A discussion about the velocity distribution commonly used as the boundary condition in surface acoustic wave numerical simulations

Abstract

Surface acoustic waves in combination with microfluidics has become an attractive research field regarding its various medical and biological applications. It is sometimes preferred to solve just the fluid domain and apply some boundary conditions to represent other components rather than performing a coupled numerical solution. To account for the piezoelectric actuation, a conventional velocity distribution built by superposing the left-going and right-going surface waves is commonly used as the boundary condition, its correctness is assessed here by comparing it to a coupled solution. It was shown that the actual leaky surface acoustic wave in coupled solution has different wavelengths in its real and imaginary parts, sometimes gets out of being sinusoidal, and has a different form compared to the superposed formula. For the phase differences other than 0 and π between the left and right electrodes, the distance between the electrodes affects the streaming and acoustic fields in the microchannel thereby leading to deviations in particle traces. Furthermore, the ratio of the horizontal to vertical components of the surface wave was extracted from the coupled solutions and compared to its previously reported values. The sensitivity analysis showed that for small particles, this ratio does not affect the streaming pattern but changes its velocity magnitude causing a time lag. For larger particles, the ratio altered the movement direction. This study suggests not replacing the piezoelectric actuation with the boundary condition to avoid inaccuracy in resulting fields that are being used in calculations of particle tracing and acoustic radiation forces.