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
When tracking down some other behavior I noticed weirdness in our WENO5-Z implementation: the alpha_lin value (sigma in the paper) can transiently become slightly larger than 1.0. This should not be happening. I'm not sure yet what the implications are (there are no obvious signs of ill behavior when this happens). If I recall correctly, the shock tube is sufficient to trigger this. The relevant variable is here.
The text was updated successfully, but these errors were encountered:
When tracking down some other behavior I noticed weirdness in our WENO5-Z implementation: the
alpha_lin
value (sigma
in the paper) can transiently become slightly larger than 1.0. This should not be happening. I'm not sure yet what the implications are (there are no obvious signs of ill behavior when this happens). If I recall correctly, the shock tube is sufficient to trigger this. The relevant variable is here.The text was updated successfully, but these errors were encountered: