Austenitic stainless steels such as 304L and 316L find employment in many industrial applications, sue to there mechanical, structural characteristics and corrosion resistance in atmospheric conditions.
Austenitic stainless steels generally is not susceptible to stress corrosion cracking(SCC) in atmospheric conditions with ambient temperature, however due to the large number of stainless steels component failure in indoor swimming pool, that cause by SCC, people started to investigate and found a phenomenon called atmospheric-induced stress corrosion cracking (AISCC).
Corrosion is a naturally occurring phenomenon most commonly associated with metals. Most discussions relating to corrosion centre on safety and economics. Corrosion is one of the main causes of component failure in industrial machinery; this can be a serious problem. Data shows that cost of corrosion in industry is about 3% of GDP; this includes prevention of corrosion, replacement of failed parts and reduction in efficiency caused by poor maintenance.
Corrosion is the universal name for destruction of material caused by the environment. It is a huge topic, the focus of this project is atmospheric induced chloride stress corrosion cracking(AISCC) on Austenistic stainless steel(ASS) types 304L and 316L.
Austenitic stainless steels are ferrous-based alloys and widely used in many engineering applications, because of their structural characteristics, good corrosion resistance and cost. Types 304L and 316L usually contain ~16 wt% chromium, this alloy element from a protective layer on the surface. Conventional wisdom predict that austenitic stainless steels are not susceptible to corrosion under normal atmospheric conditions at ambient temperature. However, in the last few decades, there have been many reports of austenitic stainless steel failures due to chloride-induced stress corrosion cracking (SCC) in ambient temperatures. This was naturally a concern and people started to study such cases.
It is generally accepted, SCC poses little threat at temperatures below 600C, however, some research has found that SCC can occur under certain conditions, at low temperatures. This is referred to as (AISCC); much work has been done to prove that austenitic stainless steels are susceptible to AISCC at temperatures below 600C with relative humidity between 10%-60% with presence of chloride deposit.
1.2. General Corrosion
Corrosion is defined as the destruction of metal by electrochemical reaction with its environment; it is a major contributor of loss of structural integrity in many engineering structures and mechanisms, and often leads to catastrophic failure. (Ref.1,2)
1.2.1. Uniform Attack
When metals are extracted form their ores, it increases the energy state of the metal (Ref.3). This results in the metal reacting with oxygen and water to form an oxide in order to reduce its energy state. This form of corrosion occurs uniformly over the metal surface, consequently causing a general thinning of the component, the rate is usually expressed in millimetres per year. Type 304L and 316L Austenitis stainless steel have a chromium rich passive layer which reduces the uniform thinning rate to a negligible level.
1.2.2. Electrochemical Reactions
Corrosion is driven by electrochemical reactions, this involves exchange of electrons between anode and cathode, this reaction takes place on a metal surface.
The most obvious case is the dissolution of metals in acidic solution
for example pure iron in hydrochloric acid(Ref.4)
Fe(s)+2HCl(aq) → FeCl2(aq) +H2(g) (1.1)
As chloride ions do not take part in this reaction, we can convert to two half equations
Fe → Fe2+ +2e- (1.2)
Equation (1.2) shows an anodic partial reaction of iron dissolution, iron is oxidised to iron ion (oxidation)
2H+ + 2e- → H2 (1.3)
Equation (1.3) shows a cathodic partial reaction of hydrogen evolution, hydrogen ions reduced to hydrogen(reduction)
Where the anodic reaction occurs, it is called an anode, and a cathode for the cathodic reaction. These two reactions must happen, simultaneously at the same rate otherwise a net charged will be produced on the metal surface, which is impossible; also the positions of the cathode and anode are not fixed, and can move across the material.
1.2.3. Passivity
In austenitic stainless steel, the uniform attack thinning rate is very low, sometimes negligible, one of the reason for the low rate is passivity. The passive state of a metal is described as an extremely low rate of metal dissolution, despite thermodynamic contributions. This is achieved by a 3-dimensional surface oxide thin film (a few nanometre thick) which isolates the metal from the corrosive environment. In Austenitic stainless steel the passive protective film mainly consists of chromium oxide Cr2O3. (Ref.5,6) such an oxide film protects the metal from uniform attack, meaning it is stainless.
Figure.1 Polarization curve active to passive transition(Ref.1)Once the electrochemical potential reaches the Passive zone of the graph steel becomes passive, forming a nanometre passive film which isolates the metal from the environment, however if the corrosion potential is too high and reaches the breakdown potential, it becomes transpassive, and localised corrosion may occur.
A Pourbaix diagram is constructed to explain stability and behaviour of metals in various environments.
Figure.2 Pourbaix diagram for iron(Douglas J. Wood San José State University College of Materials Engineering1993)
Two parameters of a Pourbaix diagram are pH and electrode potential. It predicts the regions of immunity, passivity and corrosion, it can be seen that the passivity is highly dependant on pH and corrosion potential.
However, these diagrams are only true for equilibrium conditions, and equilibrium is rarely achieved in real life situations, also it does not provide any direct kinetic information(Ref.7). It is not possible to predict the kinetic energy transfer of the reaction from the diagram, nevertheless, a Pourbaix diagram it is very often used to predict the immune and passive regions for a metal.
1.3. Stainless Steels
Stainless steels are ferrous-based alloys with minimum 11 wt% of chromium, it has excellent corrosion resistance due to the chromium rich oxide film on the surface of alloy, the film is passive and self healing in the atmosphere (Ref.8). Other alloying elements are commonly added to enhance the properties of steel, such as molybdenum in Austenitic stainless steels to improve pitting resistance
There are five main class of steels, Ferritic, Austenitic, Martensitic, Precipitation hardending Martensitic and Duplex. The classes are defined by their microstructure (Ref.8). The Austenitic class all have an FCC (face centred cubic) structure and typical alloy content of about 16% Chromium 8% Nickel.(Ref.10)
1.3.1. Austenitic Stainless Steels
They are the most common structural material in many engineering structures and mechanisms, due to their mechanical properties, corrosion resistance, cost and formability. They are ferrous-based alloys with high content of Nickel and chromium, the basic grade of ASS is Type 304, Figure.3 shows the evolution of other forms of 300 series and their alloying elements, for example by adding molybdenum and reducing carbon content in 304 it becomes type 316L. This give better pitting resistance and less sensitization on welding.
Table 1 Chemical composition of 304L and 316L austenitic stainless steel (%wt.)(Ref.8)
Carbon
Chromium
Nickel
Nitrogen
Manganese
Silicon
Iron
304L
0.03
18-20
8-10
0.1
2
0.03
balance
316L
0.03
16-18
10-14
0.1
2
0.03
balance
Figure.3 Austenitic stainless steels grades (Ref.45)
However not all the alloying elements enhance the corrosion resistance, some are added for mechanical or other purpose, they will have detrimental effect on corrosion resistance.(Ref.20) Figure.4 shows how each alloying element contributes to the corrosion resistance in ASS.
Figure.4 Effect of alloying elements (Ref.45)
1.3.2. Effects of Alloying Elements
Molybdenum
Improves resistance to pitting, especially in chloridic and sulphuric environments.
Phosphorus and sulphur
Improves machinability however it has a detrimental effect on corrosion resistance
Manganese
the role of manganese is complicated, in low concentration (1-2%) it is an austenitizer, which means it stabilises the material in the Austenitic phase, but is will have opposite effect when the concentration is too high, also for content more than 2% Mn it reduces the resistance to stress corrosion cracking.
Chromium and nickel
forms self-repairing passive oxide film on the surface of alloys, improves corrosion resistance also stabilises the FCC Phase at room temperature.
Figure.5 Boundary of metastable Austenitic for steels containing 0.1% C rapidly cooled from 11000C Keating F.H 1956 (Ref.10)Figure 5 shows the boundary of metastable Austenitic steels containing 0.1% C rapidly cooled from 11000C Keating F.H 1956
It shows that a certain amount of nickel and chromium are needed to maintain the Austenitic phase in steel.
Carbon
Generally in all steel, carbon has the effect of increasing the hardness and strength. However, in ASS, carbon also has detrimental effects, caused by carbide precipitation.
1.3.2. Sensitisation
The high content of chromium in ASS, gives a great affinity to carbon in the steel matrix, at elevated temperatures, including the temperature range for heat treatment, the carbon will diffuse and penetrate the grain, primarily along the grain boundaries. This results in precipitate of chromium carbide and causes depletion of chromium in the matrix. Consequently there is not sufficient chromium to create a passive film over the surface of the metal. Localised corrosion such as SCC or pitting may occur in this region, for this reason in some ASS the composition of carbon is reduced to below 0.03% (type 304L, 316L).(Ref.11)
Figure.6 (William D.Callister.Jr) is an illustration of chromium carbide(Cr23C6) precipitated along grain boundaries in stainless steel, and the area around, depleted of chromium(Ref.12), which means there is insufficient chromium to maintain the passivity protective layer.
1.4. Pitting Corrosion
Pitting corrosion, is a form of localised corrosion which, if widespread, is very damaging to the surface, for most of the metallic materials,(Ref.14) it is a local dissolution leading to the formation of cavities in those alloys or metal which exhibit passivity. This area has been studied for many years, it has been established that pitting is a detrimental side effect of passivity, it is an ionic process where an aggressive anion penetrates the passive film and generates a crack on the surface of a metal(Ref.9).
Nowadays, there are many existing theories to try to explain what causes pitting and the mechanisms of pitting, however, there is no consensus of opinion. There are two prevelant one is that pitting is a result of competitive adsorption between halide anions and oxygen, the other group suggest that, there is a defect on tridimensional passive films, which initiate pitting on the surface.(Ref. 14)
1.4.1. Anion Penetration and Migration Theories
In 1927 Evans suggested that the high permeation ability of Cl- through the protective oxides film is due to the small diameter of Cl- ion, and many factors which accelerate the penetration of anion, such as electrical field across the film, vacancies and defects.(Ref.15,16) It is often that the anion will meet the metal-cation that travels outward from the centre of the metal, and form a contaminate salt and agglomerate, if the geometry is right i.e. presence of micro-cavities, nucleation of a pit may happen.
However there are some arguments against anion penetration and migration theories as the pitting which happens in SO42- ClO4- solutions cannot be explained by migration theories, because the anion is too large to penetrate the oxide film, also the nucleation process is often too fast. (Ref.14)
1.4.2. Point Defect Model
During the film growth the anion and cation move across the interface, the anion reacts will metal and forms a passive film, and the cation diffusion results in dissolution and creates a metal vacancy, this metal vacancy tends to submerge into the bulk of metal, however if the metal vacancy production rate is greater than the submersion rate, the vacancy will start to accumulate at the metal-oxide interface, when the void is greater than a critical size it may result a local collapse, this kind of film breakdown can happen without the presence of halide anions.(Ref. 17)
Relationship between pitting and stress corrosion cracking
There are many examples in industry of Initiation of SCC from corrosion pits in type 316 and 304 Austenitic stainless steel. It has been observed that SCC nucleates from corrosion pits. (Ref. 18)
There have been many studies on SCCs on Austenitic stainless steel with MgCl2 solution and all yield similar results; almost all SCC begin at the corrosion pits, apart from some cases with high MgCl2 concentration and temperature. In those case crack a tend to propagate along smooth metal surfaces and along slip planes. (Ref. 19)
1.5. Stress Corrosion Cracking
Stress corrosion is a term used to describe failures of materials caused by the synergistic effect of mechanical and chemical forces, is usually occur in bulk solution, the critical stress required to cause SCC can be very low, it can be much lower than the yield stress, if the environment is right. However SCC is controlled by many different factors, the optimum environment to cause SCC is different for different alloys, even within the same alloy different heat treatments can change there susceptibility to SCC, so there are infinite alloy/environment combinations, and this is a typical list of combined alloy/environment systems exhibiting SCC.(Ref.25)
TABLE 2 Alloy/Environment systems exhibiting SCC, redrawn from "Russel H. Jones: Stress-corrosion cracking materials performation and evaluation" P2
Alloy
Environment
Carbon steel
Hot nitrate, hydroxide, an carbonate/bicarbonate solutions
High-strength steels
Aqueous electrolyse, particularly when containing H2S
Austenitic stainless steels
Hot, concentrated chloride solutions: chloride contaminated steam
High-nickel alloys
High-purity steam
Alpha-brass
Ammoniacal solutions
Titanium alloys
Aqueous Cl- Br- and I- solutions: organic liquids N2O
Magnesium alloy
Aqueous Cl- solutions
Zirconium alloys
Aqueous Cl- solutions: organic liquids: I2 at 350 0C
Table 2 shows the typical environment causing SCC in different alloys and environments. It shows austenitic stainless steels susceptible in hot,concentrated chloride solutions, also the environment can be very different for different alloys.
SCC is a delayed failure process, the crack initiate and propagates at a very slow rate, until the stress in the remaining component exceeds the fracture strength and fractures. The problem with SCC is the initiation and propagation of crack can occur with very little outside evidence, so the component can fail catastrophically without any warning.
1.5.1. Crack Initiation
SCC usually starts from pit nucleation, cracks initiate at the corrosion pit (Ref.19),it has been proven that in many cases the pitting potential is the same as the SCC potential (Ref 26,27), thus the pitting and cracking are affected by the same parameter whether it be geometry of pit, stress, chemistry of the material.
However, a crack nucleating from a pit isn't the only way SCC initiates; some studes show that stress corrosion crack can initiate on a smooth surface, Kowaka and Kudo have experimented with 18Cr-8Ni stainless steel with 25% and 45% MgCl2 and have shown that SCC initiate at the bottom of pits at 25%MgCl2 solution. While SCC nucleated on a smooth surface in 45%MgCl2(Ref.28) can initiate on pre-existing cracks caused by fabrication defects.(Ref.25)
Although we know what causes crack initiation, the model for crack initiation has not been well developed, because crack initiation has not been precisely defined.
Crack initiation processes generate large numbers of micro cracks on the surface, which continue to grow and coalesce, the crack tips grow towards each other and coalesce to form larger cracks, not all the micro crack succeed in coalescing, most of them die out during the process. These cracks don't have any effect to the later stage of SCC. Once the large crack reach the critical threshold stress intensity, it moves on to next stage, crack propagation, this mechanism explains why the fractography on SCC specimen usually shows a branched pattern.(Ref.29)
1.5.2.Crack Propagation
There are several proposed mechanism for crack propagation which can be divided into slip dissolution and mechanical fracture models, in metallic materials.
1.5.2.1. Slip Dissolution Model
Figure.7 schematic illustration of slip dissolution mechanism (Ref.25) It is assumed that the crack propagates by dissolution of metal at the crack tip and form a active path by both chemical and mechanical interaction. There are serious considerations and debate surrounding this model, some literature shows that local plastic deformation at crack tips will create a new surface(Ref 30,31), where the bare metal will expose to the environment and start to dissolves. This results in the extension of the crack, and there are two different assumptions, if the film rupture rate is faster than the rate of repassivation the crack will continue to grow, if not, the crack tip will arrest by repassivation and the crack will propagate periodically by the emergence of slip step.(Ref.25)
This model works well on inter-granular corrosion and is widely accepted as correct in such cases, however, it can not explain the high cracking rate in austenitic stainless steels which SCC is transgranular. Transgranular SCC fracture usually has a flat surface and the crystallographic orientation show there is very little dissolution during the crack propagate, it is all against the film-rupture model so in general it is not accepted as mechanisms for transgranular SCC. (Ref.32)
Mechanical fracture models
This type of model assumes that the cracks propagate by dissolution and the remaining object fails by mechanical forces.
1.5.2.2. Tarnish Rupture Model
It was the first proposed to explain transgranular SCC, and it has been modified since in other studies and can be applied to inter-granular SCC. This model proposes that the passive film fractures under the applied stress, the bare metal is exposed to environment and the surface film reforms preferentially along grain boundaries, and the grain propagates by alternating film fracture and reform. (Ref.33,36)
Figure.8 schematic illustration of Tarnish rupture mechanism (Ref.25)There is some concern regarding this model, because it may not be the case that for all metals that there is preferential penetrate along the grain boundary, nevertheless there are no experimental result to completely confirm or reject this model.
1.5.2.3. Corrosion Tunnel Model
This model assumes that a long tunnel foms across the metal, and the corrosion tunnel will keep grow in diameter and length direction until the remaining ligament fails. Under tensile stress the tunnels will tend to become a flat shots, as is observed in asutenitic stainless steel(Ref.34). This model explains what happens in transgranular SCC in terms of formation and mechanical separation of corrosion slots.
1.5.2.4. Film Induced Cleavage Model
This model was first proposed by Edeleanu and Forty (Ref.37). It suggests that in a stress corrosion crack, between the metal and environment, there is a brittle thin layer. Under a tensile stress the brittle film ruptures, and the crack crosses the film/metal interface. It enters into the ductile metal matrix with little loss in velocity, the crack will propagate (the velocity can be up to hundreds meters per second) and then suddenly blunt and arrest. This is followed by forming a new protective film by de-alloying.
Figure.9 schematic illustration of Film induce cleavage mechanism (Ref.2)The critical point on this model is the brittle crack continuing to propagate in a ductile matrix. This is the model that most believe explains what happens in chloride induced SCC of austenisite stainless steel.
1.5.3. Hydrogen Embrittlement
There are many anodic based mechanisms in corrosion processes such as active dissolution, removal material from crack tip and de-alloying. Because an anodic reaction must have a corresponding cathodic reaction, in most case hydrogen is generated as a product of cathodic reaction in corrosion process. Hydrogen diffuses interstitially through the crystal lattice and restricts the motion of dislocations (Ref.21), this reduces the ductility of the metal. Research has found that, high strength steels are susceptible to hydrogen embrittlement. However Hydrogen Embrittlement only becomes significant when there are high hydrogen concentrations (Ref.23 24). It has been suggested that a small addition of nitrogen to a solid solution in ASS improves their resistance to Hydrogen embrittlement.(Ref.22)
1.5.4. Summary of SCC
There are certain conditions required for SCC to occur, they are mainly electrochemical, mechanical and metallurgical factors,(Ref.25)
Electrochemical parameters
Mechanical parameters
Metallurgical factors
pH
stress
localized microchemistry
oxidizing potential
stress intensity
bulk composition
impurity concentration
strain rate
deformation character
temperature
yield strength
SCC is a sophisticated corrosion process that happens in the material, because it is controlled by many factors. There isn't a universal model that applies for all materials, we need to apply different models for different materials in order to explain and understand what really happens in each case. Also, one thing we need to beware of, it not all SCC models a valid, some of then a just a conjecture and never prove by experiment , so far film induced cleavage model is the most plausible mechanism that can best explain chloride induced stress corrosion cracking in Austenistic stainless steel.
1.6. Atmospheric Induced Chloride Stress Corrosion Cracking(AISCC)
AISCC is one type of SCC which occurs in atmospheric conditions at ambient temperature. The difference between AISCC and SCC is, traditional SCC usually occurs in bulk solution and whereas AISCC occurs in thin electrolyte layers on the surface of metals and causes cracking. Conventional wisdom predicts that SCC in austenistic stainless steel usually occurs at temperatures over 60 OC, this is generally accepted (Ref.38), however in last two decades there have been several reports, of SCC failures in stainless steel components at temperatures far below 60 OC. this has prompted investigation to determine what condition aid this low temperature SCC.(Ref.39)
1.6.1. AISCC Failure
In the last few decades many stainless steel components have failed during their service life time including many parts in indoor swimming pools, such as bolts and hooks to wire. Some cause serious problems such as ceiling collapses; this kind of failure has happened in many areas including the UK, USA, Denmark, Sweden and Switerland (Ref.39). In all reported cases components show evidence of SCC, transgranular and branched crack trace, also acidic chloride deposits were found on the surface. It is believed that it is some effect of cold-worked hardening, because most of the failed parts had been cold worked or subject to additional stress during installation.(Ref.40)
Apart from the swimming pool atmosphere, many low temperature SCC have occurred in sensitized stainless steel due to interganular SCC. In 1987 Umemura F,Matukura S,Nakamura H, Kawamoto T, reported on experiments on unsensitized Type 304 stainless steel in saturated MgCl2 solution, originally with the specimens slow completely immerse, and then they withdrew the specimens and allowed them to dry completely under controlled relative humidity(RH). This resulted in transgranular SCC at 30 OC. Similar tests using NaCl did not crack. (Ref.41)
This is obviously contradicts the conventional wisdom because the operating temperature is far below the threshold temperature(60 OC) for conventional SCC in austenitis stainless steel. However that is not always the case, study has shown that when a specimen is exposed to a marine atmosphere, in Kure beach for example, it is full of chloride and other contaminants but hasn't shown signs of SCC after 26 years (Ref.42). This explains that the problem is restricted to a specific environment.
1.6.2. AISCC Study
T.Prosek, A.Iversen,C.Tazen, have investigated, low temperature stress corrosion cracking of stainless steels in atmospheric conditions with the presence of chloride deposits. They found that type 304 and 316 ASS were susceptible to SCC at 30%RH and 50 OC with droplets of magnesium or calcium chlorides under non-washing conditions. They also found that the aggressiveness salt deposits increased with decreasing RH and increasing temperature. SCC is controlled by the equilibrium chloride concentration in the surface, although the chloride deposit become harmless in terms of SCC when the RH is higher than70% or between 5-10%. This is because it stops the formation of concentrated chloride solutions, but it is still a problem when the RH is around 30%(Ref.40)
Shoji et al's, work showed, the effect of relative humidity and type of chloride salt in low temperature SCC on stainless steels. The summary of Shoji's main findings, wer that cracking was most prevalent when RH was in equilibrium with humidity of saturated solution, at ambient temperature. Under these conditions magnesium, calcium and zinc chlorides caused cracking within a range of RH approximately25-45% for MgCl2,15-45% for CaCl2 , and 8-40% for ZnCl2 however sodium chloride did not cause cracking.(Ref.13,44)
Table3. Equilibrium concentration of saturated Cl solution at room temperature (Ref.13)
Chloride
Saturated Concentration, wt% of Cl
Equilibrium humidity of saturated Cl solution %
Sodium Chloride (NaCl)
16
75
Magnesium Chloride (MgCl2)
27
33
Calcium Chloride (CaCl2)
29
31
Zinc Chloride (ZnCl2)
42
10
The equilibrium humidity of a saturated solution of magnesium and calcium chloride is around 30%, also the saturated concentration of chloride is much higher than sodium chloride, this means both magnesium and calcium chloride have a higher sensitivity to SCC. This agrees with the experimental results of both Shoji et al and T.Prosek, A.Iversen and C.Tazen. Zinc Chloride generates SCC at relatively low RH in comparison to the other chlorides. It was found that the corrosivity of a chloride deposit corresponds to the equilibrium humidity of Cl saturated solution.
Many experiments have been carried out by many researchers, the result are very similar and show type 304 and 316 ASS susceptible to SCC with specific environmental conditions at ambient temperature. However the mechanism is still unclear and there isn't a good model to predict AISCC, nevertheless investigation into AISCC show that this problem is caused by RH and type of chloride.
1.6.3. AISCC Problems
In 1.6.1. it was mentioned that AISCC causes many failures. This has now become a great concern to the nuclear industry, because type 304L and 316L stainless steels are chosen as container materials for radioactive waste in UK, The storage period can be up to 100 years (Ref.43), any leakage of radioactive materials can lead to catastrophic environmental consequences. Originally type 304L and 316L stainless steels were selected because of their corrosion resistance, especially the resistance to localised corrosion at ambient temperatures. However it has been demonstrated these materials are susceptible to AISCC at ambient temperatures, without even considering the effects of nuclear waste containment, stainless steel is widely used in many engineering applications, meaning there is an immediate need to widen our understanding of AISCC to prevent failure of steel materials at ambient temperature.
Conclusion
It has been confirmed that type 304L and 316L stainless steels are susceptible to AISCC over a humidity range of 10%-60% at ambient temperature, with presence of chloride deposits. Amongst these chloride deposits magnesium chloride, and calcium chloride have a greater effect than sodium chloride.
Most of the demonstrations were carried out using droplets or droplet solutions depositing salt on U-bend specimen. This is not analogous to real life situations. My focus in this project is to mimic aerosol deposition of a salt layer on a stainless steel surface to investigate AISCC in type 304L and 316L stainless steels.
