Study on Concrete Mesoscopic Damage Under Dynamic Loading Using CT Image Segmentation and Gray Level Co-occurrence Matrix Eigenvalue

The process of concrete internal damage development to fracture is completely recorded by series CT images of concrete under dynamic loading.However,macroscopic damage of concrete has occurred when the microscopic cracks can be observed at CT image.The core of concrete CT image analysis is to excavate the microscopic damage information in concrete damage stages.In this research,image segmentation technology and gray co-occurrence matrix theory(GLCM) were used to extract the porosity and four eigenvalues(contrast,energy,correlation and homogeneity) respectively to study the change law of these parameters with stress in different cross sections under uniaxial dynamic compression using CT.Results show that porosity presents a general trend of monotonous increase with the increase of stress,while the porosity in some parts shows little change with the increase of stress when the stress level is small,which reflects the influence of concrete density compression effect.Contrast increases and decreases with the increase of stress at low stress level due to the influence of density compression effect.The index of energy is not sensitive to stress change.Correlation and homogeneity increase monotonously with the increase of stress,which are more sensitive for reflecting the whole damage process of concrete.This study shows that the sectionalized porosity,homogeneity and correlation of GLCM is helpful to reflect the mesoscopic damage process of concrete under dynamic load.