Multimaterial Metamaterial Inverse Design via Machine Learning for Tailorable and Reusable Energy Absorption.

Area of Science:

• Materials Science
• Mechanical Engineering
• Computational Science

Background:

• Growing demand for energy-absorbing structures with customizable mechanical properties.
• Existing design methods struggle with multi-objective optimization and material integration.

Purpose of the Study:

• To develop a machine learning-driven inverse design framework for tailorable energy-absorbing metamaterials.
• To co-optimize material selection, lattice topology, and geometric parameters for specific performance targets.

Main Methods:

• Utilized a 181-dimensional parameter space for optimization.
• Integrated multimaterial compatibility (TPU/resin/NiTi/Al alloy) with body-centered cubic (BCC) lattices.
• Conducted over 20,000 finite element analysis simulations.
• Employed artificial neural networks and genetic algorithms for inverse design.

Main Results:

• Achieved precise control over specific platform stress (0.015–4.05 MPa) and specific energy absorption (0.049–23.377 J/g).
• Demonstrated high reusability (over 50% recovery after five cycles) for NiTi alloy metamaterials.
• Validated the framework through additive manufacturing and experimental characterization.

Conclusions:

• The proposed framework effectively bridges the gap between customizable energy absorption and structural reusability.
• Machine learning enables efficient inverse design of advanced energy-absorbing metamaterials.
• The developed method offers a pathway for creating tailored, high-performance energy absorption solutions.