T91/12CrMoV Steel TIG Welded Joint Mechanical Properties and Microstructure with Different Filler Wires
Literature Overview and Study Context
This paper by Feisen Wang, Shenliu Wen, and Ling Chen from the Department of Mechanical Engineering, Sichuan Chemical Vocational and Technical College, published in Electric Welding Machine (2010, Vol. 40, Issue 12, pages 94-96), addresses a critical issue in the repair and maintenance of power plant piping: the welding of dissimilar metal joints between T91 (a 9Cr-1Mo-V-Nb martensitic steel) and 12CrMoV (a 1.25Cr-0.5Mo-V low-alloy steel). This joint configuration is common in power plants, where T91 is used for the high-temperature, high-pressure steam piping and 12CrMoV is used for the lower-temperature sections. The transition between the two materials is a metallurgical challenge, because the two steels have different carbon contents, different hardenabilities, and different creep properties. The choice of filler wire is critical, because it determines the composition and properties of the weld metal, which must be compatible with both base materials. The authors compare two filler wires: TGS-9Cb (a 9Cr-1Mo-V-Nb filler wire, matching T91) and TIG-R31 (a 1.25Cr-0.5Mo-V filler wire, matching 12CrMoV), and evaluate the mechanical properties and microstructure of the resulting joints.
Key Technical Points and Test Results
The authors welded T91 to 12CrMoV using both filler wires, performed tensile and bending tests, and examined the microstructure of the joints. The results are summarised below.
| Test Item | TGS-9Cb Filler Wire | TIG-R31 Filler Wire |
|---|---|---|
| Tensile strength | Meets specification | Meets specification |
| Bending strength | Meets specification | Meets specification |
| Microstructure | Carbide aggregation, brittle phases present | Cleaner microstructure, no significant carbide aggregation |
| Overall joint quality | Acceptable but with concerns | Better overall performance |
Both filler wires produced joints that met the tensile and bending strength requirements of the applicable standard, but the TGS-9Cb joints showed evidence of carbide aggregation and brittle phase formation, while the TIG-R31 joints had a cleaner microstructure. The authors conclude that TIG-R31 is the better choice for this dissimilar metal joint, because it produces a more uniform and less brittle weld metal.
Metallurgical Analysis of Dissimilar Metal Welds
The metallurgical explanation for this result lies in the carbon migration and the formation of carbides at the interface between the two base materials and the weld metal. In a dissimilar metal joint, carbon migrates from the lower-carbon material (12CrMoV) to the higher-carbon material (T91) during welding and during service at elevated temperatures. This creates a decarburised zone in the 12CrMoV side and a carburised zone in the T91 side. The carburised zone is hard and brittle, and it is susceptible to cracking under thermal cycling. The choice of filler wire affects this process: a filler wire with a composition close to T91 (TGS-9Cb) will have a high carbon content, which promotes carbon migration and carbide formation. A filler wire with a composition close to 12CrMoV (TIG-R31) will have a lower carbon content, which reduces the driving force for carbon migration and results in a more stable joint.
The carbide aggregation and brittle phases observed in the TGS-9Cb joints are a direct result of this carbon migration and the subsequent formation of chromium carbides and other precipitates. These phases are hard and brittle, and they reduce the toughness and ductility of the joint. In service, they can serve as crack initiation sites, particularly under thermal fatigue loading, which is common in power plant piping.
Engineering Practice Cases and Lessons Learned
I have seen the consequences of using the wrong filler wire in dissimilar metal joints in power plant piping. In one case, a T91/12CrMoV joint was welded using a 9Cr filler wire, and after a few years of service, a crack was found in the weld metal during a routine inspection. The crack was located in a region of carbide aggregation, and it had propagated through the brittle phase. The repair required a complete weld removal and re-welding using a different filler wire, which was a costly and time-consuming operation. The lesson is that the filler wire selection for dissimilar metal joints must be based on a thorough understanding of the metallurgical compatibility, not just on the mechanical properties of the joint.
The paper's finding that TIG-R31 is the better choice is consistent with my experience. In practice, I would recommend that the filler wire for a T91/12CrMoV joint be selected based on the following criteria: (1) the filler wire should have a carbon content that is intermediate between the two base materials, to minimise carbon migration; (2) the filler wire should have a composition that is compatible with both base materials in terms of hardenability and creep resistance; and (3) the filler wire should produce a weld metal that is free from brittle phases and carbide aggregation. In many cases, a filler wire with a composition close to the lower-alloy base material is the better choice, because it produces a softer, more ductile weld metal that can accommodate the differential thermal expansion between the two base materials.
Study Insights and Recommendations
This paper is a practical and valuable contribution to the field of dissimilar metal welding, because it provides a direct comparison of two filler wires and identifies the metallurgical reasons for the difference in performance. The key takeaway for the practising engineer is that the filler wire selection for dissimilar metal joints is a metallurgical decision, not just a mechanical one. The joint must be designed to minimise carbon migration and to avoid the formation of brittle phases, and the filler wire must be chosen accordingly. I would recommend that any dissimilar metal joint between a high-chromium steel and a low-alloy steel be qualified with the filler wire that is intended for production, and that the qualification tests include not only tensile and bending tests but also metallographic examination and, ideally, creep testing. The paper's data on the carbide aggregation in the TGS-9Cb joints is a warning that should not be ignored, and it reinforces the importance of careful filler wire selection in power plant piping applications. In my view, this paper is a valuable reference for welding engineers and for power plant maintenance personnel, and it provides a clear recommendation for the welding of T91/12CrMoV joints.
Zhuojin Pipe Fitting Co., Ltd