2016 operation

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

Re: 2016 operation

Post by jmc2000 » Sun Jul 24, 2016 7:47 pm

mfb wrote: The BCMS scheme now reliably delivers 120% design luminosity.
Also a noticeable improvement in initial beam life time from 5 to 7 hours, giving an incredible 100\fb over just 2.5 hours. Fingers crossed they may even take it to 130%, but the current problem with cryogenics might be a sign of this being too unreliable in the long run.
JMc

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

Re: 2016 operation

Post by mfb » Sun Jul 24, 2016 8:15 pm

jmc2000 wrote:
tomey36 wrote:https://lpc.web.cern.ch/lumiplots_2016_pp.htm

Is this what you mean?
There is also the statistics page which is linked on the portal.
Something more like this:
https://op-webtools.web.cern.ch/vistar/ ... p?usr=LHC3

But a luminosity plot for a particular fill rather than over the last 24 hours.

JMc
https://op-webtools.web.cern.ch/vistar/ ... LUMINOSITY ?

I don't think 120% is the best they can achieve. And heat load tends to go down over time.

Edit: time lapse of the LHC lumi screen for 24 hours.

sciing
Posts: 27
Joined: Mon Dec 07, 2009 9:18 am

Re: 2016 operation

Post by sciing » Mon Jul 25, 2016 11:46 am

The turn around time decreased now to 3 hours, great!!
In the morning meeting slides are some curves for optimal fill length for 4h and 5.8h TAT for different fills.
slide 15
https://indico.cern.ch/event/549118/con ... L16_WH.pdf
An always optimum length has not much benefit. Between 8-25h the curve is nearly flat.
There is much loss by long TATs or fills shorter than 8h.

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

Re: 2016 operation

Post by mfb » Sun Aug 07, 2016 11:25 pm

Performance has been worse since the MD block, lower time in stable beams and the initial luminosities are lower as well. We'll probably get some improvement from the increased number of bunches soon.

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

Re: 2016 operation

Post by jmc2000 » Mon Aug 08, 2016 1:07 am

Recorded integrated luminosity/week ~1.6fb compared to ~2.5fb before the MDs which still looks pretty good, considering that there was even talk of replacing magnet D that would have taken 8 days, plus a few hundred hours of intensity ramp up.

I wonder how much of the recent MD will be used over the coming weeks to improve performance?

JMc

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

Re: 2016 operation

Post by mfb » Mon Aug 08, 2016 7:06 pm

The number of bunches went up for this run (2174), the luminosity improved only marginally. On the other hand, we have this weird ATLAS/CMS discrepancy, no idea which value to trust for overall normalization.

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

Re: 2016 operation

Post by jmc2000 » Wed Aug 10, 2016 4:07 pm

Suspected short on dipole 31L2 which could be bad news in taking 30 days to replace:

http://www.symmetrymagazine.org/article ... rom-itself
“In a bad accident, the beam could go off course and drill a hole through one or two magnets,” says Schmidt. While this would not destroy the LHC, it would still require time and money for repair. Replacing a dipole magnet, for which CERN has 30-40 spares, would take 30 days. A more complicated repair, or replacement of a less common component, would take longer.
Fingers crossed it isn't this serious.

JMc

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

Re: 2016 operation

Post by mfb » Wed Aug 10, 2016 4:16 pm

It is an electric problem - the beam drilling a hole through the magnet would look differently.

Here is the last beam as it arrived at the beam dump (the image will change once they have the next dump, but then the problem has been fixed).

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

Re: 2016 operation

Post by jmc2000 » Wed Aug 10, 2016 5:59 pm

mfb wrote:It is an electric problem - the beam drilling a hole through the magnet would look differently.
A short on the winding of a dipole magnet will mean it being warmed up to room temperature, replaced, then recooled as with a drilled example - 30 days repair with most of the days needed for the warming and cooling of the helium for the dipole magnet, I'm assuming here :shifty:

JMc

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

Re: 2016 operation

Post by mfb » Thu Aug 11, 2016 9:27 pm

The problem is external. Current VISTARS message:
PS vacuum issue: no beam from PS until tomorrow.
Mitigation for RB.A12 under implementation to allow physics as soon as beam back in PS

More details at 8:30 meeting on Friday

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

Re: 2016 operation

Post by dukwon » Fri Aug 12, 2016 9:51 am

Quoting the relevant parts of today's morning meeting slides (which is most of it...)
Facts: Two quenches while ramping down RB.A12 from 6 kA with -10 A/s
• 10 June 2016 @ 547 A
• 3 Aug 2016 @ 295 A
The observations could be explained with a model in which an inter-turn short is included.
• The short is present at high current and absent at lower current
• The transition generates a voltage variation that the QPS interprets as a quench

Analysis of measurements, discussions, actions (Thursday)
• No signs of a (change in) inter-turn short observed in the cycles performed on 10/11 Aug 2016 and FPA tests at 760 A. Short might permanently be present (or not), but seems ‘stable’.
• A quench at high current in a magnet with an inter-turn short will destroy the magnet.
• Possibly a FPA at high current causes significant voltage on a magnet with an inter-turn short, resulting in a QPS trigger and subsequent heater firing, which will then also destroy the magnet.
• CONCLUSION: For operation in presence of an inter-turn short, we should reduce as much as possible the probability of a quench of magnet A31L2 as well as the probability of a FPA in sector 12.

Actions and mitigation measures
• Leave diagnostic system installed, to maximize probability of recording relevant events
• Remove global protection mechanism for IPQs and RQD/F in S12 to limit the likelihood of FPA in RB circuit (disabling preventive shutdown of adjacent circuits)
• Reduce BLM thresholds on complete MB and MQ arc families to end of Run1 values to minimize risk of magnet quenches due to UFOs
• Increase QPS thresholds on A31L2 from 100 mV to 235 mV to minimize risk of heater firings during FPA (or change of inter-turn short)

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

Re: 2016 operation

Post by mfb » Sat Aug 13, 2016 11:41 am

Okay, that is... interesting.

For now the LHC is running again, initial luminosity this morning was a bit above the design luminosity.

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

Re: 2016 operation

Post by mfb » Tue Aug 30, 2016 6:08 pm

Progress has been a bit slower the last weeks, but we are surpassing the initial integrated luminosity goal for this year these days. ATLAS shows 24/fb, CMS shows 27/fb, at least one calibration is off.

ATLAS and CMS

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

Re: 2016 operation

Post by jmc2000 » Sat Sep 03, 2016 9:06 pm

New record for instantaneous luminosity for CMS at ~13.5 hz/nb ! and it's back to averaging 2.5\fb/week as in early August. But it's difficult to know how much out of calibration CMS is, and whether the record is real.

Fingers crossed, >35\fb for 2016 is looking realistic which is x3 the data presented at ICHEP 2016 :)

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

Re: 2016 operation

Post by mfb » Sun Sep 04, 2016 5:01 pm

Comparing the different sources, always for delivered:

LHC statistics: ATLAS 27.2, CMS 27.77, LHCb 1.35
LHC lumi plots: ATLAS 27.1, CMS 28.8, LHCb 1.28

ATLAS: 26.6
CMS: 28.97
LHCb: 1.35

"Between 27 and 29 inverse femtobarn" is probably a reasonable guess for the dataset so far. In the schedule we have ~5 weeks more data-taking. Maybe 4-4.5 if we include two ramp-ups after TS2 and MD4. Looks like 35-40/fb by October 25th.


Combined with the 2015 dataset, LHCb should have a bit more B- and C-hadrons than in run 1 now.

Post Reply