DAMAGE DETECTION IN CANTILEVER BEAMS USING ARTIFICIAL NEURAL NETWORKS

摘要

In Cantilever Beams when dynamic loading is applied that is varying the load with respect to time, which may result in cracks. It is a challenge to know presence of cracks. Hence the varying frequencies are obtained from ANSYS. We use these results to train a Neural Network and find the presence of cracks. The presences of damages change the physical characteristics of a structure which in turn alter its dynamic response characteristics. Therefore there is need to understand dynamics of damaged structures. Damage depth and location are the main parameters for the vibration analysis. The application of these beams is in gas turbine blades but it is a tough task to create blade and then to do analysis so we will consider a cantilever beam which resembles the turbine blade and do analysis by considering some notches and then find the varying natural frequencies.