Improved performance after MD and TS?

The place to discuss the LHC. Commissioning, operation, issues, events ....
Post Reply
jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Improved performance after MD and TS?

Post by jmc2000 » Sun May 08, 2011 3:12 pm

This coming Sat, the LHC will be back in fills for physics, so afer the technical stop and machine development phase should be expect any improvement in the performance of the machine?

I've also noticed that the plan was to get to 900 bunches with 75ns, whereas they were using 50ns so will they stop below 900?

User avatar
PhilG
Posts: 135
Joined: Sun Dec 06, 2009 9:58 pm
Contact:

Re: Improved performance after MD and TS?

Post by PhilG » Sun May 08, 2011 7:54 pm

I don't have any inside info and the planning meetings are restricted access now, but I understand that the plan is to keep stepping up bunches until they reach about 1380 using 50ns.

There has not been much info about how all the MD tests will affect that plan. It was noticed that they tried out higher intensity bunches with up to 170 billion protons (ultimate intensity). These are the highest they can go with the current hardware, but they have been using 115 billion (nominal intensity) for physics so far. This might mean that they could try to increase the bunch intensity at some point, but when and by how much has not been made public. They may just have been using these bunches to analyse beam stability with no plan to use them for physics in the near future.

One of the MD tasks was unsqueezing tests for TOTEM and ALFA so they are likely to schedule in a physics run for them soon. These are different physics experiments aside from the main proton physics program. We also know they planned a Van der Meer scan after the technical stop. It would be nice to know a little more of the official plan but speculation and surprises are fun too. :dance:

pcatom
LHCPortal Guru
Posts: 68
Joined: Fri Nov 27, 2009 8:33 pm
Location: Geneva

Re: Improved performance after MD and TS?

Post by pcatom » Mon May 09, 2011 8:23 pm

I don't have any inside info and the planning meetings are restricted access now, but I understand that the plan is to keep stepping up bunches until they reach about 1380 using 50ns.

There has not been much info about how all the MD tests will affect that plan. It was noticed that they tried out higher intensity bunches with up to 170 billion protons (ultimate intensity). These are the highest they can go with the current hardware, but they have been using 115 billion (nominal intensity) for physics so far. This might mean that they could try to increase the bunch intensity at some point, but when and by how much has not been made public. They may just have been using these bunches to analyse beam stability with no plan to use them for physics in the near future.

One of the MD tasks was unsqueezing tests for TOTEM and ALFA so they are likely to schedule in a physics run for them soon. These are different physics experiments aside from the main proton physics program. We also know they planned a Van der Meer scan after the technical stop. It would be nice to know a little more of the official plan but speculation and surprises are fun too. :dance:
After the technical stop the weekend will (all being well!) be spent re-establishing physics conditions (test ramps, loss maps etc. ) and doing systematic VdM scans to calibrate the luminosities. After that the priority will be first of all to re-establish 768 bunches, do a few fills like that - then move up to the next step - which i think is 912 bunches. With this the 10+33 cm-2 s-1 should be within reach. After that - continue pushing up ... over 1300 bunches are possible with the 50ns schemes.

Dedicated periods for the 90m physics will be arranged - but not scheduled yet.

The studies with very intense bunches are for the future - no plans to push higher bunch intensities into operation for the moment.

Is that enough detail??

tomey36
Posts: 108
Joined: Wed Apr 14, 2010 11:39 pm

Re: Improved performance after MD and TS?

Post by tomey36 » Tue May 10, 2011 1:28 am

Dude your awesome! I just have one question, do have any information on the ATS studies they did. how are they differnt from the normal squeez operation?

User avatar
PhilG
Posts: 135
Joined: Sun Dec 06, 2009 9:58 pm
Contact:

Re: Improved performance after MD and TS?

Post by PhilG » Tue May 10, 2011 11:22 am

pcatom wrote:
Is that enough detail??
Excellent, Thanks for the clarification

pcatom
LHCPortal Guru
Posts: 68
Joined: Fri Nov 27, 2009 8:33 pm
Location: Geneva

Re: Improved performance after MD and TS?

Post by pcatom » Tue May 10, 2011 6:34 pm

Dude your awesome! I just have one question, do have any information on the ATS studies they did. how are they differnt from the normal squeez operation?
The minimum Beta* in the high luminosity IP's is limited by the aperture in the triplets, but also by the chromatic aberrations introduced by the very tight squeeze. Basically, even with larger aperture triplets, the minimum Beta* is around 40cm since the lattice sextupoles run out of strength to correct the chromatic aberations coming from the squeezed IP's.

For the upgrade of the LHC larger aperture triplets are planned. it would be nice to be able to squeeze the beams much further than the present limit. The ATS is a novel scheme to bypass the strength limit of the existing lattice sextupoles and allow a much smaller beta*. Values of around 15cm should be possible with round beams (x10 lower than the present beta*).

The accelerator physics is quite complicated - there was a presentation in Chamonix 2011 on this subject ... Basically the optics in the two arcs adjacent to each high luminosity IP are distorted in such a way as to increase the effective strength of the lattice sextupoles in these arcs. This can then be used to correct the chromatic aberrations. The price to pay is a reduction in aperture in these arcs. However, at high energy, the aperture in the arcs is very large - the aperture limitation in the machine is clearly in the triplets. Giving up some aperture margin is not really a problem.

The ATS scheme therefore involves modifying the optics over a large fraction of the machine and is a much more complex undertaking than the standard insertion squeeze.

By the way, you might ask why the lattice sextupoles cannot just be replaced with stronger ones ... each quadrupole in the arcs is equipped with a lattice sextupole which are arranged into regular families. Replacing the sextupoles with stronger field ones (assuming you could make them) would therefore involve changing every arc quadrupole in the LHC ... big job.

Hope this helps

tomey36
Posts: 108
Joined: Wed Apr 14, 2010 11:39 pm

Re: Improved performance after MD and TS?

Post by tomey36 » Tue May 10, 2011 8:39 pm

Yes thank you for that detailed explanation :) It never ceases to amaze me how complex of a machine the LHC is, just when i think i got a handle on things you all add some new process lol.

jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Re: Improved performance after MD and TS?

Post by jmc2000 » Wed May 11, 2011 12:27 am

pcatom wrote:
Is that enough detail??
Hey, thanks for the useful info.

At the Tevatron, they've managed to limit 1 quench to around 20 fills. Do you know what the figure is for the LHC and if they have a realistic chance of reducing it? They were also managing to get remarkable fill times of 70% so I'm curious to know if the guys at Tevatron are consulting the LHC guys on how they managed this.

Good luck with the progress at the LHC

pcatom
LHCPortal Guru
Posts: 68
Joined: Fri Nov 27, 2009 8:33 pm
Location: Geneva

Re: Improved performance after MD and TS?

Post by pcatom » Thu May 12, 2011 4:46 pm

At the Tevatron, they've managed to limit 1 quench to around 20 fills. Do you know what the figure is for the LHC and if they have a realistic chance of reducing it? They were also managing to get remarkable fill times of 70% so I'm curious to know if the guys at Tevatron are consulting the LHC guys on how they managed this.
There has never been a beam induced quench in the LHC during normal physics operation. There have been quenches with beam; the first was right at the start of commissioning - firing a pilot bunch directly into a magnet at injection. Another case was when there was a flashover of the injection kickers and 36 nominal bunches were mis-injected (a known failure case). At high energy the only quench was a test to try to calibrate the quench limit vs beam loss threasholds for the fast loss cases.

This is not to say that the LHC will never have beam induced quenches - but up to now the sophisticated protection and collimation systems are doing their job rather well.

User avatar
PhilG
Posts: 135
Joined: Sun Dec 06, 2009 9:58 pm
Contact:

Re: Improved performance after MD and TS?

Post by PhilG » Thu May 12, 2011 9:08 pm

If the question is about performance perhaps it should have been about any premature end to a fill, not just quenches.

jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Re: Improved performance after MD and TS?

Post by jmc2000 » Thu May 12, 2011 11:50 pm

PhilG wrote:If the question is about performance perhaps it should have been about any premature end to a fill, not just quenches.
Yes, I'm interested in the causes for a fill to be abandoned and the beam having to be dumped/quenched. Quenched and dumpled are the same thing, correct?

If you look at slide 15 in this talk given by Alexander Valishev for the Tevatron group, it has a very interesting run down of what was responsible for quenches:

http://www.c-ad.bnl.gov/pac2011/proceed ... 2_talk.pdf

Beam, component, cryo seem to be the major three with 154 quenches per 1200 stores which still looks pretty reliable.

It's going to be an exciting next few weeks to see what the talented operation's people at the LHC manage to do to get around new problems as they push for higher luminosity and run times.

User avatar
PhilG
Posts: 135
Joined: Sun Dec 06, 2009 9:58 pm
Contact:

Re: Improved performance after MD and TS?

Post by PhilG » Fri May 13, 2011 6:37 am

Quenched and Dumped are not the same thing. I think you are interested in cases where the beam has to be dumped before the run is complete. This happens for lots of reasons. A quench is a more serious event where the energy in the magnets has to be dumped quickly.

There are some important differences between the Tevatron and the LHC that are important if you are going to compare performance efficiency. The Tevatron uses anti-protons in one beam and it takes a long time to build up a sufficient store of anti-protons. This means that the Tevatron can only do one good run each day. The luminosity lifetime at the Tevatron is 5 hours but they often run for over 15 hours. There is much less data collected in the last half of the run so it is not as efficient as you might hope.

The LHC on the other hand does not need anti-protons and it can build new proton bunches very quickly. This means that if a fill is dumped it can get back up and running quite quickly. Also, the luminosity lifetimes have been much longer than the Tevatron, more like 15 to 20 hours I think. This means that in a long run they continue to collect lots of data for much longer. Potentially the LHC is therefore much more efficient than the Tevatron, but all depends on how often they lose the fill and how much delay there is in restarting. They are still at an early commissioning phase so problems are quite frequent. The Tevatron has had 25 years to iron out its problems.

Post Reply