Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.751993
Title: Dynamic refinement and boundary contact forces in smoothed particle hydrodynamics with applications in fluid flow problems
Author: Feldman, Jonathan
Awarding Body: University of Wales, Swansea
Current Institution: Swansea University
Date of Award: 2006
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Smoothed Particle Hydrodynamics (SPH) is a relatively new, simple and effective numerical method that can be used to solve a variety of difficult problems in computational mechanics. It is a fully Lagrangian meshless method ideal for solving large deformation problems such as complex free surface fluid flows. This research was carried out with the support of BAE Systems and falls into two distinct areas. Firstly to investigate new methods for treating fixed boundaries and secondly to investigate refinement algorithms which allow for both sparsely and densely populated regions of particles within the same computational domain. Much work has been done in the modelling of particle-boundary interactions in SPH since the governing equations do not naturally incorporate essential boundary conditions. In this research a new technique for calculating boundary contact forces is developed. The forces are obtained from a variational principle and as such conserve both the linear and angular momentum of the system. The boundaries are explicitly defined using this new approach and so the need for additional boundary particles is removed. In the past most SPH derivations have been based on a uniform distribution of particles of equal mass. This leads to large simulations with many particles and long run times. In other mesh based schemes it has become common place to use mesh adaptivity to improve numerical results and reduce computation times. With a corresponding refinement strategy SPH can gain these same advantages. In this research a refinement strategy based upon particle splitting is developed. Candidate particles are split into several 'daughter' particles according to a given refinement pattern centred about the original particle position. Through the solution of a non-linear minimisation problem the optimal mass distribution for the daughter particles is obtained so as to reduce the errors introduced into the underlying density field. This procedure necessarily conserves the mass of the system. The unique daughter particle velocity configuration that conserves the linear and angular momentum of the system is also identified. The conclusion of the research was the successful implementation of these improvements into the existing SPH framework. As a result the capability and flexibility of the code is greatly increased and the computational expense needed for running large simulations has been reduced.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.751993  DOI: Not available
Share: