You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is obviously not a very likely scenario, as we run on meshes much bigger than this, but it may be a useful case for testing computational intensive modules (and writing unit tests etc).
The crux of the problem seems to come down to a check that, for each dimension, does:
const float px = (nx>1) ? g->rdx : 0;
If n_=1 it will set p_ to be 0. This is then used in a division (alphadt = 0.3888889/( px*px + py*py + pz*pz ); ) so inf/nans propagates throughout anything that touches the fields.
The text was updated successfully, but these errors were encountered:
This is obviously not a very likely scenario, as we run on meshes much bigger than this, but it may be a useful case for testing computational intensive modules (and writing unit tests etc).
The crux of the problem seems to come down to a check that, for each dimension, does:
const float px = (nx>1) ? g->rdx : 0;
If
n_=1
it will setp_
to be 0. This is then used in a division (alphadt = 0.3888889/( px*px + py*py + pz*pz );
) soinf/nans
propagates throughout anything that touches the fields.The text was updated successfully, but these errors were encountered: