HiFiLES / HiFiLES-solver

High Fidelity Large Eddy Simulation Solver
Other
173 stars 131 forks source link

Diverging #20

Closed sethhensel closed 10 years ago

mlopez14 commented 10 years ago

Likely the time step is too large. When starting from uniform flow conditions, the boundaries will generate strong compression waves.

While the initial time step might be very small, it is possible to restart the computations, once the transients have disappeared, with a larger time step.

In solving the Navier-Stokes equations, we suggest using a time-step of the order of

(h/N^2) (C/ ( abs(u) + abs(c) ) + N^2 mu/h ) )

where h is the cell size, u is the local speed of the fluid, c is the speed of sound, N is the order of polynomial used in the simulation, mu is viscosity of the flow, and C is a constant that depends on the time-stepping scheme.