Abstract:This review article highlights state-of-the-art data-driven techniques to discover, encode, surrogate, or emulate constitutive laws that describe the path-independent and path-dependent response of solids. Our objective is to provide an organized taxonomy to a large spectrum of methodologies developed in the past decades and to discuss the benefits and drawbacks of the various techniques for interpreting and forecasting mechanics behavior across different scales. Distinguishing between machine-learning-based and model-free methods, we further categorize approaches based on their interpretability and on their learning process/type of required data, while discussing the key problems of generalization and trustworthiness. We attempt to provide a road map of how these can be reconciled in a data-availability-aware context. We also touch upon relevant aspects such as data sampling techniques, design of experiments, verification, and validation.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the modeling of constitutive relations in solid materials, especially how to use data - driven methods to discover, encode or simulate the constitutive laws (CL) that describe the response of solid materials. Specifically, the article aims to: 1. **Provide a systematic classification and review**: Classify and summarize various data - driven methods developed in the past few decades, including machine - learning - based (ML) methods and model - free methods. These methods are used to explain and predict mechanical behaviors at different scales. 2. **Discuss the advantages and disadvantages of various techniques**: Analyze the differences in interpretability and required data types for each method, and also explore key issues such as generality and credibility. 3. **Provide a roadmap for data - driven constitutive modeling**: In the context of considering data availability, make suggestions on how to coordinate different methods, and cover aspects such as data sampling techniques, experimental design, verification and validation. 4. **Combine physical constraints and data**: Emphasize the importance of combining physical constraints with data to reduce the need for large data sets and improve the generalization ability of the model. ### Specific problems - **Limitations of traditional methods**: Traditional constitutive modeling mainly relies on limited experimental observation data, which limits its generalization ability under complex conditions. - **Opportunities and challenges brought by emerging experimental techniques**: The development of full - field experimental methods such as digital image correlation (DIC), X - ray computed tomography (CT) and digital volume correlation (DVC) enables the acquisition of more abundant data, but also brings the challenge of how to handle large amounts of data. - **Application prospects of data - driven methods**: Use big - data analysis and information - mining techniques to improve constitutive modeling, especially in the applications in mechanics and materials science. ### Goals Through the above efforts, the paper hopes to provide guidance for future research and promote the development of data - driven constitutive modeling in the field of solid mechanics, especially in the following aspects: - **Improve prediction accuracy**: By combining data and physical knowledge, achieve efficient and high - precision prediction. - **Reduce costs**: Reduce experimental costs in industrial applications. - **Promote material innovation**: Accelerate the development and optimization of new materials. In summary, the core problem of this paper is to explore how to use modern data - driven techniques to improve the constitutive modeling of solid materials, so as to better understand and predict the behavior of materials.

A review on data-driven constitutive laws for solids

A Thermodynamics-Informed Neural Network for Elastoplastic Constitutive Modeling of Granular Materials

Data-driven computational mechanics: comparison of model-free and model-based methods in constitutive modeling

Model-data-driven constitutive responses: Application to a multiscale computational framework

Neural Networks for Constitutive Modeling -- From Universal Function Approximators to Advanced Models and the Integration of Physics

Physics-informed Data-driven Discovery of Constitutive Models with Application to Strain-Rate-sensitive Soft Materials

Constitutive model characterization and discovery using physics-informed deep learning

The Language of Hyperelastic Materials

Model-free data-driven methods in mechanics: material data identification and solvers

Multiscale computational solid mechanics: data and machine learning

Insight into constitutive theories of 4D printed polymer materials: a review

Thermodynamically Consistent Machine-Learned Internal State Variable Approach for Data-Driven Modeling of Path-Dependent Materials

Learning Physics-Consistent Material Behavior Without Prior Knowledge

Extreme sparsification of physics-augmented neural networks for interpretable model discovery in mechanics

Model-free Data-Driven simulation of inelastic materials using structured data sets, tangent space information and transition rules

Model-free Data-Driven inference in computational mechanics

Automated model discovery of finite strain elastoplasticity from uniaxial experiments

Unbiased construction of constitutive relations for soft materials from experiments via rheology-informed neural networks

Learning the physics-consistent material behavior from experimentally measurable data via PDE-constrained optimization

Machine Learning for Structural Materials

Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage