|
Pitting Corrosion Diagrams for Stainless
Steels
By: J W
Fielder and D R Johns,
British Steel Technica, Swinden Laboratories,
Moorgate Rotherham
Condensed from UK Corrosion ’89 Blackpool,
November 8-10, 1989.
Introduction
Stainless steels can be
prone to various different types of corrosion, dependent on the
combination of alloy composition and operating environment. In service
situations the localized corrosion mechanism most commonly encountered
is almost certainly pitting corrosion. It can be extremely difficult
to predict the incidence of pitting corrosion, particularly as the
complexity of the corrosive environment increases, and even more so
when fluctuations occur in the environmental parameters, as often
arises in process plant. It is clearly not possible to accumulate data
by evaluating the performance of the wide range of stainless steels in
the vast number of different and often variable corrosive environments
found in industry.
Various guides relating
to the pitting performance stainless steels have been published over
the years, perhaps the most common being the pitting index formula, ie
PI = %Cr + (3.3 x %MO) +
(X%N)
Where X = 11 to 30
This provides a good
qualitative indication of the comparative pitting resistance of
different alloys, but cannot be easily used to give a quantitative
prediction of performance in relation to a particular corrosive
environment. More useful information is provided by laboratory
immersion testing or by critical pitting temperature data, but these
are generally limited to a few standard test solutions, such as 10%
FeCl3 which has been shown to be of very limited relevance
to most service environments. In contrast, testing in more
representative solutions either requires unacceptably long exposure
periods, or provides insufficient information to form the basis of an
accurate prediction.
Hence, performance must
be predicted from an understanding of corrosion behavior in a
restricted range of environments which form the basis of most service
conditions. The effect of other significant environmental parameters
can then be superimposed onto the base data, using a combination of
available information relating to the effect of such parameters, and
service experience.
The most significant
environmental conditions which influence the pitting corrosion
behavior of stainless steels are the chloride ion concentration,
temperature and pH level. This article describes a program of work in
which the pitting corrosion susceptibility of a range of stainless
steels has been determined by a simple electrochemical method within a
matrix of these key environmental parameters. The resultant data have
been processed to produce pitting corrosion engineering diagrams which
are designed to provide corrosion engineers with a more systematic aid
to materials selection than currently available data. The translation
of the raw electrochemical data directly into useful engineering
diagrams has been simplified and considerably accelerated by the
development of a computer program.
Experimental Procedure
The materials selected
for examination were representative of the range of standard and high
alloy stainless steels. The test solutions were prepared freshly for
each test, and the pH and chloride ion concentration was adjusted
using the sodium chloride and hydrochloric acid, except at low
chloride ion concentrations where monochloroacetic acid was used. Four
chloride ion levels of 0.03, 0.1, 1.9 and 10% were used, each at four
pH levels of 7, 4.5, 3 and 1.5. Slight acidity of the distilled water
was neutralized by sodium hydroxide for test solutions at pH = 7.
Test samples were
suspended in the electrochemical cell using the attached wire and were
connected to the potentiostat, along with the reference electrode and
a platinum counter electrode. The test solution within the cell
covered each electrode and was deaerated with nitrogen, and solution
temperatures of 25, 35, 45 or 55 C were used, controlled by immersion
of the cell in regulated water baths and measured by a thermometer in
the actual test cell. The test sample was initially polarized to -400
mV (v SCE) for 1 minute, after which the potential raised at 1 mV/sec.
Current flowing in the cell was monitored, and the test was continued
until the anodic current density at the specimen exceeded 500 uA/cm2
resulting from either pitting, crevicing at the metal/paint interface
or transpassivity. The potential at which this occurred was noted as
the pitting potential unless pitting, or any other form of corrosion,
was not evident ton the test surface, in which case transpassivity was
assumed. All samples were examined visually after testing, to ensure
that a pit had developed when indicated by the test result. If
crevicing was apparent, the result was discarded and the test repeated
with a repeated surface.
Tests were performed in
order of decreasing severity of the test solution. Where it was
evident that particular alloy was not susceptible to pitting
corrosion, no further testing under less severe conditions was carried
out.
Experimental Results
The anodic polarization
determination produced a characteristic relationship between potential
and the measured current density, the most important feature being a
rapid rise in current density above a certain critical potential, as
shown in
Figure 1. Where this potential exceeded 1000 mV transpassivity was
assumed. For lower potential values the specimen test surface was
examined carefully for evidence of pitting. If crevice corrosion was
observed at the pain/metal interface, the result was discarded and the
surface reprepared and retested. The onset of crevice corrosion, which
occurred only occasionally, was generally apparent from the
unexpectedly low break potentials on the anodic polarization curve. In
general, pit initiation, as revealed by the first increase in current
density from the passive level, was followed by sustained and rapid
propagation and, hence, rapid current rise. It was found, in
particular with the more resistant alloys, the early pit propagation
could be slow, with particular repassivation occurring and a
correspondingly slow rise in current with increasing potential.
When comparing a wide
range of alloy compositions, differences in pitting behavior are not
adequately taken into account if the pitting potential was taken as
the potential of the first current increase. Experience shows that,
once a current density of 500 uA/cm2 was reached,
propagation would be sustained. The potential at which this current
density was reached was, thus, taken as the pitting potential and
represents the potential at which and equal level of irreversible
pitting corrosion was occurring in a given environment rather than the
first, perhaps transient, evidence of passive film breakdown.
Relationships between
pitting potential and the major environmental test parameters was
represented graphically, as shown in
Fig. 1. The effect of pH is shown in
Fig. 1 a), and was observed to be small at levels above pH = 5,
particularly at the higher chloride levels where the pitting potential
was generally of a low order. A progressive reduction in pH below this
level resulted in a more significant decrease of the pitting
potential, with a minimum at about pH = 3. At lower pH levels the
onset of general acid corrosion was evident and resulted in an
apparent increase in the pitting potential. This behavior was
particularly evident at the lower chlorine ion concentrations.
The effect of increasing
temperatures was clearly to reduce the pitting potential, as shown in
Fig. 1b), and the relationship between pitting potential and
temperature was approximately logarithmic. It was notable that the
difference in pitting potential between the different alloys decreased
with increasing temperature as shown in
Fig. 1c).
Overall the most
significant parameter with respect to the pitting potentials of the
alloys was the chloride ion concentration, and a characteristic
relationship was apparent (See
Figure 2a). This took the form of an initially high rate of
decrease in pitting potential with increasing chloride levels in the
range 300 to 1000 ppm, but with the rate reducing progressively on
further increases to 19000 and 10000 ppm chloride ion concentrations.
Of particular importance was the fact that the relationship was close
to logarithmic such that a plot of pitting potential against the
logarithm of the chloride ion concentration produced essentially a
straight line for each of the pH, temperature and alloy combinations
examined, as illustrated by the example in
Fig. 2b).
A feature noted with all
the materials examined, regardless of alloy content, was the marked
change in the other wise consisted relationship between pitting
potential and the environmental conditions which was apparent at pH
levels below pH = 3. This was particularly evident at the lower
chloride ion concentrations as shown in
Fig. 1a), and was attributed to the development of general acid
corrosion. It was concluded that these results should be excluded from
the data base to be used in the construction of the engineering
diagrams.
To process the results
into a form suitable for constructing the engineering diagrams, the
consistent and reproducible linear relationship between the pitting
potential and the logarithm of the chloride ion concentration was
used. To interrelate all the environmental parameters, isopotential
curves were produced of the logarithm of the chloride ion
concentration against temperature at constant pH levels using
calculated best straight line relationships.
The curves indicate the
limiting chloride ion concentrations, pH and temperature for the
development of pitting corrosion for a given material in an
environment at a given potential. The main requirement in progressing
to useful engineering diagrams is that the potential of the system
must be known.
NOTE:
See InterCorr-CLI International's
capabilities in
electrochemical testing and material evaluation.
Figure 1 - Typical graphical representations of the relationship
between pitting potential and the major environmental and
compositional parameters
Figure 2 - Relationship between
pitting potential and chloride ion concentration on a a) linear and b)
logarithmic scale for type 316L
By: J W Fielder and D
R Johns,
British Steel Technica, Swinden Laboratories,
Moorgate Rotherham
Condensed from UK Corrosion ’89 Blackpool,
November 8-10, 1989.
|
