[FLASH-USERS] small numbers in guard cellsod

[Klaus Weide]

On Tue, 23 Sep 2014, you wrote:

>I am trying to setup a simulation using FLASH code. It runs well in a test 
>problem, Sod Shock-Tube, using the original initial parameters. But when I 
>run it using physical parameters in units of "cgs", it raises a warning:

Hi,

Some remarks without looking at any details of your failures:

You have lowered your initial values of density and pressure by many
orders of magnitude, with respect to the default flash.par. That
should be fine, IF you also adjust the values of several runtime
parameters that are used as lower bounds on some variables.
In particular, you should set:

  smlrho     # threshold value for density
  smallp     # threshold value for pressure
  smallT     # threshold value for temperature
  smallE     # threshold value for energy

These should usually all be significantly smaller than the smallest
values of the corresponding variables that you expect to occur
physically.

If you do not set them, default values of 1.E-10 will be used, which
is clearly not appropriate for your application.

You may also have to adjust other runtime parameters to your application.
In particular, the timestep-related parameters
   dtinit
   dtmin
   dtmax
come to mind.

Finally, I recommend testing with lower resolution first (NOT lrefine_max=8)
in order to sort out those issues.

Klaus