Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.628896
Title: A solution adaptive grid (SAG) for incompressible flows simulations : an attempt towards enhancing SAG algorithm for large eddy simulation (LES)
Author: Kaennakham, S.
Awarding Body: Coventry University
Current Institution: Coventry University
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
A study of the use of solution adaptive grid (SAG) method for simulations of incompressible flows is carried out in this work. Both laminar and turbulent types of flows are chosen. Investigation on laminar flow simulation starts with mesh adaptation criteria that are based on strong changes of some selected flow parameters; pressure and velocity components. Three most common laminar types of flows are studied; flow in a circular pipe, flow in a channel with sudden expansion and flow in a cavity with a moving lid. It is found that with the use of SAG, a reduction in both computational grid nodes and CPU time can be obtained when compared to those of fixed grid while satisfactory solutions are also achievable. Nevertheless, the refinement criteria setup procedure reveals inconveniences and requirement for several judgments that have to be defined ‘ad hoc’. This hence, makes the refinement criteria dubious for real engineering applications. For the study of turbulent flows with large eddy simulation (LES) and implicit filtering, examination of literature reveals that the lack of connections between the filter width and a physical scale has made LES somewhat unclosed, i.e. in a physical sense. In addition, it is known that numerical and modelling errors are always combined and it is difficult to study each of them separately making the total error magnitude difficult to control. Since both error types are characterised by the grid size, LES users very often find cases where a finer mesh no longer provides better accuracy. An attempt to address this ‘physical’ enclosure property of LES and its complication to implement/setup in FLUENT begins with the construction of a new refinement variable as a function of the Taylor scale. Then a new SAG algorithm is formed. The requirement to satisfy a condition of the selected subgrid scale (SAG) model, the Smagorinsky model, is taken into consideration to minimize the modeling error. The construction of a new refinement algorithm is also aimed to be the key to studying the interaction between the two types of error and could lead to the means of controlling their total magnitude. The validation in terms of its effectiveness, efficiency and reliability of the algorithm are made based on several criteria corresponding to suitability for practical applications. This includes the simplicity to setup/employ, computational affordability, and the accuracy level. For this, two different turbulent flow types that represent different commonly found turbulent phenomena are chosen; plane free jet and the flow over a circular cylinder. The simulations of the two cases were carried out in two dimensions. It is found that there are two key factors that strongly determine the success of the algorithm. The first factor is the Taylor scale definition, with literature only available for the turbulent plane jet study, for which good level of accuracy is expected. Unfortunately, this is not true for the flow over a circular cylinder, indicating a need for further analytical work. The second encountered difficulty results from limited access to software codes, which makes it impossible to implement the proposed scheme. As a result, the algorithm formulation needs be modified with carful judgment. Nevertheless, overall results are in reasonably good agreement with their corresponding experimental data.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.628896  DOI: Not available
Keywords: engineering mathematics, fluid dynamics, incompressible flows, flow simulation, solution adaptive grid
Share: