Abstract:Damage evolution of rock under loading can be expressed to some extent by the deformation and crack propagation of the specimen surface,which could be observed or analyzed with the digital images recorded from the specimen surface during loading. However,current methods to evaluate deformation from digital images are either qualitative(semi-qualitative) or quantitative,and very complicated computation and special software are demanded(for example,digital speckle correlation method). In this paper,a very simple quantitative analysis method for digital images recorded from loaded rock specimen to evaluate its damage evolution is proposed. To realize the method,a reference image is firstly selected from the series of digital images;and then the gray correlation of all other images is calculated using a pre-defined calculation scheme. When the image coordinate is fixed,the deformation will change the pixel aray,therefore,the correlation will decrease when there is a large deformation. Likewise,intensity of crack position will change sharply because of the existence of new exposure part by crack,then there is a sharp decrease of the gray correlation. Therefore,the distribution of gray correlation will reveal the damage distribution. The validity of this evaluation method is verified through two simulated image pairs,one deformation and one crack. Then,a series of surface images recorded from an en-echelon rock structure are analyzed using this method. The result shows that the spatial and temporal parallelism exists between the evolution of the gray correlation and the deformation localization. Therefore,gray correlation is expected to be a useful tool to evaluate the damage evolution of rock.