Overview of Stress Corrosion Cracking in Stainless Steel: Electronic Enclosures in Extreme Environmental Conditions Stephani Gulbrandsen
Introduction: Electronic Enclosures 316L stainless steel is found in the electronics industry as an enclosure for electronic assemblies in corrosive environments. Enclosures protect electronics from the environment, which is particularly useful in corrosive environments, such as salt water. Examples of industries that benefit from stainless steel enclosure of their electronics include cow milking, food processing, car and window washing, and oil refineries. Unfortunately, these stainless steel enclosures can be susceptible to corrosion and stress corrosion cracking (SCC). General corrosion is more evident because it occurs over the entire surface. SCC is more insidious because it occurs on localized areas of the stainless steel enclosure and can be barely noticeable. SCC can cause severe failure, which results in exposure of the electronics assembly to the harsh environment. Overview of Stress Corrosion Cracking (SCC) Material, environment, and stress are the three key factors that impact SCC. Only under very specific combinations of these three variables can SCC occur, and adjusting just one of the variables can change the regime from SCC to no SCC or vice versa as can be seen in Figure 1. The stress in the system needs to be tensile. It can be applied or residual and it causes cracks to form at the surface and propagate through the material. Material Environment SCC Stress Figure 1: Three factors that impact SCC. SCC is only possible in selective environments where a passive film forms at the metal surface. Under tensile stresses if the environment is too aggressive and the film formation rate is very slow then general corrosion occurs. If the rate of surface film formation is very fast, then the passive film at the metal surface will immediately repassivate upon rupture [1, 2]. In SCC, the surface is able to passivate, but when the film ruptures the surface of the sample temporarily remains exposed to the environment and locally corrode. Examples of all three conditions are presented in Figure 2, which shoes a 316L stainless steel base metal that forms a chromium oxide (Cr2O3) passive film. The circled semi-stable film behavior is the one that leads to SCC.
Figure 2: Types of passive film behavior; dependent on film stability in an environment. Stress corrosion cracks can initiative at abnormalities on the metal surface, including pits, grain or phase boundaries, inclusions, secondary phases, or physical defects such as scratches. Once a crack initiates it can propagate intergranularly or transgranularly through the microstructure. Crack growth occurs when reactions occur faster at the crack tip than on the metal surface or crack sides [2]. Once the crack reaches a critical crack length it continues to propagate through the metal and the remainder of the fracture surface fails. The fracture surface is indicative of brittle failure, and other fractographic features are dependent on the SCC mechanism. Material: 316L Stainless Steel Stainless steel, particularly 316L, is widely used in highly corrosive environments, particularly in saltwater (i.e. aqueous chloride) applications. It is more corrosion resistant than 304L due to added molybdenum (Mo). However, this addition comes with an increased cost. The nominal composition of 316L can be seen in Table 1. It is an iron (Fe)-based, low carbon (C) stainless steel with alloying elements to promote corrosion resistance, stabilize the austenitic phase, and maintain adequate strength and ductility. Possible crack initiation sites in 316L are secondary phases, inclusions, and stress induced localized areas of film rupture. Table 1: Nominal composition of 316L austenitic stainless steel. C Mn Si P S Cr Mo Ni N Fe 16.00 2.00 10.00 0.03 2.00 0.75 0.045 0.03 0.10 Balance 18.00 3.00 14.00
Nickel (Ni) is added to increase impact strength of the steel and stabilize the face centered cubic (FCC) austenitic phase [3]. This phenomenon is supported by the Schaffler diagram and Cr and Ni equivalents. The equivalency equations can be seen in Figure 3. For 316L the nickel equivalent ranges from 12 to 16, and the chromium equivalent ranges from 19 to 22. This places 316L in the shaded box in Figure 3. 316L is on the boundary between austenitic and ferritic stainless steels. Austenitic stainless steels are known for corrosion resistance and ferritic stainless steel are known for strength. Because 316L lies on the edge of these two microstructures, it maintains adequate strength and corrosion resistance. Figure 3: Schaffler diagram for microstructure of steels; 316L falls in the shaded blue rectangle. Both Mo and Cr are added to increase corrosion resistance. The Cr promotes passivation of the surface by forming the protective Cr2O3 film. As the Cr content is increased, the film becomes more stable and the corrosion rate decreases [4]. It is unclear how Mo enhances the corrosion resistance in chloride containing environments [3,5]. However, it has been proposed that Mo could thicken the existing passive film, increase the surface affinity for oxygen, which decreases propensity for chloride adsorption, or form a more protective secondary film that is a glassy structured amorphous oxide [6]. Some additives in stainless steel, like sulfur (S) and silicon (Si), can be beneficial and detrimental to corrosion resistance. S decreases pitting, however once active dissolution occurs it can accelerate this process and poison repassivation [7]. Similarly, Si is added to 316L to reduce pit initiation, however, once pits form, Si promotes growth [3]. A concern with adding S is that it can form inclusions, which are anodic relative to the surrounding S-free matrix [8]. S combines with manganese (Mn) to form manganese sulfide (MnS). MnS inclusions are active, particularly in acidic, aqueous, chloride containing environments where they dissolve and form pits on the sample surface [8, 9]. This dissolution creates localized anodic areas on the metal surface, which attract chloride ions, promoting pit formation [10]. S particles can also complex with titanium (Ti) and Cr, however these sulfides are insoluble and not reactive in aqueous environments, so do not pose a problem to pit initiation [10].
Oxide inclusions are the other type of inclusion common in 316L and they are more stable than sulfide inclusions. They typically contain Cr, Mn, Ti, and aluminum (Al) with a hypothesized titanium dioxide-manganese oxide-aluminum oxide-chromium oxide (TiO2-MnO-Al2O3-Cr2O3) structure [8, 9]. These compounds do not readily corrode or act as pit initiation sites [9]. Like inclusions, secondary phases can become crack initiation sites in stainless steels. The two types of secondary phases found in 316L are chromium carbides and sigma phase. Sigma phase can be susceptible to corrosion in aqueous, acidic, chloride environments [11]. It forms in Mo containing stainless steels between 540 and 1000 C preferentially at grain boundaries [3]. This impairs corrosion resistance at the grain boundaries by depleting surrounding areas of the film stabilizing Mo [3]. The other type of secondary phase common to stainless steels is chromium carbide ((Fe,Cr)23C6) which precipitate at grain boundaries between 425 and 875 C [3]. Since 316L is being used for this study, it contains less than 0.03 weight % C, which slows the carbide formation reaction down to prevent formation in the above temperature range [3]. As a result, this phase is not a possible crack initiation site. The constituents in 316L significantly increase the corrosion resistance of the alloy, however the secondary phases and inclusions that can form in the material during processing and manufacturing, can promote SCC. Environment: Aqueous Chloride As noted in Figure 1, environment is another key factor in understanding the susceptibility to SCC. Chlorides are known to promote SCC in aqueous environments for austenitic stainless steels such as 316L because of the aggressive nature of the ion. Halide ions initiate cracks via pit formation on the surface [2]. As the chloride content increases, the tendency for pitting also increases [12]. Chloride ions preferentially initiate pits at S-rich inclusions, particularly MnS inclusions [7, 8, 10]. When the passive film ruptures at a MnS inclusion, chloride ions adsorb to the inclusion, causing dissolution, as presented in Figure 4 [8]. After the inclusion dissolves the metal surface should repassivate, but if a chloride or salt film forms where the pit had been this can lead to pit propagation, especially if there is a locally increased concentration of chloride within the pit [10]. Alternately, the pit is a stress concentration on the sample surface and the tensile stress preferentially ruptures the reformed passive film at this location, exposing base metal and allowing for a crack to initiate and propagate.
Figure 4: Chloride pit initiation process at MnS inclusions on 316L stainless steel surface [10]. Stress: Tensile Stress, the third variable in SCC, varies depending on the physical state of the material, both internally and externally. To promote SCC, the stress needs to be tensile. It can be a stress applied to the part or it can be a residual stress that is in the part from forming or processing. As the crack initiation site grows, it eventually reaches a critical crack length for the tensile stress being experienced, and the crack propagates through the material, causing failure.
SCC Mechanisms The way in which crack initiation and propagation occur is grouped into SCC mechanisms. A list of generalized mechanisms that many material-environment-stress systems fall under is presented in Table 2. The mechanisms are categorized based on whether they initiate because of metal dissolution or a mechanical process. Understanding how the mechanism works and the fractographic features help identify which mechanism applies to which systems. Table 2: SCC mechanisms [13]. Type Title Mechanism Fractographic Features Dissolution Film-rupture Stress opens crack and Crack-arrest marks ruptures protective film; Intergranular tip can remain active or repassivate and rupture by stress or slip steps Dissolution Active path Composition difference in microstructure causes Intergranular dissolution; could be along grain boundary Mechanical Corrosion tunnel Cracks form at slip steps emerged on metal surface Mechanical Adsorption-enhanced plasticity (i.e. liquid metal embrittlement) Chemisorption from environment embrittles metal Mechanical Tarnish-rupture Brittle film forms on surface and fractures, exposed metal reacts and reforms film, process repeats Mechanical Film-induced cleavage Film on surface cracks and crack propagates into metal, crack blunts and arrests; tip can remain active or repassivate with repetitive process Mechanical Adsorption-induced brittle fracture Adsorption of environmental species lowers interatomic bond strength and stress Microvoid coalescence on peaks Grooved fracture surface Slots Transgranular Slip at crack tip Voids ahead of crack Transgranular Intergranular (show crack-arrest marks) Crack-arrest marks Discontinuous crack propagation Brittle cracking in metal Cleavage-like fracture Continuous crack propagation
Mechanical Mechanical Adsorption-induced brittle fracture Hydrogen embrittlement needed for cleavage Adsorption of environmental species lowers interatomic bond strength and stress needed for cleavage Hydrogen absorbed by metal Continuous crack propagation Slip at crack tip Voids ahead of crack The stainless steel-aqueous chloride environment fits best with a dissolution film-rupture model SCC mechanism. When the Cr2O3 passive film on the 316L stainless steel surface cracks, the bare metal is exposed and exposed MnS inclusions preferentially dissolve from reacting with the chloride ions causing a pit to form on the surface. This pit creates a localized area of stress concentration that can remain active or repassivate and recrack depending on the environmental conditions. As the pit grows, the threshold stress needed to initiate cracking decreases until the applied stress is able to cause a crack to form and grow on the sample surface. An example of chloride induced SCC for 316L is presented in Figure 5. The cracks initiate from the pit along the sample surface and propagate intergranularly through the material, a fractographic indication of the dissolution film-rupture model. Pit Intergranular SCC cracks Figure 5: 316L stainless steel that underwent chloride-induced SCC [14]. Testing for SCC Susceptibility Mechanical testing is crucial to evaluate SCC of a given material in an environment under known stress conditions. When determining the most appropriate test to use it is important to look at the sample type, test conditions and constraints, and test set-up. It is important to take preventative actions so the test apparatus does not interfere with the test.
The two types of SCC test samples are smooth and notched or precracked specimens. When a sample contains a notch or precrack, the stresses are concentrated at this spot along the sample, increasing the chances for cracking at a known location [15]. If cracking were to occur, it would likely start at the notch or precrack. Also, a precrack provides information on the propensity for a crack to propagate in a given environment [15]. If a crack is already present on a sample surface, when immersed in solution if the environment promotes SCC the crack will grow, whereas if it does not cause SCC the crack will remain the same length. SCC tests can be held under constant strain or constant load conditions. Most, including slow strain rate tensile tests, bent-beam tests, and U-bend tests have constant strain conditions, while C-ring tests can be under constant strain or constant load [15]. Slow strain rate tensile tests use a tensile dog bone sample that is pulled at an initial strain rate of 10-5 to 10-7 s -1 to accelerate the SCC phenomenon. If the strain rate is too fast, the sample will fracture without SCC because there is not enough time for stress corrosion cracks to initiate on the surface [15]. As cracks grow, the applied stress decreases because the load bearing area decreases [15]. In bent-beam tests, a rectangular sample is placed in a 2-,3-, or 4-point loaded sample holder that curves the sample introducing constant strain. The sample and holder are placed in the test enviornment. These tests only test stress levels below the elastic limit for the sample and multiple samples can be placed in the same environment at the same time [15]. In all three sample holders, the maximum stress occurs at the midpoint of the convex surface and there is zero stress at the ends of the sample in the holder [15]. Of the three options, the 4-point loader is preferred because it creates uniform longitudinal stress on the convex surface [15]. The 3-point loader can introduce extra pressure at the center due to the central support. U-bend tests, the third type of constant strain test, consist of a rectangular bar specimen bent 180 into the shape of a U and placed in the test environment. The stress conditions of the sample are unknown, and only ductile samples that do not crack with bending can be used [15]. These samples are typically in a high stress state that causes quick crack propagation [15]. The final type of test, C-ring tests, can be constant load or constant strain conditions. For this test, a C-shaped specimen is machined from the sample, and a screw is placed between the two ends of the C and tightened to produce the desired test conditions [15]. Once the sample is prepared it is placed into the test environment. Of these tests, only the slow strain rate test will definitely fail and it takes the shortest amount of time. The time will be less than or equal to the time it takes the material to fail in an inert environment. The other three tests require periodic visual analysis of the samples to look for crack initiation. Oftentimes, this includes interrupting the test, removing the sample from the environment, visually analyzing it, then returning it to the environment if no cracks are present. These tests can last on the order of 5-10 weeks, though multiple samples can be placed in the same test environment. In slow strain rate tests the sample is not loaded until after it comes in contact with the environment. The sample can be preloaded with stress once it is set up in the test apparatus. The bent-beam, U-bend, and C-ring tests are typically put in the sample holder and then placed in
the test environment. The samples are stressed initially outside of the test environment, which could impact SCC behavior. Finally, it is important in all test set-ups to pay attention to the occurrence of crevices and the potential for galvanic corrosion caused by the sample holder. Of the types of tests listed, slow strain rate tensile tests are the preferred test because they take less time (although that means the SCC process is accelerated) and the sample does not experience any types of stress until placed in the test environment. The test apparatus holds the ends of the specimen and does not come in contact with the narrow region of the specimens that experience the highest stress, making it less likely for crevices to occur. Depending on the test setup, it is possible to only expose the gage length of the sample to the test environment, which would eliminate the potential for galvanic corrosion. Summary SCC is a complicated process that requires the proper combination of material, environment, and stress to occur. 316L stainless steel, a common adequately corrosion resistant metal, is susceptible to SCC in aqueous, chloride containing environment. It is commonly used in the electronics industry as an enclosure to protect electronic assemblies from the hard environment. These enclosures can be susceptible to SCC and failure, especially in aqueous chloride environments. Typically, the cracks initiate from pits formed at MnS inclusions on the sample surface. As these pits remain active and grow, they eventually reach a critical length that allows a crack to propagate through the metal. This critical crack length is dependent on the magnitude and direction of the tensile stress. The SCC mechanism follows a film-induced cleavage model, with fracture surfaces showing evidence of brittle cracking, crack arrest marks, and discontinuous cleavage-like fracture. There are a number of mechanical tests that be performed to test a materials SCC susceptibility in a given environment, including slow strain rate, bent-beam, U-bend, and C-ring tests. References 1 R. C. Newman. Review and hypothesis for the stress corrosion mechanism of carbon steel in alcohols. Corrosion, vol. 64, pp. 819-823, 2008. 2 R. N. Parkins. Predictive approaches to stress corrosion cracking failure. Corrosion Science, vol. 20, pp. 147-166, 1980. 3 Austenitic and ferritic stainless steels in Uhlig s Corrosion Handbook, 2 nd ed. M. A. Streicher. R. W. Revie, ed., 2000, pp. 601-650. 4 Y. F. Cheng, J. Bullerwell, F. R. Steward. Electrochemical investigation of the corrosion behavior of chromium-modified carbon steels in water. Electrochimica Acta, vol. 48, pp. 1521-1530, 2003. 5 Stress-Corrosion Cracking Materials Performance and Evaluation, A. J. Sedricks. R. H. Jones, ed., 1992, pp. 91-130.
6 M. Urquidi-Macdonald, D. D. Macdonald. Theoretical analysis of the effects of alloying elements on distribution functions of passivity breakdown. Journal of the Electrochemical Society, vol. 136, pp. 961-967, 1989. 7 J. Stewart, D. E. Williams. The initiation of pitting corrosion on austenitic stainless steel: on the role and importance of sulphide inclusions. Corrosion Science, vol. 33, pp. 457-474, 1992. 8 J. E. Castle, R. Ke. Studies by auger spectroscopy of pit initiation at the site of inclusions in stainless steel. Corrosion Science, vol. 30, pp. 409-428, 1990. 9 M. A. Baker, J. E. Castle. The initiation of pitting corrosion of stainless steels at oxide inclusions. Corrosion Science, vol. 33, pp. 1295-1312, 1992. 10 M. A. Baker, J. E. Castle. The initiation of pitting corrosion at MnS inclusions. Corrosion Science, vol. 34, pp. 667-682, 1993. 11 R. Chaves, I. Costa, H. G. demelo, S. Wolynec. Evaluation of selective corrosion in UNS S31803 duplex stainless steel with electrochemical impedance spectroscopy. Electrochimica Acta, vol. 51, pp. 1842-1846, 2006. 12 Corrosion of Stainless Steels, 2 nd ed. A. J. Sedricks, John Wiley & Sons Inc.: New York, 1996, pp. 148. 13 Stress-Corrosion Cracking Materials Performance and Evaluation, A. J. Sedricks. R. H. Jones, ed., 1992, pp. 1-40. 14 H. Ahluwalia. Stress corrosion cracking. Internet: http://csidesigns.com/flowgeeks/stresscorrosion-cracking/ [September 6, 2012]. 15 Stress Corrosion Cracking Test Methods, A. J. Sedricks, B. C. Syrett, ed., NACE: Houston, TX 1990.