JOURNAL OF MATERIALS SCIENCE LETTERS 11 (1992) 1085-1086
On the meaning of thresholds in environmentally assisted cracking
J. TORIBIO, A. M. LANCHA
Department of Materials Science, Polytechnical University of Madrid, ETSl Caminos, Ciudad Universitaria,
28040 Madrid, Spain
Experimental results on environmentally assisted
cracking are usually given in terms of fracture
mechanics parameters, as a plot of crack growth rate
versus stress intensity factor. This curve has three
stages representing different physical processes of
environmental cracking: stage I corresponds to the
threshold level or minimum load required to propa-
gate the crack in a given environment; stage II shows
a more-or-less stress intensity-independent growth
process; in stage III there is an approach to the K~
level, and the crack growth rate rises sharply up to
the fracture point. Each of the three stages can be
modelled according to the nature of the crack
growth rate at that level of loading [1].
From the engineering design point of view it is
very interesting to determine, as accurately as
possible, the threshold stress intensity level for the
given aggressive environment (Kth). It may be
denominated according to the specific mechanism of
crack growth [Kiscc for stress corrosion cracking
(SCC) and KIHAC for hydrogen-assisted cracking
(HAC)]. Calculations of stress intensity threshold
levels for hydrogen-induced crack growth on the
basis of hydrogen diffusion models can be found in
[2, 3].
This letter analyses the meaning of thresholds in
environmentally assisted cracking. Defining the
threshold as the stress intensity level below which no
crack propagation takes place, it would be expected
for this concept to have an intrinsic character,
dependent only on the specific material and environ-
ment, but not on mechanical variables. However,
this letter makes clear that precracking conditions
can significantly affect the environmentally assisted
cracking threshold.
To achieve this, crack growth rate tests were
performed on a commercial eutectoid pearlitic steel
in an aqueous solution which promoted environ-
mentally assisted cracking. Prismatic specimens
3.85 mm × 7.70 mm× 100.00 mm were machined
from 12 mm diameter commercial bars, with a
starter notch in its central section perpendicular to
the bar axis. Precracking was produced by axial
fatigue in air environment before the fracture test in
aggressive environment. Load was applied by means
of a three-point bending device with a distance
between supports of 61.60 mm. Strain was increased
stepwise to obtain different points of the curve crack
growth rate versus stress intensity factor. The
aggressive environment was an aqueous solution of
1 g l-lCa(OH)2 plus 0.1 g 1-1NaC1 (pH 12.5). Test-
ing was performed at room temperature (between 16
and 22 °C) and at constant potential by using an
0261-8028 © 1992 Chapman & Hall
electrochemical device consisting of a potentiostat
and a classical three-electrode assembly: metallic
sample (work electrode), saturated calomel elec-
trode (SCE, reference electrode) and platinum wire
(counterelectrode), as described in [4]. Two poten-
tials were used: -600 mV versus SCE (anodic) and
-1200mV versus SCE (cathodic). The former
corresponds to anodic dissolution or pure SCC,
whereas the latter refers to HAC [5].
To evaluate the influence of the precracking
procedure, two types of samples were prepared
using different fatigue precracking loads during the
last step Oust before the fracture test). The max-
imum stress intensity factors during fatigue pre-
cracking were Kmax = 0.25Kit and 0.50Klc , where
Kit is the critical stress intensity factor for the
prismatic three-point bending specimens used in the
experimental programme (a value greater than the
fracture toughness of the material Kic, since the
thickness is not enough to fulfil the plain strain
condition requirements). As a consequence there
were two different distributions of compressive
residual stresses in the vicinity of the crack tip, and
therefore two distinct degrees of environmentally
assisted cracking.
Results for E = -1200 mV versus SCE (cathodic
regime: HAC) are plotted in Fig. 1. For a maximum
fatigue precracking load Km,x = 0.25K1~ the
threshold stress intensity factor for HAC is
KIHAC = 0.35Klc. Below this value no crack propa-
gation is observed. For higher values of KI there is a
plateau in the curve da/dt versus KI for which the
crack growth rate remains constant and equal to
1.5 x 10 -7 ms -1 (slow crack growth rate) for the
different values of K I. When KI exceeds 0.55Klc
there is a sudden increase in the crack growth rate,
up to about 10 -3 m s-1. This part of the curve is not
horizontal, but slopes slightly. For higher KI values
final fracture takes place. For a maximum fatigue
precracking load Kma x -0.50Klc the threshold
stress intensity factor for hydrogen-assisted cracking
is KInAC = 0.58K10. No plateau is observed for
higher KI, but a sudden increase to about 10 -3 m s -I
is observed, as in the previous curve. It is important
to notice the large difference in the threshold value
(0.35K1c compared with 0.58Klc) under cathodic
regime, depending on the maximum stress intensity
factor, Kmax, during the last step of fatigue precrack-
ing.
Experimental results for E---600mV versus
SCE (anodic regime: SCC) are shown in Fig. 2. For
a maximum fatigue precracking load Kmax=
0.25K1o the threshold stress intensity factor for SCC
1085