CLADDING TECHNOLOGY SHANXI CO., LTDFuzzy Control System for MIG Welding Fusion Width Regulation
Literature Overview
The 1999 publication by Zhang Jiaying, Jiang Liping, and Zhang Xianghong in Transactions of the Welding Institute of China presents a fuzzy control system designed to regulate the fusion width in MIG welding operations. The research was conducted collaboratively across Beijing Institute of Arms Engineering, Beijing Institute of Petrochemical Technology, and the Beijing Military Medical Research Institute, reflecting a multidisciplinary approach to welding process automation. At that time, fuzzy control theory was gaining traction in Chinese welding research circles as a practical alternative to classical PID controllers, particularly for nonlinear, time-varying systems such as the welding arc.
Core Technical Concept
The fundamental challenge addressed in this work is the precise control of weld bead fusion width during MIG (Gas Metal Arc Welding) operations. Fusion width is a critical geometric parameter that directly influences weld quality, mechanical properties, and dimensional accuracy. In conventional welding control, operators rely on manual adjustment or simple feedback loops that struggle with the dynamic coupling between welding current, voltage, travel speed, and resulting bead geometry.
The fuzzy control approach proposed here employs a knowledge-based inference system that maps input variables—typically welding current (I), arc voltage (U), and travel speed (V)—to a desired fusion width output through a set of fuzzy rules. The system architecture follows a standard fuzzy logic framework:
| Component | Function | Typical Implementation |
|---|---|---|
| Fuzzification | Converts crisp inputs to fuzzy sets | Triangular or trapezoidal membership functions |
| Rule Base | Encodes expert knowledge | IF-THEN rules derived from welding experience |
| Inference Engine | Applies fuzzy logic operations | Mamdani or Sugeno type |
| Defuzzification | Converts fuzzy output to control signal | Center of gravity method |
The key innovation lies in the application of fuzzy logic to real-time fusion width feedback. The system likely employs an optical or capacitive sensor to measure bead width during welding, compares it with the target value, and adjusts welding parameters through the fuzzy controller.
Process Analysis and Engineering Relevance
In MIG welding, the relationship between process parameters and fusion width is inherently nonlinear. The fusion width (W) can be approximated by empirical relationships such as:
W ∝ (I / V)^(0.5) × (1/V_travel)^(0.3)
However, this relationship varies significantly with:
- Base metal thickness and thermal conductivity
- Shielding gas composition (Ar/CO2 mixtures)
- Wire diameter and stickout length
- Joint geometry and fit-up
The fuzzy controller offers advantages over PID control in this context because it does not require a precise mathematical model of the welding process. Instead, it leverages the accumulated experience of skilled welders encoded as linguistic rules. For example:
- IF fusion width is "too narrow" AND current is "low" THEN increase current "moderately"
- IF fusion width is "too wide" AND travel speed is "slow" THEN increase travel speed "slightly"
Defect Analysis and Countermeasures
From a quality control perspective, improper fusion width leads to specific welding defects:
| Defect Type | Cause | Consequence | Fuzzy Control Countermeasure |
|---|---|---|---|
| Excessive penetration | Too wide fusion width | Burn-through, undercut | Reduce current, increase travel speed |
| Incomplete fusion | Too narrow fusion width | Cold lap, poor joint strength | Increase current, decrease travel speed |
| Excessive reinforcement | Excessive deposition | Stress concentration | Reduce wire feed rate |
| Porosity | Wide fusion zone | Reduced toughness | Optimize gas flow, control arc length |
Integration with Engineering Practice
In modern manufacturing environments, the principles outlined in this 1999 research have evolved significantly. Contemporary welding cells employ multi-sensor fusion approaches combining optical (CCD-based), ultrasonic, and electrical measurements for comprehensive weld monitoring. However, the fundamental fuzzy control philosophy remains relevant, particularly in:
- Robotic welding cells where parameter adaptation is required for varying joint conditions
- Thick-section welding where multi-pass procedures require dynamic parameter adjustment
- Cladding operations where dilution control is critical for maintaining overlay composition
For cladding and weld overlay applications specifically, the fusion width control concept translates directly to dilution rate management. In stainless steel cladding on carbon steel substrates, maintaining fusion width within a narrow window (typically 2-3 mm per pass) is essential to limit carbon steel dilution to the overlay layer, thereby preserving corrosion resistance.
Key Questions and Reflections
Several questions arise from studying this early fuzzy control work:
- What is the achievable accuracy of fusion width control using fuzzy logic alone, compared to modern data analysis approaches?
- How does the system handle disturbances such as wire feed rate fluctuations or gas flow variations?
- What is the computational burden of real-time fuzzy inference on 1999-era microprocessors?
The research represents an important milestone in Chinese welding automation research, bridging the gap between theoretical control engineering and practical welding applications. Its methodology remains instructive for engineers designing adaptive welding systems today, particularly when sensor data is noisy or process models are uncertain.
Study Insights and Implications
The enduring value of this work lies in its demonstration that intelligent control strategies can compensate for the inherent variability of arc welding processes. For engineers working in bimetal product manufacturing, the fuzzy control philosophy suggests that overlay thickness and dilution control can benefit from similar adaptive approaches. Rather than relying solely on pre-programmed parameter sets, real-time feedback systems that continuously adjust welding conditions based on measured outcomes can significantly improve consistency and reduce rework rates in production environments.