2014 machine commissioning and 2015 start up

The place to discuss the LHC. Commissioning, operation, issues, events ....
User avatar
Tau 2
Posts: 34
Joined: Sat Apr 04, 2015 8:17 pm
Location: Heemskerk

Re: 2014 machine commissioning and 2015 start up

Post by Tau 2 » Fri Apr 17, 2015 12:41 pm

I'll try to explain what I understood from the elogbook in the first weeks of commisioning. Most of the systems mentioned here are being adjusted, tested, repaired when needed, and procedures for using them are written.
First, there are the big working magnets: the bending dipoles and focusing quadrupoles. These are more or less fixed (but they change during a ramp, see below).

Then to keep the beam in the tube, you basically need four devices:
- The BPM (Beam Position Monitor), that measure the position of the beam
- The collimators, that stop particles that go too far off (these can be moved very precisely)
- The orbit corrector magnets (MCBH and MCBV) to correct the beam
- The Beam Loss Monitors BLM, that detect if the beam has hit the tube or a collimator
Now the problem is that the BPMs are not simple devices that give exact positions in millimeters. To calibrate them, you have to compare what they measure by using the collimators to touch the beam, and see when the BLM detects it. Also, the signals of the BPM are compared to each other (with a beta beating measurement).

Once you have done all that, it is time to control the beam a bit further, by doing the more advanced adjustments:
- tune (oscillation of the beam around the ideal line)
- dispersion (sideways movement particles that go slightly faster or slower)
- chromaticity (change of the tune for faster or slower particles)
- RF (keeping the bunches into buckets so they can be accelerated, and moved together in the longitudinal direction)

Since all these things are related to each other, and all of these are separate systems, a lot of work is spent to make everything work nicely together.
And of course, during a ramp, everything has to continuosly adjusted, since almost everything changes then. That's why they do "practice ramps" with a non-dangerous amount of protons.

Two very important systems are used at the beginning and the end:
- injection (with the MKI magnets): if the beam is not put nicely into the machine, there is no way improve the quality
- dumping (first with the MKD magnets to get it out, then the BTVDD and BPMD to see if it worked, and the dump block to absorb the energy). The dumping system must safely take the beam out of the machine under all circumstances, whatever is the reason of the dump: It may not fail even if anything else fails. Only if the dump system works flawlessly, a more dangerous beam can be injected.

(And then, there are a few more control systems, such as the ADT, that electrostatically nudges the beam a tiny bit, the skew quadrupoles to adjust horizontal-vertical interactions, and the sextupoles, octupoles, decapoles and even 12-poles for precisely controlling the beam, and the special squeezing magnets for the experiments, and the undulator, and so on...)

I hope this helps you understanding the 8 weeks a bit more.
- Tau

josch222
Posts: 78
Joined: Sat Dec 12, 2009 4:10 pm

Re: 2014 machine commissioning and 2015 start up

Post by josch222 » Tue Apr 21, 2015 5:59 am

pcatom wrote:
Tau 2 wrote:But the good news is that these 8 weeks are going to be very exciting!
In a way, they're more interesting than the production runs.
I love it see how they manage to learn from every measurement (and every mistake) they make.
One example is how they handled the earth fault in sector 23: it must have been pretty scary to decide to blow it up, but they really made sure they knew what they were doing! Those things make for pretty interesting stories to follow.
Yes ... it was a bit scary - imagine putting 50kW through 2 wires with a cross section of 0.25mm2 even if it was only for 1/100th of a second. Then imagine doing it in a bath of liquid helium ...
I think the helium helped to prevent metal vapor or soot (from isolation material) deposition on
nearby surfaces. I have seen this in vacuum relays, after some times switching high currents
you can measure much lower isolation resistance between the open contacts.
On the other hand you may have to deal with sudden pressure rise and shock waves in liquids
(oh, btw, what about shock waves in superfluid liquids?)

I have used the basic technique of burning away shorts with a PCB that showed one between two traces.
It was not possible to locate it, even with a microscope. Had a resistance in the low Ohms
between two traces a few 1/10mm apart. There was no need to "rescue" that PCB, I just tried to
find out what the failure was. So I simply took a power supply with a large
output capacitor and burned it away. I expected to be able to see the location after that,
there should have been traces under the solder mask or some measurable lower isolation resistance.
But nothing to see. I think it was a single, very fine fiber of copper.
I switched PCB manufacturer after that.
Data from the system it was designed for can be seen online for a few hours today, it is some kind
of complicated fuel gauge:
http://www.solarimpulse.com/widget-energy

dukwon
Posts: 163
Joined: Fri Apr 10, 2015 1:04 pm
Contact:

Re: 2014 machine commissioning and 2015 start up

Post by dukwon » Wed Apr 22, 2015 1:15 pm

Some unfortunate news from the LHC morning meetings today and yesterday (available here https://indico.cern.ch/event/385865/).

There is an "Unidentified Lying Object" (ULO) of ~4 mm height and ~14 mm width sitting somewhere in one of the beam pipes. This could be large enough a problem to require an intervention to remove it. This may mean a long delay :(

User avatar
Tau 2
Posts: 34
Joined: Sat Apr 04, 2015 8:17 pm
Location: Heemskerk

Re: 2014 machine commissioning and 2015 start up

Post by Tau 2 » Wed Apr 22, 2015 4:25 pm

It might be not so bad. I see no plans to stop the operations (yet).
For details, see the thread "MUFO?"
- Tau

dukwon
Posts: 163
Joined: Fri Apr 10, 2015 1:04 pm
Contact:

Re: 2014 machine commissioning and 2015 start up

Post by dukwon » Tue May 05, 2015 2:02 pm

We had (non-accidental) collisions today at 450 GeV :D

Shame we at LHCb couldn't complete the SMOG test (SPS is down, have to wait until tomorrow) but we got some time-alignment data, which is nice.

User avatar
Tau
LHCPortal Guru
Posts: 200
Joined: Tue Jan 05, 2010 10:16 pm
Location: Heemskerk, Netherlands

Re: 2014 machine commissioning and 2015 start up

Post by Tau » Fri May 29, 2015 9:28 am

And yesterday we had collisions with a beta of 80 cm.
The beta is a measurement for the focus of the beam.
If they say "we are at 80 cm", they mean "the beta is 80 cm at the collision points". The lower it is, the more focusued the beams are. Previous run the lowest possible value was 80 cm, but since the stop they are aiming at 60 cm, giving about twice as many collisions.

Currently, they are working on the steering of "bunch trains": sequences of bunches that are close together. Last year they worked with bunches that were 50 ns (or 15 meters, if you want) apart, the plan is to go to 25 ns. The number of bunches is still limited in order to protect the machine, but the plan is to try trains 6 bunches of 50 ns apart and 12 bunches of 25 ns apart today.
Those trains are slightly harder to control than single bunches, so some fiddling will be needed to get them in neatly.
And then, they will continue measuring the "aperture" to make sure the beam has enough room everywhere in the machine for full energy, checking the abort gap protecting to ensure safe dumps, and other precision measurements.
So, slowly but steadily, we are getting there...
- Tau

User avatar
Tim Bergel
Posts: 29
Joined: Mon Nov 23, 2009 1:31 pm

Re: 2014 machine commissioning and 2015 start up

Post by Tim Bergel » Mon Jun 01, 2015 4:41 pm

Thank you for the information, Tau. Can you explain something? If 25 ns bunch separation does not allow you to have more bunches, then what does it achieve? From where I'm sitting it seems likely that 25 ns will be harder, so I assume there is an up-side...

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

Re: 2014 machine commissioning and 2015 start up

Post by jmc2000 » Mon Jun 01, 2015 9:22 pm

Tau wrote:And yesterday we had collisions with a beta of 80 cm.
The beta is a measurement for the focus of the beam.
If they say "we are at 80 cm", they mean "the beta is 80 cm at the collision points". The lower it is, the more focusued the beams are. Previous run the lowest possible value was 80 cm, but since the stop they are aiming at 60 cm, giving about twice as many collisions.
I'm confident that you guys at the LHC will work wonders in getting the LHC to surpass its expected performance as you did in 2010-2012.

What is the hoped for integrated liminosity/week at the end of this year compared to 1fb/week at the end of 2012?

Cheers,

Jmc

Kasuha
Posts: 570
Joined: Tue Dec 15, 2009 1:22 pm

Re: 2014 machine commissioning and 2015 start up

Post by Kasuha » Mon Jun 01, 2015 10:38 pm

Tim Bergel wrote:If 25 ns bunch separation does not allow you to have more bunches, ...
Ultimately it does. They'll just not put too many of them in right now.

User avatar
DCWhitworth
LHCPortal Guru
Posts: 599
Joined: Mon Nov 30, 2009 8:13 am
Location: Norwich, UK

Re: 2014 machine commissioning and 2015 start up

Post by DCWhitworth » Wed Jun 10, 2015 8:27 pm

Tim Bergel wrote:Thank you for the information, Tau. Can you explain something? If 25 ns bunch separation does not allow you to have more bunches, then what does it achieve? From where I'm sitting it seems likely that 25 ns will be harder, so I assume there is an up-side...
It does allow more bunches but with less protons per bunch. This will reduce 'pile up'. One of the issues they have is that as the bunches cross in the IPs they are getting multiple collisions per bunch, this can be rather hard to untangle with the detectors.

With 25ns spacing they will be able to have more bunches of less intensity and get more collisions but with less per bunch if that makes sense.

They originally struggled to get 25ns bunches working well so one of the things they did was decide to stick with the 50ns spacing but ramp up the intensity of the bunches beyond what the machine was originally intended to do. This gave them higher luminosity overall but at the cost of increasing pile up. With the higher confidence they now have in getting 25ns working they should be able to lower the average bunch intensity (in fact IIRC they can't run 25ns bunches at the same intensity as 50ns ones). But they'll end up with higher luminosity and lower pile up, so win-win.
DC

The LHC - One ring to rule them all !

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

Re: 2014 machine commissioning and 2015 start up

Post by jmc2000 » Mon Jul 27, 2015 11:59 pm

Does 25ns scrubbing also involve machine development at the same time that will ultimately aid the intensity ramp up?

Regards

jmc

User avatar
Tau
LHCPortal Guru
Posts: 200
Joined: Tue Jan 05, 2010 10:16 pm
Location: Heemskerk, Netherlands

Re: 2014 machine commissioning and 2015 start up

Post by Tau » Fri Jul 31, 2015 5:32 am

jmc2000 wrote:Does 25ns scrubbing also involve machine development at the same time that will ultimately aid the intensity ramp up?
Yes, getting those scrubbing beams in is quite a challenge. For scrubbing to be effective, they like high intensity.
Now they are using "doublet" beams that consists of pairs of bunches that are only 2.5 ns apart instead of the normal 25 or 50 ns.
These doublets give different BPM readings, and therefore needs adjustments on the beam control system. Furthermore, the higher intensity requires readjustment of the collimator and BLM settings.

PS, in case you missed it:
BPM: beam position monitor, to determine position of the beam in the tube, used for steering
BLM: beam loss monitor: to detect if the beam is out of control
collimator: barrier on the side of the beam to block stray particles
- Tau

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

Re: 2014 machine commissioning and 2015 start up

Post by jmc2000 » Fri Sep 11, 2015 8:58 pm

Fill #4349 with 745 25ns bunches is currently still looking beautifully stable with low losses after 4 hours: impressive! Should mean at least around 0.5 fb/week from now on, possibly matching 1 fb/week in 2012 by the end of the year!

Does anyone know the number of bunches and Beta* they're going to try for over the coming weeks?

Are there other ways of increasing the intensity of the beams this year?

Cheers,

JMc

mfb
Posts: 168
Joined: Sat Oct 03, 2015 7:58 pm

Re: 2014 machine commissioning and 2015 start up

Post by mfb » Sat Oct 03, 2015 8:36 pm

jmc2000 wrote:Fill #4349 with 745 25ns bunches is currently still looking beautifully stable with low losses after 4 hours: impressive! Should mean at least around 0.5 fb/week from now on, possibly matching 1 fb/week in 2012 by the end of the year!

Does anyone know the number of bunches and Beta* they're going to try for over the coming weeks?

Are there other ways of increasing the intensity of the beams this year?

Cheers,

JMc
1465 bunches now, with up to 35% design luminosity at the start of fills. Looks like ~0.1/fb per day is possible. They might add 144 more bunches after the weekend, maybe a bit more later.

1.7/fb collected so far, and 3.5 weeks left for data-taking, so ~4/fb are possible. There was the idea of decreasing beta* to 60 cm or 40 cm, but I didn't hear anything new recently so I doubt they still want to try that. Would probably need some time to set up.


In 2015, CMS seems to get a consistently lower luminosity than ATLAS. Is there some reason for the difference?

andrewp
Posts: 40
Joined: Sat Jul 17, 2010 10:10 pm

Re: 2014 machine commissioning and 2015 start up

Post by andrewp » Mon Apr 18, 2016 5:31 pm

Rampy McRampface

Post Reply