Scholarlink

<p>Steel-reinforced concrete-filled steel tubular (SRCFT) stub columns, a promising composite material, show great potential for practical engineering applications such as piers or columns. Understanding their compressive behavior and accurately predicting their strength is crucial for effective engineering design. Previous studies focused onload-carrying capacity using longitudinal reinforcement or specific section shapes like H-sections or X-sections. In practical situations, columns often require different section shapes or significant reinforcement, leading to more complex failure mechanisms. To address this, a combination of finite element models (FEMs) and 5 machine learning techniques (RF, LightGBM, AdaBoost, CatBoost, and XGBoost) were employed to accurately predict load-bearing capacity under axial load. A database of 66 experimental tests and 134 FEMs simulations was generated and analyzed. Eleven input parameters were chosen for developing ML-based models to predict the ultimate load of circular SRCFST columns under axial load. The models were trained and tested, showing improved accuracy in strength prediction. An open-source GUI based on the XGBoost model was developed to provide design engineers with precise estimates of load-bearing capacity and aid in decision-making for mix proportion under various conditions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Axial compressive behavior of reinforced concrete-filled circular steel tubular columns: finite element and machine learning modelling

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月17日

查看全文
  • Sisi Xie,
  • Haytham F. Isleem,
  • Walaa J. K. Almoghayer,
  • Ahmed Ahmed Ibrahim

摘要

Steel-reinforced concrete-filled steel tubular (SRCFT) stub columns, a promising composite material, show great potential for practical engineering applications such as piers or columns. Understanding their compressive behavior and accurately predicting their strength is crucial for effective engineering design. Previous studies focused onload-carrying capacity using longitudinal reinforcement or specific section shapes like H-sections or X-sections. In practical situations, columns often require different section shapes or significant reinforcement, leading to more complex failure mechanisms. To address this, a combination of finite element models (FEMs) and 5 machine learning techniques (RF, LightGBM, AdaBoost, CatBoost, and XGBoost) were employed to accurately predict load-bearing capacity under axial load. A database of 66 experimental tests and 134 FEMs simulations was generated and analyzed. Eleven input parameters were chosen for developing ML-based models to predict the ultimate load of circular SRCFST columns under axial load. The models were trained and tested, showing improved accuracy in strength prediction. An open-source GUI based on the XGBoost model was developed to provide design engineers with precise estimates of load-bearing capacity and aid in decision-making for mix proportion under various conditions.