Multiphysics, LTMS, Transient, Air Conditioning
The Roving Comforter (RoCo) is an innovative personal thermal management technology currently being developed at the University of Maryland as part of Advanced Research Projects Agency â€“ Energy (ARPA-e) project. It is an ultra-low cost, autonomous, battery powered air-conditioning system that provides comfort to a single person by discharging conditioned air through adjustable nozzle. To meet the stringent cost and performance targets defined by ARPA-e, careful selection of components is required for battery, robotic platform as well as the cooling mechanism. Hence a system level model which incorporates thermodynamics, electricity as well as mechanics at the same time taking cost into consideration is being developed to obtain the optimum design. This paper explores four different cooling mechanisms and discusses their transient performance. The mechanisms are categorized based on chilled water, ice water mixture, phase change material and vapor compression cycle (VCC). Preliminary calculations have revealed the water based model to have a bulk manufacturing cost of $96 on the lower while the VCC model to have $130 on the higher side. Weights of the cooling mechanism vary from 4 kg for the ice water model to 25 kg for the chilled water model. The coefficients of performance vary between 1.4 for the chilled water to 2.7 for the VCC model. System parameters to operate each mechanism for a sample two hour cooling operation are obtained from the model. The challenges faced in making each mechanism operational are discussed with their possible solutions. Experiments carried out for the battery pack reveal a total discharge time of roughly 3.5 hours. The discharge trends have been captured by a battery model. This model will be used to develop a unified battery pack for the cooling mechanism as well as the robotic functionality. The model developed for the purpose demonstrates the benefits of a unified Multiphysics model in comparison to the current methodology of using single physics models which need to be run independently. The approach is expected to provide a significant reduction in design time.