Abstract: In order to understand the macroscopic mechanical properties and experimental phenomena of propellants from the mechanism, it is necessary to carry out meso-mechanical research on propellants, and the representative volume element (RVE) is an important part of meso-mechanics research. Based on the random sequential adsorption (RSA) method, the delivery area was discretized into a background grid, and then a second random delivery was performed, which can generate a mesoscopic model with 80vol% particle volume fraction, and the model generation efficiency is also improved compared with the traditional RSA method. The method was programmed in Python language, and the secondary development interface of ABAQUS was used to realize the rapid generation of RVE model and load periodic boundary conditions. The numerical examples were used to illustrate the effectiveness of the algorithm, which provides a reference for the meso-analysis of propellants.