Simulations for the explosion and granular impact problems using the SPH method

Abstract

Simulations of explosions and granular impacts are challenging tasks to tackle using conventional mesh-based methods. In this thesis, a mesh-free technique called smoothed particle hydrodynamics (SPH) in conjunction with the Open-MP and CUDA parallel pro- gramming interfaces is introduced to tackle three-dimensional (3D) problems with large deformations. Chapter 1 gives an introduction of the SPH method and a literature review of the the- oretical improvement of SPH, landmine detonations, underwater explosions, and granular impacts. A research outline of the thesis is also presented at the end of this chapter. The basic ideas of the SPH method and some techniques which are relevant to improve the accuracy and stability of SPH, including the artificial viscosity, artificial stress, boundary implementation, neighboring particles search, and kernel gradient correction, are described in Chapter 2. In order to solve the governing equations, an elaboration of the constitutive models to update the stress tensor of soil and solid and the equation of states (EOSs) is given in Chapter 3. The simulations of the detonation and granular impact problems using the SPH method are thoroughly presented in chapters 4-7. In Chapter 4, in order to tackle 3D problems with large number of particles, the in-house SPH code is parallelized by the Open-MP programming interface. The parallel efficiency is tested by the 3D shaped charge detonation. The simulations of the 2D soil explosion and its effects on structures are investigated in Chapter 5. Based on the parallelization of the SPH code and the simulation of 2D soil explosion, the physical process of the 3D landmine detonation is studied further. The simulations of the 3D underwater explosion within cylindrical rigid and aluminium (Al) tubes including cavitation phenomenon are presented in Chapter 6. The simulations of the 3D granular impacts using GPU acceleration are presented in Chapter 7. The numerical results of SPH are compared against the experimental and other available numerical data, and it is shown that the SPH method is capable of predicting landmine detonations, underwater explosions, and granular impacts. The conclusions, novelties, and future plan of SPH research are summarized in Chapter 8.