[FLASH-USERS] Problem with Grid_bcApplyToRegion.F90 file

María Fernanda Clever Uribe clever_fisica at ciencias.unam.mx
Thu Mar 22 21:41:34 EDT 2018


Hi,



I tried Klaus' recommendation about the j index but I still don't get what
I wish for.



At the moment, following the Grid_bcApplyToRegionSpecialized.F90 boundary
condition from de ShuOsher problem, I am able to inject a wind but from all
the boundary, and I am not able to only inject the wind from a small region.



In the Grid_bcApplyToRegionSpecialized.F90 I don’t understand how I can
limit the condition to only a small part of the boundary. The counter i
goes from 1 to guard (where guard=4), and the counter j goes from 1 to je
(where je=4).



  je=regionSize(SECOND_DIR)

  ke=regionSize(THIRD_DIR)

              do i = 1,guard

                 do j=1,je

                    if (gridDataStruct==CENTER) then

                       if (ivar == DENS_VAR) regionData(i,j,1:ke,ivar) =
sim_rho

                       if (ivar == VELX_VAR) regionData(i,j,1:ke,ivar) = 0.

                       if (ivar == VELY_VAR) regionData(i,j,1:ke,ivar) =
 sim_v

                       if (ivar == VELZ_VAR) regionData(i,j,1:ke,ivar) = 0.

                       if (ivar == PRES_VAR) regionData(i,j,1:ke,ivar) =
sim_p

                       if (ivar == GAMC_VAR) regionData(i,j,1:ke,ivar) =
sim_gamma

                       if (ivar == GAME_VAR) regionData(i,j,1:ke,ivar) =
sim_gamma

                    endif

                 end do

              end do



Do I have to replace the regionSize defined in the Grid file and use
blkLimitsGC instead, or maybe use endPoints?



Or can I limit the region of the wind-injection boundary by setting some
specific parameter in the user_bnd.F90 and flash.par (something like
endPoints=?, or maybe regionSize=?).



Could anyone help me, or explain me how these arguments work in order for
me to solve this issue?

Thanks!


C.

2018-03-21 12:42 GMT-06:00 María Fernanda Clever Uribe <
clever_fisica at ciencias.unam.mx>:

> Hi;
>
> You're right, I didn't notice the index. I will try this out.
>
> Thanks!
>
>
> C.
>
> 2018-03-21 7:51 GMT-06:00 Klaus Weide <klaus at flash.uchicago.edu>:
>
>> On Tue, 20 Mar 2018, María Fernanda Clever Uribe wrote:
>>
>> > I also found an example of Grid_bcApplyToRegion.F90 and I thought that
>> this
>> > could be the file I have to use instead of Grid_bcApplyToEdge.F90,
>> because
>> > I want my boundary conditions to be only in the left corner of my
>> domain. I
>> > took the example of this file from the Grid directory and I customized
>> it.
>>
>> Yes, that makes sense.
>>
>> > I wrote a restriction over -X boundary where the jet has to be ejected,
>> but
>> > I get a blast wave in all the cells in Y axis. I attached a picture of
>> my
>> > results and also the file that I'm using.
>>
>> The condition your wrote
>>
>>                  if (i .eq. (guard-1.0)/guard) then
>>
>> does not make sense to me.
>>
>> It seems that you want different values for different valus of the SECOND
>> index of the regionData array. Currenly the whole range 1:je is treated
>> identically. It seems you have to write an explicit loop like
>>
>>    DO j=1,je
>>
>> and do different things inside the look depending on the j.
>>
>> Klaus
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20180322/bfebc15d/attachment-0001.htm>


More information about the flash-users mailing list