Shear Failure Capacity Prediction of Concrete Beam–Column Joints in Terms of ANFIS and GMDH

Abstract

Vulnerability assessment of structures in an earthquake is one of the most important topics in structural engineering. HAZUS instruction is the code widely used for assessment of structures for satisfied damage based on the interstory drift. In a structure, there are several parameters, such as forces and responses in elements, which cause damage to the structure simultaneity, and therefore, the use of only one parameter, such as interstory drift, could be unrealistic. In this paper, reinforced concrete (RC) beam–column joints are studied, with the aim of determining the maximum shear capacity of RC joints as a key parameter in the damage of RC structures. For this purpose, two strong approaches, including group method of data handling (GMDH) and adaptive neuro-fuzzy inference system (ANFIS), were used. The selected models were created based on a large experimental database. The results indicated that the considered methods are capable of determining the shear capacity of RC joints with high accuracy.