Design automation based on fluid dynamics
MetadataShow full item record
CitationA. Lashkaripour, R. Sanka, J. Lippai, D. Densmore, Design Automation Based On Fluid Dynamics. in The Proceedings of the 9th International Workshop on Bio-Design Automation, (2017)
Microfluidic devices provide researchers with numerous advantages such as high throughput, increased sensitivity and accuracy, lower cost, and reduced reaction time. However, design, fabrication, and running a microfluidic device are still heavily reliant on expertise. Recent studies suggest micro-milling can be a semi-automatic, inexpensive, and simple alternative to common fabrication methods. Micro-milling does not require a clean-room, mask aligner, spin-coater, and Plasma bonder, thus cutting down the cost and time of fabrication significantly. Moreover, through this protocol researchers can easily fabricate microfluidic devices in an automated fashion eschewing levels of expertise required for typical fabrication methods, such as photolithography, soft-lithography, and etching. However, designing a microfluidic chip that meets a certain set of requirements is still heavily dependent on a microfluidic expert, several days of simulation, and numerous experiments to reach the required performance. To address this, studies have reported random automated design of microfluidic devices based on numerical simulations for micro-mixing. However, random design generation is heavily reliant on time-consuming simulations carried out beforehand, and is prone to error due to the accuracy limitations of the numerical method. On the other hand, by using micro-milling for ultra-fast and inexpensive fabrication of microfluidic devices and Taguchi design of experiments for state-space exploration of all of the geometric parameters, we are able to generate a database of geometries, flow rates, and flow properties required for a single primitive to carry out a specified microfluidic task.
This article was accepted and presented at the 9th International Workshop on Bio-Design Automation, Pittsburgh, Pennsylvania (2017).