Monday, 19 October 2009

There's more to the LHC than bloody black holes

The LHC is cold again. This is very exciting, and also it can't come soon enough. In the absence of any actual science going on an endless stream of bollocks seems to have been coming out about the collider. The latest being this drivel about things coming from the future to... oh God I can't be bothered. It rather upsets me that the only things people really know about the LHC are that it might make a black hole and maybe something is coming through time to sabotage it. So I thought I'd talk about why this machine is ridiculously fantastic and complicated (the more likely cause of breakage).

One of the features of synchrotrons that I've always thought is amazing is the way they cool the beams. By cool I'm not talking about temperature around the beam pipe (although that's bloody cold too so that the magnets work). I'll quickly describe what it is and how people solve it, although I'm still not 100% sure how they've solved it at the LHC.

Our general collider accelerates particles around a ring using strong electric fields. The particles are bent into a circle by bending magnets and they are kept in a beam by the focussing quadropole magnets. The effect of these magnets is that if a particle is heading sideways out of the beam then they push it back in in the opposite direction. In this way the particles kind of snake around the course never straying too far out of line. The task of cooling the beam is to reduce this snaking as much as possible so that we have a really dense, straight running beam.

Electron Cooling

You can very much think of this random side-to-side motion of our particles as temperature. Our aim is to get rid of as much of this motion as possible, to cool the beam. One way that this problem has been tackled is called electron cooling. Because they're so light compared to protons (well, anti-protons in this case, negative charge) it's not so bad to send in a beam of electrons through a section of the collider at the same speed as the anti-protons but much cooler. The two beams are then left to effectively equilibrate - taking transverse energy out of the main beam.

That said, the LHC uses two proton beams (positive charge) so I'm not sure they use this. They'd have to use positrons I suppose - anyone knows please do let me know. Electron cooling is very clever either way.

Stochastic Cooling

This is the one that surprised me the most. Stochastic cooling takes a more direct approach to cooling the beam and it makes use of the fact that we're sending our particles in a circle. What you do here is you try to measure small fluctuations from the beam. If a particle has too much transverse momentum then a sensor can send a message to the other side of the ring that can (staggeringly) give the particle a sideways kick to correct the fluctuation.

Remember that our particle beam is travelling essentially at the speed of light. To get a signal to the other side of the ring in time you have to run a wire straight across the middle and rely on the fact that the particles must go around the circumference. Amazingly, you can get the signal there and you can correct the beam. I'm still a little hazy on how the sensors work but I think they detect Cherenkov radiation. Hey, I'm not a particle physicist!

What else?

There are many techniques for cooling the beam, many I don't even know about. I'm not even completely sure what they use at the LHC. What I think is amazing is that this is just one tiny detail in how these machines work and it involved a huge amount of research and refinement. I could have picked from a hundred brilliant solutions to a hundred seemingly impossible problems.

If you want to believe aliens came through time to break the LHC then go ahead. Maybe it's just not that easy smashing two beams together with so much accuracy that, to quote cern, it's "akin to firing needles from two positions 10 km apart with such precision that they meet halfway". It's hardly surprising that there have been problems. So please, in this short science interlude can we resist the urge to go crazy?

Oh, and while I'm on a bit of a rant, can everyone immediately stop referring to the Higgs as the "God particle"? Good.

UPDATE: I've been searching around for how the LHC actually cools the beams. I've found this very long document with loads of juicy technical data. From section 4.4 "Transverse damping and feedbacksystem (ADT)" I ended up at this document which is loaded with information that is mostly way over my head. From what I can gather...

It looks like they are using a form of stochastic cooling.


No comments:

Post a comment