LLM-driven design of physics-constrained constitutive models: two agents are better than one
arXiv:2605.23754v1 Announce Type: new Abstract: Developing constitutive models that capture how materials deform under load traditionally requires years of specialized expertise in continuum mechanics, machine learning, and scientific programming. Large language models (LLMs) have recently been shown to lower this barrier by generating constitutive models on demand, but existing single-agent pipelines lack systematic checks that the resulting models respect fundamental physical laws. To close this gap, we introduce the first multi-agent LLM-driven approach for constitutive model generation: a
The rapid advancement in Large Language Models (LLMs) combined with increasing computational power makes this multi-agent approach viable at this moment.
This development addresses a critical limitation of single-agent LLMs by ensuring that AI-generated models adhere to fundamental physical laws, accelerating materials science and engineering.
The process of developing complex physics-constrained constitutive models can be significantly streamlined and validated, reducing dependency on years of specialized human expertise.
- · Materials scientists
- · AI developers
- · Manufacturing sector
- · Engineering firms
- · Traditional constitutive model developers reliant on manual methods
Acceleration in the discovery and development of novel materials with precise properties.
Reduced time-to-market for products requiring advanced material science, enhancing competitive advantage.
Emergence of fully autonomous material design and optimization cycles, leading to unforeseen technological breakthroughs.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG