An Adaptive Window Shape-Based DIC Method for Large Torsional Deformation Measurement

Torsional deformation measurement is used to measure the resistance of materials to torque and is one of the basic tests for the mechanical properties of materials. The measurement of large torsional deformation based on the digital image correlation (DIC) method has the advantages of full-field measurement, noncontact measurement, and easy-to-implement, but there are also problems, such as decorrelation and out-of-view. To tackle these problems, we propose an adaptive window shape-based DIC method and a reverse matching method, based on which a large torsional deformation measurement scheme is formed. First, taking advantage of the temporal continuity of object deformation, the window shape adaptively changes with the torsion of the specimen to avoid decorrelation. To solve the problem of tracking points moving out of the camera’s field of view, the deformed subregions are reversely matched in the reference image set, which is preshot from different directions. By this, the proposed method greatly improves the measurement range, and as a tradeoff, the absolute accuracy is acceptably sacrificed. Torsion tests of cylindrical low carbon steel specimen show that the proposed method can measure large torsional deformation objects with a shear angle of 74°, and the relative error is about 0.43%.