At that time there
was an interest in internal wave propagation in the ocean and the atmosphere.
Internal waves exhibit nonlinear behavior and it was natural to describe them by
The "wave-codes" developed for the internal waves were
based on a spectral collocation method which appeared attractive due to its high
accuracy. The activity of the group consisted of developing numerical solvers
and analyzing the output of simulations. A texture-based volume render called
Viz was developed and used to explore numerical data.
The limitation of the spectral methods to box-like geometries motivated us to
explore numerical methods that could cover more complex geometries. At the same
time it was evident that more applied and realistic problems contained
dissipation. In particular, relevant engineering problems involved high
Reynolds-number flows and turbulence. To get some geometric flexibility and high
accuracy we chose the spectral element method (SEM) and started to develop a
Navier-Stokes solver based on that numerical method.
As computers grew bigger and numerical algorithms became more efficient, the
interest in solving fluid mechanical problems using numerical techniques gained
momentum and new applications came up. We have been involved in various
applications, including aero/hydro dynamics applied to airplanes and underwater
vehicles, noise generation by turbulence applied to acoustic sensors, and
aerosol dispersion in the environment or the human airways.
We concentrate our efforts within applied and basic fluid dynamics, both
experimental and computational, using the SEM code and Finite Volume codes.
Turbulence models are improved by for example utilizing structure based tensors.
Tools for analysis and visualization fluid data both from experiments and
simulations are developed and optimized. New functionality is added and built
into the texture/GPU based volume render called Voluviz.