Numerical simulation of a 2D electrothermal pump by lattice Boltzmann method on GPU

Electrothermal flow in a microfluidic system is a fast-developing technology because of the advancement in micro-electro-mechanical systems. The motion is driven by the electrothermal force generated by the AC electric field and non-uniform temperature distribution inside the system. Electrothermal force can be explored for pumps in microfluidic systems. In this paper, the lattice Boltzmann method (LBM) is used to simulate a 2D electrothermal pump. As an alternative numerical method for fluid dynamics, LBM has many advantages compared with traditional CFD methods, such as its suitability for parallel computation. With its parallel characteristic, LBM is well fitted to the parallel hardware in graphic processor units (GPU). To save computational time in parametric studies, a CUDA code was developed for executing parallel computation. The comparison of computational time between CPU and GPU is presented to demonstrate the advantage of using GPU. The effects of the frequency, thermal boundary conditions, electrode size, and gap between electrodes on volumetric flow rate were investigated in this study. It was shown that LBM is an effective approach to studying 2D electrothermal pumps on a CUDA platform.