Date of Award

2013

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Architecture

First Advisor

Ali M. Malkawi

Abstract

Computational fluid dynamics (CFD) is one of the branches of fluid mechanics that uses numerical methods and algorithms to solve and analyze problems that involve fluid flows. Computers are used to perform the millions of calculations required to simulate the interaction of liquids and gases with surfaces defined by boundary conditions. Indoor airflow simulations are necessary for building emergency management, preliminary design of sustainable buildings, and real-time indoor environment control.

The simulation should also be informative since the airflow motion, temperature distribution, and contaminant concentration is important. However, CFD computation is usually time-consuming, and not suitable for simulating real-time indoor air movement. Many researchers are concentrating on both hardware utilization and CFD algorithms, to make simulation much faster. Fast flow simulations are important for some applications in the building industry, such as the conceptual design of indoor environment, or they are coupled with energy simulation to provide deep analysis on the performance of the buildings. Such application does not require the same high level of accuracy as traditional CFD simulation because it only requires conceptual or semi-accurate distributions of the flow but within a short computing time. However, year round simulation is needed rather than the analysis of two or three extreme cases in order to help the designer investigate the problem clearly. To meet these special needs, an efficient and informative fluid simulation method is needed to provide fast airflow simulation with an inevitable but nominal compromise in accuracy.

This research provides a comprehensive workflow for the designer to simulate and analyze the annual indoor environment. In addition to the hardware acceleration deployed, fast fluid simulation algorithm is developed, and a machine learning based interpolation is used to allow the simulation coverage to be conducted annually. The outcome of this research is a methodology that allows the annual simulation time similar to the one used to perform two or three extreme cases of simulation using current methods.

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

Share

COinS