Computational reduction techniques for numerical vibro-acoustic analysis of hearing aids

Ester Creixell Mediante

Numerical vibro-acoustic analysis of hearing aids is essential for the study of current problems such as feedback paths. Significant computational challenges are encountered when solving numerical problems that require repeated calculations of large complex models. This is a problem of concern in the field of hearing aids, which are composed of a large number of small parts with complex dynamic and acoustical behavior.

This project focuses on developing computational reduction techniques to make these processes practically possible. These methods should be such that the precision of the model is kept at a desired level. The suggested approach is to divide the complex system into a set of physical subsystems, which are described by simplified substructure synthesis models.

Tasks that will benefit from this reduction are topology optimization of hearing aid parts and parametric uncertainty analysis of the vibro-acoustic response.

Supervisors: Jonas Brunskog (DTU), Jakob Søndergaard Jensen (DTU) and Martin Larsen (Oticon A/S)

To be completed: 2017
21 OCTOBER 2020