Smoothing of Digital Photoelastic Data Using Robust Outlier Algorithm

Abstract

Abstract:  With advancements in digital image processing and data acquisition, a separate branch of photoelasticity namely digital photoelasticity came into existence. Here, intensity information of the acquired image is used for the evaluation of whole-field photoelastic parameters. Digital photoelasticity provides only wrapped phasemaps of isoclinics and isochromatics and they have to be unwrapped in different ways for getting the continuous-phase values. In the case of the isochromatic phasemap, ambiguity removal prior to unwrapping is essential. In this paper, a 10-step phase-shifting methodology is proposed and a new strategy for obtaining the isochromatic phasemap free of ambiguity is demonstrated. Isoclinic unwrapping is performed by a new adaptive quality guided algorithm. Adaptive in the sense that isoclinic phase unwrapping is done autonomously even in the presence of isotropic points/π jumps occurring in the isoclinic phasemap. The isochromatic phasemap is also unwrapped using the quality guided path follower. The methodology is validated for the problem of a ring under diametral compression and later shown for three other models which have complex stress fields. Wherever possible, the parameters obtained by the new methodology are compared with analytical or numerical methods and the comparison is quite good.