Model Order Reduction for Nonlinear Microfluidics
Date Issued
2005
Date
2005
Author(s)
Chien, Chih-Min
DOI
zh-TW
Abstract
In this work, we present a methodology of generating microfluidic compact models (reduced-order model) for 3-D non-linear microchannels based on the proper orthogonal decomposition (POD) and the Galerkin method. The microfluidic governing equation, the Navier-Stokes equations, is reduced into a compact set of ordinary differential equations (i.e., the compact model) by the Galerkin condition. On the other hand, the basis functions, which will be used by the compact models, are extracted from CFD-RC full-meshed simulated results by the singular value decomposition. For the simple rectangular channel and L-shaped channel structures, the discrepancy between results (steady-state and transient) of the reduced-order models and the full-meshed CFD-RC models is less than 1%. Furthermore, our preliminary results show that the typical speedup is greater than 1000. We also demonstrate that the compact microfluidic models are modular. In other words, it is possible to create the compact model of a complicated channel by assembling the compact models and the basis functions of various types of simple channels, without performing full-meshed CFD-RC computations for the complicated channel. The error between of the assembled model and the full-meshed CFD-RC model is negligible. This modular capability indicates that the compact models generated by this methodology are ready for the system-level analysis of microfluidics.
Subjects
非線性流體
精簡模型
流道模組
model order reduction
nonlinear fluidics
macromodel
system-leavel modeling
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-94-R92522704-1.pdf
Size
23.53 KB
Format
Adobe PDF
Checksum
(MD5):496c248f95f410ad71bfa4c3f6561be8