Date of Award:

5-2012

Document Type:

Dissertation

Degree Name:

Doctor of Philosophy (PhD)

Department:

Mechanical and Aerospace Engineering

Committee Chair(s)

Barton L. Smith

Committee

Barton L. Smith

Committee

Robert Spall

Committee

Heng Ban

Committee

Aaron Katz

Committee

Blake Tullis

Abstract

Computational modeling is of particular interest to science and engineering for the improvement of design and development of products and research of physical phenomena. However, confidence in a computational model must be validated prior to it's application through comparison to experimental data. The nuclear power industry has interest in the application of computational modeling to plant design, safety, and development for the increased understanding of heat transfer and fluid dynamics. Fluid dynamics, particularly time-varying phenomena within the reactor core, has a strong effect on heat (energy) transfer and transient accident scenarios of a nuclear power plant. While this work was funded by Idaho National Labs, this research is also applicable to engineering design of heat exchangers. This research began May of 2009.

The research within this dissertation demonstrates the validation of various time-varying quantities predicted by computational fluid dynamics models (CFD) with the use of experimental measurements. Several CFD turbulence models are implemented and experimental fluid velocity and pressure data are acquired using non-invasive measurement techniques for flow through a channel containing cylinders (confined cylinders). The CFD and experimental geometries and inlet and boundary conditions are identical and model the geometries found within a nuclear reactor core. Validation of the CFD models is assessed and it is found that while some models can accurately predict the time-varying quantities considered, the majority demonstrate deficiencies. Higher fidelity models typically predict these quantities with increased accuracy. The uncertainty for both the computational and experimental results are estimated and discussed.

Uncertainty estimation of both CFD and experimental results is required for meaningful validation. However, uncertainty from the non-invasive velocity measurement technique, particle image velocimetry (PIV), is inadequate for validation purposes, particularly random time-varying flow fluctuations (turbulence). Within this dissertation, the uncertainty from four dominate error sources are demonstrated, calculated, and analyzed for PIV. This method of estimating instantaneous, mean, and fluctuation uncertainty for PIV is demonstrated with significant accuracy and is applied to the confined cylinders experimental results for the validation study.

Checksum

3141d280e160f942e517229dbbe58a98

Comments

This work made publicly available electronically on May 9, 2012.

Included in

Philosophy Commons

Share

COinS