Bhatt P P, Kodur V K R, Naser M Z
Walter P. Moore, Kansas City, MO, USA.
Department of Civil and Environmental Engineering Michigan State University, MI, USA.
Data Brief. 2024 Jan 5;52:110031. doi: 10.1016/j.dib.2024.110031. eCollection 2024 Feb.
Machine learning (ML) has emerged as an efficient and feasible technique for tackling engineering problems. Despite the numerous advantages, the implementation of ML for evaluating the fire resistance of structural members is relatively scarce, primarily due to the lack of a reliable database with a substantial number of data points. To address this knowledge gap, this paper presents a comprehensive database on the fire performance of fiber reinforced polymer (FRP) strengthened reinforced concrete (RC) beams. The database comprises over 21,000 experimental and numerical data points with varying parameters, including various geometric dimensions, FRP-strengthening levels, steel reinforcement ratio, insulation thickness and configuration, material properties, and applied load levels. The database can be implemented to train ML algorithms for developing autonomous models for predicting the fire resistance of FRP-strengthened concrete beams with varying parameters.
机器学习（ML）已成为解决工程问题的一种高效且可行的技术。尽管有诸多优点，但将机器学习用于评估结构构件的耐火性的应用相对较少，主要原因是缺乏一个包含大量数据点的可靠数据库。为了填补这一知识空白，本文提出了一个关于纤维增强聚合物（FRP）加固钢筋混凝土（RC）梁火灾性能的综合数据库。该数据库包含超过21000个具有不同参数的实验和数值数据点，这些参数包括各种几何尺寸、FRP加固水平、钢筋比率、隔热层厚度和配置、材料性能以及施加的荷载水平。该数据库可用于训练机器学习算法，以开发用于预测不同参数的FRP加固混凝土梁耐火性的自主模型。
