Semi-Comfortable Question: Personal Particle Accelerator

Back in the comments of one of the "Uncomfortable Question" threads, Matthew Jarpe asked (as background research for a new novel):

If someone were to hand you the keys to your own particle accelerator and you could do any experiment you wanted, what would it be?

Well, if somebody just gave me the keys to CERN, and left for the weekend, I'd be sorely tempted to steal some vacuum pumps and digital electronics. Because, dude, they've got some awfully nice stuff, and they'd hardly miss it...

I assume that the question is really intended to be what sort of particle physics experiment would I do if given total control of my own accelerator? In which case, the answer is "It doesn't matter. There's only one experiment you can do with a particle accelerator."

That's being a little unfair, but it's true: a particle accelerator does one thing, and one thing only. It accelerates specific particles to speeds close to the speed of light, and slams them into each other. When they design accelerators, they have specific particle types in mind, and specific energies in mind. If you're given total control over a particular accelerator, you really only have a binary decision: either you turn it on, and slam those particles together at those energies, or you turn it off, and go hiking in the Alps.

"Wait a minute," you say, "What about all those different experimental collaborations? D0 and ATLAS and all the others? If all you can do is slam specific particles together at particular energies, how are they doing different experiments?"

The fundamental experiment done by all these collaborations is the same: slam two particles together at a fair fraction of the speed of light, and see what comes out. The difference between them is in what sort of detectors they use to look at the stuff that comes out. Different experiments have detectors that are optimized to look for different things, which you can sometimes deduce from the acronyms: the Compact Muon Solenoid (CMS) detector being built at the LHC is particularly designed to catch muons, for example. In other cases, the only thing you can deduce from the acronym is that the people in charge of naming major physics experiments need to be beaten with sticks: ATLAS and ALICE, I'm looking at you.

What you can detect most effectively determines what sort of physics you can investigate. If you set your detector up to be really good at catching muons, that means it's more useful for studying processes that are likely to produce muons, and probably not so useful for studying things that produce lots of quarks. Other detectors are set up to be really good at catching particles that go in a particular direction (LHCf, for example), or to try to cover all possible directions, at the cost of slightly lower overall efficiency. All of these are looking at the same basic experiment, but they can be used to study different sub-sub-fields of particle physics.

There are also different "experiments" done with a single detector, but these exist mostly as code. The detectors are set up to track and record the particles leaving a collision event, and they write terabytes of data into computer storage. Different types of particles are detected by sifting through this mountain of data looking for a particular type of event. If you know that the short-lived exotic particle you are trying to find decays to produce two protons, six muons, and a duck, you write code to search through all the millions of events that have been recorded by a particular detector, and pull out those events where the detectors found two protons, six muons, and a confused waterfowl. You can then look at the number of such events that occurred for various energies. That's what all those graphs that Tomasso Dorigo is always posting (see here, for example) are: the number of collisions that produced a particular set of particles for a given energy of the things that were collided together.

New particles are generally detected indirectly, because they don't last very long before they fall apart into other things. You find new particles by looking for particular outcomes at various energies, and seeing a big peak in some narrow energy range. If you never see two protons, six muons, and a duck except when the colliding particle energy is between 1500 and 1520 GeV, then you know that they're the result of the decay of some new exotic particle with a mass of 1510 GeV/c², and you can start pricing flights to Stockholm. If you see ducks at all sort of different energies, then something else is going on-- somebody probably left a window open.

Redefining Matt's question a bit, then, to be "What sort of thing would you try to detect, given unfettered access to the detector and dataset of your choice?" I think I'd go with one of the black holes that Peter Steinberg talked about the other day. Because, dude, black holes!

So, there's my slightly snarky outsider's explanation of how particle physics works. There are some real experts reading this blog, who will no doubt take offense at my characterzation of the field, and I encourage them to leave corrections in the comments. And, while you're at it,. feel free to answer Matt's question, too...