Incompressible flow over a three-dimensional cavity

This thesis investigates, by using a computational fluid dynamics (CFD) method, the physics of incompressible flow passing over a three-dimensional cavity on a solid surface. This study has been divided into two parts: 1) incompressible laminar flow passing over a three-dimensional cavity; and 2) incompressible laminar-to-turbulent transition over a three-dimensional cavity.<br/><br/>In the second part, a CFD code for the solution of the three-dimensional unsteady incompressible Navier-Stokes equations is developed. This is used to simulate incompressible laminar flow past a rectangular cavity. Typical result of simulation for varying Reynolds numbers, inflow condition and cavity geometry are presented.. Unsteady vorticial structures and shear-layer oscillations are observed within and around<br/>the cavities, and the flow becomes highly unsteady at high Reynolds numbers.<br/><br/>The third part examines the laminar- o-turbulent transition over three-dimensional cavities at high Reynolds numbers. This part itself includes two part: 1) simulation of the transitional flow and 2) prediction of the transitional areas in the shear layer of the cavity. A CFD code based on the large eddy simulation approach is developed for the simulation study. For the prediction of the areas of transition, a new:<br/>spectral-entropy based method is proposed. The product of the spectral entropy and energy of the pressure fluctuation is used for predicting the location of transition in the shear-layer flow. Typical results of simulation of incompressible laminar flow past a rectangular cavity for a range of Reynolds numbers are presented. The results indicate that the transition to turbulence occurred in the shear layer at some<br/>high Reynolds numbers ( e.g. R-e = 50, 000 100}000). The lambda structures and the “half mushroom” structures are observed in the transitional flows. The new method for predicting the transition is tested, and shows a good correlation with the results obtained based on large eddy simulation.<br/>

Name dc:type.qualificationname

Doctor of Philosophy

Level dc:type.qualificationlevel

Doctoral Thesis

Grantor dc:publisher.institution

Queen's University Belfast

Year dc:date.issued

2003

Author dc:creator

• Yao, Hui

Language dc:language

eng

Identifier

oai:pure.qub.ac.uk/portal:studenttheses/01008e46-c3ec-4eda-9ca5-2e8e789ff490

OAI identifier oai:identifier

oai:pure.qub.ac.uk/portal:studenttheses/01008e46-c3ec-4eda-9ca5-2e8e789ff490