Inverse methodology for identification the thermal diffusivity and subsurface defect of CFRP composite by lock-in thermographic phase (LITP) profile reconstruction

An inverse analysis methodology is developed to simultaneously identify the thermal diffusivities and subsurface defect of carbon fiber reinforced polymer (CFRP) laminate through lock-in thermographic phase (LITP) profile reconstruction. A hybrid algorithm that combines simulation annealing algorithm (SA) and Nelder–Mead simplex search method (NM) is proposed to minimize the objective function constructed by LITP profile for determination of the thermal diffusivities (αx and αz) and subsurface defect characteristics (size D and depth HD). For the hybrid method, SA is used to find a better feasible initial-guessed solution for NM. Numerical examples show that LITP has the merit of identifying the thermal diffusivities of CFRP laminate by inverse analysis. In comparison with SA and NM, respectively, the hybrid method can be converged to yield good estimates faster and more efficiently. Experimental and simulation LITP of defective regions are used to determine the thermal diffusivities and the size and depth of subsurface defect. The results show that the estimated thermal diffusivities and the size and depth of defect are in good agreement with the reference values, and the relative deviations are less than 5%. Nevertheless, this approach can only be applied for the determination of thermal diffusivities of CFRP laminate with subsurface defect.