Wear Resistance Design of Laser Cladding Ni-Based Self-Fluxing Alloy Coating Using Machine Learning

Area of Science:

• Materials Science
• Mechanical Engineering
• Computational Science

Background:

• Laser cladding is crucial for creating wear-resistant coatings.
• Optimizing Ni-based self-fluxing alloy (SFA) coatings requires balancing composition and processing parameters.
• Current design methods can be inefficient and time-consuming.

Purpose of the Study:

• To apply machine learning for improving the collaborative design of laser cladding Ni-based SFA wear-resistant coatings.
• To develop an efficient optimization system for alloy composition and processing parameters.
• To provide methodological insights for laser processing applications.

Main Methods:

• Constructed a comprehensive database from literature data.
• Utilized feature correlation analysis (Pearson's) and feature importance (Random Forest - RF).
• Evaluated five classical machine learning algorithms and combined RF with a Genetic Algorithm (GA) for optimization.

Main Results:

• Identified Carbon (C) and Boron (B) as significant elements influencing wear properties.
• The RF-GA optimization system successfully designed multiple composition and process plans.
• Optimized alloy exhibited superior wear resistance (friction coefficient of 0.34) due to enhanced strengthening effects and hard phase content.

Conclusions:

• Machine learning, particularly the RF-GA system, offers an efficient approach to optimize laser cladding processes.
• The study provides valuable theoretical support and methodological insights for developing advanced wear-resistant coatings.
• This approach can be extended to optimize other laser processing applications.