[FLASH-USERS] confused about st_energy parameter in Stir unit

[Slavin, Jonathan]

Hi Christoph,

Thanks for your detailed reply. That clears things up substantially, though
as you say, some experimentation will be required. I do think that the
documentation could be clearer on this point.

Regards,
Jon

On Thu, Jan 3, 2019 at 5:22 PM Christoph Federrath <
christoph.federrath at gmail.com> wrote:

>
> Hi Jon,
>
> st_energy is meant to be the specific (per unit mass) energy injection
> rate, i.e., also per time (usually referred to as epsilon in the
> literature; see e.g., Frisch 1995). However, in practise, one normally
> wants to set a target turbulent Mach number, as this is usually what
> determines the physical behaviour of the system and can be compared to
> previous studies. Setting the Mach number requires some experimenting with
> st_energy and noting that it is proportional to Mach^3 (see notes in
> source/Simulation/SimulationMain/StirFromFile/forcing_generator/forcing_generator.F90).
> Since the resulting Mach number depends on many factors, most importantly
> st_energy, but also the mixture of solenoidal and compressive modes as well
> as the dissipation properties of the numerical hydro/MHD solver, some
> experimenting and iterating (setting st_energy close to what you think is
> required for your proposes and then measuring the Mach number;
> source/Simulation/SimulationMain/StirFromFile/IO_writeIntegralQuantities.F90
> does that in column 9 of the resulting .dat file and adjusting st_energy
> accordingly in a follow-up run) is required to get to the target Mach
> number. Hope this helps; please let me know if you have any questions on
> the StirMain/FromFile unit.
>
> Best regards,
> Christoph
>
> On 4 Jan 2019, at 06:51, Slavin, Jonathan <jslavin at cfa.harvard.edu> wrote:
>
> Hi,
>
> I'm looking to use the Stir unit in a simulation using the FromFile
> implementation. I've been struggling a bit with setting the input
> parameters.  I think I understand most of them, but the st_energy one is
> confusing me.  When I follow through the code it seems like the only
> dimensional parameters used in setting the acceleration are st_aka and
> st_akb which in turn are proportional to st_OUphases.  Those are
> proportional to the variance used to create them which is proportional to
> st_OUvar. Finally st_OUvar = sqrt(st_energy/st_decay).  So, if I've
> followed that correctly the acceleration is proportional to
> sqrt(st_energy/st_decay).  It seems that st_decay has units of time (s),
> and acceleration should of course have units of distance/time**2 (cm/s**2),
> which leads to st_energy having units cm^2/s^3, which seems odd to me. I
> can see dividing out the mass, which leaves cm^2/s^2, but the extra s
> dimension is strange. Should st_energy be interpreted as power per unit
> mass added by stirring?
>
> Regards,
> Jon
> --
> Jonathan D. Slavin
> Astrophysicist - High Energy Astrophysics Division
> Center for Astrophysics | Harvard & Smithsonian
> Office: (617) 496-7981 | Cell: (781) 363-0035
> 60 Garden Street | MS 83 | Cambridge, MA 02138
>
>
>

-- 
Jonathan D. Slavin
Astrophysicist - High Energy Astrophysics Division
Center for Astrophysics | Harvard & Smithsonian
Office: (617) 496-7981 | Cell: (781) 363-0035
60 Garden Street | MS 83 | Cambridge, MA 02138
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20190104/f67c95f2/attachment-0001.htm>