Machine learning based optimization of titanium electropolishing using artificial neural networks and Taguchi design...

Area of Science:

• Materials Science and Engineering
• Surface Engineering
• Sustainable Manufacturing

Background:

• Electrochemical surface treatments like electropolishing are crucial for material finishing.
• Optimizing process parameters is complex due to nonlinear interactions.
• Developing eco-friendly electrolytes is essential for sustainable manufacturing.

Purpose of the Study:

• To develop an integrated optimization framework for electropolishing titanium using an eco-friendly deep eutectic solvent.
• To systematically analyze the effects of key process parameters on surface roughness.
• To achieve a minimum surface roughness through optimized conditions.

Main Methods:

• Combining artificial neural networks (ANN) and Taguchi robust design.
• Utilizing a multilayer perceptron-based ANN model for prediction.
• Employing Gaussian noise for data augmentation to address data scarcity.
• Systematic analysis of applied voltage, processing time, temperature, electrolyte composition, and water concentration.

Main Results:

• The ANN model achieved a high predictive accuracy (R² = 0.981).
• Optimal conditions identified: 20 V, 21 min, 1:4 electrolyte ratio, 0% distilled water, 42°C.
• Experimentally confirmed minimum surface roughness of 4.162 nm.

Conclusions:

• The integrated framework effectively optimizes electropolishing processes.
• The study demonstrates a practical approach for environmentally sustainable manufacturing.
• Quantitative analysis of nonlinear complexity in electrochemical treatments is achieved.