Session: 12-21-02: General: Mechanics of Solids, Structures and Fluids

Paper Number: 100247

100247 - Correcting for Dic Speckle Pattern Inversion at High Temperature Using Color Cameras 

When performing mechanical characterization at extreme temperatures (e.g. when characterizing materials for rocket nozzles), strain gages are too easily damaged and are unable to monitor strain for the full duration of tests. To avoid this environmental challenge, Digital Image Correlation (DIC) has the advantage that it is non-contacting and is thus not vulnerable to the same damage as strain gauges, while also collecting data in full field resolution. However, for some combinations of paint and substrate materials at extreme temperatures, a phenomenon known as speckle pattern inversion can occur in which portions of the speckle pattern which appeared darker at room temperature appear brighter at high temperature, and vice versa. This occurrence is understood to be caused by differences in emissivity between the paint and the specimen, and the resulting inverted pattern is understood to be a superposition of reflected and emitted light. In experiments where data is collected over large temperature ranges, speckle inversion can prevent DIC algorithms from correlating properly between low and high temperature images.

 

In this work, we present a new method to eliminate speckle pattern inversion by using blue lights and a color camera. The new technique is compared against two others in published literature. In Method A, two images are recorded at high temperature: one with light sources turned on, which contains both the supplied light reflected by the specimen plus additional light emitted at high temperature. The second image is recorded at high temperature with the light sources turned off, which contains only the emitted light. By subtracting the two images, Method A produces an artificial third image which contains only reflected light, and can therefore be correlated against room temperature images which also contain only emitted light. However, one key drawback of Method A is that the two images must be recorded at different times (one with lights on, the other with lights off) and nominally no motion between them, which limits Method A to relatively slow tests. In Method B, the camera is outfitted with a blue optical bandpass filter, which is well-known to screen out light emitted by the specimen to prevent it from reaching the camera. Since only the emitted light is inverted, this effectively screens out the speckle pattern inversion. The drawback of method B is that, relative to images collected without a filter, the addition of the filter can introduce thick-glass distortions which may result in artificial strains.

 

In the newly presented Method C, unfiltered images collected and processed using blue lights and a color camera, which behaves effectively as 3 overlapping arrays of Red, Green, and Blue pixel sensors. Thus, although the high temperature image from the blue channel contains both emitted and reflected light, the high temperature image from the green channel contains only emitted light and the room temperature image from the blue channel contains only reflected light. The high temperature image is then artificially rescaled such that its distribution of pixel intensities better matches the room temperature blue image, thereby creating an artificial blue image at high temperature which contains only reflected light. The new method is favorable to Method A because it does not require subtracting images and can be applied to faster experiments, and is favorable to Method B because it does not require a bandpass filter.

Presenting Author: Ryan Berke Utah State University

Presenting Author Biography: Ryan Berke is an Associate Professor of Mechanical Engineering at Utah State University. He earned his BS from the University of Maryland and his PhD from the Ohio State University, both in Mechanical Engineering. He has also worked as a postdoctoral researcher at University of Illinois and a summer faculty fellow at AFRL, both in Aerospace Engineering. His research focus is experimental solid mechanics for extreme temperature environments, specializing in full-field camera-based measurements, with applications geared towards the aerospace and nuclear industries.

Authors:

Ryan Berke Utah State University
Lindsey Rowley Utah State University
Prasenjit Dewanjee Utah State University
Thinh Thai Van Lang University