Automated counting of off-axis tunnelling cracks using digital image processing

An automated method for counting propagating matrix tunnelling cracks for use in mechanical testing of Glass Fibre Reinforced Plastic (GFRP) laminates under quasi-static and fatigue loading is presented. White light images are captured from specimens during the loading. The transmitted light is used to detect the cracks in the images, which are then processed to count the cracks as they develop and grow through the duration of the test. The reproducibility and accuracy of the image processing is demonstrated using simulated transverse crack densities and patterns. The methodology is demonstrated and validated experimentally using two different laminate stacking sequences of the type [0/-θ/0/ θ]_s. The results related to the crack density evolution are shown to be consistent with results from the literature.