[FLASH-USERS] energy inaccuracy in Sedov simulation

[Robert Fisher]

Hi Jonathan, Marissa, Tomek :

  Allow me to amend my prior statement -- for the point-blast case, the
code is defining the pressure of the point-like initial condition as

  sim_pExp    = (sim_gamma-1.) * sim_expEnergy / sim_vctr

in Simulation_init.F90. The parameter sim_vctr is in turn defined as the
volume, area, or length within a region of sim_rInit in 3D, 2D, and 1D,
respectively. The total internal energy [image: \int e_{\rm int} dV] = [image:
\int p / [(\gamma - 1) \rho] dV] should identically return a total energy
of sim_expEnergy, *for an initial density of 1, on a continuous domain*.
Because of discretization, however, this will not be rigorously true, but
you should be able to achieve as close an approximation as you want by
increasing the effective resolution of the initialization -- that is,
either increasing the sim_rInit parameter or adding more resolution.

  All of the previous caveats for point-like initialization still apply.

  Best wishes,

  Bob

On Tue, Jul 26, 2016 at 11:25 AM, Marissa Adams <madams at pas.rochester.edu>
wrote:

> We understand that. The total energy is less than the input value. This is
> the problem.
>
> On Tue, Jul 26, 2016 at 11:23 AM, Tomasz Plewa <tplewa at fsu.edu> wrote:
>
>> The relevant data is the total energy shown in the first line of the dat
>> file. That should stay constant with relative accuracy better than 1e-10.
>>
>> The actual initial value of the total will vary as one changes problem
>> parameters.
>>
>> Tomek
>> --
>>
>> On 07/26/16 10:30, Slavin, Jonathan wrote:
>>
>> Hi all,
>>
>> I've been running supernova remnant evolution simulations.  The
>> initialization is done with code taken from the supplied Sedov simulation
>> with certain small changes.  One issue that I've run into is that the total
>> energy is not very accurate.  That is, though the energy is conserved, the
>> total energy, initially input as thermal energy, is below that which is
>> given in the sim_expEnergy parameter by ~5%.  Has anyone else run into
>> this?  I don't think that it has to do with any of the changes that I've
>> made to Simulation_initBlock or Simulation_init.  I'm wondering if
>> increasing the sim_nsubzones parameter might help.  Any advice would be
>> appreciated.
>>
>> Thanks,
>> Jon
>>
>> --
>> ________________________________________________________
>> Jonathan D. Slavin                 Harvard-Smithsonian CfA
>> jslavin at cfa.harvard.edu       60 Garden Street, MS 83
>> phone: (617) 496-7981       Cambridge, MA 02138-1516
>> cell: (781) 363-0035             USA
>> ________________________________________________________
>>
>>
>>
>
>
> --
> Marissa Adams                         E-mail: madams at pas.rochester.edu
> Graduate Student                      Ph: (585) 402-5779
> Department of Physics & Astronomy     Website:
> http://www.pas.rochester.edu/~madams/
> University of Rochester
> 478 Bausch & Lomb Hall
> P.O. Box 270171
>
> --
> Dr. Robert Fisher
> Associate Professor / Graduate Program Director
> University of Massachusetts/Dartmouth
> Department of Physics
> 285 Old Westport Road
> North Dartmouth, Massachusetts 02747
> robert.fisher at umassd.edu
> http://www.novastella.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20160726/4f3eb8a6/attachment-0001.htm>