open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

in the dico file the non-conservative PSI scheme (former nr.6) isn't listed as selectable scheme anymore in the new keyword SCHEME FOR ADVECTION OF VELOCITIES.
N.b., allocating the number 6 this scheme still seems to work.

I'm curious if it is considered as obsolete now and why?
As far as my experience goes, it is the only FE scheme, besides the MOC scheme, which works well for dam-break type flows or e.g. flow over a weir involving high Froude numbers and transitions between subcritical and supercritical flows. With the other options for simulating such flows like the schemes 3(4) or 13(14) I had always bad experiences although the manual says: "For dam break studies 14;5 is recommended." My experience with the latter schemes is rather different in that they work well for flows incorporating low Froude numbers by producing lower numerical diffusion comparatively.

I would be glad for some elucidations!

Best regards,
Clemens

Hello Clemens,

Yes, we do not like too much keeping a scheme that is not mass-conservative, except the MOC, however I did not know (or remember) that it had been removed in the dictionnary. We indeed thought that the NERD scheme (14) would be better, as it is mass-conservative and works on tidal flats. I think that scheme 6 is OK for velocities, but should not be used for a tracer.

With best regards,

Jean-Michel Hervouet