The LHC: A Giant Bent Klystron Tube?

[artis]

Now the text tells that "By contrast, protons with slightly different energies arriving earlier or later will be accelerated or decelerated so that they stay close to the desired energy. In this way, the particle beam is sorted into packs of protons called "bunches".
Here it refers to the main accelerating cavities of the LHC, their working principle sounds exactly like that of a klystron.

So here is my question , would it be fair to say that the LHC itself works like a giant bent/circular klystron tube?
Here is the way I understand it,
first they make the protons, then they are accelerated by linacs and "injected" into the synchrotron ring/s from where they are fed into the largest main ring?
This large main ring (the one 27km in circumference) is a circular tube with evenly spaced RF cavities working at 400Mhz and bending quadropole magnets so after enough rounds trips by the protons in the main ring they get bunched up by the cavities and accelerated to reach their maximum energy?What I gather from CERN's homepage is that initially they only had the much smaller proton synchrotron ring and then they built the larger "super proton synchrotron" which I guess was just another synchrotron but larger and more powerful and then they built the now famous 27km main ring which I gather is just another synchrotron but now on "steroids".Finally is it true that both the larger and smaller LHC rings are all synchrotrons and have the same basic function just that their size and power and parts count differ? differences in details like for example I read the smaller sps ring uses ordinary "room" temp magnets while we know the main ring uses superconducting ones.

 

[mfb]

Not everything that accelerates particles in bunches is a klystron. The particles are grouped into bunches early in the linear accelerator already. You can't accelerate a continuous beam to useful energies for the LHC. Apart from the very first steps everything uses bunches.

The main ring has a short section of RF cavities at one place - they are not the limit for anything. Most of the length is taken up by dipole magnets bending the beam into a circle. How fast they can ramp up to the full strength determines the time the LHC needs for acceleration.
Quadrupole magnets are used for focusing.

Finally is it true that both the larger and smaller LHC rings are all synchrotrons and have the same basic function just that their size and power and parts count differ?

Two of them even have it in their name. Yes.

Normal conducting magnets can be ramped much faster. The PS has a cycle time of less than a second, the SPS can accelerate particles a few times in its ~10 second cycles. Compare this to the ~20 minutes the LHC needs to ramp up its superconducting magnets.
Running the LHC with normal conducting magnets would consume way too much energy to be practical.
Running the SPS with superconducting magnets would make filling the LHC a pain (you need many SPS cycles to fill the LHC rings).

 

[artis]

@mfb I remember reading that they indeed bunch up the particles way before they enter the main ring, I guess they also synchronize the arrival of the bunches as they enter the main ring as otherwise not synchronizing it many bunches would be de-accelerated by flying through the cavities and the wrong time ?

but with respect to the klystron ring as I called it, leaving time and energy efficiency out of the equation, if you would simply introduce a steady stream of protons into the ring and then let them fly around , wouldn't eventually the cavities cause bunching up of the protons and also the acceleration of the bunches? So from your text I take that instead of the 16 cavities being evenly spaced around the ring they are all located in a row in a single place?
The quadropole focusing magnets must be evenly spaced around the ring after a number of dipole magnets right? as I would imagine after some curvature the beam needs some focusing done on it?
As for your last paragraph , hmm this is interesting, I always thought that the reason why the main ring takes about 20 mins for the protons to reach their max speed is because they can't achieve it from a single pass through the RF cavities and so need multiple passes to gain the full available energy, but your saying that the bending and focusing magnet strength is at play, or are both of these factors at play?

IIRC superconducting magnets could lose their superconductivity and return to normal state if either the field gets too strong or increases too fast?
But I was under the impression that the ring magnets are on and "ramped up" like all the time , why would they need to ramp them up while accelerating at the same time?thanks.

 

[mfb]

Yes of course everything needs to be synchronized.
This is an interesting challenge in the lead/proton mode (lead ions in one ring, protons in the other) as they have a different speed. At injection it's impossible to keep them synchronous to each other, so the LHC cannot collide them at the injection energy. After bringing them to the full energy the speed is close enough to synchronize their revolutions again.

So from your text I take that instead of the 16 cavities being evenly spaced around the ring they are all located in a row in a single place?

Yes. The protons pass the cavities millions of times during the acceleration process, gaining a tiny bit of energy each time, it doesn't matter much where they are. Having them in the same place is easier to work with. They could easily install more of them, but it wouldn't help, the limits are from the magnets. This is different at electron-positron colliders where synchrotron radiation is significant and the particles lose a significant fraction of their energy each revolution.

Quadrupole magnets need to be everywhere along the ring, otherwise you would lose the particles.

But I was under the impression that the ring magnets are on and "ramped up" like all the time , why would they need to ramp them up while accelerating at the same time?

Check how the bending radius depend on the energy and the magnetic field strength.

 

[artis]

@mfb I take that your mentioned electron positron collider produced synchrotron radiation is as you say significant because electrons having the same strength of electric charge but being much less heavy can be accelerated to much higher speeds (relativistic? ) so at such speeds bending them by say the same angle as LHC protons would produce much more EM radiation?As for the "bending radius" the way I understand it is that for a fixed radius and fixed particle weight (like that of proton) any increase in it's KE aka speed would require an increase in the deflection B field strength.
So at some point I guess the bending magnets cannot keep up with the speed and the protons would simply run into the walls and be lost eventually?

 

[mfb]

It's all ultra-relativistic for the high energy accelerators, but synchrotron radiation scales with the gamma factor which is much higher for electrons and positrons.

As for the "bending radius" the way I understand it is that for a fixed radius and fixed particle weight (like that of proton) any increase in it's KE aka speed would require an increase in the deflection B field strength.

Right, that's why you need to ramp up the magnets together with increasing the energy.

So at some point I guess the bending magnets cannot keep up with the speed and the protons would simply run into the walls and be lost eventually?

And that's limiting the energy for proton accelerators.

 

[artis]

@mfb Oh right now I got it, the reason the deflection B field must be ramped up proportionally to the proton speed is because if it was steady from the beginning the not yet fast enough protons would get too much deflected inwards and instead of crashing into the outer parts of the tube they would crash into the inner one.How precise does the magnet field increase need to be around the whole circle because a mismatched field strength between say one part of the torus with another might result in the proton bunch crashing into the wall?
ps. I guess what we need is a straight tube , nature doesn't like bends... but the problem is that the cavities cannot transfer enough force on the small protons with just a single pass even if those cavities would be like hundreds stacked in a row right?
If I'm not mistaken the largest part of the whole LHC electric bill goes to the bending magnets and the cryogenics to keep them superconducting?

 

[mfb]

How precise does the magnet field increase need to be around the whole circle because a mismatched field strength between say one part of the torus with another might result in the proton bunch crashing into the wall?

The average curvature radius is ~3 km. The beam pipe is about a centimeter wide. That's means the magnetic field needs to match the energy better than one part in 10⁵.

but the problem is that the cavities cannot transfer enough force on the small protons with just a single pass even if those cavities would be like hundreds stacked in a row right?

Cavities for the ILC are reaching ~35 MeV/m. Divide 7 TeV by 35 MeV/m and you get a linear accelerator 200 km long. And then you need two of them. It also means protons are single-use now: At 10¹¹ protons every ~30 ns the beam power would be 3.7 TW per beam. You can recover some of that energy, but its still completely unrealistic in every aspect.

If I'm not mistaken the largest part of the whole LHC electric bill goes to the bending magnets and the cryogenics to keep them superconducting?

Yes. It needs less power than LEP despite having much stronger magnets because superconducting magnets are so energy-efficient.

 

[artis]

@mfb or others, I just came to think of one more thing while reading about detectors, the beam pipes are going through the detectors (obviously) , detectors like CMS or ATLAS, are all detectors located on/around the beampipe or are some adjacent to it?

But here is the thing, as we spoke earlier here , I understood that once the SPS fills the main ring which has two beam pipes where protons fly in opposite directions through each pipe , then the bending/focusing magnet system has to slowly adapt it's strength as the particles get accelerated , while this is happening the particles are not colliding obviously but circling around the beam pipes gaining energy while passing accelerator cavities each time.
Say now we reach the point where particles are up to speed , is there a special magnet system before or inside the detectors that are around the beam pipes like ATLAS that are then switched on to "steer" the opposite beams closer together so that they actually collide within the detector?

Because as I understand that during acceleration the particles pass through (or around?) the detectors as they should in order to not lose energy but then when they are at their top speed they are (diverted? ) so that they collide, can you please give me more insight on how this works?

 

[BvU]

Hi,

A bit strange to continue such an old thread... here is my two cents -- from memory :

SPS fills the main ring which has two beam pipes where protons fly in opposite directions through each pipe

The last collider with two ring pipes was the ISR at CERN (Intersecting Storage Rings). So pp collisiions could be studied there.

Next step was SPS which is a synchrotron: one beam would be accelerated and then deflected to off-ring targets.

To use the SPS as a storage ring, to study proton-antiproton collisions in around-beampipe detectors, first the stochastic cooling process was developed (Simon van der Meer Nobel prize) to allow usable antiproton beams.
In a storage ring the collision points are straight sections where the beams are focused (with quadrupole magnets) to achieve a maximum of collisions.

Next was LEP (electron-positron) which was transmogrificated to LHC (proton-antiproton)

are all detectors located on/around the beampipe or are some adjacent to it?

around is better: you want to cover as many decay products as possible.Google around at CERN. They do their utmost to explain everything.

 

[mfb]

The last collider with two ring pipes was the ISR at CERN (Intersecting Storage Rings). So pp collisiions could be studied there.

You mean "the first"?
LHC has two beam pipes, too.

All the main detectors surround the beam pipe because you need to reconstruct particles in all directions. Transverse momentum sums to zero: Whenever there is a particle flying down there is something flying up as well. Missing that would be really bad. Even if it's a neutrino which escapes undetected you want to use the absence of a signal in your detector as information. A few specialized detectors are at one side of the beam pipe only.

The beam pipe close to the detectors has steering magnets (just like every other place of the ring). While filling the LHC and while accelerating the particles the beams are steered to avoid each other. Only once their parameters are suitable for collisions they are steered into each other.

Collisions don't reduce the energy of protons. Significant interactions completely remove the protons from the beam, or even destroy the protons altogether. But colliding bunches make the beam optics more complicated.

 

[artis]

@BvU Well the thread is not that old + I'm on topic (I hope).
Yes I am reading the CERN site quite often, hey have many PDF's with technicalities in them that I like.

@mfb I read about the detectors themselves , true it seems they are (the main ones) symmetrical and around the beampipe in order to catch the particles/ showers. The construction of them requires another separate thread which i might make.
As for what you said about the transverse momentum I gather essentially that in a bunch collision , the protons either smash and convert to other particles or they miss and then what? What happens to the ones that miss and don't get converted do they simply hit the beampipe or are they steered by magnets back into the pipes with some lower velocity and are essentially reused?

A proton at that energy even if it misses the other one for a head on collision say it strikes the pipe wall , wouldn't it still produce some radiation?Oh and also since the main detectors are all located some distance apart but on the same beampipes is it that they can only "fire" one detector at a time for a given prepared (accelerated to max energy) proton bunch?
From what I read on their site about the time of preparation , given everything goes perfect I suspect they can at maximum fire just a few shots a day ?

 

[mfb]

Out of over 100 billion protons in a bunch just a few tens of protons collide "hard". A similar number gets deflected notably and hits the wall somewhere - that leads to some secondary radiation, sure. The rest just keeps going with the same energy and momentum as before. The LHC beams are stored in collision mode for hours, sometimes even for over a day. During the whole time the protons make 11,000 turns per second, crossing a bunch from the opposite beam in at least 2 but usually 4 of the big detectors each time. All detectors have collisions the whole time until the beam gets dumped.

 

[artis]

@mfb That I did not imagine that there might be so much protons within a single bunch, and for 4each collision there are two oppositely incoming bunches so that number x2?

I am reading CERN's page
https://home.cern/news/news/accelerators/lhc-report-full-house-lhc
It says the beams size is 2.5 micrometers down from 3.5. I think that the size of the beam is what they relate to the so called "luminosity" , can this be thought of to being similar to the old CRT's where I had to adjust the focusing to make the electron beam sharper so that more electrons were hitting a smaller spot making it brighter on the screen? Just an analogy that came to mind from some CRT experience.

But the 2.5 micrometers size is meant for the transverse size of the beam I suppose? Could I call it the beam's cross section? I can't seem to find the longitudinal beam size (bunch size) I only read each bunch is separated in time and that is 25ns.
So while the LHC beams are stored in the collision mode as you say, it means that this whole time the RF cavities work to recirculate the bunches? From what you say I take that during each collision so few protons get affected that the bunch is largely intact after the collision and the spacing between bunches is also not disrupted so they can then just recirculate , do they fill in new protons from the SPS in the main ring during this time?

 

[BvU]

You mean "the first"?
LHC has two beam pipes, too.

Oops

I'm on topic (I hope).

Yep ! And you ask good questions, so keep it up !

 

[mfb]

I have wrote a few articles about the LHC and how it collides protons:
First one, also links to the others.

2.5 micrometers is the transverse beam size, yes. Along the beam direction it's a few centimeters.
It's a little bit like focusing a beam in a CRT, yes.

So while the LHC beams are stored in the collision mode as you say, it means that this whole time the RF cavities work to recirculate the bunches?

They only need to cover energy loss from synchrotron radiation, which is quite small.

do they fill in new protons from the SPS in the main ring during this time?

Check the earlier discussion about the magnet strength and energy again.

 

[artis]

Ok @mfb I read the links you gave. Although the questions I'm about to ask haven't been asked before I think.
Please look at this link.
https://indico.cern.ch/event/320137...ts/617127/849184/DetectorLectureSalvatore.pdf

1) In page 98 it shows the table of most frequent particle collisions and decays I suppose, so seems like inelastic scattering is the dominant, what is the second denoted by bb(anti)?
2) Then at page 106 it is written the approximate bunch length for two different energies, so at 450GeV the length for a single bunch is 11.2cm and for 7TeV the bunch length is down to 7.6cm, is this due to relativistic reasons where the proton electric field is decreased in the forward direction and increased in the transverse direction to motion?
If this is so , then how does that change the bunch/bucket interaction with the accelerating cavity? Are the phase changed or adjusted in the cavities as the proton bunch length changes or is it kept the same and as long as the arrival timing is correct the bunches get accelerated irrespective of their length?3) Then please go to page 114, it shows there the signatures of the different particle jets as they pass through the layers of semi strips, calorimeter absorbers/scintillators etc , now I see the letters u,d,s,b and c which supposedly should refer to the types of quarks , also gluons are mentioned, although it seems that only 5 quarks are named, namely, "up", "down", "charm", "strange" and "bottom" but the mention of "top" quark is missing, why is this so?
Are there no "top" quarks present in the LHC showers? Has this anything to do with the much higher mass of the top quark compared to all the other quarks? 4) Oh and finally , please see this paper
https://cds.cern.ch/record/1213275/files/p141.pdf
Of the various types mentioned I am most interested in the beam current transformer, see starting from page 8.
Are they employed in the LHC main ring? And can they be considered similar to the current transformers present in AC power lines for current monitoring? It seems to me they should be similar, since the proton bunch is just a pulse of current?
Also they are talking about a gap in the beam pipe which is necessary I think because otherwise the beampipe would block most of the B field necessary for readout?
Although what is most interesting is how is the gap in the beam pipe carried out given the beampipe must be leaktight and hermetical as it is under vacuum?
Maybe someone can share some specific as well as general info here on how pipe gaps for transformers or ordinary pipe ends are brought together to form seals in the LHC? As I think the LHC main rings consist of sections.

thanks.

 

[mfb]

what is the second denoted by bb(anti)?

There is only one particle that's commonly called "b". And you already identified the other as its own antiparticle. This is the cross section for collisions that produce b quark pairs (plus whatever else).
Note that all the categories there are part of the inelastic cross section.

The field of individual protons is irrelevant for the overall bunch. You would need to go through accelerator designs to find the reason why the bunch can be a bit shorter at higher energy. The bunches are smaller than the RF buckets so they can always arrive at the right time.

but the mention of "top" quark is missing, why is this so?

What is the lifetime of a top quark? What happens then?

Are they employed in the LHC main ring?

I haven't seen them but can't rule it out. Check LHC design papers.

 

[artis]

@mfb Well the top quark has a said lifetime of 5×10−25 s so with that I think it decays before it hits the walls of the beampipe within a detector , yes by the way are the walls of the beampipe thinner in the part that's inside the detetctor? Or dot he wall thickness plays no role given the energy of the collision so all the particles can escape and continue through the detector? So again would it be fair to say that the top quark is a particle much like the W+- and Z bosons that are only ever "seen" by analyzing the interactions and energies of the known basic particles actually interacting in the detector because the top quark as well as the bosons never make it to the detector in the first place?What about the bottom quark and it's anti bottom counterpart, do these make it to the detector given they decay from the top quarks which themselves surely don't make it?

 

[mfb]

by the way are the walls of the beampipe thinner in the part that's inside the detetctor?

Thinner and made out of beryllium. You can find this in the detector design reports. Beryllium comes with a smaller chance of interactions in the beam pipe.

So again would it be fair to say that the top quark is a particle much like the W+- and Z bosons that are only ever "seen" by analyzing the interactions and energies of the known basic particles actually interacting in the detector because the top quark as well as the bosons never make it to the detector in the first place?

Sure.

What about the bottom quark and it's anti bottom counterpart, do these make it to the detector given they decay from the top quarks which themselves surely don't make it?

This is again something you can calculate yourself.

 

[artis]

@mfb just give me a hint on how would I calculate that ? I guess I would first need to know the original particle , in this case the top quark although I read that first a gluon is created at the collision of two high energy protons then from the gluon decay top and anti top quark are created which then themselves decay to W+ and W- and bottom and anti bottom respectively.

Now the decay times are given for each of these particles , except for the said gluon, can you tell me what happens there? Is that gluon purely theorized?

But knowing the decay time is not enough , I think I also need to know the speed of the particle to know whether it will be able to hit the beampipe while it exists or travel further outside. How do I , or anyone else for that matter know the speed of the particles, given the first particles that can actually be observed are 2 to 3 decayed and unobserved particles away from the actual first proton interaction?

 

[mfb]

It's a virtual gluon, it doesn't exist. The top quarks decay after flying less than the diameter of a proton - completely negligible. They decay with such a high energy that the b mesons travel at close to the speed of light. You'll need some very rough energy estimate for them to account for time dilation. Assume that they get e.g. half of the top mass as energy, you are not writing a publication here and being off by a factor 2-3 doesn't matter. That's good enough for a rough estimate. Then you look up the lifetime of B mesons and calculate how far they fly.

 

[artis]

@mfb this confuses me , so after a proton - proton collision (assuming the energy was high enough) how many top quarks emerge? One or two? And if two then is one the top and the other anti top all times?
https://en.wikipedia.org/wiki/Top_quark#/media/File:Top_antitop_quark_event.svg
The reason I ask is because of what I see in this picture.As the emerged top quark decays, it decays to a b meson and W boson, The W boson being + and -, is it the case that the W+ is always created from the top quark while W- from the anti top quark aka top quarks antiparticle?

So the top quark decays to a W boson and a b meson?

https://en.wikipedia.org/wiki/B_meson
It seems there are 4 types of b mesons , what determines which ones get produced when?So the speed of the b meson aka it's KE is that of the energy given to it by the top quark decay (say half of 172GeV) minus it's own rest mass about 5-6 MeV?
So I could calculate it's KE by the formula KE=mc2-m0c2 ?

 

[mfb]

This forum cannot replace a course in particle physics. Questions about details or descriptions you don't understand are perfectly fine, but you'll need to learn at least the basics from a book or similar.