Experimental investigation on strength-weakening effect and fracture behavior of red sandstone under coupled static and dynamic cyclic loadings

Underground rock masses are often exposed to dynamic disturbances due to seismic waves and blasting exca-vation disturbances, resulting in significant changes in their physical and mechanical properties. To reveal the strength-weakening effect of rock under the coupled stress, the Brazilian disc tests of red sandstone were carried out by coupling pre-static and dynamic disturbing. The impacts of preloading, loading wave amplitude, dynamic disturbance frequency, and cycle number were studied on the weakening of Brazilian disc splitting strength. Furthermore, the fracture behaviors were characterized and discussed from the fracture morphology and fractal dimension. The results show that the loss ratios are positive under dynamic disturbance loading, indicating that the Brazilian splitting strength is weakened. For the same load wave amplitude, the loss of strength and P-wave velocity under 70 % preloading is more than twice that under 55 % preloading. At high preloading, the maximum strength loss and wave velocity losses are 13.7 % and 19.2 %, respectively. Comparing the disturbance degree of the sample with the increase of frequency by 1 Hz, cycle number by one time, and load wave amplitude by 1 % PS, it is concluded that the decline of wave velocity for the rock sample with the increase by 1 % PS in load wave amplitude is the largest, followed by the frequency and cycle. Only micro-cracks appear on the fracture surface under pure static loading. However, when the dynamic load disturbance is applied, the cement wrapped with debris falls off, besides micro-cracks. And with the increase of amplitude, the area of bare debris also progressively intensifies. Meanwhile, the width and number of micro-cracks gradually increased, with the broadest micro-crack reaching 18.2 mu m. Preloading, loading wave amplitude, loading frequency, and cycles are positively correlated with fractal dimension D. The higher the D value, the more broken the sample.