Rory Roberts (Advisor), Scott Thomas (Committee Member), J. Mitch Wolff (Committee Member)
Master of Science in Engineering (MSEgr)
A statistical based verification and validation process is applied to the transient modeling of a shell and tube heat exchanger. A generic model of a heat exchanger was developed based on first principles as a sub-system of a larger thermal system model. This model was originally created without any experimental data, as it was not readily available. To provide the data necessary to apply the validation process, a thermal emulator was designed and built that allowed control of all system inputs to the heat exchanger, while also providing the instrumentation to record all required data. A wide test matrix was chosen to fully encompass the expected operational envelope of the heat exchanger. Focus on the collection of experimental data was the minimization of uncertainty, as these uncertainties were amplified once they were propagated through the validation process. The validation process encompasses the completion of sensitivity and uncertainty analyses, uncertainty propagation, verification, and validation. Once these steps were completed using a set of non-ideal experimental data, uncertainty in the transient heat exchanger model is quantified. This manuscript proposes a way to complete the validation process without replicate data sets by utilizing known information about the physical process. At the completion of the process, both uncertainties and model form error are quantified for the system outputs and a statistical validation metric is applied. These outputs help to define whether or not the model captures the physical process to a satisfactory degree while also highlighting avenues for improvement if the uncertainty is deemed too large for the intended application.
Department or Program
Department of Mechanical and Materials Engineering
Year Degree Awarded
Copyright 2015, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.