Correlational fractal characteristics and damage progression of granite with different grain sizes based on acoustic emission monitoring

Mineral grain size caused by the geological diagenetic process is one of the fundamental reasons leading to the differences in rock mechanical behaviors, which brings challenges to the stability of underground engineering, especially in rock masses with continuous grain size changes. This study conducted uniaxial compression tests on granite samples obtained from the same invaded anticline with different grain sizes. During the tests, the acoustic emission (AE) signals were acquired synchronously. The correlation between damage evolution and AE data was then studied by frequency-spectrum analysis, fractal theory methods, and G-P algorithms. The mechanism of grain size effect on granite failure evolution was discussed. The results reveal that as the grain size increases, the strength and the elastic modulus decrease while the plastic deformation characteristic is enhanced. Grain size also affects the frequency distribution of AE signals and its time-varying characteristics. With the increase in grain size, the proportion of high-frequency signals declines during loading. The fractal dimension value of dominant frequency during loading fluctuates within the range from 0.5 to 2.25 and generally shows an overall ‘rise-fall-rise-fall’ mode. In addition, as the grain size increases from fine-grained to medium-grained, the fractal dimension changes from an increasing to a decreasing trend earlier in the compaction and elastic stages. Especially in coarse-grained samples, after a brief increase, the fractal dimension decreases continuously until approaching the overall failure. The frequency domain characteristics and correlation fractal variation of granites reflect the damage and crack evolution inside the rock sample. More scattered and unordered micro-cracks generated in samples with finer grains facilitate the dispersion of applied stress and the reduction of local stress concentration, resulting in higher strength and stiffness of the sample.