In the driftingsnow branch, we establish an adaptive time step for the explicit snow drift scheme in order to simultaneously minimize computational expense and preserve numerical stability.
Currently we have implemented a particle slow down term. However, this has not yet been accounted for when calculating the sub time step. In the case of a constant slowdown, this could be fixed with
double sub_dt = std::min(C_max * dx / (sqrt(2.) * grid_VW.grid2D.getMax() - particle_slowdown), dt); // Sub time step
or in the case of a multiplicative slowdown
double sub_dt = std::min(C_max * dx / (particle_slowdown * sqrt(2.) * grid_VW.grid2D.getMax()), dt); // Sub time step
In the
driftingsnowbranch, we establish an adaptive time step for the explicit snow drift scheme in order to simultaneously minimize computational expense and preserve numerical stability.Currently we have implemented a particle slow down term. However, this has not yet been accounted for when calculating the sub time step. In the case of a constant slowdown, this could be fixed with
or in the case of a multiplicative slowdown