Closed marcusfuchs closed 8 years ago
I was just thinking about the same, since I always get a warning that this value is not initialized and hence set to 0. Maybe it would be more logical, at least if a starting value for the mass flow is known, to set tau
to length/v
, but I guess the proposed solution is more than adequate.
Closed by #45
In one of my applications, I see significant speed-ups for initialization and simulation just be setting
start=0
fortau
,inp_a
andinp_b
inPDETime_massFlow
. I'll push this to a branch and will issue a PR. If this has no negative effects for anyone else, I think this is a useful improvement.