![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Entry tags:
Educational stuff
Since it's April Fools Day I suppose I could give you some ridiculous whacked out version of my life, but since that's business as usual around here, I thought I would give you a serious and educational explanation of some of what I do, because I think it's neat.
Experience shows that I will cut my small readership in half, but hey, I'm working for free.
The analogy I prefer to use when discussing particle physics is that of attempting to restore a precious Ming Vase. Generally, there are three steps in this procedure. First, you take the vase to the top of a twenty story building. Second, you throw it off the roof. Third, you try to put it back together by collecting the pieces and figuring out what went where.
Particle physics has similar steps. As I've mentioned earlier, particles interact with each other through the good offices of even more particles. Two electrons repel each other because they exchange what we call a "virtual" photon. This is a particle that doesn't really exist, it travels in the brief time and space allowed by Heisenberg, and connects two electrons to each other. Really it's just a name for the little packet of momentum that's carted between the two particles. There really isn't a photon there, because there's no energy given to create it. However, if you have enough background energy floating around, you can actually create a photon, a real live photon, that has the ability to go shooting off into space and hit other things.
Creating a photon is hardly a difficult feat, after all, most photons have very low energies. However, these virtual interactions can create other heavier particles, which take a lot more effort to create out of nothingness. The top quark for instance, the most recent fundamental particle to be discovered, has mass 175 times greater than that of a proton. It takes a great deal of energy to create one of those, especially since they often have to be produced in pairs. To do that, physicists have to create a system in which there is a great deal of energy involved in collisions between particles.
The first step is climbing up the building. This involves building an accelerator which will allow you to give particles these high energies. This is not an easy process. It takes years to design an accelerator these days because the energies are so high. Construction takes longer. To give you an example, the Large Hadron Collider (LHC) at CERN in Geneva will have a functional 27 kilometer long circular track buried underground. The entire beamline, all 27 kilometers, is directed by a series of superconducting magnets, which will all be kept at a temperature of 1.8K (~ -271 Celsius) by the largest Liquid Helium production facility in the world. The magnets themselves are coils woven from strands, each strange being a bundle of filaments, in other words one of those cables made up of smaller wires. If unwound, the strands alone would stretch for 240,000 km, six times the circumference of the Earth. And this is just for the beam. The amount of engineering barriers that have to be overcome for the detectors themselves is staggering.
Then you have to drop the vase over the side. This is the easiest part. Once everything's in place, switch on the beam and you have your collisions. A lot of fine tuning is involved here.
Then it's actually time for physics. When two particles collide, they produce a huge amount of byproducts.
Here's an event display from CDF:
And another (this one showing the event display that physicists actually look at):
So there are literally hundreds of pieces that fall out of a collision. This is where the vase analogy really matches. A physicist has to reassemble those pieces into a full fledged event. At CDF, we collide protons and anti-protons together. A proton is made out of three quarks, as is its antiparticle, so we're really colliding six quarks, who are going to make a whole sequence of particles that will bounce around the detector. We have no idea what they're doing in there, so we have lined the collision chamber with a series of detectors.
Our detectors are broken into two parts; Trackers, dense systems that register hits whenever a particle comes close to them, allowing us to get a fairly good idea of the path a particle took, and Calorimeters, who record how much energy each particle dumps as it passes by. If you add up all the hits in a Tracker, you can get an idea of the path a particle took. If you add up all the energy lost in the Calorimeter, you can get a fairly good idea of how much energy the particle had to start with. Eventually, once you add up all of these pieces, you can grab a fairly good idea of what happened.
Particle physicists are looking for very rare collisions for the most part. To aid them in this, they use beams that have tremendous power. The LHC beam will be colliding particles together at 600 million times a second. They make up for rarity with sheer numbers, eschewing finesse (this is a relatively inefficient way to do a single analysis, fine tuning works better. However, for a project that can investigate a hundred different possibilities, this is the only sensible solution). Most of these events are discarded as uninteresting by the trigger system, which only looks for interesting pieces. The trigger is assisted in this by the Data Aquisition System (DAQ) a piece of software (and hardware) so fundamentally complicated that it defies description. The DAQ for the Compact Muon Solenoid at LHC will use more bandwidth than is consumed by the US's international internet connections (we have a one Terabit switch, and we mean to use it).
Here's an image of CMS, with all it's components outlined:
Needless to say, what comes out of the detector is a complete mess. Even triggering only on interesting data, you get a useless pile of junk for the most part. Data is stored in several different data streams, integrated into datasets. Each dataset holds information about a certain series of events (high-pt leptons or b-jets or whatever), and when it comes time to do an analysis, physicists scan a dataset for something that looks interesting (the signature of what they're looking for). Now is a good time to say that our datasets are enormous. CMS is expected to produce actual data at the relatively low rate of 100MB/sec (approximately fifteen minutes of moderately high quality video per second), but the massive ALICE detector will be operating at 1,250 MB/sec. These detectors will also be operating for years. By the time we've finished out operational run, there will be datasets of tremendous size, and all of it will need to be accessed. You can imagine that we use a lot of computing power searching through all that data looking for the three or four interesting events (out of billions) in the data run.
As to what we're looking for, and how we do it, well, that will have to wait for another time.
Experience shows that I will cut my small readership in half, but hey, I'm working for free.
The analogy I prefer to use when discussing particle physics is that of attempting to restore a precious Ming Vase. Generally, there are three steps in this procedure. First, you take the vase to the top of a twenty story building. Second, you throw it off the roof. Third, you try to put it back together by collecting the pieces and figuring out what went where.
Particle physics has similar steps. As I've mentioned earlier, particles interact with each other through the good offices of even more particles. Two electrons repel each other because they exchange what we call a "virtual" photon. This is a particle that doesn't really exist, it travels in the brief time and space allowed by Heisenberg, and connects two electrons to each other. Really it's just a name for the little packet of momentum that's carted between the two particles. There really isn't a photon there, because there's no energy given to create it. However, if you have enough background energy floating around, you can actually create a photon, a real live photon, that has the ability to go shooting off into space and hit other things.
Creating a photon is hardly a difficult feat, after all, most photons have very low energies. However, these virtual interactions can create other heavier particles, which take a lot more effort to create out of nothingness. The top quark for instance, the most recent fundamental particle to be discovered, has mass 175 times greater than that of a proton. It takes a great deal of energy to create one of those, especially since they often have to be produced in pairs. To do that, physicists have to create a system in which there is a great deal of energy involved in collisions between particles.
The first step is climbing up the building. This involves building an accelerator which will allow you to give particles these high energies. This is not an easy process. It takes years to design an accelerator these days because the energies are so high. Construction takes longer. To give you an example, the Large Hadron Collider (LHC) at CERN in Geneva will have a functional 27 kilometer long circular track buried underground. The entire beamline, all 27 kilometers, is directed by a series of superconducting magnets, which will all be kept at a temperature of 1.8K (~ -271 Celsius) by the largest Liquid Helium production facility in the world. The magnets themselves are coils woven from strands, each strange being a bundle of filaments, in other words one of those cables made up of smaller wires. If unwound, the strands alone would stretch for 240,000 km, six times the circumference of the Earth. And this is just for the beam. The amount of engineering barriers that have to be overcome for the detectors themselves is staggering.
Then you have to drop the vase over the side. This is the easiest part. Once everything's in place, switch on the beam and you have your collisions. A lot of fine tuning is involved here.
Then it's actually time for physics. When two particles collide, they produce a huge amount of byproducts.
Here's an event display from CDF:
And another (this one showing the event display that physicists actually look at):
So there are literally hundreds of pieces that fall out of a collision. This is where the vase analogy really matches. A physicist has to reassemble those pieces into a full fledged event. At CDF, we collide protons and anti-protons together. A proton is made out of three quarks, as is its antiparticle, so we're really colliding six quarks, who are going to make a whole sequence of particles that will bounce around the detector. We have no idea what they're doing in there, so we have lined the collision chamber with a series of detectors.
Our detectors are broken into two parts; Trackers, dense systems that register hits whenever a particle comes close to them, allowing us to get a fairly good idea of the path a particle took, and Calorimeters, who record how much energy each particle dumps as it passes by. If you add up all the hits in a Tracker, you can get an idea of the path a particle took. If you add up all the energy lost in the Calorimeter, you can get a fairly good idea of how much energy the particle had to start with. Eventually, once you add up all of these pieces, you can grab a fairly good idea of what happened.
Particle physicists are looking for very rare collisions for the most part. To aid them in this, they use beams that have tremendous power. The LHC beam will be colliding particles together at 600 million times a second. They make up for rarity with sheer numbers, eschewing finesse (this is a relatively inefficient way to do a single analysis, fine tuning works better. However, for a project that can investigate a hundred different possibilities, this is the only sensible solution). Most of these events are discarded as uninteresting by the trigger system, which only looks for interesting pieces. The trigger is assisted in this by the Data Aquisition System (DAQ) a piece of software (and hardware) so fundamentally complicated that it defies description. The DAQ for the Compact Muon Solenoid at LHC will use more bandwidth than is consumed by the US's international internet connections (we have a one Terabit switch, and we mean to use it).
Here's an image of CMS, with all it's components outlined:
Needless to say, what comes out of the detector is a complete mess. Even triggering only on interesting data, you get a useless pile of junk for the most part. Data is stored in several different data streams, integrated into datasets. Each dataset holds information about a certain series of events (high-pt leptons or b-jets or whatever), and when it comes time to do an analysis, physicists scan a dataset for something that looks interesting (the signature of what they're looking for). Now is a good time to say that our datasets are enormous. CMS is expected to produce actual data at the relatively low rate of 100MB/sec (approximately fifteen minutes of moderately high quality video per second), but the massive ALICE detector will be operating at 1,250 MB/sec. These detectors will also be operating for years. By the time we've finished out operational run, there will be datasets of tremendous size, and all of it will need to be accessed. You can imagine that we use a lot of computing power searching through all that data looking for the three or four interesting events (out of billions) in the data run.
As to what we're looking for, and how we do it, well, that will have to wait for another time.
no subject
no subject
(Anonymous) 2006-04-02 04:25 am (UTC)(link)That's the problem, the only people I have to practice on these days are physicists...
no subject
no subject
no subject
no subject
I've seen very similar event displays in presentations from Chem. professors. I'm almost positive that I've seen the image of CMS before.
no subject
I'm not sure if my students have fond memories of me or not.
no subject
Nifty pictures, though.
no subject
Interesting pics, though!
no subject
(Hey, I like thr physics posts. I read them.)