Abstract: In this paper, materials such as plaster, pumice and diatomite are used to simulate water-bearing porous rock as similar material specimen. The experiment adopts four-factor (cement/plaster/diatomite ratio, the standard ratio of pumice sand, bone-glue ratio and barite powder ratio), five-level orthogonal experimental design to measure the density, uniaxial compressive strength, porosity, elastic modulus and softening factor. The measurements are conducted to determine whether the simulation materials can be used to simulate water-bearing rocks. With the use of direct analysis method, we determine how various factors affect physical and mechanical nature of the specimens. Analysis results show that: Due to different material ratios, the strength of the specimens is widely distributed. Comparing with the parameters of the porous layer, under certain similar conditions, we find that the specimens are with similar properties to the porous layer rock. Diatomite ratio is negatively correlated with density, compressive strength, elastic modulus and porosity, and it is positively correlated with softening coefficient. Pumice ratio is negatively correlated with density, compressive strength, elastic modulus and softening coefficient, and it is positively correlated with porosity. Therefore, addition of the two materials can simulate the porous nature of the rock sample well. Within the allowable range, the established empirical equation to simulate porous water-bearing rocks can be used in engineering practice.

Key words: similar material, porous water-bearing stratum, orthogonal design method, regression analysis, empirical equations