Giant gas streams in the mass media

More detailed article about the giant gas stream now on Vice. Press releases are cool.
http://motherboard.vice.com/read/colliding-galaxies-left-a-stream-of-gas-26-million-light-years-long

Galaxy Simulations As Abstract Art

Not another FLASH simulation today (perhaps tomorrow). This one's a particle-based sim of a star-forming gas blob being hit by a hot wind. Instead of animating it, every frame from the simulation is shown at once, so it shows the particle trajectories.

Gas is white, stars are the very bright hard lines, and dark matter is in green because green is nice.

A cosmic jellyfish

Simulation I set running before I went on holiday. This one has a much higher density surrounding medium than the last one did, so the effects are much more dramatic. As with the last one, it's an otherwise stable cold, dense gas disc face-on to a 1,000 km/s wind of hot, low density gas.

Unfortunately I forgot to change the density of the wind. So, although the surrounding medium is initially much denser, as intended, it gets quickly replaced by a much lower density medium.

Interestingly, even after the medium is replaced with the uniform low-density gas, the cold gas removal seems to happen in pulses. That wasn't what I was expecting at all.

A first test of ram pressure stripping

Nothing to do with morally dubious barnyard antics, RPS is simply gas stripping from one body (in this case a cold, dense disc) moving through a medium (which in this case is very hot and very low density). As the dense gas moves, "ram" pressure from the medium builds up, which can be strong enough to push the dense gas out of its happy stable orbit.

Here the medium is very very thin, and although it's a million Kelvin and moving at 1,000 km/s, it's only strong enough to push the outermost (lowest density) gas out of the disc. The colours are logarithmic, the long cone of gas which develops is fairly pathetic compared to the gas which remains in the disc.

This one is really jumping the gun - the "stable gas disc" has its gravity modelled purely by a fixed analytic potential (no self gravity) and no stars, let alone star formation. But it's a taste of things to come.

Congratulations Dr Taylor, it's a disc !

More FLASH experiments. Stage 1 (produce a stable gas disc that sits there, quietly minding its own business) is almost complete. Here it is after a billion years (5 rotations) - it resolutely refuses to disintegrate like the last one did. Since it doesn't do a lot, this one is just a rotation movie of the last output frame from the simulation.

To make things more interesting, the transparency here is equivalent to density, while the colour shows temperature. I used a logarithmic scale for the density, so the hot gas around the disc is actually much, much less dense than it looks here. That's why the inner part of the cold disc appears to be hidden. It would probably be quite interesting to watch the temperature evolve with time, but this is just a crude hack so that would take a bit longer to code.

The reason the last disc went bbblewwwerrrgh (technical term) is probably because FLASH has problems with certain boundary conditions. The mass of gas increased tenfold over the simulation, and the disc just gave up. This time, using different boundary conditions (the space is a hypertorus, so anything that hits one edge magically re-appears on the opposite edge), mass is conserved, and everything is basically tickety-boo. Apart from a density increase in the center, which is slightly annoying.

Stage 1 is our control experiment. When that's done, we'll be able to chuck other galaxies at it and see what happens, and who knows what else. We've got to be sure that the disc is stable if left alone, otherwise our interacting galaxy simulations wouldn't tell us very much.

If you're gonna fail, fail spectacularly

Today's FLASH experiments ended in disaster again, but not as much disaster as the last attempt. Making the ambient medium a thousand times less dense seems to have helped. Now the disc stays respectably stable for about 400 Myr, at which point it... err... well it does something wrong. It doesn't collapse, or explode, or tear itself to shreds... I don't know how to describe it. Watch for yourself, it's very pretty.

The problem may be that the boundary of the simulation is too small. FLASH abhors a vacuum, so there has to be some ambient medium (albeit of arbitrarily low density) and a specified computational volume. Early on, the disc expands just a little bit and brushes against the edge of the simulation volume. Then it collapses again. To fill in the gap that would be left by the collapsing gap, FLASH inserts gas of equal density to the gas that last contacted the edge. That may be causing a runaway effect, since the mass at the end of the simulation is 100 times the mass at the start. Ooops.

Beautiful Failures

Efforts to learn the FLASH hydrocode continue with an attempt to setup a simple, stable gas disc in a gravitational potential. The dense, cold gas sits inside a very low-density hot medium. The dense gas is not self-gravitating; gravity is only modelled by a fixed potential that simulates a dark matter halo, the gas itself, and the stars. That should be fine as long as the gas is stable.

... but for some reason, it isn't. The top picture shows a complete gas disc which pretty rapidly starts to expand in the center, and then the whole thing collapses. The resident FLASH expert thinks it might be due to the low resolution of the simulation causing artificial viscosity (drag) in the center, so the density builds up and so does the temperature and pressure.

Solution ? Take out a chunk of the gas in the center (second picture down). That should remove the densest gas where the problem is worse, and wouldn't be unrealistic since many spiral galaxies don't have much gas in their centers. But it didn't work.

Oh well. Maybe the hole wasn't big enough. Try a bigger hole (middle) ? Nope. Same problem.

Hmm. Perhaps if we go for broke and try just using a thin ring of gas (fourth one down). Nope. Looks like the hot thin medium (sounds like a dodgy late-night TV channel) is collapsing to very high densities due to the gravitational potential. Why this means the dense gas ring collapses I'm not sure.

In for a penny, in for a pound.. how about removing the dense gas entirely (last one) ? Nope. Same runaway density problem. Well, it's only been a day. And at least they're pretty.


The initial conditions were modelled by my officemate. Different colour schemes were used for each rendering, because I wanted too. I believe the last two use a logarithmic stretch, that's probably why they looks so much worse than the others at the end. Simulations were rendered in Blender.

PAPER ACCEPTED ! I HAS ALL THE WIN !

PAPER ACCEPTED ! I HAS ALL THE WIN !

Err, I mean, this is an important breakthrough which will revolutionise science and push the boundaries of human knowledge !

... no, that's not right either...

This is a fairly interesting discovery which gently prods the boundaries of human knowledge. We found a bunch of galaxies with some extremely long gas streams, which is kindof neat. We're doing a press release on this one, but you heard it here first.

I'll have more to say on this, of course. Placeholder post intended to be replaced with a better summary.
http://arxiv.org/abs/1407.0016

Website update !

Website update !

I added a gallery of the latest Orion ship :
http://www.rhysy.net/DiscoveryOrion/

Added an archive of all 18 versions of FRELLED, retroactively named as Farscape characters :
http://www.rhysy.net/frelled-archive.html

Added a bunch of work pages. Research interests :
http://www.rhysy.net/research.html

Publications with short abstracts :
http://www.rhysy.net/publications.html

Teaching and outreach :
http://www.rhysy.net/teaching-outreach.html

Small repository of useful code I've written :
http://www.rhysy.net/programs.html

My first simulation

I'm learning to use the FLASH hydrodynamics code to simulate galaxies. I'm starting with a deliberately silly test - a uniform density cylinder moving in a dense medium. It stops moving almost immediately and then a bunch of funky fluid-stuff happens. Rendered in Blender.