Evaluation of Mechanical Properties of Composite Solid Propellant by Genetic Engineering of Materials: Part B: Modeling of Minimum Representative Unit for Multi‐Modal Particle Packing Structure Based on Formulation of Composite Solid Propellant. Issue 4 (10th February 2021)

• Record Type: Journal Article
• Authors: Dong, Ge; Zhang, Wei; Zhou, Xing; Bao, Tong; Deng, Lei
• Is Part Of: Propellants, explosives, pyrotechnics. Volume 46:Issue 4(2021)
• Page Start: 654
• Page End: 662
• Publication Date: 2021-02-10
• Subjects: Particle packing structure -- Minimum representative unit -- Multi-modal particles -- L−S algorithm -- Genetic algorithm; Propellants -- Periodicals; Explosives -- Periodicals; 662.2
• DOI: 10.1002/prep.202000178
• Languages: English
• ISSNs: 0721-3115

Abstract

Abstract: To show the actual particle packing structure of AP/Al/HTPB propellant with multi‐modal fillers, a minimum representative unit of particle packing structure is established. In the minimum representative unit, the particle size of fillers in each mode is identical to the value in the propellant formulation. Meanwhile, on the basis of the amount series of fillers in the per unit propellant, the amount of fillers in each mode is respectively divided by that of the fewest filler, thus the amount series of fillers in the minimum representative unit is obtained. Further, with the condition that the average density of the unit is the same as that of the propellant, the side length of the minimum representative unit could be uniquely determined. In order to generate the particle packing structure with particles evenly distributed in the unit, a two‐step algorithm is put forward. In the first step, the genetic algorithm is adopted to search for an evenly and non‐overlapping distribution for the few and coarser particles. In the second step, the residual finer particles are randomly placed in the voids among the coarser particles by a modified L−S algorithm. Results showed that the particles are distributed more evenly in the packs generated by the two‐step algorithm than in packs generated by the classic L−S algorithm. Abstract :