| Scientific Objectives |
Nevertheless, MEMS has several formidable technical hurdles to overcome. Fluid behavior, in particular surface tension, is significantly different at these small scales. This weak surface force gains importance and dominates at micron length scales due to the high surface area to volume ratio [1]. For example, the pressure required to force a bubble through a circular micro-channel into a water reservoir can exceed 100 kPa (~1 atmosphere), a pressure well above the specifications of many micropumps [2]. Obviously, tasks such as filling a channel and purging a gas bubble are not trivial. Unfortunately, o generic set of design rules for the exact geometry of filling-friendly microfluidic structures exists up to now. [3].
Fuel cells, like MEMS devices, are affected by surface tension phenomena. New cell designs incorporate smaller channels, down to 5 um, to enhance transport and improve efficiency. However, gas slugs can form in channels of electrode membranes due to sloshing or chemical reactions resulting in a decrease in efficiency [4]. These readily block these smaller channels and are difficult to flush out. Utilization of capillary forces, channel geometry, or other effects must be done in order to achieve optimal cell performance and reliability[5].
|  Fig 1: Cross section of DRIE[8] |  Fig 2: KOH anisotropic etch[9] While progress has been made in understanding micro-capillary phenomenon, MEMS component design has not taken advantage of this knowledge to address two-phase flow blockage and liquid filling in channels and channel junctions. The objective of this proposed research is to investigate how channel geometry and material selection affect bubble formation, stability, and breakup. Surface tension modifying techniques, such as electro-wetting and hydrophobic/hydrophilic coatings, will be also considered. Ultimately, optimal designs and practical recommendations will result in simple, cost-effective, reliable MEMS devices. In phase one, an extensive literature survey is done to catalog different methods of etching. Manufacturing techniques such as DRIE (Deep Reactive Ion Etching), wet etching, conventional machining, and others produce different etching patterns, have depth limitations, require certain materials, and have significantly different production costs. For example, DRIE creates deep (> 500 um) rectangular channels (Fig. 1), while wet etching is less easily controlled and forms shallow trapezoidal or semicircular grooves (Fig. 2). However, DRIE requires a large up-front investment (~$1M) and requires expensive chemicals for production [6]. In contrast, the cost of wet etching is up to 2 orders of magnitude less expensive than DRIE [7]. Other techniques such as micromachining are even more economical but produce different cross-sections than described above. Complimenting the literature research, phase two considers etching requirements and the structure of MEMS components. Particular components may require particular cross-sections and aspect ratios for functionality and performance. Furthermore, this will establish particular channel geometries of interest. In addition to geometry, the contact angle of the fluid can be adjusted to change stability properties. This can be done using four different methods: material selection, coatings, thermo-capillary phenomena, and electro-wetting. For example, a Teflon-coated channel leading from a water reservoir will require a large backpressure to fill; while a hydrophilic one will fill voluntarily. In addition, active control of the interface can be achieved using electro-wetting. An electrical potential across the bubble will result in an effective quadratic decrease of the solid-liquid surface tension [1]. While this relationship (Young-Lippman equation) is phenomenological, other effects can be incorporated using variational techniques on Maxwell equations [10]. After the survey, Phase IV will use Surface Evolver to calculate the gas-liquid interface and determine its static stability for a variety of geometries, contact angles, surface tension modifying techniques, and bubble volumes. Surface Evolver, a NSF-funded software package, computes minimal surfaces that are formed by surface tension, gravity, and other energies and that are subject to various constraints [11]. Critical contact angle and capillary instability phenomena in complex geometries . such as satellite propellant management devices, arbitrary packed sphere beds, and channel junctions . are computed with ease. Practical modifications to existing topologies will be tested. In Phase V, the dynamic progression of the bubble is found using an existing CFD code with a level-set modification. The dynamical results can be compared to a static solution found in Surface Evolver to check for numerical convergence. Ultimately, this will give further insight into bubble breakup and instability. Selected results will be tested experimentally in Phase VI. Pressure can be applied externally to move a series of liquid slugs in the microchannel. If a glass cover plate is used, techniques such as pressure sensitive paints or microPIV will be used to study transient behavior. Using these results, the accuracy of CFD will be assessed. Finally, experimental and computational results will be collected and contrasted. Designs will be examined for both two-phase flow instability and manufacturability. Therefore, a performance metric for each design will be established: (1) modes of interface stability, (2) pressure or electro-wetting power required for gas/liquid slug removal, (3) estimated costs, time, and issues for manufacturing and (4) practical channel/junction applications. Originality of ProposalWhile microfluidics is not branded as an aerospace topic, the aerospace industry and NASA have laid the foundation of low Bond number (surface tension dominate) fluid mechanics. Prof. Collicott's research group has mainly focused on fluid behavior in satellite fuel tanks and control using vane structures. He has shown Surface Evolver to be highly versatile with complex geometries and this has lead to diverse work in two-phase stability, such as pulmonary branches and sphere-bed layers [12]. While building upon the expertise of his research group, this proposal is unique due to its focus on microdevices and electro-capillary phenomena.References
Powered by BOINC and Surface Evolver |