nextuppreviouscontents
Next:Study of the amplificationUp:The case with advectionPrevious:Natural configurationContents

Reasons of failure

The simulation didn't give us the hoped results. Not any instabilities developed. The reason of this failure are :
  1. We are not in the conditions of development of the instability.
  2. A shearing layer exists, so the theoretical study of stability is not more valid
  3. The absolute instability of Kelvin-Helmhotz is became a convective instability.
The boundary conditions are too proper to allow the development of an instability. A convective instability just amplifies and convies the perturbation. So, as the boundary conditions are very proper, the observation of instabilities is very hard.
There are two possibilies :
  1. The use of a very long domain to let the computationnal errors initialize the instability
  2. The introduction of a little perturbation directly in the boundary condition
We haven't had the time to test this two possibilities, but the introduction of a perturbation is a good representation of what happends in a real situation. So the study of this case would be a interesting thorough examination.


nextuppreviouscontents
Next:Study of the amplificationUp:The case with advectionPrevious:Natural configurationContents
Stephanie Terrade

Julien Delbove
2000-11-06