[FLASH-USERS] How does FLASH3 calculate temperatures for checkpoint and plot dumps?

[Dave]

Hi Mark,

It looks like it's setting density, pressure *and* eint in 
Simulation_initBlock. It doesn't look like I've got eosmode set 
anywhere, so it should be the default mode_dens_eint ?

The weird thing is that it the second iteration dumps less silly 
temperatures - so it may be that eos isn't being called on all blocks 
during initialization or something?

It isn't the feedback btw - it still seems to be broken if I turn it off 
entirely.

Actually - could you send me your simulation_initblock.f90 ? I want to 
see how/if you include the eos call in there

Cheers,
-Dave

On 16/12/2011 3:55 p.m., Mark Richardson wrote:
> Hi Dave,
>
> Temperature may be set in EOS calls provided two things:
>     The eosmode is NOT set to mode_dens_temp.
>     And either you're setting dens and pressure initially an eosmode is
> set to mode_dens_pres OR you're setting the density and energy
> initially with eosmode set to mode_dens_eint.
>
> Note, if you are setting dens and energy, then I suspect they should
> be right as your pressure is set the same way as your temperature.
>
> If you are somehow setting density and pressure then this would
> explain it as te default eosmode is mode_dens_eint.
>
> Finally, if you are adding feed back etc in by hand, you'll want to
> finish that off with a EOS call. I can send a file that does this (my
> simulation_initBlock), and in fact you should be setting it at your
> initialization, too.
>
>     -Mark
>
> Sent from my iPhone
>
> On 2011-12-16, at 2:38 PM, Dave<david.john.williamson at gmail.com>  wrote:
>
>> Hi,
>>
>> When examining the checkpoint file from a simple test run of a galaxy, I'm finding that the pres and dens values are what I expect, but the temp values seem to be way off. In an (unrealistic) supposedly isothermal test, the outer parts of the galaxy are ~10^7 K, as they should be, but there is a sharp discontinuity in the centre, where the temperature drops to ~10^7 K, according to the temp scalar - even though the pres and dens values are what they should be.
>>
>> I'm trying to track down where in Flash the temperature is calculated from internal energy etc for output, to determine if this is just being dumped out wrong, or if the internal energy itself is wrong. Any idea where that goes on?
>>
>> Cheers,
>> -David