Smoothing of Digital Photoelastic Data Using Robust Outlier Algorithm

Abstract

Photoelasticity is an optical technique for experimental stress analysis. It is widely used for 2-D and 3-D analysis of components for getting the information of principal stress difference (isochromatics) and principal stress direction (isoclinics) at every point in the domain. With the advent of digital computers, recording of images as intensity data became easier and a separate branch of photoelasticity namely digital photoelasticity came into existence [1]. In digital photoelasticity intensity information of the captured image is used for evaluating the isoclinic and isochromatic parameters. Thus, in principle one gets values of isoclinic (gq) and isochromatic (N) for the whole-field in the form of phase map [1]. Phase shifting / polarization stepping techniques are widely used in digital photoelasticity for getting the phase maps [1]. But these phase maps are wrapped and has to be unwrapped for getting continuous phase values. Both isoclinic and isochromatic phase maps have interdependence and this interaction affect their evaluation. For stress separation studies, one requires both isochromatics and isoclinics accurately free of any kinks in the domain.