Thermal management challenges for microelectronics are a major issue for future integrated circuits, thanks to the continued exponential growth in component density described by Moore¡¯s Law. Current projections from the International Technology Roadmap for Semiconductors predict that local heat fluxes will exceed 1 kW/cm2 within a decade. There is thus an urgent need to develop new compact, high heat flux forced-liquid and evaporative cooling technologies.
Thermometry techniques that can measure temperature fields with micron-scale resolution without disturbing the flow of coolant would be valuable in developing and evaluating new thermal management technologies. Specifically, the ability to estimate local convective heat transfer coefficients, which are proportional to the difference between the bulk coolant and wall surface temperatures, would be useful in developing computationally efficient reduced-order models of thermal transport in microscale heat exchangers.
The objective of this doctoral thesis is therefore to develop and evaluate non-intrusive optical thermometry techniques to measure wall surface and bulk liquid temperatures with O(1-10 micronmeter) spatial resolution. Intensity-based fluorescence thermometry (FT), where the temperature distribution of an aqueous fluorescent dye solution is estimated from variations in the fluorescent emission intensity, was used to measure temperatures in steady Poiseuille flow at Reynolds numbers less than 10. The flow was driven through 1 mm square channels heated on one side to create temperature gradients exceeding 8 ¡ÆC/mm along both dimensions of the channel cross-section. In the evanescent-wave fluorescence thermometry (EFT) experiments, a solution of fluorescein was illuminated by evanescent waves to estimate the solution temperature within about 300 nm of the wall. In the dual-tracer FT (DFT) studies, a solution of two fluorophores with opposite temperature sensitivities was volumetrically illuminated over most of the `cross-section of the channel to determine solution temperatures in the bulk flow. The accuracy of both types of FT is determined by comparing the temperature data with numerical predictions obtained with commercial computational fluid dynamics software. The results indicate that EFT can measure wall surface temperatures with an average accuracy of about 0.3 ¡ÆC at a spatial resolution of 10 micronmeter, and that DFT can measure bulk water temperature fields with an average accuracy of about 0.3 ¡ÆC at a spatial resolution of 50 micronmeter in the image plane. The results also suggest that the spatial resolution of the DFT data along the optical axis (i.e., normal to the image plane) is at least an order of magnitude greater than the depth of focus of the imaging system.
| Identifer | oai:union.ndltd.org:GATECH/oai:smartech.gatech.edu:1853/43754 |
| Date | 22 December 2010 |
| Creators | Kim, Myeongsub |
| Publisher | Georgia Institute of Technology |
| Source Sets | Georgia Tech Electronic Thesis and Dissertation Archive |
| Detected Language | English |
| Type | Dissertation |
Page generated in 0.0456 seconds