Optimizing Microstructure on Surface of Patterned Membrane by Smart Algorithm

Abstract

Patterned membranes show excellent antifouling capacity, compared with a flat membrane under the same conditions. However, it is not clear whether the patterned membranes have been optimized sufficiently. Therefore, we applied computational fluid dynamics (CFD) and a smart algorithm for optimizing a patterned membrane. The patterned membrane was linear and its cross section was an isosceles triangle. Compared to a reported patterned membrane (MR), CFD and the smart algorithm suggested that the length of the bottom line and the height of the isosceles triangle on the optimized membrane (MO) should be increased from 400 µm to 560 µm and from 200 µm to 400 µm, respectively. Furthermore, the distance between each two adjacent triangles should be extended from 800 µm to 910 µm. The MO showed a longer effective filtration length and had a higher average shear stress on the effective filtration length. Furthermore, particle motion showed that much fewer particles accumulated on MO. These suggested that MO is with a better antifouling capacity, which was in accordance with particle accumulation potential that was defined specifically and experimental results. This study showed that it is feasible to apply CFD and a smart algorithm to explore an optimum patterned membrane, which is more efficient than an approach that completely bases on experimental works.

Keywords: CFD, membrane fouling, patterned membrane, smart algorithm, microstructure