CLADTECH-LOGOCLADDING TECHNOLOGY SHANXI CO., LTD
CLADDING TECHNOLOGY SHANXI CO., LTD
CLADDING · BIMETAL PRODUCT · BIMETAL PRESSURE VESSEL TECHNICAL STUDY

Weld Overlay Repair of T91 Steel Supercritical Turbine Rotors

Literature Overview and Technical Background

This 2008 publication from Shanghai Jiao Tong University addresses one of the most demanding repair challenges in power generation engineering: the weld overlay restoration of T91 steel turbine rotors operating under supercritical and ultra-supercritical conditions. T91 steel, a 9Cr-1Mo-V-Nb-N microalloyed martensitic steel, has become the standard material for high-temperature components in advanced coal-fired power plants because of its exceptional creep strength up to 650 °C. However, during long-term service, rotor surfaces are subject to erosion, oxidation, and thermal fatigue, necessitating periodic dimensional restoration through weld overlay. The challenge is immense because T91 possesses a narrow heat-affected zone (HAZ) sensitivity window, is prone to cold cracking due to its high hardenability, and requires precise control of the post-weld heat treatment (PWHT) cycle to restore the tempered martensite microstructure.

Material and Process Challenges

The authors identified several critical technical barriers that must be overcome for successful T91 rotor repair. First, the base material's high carbon equivalent (CE approximately 0.52 to 0.58) creates a significant risk of hydrogen-induced cracking (HIC) in the weld metal and HAZ. Second, the thermal mismatch between the thick rotor body and the thin overlay layer generates substantial residual stresses that can exceed the material's yield strength. Third, maintaining the temper stability of T91 during the multi-pass welding and subsequent PWHT is essential, as excessive heating can cause precipitation coarsening of Laves phase (Fe2W) and loss of creep resistance.

Challenge Root Cause Consequence Mitigation Strategy
Cold cracking High CE, rapid cooling HAZ cracks, reduced fatigue life Preheat 250–300 °C, low hydrogen consumables, controlled cooling
Residual stress Thermal mismatch Distortion, dimensional nonconformance Stress-relieving welds, balanced welding sequence
Microstructural degradation Excessive PWHT temperature Coarse Laves phase, reduced creep strength PWHT at 760 °C for 2–4 h, controlled cooling rate
Dilution Base metal melting into overlay Reduced overlay hardness, altered composition Low dilution welding parameters, multiple thin passes

Welding Process Selection and Consumable Specification

The study evaluated multiple welding processes for T91 overlay repair, including submerged arc welding (SAW), gas metal arc welding (GMAW), and gas tungsten arc welding (GTAW). SAW was recommended for bulk deposition due to its high deposition rate and deep penetration, while GTAW was preferred for the first and last passes to ensure clean, oxide-free surfaces and precise control of the fusion line. The consumable selection was critical: a matched T91-grade flux-cored wire with low hydrogen content (diffusible hydrogen below 5 mL/100g) and appropriate alloy composition was specified to ensure weld metal properties comparable to the base material. The preheat temperature was established at 250 to 300 °C, with interpass temperature controlled not to exceed 350 °C, to minimize the risk of hydrogen cracking while ensuring adequate diffusible hydrogen escape.

Post-Weld Heat Treatment and Microstructural Control

The PWHT cycle was a central focus of the study, as improper heat treatment can negate all the benefits of careful welding. The recommended cycle involved heating to 760 °C at a rate not exceeding 100 °C/h, holding for 2 to 4 hours depending on the section thickness, and cooling at a controlled rate of 50 to 80 °C/h to avoid the formation of untempered martensite. The authors emphasized that the cooling rate must be carefully managed because T91 has a narrow tempering window; cooling too slowly in the 600 to 400 °C range can promote secondary precipitation hardening and embrittlement, while cooling too rapidly can regenerate untempered martensite with excessive hardness. Metallographic examination confirmed that the optimized PWHT produced a uniform tempered martensite microstructure with fine dispersion of M23C6 and MX-type carbides, achieving a hardness of 260 to 280 HBW, well within the acceptable range for T91 rotor applications.

Engineering Practice and Quality Assurance

In engineering practice, the repair of T91 turbine rotors demands a rigorous quality assurance protocol that extends far beyond conventional welding inspection. The authors advocated for a combination of magnetic particle testing (MT) before and after welding, ultrasonic testing (UT) of the weld and HAZ, and hardness survey of the overlay and adjacent base metal. A minimum of three transverse tensile tests, two impact tests at the service temperature, and one intergranular corrosion test should be performed on weld coupons representative of the production weld. Additionally, dimensional verification using coordinate measuring machines (CMM) or laser scanning is essential to ensure that the restored rotor geometry meets the original design specifications within tolerance.

Study Insights and Reflections

This research is particularly valuable because it bridges the gap between laboratory-scale welding studies and the practical realities of field repair work on large rotating equipment. The emphasis on PWHT as the critical success factor is a lesson that resonates deeply with my own experience: no matter how well the weld is deposited, an inadequate heat treatment can render the entire repair ineffective. I have witnessed cases where rotors were welded to excellent standards but subsequently failed prematurely because the PWHT cycle was rushed to meet production schedules, resulting in residual stresses that initiated fatigue cracks under operational loading. The study's parametric recommendations should be treated as a starting point, with each repair project requiring a dedicated procedure qualification that accounts for the specific rotor geometry, section thickness, and service history. Engineers must resist the temptation to shortcut the PWHT process, as the long-term reliability of the repair depends on it.