Data-driven material models benchmarked on classic Treloar dataset

A new paper benchmarks several data-driven constitutive modeling frameworks for hyperelasticity using the classic Treloar dataset. The study compares Constitutive Artificial Neural Networks, Physics-Augmented Neural Networks, Material Fingerprinting, and Efficient Unsupervised Constitutive Law Identification & Discovery. While all methods demonstrated strong fitting performance, the research highlights their respective strengths, limitations, and trade-offs between predictive accuracy and model complexity, offering practical guidance for their application. AI

IMPACT Provides practical guidance for selecting and implementing data-driven constitutive models in material science research.

RANK_REASON The cluster contains an academic paper detailing a benchmark comparison of different machine learning models for material science. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Constitutive artificial neural networks
• Efficient Unsupervised Constitutive Law Identification & Discovery
• Material Fingerprinting: Understanding how differences in geology impact metallurgical plant performance
• Physics-Augmented Neural Networks
• Treloar

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →