Quantifying the sequential fine cracks of strain-hardening cementitious composites (SHCC): An automatic approach for temporal and spatial assessment.

Strain-hardening cementitious composites (SHCC) are promising construction materials that enhance structural durability by forming multiple fine cracks under loading. Compared to conventional concrete, the cracks in SHCC are denser and tighter, making their quantification challenging. To thoroughly analyze and model the cracking behavior of SHCC, automatic algorithms with high accuracy are desired. In this paper, a workflow based on deep learning algorithms and 3D connectivity analysis for characterizing the crack formation and development process in SHCC under tensile loading is proposed. Temporal and spatial crack information, such as the crack number and individual crack width development during the loading process as well as the cracking strength distribution along the tensile specimens, can be obtained with minimal effort. The established workflow can provide more information for evaluating the overall cracking behavior of SHCC with high efficiency, to facilitate multiscale modeling and micromechanics-based design of the material. • A workflow for tracing individual tensile crack development in SHCC is established. • Temporal and spatial crack information analysis in SHCC tensile specimens is demonstrated. • The application of crack information in multiscale modeling of SHCC is highlighted. • Proposed workflow is suitable for diverse composite systems' cracking behaviors. [ABSTRACT FROM AUTHOR]