Failure Analysis of Stellite6 Hardfacing on Steam Turbine Valve Bodies
Literature Overview and Significance
This paper from the Henan Provincial Institute of Boiler and Pressure Vessel Inspection addresses a failure mode that I have seen too many times in the field: the spalling of cobalt-based Stellite6 hardfacing from a COSTE steel valve body after six years of service. The authors systematically investigated the root cause using microstructural analysis and hardness testing, and their findings point to a mechanism that many engineers overlook during the welding procedure qualification phase. For those of us who work with high-temperature steam service components, this is a classic case of dilution-related degradation that develops over years rather than days. The paper is particularly valuable because it connects the welding process parameters directly to the long-term service behaviour, which is exactly the kind of thinking that separates a good hardfacing job from a premature failure.
Root Cause Analysis: Dilution and Interdiffusion
The fundamental issue identified in this study is the interaction between the COSTE steel substrate and the Stellite6 deposit during the welding process. When using shielded metal arc welding (SMAW) to deposit Stellite6 onto a carbon steel or low-alloy steel base, the dilution of the deposit by the base metal is inevitable. The authors found that iron from the COSTE steel migrates into the Stellite6 layer near the fusion boundary during welding, creating a concentration gradient that persists after solidification. This is not merely a cosmetic issue; it sets the stage for subsequent degradation.
During six years of service at elevated temperature, the iron continues to diffuse from the substrate into the hardfacing layer over long distances. This is a time-dependent process that follows Fickian diffusion kinetics, and the driving force is the concentration gradient established during welding. As the iron content in the Stellite6 layer increases beyond a critical threshold, the original face-centred cubic alpha-Co(Cr,Fe,W) solid solution becomes thermodynamically unstable. The authors identified that this supersaturated solid solution undergoes a eutectoid transformation, producing two brittle phases: the body-centred cubic alpha-prime-FeCo(Cr,W) phase and the tetragonal sigma-CrCo(Fe,W) phase. Both of these intermetallic compounds are hard and brittle, with limited ductility, and their formation within the hardfacing layer significantly degrades its toughness and resistance to thermal fatigue.
The transformation mechanism can be summarised as follows:
| Parameter | As-Deposited Condition | After 6 Years Service |
|---|---|---|
| Fe content near interface | Elevated due to dilution | Significantly increased by diffusion |
| Matrix phase | alpha-Co(Cr,Fe,W) FCC | Partially transformed |
| Secondary phases | None observed | alpha-prime-FeCo(Cr,W) BCC + sigma-CrCo(Fe,W) tetragonal |
| Microhardness | Relatively uniform | Localised hard spots with brittle character |
| Fracture behaviour | Ductile | Brittle, intergranular |
Metallurgical Interpretation and Practical Implications
The eutectoid transformation described in this paper is a classic example of a diffusion-controlled phase change that occurs when the alloy composition drifts outside the intended specification. In the Stellite6 system, the nominal composition is approximately 28-32% Cr, 4-6% W, 1-1.2% C, and the balance cobalt. The iron content is typically limited to less than 3% in the virgin alloy. However, when welding onto a steel substrate, the dilution can easily raise the iron content to 10-15% or even higher near the fusion line, depending on the welding parameters and the number of layers deposited.
The practical lesson here is that dilution control is not just a quality issue; it is a life-limiting factor for hardfaced components. In my experience, the most effective way to control dilution when hardfacing valve bodies is to use a buttering technique. This involves depositing a layer of nickel-based alloy or a high-purity cobalt-based alloy directly onto the steel substrate before applying the final Stellite6 layer. The buttering layer acts as a diffusion barrier and reduces the iron dilution into the functional hardfacing surface. Additionally, using the lowest practical heat input and maintaining a short arc length during SMAW can help minimise the depth of penetration and thus reduce dilution.
The authors also noted that the transformation to brittle phases is accelerated by the combination of long-term thermal exposure and cyclic mechanical loading. This is particularly relevant for steam turbine valves, which experience both steady-state creep conditions and transient thermal stresses during start-up and shut-down cycles. The brittle sigma phase has a coefficient of thermal expansion that differs from the surrounding matrix, which generates localised stress concentrations when the component is thermally cycled. Over time, these stresses nucleate microcracks that propagate along the brittle phase boundaries, ultimately leading to the observed spalling.
Engineering Practice and Lessons Learned
From a practical standpoint, this failure analysis highlights several critical considerations for anyone involved in the hardfacing of steam turbine components. First, the welding procedure qualification must include not just mechanical testing of the as-welded deposit but also a careful assessment of the dilution zone. This should involve chemical analysis of the deposit at various distances from the fusion line, using energy-dispersive X-ray spectroscopy or wavelength-dispersive X-ray spectroscopy in the scanning electron microscope. A dilution profile that shows an iron content exceeding 5% at a depth of 2 mm from the surface is a red flag that should trigger a revision of the welding parameters.
Second, the selection of the hardfacing alloy must be matched to the expected service conditions. In high-temperature steam service where the operating temperature exceeds 500 degrees Celsius, the resistance to iron diffusion becomes a critical property. While Stellite6 is an excellent choice for its hardness and wear resistance, its susceptibility to iron diffusion at elevated temperatures is a known limitation. Alternative alloys such as Stellite21, which has a lower carbon content and higher nickel content, may offer better resistance to dilution-related degradation. Similarly, nickel-based hardfacing alloys such as NiCrBSi or NiCrWSi systems can be considered where the service temperature is high and iron diffusion is a concern.
Third, the inspection regime for hardfaced components should include periodic non-destructive testing to detect early signs of degradation. Ultrasonic testing can be used to detect disbonding at the interface, while surface replication techniques combined with metallographic examination can reveal the onset of sigma phase formation before it reaches a critical level. In the case described in this paper, the failure was detected only after the hardfacing had already spalled, which represents a late-stage event. An effective condition monitoring programme should aim to detect the degradation at an earlier stage, when corrective action is still feasible.
Key Questions and Reflections
This paper raises an important question that I have often debated with colleagues: should the welding procedure qualification for hardfacing include a simulated service ageing test? The answer, in my view, is yes for critical components such as turbine valves, but the practical challenge is that such tests are time-consuming and expensive. A compromise approach is to use accelerated ageing at a temperature above the service temperature but below the transformation temperature, combined with thermodynamic modelling to predict the phase stability. The CALPHAD approach, for example, can be used to calculate the equilibrium phase fractions as a function of composition and temperature, which provides a useful screening tool before committing to long-term testing.
Another reflection relates to the role of welding consumable selection. In the case described, the authors used SMAW with a Stellite6 electrode. In modern practice, laser cladding and plasma transferred arc (PTA) welding are increasingly used for valve hardfacing because they offer lower dilution and more precise control of the deposit geometry. However, these processes require different welding parameters and produce different microstructures. The failure mechanism described in this paper is not unique to SMAW; it can occur with any process that results in significant dilution. The key is to verify the actual dilution achieved in production and to ensure that the composition of the deposit, after dilution, remains within the acceptable range for the intended service conditions.
Study Insights and Implications
The most valuable insight from this study is that hardfacing failures are often not caused by a single factor but by the interaction between welding-induced dilution and service-induced diffusion. The authors have clearly demonstrated that the iron content in the deposit is the critical parameter that controls the phase stability of the Stellite6 layer. This means that the welding procedure qualification should focus not only on achieving the required hardness and wear resistance but also on maintaining the chemical composition of the deposit within a narrow window. In practice, this requires careful control of the welding parameters, the number of layers deposited, and the interpass temperature. It also requires a thorough understanding of the dilution behaviour of the specific welding process being used.
The sigma phase formation described in this paper is a reminder that even the most well-established hardfacing alloys have their limits. The Stellite family of alloys was developed for applications where wear resistance is the primary concern, and their high-temperature performance is a secondary consideration. When these alloys are used in high-temperature service, the engineer must be aware of the potential for phase transformations that can embrittle the deposit over time. This is particularly important in the power generation industry, where components are expected to operate for decades without failure.
Conclusion and Reference Value
This paper makes a significant contribution to the understanding of hardfacing degradation mechanisms in high-temperature service. The authors have provided a clear and well-supported explanation of the failure process, from the initial dilution during welding to the final spalling after six years of service. The identification of the eutectoid transformation as the root cause is particularly valuable, as it provides a specific metallurgical mechanism that can be used to predict the service life of hardfaced components. For the practicing engineer, the key takeaways are the importance of dilution control, the need for a robust welding procedure qualification that includes chemical analysis of the deposit, and the value of periodic condition monitoring for critical hardfaced components. The paper is a useful reference for anyone involved in the hardfacing of steam turbine components or similar high-temperature service equipment.
Zhuojin Pipe Fitting Co., Ltd