A few laboratory studies reported small amounts of
intergranular cracking of irradiated stainless steels in
SSR tests in %300  C inert environments49 although

in many related experiments50,51 no intergranular
failure was found. Small amounts of intergranular
cracking in inert tensile tests are not surprising, and
since the early 1990s, the plant and laboratory
IASCC data show that cracking is environmentally
assisted and follows a well-behaved continuum in
response over ranges in fluence, corrosion potential,
temperature, stress, etc.8,15–17
Factors that distinguish PWRs from BWRs
include their higher operating temperature, %10Â
higher maximum neutron fluence in core structural
components, higher hydrogen fugacity, and borated–
lithiated water chemistry (including the possibility of
localized boiling and thermal concentration cells in
crevices from gamma heating, which could lead to
aggressive local chemistries). The possible role of RIS
of Si may be especially important in accounting for
the limited difference in SCC response at high
potential (BWR) versus low potential (PWR) at
high fluence.52,53
Brown and Gordon37,38 (Figure 4) accumulated
and analyzed data for cracking in Alloy 600 shroud
head bolts (first observed in 1986) as well as stainless
steel safe ends (first observed in 1984) and in-core
instrumentation tubes (first observed in 1984) with a
focus on components that were creviced, a factor
known to exacerbate cracking.7,37,38 The highest radiation exposure occurred for the intermediate range
and source range monitor (IRM/SRM) dry tubes,
which contain flux monitors housed in thin-walled,


185

annealed stainless steel tubes. Cracking initiated
in the crevice between the spring housing tube
and the guide plug at fluences between 0.5 and
1.0 Â 1022 n cm–2 (E > 1 MeV). Wedging stresses
from the thick oxide observed in the crevice were
implicated, since other (applied and residual) stresses
were negligible. The primary variable from plant
to plant is the average coolant conductivity, which
correlates strongly with cracking incidence
(Figure 4(a)–4(c)). Each point in Figure 4 represents
inspection results for one BWR plant, and data are
normalized using reactor operating time (i.e., percentage of components with intergranular cracks
divided by the online exposure time). The scatter in
Figure 4(a) was attributed to variations in fluence
and specific ion chemistry, as well as limitations in
the resolution of underwater visual inspection. Scatter can also result from short-term excursions in
conductivity, which is not adequately reflected in
the average, as identified in Figure 4(c).
Correlations between IASCC and conductivity
were also reported for cracking in shroud head bolts
(Figure 4(c)) and creviced safe ends (Figure 4(b)).
The strong influence of conductivity on cracking of
stainless steel has also been shown in laboratory tests
and plant recirculation piping, where predictive
modeling8,12,13,15,54,55 has been compared to field
data on the operational time required to achieve a
detectable crack (typically, 10% of the wall thickness). Preliminary prediction of the shroud head bolt
cracking response8 also provides reasonable agreement with observation (Figure 4(c)).

High-strength, nickel-base alloy components15,22
have also experienced IASCC (Table 1), with many
incidents in lower radiation flux regions (e.g., where
the end-of-life fluence is below %5 Â 1019 n cm–2)
such as cracking of Inconel X-750 jet pump beams
in BWRs. Inconel X-750 cracking has also occurred
extensively in PWR fuel hold-down springs, which
attain an end-of-life fluence of 1–10 Â 1021 n cm–2; it
is proposed that cracking has been aggravated by
vibrational stresses (corrosion fatigue). The effects of
irradiation on IASCC in high-strength, precipitationhardened nickel-base alloy components as well as in
stainless steels have not been characterized.
BWR core shrouds8,15,16,39 and PWR baffle
bolts40,41 are the two most common examples of
IASCC although susceptibility clearly exists in
other areas, such as control blade components, fuel
components, the BWR top guide, etc. SCC in the
BWR core shroud occurs almost exclusively near the
welds (both circumferential and vertical), and


186

Irradiation Assisted Stress Corrosion Cracking

initiation is observed from both the inside (ID) and
outside (OD) surfaces (the shroud separates the
upward core flow from the downward recirculation
flow that occurs in the annulus between the shroud
and the pressure vessel). This large-diameter welded

‘pipe’ has little active (DP) loading, and its susceptibility arises primarily from weld residual stresses and
weld shrinkage strains.56–58 Cracking is observed in
both low fluence and moderate fluence areas, and the
extent of the enhancement in SCC susceptibility by
irradiation is limited because, while RH and RIS
occur, radiation creep also relaxes the weld residual
stress. SCC predictions for a BWR core shroud that
account for the damaging effect of RIS and RH and
the beneficial effect of radiation creep relaxation are
shown in Figure 8 and illustrate the complexity of
the interactions of these phenomena in the evolution
of cracking. Predictions also indicate that, if SCC
does not nucleate early in life (e.g., below 0.5 dpa),
for example, from high coolant impurity levels or
severe surface grinding, susceptibility tends to
decrease with fluence in the shroud welds because
of radiation-induced creep relaxation (although
many shroud welds are in very low flux areas).
The last decade has also seen extensive failures of
PWR baffle bolts40,41 although large plant-to-plant
and heat-to-heat differences are observed. Most baffle bolts are fabricated from type 316 stainless steel
cold-worked to %15% to increase their yield
strength. The complex baffle former structure exists
in a PWR because their fuel does not have a surrounding ‘channel,’ so the baffle former structure
must conform closely to the geometry of the fuel to
provide well-distributed water flow. The baffle former plates are usually made from annealed material,
typically type 304 stainless steel. Because of their
proximity to the fuel, very high fluences can develop –
up to $80 by the end of the original design life. The
high gamma flux produces significant heating in the

components, in some instances estimated at þ40  C,
especially in designs where the PWR coolant does
not have good access to the bolt shank. While the
heat-to-heat variations are not understood, it is clear
that plants that load-follow (and therefore undergo
power level changes and thermal cycles) are much
more prone to baffle bolt cracking. Another aggravant
is the thermal gradient and possible boiling (resulting
in altered water chemistry) in the shank area of the
bolt if the coolant access is restricted. However, primary factors must be the very sizeable stress relaxation that occurs early in life (e.g., during the first

5 dpa), followed by preferential radiation swelling of
the annealed baffle plates over the cold-worked baffle
bolts, which will cause reloading. The dynamic equilibrium between swelling and radiation creep, which
determines the ‘reloading’ stress in the bolt, is likely a
complex function of many parameters, including
local neutron flux, temperature, baffle plate geometry, and composition.
The number of IASCC incidents continues to
grow, and there can be no question that many LWR
components are susceptible. Strategies to mitigate
IASCC (e.g., NobleChem™(46–48)) and manage
IASCC (e.g., by showing some IASCC could be tolerated, installing mechanical restraints to mitigate
the impact of IASCC in BWR shrouds, or selectively
inspect and replace baffle bolts) have been successful.
IASCC field experience has led to the following
trends and correlations:
 Intergranular cracks associated with radiation
effects on solution-annealed stainless steel were
once thought to occur only at fluences above
%0.3 Â 1021 n cm–2. But significant intergranular

cracking in BWR core shrouds (which do not have
thermal sensitization) occurs over a broad range
of fluences, showing that a true fluence threshold
does not exist.15,16 The observations of SCC in unirradiated, unsensitized stainless steel (with or without
cold work) also undermine the concept of a threshold fluence below which no SCC occurs. This also
holds for thresholds in corrosion potential, water
impurities, temperature, etc.8,54,59,60
 SCC susceptibility is affected by fluence in a complex fashion. SCC in BWR shrouds and PWR
baffle bolts does not always correlate strongly
with fluence, and one important reason for this is
that radiation creep produces relaxation of the
stresses from welding and in bolts. The need to
account for many changing factors is necessary in
interpreting and predicting SCC.
 Most early incidents involved high stresses or
dynamic strains, but cracking has since been
observed at quite low stresses at high fluences
and longer operating exposure. Laboratory and
field data indicate that IASCC occurs at stresses
below 20% of the irradiated yield stress, and at
stress intensities below 10 MPa m1/2.
 Extensive laboratory and field data show that corrosion potential is a very important parameter, with
its effect being consistent from low to high fluence,
except in some high fluence materials and/or


Irradiation Assisted Stress Corrosion Cracking












under high stress intensity factor conditions. Materials prone to high radiation-induced changes in
grain boundary Si level may exhibit a very limited
effect of corrosion potential.52,53 There is no evidence of threshold potential, and indeed irradiated
materials exhibit IASCC in deaerated water.
Impurities, especially chloride and sulfate, strongly
affect IASCC in BWR water (Figure 4). As noted
by Brown and Gordon,37 this correlation applies
equally to low and high flux regions and to stainless
steels (Figure 4(a) and 4(b)) and nickel-base alloys
(Figure 4(b)). Indeed, the correlation closely parallels that from out of core.8,12,13,15,38,54,55 At higher
levels, the same impurities can affect SCC in
PWRs. Similarly, if high corrosion potential conditions form in the PWR primary (where B and Li are
present), high growth rates can result.9,14
Crevices exacerbate cracking primarily due to
their ability to create a more aggressive crevice
chemistry from the gradient in corrosion potential
(in BWRs) or in temperature (most relevant to
PWRs). Crevices can also produce stress and strain
concentration.
Cold-working often exacerbates cracking (especially abusive surface grinding), although it can
also delay the onset of some radiation effects.
Temperature has an important effect on IASCC,

enhancing both crack initiation and growth rate.
Preexisting grain-boundary carbides or chromium
depletion is not required for susceptibility
although furnace-sensitized stainless steels are
clearly highly susceptible to cracking in-core. Cr
depletion will develop or be magnified by irradiation, and increase IASCC susceptibility, although
its effect is most pronounced in pH-shifted environments, as can develop when potential or thermal
gradients exist. The role of N, S, P, and other grainboundary segregants is less clear.
IASCC is enhanced at a fluence that is dependent
on applied stress and strain, corrosion potential,
solution conductivity, crevice geometry, etc. At
sufficiently high conductivities, cracking has been
observed in solution-annealed stainless steel in
the field (Figure 6(a) and 6(b))37 and in the laboratory.8,12,13,15,54,55 Thus, while convenient in
a practical engineering sense, the concept of a
‘threshold’ fluence (or stress, corrosion potential,
etc.) is scientifically misleading8,52,59,60; cracking
susceptibility and morphology are properly considered an interdependent continuum over many
relevant parameters.

187

5.08.3 Irradiation Effects on Water
Chemistry
5.08.3.1 Radiolysis and Its Effect on
Corrosion Potential
SCC susceptibility is fundamentally influenced by
corrosion potential, not by the oxidant and reductant
concentrations per se.8,15,61 The lower H2 concentration in BWRs distinguishes BWRs from PWRs
because it permits the radiolytic formation of oxidants. Above %500 ppb (5.6 cm3 kgÀ1) H2, radiolytic

formation of oxidants is effectively suppressed, and
the corrosion potential remains close to its thermodynamic minimum (which is a function of temperature, H2 fugacity, and pH). BWRs cannot achieve this
H2 level in the core because H2 partitions to the
steam phase, which begins to form about a quarter
of the way up the fuel rods. Thus, this section is
primarily relevant to BWRs.
Radiolysis and the presence of oxidizing species
require that many sequential and nonlinear dependencies that must be considered, for example, radiation flux produces oxidizing and reducing species,
the corrosion potential is controlled by multiple
reactions of these species, the crack chemistry is
nonlinearly dependent on the corrosion potential
and the dissolved ions present, and the SCC growth
rate is a nonlinear function of the crack chemistry.8,61 The relationship between radiation flux and
SCC cannot easily be determined empirically, but
rather requires a fundamental understanding of each
subprocess.
Water is decomposed by ionizing radiation into
various primary species62–65 including both radicals
(e.g., eÀ
aq, H, OH, HO2) and molecules (e.g., H2O2, H2,
O2), which can be oxidizing (e.g., O2, H2O2, HO2) or
reducing (e.g., eÀ
aq, H, H2). The predominant species
that are stable after a few seconds are H2O2 and H2,
with O2 forming primarily from the decomposition of
H2O2. Because H2 partitions to the steam phase and
H2O2 is not volatile, %87% of the water that is
recirculated in a BWR (%11–14% of the core
flow becomes steam) is oxidant rich. H2 is introduced
in the feed water, which mixes with the recirculated

water near the top of the annulus (the region of
downflow between the core shroud and pressure
vessel).
The concentrations of radiolytic species are
roughly proportional to the square root of the radiation flux in pure water. The radiation energy versus
intensity spectrum influences the concentration of


Irradiation Assisted Stress Corrosion Cracking

0.4
0.3
Corrosion potential (VSHE)

each radiolytic species, which is described in terms of
a yield, or G value (molecules produced per 100 eV
absorbed by water). In LWRs, the G values for most
species are within a factor of %3 for fast neutron
versus gamma radiation. Despite this similarity, the
influence of fast neutron radiation is much stronger
than gamma radiation primarily because the energy
deposition rate, or mean linear energy transfer
(LET), is greater (40 eV nm–1 for fast neutrons versus
0.01 eV nm–1 for gamma radiation63). Also, the neutron flux in LWRs (e.g., %1.03 Â 109 Rad h–1 core
average and %1.68 Â 109 Rad h–1 peak in a BWR4 of
51 W cm–3 power density) is also higher than the
gamma flux (%0.34 Â 109 Rad h–1). Indeed, the moderate gamma levels present in the downcomer in the
outside annulus of a BWR core actually promote
recombination of hydrogen and various oxidants.62,65
This is a key element in the effectiveness of HWC,

and BWRs that have a wider annulus and lower
gamma flux near the pressure vessel respond less to
a given H2 addition. The contribution of thermal
neutrons and beta particles to radiolysis is small
in LWRs.
As in many electrochemical processes, the
integrated effects of various oxidants and reductants
on environmental cracking is best described via
changes in corrosion potential, which controls the
thermodynamics and influences the kinetics of most
reactions. Since electrochemical potentials are logarithmically dependent on local oxidant, reductant,
and ionic concentrations (via the Nernst relationship,
’ ¼ ’o þ RT/nF ln [products/reactants], where ’ is
electrochemical potential, R is the gas constant, T is
temperature in kelvin, n is the number of moles, and F
is Faraday’s constant), radiation-induced increases in
concentration of various species by many orders of
magnitude may have comparatively small effects on
the corrosion potential in hot water. Further, corrosion potentials are mixed potentials involving a balance of anodic and cathodic reactions on the metal
surface. At low oxidant concentrations, the rapid drop
in corrosion potential to %À0.5 Vshe (Figure 11(a))
results from mass-transport-limited kinetics of oxidants to the metal surface.
The relationship between dissolved oxygen and
corrosion potential in hot water as a function of
radiation type and flux is shown in Figure 11(a),
in which the connected points represent data
obtained in controlled radiation on/off experiments.
The data from these latter experiments are shown in
Figure 11(b) in terms of a radiation-induced shift in
potential. The curves in Figure 11(a) represent the


2 points, 200 ppb H2O2
1 point, 31ppb Cu2+

0.2
0.1

0
–0.1

200 ppb H2

H2Wc
BWR top core n + g
Proton rad. = BWR peak
–1
30–200 MGy h g

–0.2
–0.3

–1
0.3–2 MGy h g
Pt, various rad levels

–0.4
–0.5

Open symbol = no radiation
Lines indicate paired data (e.g., g on/off)

Band of data for unirradiated conditions

–0.6
–0.7

100

(a)

101

102

103

104

Dissolved oxygen (ppb)

0.3
0.2
Dfc with radiation (VSHE)

188

0.1

200 ppb H2O2

0

–0.1

200 ppb H2
2 points

–0.2

Proton, ECP in crevice
BWR top core n + g
Proton rad. = BWR peak
30–200 MGy h–1g
0.3–2 MGy h–1g
Pt, various rad levels

–0.3
<100
(b)

102
101
103
Dissolved oxygen (ppb)

104

Figure 11 (a) Effect of radiation on the corrosion
potential of type 304 stainless steel in water at 288  C.
The curves denote the range of typical values in the
unirradiated corrosion potential data (reproduced from
Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al.

In Proceedings of ICONE-4 Conference; ASME
International: New York, NY, 1996). (b) Effect of radiation
on the shift in corrosion potential from the value under
unirradiated conditions for type 304 stainless steel in water
at 288  C. With the exception of the boiling-water reactor
measurements, all data were obtained in controlled
radiation on/off experiments (reproduced from
Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al.
In Proceedings of ICONE-4 Conference; ASME
International: New York, NY, 1996). Curves in (b) show the
trends in the proton-irradiated data, where the effects of
radiation (on/off and over a range of fluxes) were evaluated
for a variety of dissolved O2 and H2 concentrations under
otherwise identical conditions. H2Wc, Hydrogen Water
chemistry.

scatter band for the data obtained under unirradiated
conditions. Similar scatter also exists in the irradiated
corrosion potential data in Figure 11(a) and comprises contributions from both real effects and experimental error. High radiation flux experiments were


Irradiation Assisted Stress Corrosion Cracking

performed by Taylor66 and Andresen et al.67 using
multiple, fundamentally different types of radiationhardened reference electrodes. In-reactor measurements have also been performed by Indig68,69 using
multiple, radiation-qualified silver chloride reference
electrodes.
Figure 11(a) and 11(b) show that little, if any,
elevation in corrosion potential results from irradiation sources that do not include neutrons or simulate
their contribution (e.g., using high-energy protons67).

Some studies using gamma radiation15,66 showed a
significant decrease in corrosion potential, especially
in the intermediate (e.g., 10–200 ppb) range of dissolved oxygen. This is consistent with enhanced
recombination of oxidizing and reducing species,
which occurs in the downcomer region of BWRs.65
In instances where neutrons or protons have been
used, a consistent, significant elevation in corrosion
potential is observed. This is more pronounced at low
dissolved oxygen concentrations with no dissolved
hydrogen (Figure 11(b)), where increases of over
þ0.25 V occur. At higher inlet oxygen concentrations
(e.g., %200 ppb), the data still show a significant
shift (typically þ0.1 to 0.15 V) in corrosion potential for radiation conditions representative of peak
LWR core fluxes (Figure 11(b)); less increase is
observed for inlet oxygen concentrations associated,
for example, with air saturation (%8.8 ppm O2) or
oxygen saturation (%42 ppm O2 at STP). A similar
elevation in corrosion potential is observed for
additions of hydrogen peroxide (200 ppb H2O2,
Figure 11(a)), which suggests that H2O2 may be a
major factor in increasing the corrosion potential
under irradiated conditions.
In-core, in situ measurements in BWRs show that
the corrosion potential, which is %þ0.2 to þ0.25 Vshe
in normal water chemistry (NWC), can be decreased
by >0.5 V by sufficient additions of dissolved hydrogen in a BWR.69 This is corroborated by other measurements67 (Figure 11(b)), which show very little
radiation-induced elevation in corrosion potential
when the fully deaerated inlet water contains moderate dissolved hydrogen (>200 ppb H2, 0 ppb O2). At
high H2 levels, the core becomes reducing, and the
small concentration of 16N (transmuted from 16O)

changes from soluble NOÀ
3 to volatile NOx and
NH3, causing a large increase in radiation level in
the steam lines and turbine.
The effect of radiation on the corrosion potential
within a crack or crevice has also been of interest,
with the possibility that a net oxidizing environment
in the crack could be created that could elevate the

189

corrosion potential above the potential at the crack
mouth. In the absence of radiation, measurements in
high-temperature water in artificial crevices (e.g.,
tubing),70,71 at the tip of growing cracks,72 and of
short crack growth behavior73 show that the corrosion potential remains low (i.e., À0.5 Æ 0.1 Vshe in
pure water at 288  C) for all bulk oxygen concentrations, indicating that complete oxygen consumption
occurs within the crack. Measurements of radiation
effects in crevices67 show that the elevation in corrosion potential is limited to <0.05 V (’c < À0.45 Vshe)
in-core; this is consistent with interpretation of available corrosion potential data on free surfaces.8,15,16
These small changes will not significantly affect the
%0.75 V (þ0.25 Vshe (near mouth) minus À0.5 Vshe (in
crack)) potential difference in the crack under irradiated normal BWR water chemistry conditions. The
potential difference, along with other factors, controls the enhancement mechanism that can lead to
an increased anion activity and altered pH at the
crack tip.12,13,54,55
5.08.3.2 Effects of Corrosion Potential
on IASCC
Preirradiated stainless steels were evaluated in
SSR tests50,74 in hot water using additions of oxygen

and/or hydrogen peroxide to elevate the corrosion
potential to simulate the effect of radiation. Jacobs
et al.50 showed that IASCC in stainless steel irradiated
to %3 Â 1021 n cm–2 was strongly affected by dissolved oxygen (and, by inference, corrosion potential)
(Figure 5(b)). Ljungberg74 also evaluated preirradiated materials in SSR tests and observed decreasing
average crack growth rates with decreasing corrosion
potential. Less IASCC occurs at low corrosion potential, but crack growth rate tests (discussed later) and
other tests confirm that IASCC does not vanish at
low corrosion potential (Figures 5 and 9).
Continuously monitored fracture mechanics
specimens were installed in the recirculation piping
and core of the Nine Mile Point Unit 1 BWR.
Furnace-sensitized type 304 stainless steel specimens
in both locations showed higher growth rates in
the core, where the corrosion potential was higher
(Figure 12). Specimens of annealed type 304 and
304L stainless steel showed growth after different
fluences but, once growth initiated, the growth rates
at a given fluence were identical. The delay time
(fluence) and differences in delay time were attributed to the need to transition from a transgranular
fatigue precrack to IG SCC.


190

Irradiation Assisted Stress Corrosion Cracking

104

100

• Furnace sensitized 304 st. st.
• 27.5 MPa m1/2

Crack length (mm)

100

Nine Mile Point 1 BWR
DCB specimens for monitoring SCC

96
94

In—core

Observed
Theory

Nine Mile Point 1 BWR
DCB specimens for
monitoring SCC
C304SEN, thermally
27.5 MPa m1/2
sensitized 304
C304ST, solution
treated 304

102

92

90

Crack length (mm)

98

98
96
94

C304L, low
carbon 304

92
90

88
88
86

Recirculation line

86

84

84
0

(a)


500

1000 1500 2000 2500 3000 3500 4000
Time (h)

0

2000

4000

6000

8000

10 000

12 000

Time (h)

(b)

Figure 12 Data for fracture mechanics specimens of type 304 stainless steel exposed in the high flux region of the
core and in the recirculation line of the Nine Mile Point Unit 1 BWR. All specimens were precracked and
wedge loaded to an initial stress intensity factor of %27.5 MPa m1/2. (a) Comparison of predicted and observed crack length
versus time for furnace-sensitized type 304 stainless steel specimens in the core and recirculation line. (b) Crack
length versus time for one furnace-sensitized and two annealed specimens of type 304 stainless steel in the core. Adapted
from Andresen, P. L.; Ford, F. P.; Murphy, S. M.; Perks, J. M. In Proceedings of 4th International Symposium on

Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Theus, G. J., Eds.;
NACE: Houston, TX, 1990; pp 1–83; Andresen, P. L.; Ford, F. P. In Proceedings of 7th International Symposium on
Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995;
pp 893–908; Was, G. S.; Andresen, P. L. Corrosion 2007, 63(1), 19–45; Andresen, P. L.; Ford, F. P. In Corrosion/89; NACE:
Houston, TX, 1989; Paper no. 497; Andresen, P. L.; Ford, F. P. International Cooperative Group on Irradiation Assisted
Stress Corrosion Cracking (ICG-IASCC) Minutes, Apr 1989.

Figures 2 and 13 present the first laboratory SCC
growth rate testing ever performed on irradiated
(4 dpa) type 304SS, which was well behaved at high
corrosion potential and showed a large change in
growth rate at lower corrosion potential. One evolving concern for high-fluence stainless steels is the
prospect of a major loss of the effect of corrosion
potential and stress intensity factor, as has been
observed in unirradiated alloys possessing high
Si.52,53 The detrimental effect of Si is likely associated
with its ability to oxidize at all LWR-relevant potentials, coupled with its relatively high solubility in hot
water (i.e., it does not form a protective oxide like
Cr2O3 or Fe/Ni oxides/spinels).
These data are compared with other irradiated
and unirradiated data in Figure 2 based on simultaneous measurements of corrosion potential and crack
growth rate in fracture mechanics specimens; the
accompanying curves represent model predictions.8,13,15,16,54,55,75,76 Clearly, the in situ data compare favorably with the spectrum of unirradiated
data, and data obtained on a fracture mechanics
specimen of furnace-sensitized type 304 stainless
steel using high-energy proton irradiation to simulate the mix of neutron and gamma radiation present
in power reactors.

5.08.4 Irradiation Effects on
Microchemistry and Microstructure

5.08.4.1

Radiation-Induced Segregation

RIS describes the redistribution of major alloying
elements and impurity elements at point defect
sinks.77–87 Because IASCC is intergranular, the sinks
of greatest interest are the grain boundaries. RIS is
driven by the flux of radiation-produced defects to
sinks and is therefore fundamentally different from
thermal segregation or elemental depletion due to
grain boundary precipitation. Vacancies and interstitials are the basic defects produced by irradiation and
can reach concentrations that are orders of magnitude higher than those at thermal equilibrium, resulting in large increases in diffusion rates. If the relative
participation of alloying elements in the defect fluxes
is not the same as their relative concentration in the
alloy, then a net transport of the constituents to or
from the grain boundary will occur (Figure 14). This
unequal participation of solutes in the vacancy and/
or interstitial fluxes to sinks results in either enrichment or depletion of an alloying element at the grain
boundary. The species that diffuse more slowly by the
vacancy diffusion mechanism are enriched, and the
faster diffusers become depleted.


Irradiation Assisted Stress Corrosion Cracking

191

250
12.59

2.7 × 10−8 mm s–1

12.53

150

To 0 ppb O2@1508 h

Crack length (mm)

12.55
1.3 × 106
mm s–1

12.51
12.49

100

Constant K

Dissolved oxygen (ppb)

200

12.57

50

CT#2 - irradiated 304SS, 4 dpa

21 MPa m1/2, 220 ppb O2, 288 ЊC, 20 ppb SO4

12.47
12.45
1475

0
1495

1515

1535

1555

1575

1595

Time (h)
Figure 13 Crack length versus time for Type 304 stainless steel irradiated to 4 dpa and tested in at 21 MPa m1/2 in water at
288  C. A large reduction in crack growth rate is observed as the dissolved O2 and corrosion potential are decreased.
Reproduced from reproduced from Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al. In Proceedings of ICONE-4 Conference;
ASME International: New York, NY, 1996.

Cv

Ci

Jv

JA

Ji
JB

JB

(a)

JA

(b)

CB
CA0,B
CA

(c)
Figure 14 Schematic illustration of radiation-induced
segregation in a binary 50A–50B system showing (a) the
development of the vacancy concentration profile by the flow
of vacancies to the grain boundary balanced by a equal and
opposite flow of A and B atoms, but not necessarily in equal
numbers, (b) the development of the interstitial concentration
profile by the flow of interstitials to the grain boundary
balanced by a equal and flow of A and B atoms migrating as
interstitials, but not necessarily in equal numbers, and (c) the
resulting concentration profiles for A and B. Reproduced
from Was, G. S. Fundamentals of Radiation Materials
Science: Metals and Alloys; Springer: Berlin, 2007.


Enrichment and depletion can also occur by association of the solute with the interstitial flux. The undersized species will enrich, and the oversized species will
deplete.80 The magnitude of the buildup/depletion
is dependent upon several factors such as whether
a constituent migrates more rapidly by one defect
mechanism or another, the binding energy between
solutes and defects, the dose, dose rate, and the temperature. RIS profiles are also characterized by their
narrowness, often confined to within 5–10 nm of the
grain boundary, as shown in Figure 15 for an irradiated stainless steel.
Segregation is a strong function of irradiation
temperature, dose, and dose rate (Figure 16). Segregation peaks at intermediate temperatures since a
lack of mobility suppresses the process at low temperatures, and back-diffusion of segregants minimizes
segregation at high temperature (where defect concentrations approach their thermal equilibrium
values). For a given dose, a lower dose rate results in
a greater amount of segregation. At high dose rates, the
high defect population results in increased recombination which reduces the number of defects that are able
to diffuse to the grain boundary. Figure 16 shows the
interplay between temperature and dose rate for an
austenitic stainless steel. RIS occurs in the intermediate temperature range, and this range rises along the


Irradiation Assisted Stress Corrosion Cracking

22

18

4
JEOL 2010F
0.75 nm probe


16

3

14
12
10
8

Ni

2

Si

1

P
0

6
–20

–15 –10 –5
0
5
10 15
Distance from grain boundary (nm)


20

Figure 15 Compositional profiles across grain
boundaries obtained by dedicated scanning transmission
electron microscopy from a low-strain, high-purity 348
stainless steel swelling tube specimen irradiated to
3.4 Â 1021 n cm–2 in water at 288  C in a boiling-water
reactor. Composition profiles were measured using
a field-emission gun scanning transmission electron
microscope. Reproduced from Jacobs, A. J.;
Clausing, R. E.; Miller, M. K.; Shepherd, C. M. In
Proceedings of 4th International Symposium on
Environmental Degradation of Materials in Nuclear Power
Systems – Water Reactors; Cubicciotti, D., Ed.; NACE:
Houston, TX, 1990; pp 14–21.

temperature scale with increasing dose rate to compensate for the higher recombination rate.
In Fe–Cr–Ni alloys, the vacancy exchange
(inverse Kirkendall) mechanism successfully explains
the observed major element segregation.84,85 Studies
have shown that nickel segregates to grain boundaries
while chromium and iron deplete. The directions of
segregation are consistent with an atomic volume
effect in which the subsized solute migrates preferentially with the interstitial flux, and the oversized
solute participates preferentially in the vacancy flux.
The results are also consistent with the diffusivity of
the solutes in Fe–Cr–Ni, in which Ni is the slow
diffuser, Cr is the fast diffuser, and Fe is intermediate.
In commercial austenitic stainless steels, chromium
depletes at grain boundaries and nickel enriches,

while iron can either deplete or enrich according to
the magnitude of the diffusion coefficient relative to
the other solutes.88
RIS increases with neutron dose in LWRs and
saturates after several ($5) displacements per atom
in the 300  C temperature range. Figure 17 shows
grain boundary chromium depletion for austenitic

Homologous temperature (T/Tm)

5
Cr

Si or P concentration (wt%)

Cr or Ni concentration (wt%)

20

1105

0.8
0.7

Back diffusion
of vacancies

933

0.6


761
589

0.5
0.4

Radiation-induced
segregation

416

LWR
0.3 peak flux region
0.2
0.1
10–8

244

Recombination of
vacancies and interstitials
10–7
10–6
10–5
10–4
–1
Radiation flux (dpa s )

72


Temperature for g stainless steel (ЊC)

192

–101
10–3

Figure 16 Dependence of radiation-induced segregation
on homologous temperature and dose rate for austenitic
stainless steels.

stainless steels as a function of dose.89–97 As the slowest diffusing element, nickel becomes enriched at the
grain boundary. Since iron depletes in 304 and 316
stainless steels, the nickel enrichment makes up for
both chromium and iron depletion and can reach
very high levels up to $30 wt%.
Minor alloying elements and impurities also segregate and have been implicated in the IASCC process.
Mn and Mo strongly deplete at the grain boundary
under irradiation,98 but neither is believed to be a
significant factor in IASCC. Minor alloying or impurity
elements such as Si and P also segregate under irradiation. Silicon strongly enriches at the grain boundary
to as much as ten times the bulk (0.7–2.0 at.%) composition in the alloy99 and can be important in IASCC.
Phosphorus is present at much lower concentrations
and is only modestly enriched at the grain boundary
because of irradiation.83,98 Phosphorus tends to
segregate to the grain boundary following thermal
treatment, which reduces the amount of additional
segregation to the grain boundary during irradiation,
making the contribution due to irradiation difficult to

detect.98 Undersized solutes such as C, B, and N should
also segregate, but there is little evidence of RIS,
due in part to the difficulty of measurement. Another
potential segregant is helium, produced by the transmutation of 10B. The mobility of He is low at LWR core
temperatures, but the opportunity for accumulation
at the grain boundary is increased by segregation of
B to the boundary. Overall, the behavior of these minor
elements under irradiation is not well understood.


Irradiation Assisted Stress Corrosion Cracking

Fast neutron fluence (E > 1 MeV) ϫ 1025 n m–2
Grain boundary Cr concentration (wt%)

26

0

2

4

6

8
304 (82)
304 (13)
304 (91)
304 (92)

304 (93)
316 (82)
316 (94)
348 (91)

24
22
20

10

12

14

HP 304 (95)
CP 304 (95)
CP 316 (95)
CP 304 (17)
CP 316 (17)
HP 316 (17)
CP 304 Protons 96
CP 316 Protons 96

18
16
14
12
10
0


5

10
Dose (dpa)

15

20

Figure 17 Dose dependence of grain boundary
chromium concentration for several 300-series austenitic
stainless steels irradiated at a temperature of about 300  C.
Adapted from Asano, K.; Fukuya, K.; Nakata, K.; Kodama, M.
In Proceedings of 5th International Symposium on
Environmental Degradation of Materials in Nuclear Power
Systems – Water Reactors; Cubicciotti, D., Simonen, E. P.,
Gold, R. E. Eds.; American Nuclear Society (ANS):
LaGrange, IL, 1992; p 838; Jacobs, A. In Proceedings of 7th
International Conference on Environmental Degradation of
Materials in Nuclear Power Systems – Water Reactors;
NACE: Houston, TX, 1995; p 1021; Jacobs, A. J.;
Wozadlo, G. P.; Nakata, K.; et al. In Proceedings of 6th
International Symposium on Environmental Degradation of
Materials in Nuclear Power Systems – Water Reactors;
Gold, R. E., Simonen, E. P., Eds.; The Minerals, Metals, and
Materials Society (TMS): Warrendale, PA, 1993; p 597;
Kenik, E. A. J. Nucl. Mater. 1992, 187, 239; Nakahigashi, S.;
Kodama, M.; Fukuya, K.; et al. J. Nucl. Mater. 1992,
179–181, 1061; Jacobs, A. J.; Clausing, R. E.; Miller, M. K.;

Shepherd, C. M. In Proceedings of 4th International
Symposium on Environmental Degradation of Materials in
Nuclear Power Systems – Water Reactors; Cubicciotti, D.,
Ed.; NACE: Houston, TX, 1990; pp 14–21; Walmsley, J.;
Spellward, P.; Fisher, S.; Jenssen, A. In Proceedings of 7th
International Symposium on Environment Degradation of
Materials in Nuclear Power System—Water Reactors;
NACE: Houston, TX, 1997; p 985; Was, G. S.; Busby, J. T.;
Gan, J.; et al. J. Nucl. Mater. 2002, 300, 198.

Oversize solutes can affect the microchemistry or
microstructure of the alloy, thereby altering the
IASCC susceptibility. They are believed to affect
RIS by acting as vacancy traps, thereby increasing
the effective recombination of vacancies and interstitials and thus reducing RIS. Kato et al.100 conducted
electron irradiations of several stainless steels at temperatures of 400–500  C up to 10 dpa. Results showed
that some solutes (Zr and Hf) consistently produced a
large suppression of radiation-induced chromium

193

depletion, while others resulted in less suppression
or suppression at only certain temperatures. Fournier
et al.101 conducted irradiation of 316 containing Hf or
Pt using 3-MeV protons (400  C) and 5-MeV Ni ions
(500  C). Ni irradiations showed little effect of the
oversize impurity in reducing grain boundary chromium depletion (Cr depletion increased in the case
of Hf), but proton irradiation showed a significant
suppression of RIS of chromium at low dose (2.5 dpa)
with the effect diminishing at higher (5.0 dpa) dose.

Pt had a smaller effect on Cr. Ti and Nb similarly
produced little change in the grain boundary chromium concentration after irradiation with 3.2-MeV
protons to 5.5 dpa at 360  C. In Zr-doped 304SS,
there were no consistent results of suppression of
grain boundary chromium after 3.2-MeV proton irradiation to 1.0 dpa at 400  C.102 Neutron irradiation at
very low dose (0.5 dpa) shows a small effect of Ti and
Nb on grain boundary Cr.103 In all, the data on the
effect of oversize solutes on RIS of chromium are
inconsistent.
RIS is understandably implicated in IASCC of
stainless steels, especially in oxidizing environments,
in part because of the effect of thermal sensitization
in extensive data from lab and plant operational
experience.15–20,37,38,104,105 As shown in Figure 17,
grain boundary chromium depletion during irradiation can be severe. Figure 18(a) shows a correlation
between grain boundary chromium level and IGSCC
susceptibility in stainless steels where the grain
boundary depletion is due to thermal sensitization.106
Much data have been accumulated to support the
role of chromium depletion as an agent in IGSCC
of austenitic alloys in oxidizing conditions. Numerous studies show that, as the grain boundary chromium level decreases, intergranular SCC increases.
Typical chromium-depleted zone widths are of the
order 100–300 nm full width at half-maximum
(FWHM), providing a significant volume of depleted
material adjacent to the grain boundary.
Figure 18(b) shows a similar correlation between
grain boundary chromium level and IASCC susceptibility as measured by the percentage IG cracking
on the fracture surface during SSR experiments.
A major difference between Cr depletion profiles
resulting from RIS and those due to precipitation

reactions is that the width of the RIS profiles can
be as much as 2 orders of magnitude smaller, typically 5–10 nm. There is large scatter in the data
that makes a direct correlation difficult to support,
and differences in testing conditions undoubtedly
contribute.


194

Irradiation Assisted Stress Corrosion Cracking

%IGSCC in slow strain rate test

100
80
60

e° = 2 ϫ 10–7 S–1

␧° = 1 ϫ 10–6 S–1
40
Alloy 600
SSR tests,
23 ЊC sulfuric
acid

20

Type 304SS
SSR tests, 288 ЊC

8 ppm O2 water

0
4
(a)

6
8
10
12
14
16
18
20
Minimum grain boundary chromium concentration (wt%)

100

HP 316 SS with single
solute additions108
3xx SS17

80

%IG

60
40
20
0

10
(b)

12

14
16
18
20
Grain boundary Cr content (wt%)

irradiation. However, it should be noted that these
alloys were not irradiated, and this difference may be
important in the relevance of such experiments to
IASCC. Using 1.5–5% Si stainless steels of both
standard (e.g., 304L) base composition and synthetic
irradiated grain boundary composition, Andresen has
observed significantly increased growth rates, reduction in the benefit of lowering corrosion potential,
and very little effect of stress intensity factor between
27 and 13 ksi in1/2.
The data on impurity segregation effects on
IASCC remain inconclusive. Extensive experiments
have been conducted to isolate the effect of particular
impurities such as S, P, C, N, and B in IASCC, but
none have yielded unambiguous results. Sulfur has not
been found to segregate under irradiation, and, while
P thermally segregates to a significant extent, irradiation-induced P segregation is small in comparison. C,
N, and B cannot be measured in STEM; N and B are
very difficult to identify in AES; and C is a common
contaminant. Overall, it has been a challenge to establish a link between impurity element segregation and

IASCC in austenitic stainless steels.

22

Figure 18 Effect of grain boundary Cr content on
intergranular stress corrosion cracking for (a) sensitized
stainless steel and Alloy 600 and (b) irradiated stainless
steels. Slow-strain-rate tests are tests in which the
specimen is monotonically strained versus time.
Reproduced from Bruemmer, S. M.; Was, G. S. J. Nucl.
Mater. 1994, 216, 348.

Among the minor alloy elements, only Si is known
to segregate to high levels, and Si segregation is
correlated with IASCC. Experiments by Busby
et al.107,108 on a high-purity 316 base alloy doped
with 1 wt% Si showed severe IASCC in NWC and
in primary water after irradiation to 5.5 dpa at 360  C.
STEM measurements of grain boundary Si confirm
levels up to 6 wt%. Past studies comparing Auger
electron spectroscopy (AES) and scanning transmission electron microscopy (STEM) results have shown
that the actual concentration of Si at the grain boundary plane may be as high as 15–20 wt%. Though the
electron beam probe in STEM is very small, the
measurement underpredicts the concentration at
the grain boundary by a factor of 2 to 3. Yonezawa
et al.109–112 and Li et al.113 have provided extensive
evidence to show that increased Si in stainless steel
results in increased IGSCC in alloys tailored to imitate the composition of grain boundaries under

5.08.4.2 Microstructure, Radiation

Hardening, and Deformation
5.08.4.2.1 Irradiated microstructure

The microstructure of austenitic stainless steels
under irradiation changes rapidly at LWR service
temperatures. Point defect clusters (called ‘black dot
damage’ when electron optics could not resolve the
details) begin to form at very low dose, dislocation
loops and network dislocation densities evolve with
dose over several displacements per atom, and the
possibility exists for the formation and growth of
He-filled bubbles, voids, and precipitates in core
components in locations exposed to higher dose and
temperatures.114–120 Below 300  C, the microstructure is dominated by small clusters and dislocation
loops. Near 300  C, the microstructure contains
larger faulted loops plus network dislocations from
loop unfaulting and cavities at higher doses.
The primary defect structures in LWRs are
vacancy and interstitial clusters and Frank dislocation
loops. The clusters are formed during the collapse of
the damage cascade associated with primary and secondary atom collisions after an interaction with a
high-energy particle. The larger, faulted dislocation
loops nucleate and grow as a result of the high mobility of interstitials. The loop population grows in size
and number density until absorption of vacancies and


Irradiation Assisted Stress Corrosion Cracking

interstitials equalize, at which point the population
has saturated. Figure 19 shows the evolution of loop

density and loop size as a function of irradiation dose
during LWR irradiation at 280  C. Note that saturation of loop number density occurs very quickly,
by $1 dpa, while loop size continues to evolve up
to $5 dpa. The specific number density and size are
dependent on irradiation conditions and alloying elements, but the loop size rarely exceeds 20 nm and
densities are of the order of 1 Â 1023 m–3.
The hardening process and IASCC susceptibility
are influenced by small defects. The traditional view
that small defect clusters are predominantly faulted
interstitial loops and vacancy clusters121 may be inaccurate. Analysis of recent postirradiation annealing
experiments by Busby et al.122 and Simonen et al.123
suggests that there are at least two types of defects
with different annealing characteristics: vacancy and
interstitial faulted loops, each with different annealing
kinetics. The step change in hardness as a function
of annealing time suggests that the density of vacancy
loops is perhaps much higher than previously believed,
and higher than the density of interstitial loops.122
Above 300  C, voids and bubbles may begin to
form, aided by the increased mobility of point defects
at the higher temperature. The dislocation structure
will evolve into a network structure as larger Frank
loops unfault. The reduction in the sink strength
of the dislocation loops aids in the growth of voids
and bubbles. While their size and number density
increase with temperature, the dislocation microstructure continues to be the dominant microstructure component over the temperature range expected
for LWR components (<350  C).

Loop density (m–3)


11
1023

Loop density

1022

Loop size

10
9
8
7

1021

1020
0

6

Type 304SS heats
Type 316SS heats
1

2

3
4
5

6
Irradiation dose (dpa)

5
7

8

Average loop diameter (nm)

12

4

Figure 19 Measured change in density and size of
interstitial loops as a function of dose during light water
reactors irradiation of 300-series stainless steels at
275–290  C. Reproduced from Bruemmer, S. M.; Simonen,
E. P.; Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L.
J. Nucl. Mater. 1999, 274, 299.

195

Irradiation can also accelerate or retard the
growth of second phases, modify existing phases, or
produce new phases, although these processes are
more pronounced above 400  C. In stainless steels,
the principal second phase is chromium carbides,
which are stable under irradiation. In high-strength
Ni-base alloys, the second phases can undergo several types of transformations: g0 can dissolve, g00 can

dissolve and reprecipitate, and Laves phase can
become amorphous. A key factor in phase formation
in austenitic stainless steels under LWR operating
conditions is RIS, which can induce the formation
of phases by exceeding the local solubility limit. Was
et al.124 irradiated a high-purity stainless steel containing 1 wt% Si with 3.2-MeV protons to 5.5 dpa at
360  C. They observed the formation of g0 (Ni3Si) in
the matrix but not on the grain boundary, which is
puzzling since the concentration of both Si and Ni is
higher in the boundary. This was also observed in a
similar alloy irradiated with neutrons to 7 dpa at
300  C.125 g0 is a coherent precipitate that can significantly strengthen the matrix and has the potential to
alter the deformation behavior in the unirradiated
and irradiated conditions.
Oversize solutes can also affect the irradiated
microstructure by mechanisms similar to RIS. Proton
and nickel ion irradiations show that the addition of
Hf to a 316SS-base alloy increased loop density,
decreased loop size, and eliminated voids.101 Platinum addition to 316SS resulted in no change in loop
density and a small increase in loop size, but
increased void size and density. The good agreement
between proton and Ni ion irradiation results indicates that the major effect of the oversized solute is
not due to the cascade (where there are large differences between proton and nickel ion irradiation),
but rather is due to the post-cascade defect partitioning to the microstructure evolution. Electron
irradiation experiments by Watanabe et al.126 and
proton irradiation experiments by Was et al.124
showed that stainless steel with Ti additions had
slightly lower dislocation loop densities and larger
sizes compared to the base alloy. Nb increased only
the loop size. In contrast to the base alloy, neither

the Ti nor the Nb-doped alloys formed voids under
the conditions tested. Zirconium addition to 304SS
resulted in reduced hardness, decreased loop density, and no change in loop size in proton irradiation
to 1.0 dpa at 400  C and compared to the base
alloy.127 Zirconium-containing samples also had a
lower void density with no change in void size as
compared to the base alloy.


196

Irradiation Assisted Stress Corrosion Cracking

5.08.4.2.2 Radiation hardening

1200
1000

Yield strength (MPa)

The dislocation loop microstructure formed during
irradiation hardens the alloy, which correlates with
increased SCC susceptibility. Under an applied
stress, dislocation network interacts elastically with
the dislocation loops, producing an increase in the
yield strength of the alloy, which can be detected in
tensile tests or indentation hardness measurements.
This is accompanied by a decrease in elongation,
which is affected more than reduction in area because
necking occurs relatively early in most tensile tests of

irradiated stainless steel. Fracture toughness also
decreases with fluence.
The increase in yield strength with dose in 300
series stainless steels irradiated around 300  C is
shown in Figure 20. The yield strength can reach
values up to five times the unirradiated value by about
5 dpa, and its increase follows a square root dependence on dose. Both the source hardening model128
and the dispersed barrier hardening model129 provide reasonable correlations between hardening
and the dislocation loop microstructure. In the dispersed barrier hardening model, the increase in hardness is proportional to (Nloop  dloop)1/2, where Nloop
is the loop number density and dloop is the loop
diameter.
Deformation changes dramatically with radiation.
Homogeneous deformation at low dose is replaced by
heterogeneous deformation at higher doses as the
defect microstructure begins to impede the motion
of dislocations. Plasticity becomes localized to narrow channels that have been cleared of defects by
preceding dislocations, providing a preferred path for
subsequent dislocation motion. The channels are
very narrow (<100 nm) and closely spaced (<1 mm)
and typically run the full length of a grain, terminating at the grain boundaries.130 Dislocation channeling
can cause localized necking and a sharp reduction in
uniform elongation.131
Hardening has been cited as a key factor in IASCC
susceptibility. Figures 2 and 21(a) show that increasing cold work and yield strength increases the crack
growth rate in nonsensitized austenitic stainless steel
tested in 288  C BWR NWC.132 This dependence
carries over to irradiated material. Figure 21(b)
shows a correlation between yield strength and susceptibility to IASCC in SSR tests although the correlation is complicated by other radiation-induced
changes, especially RIS. The correlation is better at
high values of yield strength (>800 MPa) and at very

low values (<400 MPa), with more scatter at intermediate values (400–800 MPa).

800
600
400
200
0

0

2

4
6
Dose (dpa)

8

10

Figure 20 Irradiation dose effects on measured
tensile yield strength for several 300-series stainless
steels, irradiated and tested at a temperature of about
300  C. Adapted from Singh, B. N.; Foreman, A. J. E.;
Trinkaus, H. J. Nucl. Mater. 1997, 249, 103; Seeger,
A. In Proceedings of 2nd United Nations International
Conference on the Peaceful Uses of Atomic Energy;
United Nations: New York, NY, 1958; Vol. 6, p 250;
Bruemmer, S. M.; Cole, J.; Carter, R.; Was, G. S. In
Proceedings of 6th International Symposium on

Environmental Degradation of Materials in Nuclear
Power Systems – Water Reactors; Gold, R. E.,
Simonen, E. P. The Minerals, Metals, and Materials Society
(TMS): Warrendale, PA, 1993; p 537; Bloom, E. E.
In Radiation Damage in Metals; Peterson, N. L.,
Harkness, S. D., Eds.; ASM International: Materials Park,
OH, 1975; p 295.

Hardening alone is not sufficient to explain
IASCC. Annealing experiments tracked the change
in hardness and dislocation loop microstructure versus annealing condition.133 Figure 22 shows the
change in hardness and the change in the dislocation
loop line length (Navg  davg) along with the change
in IASCC susceptibility in SSR tests as annealing
progresses.120,134,135 Except for very short annealing
times, the loop line length and the hardness closely
track each other, as expected if the loop structure is
controlling the hardening. At very small values of
(Dt)1/2, the hardening remains flat before decreased
with annealing time. While both hardening and
cracking are reduced with increased annealing, the
behavior of the hardness does not fully explain the
response.
Busby et al.122 postulated that the removal of very
small defects at short annealing times may account
for the IASCC behavior. Short times at high temperature (500  C) may preferentially remove the small


Irradiation Assisted Stress Corrosion Cracking


10–7
10–8
10

100
Stress corrosion cracking,
nonsensitized austenitic stainless steels
in simulated BWR water 288 ЊC

HP 316SS with single
solute additions 108
3xxSS 17

80

304, 1.4301
347, 1.4550
321, 1.4541
316Ti, 1.4571

60
% IG

Stress corrosion cracking growth
rate, Δa/Δt, (m s–1)

10–6

197


–9

40
10–10
20
10–11
Intergranular

10–12
0
(a)

200

increasing fraction of transgranular SCC

400 600 800 1000 1200 1400 1600
Yield strength, Rp0.2 (MPa)
(b)

0
200

400

600

800

1000


Yield stress (MPa)

Percentage of as-irradiated feature
remaining after heat treatment

Figure 21 Effect of yield strength on intergranular stress corrosion cracking. (a) Crack growth rate of cold-worked,
unirradiated 300-series stainless steels tested in 288  C simulated boiling-water reactor (reproduced from Speidel, M. O.;
Magdowski, R. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear
Power Systems – Water Reactors; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; The Minerals, Metals & Materials
Society: Warrendale, PA, 1999; p 325) and (b) percentage intergranular stress corrosion cracking in slow strain rate
tests on 300-series stainless steels where hardening is by irradiation (adapted from Bruemmer, S. M.; Simonen, E. P.;
Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater. 1999, 274, 299; Busby, J. T.; Kenik, E. A.; Was, G. S.
J. Nucl. Mater. (in press)).

120
RIS

100

IASCC
Hardness
Loop line length
RIS

80
60
40

Hardness

and loop
line length
IASCC

20
0
0

0.0005 0.001 0.0015 0.002 0.0025 0.003
Iron diffusion distance, (DFet)1/2, (cm)

Figure 22 Removal of radiation-induced segregation
and dislocation microstructure as measured by loop line
length and hardness with extent of annealing as
measured by (Dt)1/2 for iron, where D is the diffusivity of
Fe and t is time. This relationship accounts for annealing
at different times and temperatures. The effect on
irradiation-assisted stress corrosion cracking in slow
strain rates is also shown. Adapted from Busby, J. T.;
Was, G. S.; Kenik, E. A. J. Nucl. Mater. 2002, 302, 20;
Edwards, D. J.; Simonen, E. P.; Garner, F. A.;
Greenwood, L. R.; Oliver, B. M.; Bruemmer, S. M. J. Nucl.
Mater. 2000, 317, 32; Katsura, S.; Ishiyama, Y.; Yokota, N.;
et al. In Corrosion/98; NACE: Houston, TX, 1998;
Paper no. 132; Jacobs, A.; Wozadlo, G. P.; Gordon, G. M.
Corrosion 1995, 51, 10, 731.

defect clusters either by annihilation or by spontaneous dissociation. The dislocation loops may absorb
the free vacancies and interstitials, thus adding to
their line length. The loss of the small defect clusters

will be offset by the growth of Frank loops producing
no net change in measured hardness or yield strength.
Despite a lack of hardness change, this process may
alter the deformation mode at the local level by
removing the small obstacles to dislocation motion,
thus changing the character of localized deformation.
Hash et al.136 showed that hardening is not the sole
factor in IASCC, by testing a series of samples of
commercial purity 304SS with nominally the same
hardness but with different contributions from cold
work and proton irradiation. At the extremes were a
sample that was cold-rolled to a 35% reduction in
thickness and no irradiation and one with 1.67 dpa
irradiation and no cold work. Three samples had
varying amounts of cold work (10, 20, and 25%) and
corresponding amounts of irradiation dose (0.55,
0.25, and 0.09 dpa) to give a hardness level that was
within 5% over all samples. SCC susceptibility was
measured by the amount of IG cracking in an SSR
test in 288  C BWR NWC. IASCC susceptibility was
not constant, as would be expected if hardness were
the only factor, with cracking observed in only the
two highest dose samples (0.55 dpa with 10% cold


198

Irradiation Assisted Stress Corrosion Cracking

work and 1.67 dpa with no cold work), irrespective of

their hardness, Figure 23. The amount of cracking in
the lower dose sample was higher than in a companion sample at the same dose but without cold work,
indicating that cold work can enhance the IASCC
susceptibility. These results also suggest that RH may
promote crack initiation, as it does crack growth.
Combined with the annealing results, they suggest
that other factors besides the hardness level and yield
strength – such as RIS and deformation mode – play a
role in the IASCC process.
5.08.4.2.3 Deformation mode

Results of IASCC experiments on proton-irradiated
samples over a wide range of doses and alloys have
consistently showed that high Ni alloys have high
IASCC resistance in SSR tests. In particular, Ni concentrations >18 wt% are highly resistant to IASCC
compared to 304SS with 8 wt% Ni. Kodama et al.137
showed that there is a good, but not perfect, correlation between Ni equivalent and IASCC. Notable
exceptions are alloy 800138 and a Fe–20Cr–25Ni–
1Nb alloy used in AGR that experienced IGSCC.139
Ni may affect IASCC directly through a change in
composition or indirectly by changing the slip character. Higher nickel content in stainless steel increases
the stacking fault energy (SFE) significantly, producing a change in slip from planar to wavy. Swan et al.140
studied the effect of SFE on slip behavior using a
series of Fe–18Cr–xNi alloys where 8showed that for the 8% Ni alloy, the slip was entirely
planar and, as Ni increases, the cross-slip increases. By
20% Ni, there was no evidence of planar slip and the
deformation microstructure consisted of a web of
Dose (dpa)

0

0.09

0.25

0.55

dislocation tangles, which is evidence of wavy slip in
a high SFE material. Figure 24 shows that planar slip
yields greater dislocation interaction with grain
boundaries than does a wavy slip.
SFE has been linked to SCC resistance in stainless
steels by Thompson and Bernstein, who found that
increasing SFE correlates well with increased reduction in area and decreased SCC susceptibility.141 The
IASCC susceptibility of several irradiated stainless
steels is plotted as a function of SFE in Figure 25
and shows that there is a good correlation between
SFE and IASCC using Rhode’s142 and Schramm’s143
correlations for SFE. These correlations differ in the
elements included and the weights given in terms of
nickel equivalent (NiEq), but are unable to account
for some minor elements and therefore may deviate
substantially from the true SFE in some cases. There
is much scatter in the Ni equivalent plots, similar to
that shown in the IGSCC dependence on grain
boundary Cr and yield strength. One potential source
of the scatter is the inherent variability in initiationdominated phenomena. Another is that, while grain
boundary Cr and yield strength are measured, NiEq
and SFE are generally calculated and with considerable uncertainty. So they may be useful in identifying

transitions in behavior, but are not as reliable
quantitatively.
Microstructure can also influence the deformation
mode. Farrell et al.144 conducted neutron irradiation
of 316 stainless steel at temperatures between 65
and 100  C to doses of <1 dpa, and then characterized the deformation behavior. With increasing dose,
the propensity for dislocation channeling increased
(Figure 26). The volume of material occupied by
channels increases rapidly with dose and saturates

1.67

Extent
of IG
cracking

1mm

1mm
Low SFE ® planar

Cold work 35%

e

a

25%

20%

10%

High SFE ® wavy slip

0%

Figure 23 Degree of irradiation-assisted stress corrosion
cracking in Type 304 stainless steel samples with the same
hardness but with different combinations of hardening by
cold work and irradiation using 3.2-MeV protons at 360  C.
Reprinted, with permission, from 21st International
Symposium on Effects of Radiation on Materials,
copyright ASTM International, 100 Barr Harbor Drive,
West Conshohocken, PA 19428.

Grain boundary

Grain boundary

Figure 24 Micrographs and corresponding schematic
illustrations of planar and wavy slip in the vicinity of a grain
boundary (micrographs from Swann140).


Irradiation Assisted Stress Corrosion Cracking

at <0.5 dpa. These data show that the defect microstructure created by irradiation can induce planar
deformation in the form of narrow dislocation
channels.

The importance of slip localization in IASCC may
be in the way in which the dislocations interact with
the grain boundary. In planar slip, well-defined and
separated slip bands or dislocation channels (for irradiated materials) transmit dislocations during plastic
deformation. These slip bands or channels terminate
at grain boundaries where dislocations are fed into
the grain boundary region versus forming a tangled
dislocation network within the grain. In the more
traditional view of dislocation infusion into grain
boundaries, the pileup of dislocations in the intersecting channel creates progressively higher stresses
at the grain boundary at the head of the pileup. If the
stress exceeds a critical value, separation of the grain
boundary could occur according to a Stroh (wedge)
cracking type of mechanism.145 This cracking process
could occur regardless of the environment and may
in fact be the mechanism that occurs in some of the
IG cracking observed in very highly irradiated steels
in inert environment.131 At lower fluences, the stress
at the grain boundary may promote rupture of the
oxide film, leading to exposure of the metal to the
solution and subsequent corrosion and IASCC.
Alternatively, deformation could occur in the
boundary plane, which would rupture the oxide
film and promote IASCC. Alexandreanu146 showed
that dislocation absorption by grain boundaries in

100

IGSCC (%)

80
60
SFE = 1.2 + 1.4% Ni + 0.6% Cr
+ 17.7% Mn – 44.7% Si
Rhodes and Thompson142

40
20
0
0

(a)

50
100
Stacking fault energy (mJ m–2)

150

100

IGSCC (%)

80
60
SFE = –53 + 6.2% Ni + 0.7% Cr
+ 3.2% Mn + 9.3% Mo
Schramm and Reed143

40

20
0
0

(b)

50
100
150
Stacking fault energy (mJ m–2)

200

Figure 25 Irradiation-assisted stress corrosion cracking
susceptibility as measured by percentage IG cracking as a
function of stacking fault energy determined using
(a) Rhode’s correlation (reproduced from Rhodes, C. G.;
Thompson, A. W. Met. Trans. 1977, 8A, 1901) and
(b) Schramm’s correlation (reproduced from Schramm, R. E.;
Reed, R. P. Met. Trans. A 1975, 6A, 1345).

0.3

1015

ND (cm–2)

0.15

1014

Channel area

0.1

ORNL NERI
316 SS
Neutron irradiated

0.05
0

0.5

1
Dose (dpa)

1.5

2

0.04

0.03

0.02

0.01

Channel area (nm2 nm–2)

0.2

0.05
Strain hardening exponent (n)

0.25
Strain hardening
exponent (n)

1013
0

199

0

Figure 26 Variation in dislocation channel area, dislocation loop line length, and strain hardening exponent as a function of
dose for neutron-irradiated 316SS. ‘ND’ is the product of the number density and diameter of Frank dislocation loops.
Reproduced from Farrell, K.; Byun, T. S.; Hashimoto, N. Mapping flow localization processes in deformation of irradiated
reactor structural alloys, Report ORNL/TM-2002/66; Oak Ridge National Laboratory, July 2002.


200

Irradiation Assisted Stress Corrosion Cracking

nickel-base alloys leads to deformation in the grain
boundary, usually by boundary sliding, and promotes
IGSCC. He observed random grain boundaries
exhibited more sliding, and boundaries where sliding

occurred were four times more susceptible to IGSCC
in primary water at 360  C than boundaries that did
not slide. This same process could explain how dislocation injection into grain boundaries by planar slip
or dislocation channeling can result in IASCC.
The concept that slip planarity controls the deformation mode and IASCC is consistent with observations of IASCC in 304 and 316SS. 316SS has, in some
cases83 been shown to be more resistant to IASCC
than is 304SS, which is consistent with its higher SFE.
Observations by Bailat et al.147 on neutron-irradiated
samples and by Busby et al.148 on proton-irradiated
samples show clear dislocation channel patterns on
the surface of 304SS samples that cracked in SSR
tests but an absence of those patterns on the 316SS
samples that did not crack. In addition, SSR tests on
high Ni alloys (with SFE of $40 mJ m–2) showed no
IG cracking and also no dislocation channeling.
Further support for the influence of slip planarity

on IASCC comes from the annealing studies cited
earlier. On samples that underwent either a short,
low-temperature anneal or were tested in the asirradiated condition, IG cracking was accompanied
by a high density of dislocation slip bands on the
surface of $65 mmÀ2. Annealing at 500  C for
45 min resulted in the elimination of IG cracking
and a reduction in the surface slip band density
to $25 mm–2.
In the annealing experiments described earlier,
hardening by irradiation was much more effective in
initiating IGSCC than was cold work, where the
dislocation structure will likely be more cellular
than planar. Bruemmer et al.149 showed a correlation

between slip band intersection with the walls of a
growing crack and the accompanying steps on the
oxide on the walls. Figure 27(a) shows a TEM
micrograph of SCC in a cold-worked 316 stainless
steel baffle bolt. The slip bands intersect the walls of
the narrow crack at 45  to the crack growth direction.
The slip band–crack wall intersections are also coincident with steps in the oxide as shown schematically
in Figure 27(b). This suggests that the flow of

Metal
m/o interface

Oxide

100 nm
Crack wall
oxide growth
Crack

Grain boundary
deformation
by dislocation

Figure 27 Micrograph of (a) deformation bands intersecting a crack in a baffle bolt (reproduced from Thomas, L. E.;
Bruemmer, S. M. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power
Systems – Water Reactors; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; Metallurgical Society of the American Institute of
Mining, Metallurgical, and Petroleum Engineers (AIME): Warrendale, PA, 1999; p 41). (b) A schematic of the role that the
deformation bands may be playing in irradiation-assisted stress corrosion cracking.

Irradiation Assisted Stress Corrosion Cracking

250

200
Stress (MPa)

dislocations along the slip steps and into the grain
boundary may have been responsible for discontinuous growth of the crack along the grain boundary.
Thus, low SFE and irradiation can both lead to
planar or localized deformation, which terminate at
grain boundaries, which must be accommodated by
the grain boundary. This results in shear strain that
can rupture the oxide film and promote the initiation
of intergranular cracks.

201

150

100

50

5.08.4.3 Radiation Creep and Stress
Relaxation

0
0

5
10
Displacement level (dpa)

15

Figure 28 The effects of radiation-induced creep on load
relaxation of stainless steel at 288  C. Reprinted, with
permission, from 16th International Symposium on Effects
of Radiation on Materials, copyright ASTM International, 100
Barr Harbor Drive, West Conshohocken, PA 19428.

1.0
0.9

Fraction of stress remaining

At LWR temperatures, radiation creep results from
diffusion of the radiation-produced vacancies and
interstitial atoms to dislocations, enhancing the
climb-to-glide process that controls time-dependent
deformation. Radiation creep can be both beneficial
and detrimental. Benefits accrue from relaxation of
constant displacement stresses, for example, weld
residual stress and in loaded bolts and springs. However, under these conditions – and more so under
constant load conditions – radiation creep also
induces elevated creep rates, including grain boundary sliding, that help initiate and sustain SCC.
Figures 28 and 29 show examples of load relaxation under constant displacement conditions, a process that is quite reproducible over a wide range of
materials and loading modes, and generally produces sizeable (>50%) load relaxation within a few
displacements per atom. Thus, for example, in areas

of the BWR shroud that receive a moderate neutron
flux, if SCC initiation does not occur early in life
(e.g., by 1 dpa), the relaxation in residual stress
should diminish the likelihood of cracking later in
life. Because the effect of relaxation is significant, it
tends to offset the detrimental effects of RIS and
RH. Thus, it is not surprising that the incidence of
SCC in BWR shroud welds, where the neutron flux
can vary by 2 orders of magnitude because of the
varying proximity of the fuel, does not show a strong
correlation with fluence.
Radiation creep relaxation also affects PWR baffle
bolts, which are subject to large variations in fluence
and temperature.40,41 Baffle bolts in high flux regions
can accumulate more than 3 dpa yearÀ1, and thus the
preload will rapidly decrease during the first several
years. Thus, SCC probably initiates early in life
(before significant radiation creep relaxation occurs)
or later in life when reloading occurs from differential swelling in the (annealed) baffle plates relative to
the (cold-worked) baffle bolts.

Preload
23.6 N
36.5 N

0.7

0.5

Walters -1977

370 ЊC, EBR-II
X-750 springs
0.3

0

1

2
3
Neutron dose (dpa)

4

5

Figure 29 Radiation creep relaxation of X-750 springs at
370  C. Reproduced from Walters, L. C., Ruther, W. E.
J. Nucl. Mater. 1977, 68, 324.

While difficult to prove, the elevated and sustained
deformation rates associated with radiation creep can
only accentuate susceptibility to SCC. Estimates of
crack tip deformation rates15 indicate the radiation
creep is not a large contributor in actively growing
cracks, but rather it is expected to promote crack
initiation and to sustain crack growth (or promote
crack reinitiation, if an existing crack does arrest).