Fuzzy Comprehensive Evaluation of Strip Electrode Surfacing Forming Quality Using Matlab-FIS

Literature Overview

This paper, published in 2013 in the Welding Journal (Vol. 34, Issue 8, pp. 89-91) by Guo Xiao, Xu Kai, and Zou Liwei from the Harbin Welding Research Institute of the China Academy of Machinery Science and Technology, presents a fuzzy comprehensive evaluation model for strip electrode surfacing forming quality based on the Matlab-FIS (Fuzzy Inference System) platform. The research was supported by the National Science and Technology Major Project funds (2012ZX060004-21; 2011ZX04016-061). The work addresses the challenge of quantitatively assessing the geometric quality of surfacing layers produced by strip electrode welding, a process widely used in building up thick sections and restoring worn surfaces in heavy equipment manufacturing.

Core Methodology and Model Construction

The authors employed the Fuzzy Analytic Hierarchy Process (FAHP) to construct a multi-criteria evaluation framework. Four key geometric indicators were selected to characterize the surfacing layer forming quality: bead thickness, bead width, edge contact angle, and straightness. Each indicator represents a distinct aspect of the deposited layer geometry that affects both the aesthetic quality and the functional performance of the surfacing.

The evaluation model was structured hierarchically. At the top level is the overall forming quality, which is decomposed into the four geometric indicators at the second level. The FAHP method was used to construct fuzzy consistent judgment matrices, from which the weight coefficients for each indicator were determined. Membership functions for each indicator were established based on practical welding experience, translating measured geometric values into fuzzy membership degrees between 0 and 1.

Fuzzy Inference System Implementation

The Matlab-FIS platform was used as the computational engine for the fuzzy evaluation model. The fuzzy inference process involves three main steps: fuzzification of the input geometric measurements, application of fuzzy rules to determine the membership degree of each quality grade, and defuzzification to obtain a crisp evaluation score. The fuzzy rules encode expert knowledge about how combinations of geometric measurements correspond to acceptable or unacceptable forming quality.

The validation of the model was conducted through experimental welding trials. The surfacing layers produced under different process conditions were measured for all four geometric indicators, and the fuzzy evaluation scores were compared with direct visual and instrumental observations. The results showed good agreement between the model predictions and the actual assessment outcomes, confirming the validity of the approach.

Technical Significance and Process Understanding

Strip electrode surfacing is a high-deposition-rate process that uses a continuous strip of electrode material fed into the arc. The process parameters that influence forming quality include current, voltage, travel speed, stick-out length, and shielding gas flow rate. The geometric quality of the deposited layer is sensitive to these parameters, and small variations can lead to significant changes in bead profile.

The use of fuzzy logic for quality evaluation is particularly appropriate in this context because the relationship between process parameters and geometric outcomes is often nonlinear and difficult to describe with precise mathematical equations. Fuzzy logic accommodates the inherent uncertainty and imprecision in welding process monitoring, allowing for evaluation based on approximate reasoning that mirrors how experienced welders assess bead quality visually.

Engineering Practice and Quality Control Integration

For production environments where strip electrode surfacing is used to build up thick sections on large components such as ship hulls, pressure vessels, and heavy machinery parts, the ability to quantitatively assess forming quality is essential for maintaining consistent product quality. The fuzzy evaluation model can be integrated into online or offline quality control systems, providing a standardized scoring mechanism that reduces reliance on subjective visual inspection alone.

The model also has value in process optimization. By systematically varying process parameters and recording the resulting geometric measurements, engineers can use the fuzzy evaluation scores to identify the parameter windows that produce the best forming quality. This data-driven approach complements traditional trial-and-error methods and can accelerate the qualification of new surfacing processes.

Study Insights and Reflections

This work represents a meaningful contribution to the field of welding quality assessment by combining fuzzy mathematics with practical engineering knowledge. The selection of four geometric indicators provides a comprehensive yet manageable framework for evaluating surfacing layer quality. The use of Matlab-FIS as the computational platform makes the model accessible to engineers familiar with MATLAB, facilitating adoption in research and development settings. The approach could be extended to include additional quality indicators such as dilution rate, microhardness distribution, and surface roughness, creating a more holistic evaluation system. The validation against experimental results provides confidence in the model's predictive capability, making it a practical tool for quality control in strip electrode surfacing operations.