Analysis and prediction methods for the static properties of reinforced concrete beams under fire

Abstract

In order to study the thermal properties and static performance of reinforced concrete (RC) beams exposed to fire and reveal the relationship between fire exposure time, concrete spalling, bond slip and mechanical properties, the numerical simulation, theoretical analysis and machine learning methods were adopted to study the effects of high temperature on thermal and static mechanical properties of RC beams and the prediction methods of residual bearing capacity were proposed in this paper. Fire tests and static load tests after natural cooling were carried out on 4 RC simply supported beams. The refined numerical analysis model of RC beam was established and its validity was verified by the fire test. The effects of fire exposure time and characteristic parameters of concrete spalling on temperature fields and interfacial bond properties were simulated and analyzed. The residual flexural capacity were calculated and compared with measured and simplified theoretical values. Finally fire exposure time, concrete spalling depth and local spalling area ratio were selected as variables to design 250 groups of sample data, and sample database was established based on the refined numerical analysis model. Several methods for predicting residual flexural capacity after fire were proposed. Their accuracy and reliability were compared and analyzed, which can provide a method reference for safety estimation and analysis of post-disaster engineering.