Choose timezone
Your profile timezone:
20th quench behavior team meeting
→
Europe/Zurich
DRAFT
Date: 2018-01-08, 09.00-10.30
Place: 927
Presents: G. Willering, P. Hagen, E. Todesco, M. Modena, P. Ferracin, S. Le Naour, D. Tommasini, A. Devred, M. Pojer, O. Bruning, M. Bajko, J. C. Perez
Hardware commissioning campaign of December 2018 [M. Pojer]
Analysis of sector 12 retraining [E. Todesco]
Ezio presents the hypothesis used to provide the estimate of 550 quenches to reach 7 TeV for the whole LHC. This estimate relies on the data of HC campaigns to 11.08 kA of 6 sectors, and to about 11.5 kA for two sectors. Data are split by manufacturer, second quenches (i.e. magnet quenching second time ) are removed and analyzed separately, and the production of split in five batches: one of 1000 series, a slow and a fast for 2000 series, and a slow and a fast for the 3000 series. The slow batch of the 3000 series also features almost all the second quenches. The data of each batch are interpolated using a modified Gaussian distribution, and extrapolation can be given at 7 TeV. Note that for the 1000 series and for the fast 2000 series batch, the few data do not allow any significant fit; therefore an arbitrary rate of 10% was taken as probability to quench to 7 TeV.
The sector 12 is made by an extremely uneven combination of batches. It contains 95 2000 series magnets, almost all from the fast batch. Moreover there are 50 1000 series magnets, and only 9 3000 series magnets.
For the 50 1000 series magnets, 5 quenches were expected to 7 TeV, that can be interpolated to 2 quenches up to 11.415 kA. Instead, 7 quenches were observed (plus one second quench). This revealed an unexpected batch of slow magnets, training above 11.2 kA, between 1350 and 1416 (last part of the production). The training of the other two sectors above 11.08 kA done in the previous years did not show any similar feature in the first part of the 1000 series production, so the phenomenon seems localized in this batch.
For the 95 2000 series magnets, only two quenches were observed. Therefore the assumption of a negligible training of these magnets (10%) is confirmed.
For the 9 3000 series magnets, the need of training is confirmed. We had 4 first quenches and one second quench, in line with the expectations.
Summarizing, the training campaign confirmed two positive features that give no evidence of showstoppers to reach 7 TeV:
-absence of single magnets blocking the training
-very high fraction of first quenches (85%)
On the other hand, the training showed the existence of a batch of 1000 series magnets heavily quenching above 11.2 kA. This batch appears to be limited to the very last part of the 1000 series production, namely about 70 magnets from 1350 to 1416.
Analysis of sector 12 retraining [P. Hagen]
Per recalls the most relevant features of the training campaign:
-a rate of 2 quenches per day, a total of 13 events in 7 days, with two events involving independent quenches in two different magnets
-a surprise in the 1000 series behavior, that until today always stayed in the shadow of the other two productions: half of the quenches belong to this series, even though the sector contains 1/3 of these magnets.
- a visible as a change of slope above 11.2 kA
Discussion and future actions: [all]
- check the behavior of the last batch of 1000 series in the SM18 data and in the memory of the engineers following the production
- review the estimates accounting for the new batch seen in 1000 series
There are minutes attached to this event.
Show them.
