Overlay Welding Repair of T91 Steel in Ultra-Supercritical Turbine Rotors

Literature Overview

This paper, published in 2008 in the journal Materials in Mechanical Engineering by researchers from Shanghai Jiao Tong University, addresses a critical maintenance challenge in power generation: the repair of abnormally worn surfaces on ultra-supercritical steam turbine rotors made of T91 steel. T91 is a 9Cr-1Mo-V-Nb martensitic precipitation-strengthened steel that has become the standard material for turbine components operating at steam temperatures above 600°C. The authors employed submerged arc overlay welding using 2CrMo electrode wire under optimized process parameters, and systematically investigated how the chemical composition of the overlay layer—particularly chromium content—influences microstructure and mechanical performance.

Core Technical Findings

The key result is that the overlay layer produced by submerged arc welding with 2CrMo wire exhibits comprehensive mechanical properties that surpass those of the T91 base metal. This is significant because T91 base metal, while offering excellent high-temperature strength and creep resistance, can suffer from localized wear under abnormal operating conditions such as blade vibration, foreign object ingestion, or bearing misalignment.

The most notable finding is that when the chromium content in the overlay layer is controlled to approximately 2%, the wear resistance improves dramatically. This observation is technically meaningful because chromium at this concentration promotes the formation of fine Cr carbides (primarily Cr7C3 and Cr23C6) that provide effective hardening without introducing excessive brittleness. The balance between hard phase volume fraction and matrix toughness is critical—too little chromium yields insufficient carbide precipitation, while excessive chromium can lead to coarse, brittle carbide networks that compromise fatigue resistance.

Process Analysis and Engineering Practice

Submerged arc welding was selected for this repair application for several practical reasons. The process provides deep penetration with low dilution rates, which is advantageous when building up material on a worn rotor surface. The flux cover ensures a controlled atmosphere, minimizing oxide inclusions that could initiate cracks under cyclic thermal loading. However, the selection of 2CrMo wire rather than a T91-matched consumable is a deliberate engineering decision. Matching T91 composition in a single-pass overlay would be extremely difficult due to the need for precise Nb and V additions, and the resulting overlay would likely require a post-weld heat treatment to restore the precipitation-strengthened microstructure—a constraint that is often impractical for field repairs.

The use of 2CrMo wire offers a pragmatic compromise: the overlay provides adequate hardness and wear resistance for the specific service condition (mechanical wear rather than high-temperature creep), while maintaining compatibility with the base metal's thermal expansion characteristics. The ~2% Cr threshold identified by the authors suggests that the dilution ratio between the 2CrMo wire and T91 base metal naturally produces an optimal composition window. This finding has direct implications for field repair procedures, where consumable selection and layer thickness control become critical process variables.

Key Reflections and Implications

The paper raises an important question about the long-term integrity of dissimilar overlay repairs on turbine rotors. While the overlay layer demonstrates superior room-temperature mechanical properties, the thermal cycling behavior during turbine start-up and shutdown is not addressed. The coefficient of thermal expansion mismatch between the overlay and base metal could generate interfacial stresses that, over thousands of thermal cycles, may lead to fatigue cracking at the fusion boundary. In engineering practice, any repair of this nature should be accompanied by a residual life assessment that accounts for both wear resistance improvement and potential fatigue degradation.

Furthermore, the study's emphasis on chromium content as the primary variable controlling wear performance suggests that future optimization could explore multi-element alloying strategies—incorporating Mo, Nb, or Ti—to further enhance both hardness and high-temperature stability. The work provides a solid foundation for developing standardized repair procedures for T91 turbine components, but the engineering community should approach its findings with awareness of the service-specific limitations inherent in any single-variable study.

Summary

This study demonstrates that submerged arc overlay welding with 2CrMo wire is a viable repair method for worn T91 turbine rotor surfaces, with the overlay layer achieving superior mechanical properties to the base metal. The critical finding that approximately 2% chromium content maximizes wear resistance provides a clear process control target for field applications. However, engineers must supplement this data with fatigue and thermal cycling assessments before deploying such repairs in service-critical ultra-supercritical turbine components. The work represents a practical bridge between laboratory metallurgy and field-level power plant maintenance, and its findings should inform the development of industry-standard repair specifications for next-generation turbine materials.