For particle-filled micro model, indeterminacy and its influence on equivalent macro performance are emphasized by domestic and foreign researchers. Taking the high filling-proportion of propellant into account, randomcity of particle amount and nonintegrated filling-proportion was discussed based on duality of random sample. Then subdomains-centroid method was put forward to characterize the space distribution of given number of particles. Subsequently, user-defined technique of Abaqus/Python was used to study indeterminacy and its influence, in the frame of Matlab software. The results of verification test indicate, 1) randomcity of space distribution results in of equivalent modulus, on the condition of given number, and 2) subdomains-centroid method is intuitionistic and effective to characterize space distribution.