[FLASH-USERS] Question about simulation time interval delta t

[Suoqing Ji]

Hi Siyi,

I can provide you a quick one, and others might have more comprehensive answers.

The time step is determined at each step by the subroutine Driver_computeDt, which computes the timestep required by all of the units used, and take the minimum dt over different units. Say, Hydro_computeDt gives dt1 for hydro unit, and Particles_computeDt gives dt2 for particle unit and so on, then final dt = min(dt1, dt2, …)

How dt is computed for each unit varies. As a classical example, the timestep for hydro obeys the Courant-Friedrichs-Levy condition. Physical quantities in each cell corresponds to a certain value of dt under this condition, and Hydro_computeDt computes dt for each cell, selects the minimum dt over the entire domain, and return it to Driver_computeDt.

Hope this helps!

Best wishes,
—
Suoqing Ji
Ph.D Candidate
Department of Physics
University of California, Santa Barbara
http://physics.ucsb.edu/~suoqing

On May 22, 2017, 7:50 PM -0700, Siyi Yu <siyiyu at lbl.gov>, wrote:
> Dear Flash users and developers,
> Permit-me to ask a little question about the simulation delta t. How the system defines the time interval it takes? When I launch a simulation, it appears that delta t is not a constant and decrease as the simulation goes on. Could I know why please?
> Best
> Siyi YU
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20170522/1c8c69f2/attachment.htm>