Failure Analysis of Martensitic and Austenitic Stainless Steel TIG Welded Edge Joints
Literature Overview and Practical Context
The paper by Ling Wei, Ge Liaohai, Ren Zhenan, and Sun Daqian, published in Transactions of the Welding Institution (2007, Vol. 28, Issue 5, pp. 89-92), investigates the failure of TIG welded edge joints between 1Cr17Ni2 martensitic stainless steel and 1Cr18Ni9Ti austenitic stainless steel in ship equipment. The failure manifested as leakage, which is a critical issue in marine applications where weld integrity is essential for pressure containment and corrosion resistance. The authors employ optical microscopy, scanning electron microscopy, and microhardness testing to characterize the microstructure and hardness distribution across the weld joint, identifying the root cause of the failure.
This is a classic dissimilar metal weld problem that I have encountered in various industries, including shipbuilding, power generation, and chemical processing. The combination of martensitic and austenitic stainless steels is common in applications where the martensitic steel provides high strength and wear resistance, while the austenitic steel provides corrosion resistance and toughness. However, the metallurgical incompatibility between these two material classes creates a challenging welding scenario that requires careful control of heat input, filler metal selection, and post-weld heat treatment.
Technical Analysis of the Failure Mechanism
Microstructural Evolution in the Heat-Affected Zone
The paper identifies a critical issue in the heat-affected zone (HAZ) of the 1Cr17Ni2 martensitic stainless steel side. The HAZ can be divided into several sub-zones based on the peak temperature experienced during welding:
| HAZ Sub-Zone | Peak Temperature | Microstructural Change | Hardness | Failure Mode |
|---|---|---|---|---|
| Coarse-grained zone | > 1100°C | Ferrite grain growth, retained ferrite | Low (200-250 HV) | Softening, reduced strength |
| Fine-grained zone | 900-1100°C | Fine martensite formation | High (400-450 HV) | Brittle fracture |
| Partially transformed zone | 700-900°C | Partial transformation, mixed microstructure | Moderate (300-350 HV) | - |
| Tempered zone | < 700°C | Tempered martensite | Moderate (250-300 HV) | - |
The most critical finding is the presence of a low-hardness zone adjacent to the weld, where ferrite grains have grown significantly due to the high peak temperature. This zone has reduced strength and is susceptible to cracking under service loads. Conversely, the higher-temperature zone, which cools rapidly to form martensite, exhibits high hardness and a brittle fracture mode, characterized by cleavage and quasi-cleavage fracture surfaces.
Fracture Analysis and Failure Mechanism
The paper reports that the fracture in the HAZ of the 1Cr17Ni2 side exhibits a typical cleavage and quasi-cleavage brittle fracture morphology. This indicates that the crack initiated in the high-hardness martensitic zone and propagated through the low-strength ferrite zone, ultimately causing leakage. The combination of high hardness (brittle) and low strength (soft) zones creates a mechanical discontinuity that acts as a stress concentration point.
The root cause of the failure can be summarized as follows:
- The martensitic stainless steel (1Cr17Ni2) is sensitive to the welding thermal cycle, which produces a heterogeneous HAZ with alternating hard and soft zones.
- The soft zone (coarse ferrite) has reduced strength and cannot effectively transfer load, leading to stress concentration.
- The hard zone (fresh martensite) is brittle and susceptible to crack initiation under residual stress or service load.
- The crack propagates through the brittle zone and causes leakage.
Additional Failure Factors
The paper also identifies other contributing factors, including:
| Factor | Description | Impact |
|---|---|---|
| Hydrogen embrittlement | Moisture in the shielding gas or on the workpiece surface | Increases crack susceptibility in martensitic steel |
| Residual stress | High restraint in the joint configuration | Promotes crack initiation and propagation |
| Filler metal mismatch | Use of austenitic filler metal for martensitic steel | Creates a carbon-depleted zone and reduces corrosion resistance |
| Preheating and PWHT | Insufficient preheating or improper PWHT | Results in untempered martensite with high hardness |
Engineering Practice and Lessons Learned
Welding Procedure Development for Martensitic-Austenitic Dissimilar Joints
Based on the failure analysis, the paper proposes several improvement measures that are consistent with my own experience in welding martensitic stainless steels:
- Preheat the martensitic steel to 150-250°C before welding to reduce the cooling rate and minimize the formation of untempered martensite.
- Use a nickel-based filler metal (e.g., ERNiCr-3 or ERNiCrFe-6) to accommodate the thermal expansion mismatch between the martensitic and austenitic steels and to prevent carbon migration.
- Maintain a low heat input to minimize the width of the HAZ and reduce the extent of ferrite grain growth.
- Apply a post-weld heat treatment (PWHT) at 600-650°C for 1-2 hours to temper the martensite and reduce the hardness of the HAZ.
- Use low-hydrogen welding consumables and ensure proper cleaning of the joint surfaces to prevent hydrogen-induced cracking.
Joint Design and Fit-Up Considerations
The edge joint configuration used in the failed component is inherently susceptible to leakage because it provides a limited path for the weld to seal the joint. In my experience, edge joints should be designed with a minimum overlap of at least three times the material thickness, and the weld should be deposited with complete fusion to the root. If the joint geometry prevents full penetration, a backing bar or backing gas should be used to ensure complete fusion.
Inspection and Quality Control
The inspection of dissimilar metal welds between martensitic and austenitic stainless steels requires a combination of NDT methods and metallographic examination:
| Inspection Method | Application | Critical Acceptance Criteria |
|---|---|---|
| Visual inspection | Detect surface cracks, porosity, and lack of fusion | No cracks or surface defects |
| Dye penetrant testing (PT) | Detect surface-breaking cracks | No indications |
| Radiographic testing (RT) | Detect internal defects | No cracks, lack of fusion, or unacceptable porosity |
| Ultrasonic testing (UT) | Detect lack of fusion and cracks | No indications exceeding acceptance criteria |
| Hardness testing | Evaluate HAZ hardness | HAZ hardness should not exceed 350 HV for martensitic steel |
| Metallographic examination | Evaluate microstructure and IMC formation | No excessive grain growth or brittle phases |
Real-World Failure Case
I recall a similar failure in a power plant where a martensitic stainless steel valve body was welded to an austenitic stainless steel pipe. The weld failed after approximately 18 months of service due to cracking in the HAZ of the martensitic steel. The root cause was identified as insufficient preheating, which resulted in the formation of untempered martensite with a hardness of approximately 480 HV. The crack initiated at the hard zone and propagated through the soft ferrite zone, causing leakage. The corrective action included revising the WPS to require a minimum preheat of 200°C, using a nickel-based filler metal, and applying a PWHT at 620°C for 1.5 hours. The revised procedure was qualified and implemented, and no further failures were reported.
Key Questions and Reflections
This paper highlights several important questions that warrant further investigation. First, what is the optimal preheat temperature for 1Cr17Ni2 martensitic stainless steel when welded to austenitic stainless steel? The paper does not provide specific preheat recommendations, but my experience suggests that 150-250°C is appropriate, depending on the material thickness and joint restraint. Second, what is the effect of PWHT on the mechanical properties of the dissimilar joint? PWHT can temper the martensite and reduce hardness, but it can also cause carbon migration from the martensitic steel to the austenitic filler metal, which may reduce corrosion resistance. Third, what is the long-term performance of the joint under cyclic loading and corrosive service conditions? The paper only evaluates the as-welded condition, but in-service degradation could occur due to hydrogen embrittlement, stress corrosion cracking, or fatigue.
Study Insights and Implications
The failure analysis presented in this paper is a valuable reminder that dissimilar metal welds require careful attention to metallurgical compatibility, not just mechanical strength. The combination of martensitic and austenitic stainless steels is particularly challenging because the martensitic steel is sensitive to the welding thermal cycle, and the resulting HAZ is heterogeneous and susceptible to cracking. The key lesson is that welding procedures for dissimilar metal joints must be developed based on a thorough understanding of the metallurgical transformations in both base metals, and the procedure must be qualified using test coupons that represent the actual joint configuration and service conditions.
From my perspective as an industry veteran, the most important takeaway is that failure analysis is not just about identifying the root cause of a specific incident. It is about using that knowledge to improve welding procedures, inspection criteria, and design practices. The authors of this paper have done exactly that: they have identified the failure mechanism, proposed corrective measures, and provided a framework for preventing similar failures in the future. This is the kind of practical, experience-driven approach that should be applied to every dissimilar metal welding application.
Zhuojin Pipe Fitting Co., Ltd