A developed hybrid experimental–analytical method for thermal stress analysis of a deep U-notched plate

Abstract

A general framework of hybridizing experimentally-recorded load-induced thermal information by means of thermoelastic stress analysis (TSA) of a loaded structure with the Michell solution of Airy stress function in isotropic linear elasticity is proposed for in-plane stresses determination. The capability of the proposed hybrid method in separating the TSA signals into individual stresses is demonstrated by stress-analyzing a deep U-notched aluminum plate without neither knowing the entire geometry and distant loading/boundary conditions nor using supplementary experimental information. Even though no experimental data were processed at, and adjacent to, the edges of the U-notched plate, the individual stresses are determined throughout the surface of the plate. Prior processing actual experimental TSA data, the accuracy and the stability of the proposed hybrid method through adding artificial noise scatters in the processed simulated data with the number of retained terms in the generalized Airy stress expressions are firstly investigated. The effects of the superimposed noise scatters as well as the number of employed data in the evaluated stresses are also assessed. Finally, the reliability of the determined hybrid-experimental stresses is supported through a comparison with finite-element predictions and strain gauge measurements. The ease of implementation, the accuracy and the precision of the evaluated stress, and the versatility to the type of employed data make the proposed hybrid method more attractive and superior than other hybrid methods.