A machine learning framework for predicting the axial capacity of cold-formed steel face-to-face built-up channel sections at elevated temperatures

Abstract

The structural performance of cold-formed steel (CFS) built-up sections under fire is significant in structural engineering. This study presents an efficient and reliable machine learning (ML) framework for predicting the axial capacity of CFS face-to-face (FTF) built-up channel sections at elevated temperatures. Using finite element (FE) datasets from the existing literature, the axial capacity of CFS-FTF built-up channel sections at elevated temperatures is predicted using different ML models, comprising Support Vector Machine, Radial Basis Neural Network, Artificial Neural Network, Extreme Learning Machine, Convolutional Neural Network and Boosting. Finally, the Boosting prediction was interpreted by the SHapley Additive exPlanations method to determine the significance of input features. The ML-based framework proposed in this study could offer a promising alternative for researchers and engineers to efficiently and effectively predict the axial capacity of CFS-FTF built-up channel sections at elevated temperatures.