| Popis: |
The efficacy, reliability and versatility of the light emitting diode (LED) can outcompete most established light source technologies. However, they are particularly sensitive to high temperatures, which compromises their efficacy and reliability, undermining some of the technology s key benefits. Consequently, effective thermal management is essential to exploit the technology to its full potential. Thermal management is a well-established subject but its application in the relatively new LED lighting industry, with its specific constraints, is currently poorly defined. The question this thesis aims to answer is how can LED thermal management be achieved most effectively? This thesis starts with a review of the current state of the art, relevant thermal management technologies and market trends. This establishes current and future thermal management constraints in a commercial context. Methods to test and evaluate the thermal management performance of a luminaire system follow. The defined test methods, simulation benchmarks and operational constraints provide the foundation to develop effective thermal management strategies. Finally this work explores how the findings can be implemented in the development and comparison of multiple thermal management designs. These are optimised to assess the potential performance enhancement available when applied to a typical commercial system. The outcomes of this research showed that thermal management of LEDs can be expected to remain a key requirement but there are hints it is becoming less critical. The impacts of some common operating environments were studied, but appeared to have no significant effect on the thermal behaviour of a typical system. There are some active thermal management devices that warrant further attention, but passive systems are inherently well suited to LED luminaires and are readily adopted so were selected as the focus of this research. Using the techniques discussed in this thesis the performance of a commercially available component was evaluated. By optimising its geometry, a 5 % decrease in absolute thermal resistance or a 20 % increase in average heat transfer coefficient and 10 % reduction in heatsink mass can potentially be achieved. While greater lifecycle energy consumption savings were offered by minimising heatsink thermal resistance the most effective design was considered to be one optimised for maximum average heat transfer coefficient. Some more radical concepts were also considered. While these demonstrate the feasibility of passively manipulating fluid flow they had a detrimental impact on performance. Further analysis would be needed to conclusively dismiss these concepts but this work indicates there is very little potential in pursuing them further. |
