Complex Microfluidic Systems Architectures and Applications to Micropower Generation

摘要

Many highly efficient complex systems exhibit an architecture that contains a trade-off between reliability of components and the size of the system built up from such components. We advanced a novel microfluidic concept, named micro fCPU, the Microfluidic Central Processing Unit, where the key microfluidic operations are performed within a single enclosure, using active, software-based inputs rather than physical hardware changes. Parallelization of such a unit avoids necessity of multiple pumps and valves, reducing the footprint of the device. Use of such devices in parallelized architectures requires a study of reliability of individual components. In this project, we advanced such a study of trade-offs between reliability and parallelization in architectures for microfluidic power generation, optimization of which can lead to revolutionary changes in power density of such devices. On the single device side, we performed an experiment that utilized the concept of synthetic turbulence generation that can resolve the inefficiency of the single device operation. We studied the tradeoff between robustness and size of the system indicated above. We utilized the single-component prior uncertainty estimate based on experiments to simulate parallel architectures consisting of microscale units assembled into a power generator. Our study indicates that micropower generation devices can be substantially improved by using 1) the Synthetic Turbulence Generator to reduce the lenghtscale of mixing in the microcombustor 2) a linear array architecture utilizing several microcombustors, to improve robustness and the stability margin of microcombustion. The tradeoff that we studied is common for a large class of Complex Systems Architectures where reliability of components is traded against the size of the system.