|
Abstract:
The decorrelation effect of digital image correlation (DIC) can cause DIC calculation failure, and it is always been regarded as a defect of DIC, the problem seriously hinder the promotion and application of DIC in the field of fracture mechanics. Meanwhile, structures (such as steel structures) are prone to fatigue cracking under repeated loads. Fatigue crack measurement is very important for carrying out model test research and engineering problem analysis. However, the existing methods are not suitable for full-field dynamic fatigue crack measurement with high-precision. This research proposed a novel full-field dynamic fatigue crack measurement approach and its visualization by using the principle of DIC. The approach first constructs point-cloud data structure with topological relations and calculates the crack displacement fields for the captured digital images of cracks. The zero-mean normalized cross-correlation (ZNCC) criterion is employed to eliminate the vanishing points within cracked regions, and the discrete birth and death boundaries of cracks are extracted and interpolated by a presented "three-living point" algorithm. The least square method is finally utilized to convert the discrete crack boundaries into continuous crack boundaries, and as a result the dynamic varying process of crack length and width are automatically calculated. Numerical simulation and fatigue tests are carried out to verify the accuracy of fatigue crack measurement algorithm. Results show that the digital reconstruction errors of fatigue crack boundaries are within 0.5 pixel. The calculated errors of crack length and width are respectively 0.46 pixel and 0.08 pixel. Furthermore, refined measurement for the dynamic propagation process of cracks are successfully achieved in fatigue tests of welded steel joints. This research proves that, due to the advantages in accuracy, efficiency, and cost, the presented full-field dynamic fatigue crack measurement approach and its visualization using DIC technology is highly effective, and thus is applicable to laboratory measurement and engineering field testing.
|