Modeling Temperature of Fire-Damaged Reinforced Concrete Buildings Based on Nondestructive Testing and Gene Algorithm Techniques

It is a daunting task to ascertain the level of deterioration in structural concrete and embedded reinforcing steel after fire damage. Thus, this study focuses on modelling the temperature of fire-damaged reinforced concrete buildings based on nondestructive and Gene algorithm techniques. The study employed nondestructive data (rebound hammer number (RH) and ultrasonic pulse velocity (UPV)) from laboratory simulation tests on concrete elements and field forensic investigation of fire-damaged buildings to develop mathematical models. Other data used for the model was the thickness of the member. The input parameters for the model were: RH, UPV, and breadth of the member, while the temperature was the output data. The study proposed a model based on genetic expression programming (GEP), with a correlation coefficient of 0.9895. The developed model generalised the data correctly with a high degree of accuracy; it would be a useful tool for the field assessment of fire-damaged reinforced concrete buildings.