Basically what I've been doing for the past several weeks is trying, with liberal amounts of help from Bjorn (the post doc here with me if you'll recall) to meet the first major milestone of my thesis: finding a rho particle. For those who do not know, my thesis is to determine rho production as a function of multiplicity for ultraperipheral collisions. Dont worry, it's not really half as scary as it sounds.
To start, what is a rho particle? A rho particle is a kind of subatomic particle that is made up of quarks (like most matter, except for leptons like electrons, but no body likes them anyway). To be specific, the rho is a meson which means that it is made up of one quark and one antiquark. since quarks are the smallest things we know about (well, excepting strings but I really dont know much about those except that they vibrate), we consider quarks the building blocks of matter. Kind of like meat and potatoes are the building blocks of food.
Rho mesons are made up of two quarks: an up quark and an anti-down quark. Dont bother with what up and down mean, its just a flavor (yes, flavor) of quark. Most matter is up and down quarks and antiquarks, but there are also charmed, strange, top, and bottom quarks. My task is to look at the data that ALICE detectors glean from collisions at the LHC (large hardon collider for those of you who are familiar with fun typos) and try to pick out rho mesons.
The other parts of my thesis involve the terms "multiplicity" and "ultraperipheral." The multiplicity of an event (collision) is how many things are produced. If a collision produces 4 things (any things, regardless of what they are) then the event had a multiplicity of 4. Millions and millions of these collisions happen very quickly so a lot of data is produced, filtered, and stored in huge files, but various attributes such as the momentum of the particle, the path they followed (the particle's "track"), the speed of the particle (so we can get the mass from the momentum), and many others are all saved and kept together so that all this data can be analyzed.
The last thing I need to explain about my thesis is the term ultraperipheral. There are three kinds of collisions that can happen when you fling two atomic nuclei at eachother. One, they can collide directly in a central collision. This breaks each nucleus apart into a very hot soup of quarks and gluons (dont worry about what a gluon is, it doesn't matter for my purposes). Another, they can brush past eachother, just barely scraping eachothers edge in a peripheral collision. Or third, they can pass close enough by eachother to interact, but still miss in an ultraperipheral collision. I'm looking at ultraperipheral collisions because they produce the least amount of particles (lowest multiplicity) so I have a lot less noise to deal with when trying to see the rhos.
Sadly we cannot detect rho particles directly because they have very very very short half lives and decay without getting a chance to even move off the beam line in the experiment. However, they decay into particles called pions (another kind of meson) which we can detect. Most things decay into pions, so our task is to try and filter out which pions came from rho particles.
To do this, we look at something called the invariant mass. Most of you, I'm sure, are familiar with Einstein's equation E=mc^2. This quantity is, in fact, the invariant mass: the mass of the particle when it isn't moving. As things appraoch the speed of light, they get more and more massive, so we want something to look at that doesn't change with how fast something is going. This equation is actually a specific case for a particle at rest. If you throw momentum into the mix, the equation becomes E^2 = (mc^2)^2+(pc)^2 where p is the momentum and c is the speed of light.
Dont worry, those are the only two equations I'm going to throw at you. Suffice to say, you can find the mass of the original particle from the momentum and mass of the particles that it decayed into. You can then graph this, and viola! The more parent particles (particles that decayed into the pions) that have a given invariant mass, the larger a bump there will be at that mass. The rho particle has a mass of 770 MeV/c^2. This is in terms of energy (in mega electron volts) over c^2, which is much easer to deal with than converting the mass of something as small as a subatomic particle into pounds or kilograms.
So you get this:
Yes, that is a hideous mess of black. We can't see anything from that, so we would rather look at it like this:
That at least shows something, even if it is still a bit of a mess. We can see that most of our particles are where it isn't aqua, over from masses of 0 to 4. The bottom axis is the invariant mass for the particles, the right is the multiplicity of the event. Well, I'm interested in the area around 770 MeV/c^2, or on this mass scale 0.77 (as it is in GeV or giga electron volts. Insert Doc. Brown quote here involving jigawatts). So, zooming in also on the low end of multiplicity (remember, ultraperipheral events have less stuff flying out of them):
This is starting to show something, though it is still hard to tell what. So we use a handy tool called an x-projection. This takes everything and looks at it side on, as if it were all projected onto the x axis:
So we can see here several spikes that indicate different particles. Well, there are too many spikes here because the method we were using to identify different particles particles isn't very good. So we had a lot of extra other stuff here. But, there is hope because there is that large buldge between 0.6 and 0.8 which is where the rho should be! Next we tried looking at the data filtered a different way in hopes of cutting out a lot of the unneeded stuff:
That was a miserable failure. The graphs are actually identicle. Another part of our problem here is we dont really have enough events to get a good clean graph. We're fighting statistics and losing (damn you statistics). So we decided to throw out the particle identification all together and assume that EVERYTHING was pions. Surprisingly this helped a bit:
But not a whole heck of a lot. It's still got spikes, even if the around where the rho should be is a little sharper. Now, remember what I said about ultraperipheral collisions (UPC) being what I was looking at? Well none of those charts were for UPC events. So we applied yet another filter designed to look at just UPC events and we got:
Which is still a mess looked at this way, but if you look in the bottom left corner, you see something! A little spec. Lets look at it in technocolor:
Wahoo! There's definately something, and it is definately around the range that the rho should be at. So we zoom in as we want as low multiplicity as we can (the best being if only two pions were produced because that would be the cleanest signal):
Better, but for the best we can just look at events with a multiplicity of 2. zooming in there we see:
Which from here is a pretty pretty rainbow right at .77 GeV/c^2. Projecting it:
That my friends is what I've been looking for. It could still use some cleaning up, but that peak there is what we in the particle biz call a rho peak. It's a distribution over a range of masses rather than exactly at .77 because quantum mechanics doesn't like things to be well defined or make sense, but the majority of the particles of average are right around there.
So thanks for reading, if you made it through that without being bored or confused then I succeeded in not including too much information and being witty and entertaining as usual. That is pretty much how analysis here works: find something you can graph that will show your particle, and engage in a vicious battle with the data, histograms, and bugs in your code in order to force it to its knees and surrender its information. It never goes down without a fight.
Until next time!
Ciao!
No comments:
Post a Comment