Characterizing acoustic propagation in unconsolidated sand reservoirs is critical in offshore oil and gas exploration. We have simulated the acoustic field in a borehole surrounded by granular media based on nonuniform contact, using a 2nd-order in time and 10th-order in space finite-difference technique. We focus on the impact of the porosity and coordination number, grain size, and grain scale distribution on acoustic logging. Numerical simulation results show that P- and S-wave velocities decrease with increasing the porosity or decreasing the coordination number and increase with increasing the grain size. For different grain size distributed in the vertical and radial directions, the velocity and amplitude of the P-wave and S-wave are different. As reflected waves in a borehole, the arrival wave’s velocity is higher and the amplitude is stronger, whereas grains near the source or borehole axis are larger. The results of this paper provide a reference for analyzing and predicting different graded bedding formations for acoustic logging.