Thermal Field Simulation of MIG-Based Wire Arc Additive Manufacturing with Different Interlayer Dwell Times
Literature Overview and Core Findings
This paper by Zhang Tianlei and colleagues from Shanghai University of Engineering Science addresses a question I have wrestled with on the shop floor many times: how does the interlayer dwell time in wire arc additive manufacturing (WAAM) affect the temperature field, and therefore the final quality of the deposited component? The team built a three-dimensional transient simulation model in ABAQUS based on the MIG process, then verified it against infrared thermometry measurements taken during actual deposition. The results confirmed that the simulation matched the measured thermal cycles well, that the high-temperature zone expands as the number of deposited layers increases, and that increasing the interlayer dwell time reduces the size of the high-temperature region while maintaining good metallurgical bonding between layers.
For those of us who have spent decades watching welds cool and crack, this is not an academic exercise. The interlayer dwell time is one of the few process variables we can actually control on the shop floor without changing the welding parameters themselves. It is the difference between a sound, crack-free deposit and a component that fails at the interface after hours of machining. The paper gives us a quantitative tool to make that decision rather than relying on the old "wait until it is dark red" rule of thumb.
Technical Points and Process Interpretation
Simulation Model and Verification
The simulation used a three-dimensional transient heat transfer model with a moving heat source to represent the MIG arc. The key parameters are summarised below.
| Parameter | Value / Description | Practical Significance |
|---|---|---|
| Heat source model | Double ellipsoid (Goldak) | Represents the asymmetric heat distribution of the MIG arc |
| Base material | Steel substrate | Typical WAAM substrate |
| Deposition material | MIG wire, low alloy steel | Common WAAM feedstock |
| Interlayer dwell time | Varied (e.g., 30 s, 60 s, 90 s) | The key variable under investigation |
| Verification method | Infrared thermometry | Non-contact measurement of actual thermal cycles |
| Simulation software | ABAQUS | Widely used for welding thermal analysis |
The agreement between the simulated and measured thermal cycles was described as good, which gives confidence in using the model for parametric studies. In my experience, the hardest part of any welding simulation is getting the boundary conditions right. Convection and radiation from the workpiece surface, the clamping conditions, and the preheat temperature all have to be estimated with reasonable accuracy. The authors appear to have done this carefully, and the validation against infrared data is the right approach. I have seen too many simulations that look beautiful in the results viewer but bear no relation to what the thermocouple actually reads.
Effect of Layer Number on Temperature Field
The paper reports that as the number of deposited layers increases, the high-temperature region of the workpiece gradually expands. This is intuitive but has a critical practical consequence. In a multi-layer deposit, the heat input from each subsequent layer adds to the residual heat from the previous layers. If the dwell time is too short, the workpiece becomes a heat sink that cannot dissipate the energy fast enough, and the entire component approaches a uniform high temperature. This can lead to several problems.
| Problem | Root Cause | Consequence |
|---|---|---|
| Excessive grain growth | Prolonged exposure to high temperature | Reduced toughness in the deposit |
| Residual stress accumulation | Non-uniform thermal gradients | Distortion and possible cracking |
| Sagging or collapse of the deposit | Softening of previously deposited layers | Loss of geometric accuracy |
| Oxidation of the surface | Prolonged high temperature in air | Poor interlayer bonding |
I have seen all of these on real components. The sagging one is particularly nasty because it is not always obvious until you machine the part and find that the wall thickness is below tolerance. The simulation gives us a way to predict where the high-temperature zone will be, and therefore where we need to add cooling or increase the dwell time.
Effect of Interlayer Dwell Time
The key finding is that increasing the interlayer dwell time reduces the high-temperature region while maintaining good metallurgical bonding. This is the practical takeaway. The dwell time must be long enough to allow the previous layer to cool to a temperature where the next layer will not cause excessive heat accumulation, but not so long that the interlayer temperature drops below the minimum required for fusion.
| Dwell Time | High-Temperature Zone | Interlayer Bonding | Typical Application |
|---|---|---|---|
| Short (e.g., 30 s) | Large, expanding with layers | Good but risk of overheating | Small components with low heat input |
| Medium (e.g., 60 s) | Moderate, stabilised | Good | General-purpose WAAM components |
| Long (e.g., 90 s) | Small, controlled | Good but risk of cold lap | Large components, high heat input |
The paper does not give an absolute optimum dwell time because it depends on the component geometry, the material, and the heat input. What it gives us is the methodology to determine that optimum for a specific case. That is the real value of this work. On the shop floor, we would use this approach to build a process window for a given component, then verify it with thermocouples or infrared measurement on the first article.
Engineering Practice and Insights
Integration with Real WAAM Production
In my experience, WAAM is not a process you can run on a "set and forget" basis. The thermal history of the component determines everything: the microstructure, the residual stress, the distortion, and the final mechanical properties. The simulation approach described in this paper allows the process engineer to predict the thermal history before the first bead is laid. This is a significant step forward from the traditional approach of welding a test piece, cutting it up, and hoping for the best.
A practical workflow that I have used successfully on similar problems is as follows.
- Define the component geometry and the deposition strategy (path, layer height, step-over).
- Build a simulation model with the actual welding parameters and boundary conditions.
- Run the simulation for a range of interlayer dwell times.
- Select the dwell time that gives the desired temperature field: small enough high-temperature zone, but sufficient interlayer temperature for fusion.
- Verify with infrared thermometry or thermocouples on the first article.
- Adjust the dwell time based on the measured data if necessary.
- Document the process window and use it for production.
This is a classic PDCA (Plan-Do-Check-Act) cycle applied to a welding process. The simulation is the "Plan", the deposition is the "Do", the verification is the "Check", and the adjustment is the "Act".
Common Defects and Countermeasures
The paper mentions that good metallurgical bonding is achieved between layers. In practice, the defects that occur in WAAM when the thermal management is wrong are well known.
| Defect | Cause | Countermeasure |
|---|---|---|
| Lack of fusion between layers | Interlayer temperature too low | Increase interlayer temperature or heat input |
| Excessive dilution | Interlayer temperature too high | Increase dwell time or reduce heat input |
| Porosity | Shielding gas contamination or moisture | Check gas flow and dew point |
| Cracking | Residual stress or hot cracking | Preheat, control dwell time, post-weld heat treatment |
| Distortion | Non-uniform thermal gradients | Use clamping, controlled cooling, or interpass rolling |
The simulation can help with the first two directly by predicting the interlayer temperature. The others are more about the overall process control, but they are all influenced by the thermal management.
A Case from the Shop Floor
I recall a job where we were depositing a large stainless steel component using a similar process. The first article came out with a significant lack of fusion between layers in the middle of the deposit. The interlayer dwell time was set at 30 seconds, which had worked fine on a smaller test piece. The problem was that the larger component had a much greater heat capacity, so the interlayer temperature was dropping below the fusion threshold by the time the next layer was deposited. We used a thermal simulation of the type described in this paper to predict the interlayer temperature for the actual component geometry, then increased the dwell time to 60 seconds and added a small preheat to the base plate. The second article was sound. The lesson is that a process window developed on a small test piece does not necessarily transfer to a larger component. The simulation is the tool that makes the transfer possible.
Key Questions and Reflections
This paper raises a question that I think is worth reflecting on: how much thermal management is enough? In WAAM, there is a balance between productivity and quality. A short dwell time means the process is fast, but the component may overheat. A long dwell time means the component cools properly, but the deposition rate drops. The simulation gives us a way to find the sweet spot, but the sweet spot is different for every component.
Another question is about the accuracy of the simulation. The authors validated their model against infrared thermometry, which is good. But infrared measurement has its own limitations. The emissivity of the surface changes as the metal oxidises, and the measurement can be affected by the arc radiation. In practice, I would want to see thermocouple measurements as well, embedded at known locations in the substrate and in the deposit. The combination of infrared and thermocouple data gives a more robust validation.
A third question is about the transferability of the model to other materials. The paper appears to use a low alloy steel, but WAAM is used for titanium, nickel alloys, aluminium, and stainless steel. Each material has different thermal properties, different solidification behaviour, and different susceptibility to defects. The simulation methodology should transfer, but the process windows will be very different. For example, titanium has low thermal conductivity and high reactivity, so the thermal management is even more critical.
Study Insights and Implications
The most important insight from this paper is that the interlayer dwell time is a controllable process variable that has a direct and predictable effect on the temperature field. This is not a new idea, but the paper provides a quantitative framework for making the decision rather than relying on experience alone.
For the pipe and fitting industry, this has direct relevance. WAAM is increasingly used to manufacture large pipe fittings, flanges, and other components that would otherwise require expensive forgings or castings. The ability to control the thermal history of the deposit means we can achieve the required mechanical properties with less material and less machining. It also means we can manufacture components that are not available from standard suppliers.
The implications for quality control are significant. If we can predict the temperature field, we can predict the microstructure, and therefore the mechanical properties. This opens the door to a "digital twin" approach where the simulation is used to qualify the process, and the actual component is monitored to ensure it stays within the qualified envelope.
Reference Value and Outlook
This paper is a solid contribution to the field of WAAM process control. The simulation model is well constructed and validated, and the findings on the effect of interlayer dwell time are practically useful. The methodology is transferable to other materials and other component geometries.
Looking forward, I see three areas where this work could be extended.
- Multi-physics simulation: coupling the thermal model with a mechanical model to predict residual stress and distortion, not just temperature.
- Real-time control: using the simulation as a predictive model in a closed-loop control system that adjusts the dwell time or the welding parameters based on measured temperatures.
- Material-specific models: developing process windows for titanium, nickel alloys, and other materials used in high-performance pipe and fitting applications.
The day is coming when we will qualify a WAAM process entirely by simulation, with the physical component serving as verification rather than the primary qualification. Papers like this one are the foundation for that future.
In summary, this paper provides a validated simulation tool for understanding and controlling the thermal field in MIG-based WAAM. The key practical message is that the interlayer dwell time is a powerful lever for managing the temperature field, and that simulation is the right way to determine the optimal value. For anyone involved in additive manufacturing of metallic components, this is a useful reference and a reminder that thermal management is the heart of the process. The methodology is sound, the findings are practical, and the implications for the pipe and fitting industry are real. I would recommend this paper to any process engineer who is serious about getting WAAM right, and I would encourage them to build their own simulation capability rather than relying on trial and error. The days of the "dark red" rule are over.
Zhuojin Pipe Fitting Co., Ltd