Here is a very good article which details the challenges in building the LHC: link.
The beampipe is stainless and not aluminum. The most important reason is the bellows which connect different sections of the beampipe and allow for the rather substantial contraction of the beampipe when it goes from room temperature to 2K. I can not find the figures right now, but if my memory serves me right, it was on the order of 10m over the whole 27km. Such bellows can not be made of aluminum, I am told. Stainless does not outgas and hold vacuum well and also makes better welds. It is, however, awful in a radiation environment as it does activate! The experiments use berylium for the interaction regions, to keep the radiation length to a minimum, but use that for a few meters only, as it costs too much. 40meter berylium pipe, I am told, costs 40 million SFr. So people here are looking into using aluminum for 20m on both sides of the 2-3m berylium pipe in the experiments. (Note, the active length of ATLAS and CMS are like 40 meters.)
And the answer is yes, we are baking out kilometers of beamline! Meaning, heating it up to get rid of all the "muck" that is stuck inside the vacuum pipes to make sure we get rid of it before we put the beam in there. Noone around here on my floor knows how we do it actually, but everyone is pretty sure that we do...
Btw, I am a detector physicist. Not an accelarator physicist. But yes, well, we have to know quite a bit about the LHC to be able to use it!:)
Incidentally, since you work on silicon technology, it might interest you that the detectors closest to the interaction point (or collisions) are all made of silicon. In ATLAS's case there is 68m^2s of it and in CMS, it is about 120m^2s of it in total. That might sound like a lot of silicon, but it is about two buckets of sand since it is about 300 microns thick... Except, it costs a lot of money... and generates 200million bits of data at each collision and collisions are at 40MHz so the data rate is pretty damn high. (No, we dont read all of it. We cant... We read out only the interesting events at 75kHz.)
I do not know about that skwang. But chances are you have a better probability of figuring that out sitting at Fermilab, than me sitting at CERN. Ask around... People generally know these things. And let me know the answer. If it was the publisher, well, I will have to drop by grudge against Lederman...
The beam parameters for the LHC beam are listed here and yes, it is an impressive list. The experiments at IP1 and IP5, which are respectively ATLAS and CMS are listed as having a beam spot of 16.7 while the other two (ALICE and LHC-b) will have a beam spot of 70.9 as they do not need as high luminosities to achieve their perspective goals.
The beam size around the ring remains largish ~ 100 to 200 microns mostly, but then the beams are squeezed just before the collisions on both sides of the experiments. This "squeezing appraratus" happens to consist of three sets of quadrapole magnets. Incidentally, these magnets are holding up the LHC schedule as one of them failed a test.
I do not know what the beampipe is made of, but I would guess stainless steel. (Will check with machine people today.) (The beampipe itself is housed in a stainless steel casing with lots of services inside it.) The beampipe inside the experiments are made from berylium as it has be a material which has very small radiation length to avoid particles coming out of the collision interacting with it too much.
I do not know much about space charge issues but I do know that the circulating beam current is 0.582 A, and so carries a beam halo of other particles with it and wake field calculations are hard...
I happen to think that the particle physicists won't have to fight to find compelling questions in the next few years. It just may not come from collider physics though... The relatively new data on cosmological parameters will soon be so statistically significant that it will be difficult for the layman to ignore questions like "What is the nature of dark energy?" or "What is the nature of dark matter?" Afterall, anyone who realizes that only 4% of the universe is "stuff like us," has to stop and think about it. These cosmological questions can not be answered by conventional astronomical observations, and may not be answered at the LHC either. Maybe we will have to think of new experiments and methods to probe the answers to such questions.
If you are suggesting that "collider physics" is dead after a stagnant LHC run, you might be right. But, I hardly think that it means that particle physics is dead. It might force us to think of different ways of getting to the heart of the mystery! Collider physicists might have to overcome their rather large biases against working with physicists from other fields and collaborate with them to secure funding for the next physics "project," which may or may not be a collider...
This week is "the Trigger and Physics week" for ATLAS, which is one of the two major experiments at the LHC. The opening talk by the head of the collaboration clearly laid out the LHC schedule, but on slides that are not published on the agenda. The original article that is referred in the/. gist has gotten it wrong!
The LHC schedule can not be publicly released until it is approved by the CERN council, which is meeting on the 18th of June. Presumably, once approved, CERN will make a public statement about the plans.
Currently, the plan is to close the experiments for "bake-out" and readying towards a full LHC cool-down and vacuum test around end of March. "Closing the experiments" means that the beam-pipe is one sealed throughout the 27km ring, which seriously limits the movement, fixing and other assembly tasks of the detector communities, so this is a "deadline" for detectors to be "ready for data-taking".
It takes anywhere between a month or two to ready the ring for insertion of *a* beam. It is looking likely right now, that *a* beam will be inserted into the ring around mid-May. However, that is not enough for the operation of the LHC. The LHC is a Collider, so it needs *two* beams to collide. Colliding two beams within an average design beam spot of 16 microns, is no easy task after having them traveling around 27km. (Before the beams are steered the collide, they are "squeezed" to a smaller radius so that the "density" of collisions are higher. This density of collisions, is what determines the luminosity, or, the number of interactions that happen between two beams, and gives the effective high resolution power of the collider.)
Once "one" beam is commissioned inside the LHC, the other beam, traveling opposite to the first one, will be commissioned. Noone really knows how long it will take to really understand and fine-tune the path (or orbit) of the beams inside the ring, but that is what determines when the LHC will get collisions and the first real data will start flowing, if the detectors, can actually time-in and calibrate, and move/push the data off of the detectors into the Grid for analysis. Now, Lyn Evans, who is the head of the LHC commissioning has repeatedly said that he imagines that is will take at least 3 months to get collisions, once a single-beam is commissioned..
So FALL 2008 is the earliest any realist is expecting to see collisions from the LHC. Then the ball is in the detectors' courtyard to collect data continuously and efficiently, to be able to calibrate all detectors in a timely fashion, to identify and fix detectors problems, and to push the (high bandwidth) data out to the analysis farms...
First physic results out of the LHC will not be before Summer 2009... The first paper will be a boring "foo is the multiplicity of events" and the next will be "bar is the cross-section for Drell-Yan/mininum bias processes" paper. The one after that might be interesting though!!
I agree with you. I think this is a lot of hype just because people like the drama that ensues. I work at CERN and I will be delighted if Fermilab has indeed discovered the Higgs -- the work is, by non means, finished. And it also saves us from the worst case scenario that the LHC turns on and *nothing* shows up.
Incidentally, Frank Wilzcek who won the Nobel prize a couple of years back for his work on QCD gave a talk at CERN last week and painted a picture/theory, where indeed, the Higgs is hidden and although it exists, the LHC has no hope of finding it. Such a scenario is the nightmare of people like me. I am a huge fan of the work that people do in Fermilab and I am sick of/.ers suggesting this non-existant rivalry between CERN and Fermilab.
There is only ONE thing I hold against Fermilab, but actually it is against their former directory, Leon Lederman, for coining the greatest misnomer in the history of particle physics and re-naming the Higgs boson to be the "God particle." Who expected such nonsense from a Nobel Laurate?! The far fetched things that people, even physicists, do to get more publicity makes me sick.
Running Fermilab/Tevatron with its current program will contribute 'not much' to science once LHC is operational. The data that the LHC collects is far more "resourceful" than the Fermilab/Tevatron data because it has higher energy reach and higher luminosity.
So if you want to do something great for American physics and science in general, than convince them to build a Linear Collider, at high energies, which will complement the data that the LHC will take.
If the Higgs is found, still Fermilab can do very little in terms of measuring its properties!
Imagine the following (poor) analogy: Humans can "see" for the first time, and "see" the moon. Great! You have discovered the moon, but know nothing about what it is. You have no clue about the solar system...
In some way, finding the Higgs, if Fermilab has indeed done it, is all wonderful, but hardly the end of the story. I, for one, will be exteremely *seriously* dissappointed if the Nobel prize for finding the Higgs is awarded before we have measured its quantum numbers. What are quantum numbers? Well, what makes a particle, that particular particle, is its couplings to other particles, and also its SPIN. The spin of the Higgs has huge implications, cosmological and otherwise, and without that measurement, for me, it is not yet "a Higgs particle." And surprise, surprise,... the spin of the Higgs is notoriously hard to measure, a feat that Fermilab is not going to accomplish for sure, and the LHC might take about 5-10 years to make this measurement.
So anyone at Fermilab, who is dreaming about the Nobel for finding the Higgs in the next few years, "just-forget-about-it".
I agree with you. I am a BSD-supporter and used to be anti-GPL since thought it was motivated by paranoi. But now, sadly I see that it is almost a necessary means to protect the FOSS environment.
What I am wondering is how fast GPLv3 will be adopted by large players. Any ideas? Will Torvalds embrace it immediately? I hope so...
But I still don't get it. Surely Novell must have lawyers and "some" set of intelligent people there. Why did they agree to such an incredibly crappy deal?
Ok, so there are two different issues here. One is that the experiments so far do not show a statistically significant deviation from the current value. With better statistics, more observations and so on, the situation may or may not change. But the second point is the following: if indeed, there was a "big bang," well, you are required by physics laws, that indeed at the beginning of the universe, the electric charge was different from today's. At the LEP collider at CERN, with a center of mass collision energy of ~210GeV (with electrons and positrons colliding), the fine structure constant was ~1/129 as opposed to ~1/137. If we are indeed making "small replicates of the initial universe" in the particle colliders, well, then at the beginning of the universe, the electric charge was different.
Aside: If dark energy indeed rules the universe, the universe will eventually rip itself apart, changing this constant... But we wont be around to worry about that...
Re:Changes over time?
on
MacGyver Physics
·
· Score: 3, Interesting
And I should add that there are experiments worth confirming.
I might be going off the deep end here, but the fact of the matter is, the universe is expanding quite rapidly and there is nothing that says that physics constants can not change over time. One "constant" that has changed and actually is not a constant at all, is the fine-structure constant (Read this to mean: the electric charge. ) The coupling of photons to electrons change, effectively changing the electric charge with distance. Hence, the fine-structure constant is known as a running-coupling constant. There are experiments under progress right now that are trying to measure the fine-structure constant from very-far-away galaxies, or back in time. Ok, maybe I am talking about cosmological scales here, but it would be funny if humans evolved, and some billions of years later, someone reading about some experiment like Rutherford's re-did it and got different results...
Back to the subject: yep, it is pretty crucial to get undergrads to repeat old experiments, especially ones like P-violation, Moessbauer, optical-pumping, muon-lifetime, which have all contributed to our current understanding of physics as a whole. Afterall, if they continue in physics, they might be stuck on experiments like mine, where one does not get data for the next "n" years. (I consider myself a physicist now for 10 years and have not been on a running and data collecting experiment yet. I am very happy that I got to do all those old-experiments in my undergrad years. Good old junior lab... )
So thanks for the original article. I found 8 things that Mike Turner lists in his 1998 paper. As reading the article is not a possibility, off I go: (I am going to be *very* optimistic)
> 1. Origin of the expansion and definitive measure of the present expansion rate H0 (Hubble's constant).
Ok. Origin of expansion... well there are some pretty theories out there. Inflation looks like it is good shape. But a good measurement of the Hubble's constant. Err. Yeah, the precision is better than 1998. But hardly close to where one would like to it be...
> 2. Origin of the heat in the universe and a precise measure of the present temperature of the CMB.
Origin of heat in the universe?! Not really. Precise measurement of the present temperature of the CMB is DONE!:)) YEY!
> 3. Full accounting of matter and energy in the universe. From such an accounting one can infer the present rate of deceleration (or acceleration) of the expansion and the geometry of the universe.
NOT! NOT AT ALL! That would require us to understand dark energy -- and the search crew is out... May never return...
> 4. Understanding of the origin of the density inhomogeneities that seeded all the structure seen in the universe today.
If you make some wild assumptions, hand-wave rigirously... Yeah, why not?!
> 5. Understanding of the origin of ordinary matter and particle dark matter.
LOL! We dont know what caused *mass*. The elusive Higgs is not found yet! Dark matter??! Ahem. Yeah! If you believe in SUSY... Nevermind all the modified-Newtonian dynamics arguments/.ers will throw at you!
> 6. Understanding of the dynamite behind the big bang. The term "big bang theory" is a misnomerit is a theory not of the big bang event but, rather, of the events thereafter.
There is no way I am ever going to understand that. Mike Turner or someone else might. I do not want to know how much drugs that requires. But yes, why not?!
> 7. Understanding of the regularity of the universe, as evidenced by the uniformity of the CMB (temperature variations across the sky of less than one part in 104) and the statistically homogeneous distribution of galaxies.
Now, there is some sort of an understanding developing there. SDSS (Sloan Digital Sky Survey) and CMB surveys agree to some extent. Indeed.
> 8. Description of the history of the universe from the big bang event on.
Aha. Yeah. Good luck writing that book. I am sure I'll enjoy reading it.
Why not instead base the understanding of the universe on being able to understand the "scales" of the universe. For example, a very fundamental number that comes from the CMB (Cosmic Microwave Background) measurements is the number of photons to the number of baryons in the universe. In other words, the amount of light that came out of the explosion to the amount of stuff that exploded in the Big Bang. Now I am sure that's something we should be able to explain if we have solved Cosmology. Nope. No luck. Is this even in his list? Nope. The number is pretty well-measured and it is ~10^{-10}. Theories put it at 10^{-20}. Do you consider that solved??
Ok. End of rant. I actually like Mike Turner. He is one of the few people who try to explain cosmology to normal people and he does a fairly good job. I still wish I could RTA.
OK, so you go to the bottom of the confusion. The LHC runs at 40MHz. All of the detector readout in all experiments is tuned to this number. If it would be off... ouch! The catch is that there are *empty* bunches. These known as the orbit, last for a few microseconds and what most detectors do during this time is to reset their front-ends which might have beserk with radiation. But really, the orbit gap comes from the insertion mechanism of the beam from the SPS to the LHC. Accounting for the fact that there are no collisions during orbit, there are 30 million collisions in one second *on average*. Saying that there are 30 million collisions a second, results in people like you, going ahead and calculating the MHz from that... because the clause "a second" implies "constant and repetitive" to some extent. "30 million collisions in one second" would be alright as well as saying "a second on average."
I am so humbled. NYTimes writer actually replied to my e-mail on this. I suggested to the NYTimes to add "on average" to this to clear up the confusion, to which he replied with " I might have said something like that if I had had more time after I discovered this little glitch. "
Naa. We *have* thought about pointing it towards Redmond though. (They should try suing CERN. I am sure we'd love that.)
The conversation went "Oh, but we just paid so much money for the damn superconducting magnets? Why do they still eat so much power?" "Oh-oh" (Ok, there were no machine engineers around, but a bunch of physicists. The machine engineers though want to blow out the brains of physicists on a regular basis, who they consider idiots... So you see, it is all in good humor.) Luckily, we still remember how to calculate synchrotron losses. Sort of...
The collision are happening at 40MHz. (The NYTimes, got it wrong and says 30MHz. I dont know where you got the 20MHz number from. But nope.. everything at the LHC, detectors, everything runs at 40MHz.) The trick is to "trigger"...
Say, you have a huge library of books, and you only want to find those that have something to do with... computers. Then you look at the title of the book and throw it out if it is about the structure of DNA. It is sort of like that. The "trigger" looks at the event, but only part of the event -- as seen by the detectors, and checks to see if there is something that might be interesting and new in there, such as a Higgs, or some supersymmetric particle. If not, then it discards it. So the pixel detector is not read out every 40MHz. It gets what is known as the L1A trigger, and reads out at 75kHz. Ok, that's still a lot of data!
You might be interested to know that CMS (and LHCb) use technology that is very similar to what the phone companies use. CMS use switchers to distribute the events arriving at 75kHz to a large cluster of computers. The data can not be written to disk at this speed, so they reduce this data further to ~100Hz and write to disk. This is called a higher-level trigger decision and has to be done under 40msec, which is the biggest challenge there. LHCb's L1A rates are higher as their data packets are smaller and they write to disk ~1000Hz. ATLAS does something completely different and reduces data in multiple steps rather than using switchers. But no, none of this is trivial by any means.
As for the SSC in TX, I think they filled it up mostly for some stupid reason. So nope, no chance of using it ever again.
Err.... Actually, this power does not go into the electromagnets directly. The electromagnets happen to be superconducting magnets, which, once powered, do not require more current. That's not where the power goes. The power goes into keeping it cool.
18kW of synchrotron radiation is dumped into the cryogenics system. The syncrotron radiation is due to the relativistic charged particles curving under the influence of the magnetic fields, but this dumped energy needs to be extracted before it results in a quench. A quench is defined as a superconducting magnet, which has no resistivity, transitioning into the resistive phase, due to the temperature rising locally above the critical point.
Here is an interesting link to the power budget of CERN: link
As you will see, the LHC eats up little power (given its size) compared to the SPS (Super Proton Synchrotron) which has conventional magnets and has much smaller radius. The SPS delivers 450MeV protons to the LHC, which then accelerates them upto 14TeV. But the SPS eats up more power than the LHC due to its conventional magnets. Hurray for super-conductivity.
ps. you may not have realized this, but might like to know that your post resulted in an excited discussion in at least one CERN corridor...
Actually there is a blatant mistake in the NYTimes article. It says that collisions wil happen 30 million times a second.
"Again and again and again -- 30 million times a second, in fact."
Nope. The LHC runs at 40MHz.... A number that is absolutely hard-coded into the design and can not be changed... Wrote an e-mail to the NYTimes. They are generally pretty good with correcting in due time.
14 TeV is the amount of energy that is in a collision from two 7TeV beams colliding. In this case, the beam means particles (protons) accelerated to carry 7TeV of momentum. But that's just one "particle". The LHC, there are many "buckets" of particles being stored and collided and the total stored energy around the whole ring is 360MegaJoules. It is fairly easy to calculate actually:
There are 2808 bunches around the ring, each containing 1.15x10^{11} protons each with 7TeV of momentum. 7TeV = 7x10^{12} x 1.602x10^{-19} Joules. You multiply it all out, you get 362MegaJoules stored in the beam around the LHC ring.
That's 1 small cruise ship of 10,000 tons moving at 30km/hour.
450 automobioles of 2tons moving at 100km/hour.
Is enough to melt 500kg of copper. (which is actually a worry if the beams "are lost" due to a magnet quench and they hit the vacuum pipe!)
Oh, btw, the power consumption of the LHC only (excluding the detectors) is ~120MW.
Slashdot Mirror
Here is a very good article which details the challenges in building the LHC: link.
The beampipe is stainless and not aluminum. The most important reason is the bellows which connect different sections of the beampipe and allow for the rather substantial contraction of the beampipe when it goes from room temperature to 2K. I can not find the figures right now, but if my memory serves me right, it was on the order of 10m over the whole 27km. Such bellows can not be made of aluminum, I am told. Stainless does not outgas and hold vacuum well and also makes better welds. It is, however, awful in a radiation environment as it does activate! The experiments use berylium for the interaction regions, to keep the radiation length to a minimum, but use that for a few meters only, as it costs too much. 40meter berylium pipe, I am told, costs 40 million SFr. So people here are looking into using aluminum for 20m on both sides of the 2-3m berylium pipe in the experiments. (Note, the active length of ATLAS and CMS are like 40 meters.)
And the answer is yes, we are baking out kilometers of beamline! Meaning, heating it up to get rid of all the "muck" that is stuck inside the vacuum pipes to make sure we get rid of it before we put the beam in there. Noone around here on my floor knows how we do it actually, but everyone is pretty sure that we do...
Btw, I am a detector physicist. Not an accelarator physicist. But yes, well, we have to know quite a bit about the LHC to be able to use it! :)
Incidentally, since you work on silicon technology, it might interest you that the detectors closest to the interaction point (or collisions) are all made of silicon. In ATLAS's case there is 68m^2s of it and in CMS, it is about 120m^2s of it in total. That might sound like a lot of silicon, but it is about two buckets of sand since it is about 300 microns thick... Except, it costs a lot of money... and generates 200million bits of data at each collision and collisions are at 40MHz so the data rate is pretty damn high. (No, we dont read all of it. We cant... We read out only the interesting events at 75kHz.)
I do not know about that skwang. But chances are you have a better probability of figuring that out sitting at Fermilab, than me sitting at CERN. Ask around... People generally know these things. And let me know the answer. If it was the publisher, well, I will have to drop by grudge against Lederman...
The beam parameters for the LHC beam are listed here and yes, it is an impressive list. The experiments at IP1 and IP5, which are respectively ATLAS and CMS are listed as having a beam spot of 16.7 while the other two (ALICE and LHC-b) will have a beam spot of 70.9 as they do not need as high luminosities to achieve their perspective goals.
The beam size around the ring remains largish ~ 100 to 200 microns mostly, but then the beams are squeezed just before the collisions on both sides of the experiments. This "squeezing appraratus" happens to consist of three sets of quadrapole magnets. Incidentally, these magnets are holding up the LHC schedule as one of them failed a test.
I do not know what the beampipe is made of, but I would guess stainless steel. (Will check with machine people today.) (The beampipe itself is housed in a stainless steel casing with lots of services inside it.) The beampipe inside the experiments are made from berylium as it has be a material which has very small radiation length to avoid particles coming out of the collision interacting with it too much.
I do not know much about space charge issues but I do know that the circulating beam current is 0.582 A, and so carries a beam halo of other particles with it and wake field calculations are hard...
I happen to think that the particle physicists won't have to fight to find compelling questions in the next few years. It just may not come from collider physics though... The relatively new data on cosmological parameters will soon be so statistically significant that it will be difficult for the layman to ignore questions like "What is the nature of dark energy?" or "What is the nature of dark matter?" Afterall, anyone who realizes that only 4% of the universe is "stuff like us," has to stop and think about it. These cosmological questions can not be answered by conventional astronomical observations, and may not be answered at the LHC either. Maybe we will have to think of new experiments and methods to probe the answers to such questions.
If you are suggesting that "collider physics" is dead after a stagnant LHC run, you might be right. But, I hardly think that it means that particle physics is dead. It might force us to think of different ways of getting to the heart of the mystery! Collider physicists might have to overcome their rather large biases against working with physicists from other fields and collaborate with them to secure funding for the next physics "project," which may or may not be a collider...
This week is "the Trigger and Physics week" for ATLAS, which is one of the two major experiments at the LHC. The opening talk by the head of the collaboration clearly laid out the LHC schedule, but on slides that are not published on the agenda. The original article that is referred in the /. gist has gotten it wrong!
The LHC schedule can not be publicly released until it is approved by the CERN council, which is meeting on the 18th of June. Presumably, once approved, CERN will make a public statement about the plans.
Currently, the plan is to close the experiments for "bake-out" and readying towards a full LHC cool-down and vacuum test around end of March. "Closing the experiments" means that the beam-pipe is one sealed throughout the 27km ring, which seriously limits the movement, fixing and other assembly tasks of the detector communities, so this is a "deadline" for detectors to be "ready for data-taking".
It takes anywhere between a month or two to ready the ring for insertion of *a* beam. It is looking likely right now, that *a* beam will be inserted into the ring around mid-May. However, that is not enough for the operation of the LHC. The LHC is a Collider, so it needs *two* beams to collide. Colliding two beams within an average design beam spot of 16 microns, is no easy task after having them traveling around 27km. (Before the beams are steered the collide, they are "squeezed" to a smaller radius so that the "density" of collisions are higher. This density of collisions, is what determines the luminosity, or, the number of interactions that happen between two beams, and gives the effective high resolution power of the collider.)
Once "one" beam is commissioned inside the LHC, the other beam, traveling opposite to the first one, will be commissioned. Noone really knows how long it will take to really understand and fine-tune the path (or orbit) of the beams inside the ring, but that is what determines when the LHC will get collisions and the first real data will start flowing, if the detectors, can actually time-in and calibrate, and move/push the data off of the detectors into the Grid for analysis. Now, Lyn Evans, who is the head of the LHC commissioning has repeatedly said that he imagines that is will take at least 3 months to get collisions, once a single-beam is commissioned..
So FALL 2008 is the earliest any realist is expecting to see collisions from the LHC. Then the ball is in the detectors' courtyard to collect data continuously and efficiently, to be able to calibrate all detectors in a timely fashion, to identify and fix detectors problems, and to push the (high bandwidth) data out to the analysis farms...
First physic results out of the LHC will not be before Summer 2009... The first paper will be a boring "foo is the multiplicity of events" and the next will be "bar is the cross-section for Drell-Yan/mininum bias processes" paper. The one after that might be interesting though!!
Incidentally, Frank Wilzcek who won the Nobel prize a couple of years back for his work on QCD gave a talk at CERN last week and painted a picture/theory, where indeed, the Higgs is hidden and although it exists, the LHC has no hope of finding it. Such a scenario is the nightmare of people like me. I am a huge fan of the work that people do in Fermilab and I am sick of /.ers suggesting this non-existant rivalry between CERN and Fermilab.
There is only ONE thing I hold against Fermilab, but actually it is against their former directory, Leon Lederman, for coining the greatest misnomer in the history of particle physics and re-naming the Higgs boson to be the "God particle." Who expected such nonsense from a Nobel Laurate?! The far fetched things that people, even physicists, do to get more publicity makes me sick.
So if you want to do something great for American physics and science in general, than convince them to build a Linear Collider, at high energies, which will complement the data that the LHC will take.
Imagine the following (poor) analogy: Humans can "see" for the first time, and "see" the moon. Great! You have discovered the moon, but know nothing about what it is. You have no clue about the solar system...
In some way, finding the Higgs, if Fermilab has indeed done it, is all wonderful, but hardly the end of the story. I, for one, will be exteremely *seriously* dissappointed if the Nobel prize for finding the Higgs is awarded before we have measured its quantum numbers. What are quantum numbers? Well, what makes a particle, that particular particle, is its couplings to other particles, and also its SPIN. The spin of the Higgs has huge implications, cosmological and otherwise, and without that measurement, for me, it is not yet "a Higgs particle." And surprise, surprise,... the spin of the Higgs is notoriously hard to measure, a feat that Fermilab is not going to accomplish for sure, and the LHC might take about 5-10 years to make this measurement.
So anyone at Fermilab, who is dreaming about the Nobel for finding the Higgs in the next few years, "just-forget-about-it".
I, for one, welcome the Mac OS X support. - A proud Emacs user now for 11 years. (ie. not too long really!)
I agree with you. I am a BSD-supporter and used to be anti-GPL since thought it was motivated by paranoi. But now, sadly I see that it is almost a necessary means to protect the FOSS environment. What I am wondering is how fast GPLv3 will be adopted by large players. Any ideas? Will Torvalds embrace it immediately? I hope so...
But I still don't get it. Surely Novell must have lawyers and "some" set of intelligent people there. Why did they agree to such an incredibly crappy deal?
Aside: If dark energy indeed rules the universe, the universe will eventually rip itself apart, changing this constant... But we wont be around to worry about that...
I might be going off the deep end here, but the fact of the matter is, the universe is expanding quite rapidly and there is nothing that says that physics constants can not change over time. One "constant" that has changed and actually is not a constant at all, is the fine-structure constant (Read this to mean: the electric charge. ) The coupling of photons to electrons change, effectively changing the electric charge with distance. Hence, the fine-structure constant is known as a running-coupling constant. There are experiments under progress right now that are trying to measure the fine-structure constant from very-far-away galaxies, or back in time. Ok, maybe I am talking about cosmological scales here, but it would be funny if humans evolved, and some billions of years later, someone reading about some experiment like Rutherford's re-did it and got different results...
Back to the subject: yep, it is pretty crucial to get undergrads to repeat old experiments, especially ones like P-violation, Moessbauer, optical-pumping, muon-lifetime, which have all contributed to our current understanding of physics as a whole. Afterall, if they continue in physics, they might be stuck on experiments like mine, where one does not get data for the next "n" years. (I consider myself a physicist now for 10 years and have not been on a running and data collecting experiment yet. I am very happy that I got to do all those old-experiments in my undergrad years. Good old junior lab... )
*ducks and then goes back to read some article on some exotic particle which we will never find*
> 1. Origin of the expansion and definitive measure of the present expansion rate H0 (Hubble's constant).
Ok. Origin of expansion... well there are some pretty theories out there. Inflation looks like it is good shape. But a good measurement of the Hubble's constant. Err. Yeah, the precision is better than 1998. But hardly close to where one would like to it be...
> 2. Origin of the heat in the universe and a precise measure of the present temperature of the CMB.
Origin of heat in the universe?! Not really. Precise measurement of the present temperature of the CMB is DONE! :)) YEY!
> 3. Full accounting of matter and energy in the universe. From such an accounting one can infer the present rate of deceleration (or acceleration) of the expansion and the geometry of the universe.
NOT! NOT AT ALL! That would require us to understand dark energy -- and the search crew is out... May never return...
> 4. Understanding of the origin of the density inhomogeneities that seeded all the structure seen in the universe today.
If you make some wild assumptions, hand-wave rigirously... Yeah, why not?!
> 5. Understanding of the origin of ordinary matter and particle dark matter.
LOL! We dont know what caused *mass*. The elusive Higgs is not found yet! Dark matter??! Ahem. Yeah! If you believe in SUSY... Nevermind all the modified-Newtonian dynamics arguments /.ers will throw at you!
> 6. Understanding of the dynamite behind the big bang. The term "big bang theory" is a misnomerit is a theory not of the big bang event but, rather, of the events thereafter.
There is no way I am ever going to understand that. Mike Turner or someone else might. I do not want to know how much drugs that requires. But yes, why not?!
> 7. Understanding of the regularity of the universe, as evidenced by the uniformity of the CMB (temperature variations across the sky of less than one part in 104) and the statistically homogeneous distribution of galaxies.
Now, there is some sort of an understanding developing there. SDSS (Sloan Digital Sky Survey) and CMB surveys agree to some extent. Indeed.
> 8. Description of the history of the universe from the big bang event on.
Aha. Yeah. Good luck writing that book. I am sure I'll enjoy reading it.
Why not instead base the understanding of the universe on being able to understand the "scales" of the universe. For example, a very fundamental number that comes from the CMB (Cosmic Microwave Background) measurements is the number of photons to the number of baryons in the universe. In other words, the amount of light that came out of the explosion to the amount of stuff that exploded in the Big Bang. Now I am sure that's something we should be able to explain if we have solved Cosmology. Nope. No luck. Is this even in his list? Nope. The number is pretty well-measured and it is ~10^{-10}. Theories put it at 10^{-20}. Do you consider that solved??
Ok. End of rant. I actually like Mike Turner. He is one of the few people who try to explain cosmology to normal people and he does a fairly good job. I still wish I could RTA.
OK, so you go to the bottom of the confusion. The LHC runs at 40MHz. All of the detector readout in all experiments is tuned to this number. If it would be off... ouch! The catch is that there are *empty* bunches. These known as the orbit, last for a few microseconds and what most detectors do during this time is to reset their front-ends which might have beserk with radiation. But really, the orbit gap comes from the insertion mechanism of the beam from the SPS to the LHC. Accounting for the fact that there are no collisions during orbit, there are 30 million collisions in one second *on average*. Saying that there are 30 million collisions a second, results in people like you, going ahead and calculating the MHz from that... because the clause "a second" implies "constant and repetitive" to some extent. "30 million collisions in one second" would be alright as well as saying "a second on average."
I am so humbled. NYTimes writer actually replied to my e-mail on this. I suggested to the NYTimes to add "on average" to this to clear up the confusion, to which he replied with " I might have said something like that if I had had more time after I discovered this little glitch. "
I remain, very much, a fan of the NYTimes...
Here is another article on the subject.
And I got the SSC link wrong. Here it is: SSC
The conversation went "Oh, but we just paid so much money for the damn superconducting magnets? Why do they still eat so much power?" "Oh-oh" (Ok, there were no machine engineers around, but a bunch of physicists. The machine engineers though want to blow out the brains of physicists on a regular basis, who they consider idiots... So you see, it is all in good humor.) Luckily, we still remember how to calculate synchrotron losses. Sort of...
Sorry, you are right. Good catch.
Yes, the data throughput is staggering.
The collision are happening at 40MHz. (The NYTimes, got it wrong and says 30MHz. I dont know where you got the 20MHz number from. But nope.. everything at the LHC, detectors, everything runs at 40MHz.) The trick is to "trigger"...
Say, you have a huge library of books, and you only want to find those that have something to do with ... computers. Then you look at the title of the book and throw it out if it is about the structure of DNA. It is sort of like that. The "trigger" looks at the event, but only part of the event -- as seen by the detectors, and checks to see if there is something that might be interesting and new in there, such as a Higgs, or some supersymmetric particle. If not, then it discards it. So the pixel detector is not read out every 40MHz. It gets what is known as the L1A trigger, and reads out at 75kHz. Ok, that's still a lot of data!
You might be interested to know that CMS (and LHCb) use technology that is very similar to what the phone companies use. CMS use switchers to distribute the events arriving at 75kHz to a large cluster of computers. The data can not be written to disk at this speed, so they reduce this data further to ~100Hz and write to disk. This is called a higher-level trigger decision and has to be done under 40msec, which is the biggest challenge there. LHCb's L1A rates are higher as their data packets are smaller and they write to disk ~1000Hz. ATLAS does something completely different and reduces data in multiple steps rather than using switchers. But no, none of this is trivial by any means.
As for the SSC in TX, I think they filled it up mostly for some stupid reason. So nope, no chance of using it ever again.
Err.... Actually, this power does not go into the electromagnets directly. The electromagnets happen to be superconducting magnets, which, once powered, do not require more current. That's not where the power goes. The power goes into keeping it cool. 18kW of synchrotron radiation is dumped into the cryogenics system. The syncrotron radiation is due to the relativistic charged particles curving under the influence of the magnetic fields, but this dumped energy needs to be extracted before it results in a quench. A quench is defined as a superconducting magnet, which has no resistivity, transitioning into the resistive phase, due to the temperature rising locally above the critical point. Here is an interesting link to the power budget of CERN: link As you will see, the LHC eats up little power (given its size) compared to the SPS (Super Proton Synchrotron) which has conventional magnets and has much smaller radius. The SPS delivers 450MeV protons to the LHC, which then accelerates them upto 14TeV. But the SPS eats up more power than the LHC due to its conventional magnets. Hurray for super-conductivity. ps. you may not have realized this, but might like to know that your post resulted in an excited discussion in at least one CERN corridor...
Actually there is a blatant mistake in the NYTimes article. It says that collisions wil happen 30 million times a second.
"Again and again and again -- 30 million times a second, in fact."
Nope. The LHC runs at 40MHz.... A number that is absolutely hard-coded into the design and can not be changed... Wrote an e-mail to the NYTimes. They are generally pretty good with correcting in due time.
14 TeV is the amount of energy that is in a collision from two 7TeV beams colliding. In this case, the beam means particles (protons) accelerated to carry 7TeV of momentum. But that's just one "particle". The LHC, there are many "buckets" of particles being stored and collided and the total stored energy around the whole ring is 360MegaJoules. It is fairly easy to calculate actually:
There are 2808 bunches around the ring, each containing 1.15x10^{11} protons each with 7TeV of momentum. 7TeV = 7x10^{12} x 1.602x10^{-19} Joules. You multiply it all out, you get 362MegaJoules stored in the beam around the LHC ring.
That's 1 small cruise ship of 10,000 tons moving at 30km/hour.
450 automobioles of 2tons moving at 100km/hour.
Is enough to melt 500kg of copper. (which is actually a worry if the beams "are lost" due to a magnet quench and they hit the vacuum pipe!)
Oh, btw, the power consumption of the LHC only (excluding the detectors) is ~120MW.
Yeah. That's called ATLAS. Except it has a toroid as well as a solenoid technically...