| |
 |
| Paper: |
Towards Understanding Simulated Feedback in AMR and SPH Codes and the Multi-Phase Nature of the ISM |
| Volume: |
453, Advances in Computational Astrophysics: Methods, Tools, and Outcome |
| Page: |
19 |
| Authors: |
Mitchell, N. L.; Bower, R. G.; Theuns, T.; Vorobyov, E. I. |
| Abstract: |
Feedback from supernova is believed to be a key ingredient for
regulating star formation within galaxies, however modelling it
self-consistently is prohibitively expensive. Even superbubbles which
are formed from multiple supernova occuring in close proximity, are
only a few hundred parsecs across — tiny compared to the sizes of many
galaxies. Thus any simulation which aims to study the large scale
properties of galaxies, groups and clusters cannot currently resolve
the ISM into its true multi-phase nature.
In order to overcome this limitation, many cosmological simulations which are
run in both AMR and SPH codes, adopt polytropic equations of
state. These approximate the physics of the ISM below those scales
which can be resolved where the ISM splits to become multi-phase.
However we show that when identical sub-grid physical recipes
for cooling, star formation and feedback are included into both SPH
and AMR codes, they do not necessarily yield the same results. Instead, we find
that energy is dissipated far more readily in an AMR code, allowing
supernova driven winds to stall. This prevents supernova feedback in
AMR simulations from removing sufficient gas to adaquately regulate
the star formation rate. Whereas in SPH codes the winds can
remove more gas, with wind particles able to stream more freely out of
the galaxy. Determining which of these codes provides a more
physically correct description is extremely difficult, however it
clearly highlights the need for a more robust model for the ISM.
For a better understanding of the means by which energy
from feedback is redistributed within the ISM, we present our new
multi-phase chemodynamic model in the FLASH AMR code. We seperate the
ISM into a hot tenuous gas phase and an almost collisionless compact
molecular cloud component. Both phases are modelled on the adaptive
mesh, the hot gas being modelled by using the standard Euler equations
for compressible fluid dynamics whilst the collisionless component is
solved using a similar set of equations based upon the zeroth, first
and second order moments of the collisionless Boltzmann equation (the
stellar component is modelled in an identical fashion). Now
cold molecular star forming material can continue to accrete towards
the galactic centre whilst hot supernova winds can propagate outwards
through the space in between them. Such a model allows us to
investigate the exchange of mass and energy between the different
phases of the ISM along with a multitude of physical processes
including ram pressure, heat conduction and turbulence. |
|
|
|
|
|
