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

Hybrid Fuzzy and Integral Control of Weld Width in TIG Welding

Literature Overview and Core Findings

This study by Huang Shisheng and He Jianfeng from South China University of Technology, published in Control Theory and Applications in 1995, addresses a classic problem in automated TIG welding: how to control the weld bead width in real time. The weld width is the most important geometric parameter of a weld, because it determines the joint strength, the heat input and the appearance of the weld. In automated welding, the weld width must be controlled to a tight tolerance, but the process is nonlinear, time-varying and subject to disturbances such as arc length variation, joint gap variation and heat sink changes.

The authors propose a parameter self-adjusting fuzzy and integral hybrid control mode, which combines the advantages of fuzzy control and integral control. Fuzzy control provides good dynamic response and robustness to disturbances, while integral control eliminates steady-state error. The hybrid controller adjusts its parameters based on the error and the rate of change of the error, so that it can respond quickly to large errors and accurately to small errors. The simulation results demonstrate the practicality and effectiveness of the method.

Interpretation of Technical Points

The weld width control problem is fundamentally a thermal process control problem. The weld width is determined by the heat input, which is a function of welding current, arc voltage, travel speed and the thermal properties of the material. In a typical TIG welding system, the welding current is the primary control variable, and the arc voltage and travel speed are held constant. However, the relationship between welding current and weld width is nonlinear, and it changes with the joint geometry, the material thickness and the ambient conditions.

Fuzzy control is well suited to this problem because it does not require an accurate mathematical model of the process. A fuzzy controller uses linguistic rules such as "if the weld width error is large positive, then increase the current significantly" and "if the weld width error is small negative, then decrease the current slightly". The rules are based on the experience of the welding engineer, and they can be tuned to achieve the desired performance. However, a pure fuzzy controller has a steady-state error because the rule base is discrete and the output is quantised.

Integral control is the standard solution to the steady-state error problem. An integral controller accumulates the error over time and adjusts the output until the error is zero. However, a pure integral controller has poor dynamic response and can cause overshoot and oscillation. The hybrid approach combines the two: the fuzzy controller provides the fast dynamic response, and the integral controller provides the accurate steady-state performance. The parameter self-adjusting feature allows the controller to adapt to changes in the process, such as a change in the joint gap or a change in the material thickness.

Integration with Engineering Practice

The weld width control problem is directly relevant to automated pipe welding, especially in orbital TIG welding of root passes and in longitudinal welding of pipe. In my experience, the root pass is the most critical weld in a pipe joint, because it determines the quality of the entire weld. If the root pass has a lack of fusion or an undercut, the weld will fail in service, and the repair cost is high. The weld width of the root pass must be controlled precisely to ensure complete fusion with the pipe wall and a smooth transition to the cap passes.

In automated pipe welding, the weld width is typically controlled by adjusting the welding current or the travel speed. However, the weld width is difficult to measure in real time because the weld pool is molten and the arc is bright. Optical sensors can be used to measure the weld pool width, but they are sensitive to arc light, spatter and fume. The authors' approach to control is based on the weld width as the controlled variable, which implies that a weld width sensor is available. In practice, the sensor must be robust and reliable, and the control algorithm must be able to handle the sensor noise and the process disturbances.

The hybrid fuzzy and integral control approach is particularly suitable for this application because it can handle the nonlinearity and the time-varying nature of the welding process. The fuzzy controller can be designed based on the experience of the welding engineer, and the integral controller can be tuned to eliminate the steady-state error. The parameter self-adjusting feature allows the controller to maintain performance as the process conditions change, such as when the pipe is rotated and the heat sink changes.

Key Questions and Reflections

The study raises an important question about the practical implementation of the hybrid controller. The simulation results are promising, but the controller must be tested on a real welding system before it can be used in production. The welding process has many disturbances that are not present in the simulation, such as arc length variation, electrode wear, shielding gas fluctuations and material property variations. The controller must be robust to these disturbances, and the sensor must be reliable in the harsh welding environment.

Another question is the choice of the weld width as the controlled variable. The weld width is a geometric parameter that is determined by the heat input and the material properties. However, the weld width is not the only important parameter; the weld penetration, the weld reinforcement and the weld profile are also important. In practice, the weld width is often used as a proxy for the heat input, because it is easier to measure than the weld penetration. However, the relationship between weld width and penetration is not always linear, and a weld with the correct width may have insufficient penetration if the heat input is distributed incorrectly.

The economic case for the hybrid controller is strong in high-volume automated welding. If the weld width can be controlled to a tighter tolerance, the rejection rate can be reduced, and the productivity can be increased. In my experience, the rejection rate in automated pipe welding is often 5–10% due to weld geometry variations, and a good control system can reduce this to less than 1%. The savings in rework and scrap can justify the investment in the control system.

Study Insights and Implications

This paper is a valuable contribution to the field of welding automation, and it demonstrates that the weld width control problem can be solved with a hybrid fuzzy and integral controller. The parameter self-adjusting feature is particularly important, because it allows the controller to adapt to changes in the process without manual retuning. The simulation results show that the hybrid controller has good dynamic and static characteristics, and it is more effective than either pure fuzzy control or pure integral control.

From a production standpoint, the key insight is that the weld width can be controlled in real time with a combination of fuzzy logic and integral action. The fuzzy logic provides the fast response and the robustness to disturbances, while the integral action eliminates the steady-state error. The parameter self-adjusting feature ensures that the controller maintains its performance as the process conditions change. This is a practical solution to a real production problem, and it can be implemented with a standard industrial controller.

The practical guidance for fellow engineers is to consider the hybrid fuzzy and integral controller for automated TIG welding applications where the weld width must be controlled to a tight tolerance. The controller can be implemented on a programmable logic controller (PLC) or a dedicated welding controller, and it can be integrated with a weld width sensor. The controller should be tuned based on the specific welding process, and it should be validated with a full welding procedure qualification before it is used in production.

In summary, the hybrid fuzzy and integral control approach is an effective solution to the weld width control problem in TIG welding. The controller combines the fast dynamic response of fuzzy control with the accurate steady-state performance of integral control, and the parameter self-adjusting feature allows it to adapt to process variations. The simulation results demonstrate the practicality and effectiveness of the method, and the approach can be applied to automated pipe welding and other welding applications where weld geometry control is critical. The paper is a valuable reference for welding engineers and control engineers who are working on welding automation, and it reinforces the principle that the best control solutions are those that combine the strengths of multiple control strategies.