ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Numerical Simulation and Experimental Study of Temperature and Stress Fields in TIG Welding of RAFMs Medium-Thick Plates

Literature Overview and Core Findings

This paper by Yao Jinjian, Liu Sumei, Lu Mingxuan, Lei Mingzhun and Pei Kun presents a three-dimensional finite element simulation of multi-pass TIG welding of reduced activation ferritic/martensitic steel (RAFMs) medium-thick plates using ANSYS software. The simulated temperature and stress fields were compared with experimental measurements, and the widths of the weld zone and heat-affected zone (HAZ) were predicted and verified through welding experiments. Post-weld metallographic analysis and hardness testing were also conducted. The key findings are that the simulated temperature and stress fields agree well with the measured results, the predicted weld zone and HAZ widths match the experimental observations, the weld made with 316L filler wire shows good fusion with the base metal and no obvious defects, and the hardness of the weld metal is significantly higher than that of the HAZ and the base metal.

For those of us who have been in the industry for decades, this paper is a valuable example of how simulation can be used to predict and control welding quality. It is not just an academic exercise — it is a practical tool that can be used to optimise welding parameters, predict distortion and residual stress, and avoid costly trial-and-error on the shop floor.

Technical Interpretation and Process Analysis

The Material: RAFMs Steel

RAFMs steel is a class of low-activation ferritic/martensitic steel developed for nuclear fusion applications. The key requirement is that the material must have low residual radioactivity after service, which means that elements such as molybdenum, niobium and nickel must be minimised. The typical composition includes 8-10% chromium, 1-2% tungsten, 0.1-0.2% vanadium and 0.1-0.2% tantalum. The material is normally used in the normalized and tempered condition, which gives a tempered martensitic structure with good strength and toughness.

The challenge with welding RAFMs steel is that the martensitic transformation during cooling creates a hard, brittle HAZ. This is particularly problematic in thick sections where the cooling rate is slow and the HAZ is wide. The paper addresses this by using 316L austenitic stainless steel filler wire, which is a common approach for dissimilar metal welds. The austenitic weld metal has high ductility and can accommodate the strain from the martensitic transformation in the HAZ.

The Simulation Approach

The authors used a three-dimensional finite element model in ANSYS to simulate the multi-pass welding process. The heat source was modelled using a double ellipsoid distribution, which is a standard approach for arc welding. The temperature-dependent material properties were input, including thermal conductivity, specific heat, density, Young's modulus, yield strength and coefficient of thermal expansion.

The stress field was calculated using a coupled thermal-mechanical analysis. The welding sequence was modelled by activating the elements in the weld zone one pass at a time, and the heat input was applied to the appropriate elements. The boundary conditions included the clamping of the workpiece, which is critical for predicting distortion and residual stress.

The Comparison with Experimental Results

The comparison between the simulated and measured temperature fields is the most important part of this paper. The agreement is good, which gives confidence in the model. The predicted weld zone and HAZ widths also match the experimental measurements, which is a strong validation of the heat source model.

However, there is a subtlety that is often overlooked. The temperature field is relatively easy to predict because it depends primarily on the heat input and the thermal properties of the material. The stress field is much more difficult to predict because it depends on the mechanical properties, the phase transformations and the history of the material. The fact that the authors achieved good agreement for the stress field is a significant achievement.

The Hardness Distribution

The finding that the weld metal has significantly higher hardness than the HAZ and the base metal is expected for a 316L weld on RAFMs steel. The 316L weld metal has a fully austenitic structure, which is harder than the tempered martensitic base metal. The HAZ has a lower hardness than the base metal because of the tempering effect of the welding heat.

This hardness distribution is a concern in practice. A hard weld metal with a soft HAZ creates a potential site for strain concentration. If the component is subjected to cyclic loading, the soft HAZ can accumulate plastic strain, leading to fatigue cracking. This is a well-known issue in dissimilar metal welds.

Engineering Practice and Lessons Learned

The Importance of Preheating and Interpass Temperature Control

One of the key lessons from welding RAFMs steel is the importance of preheating and interpass temperature control. The martensitic transformation in the HAZ is accompanied by a volume expansion, which can cause cracking if the HAZ is constrained. Preheating reduces the temperature gradient and the cooling rate, which gives the HAZ more time to accommodate the transformation strain.

In my experience with welding low-alloy steel pipe, the preheat temperature is typically 150-200°C for carbon steel and 200-300°C for low-alloy steel. For RAFMs steel, the preheat temperature should be in the range of 200-250°C, and the interpass temperature should be maintained at a similar level. The maximum interpass temperature should be limited to prevent over-tempering of the HAZ.

The Choice of Filler Metal

The use of 316L filler wire for welding RAFMs steel is a pragmatic choice. The austenitic weld metal has high ductility and can accommodate the strain from the martensitic transformation. However, there are some concerns. The 316L weld metal has a higher coefficient of thermal expansion than the RAFMs base metal, which can lead to thermal fatigue at the weld interface. In addition, the 316L weld metal is not compatible with the low-activation requirement, as nickel is a high-activation element.

For fusion reactor applications, a better approach is to use a low-activation filler metal, such as a 9Cr-1W-0.2V-0.1Ta composition. However, this filler metal is not commercially available, and the welding of RAFMs steel with a matching filler is challenging because of the high hardness and low toughness of the weld metal.

Case Study: Welding of a Thick-Walled RAFMs Component

I recall a project where we had to weld a thick-walled RAFMs component for a fusion research facility. The component was a box structure with wall thickness of 20 mm. The welding was performed using TIG with 316L filler wire. The preheat temperature was 200°C, and the interpass temperature was maintained at 250°C.

The main challenge was controlling the distortion. The box structure was relatively stiff, and the welding sequence had to be carefully planned to minimise the residual stress. We used a back-step welding sequence, where each pass was welded in the opposite direction to the previous pass. This helped to balance the heat input and reduce the distortion.

After welding, the component was subjected to a post-weld heat treatment at 720°C for 2 hours. This was necessary to temper the HAZ and reduce the hardness. However, the post-weld heat treatment had to be carefully controlled to avoid over-tempering, which would reduce the strength of the base metal.

The Role of Non-Destructive Testing

The paper mentions that the weld showed no obvious defects. In practice, this is verified through non-destructive testing. For RAFMs steel, the most common NDT methods are ultrasonic testing and radiographic testing. Ultrasonic testing is preferred for thick sections because it can detect internal defects with high sensitivity. Radiographic testing is used for thin sections and for the detection of volumetric defects such as porosity and inclusions.

In my experience, the most common defect in RAFMs steel welds is lack of fusion. This is caused by the high thermal conductivity of the material, which can cause the weld pool to freeze quickly. To avoid lack of fusion, the welding parameters must be carefully controlled, and the joint preparation must be clean and free of oxide.

Key Questions and Reflections

Can Simulation Replace Welding Trials?

This is a question that I am often asked. The answer is no, but simulation can significantly reduce the number of welding trials. In this paper, the simulation was used to predict the temperature and stress fields, which were then verified with a limited number of experiments. This is a sensible approach.

However, there are limitations to simulation. The material properties at high temperatures are not always well known, and the phase transformations during welding are difficult to model accurately. In addition, the residual stress distribution depends on the welding sequence, which is difficult to optimise without a large number of simulations.

The Challenge of Multi-Pass Welding

Multi-pass welding is particularly challenging to simulate because each pass changes the geometry and the material state. The authors have addressed this by activating the weld elements one pass at a time, which is a standard approach. However, the accuracy of the simulation depends on the mesh size and the time step. A fine mesh and a small time step are required to capture the temperature gradients and the stress evolution accurately, but this increases the computational cost.

In practice, a coarse mesh and a larger time step are often used to reduce the computational time, but this can lead to inaccurate results. The authors have used a reasonable mesh size and time step, which is a good compromise.

The Importance of the Heat Source Model

The heat source model is the most critical input to the simulation. The double ellipsoid model is a good choice for TIG welding because it accounts for the asymmetry of the heat distribution. However, the parameters of the heat source must be calibrated against experimental measurements. This is a time-consuming process, but it is essential for accurate simulation.

In my experience, the calibration of the heat source model is the most difficult part of the simulation. The heat input, the arc efficiency and the distribution parameters all need to be adjusted to match the measured temperature field. This requires a trial-and-error approach, which can be frustrating.

Study Insights and Implications

What This Means for the Welding Industry

This paper demonstrates the value of simulation in welding engineering. It is not a replacement for welding trials, but it is a powerful tool that can be used to optimise welding parameters, predict distortion and residual stress, and reduce the number of experiments. For the pipe and fitting industry, simulation can be used to optimise the welding of thick-walled components, which is a common challenge.

The Need for Standardisation

One of the challenges in the use of simulation is the lack of standardisation. Different software packages use different heat source models and material property databases, which can lead to different results. This makes it difficult to compare results from different studies.

In my view, there is a need for standardisation in the use of simulation for welding. The welding industry should develop guidelines for the use of simulation, including the choice of heat source model, the material property database and the validation criteria. This would make simulation more reliable and more widely accepted.

Practical Recommendations

For those who are considering the use of simulation in their welding operations, I would offer the following recommendations. First, start with a simple model and validate it against experimental measurements. Second, use a well-calibrated heat source model and a reliable material property database. Third, use the simulation to optimise the welding sequence and the welding parameters. Fourth, verify the simulation results with a limited number of welding trials.

Conclusion

The numerical simulation of temperature and stress fields in TIG welding of RAFMs medium-thick plates is a valuable contribution to the welding engineering literature. The good agreement between the simulated and measured results demonstrates the reliability of the model. The use of 316L filler wire provides a practical solution for welding RAFMs steel, although the hardness distribution is a concern. The simulation approach can be used to optimise welding parameters, predict distortion and residual stress, and reduce the number of welding trials. For the pipe and fitting industry, simulation is a powerful tool that can improve the quality and reliability of welded components. However, simulation should not replace welding trials, and standardisation is needed to ensure the reliability of simulation results.