Failure Analysis of TIG Welded Joints between Martensitic and Austenitic Stainless Steels

Literature Overview

This paper, published in Transactions of the China Welding Institute (2007, Vol. 28, No. 5, pp. 89-92) by Ling Wei, Ge Liaohai, Ren Zhen'an, and Sun Daqian, investigates a real-world failure case involving a TIG-welded joint between 1Cr17Ni2 martensitic stainless steel and 1Cr18Ni9Ti austenitic stainless steel used in marine equipment. The joint experienced leakage, prompting a systematic failure analysis using optical microscopy, SEM, and microhardness measurement. This case study is directly relevant to pipeline engineers working with dissimilar stainless steel weldments, particularly in offshore and subsea applications where mixed metallurgical grades are commonly encountered.

Core Technical Findings

Microstructural Analysis of the Weld Joint

The dissimilar joint exhibits markedly different microstructural evolution on each side of the weld, driven by the fundamentally different phase transformation behaviors of martensitic and austenitic stainless steels:

The critical weakness zone is identified in the 1Cr17Ni2 HAZ at the higher temperature region, where rapid cooling from the welding thermal cycle produces high-hardness martensite. This region exhibits a typical cleavage and quasi-cleavage brittle fracture morphology on the fracture surface, making it the primary site for crack initiation and propagation.

Root Cause Analysis

The failure mechanism can be traced through the following causal chain:

1. Thermal cycle asymmetry: The martensitic side undergoes a phase transformation (austenite → martensite) during cooling, while the austenitic side does not, creating a severe microstructural and mechanical property mismatch across the joint
2. Martensite formation in HAZ: The rapid cooling rate in the 1Cr17Ni2 HAZ produces untempered martensite with high hardness and extremely low fracture toughness
3. Residual stress concentration: The differential thermal expansion between the two base metals generates significant residual stresses at the weld interface
4. Crack initiation and propagation: Under service loading (pressure, thermal cycling, vibration), cracks initiate at the brittle martensite zone and propagate through the HAZ, eventually causing leakage

Contributing Factors

Beyond the primary metallurgical mechanism, several secondary factors contributed to the failure:

• Welding procedure inadequacy: The WPS likely did not include post-weld heat treatment (PWHT) to temper the martensite formed in the HAZ
• Joint design: The butt joint configuration concentrates stress at the interface where the property mismatch is most severe
• Service environment: Marine conditions introduce chloride ions that can exacerbate cracking through stress corrosion cracking (SCC) mechanisms

Standards and Code Compliance

For subsea and offshore pipeline applications governed by DNV-ST-F101 or ASME B31.4, dissimilar stainless steel welds require rigorous qualification testing including tensile, bend, and fracture toughness tests, along with metallographic examination of the HAZ on both sides of the joint.

Engineering Practice Recommendations

Improved Welding Procedure

To prevent recurrence of this failure mode, the following improvements should be implemented:

• Preheat and PWHT: Apply preheat at 200-300°C to the martensitic side and perform post-weld heat treatment at 600-700°C for 1-2 hours to temper martensite and reduce hardness
• Filler metal selection: Use a nickel-based filler (e.g., ERNiCrMo-3 or equivalent) to dilute the weld metal composition and reduce carbon pickup in the weld metal, thereby limiting martensite formation
• Heat input control: Moderate heat input (1.5-2.5 kJ/mm) to balance fusion penetration with HAZ grain coarsening
• Multi-pass welding: Employ multi-pass welding with interpass temperature control (150-200°C) to temper each pass's HAZ martensite during subsequent passes

Inspection Protocol

For production welds of this type, the following inspection regime is recommended:

• 100% visual inspection (VT): Surface examination of both sides
• 100% magnetic particle testing (MT): Applicable to the martensitic side for surface and near-surface defects
• 100% dye penetrant testing (PT): Applicable to the austenitic side
• Radiographic testing (RT) or ultrasonic testing (UT): For volumetric defect detection
• Hardness survey: Traverse the HAZ on both sides to verify hardness profiles are within acceptable limits
• Metallographic examination: On coupon welds, verify HAZ microstructure and absence of excessive martensite

Study Insights and Implications

This failure case serves as a powerful reminder that dissimilar stainless steel welds require fundamentally different metallurgical considerations than similar-material welds. The martensitic side of the joint is inherently more vulnerable because the welding thermal cycle can induce uncontrolled phase transformations that degrade fracture toughness. In pipeline engineering, where joints must withstand decades of cyclic pressure loading and thermal cycling, the absence of PWHT on martensitic stainless steel welds represents an unacceptable risk. Every engineer involved in specifying, welding, or inspecting dissimilar stainless steel joints should internalize the lesson from this case: the HAZ microstructure, not the weld metal, is typically the governing factor in joint reliability. Future qualification programs should mandate fracture toughness testing (Charpy V-notch or CTOD) on the martensitic HAZ specifically, as conventional tensile and bend tests may not detect the reduced fracture resistance caused by untempered martensite.