[FLASH-USERS] ERROR After calling Eos

[Klaus Weide]

On Wed, 25 Aug 2010, Sean Lynch wrote:

>  *** Writing values calculated from Simulation_init ***
>  sim_rhoCloud      :   1.61000000000000003E-016
>  sim_tCloud        :    20.000000000000000     
>  sim_pCloud        :   2.67727285894459990E-007
>  sim_rhoAmbient  :   1.61000000000000000E-019
>  sim_tAmbient    :    20000.000000000000     
>  sim_pAmbient    :   2.67727285894459990E-007

Sean,

I notice that your densities are smaller than smlrho, which is used as a 
floor value for densities in various places in the code.  You should 
probably give the smlrho runtime parameter a value that is smaller than
tha smallest density value you expect to see in your domain.  Also
check other small* runtiem parameters.

I don't know whether at has anything to do with the error below.


>  ERROR After calling Eos, eosData(EOS_EINT) or eosData(EOS_DENS) are zero
>  iflag=           1           1           1           1           1           1           1           1
>  solnData(EINT_VAR,ib:ie,jb,kb)=   0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.000000000application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> 0000000        0.0000000000000000        0.0000000000000000     
>  eosData (eint+1:eint+vecLen)  =  1.00000000000000004E-010  1.00000000000000004E-010  1.00000000000000004E-010  1.00000000000000004E-010  1.00000000000000004E-010  1.00000000000000004E-010  1.00000000000000004E-010  1.00000000000000004E-010
>  solnData(DENS_VAR,ib:ie,jb,kb)=   0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000     
>  eosData(dens+1:dens+vecLen)   =   0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000        0.0000000000000000     
.....

> > - From where is this Eos_wrapped call made?
> >  Is this the very first such call?
> 
> The failure is during the very first call to Eos_wrapped. This call is made from gr_expandDomain

So this is happening at a point when only one block exists, a root block 
(of 8 by 8 cells) covering your entire domain.  Simulation_initBlock has 
been called on this block #1 (also from gr_expandDomain), right before the 
Eos_wrapped call.

Does your Simulation_initBlock perhaps not handle this correctly, 
because it does not expect to be invoked on such a coarse block?

If the above does not help, check whether the content of the block is as 
expected BEFORE the 'call Eos' in Eos_wrapped - put some print statements 
in Eos_wrapped (is suggest between 'call Eos_getData' and 'call Eos'), 
etc.

Klaus