A C++11 Implementation of a Moving Wall in the Lattice Boltzmann Method

Abstract

This thesis is part of the research on the thermodynamic properties of thorium salt and aims to provide a C++11 code to simulate fuel salt flow at various temperatures around a settling sphere and inside a container. This thesis should form the basis of a C++11 implementation on fuel salt flow in the Molten Salt Fast Reactor using the Lattice Boltzmann Method. The research into the thermodynamic properties of the thorium fuel salt is necessary in the research and development of new Generation IV Molten Salt Fast Reactors. The simulation uses the Lattice Boltzmann Method in the C++11 code language and the Palabos open-source library which are introduced in chapter 1. The governing equations in this simulation are the dynamics of the sphere, the Lattice Boltzmann Equation (LBE) and the boundary method equations as presented in chapter 2. The boundary condition chosen for this thesis is the Guo extrapolation method with the Inamuro iteration to determine the fluid force and torque on the sphere. These models were chosen since they existed in the Palabos library and were used in 3D moving wall examples. Since there was no documentation available from Palabos on the off-lattice 3D methods, building the model proved to be time consuming and was a trial and error process. The resulting code as shown by chapter 3 is still a work in progress. The challenge was to combine both a settling sphere or in other words a moving wall with fluid wall interactions, in 3D and inside an enclosed container (hence no infinite flow in any direction and a second boundary). Due to the complex nature of the problem and the lack of documentation, time ran out before results could be obtained. If the model would have generated results these would have needed to be verified and validated as presented by chapter 4. The intended simulation presented in chapter 5 was a settling sphere inside a cylindrical container with simulations for varying temperatures and Reynolds numbers. As stated earlier the model did not obtain results but the project did as stated in chapter 6, namely a good understanding of modelling fuels salts with C++ and the Lattice Boltzmann Method (LBM) Palabos library is obtained. Therefore in chapter 7 it is recommended: to either update the palabos library and expand the documentation, resulting in a library owned and maintained by Delft University of Technology for Computational Fluid Dynamics (CFD) with the Lattice Boltzmann Method (LBM) in C++, or choose a different library entirely. Any such library should make the most of the advantage that the LBM has, namely that it is a very parallel method. Proper documentation and coding standards should be implemented keeping the library readable and understandable. Hence the conclusion in chapter 8 is that the Palabos library might prove a good solution in the future if modified and documented. More research on different boundary conditions and their performance is required. More research is also required on the performance of the Palabos library compared to other libraries. Finally a performance analysis for hardware and parallelisation methods would be required to make a decision on improvements in that department.