Fuzzy Comprehensive Evaluation of Strip Cladding Forming Quality Based on Matlab-FIS

Literature Overview

This paper by Guo Xiao, Xu Kai, and Zou Liwei from the Harbin Welding Research Institute of the Chinese Academy of Machinery Science and Technology, published in the Welding Journal in 2013 and supported by the National Science and Technology Major Project (2012ZX060004-21; 2011ZX04016-061), presents a fuzzy comprehensive evaluation (FCE) method for assessing the forming quality of strip cladding (electroslag strip cladding) processes. The work applies fuzzy logic and fuzzy inference systems (FIS) implemented in MATLAB to create an objective, quantitative evaluation framework for strip cladding quality, addressing the inherent subjectivity and variability in traditional visual and dimensional inspection methods.

Core Technical Content

Strip Cladding Process Overview

Electroslag strip cladding (also known as ESW cladding or electroslag weld overlay) is a high-productivity process for depositing thick cladding layers onto steel substrates. The process uses a continuous metal strip as the filler material, which is melted in an electroslag pool and deposited onto the substrate. Key process parameters include:

Parameter	Typical Range	Effect on Forming Quality
Current	800–1500 A	Influences penetration and deposition rate
Voltage	35–50 V	Controls slag pool temperature and fluidity
Travel speed	100–300 mm/min	Affects bead width and profile
Strip width	25–75 mm	Determines single-pass coverage
Flux composition	Rutile, basic, or mixed	Influences slag properties and wetting
Preheat temperature	150–300 °C	Reduces thermal stress and cracking

Fuzzy Comprehensive Evaluation Methodology

The FCE method involves several key steps:

Identification of evaluation factors: Selection of quality characteristics that define strip cladding forming quality.
Fuzzification: Conversion of crisp measurements into fuzzy membership functions.
Weighting: Assignment of importance weights to each factor based on engineering significance.
Fuzzy inference: Application of fuzzy rules to determine the overall quality grade.
Defuzzification: Conversion of the fuzzy output into a crisp evaluation score.

Evaluation Factors and Criteria

The study identifies the following key evaluation factors for strip cladding forming quality:

Factor	Measurement Method	Weight	Quality Grade Thresholds
Bead width uniformity	Laser scanning / optical measurement	0.20	±2 mm tolerance
Bead height uniformity	Cross-sectional measurement	0.20	±1 mm tolerance
Surface profile	Coordinate measurement	0.15	Ra < 12.5 μm
Dilution ratio	Optical emission spectrometry	0.20	<25%
Bond strength	Shear test per ASTM E23	0.15	>200 MPa
Surface defects	Visual + PT inspection	0.10	No critical defects

Fuzzy Inference System Implementation

The MATLAB-FIS implementation defines the following fuzzy sets for each evaluation factor:

Excellent (E): Membership function centered at the optimal value with a narrow spread.
Good (G): Membership function centered slightly below optimal with moderate spread.
Acceptable (A): Membership function covering the tolerance range.
Poor (P): Membership function for values exceeding tolerance limits.
Rejected (R): Membership function for severely out-of-tolerance values.

The fuzzy rules follow the format: IF factor_i is grade_j THEN overall_quality is grade_k, with multiple rules combined using fuzzy operators (AND, OR) to produce a comprehensive evaluation.

Engineering Practice Integration

Quality Control Application

The FCE method provides a systematic framework for quality control in strip cladding production:

Real-time monitoring: Integration with automated measurement systems enables continuous quality assessment during production.
Process optimization: Identification of the most influential quality factors guides parameter optimization.
Acceptance/rejection decisions: Objective criteria reduce subjectivity in quality decisions.
Trend analysis: Statistical process control (SPC) integration enables early detection of quality degradation.

Comparison with Traditional Evaluation Methods

Method	Objectivity	Speed	Cost	Applicability
Visual inspection	Low	Fast	Low	Surface defects only
Dimensional measurement	Medium	Medium	Medium	Geometry only
Destructive testing	High	Slow	High	Limited to samples
Fuzzy comprehensive evaluation	High	Fast	Medium	Comprehensive assessment
data analysis-based	Very High	Fast	Medium–High	Requires training data

Case Application: Bimetallic Pressure Vessel Cladding

In the fabrication of bimetallic pressure vessels per GB/T 150 and NB/T 47002, strip cladding is often used to build up the corrosion-resistant lining on the carbon steel shell. The FCE method can be applied to evaluate the cladding quality at critical locations:

Vessel heads: Curved surfaces require special attention to bead profile uniformity.
Weld seams: Cladding over circumferential and longitudinal welds requires assessment of dilution and bonding quality.
Nozzle areas: High stress regions where cladding quality directly affects structural integrity.

Study Insights and Reflections

This research represents a significant advancement in the quality assessment methodology for strip cladding processes. The application of fuzzy logic to welding quality evaluation addresses the inherent uncertainty and imprecision in traditional inspection methods, providing a more nuanced and comprehensive assessment framework.

From an engineering practice perspective, the FCE method offers several practical advantages:

Reduced subjectivity: By converting expert judgment into mathematical membership functions, the method reduces variability between inspectors.
Multi-factor integration: The ability to simultaneously consider multiple quality characteristics provides a more holistic view of cladding quality.
Scalability: The method can be adapted to different cladding processes, materials, and application requirements.
Documentation: The fuzzy rules and membership functions create a documented, auditable quality assessment process.

However, several limitations should be acknowledged:

Initial setup complexity: Defining appropriate membership functions and weights requires expert knowledge and may be time-consuming.
Calibration requirement: The system must be calibrated against known quality outcomes to ensure accuracy.
Dynamic conditions: The method assumes relatively stable process conditions; significant process variations may require real-time rule adjustment.
Standardization: Lack of industry-wide standards for FCE application in welding limits widespread adoption.

The integration of fuzzy comprehensive evaluation with modern sensor technologies and data acquisition systems could enable real-time quality monitoring and closed-loop process control, representing a significant step toward intelligent manufacturing in the cladding industry. The methodology also aligns with the principles of statistical process control and continuous improvement, making it compatible with existing quality management systems such as ISO 9001 and ASME QME-1.

This work demonstrates the value of applying computational intelligence methods to traditional manufacturing quality problems. As the industry moves toward Industry 4.0 and digital manufacturing, such hybrid approaches combining domain expertise with computational tools will become increasingly important for maintaining and improving product quality in complex welding and cladding operations.