Planar Micro Nozzles - CFD Model Demands for Thrust Optimization

Regarding current trends in the miniaturization of satellites, an appropriate propulsion system is sought. Within the EU project PRECISE, a system analysis has shown that a MEMS (micro-electro-mechanical system) monopropellant hydrazine thruster has the potential to deliver thrust and Isp on a level that cannot be matched by other microsystems (e.g. cold gas). While manufacturers and researchers are experienced in catalyst and MEMS design, there is to date no standard or state-of-the-art in the analysis and design of high Reynolds number, micro scale, planar nozzles. Therefore, we undertook a systematic study of the modeling demands for such flows. A first part concerns the thermochemical fluid representation in CFD. We show that a constant property perfect gas representation will misevaluate thruster performance, as changes in the isentropic exponent have major impact on the theoretical exhaust velocity. The error grows with the complexity of the exhaust molecules and is more pronounced for molecules such as ammonia NH3 which is present in hydrazine decomposition. Hence, it is also important to model the gas composition as detailed as possible. Heat loss in the combustion chamber is assessed, which is very pronounced in micro scale systems. It is shown that a thus reduced chamber temperature strongly affects engine efficiency and hence also needs to be incorporated into any serious analysis. Finally, geometrical simplifications in CFD are regarded. We demonstrate that planar nozzle flows are inherently three dimensional in nature; CFD analysis of a 2D configuration will yield grossly wrong results as it cannot capture the boundary layers growing from the top and bottom wall. This effect is a lot more pronounced than the boundary layers growing from the contoured nozzle walls. All four boundary layers interact, choking the nozzle flow until it almost resembles a fully developed pipe flow.