search

fmihpc / vlasiator

Vlasiator - ten letters you can count on
https://www.helsinki.fi/en/researchgroups/vlasiator
Other
45 stars 37 forks source link

Second layer cells on outflow boundaries have bad moments / veloctiy space contents. #2

Closed ohannuks closed 3 years ago

ohannuks commented 10 years ago

I ran the Shocktest project with the full3dslice branch with outflow set on the x+. The third order solution produced a huge spike at the outflow boundary, giving the last cell a rho value of 7.46e6. The second last cell had a value of 2.75e6.

The second order solution didn't produce the same spike.

shocktest

galfthan commented 10 years ago

Lot's of cahnges have been done, and if I remember this was already not visible anymore some time ago. Let's reopen if encountered again

ykempf commented 10 years ago

Sure it is there, check pressure in the second layer of outflow in Magnetosphere runs (no Boundray_type=1 threshold), you'll be surprised. My hteory is that there is no block cleanup done in these cells but there is creation of neighbours to the layer 1 and inner cells close enough.

galfthan commented 10 years ago

Ok! re-opening

ykempf commented 10 years ago

At least it looks similar to the -x outflow in MS simulations...

ykempf commented 3 years ago

This has been fixed.