AI-Powered Very-High-Cycle Fatigue Control: Optimizing Microstructural Design for Selective Laser Melted Ti-6Al-4V

Area of Science:

• Materials Science and Engineering
• Mechanical Engineering
• Computational Science

Background:

• Additive manufacturing (AM) presents opportunities for advanced material design.
• Optimizing material properties for fatigue resistance is crucial for critical applications.
• High cycle fatigue (HCF) testing is essential for assessing material durability.

Purpose of the Study:

• To integrate machine learning (ML) into AM for enhanced material properties and fatigue resistance.
• To explore mechanistic ML techniques for tailoring microstructures.
• To accelerate the design and production of reliable, high-performance components.

Main Methods:

• Utilized ultrasonic fatigue tests to gather very high cycle fatigue data (up to 1x10^10 cycles).
• Applied mechanistic machine learning models to predict fatigue thresholds and optimize process parameters.
• Analyzed microstructural features like grain orientation and phase uniformity.

Main Results:

• ML models reduced design iteration cycles by over 50%.
• Fatigue crack propagation resistance improved by 20-30% through microstructural refinement.
• Achieved 15% weight reduction and improved yield strength in ML-designed metamaterials.
• Identified key process parameters (temperature gradients, cooling rates) influencing microstructural evolution in titanium and aluminum alloys.

Conclusions:

• ML integration in AM significantly enhances fatigue life and reliability for critical components.
• ML enables the design of lightweight, high-strength metamaterials for diverse applications.
• This research paves the way for intelligent, adaptive manufacturing systems with improved performance and efficiency.