3D heat diffusion simulation using 3D and 1D heat sources – Temperature and phase contrast results for defect detection using IRT

Abstract

Infrared thermography (IRT) has proven to be a powerful non-destructive technique and it has been successfully used for detecting defects in a wide range of applications and sectors. Numerical modeling of heat transfer and diffusion in the presence of defects combined with experimental IRT results may be used both to detect and characterize existing defects with specific geometries and depths. However, the three-dimensional (3D) nature of defects combined with the need to simulate heat transfer and diffusion phenomena in transient regime often presents many challenges for researchers. The study presented in this paper is motivated by such difficulties and is intended to contribute to the interpretation of quantitative data results collected in experimental defect detection IRT tests and help with the definition of IRT experimental set-up parameters.In this study, 3D heat diffusion by conduction in the proximity of a 3D defect is modeled using a boundary element method (BEM) formulated in the frequency domain. The defect is a crack lodged in an unbounded solid medium with null thickness. In order to overcome difficulties that occur in the presence of null thickness elements, the BEM formulation was written in terms of normal-derivative integral equations (TBEM) and known analytical solutions were used to solve the resulting hypersingular integrals. The focus of this paper is to study the influence using either a 3D (point) energy source or a 1D (planar) energy source in heat diffusion simulations performed for IRT defect detection studies. Heat field and thermal wave phase results were computed in the presence of a defect and for when there is no defect present and a comparative analysis of the results obtained for a point and a planar source was carried out. In order to contribute to defects characterization studies, the influence of the crack's characteristics such as its size, shape and placement (depth and position) was analyzed using a phase contrast approach. Other features that may be relevant in IRT experiments, such as the nature of the stimulus provided and its distance from the surface were also studied. The major findings achieved from varying those parameters are presented in the conclusions.