Photoelastic Fringe Pattern Analysis by Image Processing

This work presents a strategy for automatic analysis of photoelastic fringe patterns. The study is focused on the analysis of color type photoelastic fringes. By capturing and analyzing color fringes automatically, the difference between principal stresses of a sample are quantitatively determined. Among many techniques available in literature, the color photoelasticity is chosen as the basis for the strategy since it minimazes the operator’s intervention. One important part of this technique is the calibration process, i.e., a process in which parameters of the relation between the fringe order and the fringe color are determined. This work presents a calibration based on two different filtering processes: average and harmonic. Using an experimental setup composed by a polariscope, a load system and a CCD type camera, the developed strategy is evaluated. Results of experiments are compared with the analytical solution of a disc under diametric compression for both calibration filtering processes. Despite some deviations, related to the material used in the sample and non-uniformity in the applied load, the proposed strategy showed to be efficient to evaluate the fringe order in a range of 0.5 to 3.0.