Masters Theses

Date of Award

5-1993

Degree Type

Thesis

Degree Name

Master of Science

Major

Engineering Science

Major Professor

A. J. Baker

Committee Members

G. S. Iannelli, Dick Kelso

Abstract

The work in this thesis is the benchmarking and validation of the continuity constraint algorithm for the solution approximation to the Navier-Stokes (NS) equation system for thermal-incompressible fluid flow. The benchmark cases are used to compare the continuity constraint solution to accepted accurate solutions. The two benchmark cases are the two-dimensional, isothermal step-wall diffuser and thermal cavity problems. The continuity constraint algorithm is then validated with comparison to experimental data obtained for a full-scale, two-dimensional test room for an isothermal case and a non-isothermal case with a heated floor. The finite element spatial discretization utilizes the bilinear basis for all variables, and time integration is performed with a θ-implicit ordinary differential equation approximation algorithm. The continuity constraint solution approximation of velocity is used for the vorticity and streamfunction solutions in a post-processing operation. The results are presented in the form of velocity vector and isocline plots for the interpretation of the results. the benchmark problem showed exceptional agreement for the continuity constraint solution with the accepted solutions as long as the flow remained two-dimensional. In the step-wall diffuser simulations at Reynolds numbers involving three-dimensional flow, the absence of cross-flow effects made comparison with the experimental results unsuitable. For the room air motion simulations, the results for the isothermal and non-isothermal case showed the solution to be unsteady with the appearance of oscillations in the supply jet and the absence of a stationary flow field. The transient nature of the flow made direct comparison with the experimental results inconclusive, since the experimental data was presented in a stationary format.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS