Speckle Pattern Inversion in High Temperature DIC Measurement

Abstract

During hot fire rocket engine testing, non-contacting measurements are superior to bonded gauges because they are immune to burning, shaking loose, or damage due to the harsh testing conditions. Additionally, when compared to instruments which register at single points, Digital Image Correlation (DIC) has the added benefit in that it collects full-field displacement and strain maps over the duration of the test. However, for certain materials and paints under some circumstances of temperature and camera sensitivity, portions of the speckle pattern which were darker at room temperature may emit more light compared to the initially lighter portions of the pattern, resulting in a high temperature pattern which is inverted in comparison with that at room temperature. To address this inversion, a post-processing method is introduced wherein an inverted image containing only emitted light is subtracted from an image containing both emitted and reflected light, thereby generating an un-inverted image. The artificial high temperature image is subsequently correlated against the room temperature image to obtain full-field strains. The subtraction technique is then validated using optical bandpass filters to prevent significant amounts of emitted light from reaching the camera sensor. The two methods are mapped onto common coordinates and shown to produce comparable results. The subtraction method sufficiently mitigates speckle pattern inversion, but its key drawback is that it only works when there is negligible displacement between the subtracted images (i.e. quasi-static loading). It is therefore preferable to eliminate inversion from reaching the camera in the first place by using optical bandpass filters.