Barred Galaxies

 
So spiral arms are a rotating density wave, where the self-gravity of the disk help to "amplify" the spiral features.

How big of an amplification can we get? In some cases, the amplification can be very strong and we can get barred galaxies.

Approximately 1/3 of spiral galaxies are barred -- bars are common, but not universal. 

  • What makes a galaxy barred? 
  • How long do they last?

Numerical simulations of galaxies have shown that, by themselves, rotating disks are wildly unstable, and will spontaneously form bars. Why?

Take a disk where all the stars are rotating on perfectly circular orbits - relative to one another, the stars aren't moving. Tweak it just a little, so that there is a little bit of excess density somewhere in the disk. Nearby stars will begin to fall towards this density excess, making it even more excessive. So stars further away will begin to feel the pull, and they will fall in. And we keep going until everyone is falling towards the excess, and we have a bar.
Like this: MPEG movie showing a bare disk going bar unstable.
But we know that stars don't move on perfectly circular orbits. They move on elliptical orbits, which we can think of as an oscillation around a circular orbit. This oscillation is a form of kinetic energy -- the faster they oscillate, the more kinetic energy they have. And this kinetic energy helps them avoid falling into the bar. So the more oscillation energy, the more stable the galaxy is to avoid forming a bar.

It turns out (from dynamical theory) that the oscillation frequency is determined in large part by the rotation speed -- the faster you rotate, the faster you oscillate.

Question: we want to increase the oscillation frequency -- how do we do that?
 

In fact, we can use this as a constraint on dark matter. If disk galaxies didn't have dark matter halos, they'd be very unstable against bar formation, and all spirals should be very strongly barred, or else have very high velocity dispersion. This is known as the Ostriker-Peebles criterion.
Here's an MPEG movie of that same galaxy, this time with a massive dark halo around it.
So we know when bars form -- when the gravitational energy of the disk beats the kinetic energy of oscillation.

What about destruction of bars?

Bar destruction

Once a strong bar forms, the stars are no longer orbiting on circular orbits -- the bar has twisted the orbits so that the stars are moving on more radial orbits, passing close though the center of the galaxy. In fact, it is the stars on radial orbits which make up the bar.

To destroy the bar, we want to move the stars off radial orbits. How would we do this? Hint: think about what happens when a star moves through the center of the galaxy, and what it might meet.
 

Here's an example of bar dissolution in action (from Norman, Sellwood, and Hasan 1996):

Galaxy Evolution

Here's the rub: Strong bars can drive interstellar gas inwards, towards the center of the galaxy. Why? Gas dissipates energy during collisional shocks, and can flow inwards. Stars don't collide, so they don't dissipate energy, so there's no net infall of stars.

So put these three pieces of information together:

Can we come up with an consistant evolutionary scenario for barred galaxies and normal galaxies?