Transcript SPDPSupgrade - Indico

MC work program for SPD/PS
Or a brainstorming on SPD/PS use for the trigger at
the upgrade after a discussion with Jacques
Thanks to O. Deschamps, S. Monteil
Hugo Ruiz, Elias López
1
Outline
• Introduction
• Use of PS/SPD:
– Offline
– At L0
• Trigger and upgrade
• Plans: utility of the PS/SPD at the trigger
• MC requests
2
Intro
• The hardware of SPD/PS would probably make it to
the upgrade (http://indico.cern.ch/conferenceDisplay.py?confId=43976)
– maybe some rotations of PMTs
• SPD & PS do not harm much:
• SPD: just a first sensitive ECAL layer
• PS: 12 mm (2.5X0) lead   no degradation of sE
for e± with E>10 GeV (calo TDR)
• Some electronics, fibers to be replaced  $
3
Offline use of the SPD
• Identify conversions after the
magnet
• On a sample of B0K*g:
(LHCb-2003-091)
Fract.
Converted
(SPD =1)
Non-conv.
(SPD =0)
30%
70%
Purity
10%
(backsplash)
20%
(e in other cell)
 SPD hit
4
Offline use of the SPD
4%
7%
• Some effect in p0 mass
• Effect on B mass in B0K*g?
5
Offline use of the PS
• Identification of e-m clusters
– Current g-id uses absence of tracks only
• There is a version including PS info but not used
– e±-id includes E in PS in the likelihood, but minor
contrib. wrt track position and energy c2
• Deterioration without PS?
• PS role more important with higher track occupancy?
• Energy resolution: would only (but slightly)
improve without PS
6
SPD/PS use at L0
• PS: confirm ECAL clusters
as e-m (ECAL: 1.1 lI)
• SPD:
– distinguish e – g
– multiplicity:
– GECs (very uncertain future)
– Under study: trigger for
ggmm for lumi: 2 ms + low
SPD multiplicity
Recent global L0 optimization on
trigger benchmark channels, incl.
Bsfg , B+K+e+e-:
LO line
E T>
(MeV)
Rate
(MHz)
e
2320
170
g
p0 loc
p0 glob
3060
65
4740
4520
95
85
hadr
4040
600
by M. Gandelman, numbers by Antonio
7
Difficulty for SPD @ LO: g  e+e• Re-opt TDR material budget: 0.4 X0 magnet-Calo
– Underestimation: only M1 is 0.265 instead of the 0.22
in the simulation
7
• Probability of photon survival: e
 0.45
9
 70%
• Btw, this is what makes the SPD useful offline
• Current g and e confirmation at HLT1 uses all em clusters for both cases
• But the SPD still “doing something”, at least while
ETgETe
8
… g  e+e• Another source of “inefficiency” on g’s:
random coincidence
– Multiplicity  33 @ 21032   330 @ 21033
– Occup 330/6000  5.2% (max 16%)
• Ways out of g  e+e-:
– Still worth for the remaining 70%?
• Seems to be the case now
– (non-binary output…)
– If M1 has to be removed for other reasons…
9
Current HLT1: concept
• Use calo clusters that passed L0 as seeds for
confirmation with T stations, VELO
– N candidates:
Hadr
e
g
1.2
6.5
4.1
 GECs not useful anymore
• Refinement of L0 candidates at HLT
• J. Albrecht tried decoding relevant calo region and doing Scorrection
– No gain for HCAL if wanted high efficiency, some gain for ECAL
–
LHCb-118-2007 and http://indico.cern.ch/getFile.py/access?contribId=35&resId=0&materialId=slides&confId=33862
10
Current HLT1: ECAL alleys
• Photon alley:
M . Witek tomorrow or at http://indico.cern.ch/conferenceDisplay.py?confId=39471
– 85 % of candidates e-m clusters due to π0
• Decode relevant area to cut on shower shape: keep
90% of signal while reducing rate to 50%
• Uses fast version of Caloreconstruction by V. Egorychev
http://indico.cern.ch/conferenceDisplay.py?confId=31908
• Current L0 clustering ok for upgrade?
• Is this what is planed?
11
Trigger and upgrade
• Lumi 2·1033
• All detectors read-out at 40 MHz, no L0
– But keep possibility of throttling
• Still, trigger needs to use:
– Seeds à la L0 (calo, muons)
• Can they be really current L0 clusters?
– Reconstruction based in regions of interest
See Hans’ presentation at http://indico.cern.ch/conferenceDisplay.py?confId=43555
• Bsfg , B0K*e+e- some of the star channels
12
Possible SPD uses in the upgrade
• Set different ET thesholds for g and e
– Offline cuts:
ETg>
Ref
Bs fg
2.8 GeV
LHCb-PHYS-2007-030
Bs J/y(e+e-)f
0.75 GeV
http://indico.cern.ch/conferenceDisplay.py?confId=50764
B0 K*e+e-
0.3 GeV
A. Hicheur
Marie-Helene Schune
http://indico.cern.ch/contributionDisplay.py?contribId=2&confId=26414
• Useful to prepare trigger menues?
– 2 hadr + g, 1 hadr + e+ + e13
Plan
• Run current L0 on minimum bias and signal
events at 2·1033
• Study rate, number of e/g clusters vs ETcut with
SPD/PS/M1 there/not there
– Study some “trigger menues”?
• Difficulty to compare scenarios: what are the
limitations?
– Rate?
– Number of candidates?
14
MC
• At last calo upgrade meeting, asked for production of
MC with/without M1
– Samples:
• Minimum bias (rate)
• Bsfg (efficiency)
– Should add B0K*e+e– No idea of status
• Ask also for the 3 samples with/without SPD/PS
– Sample can also be studied for offline PID, sE… who?
15