🤖 AI Summary
Traditional mechanical structural optimization relies heavily on expert knowledge and computationally expensive finite element analysis (FEM), while existing machine learning approaches suffer from poor generalizability and dependence on large-scale labeled datasets.
Method: We propose the first zero-shot LLM-FEM collaborative framework for autonomous 2D truss generation, multi-objective evaluation, and iterative optimization—requiring no fine-tuning. A general-purpose large language model (GPT-4.1 and its lightweight variant) serves as the natural language reasoning engine, tightly coupled with a physics-based FEM module for structural validation. Convergence is guided via temperature-controlled sampling and multi-objective feedback-driven prompt engineering.
Contribution/Results: Compared to NSGA-II, our method drastically reduces FEM evaluations and accelerates convergence. GPT-4.1-mini (temperature=0.5) achieves the highest constraint satisfaction rate and optimization step efficiency. This work establishes, for the first time, the feasibility of LLMs as cross-task general-purpose structural optimizers.
📝 Abstract
Conventional mechanical design follows an iterative process in which initial concepts are refined through cycles of expert assessment and resource-intensive Finite Element Method (FEM) analysis to meet performance goals. While machine learning models have been developed to assist in parts of this process, they typically require large datasets, extensive training, and are often tailored to specific tasks, limiting their generalizability. To address these limitations, we propose a framework that leverages a pretrained Large Language Model (LLM) in conjunction with an FEM module to autonomously generate, evaluate, and refine structural designs based on performance specifications and numerical feedback. The LLM operates without domain-specific fine-tuning, using general reasoning to propose design candidates, interpret FEM-derived performance metrics, and apply structurally sound modifications. Using 2D truss structures as a testbed, we show that the LLM can effectively navigate highly discrete and multi-faceted design spaces, balance competing objectives, and identify convergence when further optimization yields diminishing returns. Compared to Non-dominated Sorting Genetic Algorithm II (NSGA-II), our method achieves faster convergence and fewer FEM evaluations. Experiments with varying temperature settings (0.5, 1.0, 1.2) and model sizes (GPT-4.1 and GPT-4.1-mini) indicate that smaller models yield higher constraint satisfaction with fewer steps, while lower temperatures enhance design consistency. These results establish LLMs as a promising new class of reasoning-based, natural language-driven optimizers for autonomous design and iterative structural refinement.