In: f90bc8b
For tau = 0 cases, the splitting has not been carried out correctly, and is just applying f = f0 at the end of the FVM step in a single timestep.
Proposed Fix:
Considering strang splitting:
- Solve df/dt + A_q df/dq + A_p df/dp = 0 for dt/2
- Apply f = f0
- Solve df/dt + A_q df/dq + A_p df/dp = 0 for another dt/2
In: f90bc8b
For tau = 0 cases, the splitting has not been carried out correctly, and is just applying f = f0 at the end of the FVM step in a single timestep.
Proposed Fix:
Considering strang splitting: