Modeling and Optimization of Fractal Dimension in Wire Electrical Discharge Machining of EN 31 Steel Using the ANN-GA...

Area of Science:

• Manufacturing Engineering
• Materials Science
• Computational Intelligence

Background:

• Wire Electrical Discharge Machining (WEDM) is a non-traditional machining process used for intricate shapes.
• Surface characteristics are critical for component performance and longevity.
• Fractal dimension offers a scale-independent metric for surface roughness analysis.

Purpose of the Study:

• To correlate and optimize WEDM process parameters for enhanced surface characteristics.
• To model the complex relationship between WEDM parameters and fractal dimension.
• To achieve superior surface texture by minimizing defects.

Main Methods:

• Utilized artificial neural network (ANN) with a feed-forward architecture and Levenberg-Marquardt algorithm for modeling.
• Employed a genetic algorithm (GA) for optimizing WEDM parameters.
• Conducted experiments based on a rotatable central composite design.
• Analyzed machined surfaces using Scanning Electron Micrography (SEM).

Main Results:

• Developed a 4-3-3-1 ANN architecture that accurately predicts fractal dimension (R-value of 0.97).
• Identified optimal machining parameters using GA, showing close agreement with experimental results.
• SEM analysis confirmed significant improvement in surface texture, with reduced formation of re-solidified globules.

Conclusions:

• The combined ANN-GA approach effectively models and optimizes WEDM parameters for improved surface quality.
• This methodology offers a pathway to achieve enhanced surface characteristics in WEDM.
• The reduction in re-solidified globules leads to a superior machined surface texture.