Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Lightweight Large Language Model-Based Multi-Agent System for 2D Frame Structural Analysis

Ziheng Geng, Jiachen Liu, Ran Cao, Lu Cheng, Haifeng Wang, Minghui Cheng

TL;DR

This paper tackles the challenge of automating 2D frame finite element modeling with LLMs by proposing a lightweight, multi-agent system built around the Llama-3.3 70B Instruct model. The architecture decouples geometric reasoning from code generation across five specialized agents, enabling end-to-end generation and execution of OpenSeesPy models with visualization support via OpsVis. Benchmark results on 20 frame problems show accuracy above 80% in most cases and clear improvements over Gemini 2.5 Pro and ChatGPT-4o, with ablation studies underscoring the importance of task decomposition. The approach delivers a practical, cost-effective pathway for automated structural analysis that can run on standard workstations, with strong implications for accelerating engineering workflows and enabling broader accessibility to FEM tooling.

Abstract

Large language models (LLMs) have recently been used to empower autonomous agents in engineering, significantly improving automation and efficiency in labor-intensive workflows. However, their potential remains underexplored in structural engineering, particularly for finite element modeling tasks requiring geometric modeling, complex reasoning, and domain knowledge. To bridge this gap, this paper develops a LLM-based multi-agent system to automate finite element modeling of 2D frames. The system decomposes structural analysis into subtasks, each managed by a specialized agent powered by the lightweight Llama-3.3 70B Instruct model. The workflow begins with a Problem Analysis Agent, which extracts geometry, boundary, and material parameters from the user input. Next, a Geometry Agent incrementally derives node coordinates and element connectivity by applying expert-defined rules. These structured outputs are converted into executable OpenSeesPy code by a Translation Agent and refined by a Model Validation Agent through consistency checks. Then, a Load Agent applies load conditions into the assembled structural model. Experimental evaluations on 20 benchmark problems demonstrate that the system achieves accuracy over 80% in most cases across 10 repeated trials, outperforming Gemini-2.5 Pro and ChatGPT-4o models.

A Lightweight Large Language Model-Based Multi-Agent System for 2D Frame Structural Analysis

TL;DR

This paper tackles the challenge of automating 2D frame finite element modeling with LLMs by proposing a lightweight, multi-agent system built around the Llama-3.3 70B Instruct model. The architecture decouples geometric reasoning from code generation across five specialized agents, enabling end-to-end generation and execution of OpenSeesPy models with visualization support via OpsVis. Benchmark results on 20 frame problems show accuracy above 80% in most cases and clear improvements over Gemini 2.5 Pro and ChatGPT-4o, with ablation studies underscoring the importance of task decomposition. The approach delivers a practical, cost-effective pathway for automated structural analysis that can run on standard workstations, with strong implications for accelerating engineering workflows and enabling broader accessibility to FEM tooling.

Abstract

Large language models (LLMs) have recently been used to empower autonomous agents in engineering, significantly improving automation and efficiency in labor-intensive workflows. However, their potential remains underexplored in structural engineering, particularly for finite element modeling tasks requiring geometric modeling, complex reasoning, and domain knowledge. To bridge this gap, this paper develops a LLM-based multi-agent system to automate finite element modeling of 2D frames. The system decomposes structural analysis into subtasks, each managed by a specialized agent powered by the lightweight Llama-3.3 70B Instruct model. The workflow begins with a Problem Analysis Agent, which extracts geometry, boundary, and material parameters from the user input. Next, a Geometry Agent incrementally derives node coordinates and element connectivity by applying expert-defined rules. These structured outputs are converted into executable OpenSeesPy code by a Translation Agent and refined by a Model Validation Agent through consistency checks. Then, a Load Agent applies load conditions into the assembled structural model. Experimental evaluations on 20 benchmark problems demonstrate that the system achieves accuracy over 80% in most cases across 10 repeated trials, outperforming Gemini-2.5 Pro and ChatGPT-4o models.

Paper Structure

This paper contains 14 sections, 16 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Benchmark
  3. Dataset overview
  4. Performance evaluation of Llama-3.3 70B Instruct model
  5. A large language model-based multi-agent system for frame structural analysis
  6. Overall system design
  7. Architecture of the geometry agent
  8. Architecture of the model validation agent
  9. Results and discussion
  10. Performance of the LLM-based multi-agent system
  11. Comparison with state-of-the-art LLMs
  12. Ablation experiment: Validity of task decomposition
  13. Runtime and costs
  14. Summary and conclusions

Figures (16)

  • Figure 1: A benchmark dataset comprising twenty representative frame structural analysis problems.
  • Figure 2: Workflow to assess the performance of Llama model in 2D frame structural analysis.
  • Figure 3: Syntax errors produced by the Llama model in frame structural analysis.
  • Figure 4: Evaluation of Llama model’s spatial reasoning capability through frame geometry generation.
  • Figure 5: System design of the LLM-based multi-agent framework for automated frame structural analysis.
  • ...and 11 more figures