The Practicality of Quality Assessment Metrics for Millimetre‐Scale Digital Image Correlation Speckle Patterns

ABSTRACT Digital image correlation (DIC), often used in multidisciplinary applications by nonexperts, relies heavily on the quality of a speckle pattern for accurate deformation tracking. Numerous metrics have been developed to assess speckle pattern quality and guide decision‐making in DIC postprocessing. This investigation considered the applicability of these metrics in the selection of an optimal speckling method for two‐dimensional DIC strain measurements, using the standard Brazilian tensile strength rock mechanics laboratory test as a case study. Four speckle patterning methods—spray paint, airbrush, stamp and laser engraving—were optimized to produce patterns of good visual quality for this specific DIC setup. Twenty specimens, five for each method, were prepared, and their speckle patterns were analysed using 10 published metrics. Each pattern was also deformed numerically to quantify the associated systematic and random errors. Overall, none of the 10 metrics demonstrated substantial agreement with the numerical experiment errors due to their failure to account for the confounding influence of pattern characteristics. Consequently, these tools are limited in their usefulness to evaluate the accuracy and reproducibility of speckle patterns with similar characteristics. Numerical simulation of pattern deformation is recommended as the most accurate resource for nonexpert DIC users to assess reproducibility and select optimal patterning methods in practical applications.