Reproduce “A review on data-driven constitutive laws for solids” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jan Niklas Fuhg, Govinda Anantha Padmanabha, Nikolaos Bouklas, Bahador Bahmani, WaiChing Sun, Nikolaos N. Vlassis, Moritz Flaschel, Pietro Carrara and 1 more
This review paper provides a comprehensive classification and critical analysis of data-driven methods for discovering and modeling constitutive laws in solid mechanics. It emphasizes integrating physical constraints into machine learning approaches to enhance model interpretability, generalization, and trustworthiness, while also proposing a verification and validation framework.
Key insights
- Integrating physical constraints and domain knowledge into data-driven models is vital for enhancing interpretability, improving generalization, and reducing the data requirements of purely data-driven, black-box ML models.
- The field is undergoing a paradigm shift from purely phenomenological, expert-driven constitutive modeling to more automated, data-driven approaches that can distill complex material behaviors from large datasets.
- Effective data sampling and experimental design are critical for the efficiency and reliability of data-driven models, particularly in situations with limited data availability.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.