Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.752193
Title: Low Reynolds number heat transfer prediction employing large eddy simulation for electronics geometrics
Author: Tyacke, James C.
Awarding Body: Swansea University
Current Institution: Swansea University
Date of Award: 2009
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
The accurate prediction of convective heat transfer within electronics systems has always been of great importance for the reliability of such systems. Current computational methods based on the Reynolds-Averaged Navier-Stokes equations do not provide reliable predictions due to the inability of current methods to capture complex time dependent flow features. This study investigates the use of time dependent Large Eddy Simulation and hybrid methods to make more reliable thermal predictions. These methods are tested on a heated ribbed channel, a heated cube in an array of cubes and a complex CPU case. A variety of models and methodologies are applied and analysed. It is apparent that the most important scales are the large vortices generated by geometrical features. Due to the low Reynolds number flows found in electronics systems, there is a relatively small range of scales to capture. This gives rise to some unpredictability in model choice and grid resolution, though consistency is much improved over traditional methods. Important sources of error are considered to be problem definition and boundary conditions for which unsteady data is not available. Use of nonlinear models and higher order discretisation did not provide adequate improvements in accuracy for the increase in computational expense. Combining Reynolds-Averaged Navier-Stokes and Implicit Large Eddy Simulation into a hybrid model seems to provide fair reliability when compared to other modelling methods on a range of grid resolutions.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.752193  DOI: Not available
Share: