Crack Characterization Of Cement-based Materials Based On Digital Volume Correlation Method

Cement-based materials are widely used in the field of building materials.Due to their own shrinkage and expansion,external load and other factors,the formation and expansion of internal cracks have an impact on their performance.It is of great significance to provide a high precision 3D crack characterization method for cement-based materials because the 3D crack morphology and crack parameters cannot be characterized by current crack detection methods.In this paper,two crack recognition methods based on digital volume correlation(DVC)are proposed,one is based on gray fitting-displacement difference and the other is based on subvolume separation.The method based on gray fitting-displacement difference mainly includes fitting crack equation,updating crack equation and calculating width parameters.The method based on subvolume separation includes the steps of searching crack parameters,calculating subvoxel displacement and updating crack parameters.In order to verify the feasibility of the above two crack identification algorithms,a type I numerical crack was designed to evaluate the algorithm,and different conditions such as noise and deformation field were designed in the numerical experiment.Finally,cracks in real samples were identified and characterized,and the following conclusions were obtained:(1)In the condition of no noise,the minimum crack width identified by the method based on gray fitting-displacement difference is about 0.20 voxel,the recognition rate is60%～70%,and the width deviation is about 0.04 voxel.The minimum crack width identified by the method based on subvolume separation is about 0.10 voxel,the recognition rate is about 90%,and the width deviation is about 0.06 voxel.(2)In the condition of CT noise,the minimum crack width identified by the crack recognition method based on gray fitting-displacement difference is about 0.50 voxel,the recognition rate is about 60%,and the width deviation is about 0.18 voxel.The minimum crack width identified by the method based on subvolume separation is about 0.10 voxel,the recognition rate is about 80%～90%,and the width deviation is about 0.16 voxel.(3)In the condition of different deformation field,the recognition rate of the crack identification method based on gray fitting-displacement difference is 60%～70%.The ability of identifying the minimum crack width does not change with the change of deformation field,but the width deviation increases with the increase of deformation field.The identification rate of the method based on subvolume separation is 80%～90%,the ability of identifying the minimum crack width decreases with the increase of deformation field,and the width deviation increases with the increase of deformation field.(4)In the condition of real crack,the crack direction and crack width can be characterized by both methods.The crack width in the u and v directions is about 0.30 voxel,and the crack width in the w direction is about 0.10 voxel.The results of the two methods show that the width of the crack in the direction of u and v increases along with the crack depth,while the width of the crack in the direction of w remains unchanged.After analysis,the algorithm based on gray fitting-displacement difference is suitable for identifying and calculating cracks with widths larger than 1voxel and above.The method based on subvolume separation can identify and calculate cracks with a width of about 0.2voxel.