A MEMS loudspeaker based on Advanced Digital Sound Reconstruction

Publikation: StudienabschlussarbeitDissertation

Abstract

The underlying concept of loudspeakers has not changed since its invention in the 1910s. Due to miniaturization and the rise of Micro-Electro-Mechanical-Systems (MEMS), the requirements for loudspeakers are ever-increasing, exhausting the possibilities of the established principle. A completely different approach is given by ultrasound pulse-based (USPB) sound generation principles like Advanced Digital Sound Reconstruction (ADSR), which focuses on MEMS applications. Here, the sound is generated by overlapping short sound pulses to reconstruct an audio signal. Due to the involvement of fast-switching actuator parts on the micro-scale, the requirements for a simulation environment are entirely different from the well-known tools. This thesis establishes an appropriate simulation workflow for USPB sound generation principles based on the Finite Element method (FEM). With this simulation framework viscous acoustics on moving domains can be accurately computed. It has been embedded in a broader existing framework to enhance its capabilities further by including additional displacement-dependent forces like the electrostatic force or a contact force.
Linearized compressible flow equations in an Arbitrary-Lagrangian-Eulerian setting are used to simulate USPB sound generation principles efficiently and accurately. Combined with a quasi-static mechanical model (linear elasticity) governing the domain movement, complex sound fields can be simulated.

Due to the small scales involved in the simulation (low µm range), continuum theory might break down. To combat this problem, an extension of the derived framework with Maxwell slip boundary conditions is proposed, effectively extending the viability of the set of equations. Here, multiple operators developed in computational fluid dynamics are revisited and tested for acoustic applications. Additionally, a Lagrangian-type formulation and a Mortar-like formulation are discussed.

The proposed framework is benchmarked with convergence studies, stability tests, and measurement comparisons, validating its applicability. Using the validated numerical framework, multiple actuators using ADSR and its sub-principles are simulated and results are compared to analytical
estimations. For the sake of completeness, ADSR is also investigated in an experimental setup to prove the principle experimentally. Besides validating the underlying principle of ADSR numerically, a realizable MEMS actuator is presented and simulated. Finally, design guidelines are extracted from the simulations and summarized, which are crucial for further development.
Originalspracheenglisch
QualifikationDoktor der Technik
Betreuer/-in / Berater/-in
  • Kaltenbacher, Manfred, Betreuer
Datum der Bewilligung21 Feb. 2024
PublikationsstatusVeröffentlicht - 2024

Fingerprint

Untersuchen Sie die Forschungsthemen von „A MEMS loudspeaker based on Advanced Digital Sound Reconstruction“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren