BOND STRENGTH PREDICTION OF UHPC–NSC INTERFACE

The application of ultra-high-performance concrete (UHPC) on top normal-strength (NSC) is a practical rehabilitation approach to maintaining degraded and damaged members. However, successful repair operation consequent adequate performance are very much dependent the ability interface between UHPC NSC present superior bonding under various surface conditions. Consequently, predicting strength bond at joining existing newly placed overlaying - with sufficient certainty has become vital required step in assessing rehabilitated structural elements. In this work, Artificial Neural Network (ANN as well Gene Expression Programming (GEP methods utilized predict substrate using comprehensive database set consisting 264 experimental data points gathered from literature. A parametric ANN analysis performed examine assess effect each parameter interfacial strength. following five factors identified key parameters through GEP analyses: curing method, age UHPC, compressive NSC, treatment, moisture developed models have good accuracy closer predictions slant shear test splitting tensile root mean square error (RMSE) values 5.0, 4.3, coefficient variation (COV) 37%, 24%, respectively.