Energy conversion using thermal transpiration : optimization of a Knudsen compressor

Abstract

Knudsen compressors are devices without any moving parts that use the nanoscale phenomenon of thermal transpiration to pump or compress a gas. Thermal transpiration takes place when a gas is in contact with a solid boundary along which a temperature gradient exists. If the characteristic length scale is on the order of, or smaller than, the molecular mean free path, then the gas flows from cold to hot regions. The nanoscale nature of this phenomenon lends itself to use in nanoscale devices where moving parts are difficult to manufacture. Additional applications include low pressure environments, such as space or vacuum, where molecular mean-free paths are long. Although the flow rates obtained from individual Knudsen compressors are small, reasonable flow rates and significant pressure rises can be attained by cascading a large number of single stages. In this thesis, we use kinetic-theory based simulations to study thermal transpiration and its application to Knudsen compressors. We simulate such flows in a variety of porous media configurations and then study the effect of various device parameters and operating conditions on the compressor performance. It is generally observed that generally Knudsen compressors are more efficient when producing a flow than when creating a pressure rise. Small Knudsen numbers and short device lengths tend to increase the mass flow rate, but decrease pressure rise. Particular attention in our investigation is paid to the compressor efficiency, where a number of efficiency measures are defined, discussed, and compared to previous work in the literature, where available. It is generally found that the Knudsen compressor requires large temperature differences to be competitive as an energy conversion device.