Cloud-Cloud Collision Simulations

Our cloud-coud collision simulations attempt to capture the more realistic scenario in which low-density, atomic / quasi-molecular clouds collide in spiral arms to produce high density molecular clouds. Although the inital clouds are seeded with some low-level turbulence, this does not decay as much as in the isolated clouds, as the collision pumps kinetic energy into the gas. These simulations are performed with Arepo, and thus cover both magnitised and non-magnitised clouds.

Technical details

Physial properties in initial conditions

The clouds have 104 solar masses, initial density of 10 cm-3, and a radius of 19 pc. They are stirred with a turbulent field with power law scaling k-4, comprising purely solenoidal modes. The turbulence has a mean velocity dispersion of 1 km/s. This turbulence is roughly transonic at our initial temperature of ~ 200-300 K. Clouds collide head on (along the x axis), which each cloud having a velocity of 3.75 km/s towards the other. The clouds are initially 19pc apart. The magnetic field is set to be Bx = 3 μG, and zero in the other dimentions. The clouds are embedded in the WNM with mean density of 0.1 cm-3.

For the CO chemistry, we use the Nelson and Langer (1999) reduced network and for the H2 chemistry we use the prescriptions in Glover and Mac Low (2007). More details can be found in Glover and Clark (2012b). The clouds start with a small ionisation fraction of H+ = 0.01, to account for the ionisation by cosmic rays, but otherwise the chemistry starts off with the hydrogen in the form of atomic hydrogen, carbon in the form of C+, and oxygen in its neutral form.

In the simulations listed below, we vary 1) the strength of the interstellar radiation field, quoted in Habing units, and expressed via G0, 2) the cosmic ray ionisation rate (CRIR), which are assumed to be free-streaming in the simulations 3) the inital turbulent velocity field, to create different 'clouds'. We will add to the list of simulations below as more become available.


Sim 1: G0 = 1.7 and CRIR = 3 x 10-17 s-1 and turbulent seed 1

The emission (ppv) FITS files: CO (1-0);  [CI] (1-0);   [CI] (2-1);   [CII];   HI (21 cm)

FITS files of physical properties (ppv): density;  temperature


Sim 2: G0 = 5.1 and CRIR = 9 x 10-17 s-1 and turbulent seed 1

The emission (ppv) FITS files: CO (1-0);  [CI] (1-0);   [CI] (2-1);   [CII];   HI (21 cm)

FITS files of physical properties (ppv): density;  temperature


Sim 3: G0 = 17 and CRIR = 3 x 10-16 s-1 and turbulent seed 1

The emission (ppv) FITS files: CO (1-0);  13CO (1-0);  [CI] (1-0);   [CI] (2-1);   [CII];   HI (21 cm);   HI (21 cm) with 100K, 20km/s line

FITS files of physical properties (ppv): density;  temperature

Emission (ppv) FITS for side-on view: Side on CO (1-0);  Side on 13CO (1-0);  Side on HI (21 cm);   Side on HI (21 cm) with 100K, 20km/s line

Original 3D (ppp) cubes that were used for RT: Gas density cube;  Dust temperature cube;  Gas temperature cube;  Magnetic field cube;  Velocity cube;   CO abundance cube;  H2 abundance cube;  H+ abundance cube;  NB: The gas / dust properties are given in CODE units. The IDL reader routines contain the units of mass, length, velocity that can be used to convert these cubes to cgs. The abundances are given relative to the number density of hydrogen nuclei, so the abundance of neutral hydrogen is given by xH = 1 - xHP - 2*xH2.