Simulation of MEMS devices using linearized compressible flow equations on moving domains

A new method of sound generation - which we call Advanced Digital Sound Reconstruction (ADSR) - is based on overlapping redirected sound pulses to form an audio signal. The redirection of these sound pulses can be achieved with highly dynamic shutter gates, where multiple cells equipped with shutter gates are actuated consecutively contributing to the overall audio signal. The application of ADSR focuses on Micro-Electro-Mechanical-Systems (MEMS) speakers, hence, simulating such devices requires the consideration of viscous effects. Furthermore, the movement of the shutter gates is vital for the operating principle, therefore, we use the linearized compressible flow equations formulated in an arbitrary Lagrangian-Eulerian (ALE) framework to incorporate all necessary effects. The movement of the domain itself is governed by an artificial quasi-static mechanical computation where a simple forward-coupling provides the new geometry for the linearized compressible flow equations. To keep the computational cost minimal, we couple to the acoustic wave equation as soon as viscous effects can be neglected. Finally, the introduced simulation framework is applied to a certain embodiment in order to simulate the operating principle of ADSR and to evaluate the influence of crosstalk between the individual cells generating sound pulses.