| Nearly a century ago, the emergence of electromechanical technology coupled with a new understanding of psychrometry led Willis Carrier to invent unprecedented systems for manifesting the potential of "man-made weather." His conceptual strategy for conditioning the thermal environment of buildings remains virtually unchanged today. But as HVAC systems were proliferating, the science of fluid mechanics was evolving, and two significant breakthroughs over the course of this century are challenging the validity of the perfectly-mixed centralized air system. The development of boundary layer theory enabled the eventual invention of computational fluid dynamics, which now provides researchers with an unprecedented window into the thermal behavior of air. And in a scenario similar to that facing Carrier in 1904, a new technology--microelectromechanical systems--is emerging that may finally enable architects to design the discrete thermal environment of a building, rather than only integrate the static shell of generalized systems.; Computational fluid dynamics offers the opportunity to explore the discrete behavior of air, but the majority of building simulations have been devoted to large scale models with high momentum air supplies. By removing the prima facie high velocity air diffuser from the room behavior, however, one can consider that the air phenomena present are essentially independent behaviors that interact with room air through their respective boundary layers. Small scale modeling of the boundary layer reveals that heat transfer from a given thermal source can be significantly altered if the characteristic length which drives convective transport is shifted. This shifting can occur by changing the location and/or orientation of the cold sink, without changing any physical parameters of the thermal source. FIDAP, a finite element CFD code, was used in this thesis to investigate the impact of characteristic length on boundary layer heat transfer, and preliminary results indicate that orientation of the cold sink has a far greater impact on characteristic length than previously assumed. As a result, direct control of boundary layer heat transfer is achievable and may be an ideal application of MEMS technology. |
