[FLASH-USERS] how to force refinement

[Klaus Weide]

On Mon, 26 Apr 2021, Slavin, Jonathan wrote:
> I think you may be right. I had thought that by using gr_lrefineMaxByTime
> and setting a higher value of for gr_lrefmaxTimeValue (i.e. a higher value
> than the initial one) that the value of lrefine_max would be increased.
> However that seems not to be happening. I've used gr_lrefineMaxByTime in
> the past to decrease the maximum refinement, but it seems that it doesn't
> work in the opposite direction.

Yes, no matter the values of other runtime parameters and 
variables, lrefine_max is always a hard limit (enforced by the 
lower-level PARAMESH routines), and its value does not change
during the simulation.

Other variables, like gr_maxRefine, can change during the simulation and 
can restrict the effective maximum refinement further (depending on
what is done in Grid_markRefineDerefine).

> I'm not sure what happens to lrefine_max, but when I print it out, I got
> some odd value (I printed with format i5 and got ****. Of course I could've
> printed it with default formatting.)

Make sure 'lrefine_max' refers to the variable in the correct PARAMESH 
module.  You should have something like

  use tree, ONLY : lrefine_max

in effect, in the location where you try to print the value.

Maybe you were trying to print before the variable 'lrefine_max' got 
initialized from the runtime parameter 'lrefine_max'? (see Grid_init.F90)

Klaus