Determining heat intensity of a fatigue crack from measured surface temperature for vibrothermography

Abstract

We propose a simulation -assisted linear-inversion method to estimate heat intensity generated at a fatigue crack during vibrothermographic nondestructive evaluation. In vibrothermography, mechanical excitation is applied to a test specimen, and the contacting crack faces generate heat which flows to the surface and can be detected by an infrared camera. The heat intensity generated at the crack is not a directly measurable quantity, and instead has to be estimated by inverting the measured crack surface temperature. This inversion is an ill-posed problem because of the inherently diffusive nature of heat flow. The proposed method in this paper attempts to estimate the crack heat intensity by minimizing the squared error between numerically modeled crack heating and the experimentally measured surface temperature of the crack. We demonstrate that with reasonable assumptions, the effect of diffusion on the heat intensity can be overcome, and it is possible to estimate the crack heat intensity. Furthermore, we reconstruct the surface temperature on two different material systems and quantify the reconstruction accuracy of the least-squares inversion algorithm.

Special Feature on Nondestructive Evaluation of Materials

“…An algorithm was developed to estimate the total heat intensity at a fatigue crack using vibrothermography by Vaddi et al [4].…”

Estimation of the area and depth of buried heat sources: an approach to characterize defects in inductive and ultrasound excited thermography

“…The basic strategy to deal with the loss of information [14][15] consists in making use of prior information about the solution (heat source). This can be knowledge of the orientation (vertical fatigue crack 16 , horizontal delamination 17 ), knowledge of the geometrical shape (half-penny crack 18 ), or knowledge about a certain property of the solution (the heat sources are scarce [19][20] , the edges are sharp 21 , etc). The amount of prior information available determines the difficulty in finding the solution: for instance, if the particular geometry of the heat source is known, the problem is reduced to a parameter estimation, with a reduced number of unknowns (the characteristic parameters of the geometry).…”