Numerical Simulation of Submerged Arc Surfacing Considering Transformation-Induced Plasticity
Literature Overview and Practical Context
Huang Qingchun and colleagues from Liaoning University of Science and Technology and China Energy Engineering Group present a sophisticated numerical simulation of the submerged arc surfacing process for roll repair, incorporating transformation-induced plasticity (TRIP) effects. The work uses CALPHAD (Calculation of Phase Diagram) methods to compute temperature-dependent material properties of the surfacing material, and COMSOL multiphysics coupling to simulate the transient temperature field, martensite phase fraction, and TRIP stress evolution. The key finding is that the temperature field distribution and martensite transformation are both sensitive to the heat transfer rate, with the weld near the base metal side cooling faster and undergoing more rapid martensite transformation, resulting in a wider distribution of TRIP stress values. The residual stress after complete cooling is 376 MPa, with a martensite phase fraction of 94%, and the model was validated by SEM observation of the roll surfacing cross-section.
Core Technical Points and Simulation Approach
This paper represents a significant advancement in welding simulation because it couples thermal, metallurgical, and mechanical analyses with TRIP effects. The CALPHAD method is used to calculate the temperature-dependent properties of the surfacing material, which is a rigorous approach that avoids the common practice of assuming constant properties. The COMSOL platform is used to solve the coupled field equations, and the element birth-death technique is presumably used to activate the weld metal elements progressively. The key insight is that martensite transformation during cooling produces a volumetric expansion that counteracts the thermal contraction, leading to a reduction in tensile stress in the HAZ. This is the TRIP effect, and it is critical for predicting residual stress accurately.
| Simulation Component | Method Used | Key Output |
|---|---|---|
| Material Properties | CALPHAD calculation | Temperature-dependent thermal and mechanical properties |
| Temperature Field | COMSOL transient thermal analysis | Temperature distribution and cooling rates |
| Phase Transformation | Martensite kinetics model | Martensite phase fraction evolution |
| Stress Analysis | Coupled thermal-mechanical with TRIP | Residual stress distribution and TRIP stress |
| Model Validation | SEM observation of cross-section | Confirmation of predicted microstructure |
Engineering Practice Insights and Lessons Learned
The residual stress value of 376 MPa with 94% martensite is a critical finding for anyone involved in roll repair by submerged arc surfacing. Rolls are typically made from high-carbon alloy steels such as 70Cr3Mo or similar, and they are subjected to high contact stresses and thermal cycling during service. A residual stress of 376 MPa in the surfacing layer is significant, and if it is tensile, it can contribute to premature cracking and spalling. The paper's finding that the TRIP effect reduces the residual stress is important because it means that the actual residual stress is lower than what a conventional simulation without TRIP would predict.
From my own experience in roll repair, I can tell you that the most common failure mode is spalling of the surfacing layer, which is caused by a combination of residual stress and cyclic contact loading. The paper's recommendation to use the simulation method to predict and minimize residual stress is sound, but I would add that the practical countermeasures are still preheat, interpass temperature control, and post-weld heat treatment. In our shop, we always preheated rolls to 200–250°C before surfacing, maintained an interpass temperature of 150–200°C, and performed a stress-relieving treatment at 350–400°C for 4–6 hours after welding. The simulation results in this paper provide a quantitative basis for these practices.
One practical issue that I want to highlight is the sensitivity of the results to the heat transfer coefficient. The paper notes that the temperature field is sensitive to the heat transfer rate, and this is a critical parameter in any welding simulation. In practice, the heat transfer coefficient depends on the workpiece geometry, the shielding gas flow, and the ambient conditions, and it is often the source of significant error. I would recommend that any engineer using this simulation method calibrate the heat transfer coefficient against experimental measurements, such as thermocouple data from a test weld.
Summary
This paper presents a rigorous numerical simulation of submerged arc surfacing with TRIP effects, using CALPHAD for material properties and COMSOL for coupled field analysis. The finding that martensite transformation reduces residual stress through TRIP is a significant contribution, and the predicted residual stress of 376 MPa with 94% martensite provides a quantitative basis for process design. For engineers involved in roll repair and similar surfacing applications, this work offers a valuable tool for predicting residual stress and optimizing process parameters to prevent cracking and spalling.
Zhuojin Pipe Fitting Co., Ltd