wiki:CoolingTurbulence

Version 34 (modified by trac, 12 years ago) ( diff )

Cooling Turbulence Wiki Page

Back of the envelope

Isotropic forcing - or acceleration at a length scale of material at a density will drive turbulent motions at a velocity that will dissipate in a crossing time giving a heating rate

In the absence of cooling, the turbulence would raise the temperature to the kinetic temperature of the turbulence in one crossing time - and the turbulence would become sub-sonic.

However, balancing this heating is a cooling term .

In addition to the dynamical time, we now have a cooling time for the shocked material. .

Of key importance is the dimensionless ratio where

If the cooling time is much less than the dynamical time - then shocked gas will quickly cool and the shocks will seem isothermal. If the cooling time is much longer than the dynamical time - then the gas as a whole will heat up.

If we plot this ratio as a function of turbulence velocity and the product of the (number) density and the driving scale we get the following image

Contour Plot

Plot showing ratios of cooling time to turbulent crossing time Note we've labelled our three runs 'A','B', & 'C'

We can also consider the combined effect of the global heating by the turbulence and the global cooling of the ambient gas. Combining the heating and cooling we arrive at which simplifies to . This gives an equilibrium temperature of for and a relaxation time of . Assuming that is an increasing function of which it is for the system will always evolve towards where .

Finally once the turbulent gas has reached the equilibrium temperature, the ratio will determine whether the turbulence will ultimately be supersonic or subsonic. We can characterize this by

The following image shows the same contours as before although over an extended range in the y-direction. Behind the contours however is a color map showing the mach number of the turbulence at the equilibrium temperature. The missing values imply that the turbulent is outside of the range of the cooling function - and no equilibrium solution exists…

Contours with Equilibrium Mach Number

Results

Results

Running three preliminary runs…

Description Run
Too Hot A 2.8e3 .28 1.0e1 1.0e4 100 18.2 9.6e4 6.6e-22 7.3e5 2.6e2
Too Cold B 1.0e-2 2.9 2.5e3 1.6e5 400 72.9 1.4e4 4.3e-23 6.6e2 6.4e4
Just Right C 8.5e-1 .55 2.0e1 4.0e4 400 36.5 9.8e4 6.7e-22 8.8e2 1.0e3
High Mach D 1.0e-5 10. 1.6e4 4.2e4 4e5 37.5 293 9.1e-27 8.3e2 8.0e7

Run A

The first run was well below the line where the cooling time equals the crossing time. By frame 35 the ambient has heated up and by frame 118 - it is clear that everything is hot and subsonic.

Frame 35

Frame 118

Run B

Run B is strongly cooling.

Frame 35

Run C

Run C is moderately cooling.

Frame 35

Column Density

Spectra

Frame 154

This image shows the velocity spectra (left) and density spectra (right) for runs A (red), B (green), and C (blue). Spectra of velocity and density from frame 154 (movie)

PDFs

The image below shows 2D joint pdf's of [v vs. T] (top row), [rho vs. T] (middle row), & [rho vs. v] (bottom row) for Runs A (left column), B (middle column), & C (right column). (movie)

Potential References

  • Seifried, Schmidt, & Niemeyer, Forced turbulence in thermally bistable gas: A parameter study arXiv September 15, 2010. (link)
    • Hmmm, this looks very similar. Probably need to read this one.
    • Enzo, 2563 mostly, with accompanying 1283 and 5123.
    • Also note references.
    • Cold: initial temperature 500—5000 K, driven down to 10s of K, depending on simulation. Also imposed temperature ceiling late in time to keep max speed reasonable. My initial impression is that this is madness, although they do claim to reach a steady-state…

Other refernces:

  • Krause & Alexander, Simulations of multiphase turbulence in jet cocoons, MNRAS 2007, 376, 465. (link)
    • 2D simulations of clumps near a KH interface.
  • Krause, Jets and multi-phase turbulence, Memorie della Societa Astronomica Italiana 2008, 79, 1162. (link)
    • 3D extension of Krause & Alexander 2007
    • 1203 and one 2403 simulation.
    • Box size: 1 kpc.
    • This is the only result on ADS for Abstract: "radiative cooling" AND "turbulence" AND "3D" AND "interstellar medium".
    • If we remove ISM, then there are 24 results, mainly relating to galaxy clusters & the intercluster medium (ICM).

Attachments (10)

Note: See TracWiki for help on using the wiki.