Assessment of core strength of concrete by artificial neural networks

The proposed work deals with the use of Ultrasonic pulse velocity technique as an alternative method to identify compressive strength core concrete. non-destructive without causing damages structure is tedious interpretation results influenced by various factors. Hence, empirical relationship developed using artificial neural network model for creating a regression between and concrete specimens. Tests were conducted on reinforced cylinders at orientation angles (0°, 45°, 90°). tests based design experiment Box-Behnken model. These trained Levenberg-Marquardt back propagation hidden layers. Results indicate that prediction grade mixes nearer two-level factorial R2 = 0.897, sum squared error found be 0.9968.