Why do things crack? A stress is applied that is strong enough to separate the material. Metals are theoretically much stronger than we find in practice, and this is due to “artifacts” within the microstructures that act as stress concentrators. Stress concentrators amplify the stress at a confined region and allow the material to crack earlier than the atomic bond strength would suggest. These “artifacts” can be things like inclusions, voids, grain boundaries, precipitates, and so on. Generally, the cleaner (less unwanted inclusions) or purer a metal is, the more resistant it is to cracking. There can be benefits from the “artifacts” mentioned above; remember, this is metallurgy and there is always a “however”.
By Justin Bekker, Metallurgical & Welding Engineer, Stress Engineering Services Canada
When do cracks cause failure? The existence of a crack may be acceptable for a long period of time; the act of cracking may relieve the stress on the part. However, it may also increase the stress on the remaining ligament (the remaining connected portion of the part that carries the load), and in this case, the crack may propagate.
Toughness is a measure of the metal’s ability to resist cracking, and this term encompasses both initiation of the crack and propagation of the crack. Austenitic stainless steels (3XX series) can maintain good toughness at low temperatures, whereas duplex, ferritic, and martensitic stainless steels have a ductile to brittle transition temperature where the toughness dramatically decreases at a certain temperature. This is why austenitic stainless steel must be used for cryogenic applications.
Inherent Forces Within a Weld
Regardless of the alloy, welds impart residual stresses on its joints. The residual stress is due to a temperature gradient between the base metal and the weld metal. The weld is hotter than the base metal, so its thermal expansion pulls against the colder base metal. Generally, the base metal has much more mass than the weld metal and confines the weld metal/HAZ so that it compressively yields when hot. Remember that if an object is allowed to freely thermally expand when heating, there is no stress on the object; but if the object is confined, it will feel compressive forces as it tries to thermally expand. Upon cooling, this results in a tensile stress within the weld and a compressive stress some distance from the fusion line, within the base metal. The stress gradient may look similar to Figure 1. Note that the residual stress from welds may be above the yield strength of the material. Welding parameters should be considered in order to minimize residual stress when it may lead to problems.
Cracking in Stainless Steel Welds
Stainless steel welds are susceptible to many cracking mechanisms that other alloys are subject to. For example, solidification cracking and liquation cracking.
Solidification cracking, sometimes referred to as hot cracking, occurs during solidification while the weld is still hot. This is a particular issue with austenitic stainless steel that comprises too little ferrite. As austenitic stainless steel solidifies from a liquid, ferrite forms and upon further cooling, most of this ferrite will eventually transform into austenite. At elevated temperatures, the ferrite is beneficial in dissolving phosphorous and sulfur, which are low melting impurities that may otherwise segregate out during solidification and lead to cracking during subsequent weld passes. Additionally, ferrite seems to act as a binder between solidifying austenitic dendrite columns, although this mechanism is not fully agreed upon. With too little ferrite, the parallel dendrite columns can split open as a crack.
The Schaeffler or WRC-1992 diagram is a graphical way to predict the ferrite content and determine the risk of solidification cracking in an austenitic stainless steel weld. For example, the chemistry composition listed in the below table would have a ferrite percentage of 7% using the formulas shown in the WRC-1992 diagram (the angled lines show the ferrite percentage). A ferrite content of 2 – 10% is recommended to avoid solidification cracking. In the case of weld metal, where dilution occurs from the base metals being joined, a slightly more complicated calculation can be carried out to determine the weld metal ferrite level by determining the ferrite level of the two base metals being joined, the filler metal, and the dilution level.
Liquation cracking occurs when a metal of lower melting temperature, like the zinc used in galvanizing or copper from the tip of a GMAW gun, comes in contact with a stainless steel weld. During solidification, the low melting temperature metal remains as a liquid at grain boundaries for a moment. The shrinkage stresses may then pull the grains apart, forming a crack.
One of stainless steel’s nemeses is chlorides. Chlorides are present in saltwater, such as swimming pools, oceans, or downhole drilling wells. Chloride ions can penetrate the protective chromium oxide layer, forming pits where accelerated corrosion may occur. This is the corrosion part of stress corrosion cracking (SCC) in stainless steels.
SCC also requires a stress to be present, which may come from loading on the component, or may occur from the residual stress of welding, as discussed earlier. The appearance of SCC is often fine colonies of parallel cracks, which may eventually form pinholes, indicating that SCC has progressed through the wall thickness of a pressure containing component. Elevated temperature exacerbates the aggressive effect of chlorides. Increasing the molybdenum and chromium content decreases the risk of chloride attack but may not fully prevent it. Switching to a duplex stainless steel further protects against chloride attack but again may not fully prevent it.
One common misconception is that hydrogen does not affect austenitic stainless steel; it does not lead to embrittlement to the same extent as carbon steels, but it may still lead to hydrogen cracking and should not be disregarded.
Corrosion in Stainless Steel Welds
Corrosion and cracking are not necessarily mutually exclusive. For example, SCC, as its name suggests, involves both corrosion and cracking. Similarly, sensitization may involve both.
Sensitization is a depletion of the chromium ions near the grain boundaries leading to corrosion along the grain boundaries. There is a specific temperature window, approximately 500 – 800°C (900 – 1500°F), where sensitization occurs; this is where chromium carbide formation is favourable. This forbidden temperature region is one of the reasons that post weld heat treatment of stainless steel is generally avoided. Obviously, a weld must be exposed to this temperature region, but it is generally for a very short interval and does not cause harm. Multipass welds that repetitively expose an area of the HAZ to this temperature region run the risk of sensitization. Joining stainless steel to carbon steel is typically done using a Type 309 filler metal containing additional chromium. The additional chromium is meant to account for the additional carbon from the carbon steel; extra chromium and carbon form chromium carbides (typically at the grain boundaries) and the intent is for sufficient chromium to be available to form the protective oxide that stainless steel is known for. If there is not sufficient chromium, then the grain boundaries are susceptible to preferential corrosion, which can result in grains essentially falling out of the material. Intergranular cracking may also occur along the corroded grain boundaries. A misconception is that the carbon steel will be more susceptible to corrosion than the sensitized weld, but that is not the case; the sensitized region will be the first area to corrode.
An alternative to using additional chromium to account for carbide formation is to use titanium or niobium, which more favorably form carbides than chromium does. Type 321 and 347 are “stabilized” with titanium and niobium, respectively, which can be used at higher temperatures. However, there is also a downfall to Type 321 and 347, and that is with multipass welds. The HAZ region adjacent to the fusion line of the first weld pass (Weld 1 in sketch) is hot enough to dissolve the titanium or niobium carbide and cools fast enough to minimize that carbide formation. A subsequent weld pass (Weld 3 in sketch) that reheats the previous HAZ at a lower temperature permits the chromium carbide formation, which is more favorable at the lower temperature. This creates a narrow band of HAZ (HAZ 3 in sketch) that has been sensitized and can preferentially corrode like a knife into the HAZ. Figure 2 below shows a cross-section of a knife-line attack.
Final Words
Similar to sensitization, dwelling between 600 – 900°C (1,100 – 1,750°F) encourages the formation of sigma phase. This is a particular concern with duplex stainless steel where it lowers the corrosion resistance and reduces toughness. The reduction in toughness makes a brittle failure more likely. The presence of sigma phase is determined using microscopy, which is a time-consuming process. The correct alloy and joining method should be carefully selected for each application. In severe cases, planned replacement intervals may be the best option.