2025-09-22 ペンシルベニア州立大学(PennState)
＜関連情報＞
- https://www.psu.edu/news/engineering/story/doing-lot-little-new-ai-system-helps-explain-laser-welding-defects
- https://www.sciencedirect.com/science/article/pii/S0890695525000756
大規模言語モデルを用いた高速レーザー溶接の物理方程式の導出 Derivation of physical equations for high-speed laser welding using large language models
Kyubok Lee, Zhengxiao Yu, Zen-Hao Lai, Peihao Geng, Teresa J. Rinker, Changbai Tan, Blair Carlson, Siguang Xu, Jingjing Li
International Journal of Machine Tools and Manufacture Available online: 20 August 2025
DOI:https://doi.org/10.1016/j.ijmachtools.2025.104320
Graphical abstract
Highlights
- A novel framework bridging sparse data and text to infer governing physical equations.
- Reveals humping as an imbalance where inertial effects dominate over capillary stabilization.
- Enables physical law discovery by integrating text and data with LLMs.
- Identify simple and interpretable equations for complex laser welding dynamics.
Abstract
It is challenging to formulate complex physical phenomena that occur in a manufacturing process, particularly when the available data are limited, rendering conventional data-driven approaches ineffective. This study aims to predict humping onset in high-speed laser welding by introducing a novel framework, namely text-to-equations generative pre-trained transformer (T2EGPT). This method leverages the capabilities of large language models (LLMs), in combination with sparse experimental data and enriched literature data, to derive an interpretable and generalizable equation for predicting humping initiation. By capturing key correlations among physical parameters, T2EGPT generates a compact and dimensionless expression that accurately predicts hump formation. The equation reveals that humping arises from the interplay between inertia-driven backward melt flow and capillary-driven surface stabilization, where inertial forces drive molten metal backward and capillary forces resist surface deformation. Compared to traditional data-driven models, T2EGPT demonstrates enhanced predictive accuracy and cross-material transferability. More broadly, this study highlights the potential of LLMs to integrate textual information with data-driven discovery, enabling the extraction of physical laws in data-scarce scientific domains.
