News and Blog

EPSRC CDT in Next Generation Computational Modelling

Applications of Computational Modelling in Acoustics and Vibration

Dr. Cheer explaining Cochlear mechanics.

Dr. Cheer explaining Cochlear mechanics.

This mornings seminar was given by Dr. Jordan Cheer about the role various roles of the Institute of Sound and Vibration Research (ISVR) research group at the University of Southampton.

The ISVR group covers 4 main areas of research:

  • Acoustics
  • Dynamics
  • Signal Processing and Control
  • Human Sciences


The Acoustics group covers many areas of research such as transport noise, audio engineering and sound perception. There are two key computational methods used to model these areas. Firstly, Finite Element Methods (FEM) is used, which can perform calculations for large domains and high frequencies. This is useful as humans hear noise up to a high frequency. It can also provide an accurate representation of the geometry and mean flow. The second main area of research in the Acoustics group uses Boundary Element Methods (BEM) for sound propagation in flow. An industrial example of this is modelling the noise radiation from an engine intake.


A key area of research of the Dynamics group is ground vibration in railway noise. This is because of new high speed rails causing large vibrations. The large vibrations are due to the speed of the train being similar to the wave speed of the sound propagation. Furthermore soft ground can cause additional vibration problems. Hence a non-linear track/ground model is implemented to model the vibration waves caused by the train in time. Also of interest is the modelling of noise vibrations from tunnels. This is of particular interest in areas such as Glasgow where the vibration from underground trains could effect historic buildings.

Signal Processing and Control

It is of great interest to control the level of vibration in systems. This may be due to noise pollution, fatigue or other problems. Therefore active vibration control solutions are developed based on the structural and acoustic response of the system. The results of the vibration control cause the dominant frequencies to be reduced. However it is possible for other, less dominant frequencies to be enhanced.

Output of model with control vs no control.

Output of model with control vs no control.

Human Sciences

Finally acoustical modelling is used for medicinal purposes, in particular Cochlear Mechanics and the effect of a cochlear implant. This is done by modelling the physical mechanisms of normal hearing and how the brain processes the signals. This helps to understand deafness and develop aids for the hearing impaired. As the ear is not merely a structural system the model must incorporate other couplings such as:

  • Mechanoelectric coupling
  • Electromechanical coupling
  • Fluid-structural coupling

Another challenge of hearing aids is the physical implants. It is found that slight variations in the position of the implant can cause relatively large variations in voltage supplied to the brain. Therefore hearing is not improved as it should be.