Cross-Scale Modeling and Optimization of a Cantilever-Type Micro/Nano Valve Based on Nonlocal Stress Theory

Cantilever-type micro/nano valve has been an essential component in micro/nano fluidic networks, how to model and optimize the valves at micro/nano scales remians a critical problem. This paper aims at developing a modeling and optimization approach for the valves, which is suitable at different size scales. Firstly, a model for micro/nano valves is developed by a continuum approach based on the nonlocal stress theory. The model could well represent size-dependent geometrically nonlinear characteristics of the valves at different size scales. Secondly, the properties of the model are discussed by numerical simulations to demonstrate the efficiency of the model for describing mechanic behaviors of the valves at different size scales. Finally, a differential evolution algorithm based multi-objective optimization scheme is employed to optimally determine the structure parameters of the valve. Pareto optimal solutions are consequently obtained, and an optimal nanovalve of 66.9 nm thickness with 1552 kHz natural frequency and 33.6 nm maximum transverse displacement is presented. All the results demonstrate that the proposed cross-scale modeling and optimization approach is efficient for the determination of structure parameters of cantilever-type micro/nano valves.