Abstract:Triply periodic minimal surfaces (TPMSs) have attracted great attention due to their distinct advantages such as high strength and light weight compared to traditional lattice structures. Most previous works focus on forward prediction of the mechanical behaviors of TPMSs. Inverse design of the configurations based on customized loading curves would be of great value in engineering applications such as energy absorption. Inspired by TPMSs, we propose the concept of the shell-based mechanical metamaterial (SMM) in this work, which possesses the main geometrical features and mechanical properties of TPMSs. A novel approach, combining machine learning (ML) for high efficiency and genetic algorithm (GA) for global optimization, is put forward to inversely design the configuration of SMM. Two strategies are introduced to develop artificial neural networks (ANNs) for the prediction of their loading curves under compression. GA is then employed to design objective configurations with customized loading curves. The connection between the loading curves and deformation modes is also illustrated to demonstrate the values of such inverse design. SMM with a strain-hardening curve tends to exhibit globally uniform deformation, while SMM with a strain-softening curve tends to present layer-by-layer deformation during compression, which is demonstrated by experiments and simulations. This work fills the blanks of inverse design of SMM with customized loading curves and contributes to the concept of structure design combining ML and traditional optimization approaches.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the inverse design of shell - based mechanical metamaterials (SMM) based on customized loading curves. Specifically, most previous studies have focused on predicting the mechanical behavior of triply periodic minimal surfaces (TPMSs), while this paper focuses on how to design the configuration of TPMSs according to the required mechanical properties. To achieve this goal, the author proposes the concept of shell - based mechanical metamaterials and combines machine learning (ML) and genetic algorithm (GA) to propose a new method for the inverse design of SMM configurations. ### Main research contents: 1. **Proposing shell - based mechanical metamaterials (SMM)**: - SMM has the main geometric features and mechanical properties of TPMSs, but has a larger design space, which can meet the needs of inverse design. - The design space of SMM includes periodicity, cubic symmetry, smoothness, and central symmetry or mirror symmetry. 2. **Using machine learning (ML) for rapid prediction of loading curves**: - An artificial neural network (ANN) is used to construct a surrogate model to efficiently predict the loading curve of SMM under compressive load. - Two strategies are proposed to construct ANN: one is that a single ANN processes all outputs (ANN1), and the other is that multiple independent ANNs process each output point respectively (ANN2). 3. **Using genetic algorithm (GA) for inverse design**: - GA is used for global optimization to search for the SMM configuration closest to the target loading curve. - The performance of candidate configurations is quickly evaluated through the trained ANN, thereby improving the efficiency of inverse design. ### Experimental results: - **Forward prediction**: - A loading curve data set of 7,000 SMM configurations is generated by the finite element method (FEM), of which 5,500 are used for training and 1,500 are used for testing. - The relative errors of ANN1 and ANN2 on the test set are 3.1% and 2.7% respectively, indicating that ANN can accurately predict the loading curve of SMM. - **Inverse design**: - The SMM configuration with a customized loading curve is successfully designed, demonstrating the effectiveness and practicality of inverse design. - The deformation modes corresponding to different loading curves are verified by experiments and numerical simulations, further proving the value of inverse design. ### Conclusion: This paper fills the gap in the field of SMM inverse design based on customized loading curves. By combining machine learning and traditional optimization methods, it provides an efficient and accurate design method. This not only helps in the design of energy - absorbing structures, but also provides new ideas for the inverse design of mechanical metamaterials.

Inverse design of shell-based mechanical metamaterial with customized loading curves based on machine learning and genetic algorithm

Inverse Design of Cellular Structures with the Geometry of Triply Periodic Minimal Surfaces Using Generative Artificial Intelligence Algorithms

Optimal Design of One-Dimensional Elastic Metamaterials Through Deep Convolutional Neural Network and Genetic Algorithm

Inverse Design of Mechanical Metamaterial Achieving a Prescribed Constitutive Curve

A Deep Learning Approach for Inverse Design of Gradient Mechanical Metamaterials

Performance-based Inverse Structural Design of Complex Gradient Triply Periodic Minimal Surface Structures Based on a Deep Learning Approach

Rapid Inverse Design of Metamaterials Based on Prescribed Mechanical Behavior Through Machine Learning.

A Deep Learning Approach for Reverse Design of Gradient Mechanical Metamaterials

Machine Learning-Accelerated Inverse Design of Programmable Bi-Functional Metamaterials

Inverse design of anisotropic bone scaffold based on machine learning and regenerative genetic algorithm

Tailoring the mechanical properties of 3D microstructures: a deep learning and genetic algorithm inverse optimization framework

3D Printing of Flexible Mechanical Metamaterials: Synergistic Design of Process and Geometric Parameters

Machine Learning-Encoded Multiscale Modelling and Bayesian Optimization Framework to Design Programmable Metamaterials

Machine learning – enabled inverse design of shell-based lattice metamaterials with optimal sound and energy absorption

Progressive design of gradually stiffer metamaterial using surrogate model

A Novel Triple Periodic Minimal Surface-Like Plate Lattice and Its Data-Driven Optimization Method for Superior Mechanical Properties

Machine Learning-Based Design and Optimization of Curved Beams for Multistable Structures and Metamaterials

A Design Framework for Gradually Stiffer Mechanical Metamaterial Induced by Negative Poisson's Ratio Property

Customizable anisotropic microlattices for additive manufacturing: Machine learning accelerated design, mechanical properties and structural-property relationships

Inverse Design of Mechanical Metamaterials Using a Point-Cloud-Based Deep Generative Model

Inverse design of Bézier curve-based mechanical metamaterials with programmable negative thermal expansion and negative Poisson's ratio via data augmented deep autoencoder