Experimental and Artificial Neural Network modeling of Natural Frequency of Stepped Cantilever shaft

The natural frequency of aluminum cantilever stepped beam (two steps) was investigated experimentally and theoretically by modeling the experimental data using artificial neural network (ANN) for different values of small and large diameters and for different lengths of larger diameter step. Two hidden layers and different number of neurons in each hidden layer were employed with the ANN. Theoretical natural frequency results using two algorithm functions (trainlm) and (trainrp) of the ANN method were compared with the experimental solution. The results showed that there was an increase in the natural frequency with the increasing of the larger diameter length of the stepped shaft and a high performance of the ANN was found to predicate the experimental results.