Application of Entropy in Crack Identification at Welding Joint With a New Smart Coating Sensor

An identification module is designed and studied to detect and evaluate the cracks at the welding joint area using a new smart coating sensor and entropy measurement. A new piezoelectric composite coating is applied at a welding joint to possibly charge the wireless data transmission module as an energy harvester. It also sends warning and dynamic signals for crack evaluation when the crack damage occurs. More specifically, entropy calculation is introduced to quantify the weak perturbations, which is caused by the material nonlinearity and crack breathing at the crack tip and hidden in the signal. In this paper, a finite element model (FEM) of a welded beam experiencing dynamic base motion is established as an example. The effects of material nonlinearity and crack breathing on structural dynamics response are simulated by creating nonlinear material property around the crack area and contact pair of crack walls, respectively. After obtaining the time domain vibration signal, crack severity is quantified using Sample Entropy. It is concluded that, even at very early stages of 5% of the beam thickness for the crack depth, the entropy variation is significant for a damaged beam compared with the healthy one.