A Hybridization of CSA DEA Approach for Detection of Multiple Transverse Crack Rotor Shaft Rotating in Fluid Environment under Axial and Bending Loading

Abstract

Rotor shaft cracks can be acknowledged as a substantial issue regulating the reliable and safe machines operation. When the crack has grown and is not discovered, unexpected catastrophe may happen, ultimately resulting the plant being shut down with numerous losses attached. When the rotor shaft is revolving in a viscous fluid environment, it becomes exceedingly challenging to analyze the size and position of the crack. In the present research work, a novel hybridized technique a clonal selection algorithm (CSA) with a differential evolution algorithm (DEA) is proposed for early detection of multiple transverse cracks and its position in the rotor shaft rotating under altered viscous fluid environment in an acceptable amount of time limit. Relative natural frequency at x-axis, relative natural frequency at y-axis, relative amplitude of vibration at x-axis, relative amplitude at y-axis and viscosity of fluid are use as input parameter and relative crack depth and crack location are used as a output parameter in the hybrid technique. To compute the amplitude and natural frequency of the cracked rotor shaft rotating under different fluid environment utilizing the stiffness matrices of crack element. An external force of fluid is computed by Navier–Stokes equation. Theoretical evaluation was executed using Matlab. To authenticate the theoretical and experimental value of natural frequency and amplitude, perform the finite element analysis by using ANSYS. For training the hybrid system, the amplitude and natural frequencies are found out using theoretical, experimental and finite element analysis for different crack depth and positions. The test results of the recommended hybrid technique are compared with finite element analysis and experimental analysis for validation, and satisfactory outcomes have been observed. Therefore the recommended hybridized CSA–DEA technique would establish an effective tool for real-time crack detection in rotor.