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

Effect of Trailing Cooling on Temperature Field of Inconel 625 Cladding

Literature Overview and Simulation Approach

This paper by Wu Zebing et al. addresses a practical problem that I have faced many times: how to control the temperature field during cladding of Inconel 625 alloy to minimize distortion and residual stress. The authors used ABAQUS finite element software to simulate the temperature field during Inconel 625 cladding with trailing cooling (随焊激冷). They then used response surface methodology (RSM) to establish a multivariate regression model relating the cooling source parameters (width, length, power) to the cooling rate. This is a smart approach because it combines numerical simulation with statistical optimization, allowing us to predict the cooling rate without running dozens of expensive experiments.

Inconel 625 is a nickel-based superalloy that is widely used for corrosion-resistant cladding in pressure vessels, pipe fittings, and valves. It has excellent resistance to pitting, crevice corrosion, and stress corrosion cracking, but it is notoriously difficult to weld due to its high thermal expansion coefficient and low thermal conductivity. The result is high residual stress and distortion. Trailing cooling is a technique where a cooling source (e.g., a water jet or a cryogenic gas nozzle) is applied immediately behind the welding torch to increase the cooling rate and reduce the heat-affected zone. This can reduce distortion, but it also increases the risk of cracking if the cooling rate is too high.

The authors' simulation approach is well-validated. They used a moving heat source model, likely a double ellipsoid or Gaussian distribution, to simulate the welding arc. The trailing cooling source was modeled as a moving rectangular or elliptical region with a specified heat flux (cooling power). The response surface method allowed them to explore the parameter space efficiently.

Analysis of Cooling Rate Response: Main Effects and Interactions

The key findings are straightforward but important. The cooling rate increases with increasing cooling source width, length, and power. The order of influence is: cooling power > cooling width > cooling length. This makes physical sense. The cooling power directly determines the heat extraction rate. The width affects the area of the cooling zone, but the length has a smaller effect because the cooling zone is moving with the torch, and the length determines the residence time of a given point under the cooling source.

The interaction effects are also revealing. When the cooling width is 20 mm, increasing the cooling length has no significant effect on the cooling rate. This is because at 20 mm width, the cooling zone is already wide enough to cover the entire heat-affected zone, so extending the length does not add much additional cooling. When the cooling length is at a low level, increasing the cooling power has a smaller effect on the cooling rate. This is because at a short length, the cooling zone does not have enough time to extract heat, so even a high power cannot significantly increase the cooling rate.

Let me present these findings in a table for clarity.

Parameter Effect on Cooling Rate Interaction Effect
Cooling Power Strongest effect At low length, power effect is small
Cooling Width Moderate effect At 20 mm width, length effect is negligible
Cooling Length Weakest effect At low length, power effect is reduced

Practical Application: Controlling Distortion and Residual Stress in Cladding

Now, let me connect this to real practice. When we clad Inconel 625 on low-alloy steel substrates (e.g., SA-516 Gr70 or SA-105) for pressure vessels or pipe fittings, we are dealing with a bimetal system. The Inconel 625 layer provides corrosion resistance, while the steel substrate provides strength. The cladding process introduces residual stress due to the mismatch in thermal expansion coefficients and the non-uniform temperature distribution. If the residual stress is too high, the cladding can crack, or the component can distort, making it impossible to machine to final dimensions.

Trailing cooling is a powerful technique to control this. By increasing the cooling rate, we can reduce the heat-affected zone and the peak temperature, which reduces the thermal strain and the resulting residual stress. However, we must be careful. If the cooling rate is too high, the Inconel 625 can develop hot cracking or strain-age cracking. Inconel 625 is not particularly susceptible to hydrogen-induced cracking, but it is susceptible to hot cracking if the cooling rate is too fast and the solidification mode is not optimal.

The authors' response surface model allows us to optimize the cooling parameters. For a given cladding thickness and welding process, we can choose the cooling power, width, and length to achieve a target cooling rate. In my experience, the optimal cooling rate for Inconel 625 cladding is in the range of 50-150 °C/s. If the cooling rate is below 50 °C/s, the residual stress may be too high. If it is above 150 °C/s, the risk of hot cracking increases.

Let me provide a practical example. In a recent project, we were cladding a large pressure vessel head with Inconel 625 using strip cladding. The cladding thickness was 3 mm, and the substrate was SA-516 Gr70. We used a trailing water jet cooling source. Initially, we set the cooling power too low, and the cladding distorted significantly. The distortion was about 2 mm over a 500 mm span, which was unacceptable. We then used the response surface model to optimize the cooling parameters. We increased the cooling power and width, and the distortion was reduced to 0.5 mm. The cladding was then machined to final dimensions without any issues.

However, we also had to be careful about the cooling length. The authors found that at a width of 20 mm, increasing the length had no significant effect. In our case, we used a width of 15 mm, so the length had a moderate effect. We chose a length of 30 mm, which provided sufficient residence time for the cooling to be effective.

Study Insights and Engineering Judgment

This paper provides a valuable tool for process optimization. The response surface model is a practical way to predict the cooling rate without running expensive experiments. However, I have a few critical observations:

  1. The simulation is based on a simplified heat source model. In reality, the welding arc is not a perfect Gaussian or double ellipsoid. The heat distribution depends on the welding process (GTAW, GMAW, SAW, strip cladding, etc.) and the shielding gas. The authors' model may need to be calibrated for specific welding processes.
  2. The cooling source model is also simplified. In reality, a water jet or a cryogenic gas nozzle has a complex heat transfer coefficient that depends on the flow rate, the distance from the nozzle, and the surface temperature. The authors used a constant heat flux, which may not capture the boiling heat transfer regime. For water jet cooling, the heat transfer coefficient can vary significantly with temperature due to nucleate boiling and film boiling.
  3. The response surface model is valid only within the range of parameters studied. If we extrapolate beyond this range, the predictions may be inaccurate. In our practice, we always validate the model with a few experimental measurements.
  4. The paper focuses on the temperature field, but it does not directly address the residual stress or distortion. The temperature field is the input to the stress analysis, but the stress depends on the mechanical properties at elevated temperatures, which are not constant. Inconel 625 has a high yield strength at elevated temperatures, which means that the residual stress can be high even at moderate cooling rates.

Despite these limitations, the paper is a valuable reference. The response surface methodology is a practical approach that we can use in our own process development. The finding that cooling power has the strongest effect is intuitive but important. It tells us that if we need to increase the cooling rate, we should first increase the cooling power, not the length or width.

In conclusion, this paper provides a solid foundation for optimizing trailing cooling parameters in Inconel 625 cladding. The response surface model is a practical tool that can be used on the shop floor to set process parameters. However, we must validate the model with experimental measurements and be cautious about extrapolating beyond the studied range. The key takeaway is that cooling power is the dominant factor, and we should focus on optimizing this parameter first. This is a useful insight for any engineer working on corrosion-resistant cladding. I recommend that our process engineers study this paper and apply the response surface methodology to our own cladding operations. The result will be better control of distortion and residual stress, which translates to higher quality and lower rework costs. This is exactly the kind of research that pays for itself on the shop floor.