Sie sind auf Seite 1von 83

Learning Objectives – Types of Corrosion

1. Understand the major classifications of corrosion and their principle

2. Develop the skills to predict the principle type of corrosion that will
occur to a material or metal given its service environment.
3. Know how to judge the suitability of materials or metals for a given
service environment.
4. Types of Corrosion

4.2 Uniform corrosion (General

4.3 Galvanic corrosion
4.4 Dezincification (Selective
leaching corrosion)
4.5 Crevice corrosion
4.6 Pitting corrosion
4.7 Inter-granular corrosion
4.8 Stress corrosion cracking and
hydrogen damage
4.9 Corrosion Fatigue
4.10 Fretting corrosion
4.11 Erosion corrosion and
cavitation damage
4.3 Galvanic Corrosion
When 2 dissimilar metals are coupled electronically:

• that with more negative electrode potential (more active metal) → more anodic

• that with more positive electrode potential (less active metal) → more cathodic

• Potential difference ∆E prior to coupling → driving force for corrosion current,

- provides change in overpotentials at anode and cathode, + IR potential drop in
l il
solution (ohm’s law:IR = ∆φ s = iA × = where σ = conductivity, l = distance)
Aσ σ
• Area effect - small anode + large cathode → catastrophic failure may result, as
reduction current from large cathode (low current density) produces high current
density at small anode, resulting in high penetration rate.

• Such problems arise repeatedly in ‘engineering compromises’, where the metals of

differing mechanical, physical and engineering properties are combined.
4.3 Galvanic Corrosion

Electrons Ammeter

Coupling dissimilar metals by

IR = ∆φs
a zero resistance ammeter is
analogous to shorting a battery Fe2+ Fe2+
O2 O2

H+ Cu2+

4.3 Galvanic Corrosion
Potential Distribution in Open Circuit Electrochemical Cell

φ a Anode - Voltmeter + Cathode


Anode corrosion potential

Cell Voltage,
U = φc -φa Eref = 0.000 V
Cathode corrosion potential

Potential of electrodes Potential in solution
4.3 Galvanic Corrosion
Potential Distribution in Shorted Electrochemical Cell
Anode -+ Ammeter - Cathode
e e-

- conductivity, σ

Cathode reversible potential

+ cathode overpotential
Potential of anode, Potential of cathode,

Anode reversible φs
potential + overpotential

Potential in solution
4.3 Galvanic Corrosion
Schematic Kinetics of Galvanic Interaction of Cu and Fe at pH 4
0.5 Cu → Cu 2+ + 2e −
Electrode potential (SHE) / V

0.4 Ecorr(Cu)
0.2 ∆ηCu
O2 + 4H+ + 4e-  2H2O
E(Cu, shorted to Fe) < ECu2+/Cu immune ∆Ecorr
-0.2 IR=∆φs
-0.3 2 H + + 2e − → H 2
-0.4 Fe → Fe 2+ + 2e −
-0.5 Ecorr(Fe) icorr(Fe)
10-4 10-3 10-2 10-1 100
| Current density / A m-2 |
A Galvanic cell is formed when two dissimilar metals are connected electrically while both
are immersed in a solution electrolyte. The more active metals or alloys of the two is
corroded preferentially by galvanic corrosion.
Corrosion of a galvanic couple between a corroding and an inert metal

Zn - Pt couple in deaerated acid

io,H2(Zn)=10-10 A/cm2
io,H2(Pt)=10-3 A/cm2


Zn Pt
Galvanic coupling Zn to Pt;
i ) shifts the corrosion potential of Zn from Ecorr to Ecouple.
ii ) increases the corrosion rate from icorr(Zn) to icorr(Zn-Pt).
iii) decreases the rate of hydrogen evolution on Zn from iH2(Zn) to iH2(Zn-Pt).

Effect of exchange current density by comparison of Zn-Pt and Zn-Au galvanic couples.

E°Au3+/Au = 1.498 V io,H2(Au) = 10-6 A/cm2

E°Pt2+/Pt = 1.2 V io,H2(Pt) = 10-3 A/cm2
icorr(Zn-Pt) < icorr(Zn-Au) < icorr(Zn)
Ecorr(Zn-Pt) > Ecorr(Zn-Au)

The reason why gold produces a less severe

galvanic effect is not related to its reversible
potential but rather to the fact that it has a
lower hydrogen exchange current density than
Effect of area ratio of cathode to anode in galvanic couple

e- e-

Zn Pt Zn

Corrosion of a galvanic couple between two corroding metals

Zn - Fe couple in a deaerated acid solution

• Coupling of Fe to Zn with equal area :

i ) Corrosion rate of Fe is reduced from icorr(Fe) to icorr(Zn-Fe).
ii) Corrosion rate of Zn increases from icorr(Zn) to icorr(Fe-Zn).
• This principles is applied to galvanizing where Fe is cathodically protected by coupling with Zn
Zn – Fe couple in a deaerated acid solution

io,H+/H2 (Zn) io,H+/H2 (Fe)

Fe→Fe2+ +2e-

Zn→Zn2+ +2e-
icorr(Fe - Zn)

icorr(Fe - Zn)
Zn Fe

10-10 10-8 10-6 10-4 10-2 10-0 log |i| Zn2+

 General conclusions on the galvanic corrosion

1) If two corroding metals are galvanically coupled, the corrosion rate of more active
potential( ie. with more active corrosion potential) is accelerated, and that of the
other metal is retarded.
2) The polarity of electrode for galvanic couple should be defined on the basis of
corrosion potential.
3) The corrosion behavior of a galvanic couple is determined by the following parameters :
Er, io, B and the relative areas of the two metals.
4) Galvanic corrosion behavior can not be predicted accurately on the basis of EMF series.
Use galvanic series to predict corrosion behavior of galvanic couple.

Corrosion of galvanic couple containing active - passive metal

Case 1 : Epp < EH+/H2

ex) Ti - Pt, Cr - Pt in acid solution

 Coupling of Ti to Pt leads to spontaneous

passivation of Ti without oxidizers.
4.3 Galvanic Corrosion
Effects of Galvanic Coupling of Steel and Copper Pipes
Steel pipe Copper pipe

Aqueous solution
Ionic current flow in solution
Fe → Fe 2+ + 2e − O2 + 4 H + + 4e − → 2 H 2O

Electron flow in pipes

Locally-enhanced rates of steel corrosion
Steel acts as anode and copper as cathode, which is made immune
locally, if copper electrode potential < ECu 2+ Cu
Solution to problem:
Insulating washers
Insulating O-ring

Aqueous solution

Steel pipe Copper pipe

4.3 Galvanic Corrosion
Effects of Galvanic Coupling of Steel and Copper Pipes
Highest iFe / penetration
rate where current path Electron
length / R lowest flow in pipes
+ −
Fe → Fe 2+ + 2e − -
e 2 O + 4 H + 4e → 2 H 2O

Ionic current (I)x flow

Relectrolyte Fe2+ X- in aqueous solution
– ohmic resistor

Steel pipe Copper pipe

Non-ohmic resistance due to reaction(s), current flowing between

electronically conducting metal and ionically conducting solution
4.3 Galvanic Corrosion
Simple Electronic Analogue of Galvanic Corrosion

φs,A Relectrolyte φs,C

R anode ∆Ecorr R cathode
Current, I
e- e-
Anode Cathode
φA φC
Steel pipe Copper pipe

Aqueous solution
Ionic current flow in solution
Fe → Fe 2+ + 2e − O2 + 4 H + + 4e − → 2 H 2O

Severest localised corrosion Electron flow in pipes

4.3 Galvanic Corrosion

Normalised potential

0.8 S9
0.7 S8
0.6 S7
0.5 S6
0.4 S5
Distance along
0.3 S4
0.2 Solution S3
Fe - Anode
Potential φs S2
1 3 5 S1
7 9 11 13 15 17 19

Distance across pipe

4.3 Galvanic Corrosion
Effect of Relative Areas of Dissimilar Materials in Galvanic Corrosion

Aqueous Electrolyte

iFe >> iO2

O2 + 4H+ + 4e-  2H2O O2 + 4H+ + 4e-  2H2O

e- e-
Less active metal Less active metal

Steel Rivet
Small area of more active metal/alloy in large area of less active metal
causes high current densities and more rapid penetration than converse.
4.3 Galvanic Corrosion
Solutions to Galvanic Corrosion Problems
4.3 Galvanic Corrosion

Galvanic Series

for seawater

• non-equilibrium potentials
• c.f. Electrochemical series of
equilibrium electrode potentials
• Dark boxes refer to active
behaviour of active-passive
4.4 Selective Leaching Corrosion
Selective Leaching Corrosion ( Galvanic Corrosion in Alloys )

• Galvanic corrosion may occur between different phases in alloys

• MnS inclusions in stainless steel – initiates pitting corrosion

• Important when one of phases is more active as a cathode, e.g.:

• Cu-containing precipitates in aluminium alloys – initiates pitting


• Fe and Cu impurities in commercial zinc

– cause large increases in corrosion rates compared to pure zinc

• Cu in brasses (Zn-Cu alloys)

4.4 Selective Leaching Corrosion
• micro-scale galvanic effect

• environment extracts one component of an alloy,

which may be a solid solution or a mixture of phases.


a) dezincification of brasses (Cu-Zn alloys)

b) loss of Cu from Cu-nickels - particularly in solutions containing

ammonium ions→ Cu(NH3)42+ / Cu(NH3)2+ (and Ni(NH3)4 or 62+ ).

c) graphitisation of cast-iron (Fe-C alloy with > 2 % C)

• graphite emerges on the surface as an active cathode for hydrogen
evolution and oxygen reduction, resulting in the selective
removal of iron from a graphite network.
4.4 Selective Leaching Corrosion
Graphitisation of Cast-Iron
O2 + 4 H + + 4e − O2 + 4 H + + 4e − O2 + 4 H + + 4e − O2 + 4 H + + 4e −
→ 2 H 2O → 2 H 2O → 2 H 2O → 2 H 2O

Fe2+ Fe2+ Fe2+

e- e- e- e- e- e-

• Oxygen reduction occurs primarily on graphite, exposed from

corrosion of cast iron (> 2 % C in Fe)
• Fe dissolves anodically: Fe → Fe 2+ + 2e -
4.4 Selective Leaching Corrosion

Graphitic Corrosion
of 0.2 m diameter
Cast Iron Pipe
Learning Outcomes
What have you learned from the lecture ?
1. You can estimate the corrosion potential and corrosion current
density resulting from galvanic coupling of dissimilar metals,
including area effects.

2. You can describe how to mitigate galvanic corrosion.

3. You can describe the essential features of selective leaching corrosion

and list some common examples.
4. Types of Corrosion

4.2 Uniform corrosion

(General corrosion)
4.3 Galvanic corrosion
4.4 Dezincification (Selective
leaching corrosion)
4.5 Crevice corrosion
4.6 Pitting corrosion
4.7 Inter-granular corrosion
4.8 Stress corrosion cracking
and hydrogen damage
4.9 Corrosion Fatigue
4.10 Fretting corrosion
4.11 Erosion corrosion and
cavitation damage

Pitting and Crevice Corrosion

4.5 Crevice Corrosion
• involves geometric problems - joints or contacts

• Often associated with small volumes of stagnant solution caused by

holes, gasket surfaces, lap joints, surface deposits and crevices under
bolt and rivet heads → differences in local mass transport rates

• requires concentration differences, e.g. ‘differential aeration cells’

• involves generation of a potential gradient, due to metal dissolution,

causes Cl- to migrate into the crevice to maintain electroneutrality

• Cl- ions are aggressive towards passive oxide films, causing their
breakdown ?
Differential Aeration Cell due to Deposit on Metal Surface

Aqueous Solution

O2 + 2 H 2 O + 4e − → 4OH −

O2 Inert deposit O2

Passive Active Passive

Fe → Fe 2+ + 2e -
4.5 Crevice Corrosion
Effect of Differences in Oxygen Mass Transport Rates on Corroding Metal

Kinetic control: inet = i0 e β aη − e − β cη )
Electrode potential (SHE) / V

2 Fe + 3H 2O → Fe2O3 + 6 H + + 6e-
Ecorr (cathode)
iL ,O2 = −4 Fk m ,O2 CO2
Passive + −
dissolution O2 + 4 H + 4e → 2 H 2O ∆Ecorr
potential, Ep
Ecorr (anode) icritical

icorr (anode) Fe → Fe 2+ + 2e - Active dissolution

log (| current density, j / A m-2 | )

4.5 Crevice Corrosion
∑ ci zi = 0 O2

Cl- O2 Fe2+ Fe2+ Fe2+

1. O2 consumed in crevice by slow passive corrosion
O2 + 4 H + + 4e − → 2 H 2O

2. Passive corrosion continues and pH falls by Fe2+ hydrolysis

Fe 2+ + 2 H 2O → Fe(OH ) 2 + 2 H +
3. Passive film breaks down in acid and rapid active corrosion starts
Fe2O3 + 6 H + + 2e − → 2 Fe 2+ + 3H 2O
4. Active corrosion reinforces H+ production by hydrolysis (2)
Fe → Fe 2+ + 2e -
1. In the beginning the metal corrodes at the same rate inside and outside the 4. The growth (or propagation) phase. Because of increased corrosion
crevice. The anodic reaction MM+ + e– and the cathodic reaction O2 + rate, the migration of Cl– increases also, which contributes to a further
2H2O + 4e– 4OH– occur on area elements distributed between each acceleration of the corrosion process. The process promotes itself, i.e. it
other all over the surface. If the metal is passive beforehand, the corrosion rate is “auto–catalytic”. Since the pH has been strongly reduced, the
is low (corresponding to the passive current density) and the consumption of hydrogen reaction 2H++2e–H2 can also possibly contribute as a second
oxygen is correspondingly low. The oxygen is gradually being consumed cathodic reaction.
within the crevice. This is also the case when the metal surfaces are passive, pH inside the crevice can reach values of 0–4 depending on the actual
provided that the crevice is so narrow and deep that oxygen is more slowly material–environment combination. At the same time, pH may increase to
transported into the crevice than it is consumed inside it. 9–10 on the metal surface outside the crevice, where the oxygen reduction
takes place. The conditions at this stage can in principle be described with
2. When the oxygen has been completely consumed inside the crevice, OH– the overvoltage curves in Figure 7.16b. The corrosion current Icorr = Io +
can no longer be produced there. Conversely, the dissolution of metal inside the IH, where Io is the oxygen reduction current outside the crevice and IH
crevice continues, supported by the oxygen reduction outside the crevice. The represents the hydrogen reaction 2H+ + 2e– H2 inside. There is a potential
concentration of metal ions within the crevice increases, and with missing OH– drop ¨E = Eo – Ei in the electrolyte from the free surface outside to the
production in the crevice, the charge equilibrium (electrical neutrality) is inside of the crevice. The area of the crevice surface is often small
maintained by migration of Cl– into the crevice. In this way an increasing compared with that of the outer surface. This means that a large corrosion
amount of dissociated metal chloride (M+Cl–) is produced in the crevice. The current is concentrated on a small area, so that the corrosion current density
metal chloride reacts with water (hydrolyzes): is very high in unfavourable cases.
M+Cl– + H2O MOH + H+Cl–.

Metal hydroxide is deposited and hydrochloric acid is formed in the crevice,

which causes a gradual reduction of the pH. This process leads to a critical
corrosion state.

3. When the environment has become sufficiently aggressive, the surface oxide
film is attacked. The crevice surfaces are transferred to an active state and the
corrosion rate increases.
4.5 Crevice Corrosion
Recommendations to minimise crevice corrosion include:
1- Selection of material:
-conventional stainless steels, with martensitic, ferritic, austenitic or ferritic–austenitic (duplex) structure, are sensitive
to crevice corrosion
- Newer high-alloy steels with high Mo content show by far better crevice corrosion properties in seawater and other
Cl-containing environments

2- Design and production. Avoid as far as possible crevices and deposition

-Use butt welds instead of overlap joints. Alternatively, overlap joints can be sealed by continuous welds
- Design in a way that promotes complete drainage and prevents accumulation of deposits. Make it easy to inspect and
3. Cathodic protection. The potential must be kept below the protection potential. An efficient method for internal
protection of pipe joints is the “resistance controlled cathodic protection”

4- Measures for preventing deposition.

a) Inspection and cleaning during and between service periods.
b) Separation of solid material from flowing media in process plants.
c) Gravel filling around buried piping and structures
4.6 Pitting Corrosion

pitting of
centrifuge part
- 316 stainless steel
4.6 Pitting Corrosion

Deep Pits in Stainless Steel

- highly localised
- less general corrosion

Shallow Pits
– more general corrosion at
lower pHs, at which stainless
steel is more active
4.6 Pitting Corrosion
Characteristics of Pitting Corrosion
• Probability of pitting increases with increasing potential
→ provides greater driving force for corrosion

Passive alloy or metal required

• pitting requires a passive external surface → large potential difference

(∆E) with active pit, causing current to flow;
if external surface is active → no driving force

• carbon steel will pit only if conditions lead to its passivation

(e.g. neutral / alkaline solutions)

• carbon steel will not pit if it is corroding generally

(e.g. acid / neutral salt solutions)
4.6 Pitting Corrosion - Theories

Absorption of halide ions causes strong increase of the ion conductivity in the oxide film so that metal
ions can migrate through the film. In this way localized dissolution occurs, and intrusions are
subsequently formed in the metal surface . Another theory is that the initial adsorption of aggressive
anions at the oxide surface enhances catalytically the transfer of metal cations from the oxide to the
electrolyte and thus causes successive local thinning of the oxide film. A third possibility is that the
attacks start at fissures in the passive layer.
Which of the mechanisms is the most effective depends on both material and environment.
4.6 Pitting Corrosion
• Localised dissolution catalysed by Cl- and H+ → pH gradients (indicator in image)
• Dissolution of reactive component of an alloy – MSx inclusions in stainless steel
• Distributed failures in protective / passive films - Cl- pitting of Al/Al2O3 or SS/Cr2O3
• O2 reduction on oxide - much larger surface area than pits → small number density
• Current density in individual pit >> that on oxide
• Deep penetration → possible catastrophic failure

Solution to problem ?
Increasing pitting resistance:
↓ type 304 stainless steel
↓ type 316 stainless steel
↓ Hastelloy C (15.5% Cr, 15.5% Mo, 3% W,
balance Ni)
↓ Titanium
4.6 Pitting Corrosion
Fissure in oxide film
O2 O2
O2 + 2 H 2 O + 4e − → 4OH − O2 + 2 H 2 O + 4e − → 4OH −
Iron Oxide Iron Oxide
e- Fe2+ e-
Steel Fe → Fe 2+ + 2e -

1. Breakdown of passive film nucleates pits. Ingress of electrolyte to

substrate results in reductive dissolution of oxide, exposing bare
metal within the pit
Fe2O3 + 6 H + + 2e − → 2 Fe 2+ + 3H 2O
2. Cathodic reduction of O2 occurs on passive film is balanced by acitve
anodic dissolution of iron within the pit
Fe → Fe 2+ + 2e -
4.6 Pitting Corrosion
∑c z
i i = 0 Electroneutrality
→ migration of Cl-
i pit >> iO2
Precipitated corrosion products
O2 O2
O2 + 2 H 2 O + 4e − → 4OH − O2 + 2 H 2 O + 4e − → 4OH −
Iron Oxide Iron Oxide

e- Fe2+ e-
Steel Fe → Fe 2+ + 2e -

3. In growing pits, hydrolysis of Fe2+ decreases pH and breaks down

passive film: Fe 2+ + 2 H O → Fe(OH ) + 2 H +
2 2

4. Cl- migrate into pit to balance charge – electroneutrality in solution

5. Cathodic reduction of O2 on passive film causes active corrosion
within the pit (2).
Externally added OH- may inhibit pitting by increasing pH in pits
Factors of general significance for pitting are:

a) pH and chloride concentration. The pitting potential and pitting resistance normally increase with
increasing pH and decreasing chloride concentration.

b) Flow velocity. When pitting is initiated, an aggressive environment is established within the pit, and this
is combined with an increased pH at the adjacent free surface. The increased pH gives an increased
resistance to initiation of new pits around the first one formed (this is the case, e.g. for unalloyed steel,
but it is more uncertain with respect to aluminum). If there is no movement of the liquid, these conditions
will be conserved, and the result is large and few pits. Liquid flow gives a higher probability for washing
away the aggressive environment in the pit, and at the same time it increases the transport of oxygen to
the active area, so that the pit may be passivated before it gets the chance to grow to a considerable size.
The alkaline layer around an active pit is also washed away to a higher extent, and pit initiation in the
neighbourhood of an active pit occurs more easily. This means that increased flow velocity results in
smaller but more numerous pits, which makes pitting corrosion less serious.

c) The gravity force. Horizontal top surfaces are often heavily pitted, while underside surfaces are hardly,
or not at all, attacked. Vertical surfaces are intermediate as to the extent of pitting. The reason is that the
aggressive environment in the pits has a higher density than pure water.

d) Cu2+ and Fe3+ ions favour pit initiation and accelerate pit growth because they lead to increased
potential . The Cu ions have a double effect because Cu is precipitated on the material surface and forms
efficient cathodes.

e) Metallurgical properties. Impurities and inclusions are important, e.g. AlFe secondary phase, which
contributes to the localization of pits on aluminum because 1) the oxide film on top of and around the
inclusion is weak, thin and stressed, 2) such inclusions are efficient cathodes. Pitting of this type will
usually last only until the pits are so large that the inclusions are falling out.

f) The insulating ability of the oxide. If the oxide insulates efficiently, the surface is inactive as a cathode.
This is to a certain extent the case for aluminum in seawater; which is the direct reason for the relative
slow growth of pits in this case.
g) Surface roughness. The main trend is that smooth surfaces get few, large pits while rough surfaces get
numerous smaller pits.

h) Temperature. Increasing temperature gives usually decreasing pitting potential and increasing liability
to pitting corrosion.

i) Galvanic contact with a more noble material increases the tendency to and the rate of pitting corrosion
(the corrosion potential is lifted).
Factors causing localized corrosion (summary)
• Heterogeneities in metal
Second phases.
Non metallic inclusions.
Non uniform microstructure : heat affected zone in weld.
Non uniform cold work.
• Heterogeneities in environment
Differential aeration.
• Important features in localized corrosion
Small anode area /large cathode area.
pH changes at the cathodic and anodic sites.
Degree of occlusion.
Cl- concentration.
Potential-pH Diagram for Fe-H2O System; 298 K, activity = 10-4

Electrode potential (SHE) / V

H+ Fe2+ Fe2O3 H2O
H2 Fe(OH)2 Fe O
-0.4 3 4
Inside pit Passivation
-0.8 outside pit
by O2 reduction

-1.6 Fe

-1 1 3 5 7 9 11 13 15
4.6 Crevice and Pitting Corrosion
Crevice Corrosion of Stainless Steels from SO2 Aqueous Absorber

304 ss – Acid 316 ss – Slightly Acid 304 ss – Alkaline

General + pitting Crevice + passive Passive
Leak from Water Pipe Caused Pitting of Petroleum Gasoline
Pipeline that Leaked into Sewer in Guadalajara, Mexico
April 1992 - explosions along 2 km of sewer killed 215, injured 1,500
people and damaged 1,600 buildings.
Learning Outcomes
What have you learned from this lecture?
1. You can describe the conditions that lead to crevice corrosion
(requires geometries that promote local composition gradients - often
associated with depletion of dissolved O2 resulting in ‘differential
aeration’, requires passive external surface, etc.)
2. You can describe the mechanism of crevice corrosion of steels with
the associated chemical reactions
3. You can list some engineering solutions to crevice corrosion
4. You can describe the conditions that lead to pitting corrosion
(Requires passive surface, requires a flaw in passivation film to act as
a pit nucleation site, chloride ions often implicated – maintain
electroneutrality, etc.)
5. You can describe the mechanism of pitting of steels with the
associated chemical reactions
6. You can rationalize some of the engineering solutions to pitting
corrosion (increasing bulk solution pH, selecting stainless steels with
increased by Mo content, etc.)
4. Types of Corrosion

4.2 Uniform corrosion

(General corrosion)
4.3 Galvanic corrosion
4.4 Dezincification (Selective
leaching corrosion)
4.5 Crevice corrosion
4.6 Pitting corrosion
4.7 Inter-granular corrosion
4.8 Stress corrosion cracking
and hydrogen damage
4.9 Corrosion Fatigue
4.10 Fretting corrosion
4.11 Erosion corrosion and
cavitation damage
4.7 Intergranular Corrosion
Intergranular corrosion is localized attack on or at grain boundaries with insignificant corrosion on
other parts of the surface. The attacks propagate into the material. This is a dangerous form of corrosion
because the cohesive forces between the grains may be too small to withstand tensile stresses; the
toughness of the material is seriously reduced at a relatively early stage, and fracture can occur without
warning. Grains may fall out, leaving pits or grooves, but this may not be particularly important.
The general cause of intergranular corrosion is the presence of galvanic elements due to differences in
concentration of impurities or alloying elements between the material in or at the grain boundaries and
the interior of the grains:
a) Impurities segregated to the grain boundaries (causing, e.g. the AlFe secondary phase in aluminum).
b) Larger amount of a dissolved alloying element at the grain boundaries (e.g. Zn in brass).
c) Smaller amount of a dissolved alloying element at the grain boundaries (e.g. Cr in stainless steel).
In most cases there is a zone of less noble material in/at the grain boundaries, which acts as an anode,
while the other parts of the surface form the cathode. The area ratio between the cathode and the anode
is very large, and the corrosion intensity can therefore be high.
In some cases, precipitates at the grain boundaries may be more noble than the bulk material; these
precipitates will stimulate grain boundary attacks by acting as efficient local cathodes (e.g. CuAl2 in
aluminum alloys).
Intergranular corrosion occurs in stainless steels and alloys based on nickel, aluminum, magnesium,
copper and cast zinc. In the following sections we shall look at the three former groups in some detail.
4.7 Intergranular Corrosion
Stainless Steels
• Expensive - no desirable physical or mechanical properties
• Corrosion resistance due to passivation is only reason for their use
1. Martensitic (Fe-Cr-C)
• 20 % Cr for ambient, > 30 % for high T, 0.2 - 2 % C
• Most common type 420 : Fe-13%Cr-0.35%-0.45%C
• Strong (due to martensite) and hard, BUT corrosion resistance < ferritic <
• Corrosion resistance better in quenched + tempered condition, then annealed
2. Ferritic (Fe-Cr, < 0.1 % C)
• need >12 % Cr for stainless steel - 20 % Cr for ambient T, > 30 % for high T
• bcc structure → soft; strengthened only by work hardening
• Most common type 430 : Fe-16-17%Cr-< 0.12%C
• used for car trim
• good atmospheric corrosion resistance
• Best stainless steel for stress cracking resistance, especially in Cl- media
3. Austenitic (Fe-Cr-Ni, 15-27 % Cr, 3-22 % Ni)
• 304 (Fe-18-20%Cr-8-12%Ni),
316 (Fe-16-18%Cr-10-14%Ni2-3%Mo),
347 (Fe-17-19%Cr-9-13%Ni1%Nb) most common types
- Mo gives better corrosion resistance in Cl-
• fcc structure, strengthened by cold working only
• 500-800°C → weld decay → grain boundary Cr23C6 and possibly Mo23C6
precipitate → intergranular corrosion
4. Precipitation Hardening Stainless Steel
• Fe-17%Cr-7%Ni-1%Al, Fe-17%Cr-4%Ni-4%Cu, Fe-17%Cr-4%Ni-3%Mo →
• very high strength - σty up to 1.3 GPa
• moderate corrosion resistance - used in aerospace applications
• expensive to use, requiring solution heat treatment + quenching + annealing to
produce hardening
4.7 Intergranular Corrosion
Metal Microstructure – Mechanical Property Relationships
Rapid cooling of liquid
→greater supersaturation
→greater nuclei number density
→finer grains
→stronger material
Edge dislocations move along slip planes,
interacting with each other and grain
boundaries → work hardening
Grain boundaries of polycrystalline material
• higher energies than bulk crystals
• lattice spacing + bond lengths differ
• fine grains < 1 mm, coarse grains > 10 mm
• 1010 - 1016 edge dislocations emerge per m2 of surface, depending on
extent of cold working
• Annealing, heat treatment at 0.3-0.4 TM , may be required to soften the
structure of cold worked / hardened materials, e.g. from rolling
4.7 Intergranular Corrosion
• Specific reactivity at grain boundaries in a metal or alloy; very
little weight loss from very small areas, but with substantial
mechanical results.
Reactive material at grain boundaries in metals or alloys results from:
a) emergent slip bands or disorganised material as a result of cold
work. e.g. cold worked 30Cu - 70Zn brass in HgNO3 + HNO3
breaks easily.
b) Depleted zones at grain boundaries by precipitation of segregates.
E.g. loss of Cr and/or Mo (if present) at grain boundaries in
austenitic stainless steels (like 316) due to precipitation of Cr23C6
andMo23C6 in as a result of heat-transformation (typically 500 -
800 C) in welding ("weld decay").
4.7 Intergranular Corrosion
“Weld Decay” Corrosion Round Welds in
Austenitic Stainless Steel Distillation
due to formation of Cr23C6 at grain
boundaries during welding, depleting Cr in
the stainless steel grains to bulk
concentrations below ca. 10 % at which
Cr2O3 provides protection.
Corrosion at Welds in Stainless Steel Distillation Column
due to formation of Cr23C6 at grain boundaries during welding

Scanning Electron Micrograph

Corrosion at Welds in Stainless Steel Distillation Column
due to formation of Cr23C6 at grain boundaries during welding
4.7 Intergranular Corrosion
Engineering Solutions to Intergranular Corrosion:
1.Subject the sensitized material (weld) to a high temperature heat
treatment to re-dissolve the segregated phases.
2.Select a steel with a low carbide content, below 0.03%, (as in 304L,
316L, and 317L, austenitic stainless steels).
3.Select a steel with a lower chromium content (as in 403, 410 and
420 martensitic steels)
4.Select a steel that contains Nb or Ti, which have a greater tendency
to form carbides than chromium, so that the Cr stays dissolved in the
steel (as in 347 and 321 austenitic stainless steels, respectively).
ferrite in
austenite grains

to ‘weld decay’ caused by Cr23C6

precipitation at grain boundaries

→insoluble Mo chlorides
4. Types of Corrosion

4.2 Uniform corrosion

(General corrosion)
4.3 Galvanic corrosion
4.4 Dezincification (Selective
leaching corrosion)
4.5 Crevice corrosion
4.6 Pitting corrosion
4.7 Inter-granular corrosion
4.8 Stress corrosion cracking
and hydrogen damage
4.9 Corrosion Fatigue
4.10 Fretting corrosion
4.11 Erosion corrosion and
cavitation damage
4.8 Stress Corrosion Cracking and Hydrogen Damage
Stress corrosion cracking (SCC), (usually) requires:
1. Static TENSILE stress – applied or residual
2. Structural sensitivity of material
3. Specific chemical environment
1. Static TENSILE stress
Macroscopically, cracks propagate normal to a tensile stress. The
stress may be applied (pulling or bending a material) or may be
residual (as a result of mechanical working or heat treatment).
2. Structural sensitivity of material
A passive film is a prerequisite for SCC (such films are weak in
tension). Softer materials (pure metals) are more resistant to SCC
than alloys (harder).
4.8 Stress Corrosion Cracking and Hydrogen Damage
2. Structural sensitivity of material (Contd.)
Microscopically, SCC may be transgranular (across grain boundaries)
or, more commonly, intergranular (following grain boundaries).

Inter-granular SCC - stainless Trans-granular SCC - brass

4.8 Stress Corrosion Cracking and Hydrogen Damage
3. Specific chemical environment
SCC only occurs at electrochemical potentials where the passive
film is less stable, eg. in Zone 1 or Zone 2 below:

Image taken from Denny A.

Jones, Principles and
Prevention of Corrosion, 2nd
ed., Prentice Hall, Upper Saddle
River, NJ, USA, 1996, pg 242.
4.8 Stress Corrosion Cracking and Hydrogen Damage
Mechanism of Crack Propagation:

Tensile stress

Passive walls
Chemical Environment Active
crack tip
Passive walls

Tensile stress
1. Crack walls and tip passive
2. Plastic strain ruptures passive film at crack tip during active corrosion
3. crack tip re-passivates
4. Repeat 1-3
Environment-Alloy Combinations for Stress Corrosion Cracking
Alloy Environment Temp.
Austenitic Hot acid chloride solutions such as MgCl2 and BaCl2 NaCl-H2O2 solutions 60-200°C
Neutral halides: Br-, I-, F- Alkaline CaCl2
stainless steels Seawater
Concentrated NaOH solutions 120°C
NaOH-H2S solutions
Condensing steam from chloride waters
For sensitised alloys (due to Cr23C6 precipitation at grain boundaries):
Polythionic acids (H2SnO6), Sulfurous acid. RT
Pressurized hot water containing 2 ppm dissolved oxygen 300°C
Ferritic H2S, NH4Cl, NH4NO3 , hypochlorite (Resistant to most environments if free of Ni
but may fail by other modes of corrosion in same media.)
stainless steels
Duplex Susceptible to same environments as austenitic stainless steels but more resistant.
(Immune to intergranular SCC in polythionic acid. Also greater
stainless steels resistance than ferritic stainless steels to other forms of corrosion.)
Martensitic Caustic NaOH solutions. (Resistant to SCC in hot chlorides. Susceptible to
hydrogen embrittlement.)
stainless steels
Carbon Steels Caustic NaOH solutions 50°C
NaOH-NaSiO3 solutions. > 255°C
Calcium, ammonium, and sodium nitrate solutions Boiling
Mixed Acids (H2SO4- HNO3) RT
HCN solutions, acidified Warm
Acidic H2S solutions
Anhydrous liquid ammonia, Carbonate/bicarbonate, Amines RT
CO/CO2 solutions All
Nickel-Cr-Fe High-temperature chloride solutions aggravating factors: pH < 4, oxidising >205°
Alloys 600, 800, species, such as dissolved O2, H2S, free sulfur. C
690 Polythionic acids and thiosulfate solutions, sensitised alloys with excess C
Caustic alkaline solutions RT
Ni-Cu Acidic fluoride solutions RT
Monel HF, Hydrofluosilicic acid
Alloy 400 Susceptible in cold-worked state. Resistant in stress-relieved state
Nickel Alloy 200/1 Caustic alkaline solutions 290°C

Cu-Zn Alloys Ammonia vapours in water RT

(Brass) Amines in water
Nitrites in water
Water, Water vapour alone
> 15% Zn (45-50% Zn, β or β+τ alloys)
Nitrate solutions
Some sulfate solutions
Aluminium Alloys Air with water vapour RT
Potable waters
NaCl solutions
NaCl-H2O2 solutions
Titanium Alloys Red fuming nitric acid RT
Hot salts, molten salts >
N2O4 260°C
4.8 Stress Corrosion Cracking and Hydrogen Damage
Stress-strain Behaviour of Carbon Steel

Stress cracking corrosion often leads to rapid catastrophic failure of

components at a fraction of what their service life would be in an
otherwise inert environment.
Stress Corrosion Cracking and Hydrogen Damage
Stress Corrosion Testing at Constant Applied Stress

Threshold stress

Log (Time to Failure)

Control of Stress-Corrosion Cracking

• Remove stress - often difficult, especially for residual stresses
• Avoid environment necessary to cause SCC
• Apply electrochemical protection, where possible
• Use a different material



Stress Corrosion Cracking and Hydrogen Damage
Hydrogen Damage
1. Electrochemical conditions that result in the production of atomic
hydrogen by proton / water reduction.
H 2O + e − → H + OH −
+ −
H +e → H
2. Hydrogen diffuses into the material ahead of crack, making it harder
and more brittle (like a ceramic).
4.8 Stress Corrosion Cracking and Hydrogen Damage
Hydrogen Damage
3. Tensile stress causes brittle failure up to zone of depleted hydrogen
(softer, pure material)
4. The crack tip acts as a stress raiser.

Tensile stress

Tensile stress
Stress Corrosion Cracking and Hydrogen Damage
Hydrogen Damage
Higher strength materials are more susceptible to hydrogen embrittlement
Tensile stress
Passive walls Cracking of
Chemical Environment Type 301
Stainless Steel
Passive walls

Tensile stress
Hydrogen embrittlement is accelerated by species that
slow the recombination reaction that forms H2, such as
S2- and As3+.
slowed by S 2 − and As 3+
H + H     
→ H 2
Corrosion fatigue (CF) is crack formation due to varying stresses combined with
corrosion. Alternatively, it may be defined as fatigue stimulated and accelerated by
corrosion. The problem is caused by tensile stresses, just as it is for SCC. The
external difference between these two forms of deterioration is only that CF develops
under varying stresses and SCC under static stresses.
Typical fatigue fractures that have occurred in non-corrosive environments show a
large smooth crack surface area where the crack has grown by fatigue and an (often
smaller) area with a rough and crystalline surface formed by fast fracture when the
maximum stress reached the ultimate strength. Characteristic beach marks on the
smooth part show the form and position of the crack front at different stages. Fatigue
fractures of cathodically protected steel in seawater are also characterized by such
beach marks..
Usually, CF under free corrosion gives beach marks too, but sometimes they may
be less visible because of corrosion products. Corrosion fatigue fractures are
normally transgranular, and contrary to “dry” fatigue often branched.
Corrosion fatigue occurs for all material–environment combinations where general
or localized surface corrosion is developed. Fatigue and CF became subject to much
attention in connection with platforms in the North Sea, because the load spectrum
caused by the waves made design against fatigue more relevant in the North Sea than
in oil fields developed earlier, e.g. in the Mexican Gulf.
4.9 Corrosion Fatigue
• Decrease in fatigue resistance due to corrosive medium + cyclic stress.
• Fatigue failure results from structural changes in an alloy due to rapid
reversal of stress - involves crack initiation and growth.
• Some materials show a fatigue limit (e.g. mild steel) but many do not.
Corrosion fatigue ‘beach’ marks
- corrosion product accumulation during
discontinuous crack growth
4.9 Corrosion Fatigue
Corrosion Fatigue - S-N Curve
• Metal fatigue results in crack propagation due to a cyclic stress
• Corrosion makes both crack initiation and propagation easier
Stress Amplitude


Threshold stress

Corrosive environment

log (cycles to failure, Nf)

In dry fatigue of smooth components, the initiation time makes the major part of the
lifetime (e.g. 90%). For components with welds (not first class), with sharp notches or
exposed to a corrosive environment, this relationship is turned upside down, so that the
initiation period is relatively short (for instance 10% of the lifetime), while the crack
growth forms the major part of the lifetime.
The crack development can be divided into four stages
1. Initiation
2. Crack growth stage I, with crack growth direction about 45o relative to the
tensile stress direction.
3. Crack growth stage II.
4. Fast fracture.
The observations dealt with above have been explained by various theories. The effect of corrosion is mainly
ascribed to three different mechanisms:
i. Formation of corrosion pits and grooves leading to stress concentrations. This explanation has to some extent
been supported by the observation that machined notches and corrosion attacks give similar effects on fatigue
ii. Mutual acceleration of corrosion and mechanical deterioration. This can be considered in connection with a
common explanation of dry fatigue initiation mechanism for some materials. During the initiation period,
local slip bands are formed by slip in crystallographic planes with maximum shear stress. This is a common
mechanism of fatigue initiation. At higher stresses, fatigue is initiated at the grain boundaries. As a result of
alternating stresses the slip along different parallel planes leads to extrusions and intrusions. This finale
implies stress concentrations at the intrusions followed by gradual crack formation in the plane of the slip
band. When the crack has grown a length of one or a few grain diameters into the material (the order of 0.1
mm) the direction and plane of crack propagation is changed so it becomes normal to the tensile stress. The
mutually accelerating effects of corrosion and mechanical deterioration in this case are explained in two
ways: a) Corrosion induced plastic deformation, i.e. corrosion induces or promotes the formation of slip
bands, extrusions and intrusions by the dissolution of dislocation locks (these are localized at sites with high
density of dislocations, which makes the material more active).
b) Deformation-induced corrosion, which is caused by the extremely high activity of material under plastic
deformation in action, and which implies formation and growth of microcracks and possibly macrocracks
by intense localized anodic dissolution at the initiation site and successively at the crack front.
iii. Absorption and adsorption of species from the environment, including absorption of hydrogen and resulting
embrittlement. One theory is that reduction of the surface energy at the crack surface (particularly at the crack
front) caused by adsorbed elements promotes the formation and growth of cracks under fatigue loading. This
theory has been supported by the observation that fatigue is accelerated in some non-corrosive liquids.
However, in cases of CF in aqueous solutions, interest is usually concentrated on the absorption of hydrogen.
Comparison of Environmentally Induced Corrosion
Characteristic Stress Corrosion Corrosion Fatigue Hydrogen-Induced
Cracking Cracking Cracking (HIC)
stress Static cyclic with some tensile Static
tensile tensile
aqueous corrosive specific to the alloy any any
temperature increase accelerates accelerates increase to room
temperature, then
pure metal more resistant susceptible susceptible

crack morphology Transgranular or Transgranular Transgranular or

intergranular branched Unbranched intergranular
Sharp tip blunt tip sharp tip
corrosion products in absent (usually) present absent (usually)
crack surface appearance cleavage like beach marks and / or cleavage like
cathodic polarisation suppresses (usually) suppresses accelerates

near maximum strength susceptible, but HIC accelerates accelerates

level often predominates
Learning Outcomes
What have you learned from this lecture?
1. You can describe the crystallographic defects that lead to
intergranular corrosion, e.g. atoms of a material are in a higher
energy state at grain boundaries relative to the bulk (strained bond
lengths and incomplete bonding) and/or where phases have
segregated (Cr23C6 and Mo23C6).
2. You can describe how and why materials become more susceptible to
intergranualar corrosion, e.g. cold working which increases the defect
density in a material, heat treatment (weld decay) that causes phases
segregation leaving behind phases that more susceptible to corrosion
(Cr and Mo deficient in the case of some stainless steels).
3. You can list some engineering solutions to intergranular corrosion.

You do not need to memorize the number designations or

compositions of stainless steels
Learning Outcomes
What have you learned from this lecture?
1. You can describe the requirements, mechanisms and consequences of
stress cracking corrosion and hydrogen damage.
2. You can describe some engineering solutions for controlling stress
cracking corrosion and hydrogen damage.
3. You can describe the requirements, mechanisms and consequences of
corrosion fatigue.
4. You can compare the characteristics of stress cracking corrosion,
hydrogen damage, and corrosion fatigue.