1. The ATLAS B physics trigger
Natalia Panikashvili
Technion Institute of Technology, Israel
University of Michigan, USA
On behalf of the
B – physics trigger group
Beauty 2005, 10th International Conference on B-Physics at Hadron
Machines, Assisi (Perugia), Italy. June

2. LHC and ATLAS detector tracking: forming the Inner Detector (ID)
Pixel
Semiconductor Tracker (SCT)
Transition Radiation Tracker (TRT)
forming the Inner Detector (ID)
muons:
Monitored Drift Tubes (MDT)
Cathode Strip Chambers (CSC)
Resistive Plate Chambers (RPC)
Thin Gap Chambers (TGC)
calorimeters: Electromagnetic Liquid Argon and Hadronic Tile detectors
Center of mass energy = 14TeV
Bunch – crossing rate = 40MHz
Luminosity:
L = ~1033 cm-2 s-1 - most of B-physics will be measured here
L = ~1034 cm-2 s-1 -
for High pT discovery
physics, however will be
used also for rare B decays
p–pbar collision = 109Hz
Bunch crossing interval = 25ns
Pileup = 23 ( L = ~1034 cm-2 s-1 )
16 April 2017

3. Events rate too high to collect all events
99% of bunch crossing end up with non-b events
The selection of physics signals requires the identification of objects that can be isolated from the high particle density environment.
Event rate 
Level-2 
Level-1 
Offline Analyses
Object
Examples of physics coverage e
Higgs, B-physics, new gauge bosons, extra dimensions, SUSY, W, top g
Higgs, extra dimensions, SUSY,
B-physics m
B-physics, Higgs, new gauge bosons, extra dimensions, SUSY, W, top
Jets
SUSY, compositeness, resonances,
16 April 2017

4. The Trigger system and Region of Interest mechanism
Reduces the high data rate by selecting interesting events through 3 steps:
H →2e + 2 2 2e
LVL1 decision made:
Muon Trigger Chambers and Calorimeter data to find e, g, t, jet, m candidates above thresholds
Identifies Regions of Interest
Processing time 2.5 ms
LVL2 uses Region of Interest data
Combines information from all detectors
Performs fast rejection.
Processing time 10 ms
Output rate ~2kHz
Event Filter
Can be “seeded” by LVL2 result
potential full event access
Processing time 1s
Output rate ~200Hz
hardware
software
16 April 2017

5. Level1: Muons Trigger Chambers: Barrel region (| h |< 1.1) : η
RPC η 
Trigger Chambers:
Barrel region (| h |< 1.1) :
Resistive Plate Chambers (RPC)
End-cap region ( 1<| h |< 2.7) :
Thin Gap Chambers (TGC)
TGC
Toroid
Muon chambers layout and curved muon tracks
To indicate a m candidate, a hit must be accompanied by hits in the other detector layers, within the coincidence window.
Low pT m - 3/4
High pT m - 3/4 + 1/2 (2/3) for barrel (end-cap)
Trigger efficiency
85% low pT ; 87% high pT
16 April 2017

6. pT of muons from different processes
Level1: Muon trigger
To reduce background from decay in flight of p/K  pT typical threshold in Level1 scheme for low luminosity (L ~1033 cm-2s-1) is 6GeV
pT of muons from different processes
Trigger rate (kHz) expected in Muon Spectrometer from LVL1 TDR. The low pT assume a luminosity of 1033cm-2s-1:
Process
Barrel
End-Cap
Combined System
Low pT
6GeV
p/K
7.0
9.8
16.8 b
1.9
2.1
4.0 c
1.1
1.3
2.4 W
0.004
0.005
0.009
Total
10.0
13.2
23.2
Only 4 kHz of these are b events!
16 April 2017

7. Example of e/g trigger algorithm:
Level-1: Calorimeter
Example of e/g trigger algorithm:
Hadronic Tile Calorimeter
EM Calorimeter
Hadronic LAr End-Cap Calorimeter
Calorimeter Trigger looking for
e/g + Jets + t objects
Using trigger towers of Hadronic
and Electromagnetic calorimeters
The requirement for a trigger object:
The RoI cluster – local maximum
The most energetic cluster > ET
Total ET in EM isolation < EM Isolation Threshold
Total ET in Hadron < Hadronic isolation threshold
16 April 2017

8. LVL1 EM/Jet RoI multiplicity vs. Energy threshold, b  m X events
Level1: Calorimeter
LVL1 EM/Jet RoI multiplicity vs. Energy threshold, b  m X events
16 April 2017

9. ATLAS B-Physics Programme
CP Violation
Control channel:
sin(2b)
DGs = GH - GL; Gs; A||; A┴; the strong phase difference d2 - d1; the weak phase fs
Measurement of Bs oscillations:
D ms= mH – mL
Rare decays
Precise measurements of the branching ration
Lb polarization measurements
Asymmetry parameter ab, Pb, life time measurements
Bc mesons
Precise determination of Bc mass, Bc life time
16 April 2017
See talk of C. Padilla “Overview of ATLAS performance for B-physics”

10. B-physics trigger strategy
How to be efficient for interesting B-physics events?
Different LVL1 trigger menus will be used at different luminosities:
Lower luminosity - L ~ 1033 cm-2s-1 LVL1 1m ( pT > 6 – 8 GeV )
Higher luminosity - L ~ 1034 cm-2s-1 LVL1 2m ( pT > 6 GeV ; lower threshold? )
single-muon
di-muon
all h b c
J/y
@1033cm-2s-1
Cross section, (nb)
LVL2:
LVL1 confirmation
m tracking
Di – m trigger
J/y, Bs,d trigger
Event triggered by LVL1 m, the
additional information from e/g + Jet Region of Interest can be used
EM RoI
Jet RoI
16 April 2017

11. Level2: m Tracking mFast - Standalone m reconstruction
Propose: Confirm the LVL1 trigger with a more precise pT estimation within a RoI
“Global Pattern Recognition” involving Trigger Chambers and Precision Chambers
“Track fit” involving drift time measurements, performed for each MDT chamber
Fast “pT estimation” via a look-up-table (LUT) with no use of time consuming fit methods
LVL2 muon standalone
Efficiency
Output rates after LVL2 standalone m
reconstruction still dominated by /K decays
LVL2 muon + ID
Efficiency
mComb - Combined m reconstruction
(Using reconstructed m and ID information)
Propose: Rejection on the p/Km
16 April 2017

12. Single m Rate estimation for barrel region (kHz)
Level2: m Tracking
IDSCAN- track reconstruction in Inner Detector
Input - Space Points (SP) found in Pixel and SCT Detectors
Output – track and SP associated with them.
Single m Rate estimation for barrel region (kHz)
K/p
decays b c
Fake
LVL1
Total
7.60
1.5
0.90
1.0
11.0
mFast
3.18
0.91
0.41
Negligible
4.5
mComb
1.1
0.68
0.35
2.13
16 April 2017

13. Level2: Di m , J/ym+m-, rare decay selection
We would like to be efficient for “gold” channels - J/y !
LVL1: 1m RoI pT (m) > 6GeV
Efficiency for both m+m- from J/y to be found in enlarged m RoI
LVL2:
Confirm m RoI using:
m stand alone reconstruction (mFast)
combine m with Inner Detector track (mComb)
Open the region (Dh x Df) around m in order to find J/y or Bs,d
Find ID tracks in selected region (IdScan )
M (m+m-) > 2.8 GeV
Extrapolate track to MS
Associate track with MS hits
Create Di - m or J/y Bs,d
EF: Refit ID tracks in Level2 RoI Vertexing
16 April 2017

14. Create the pair of tracks with opposite charge
J/y selection overview
MDT
RPC/TGC
LVL1 pT(m) > 6GeV
m RoI ( φ, η ) μ μ
Create the pair of tracks with opposite charge
16 April 2017

15. Level2: J/y  e+e- selection using EM RoI
LVL1: 1m RoI pT (m) > 6GeV
≥1 EM RoI ET > 2GeV
Bd J/y (e+e-) Ks(p+p-)
LVL2:
Confirm m RoI
Confirm e RoI using:
Calorimeter (T2Calo)
If only one e found at LVL1:
Open larger region for 2nd e
Find ID tracks in selected region
Mass Cuts
Associate the track with EM Calorimeter information
If both electrons found at LVL1:
confirmation at LVL2 inside small region about each e e+ e-
EM RoI
TRT
EF: Refit ID tracks in Level-2 RoI
Vertex reconstruction
Transverse Decay length cut
EM Calorimeter
16 April 2017

16. Level-2: J/y selection
Full Width of RoI in h and f
Efficiency for both e+e- from J/y to be found in enlarged EM RoI
16 April 2017

17. Level2: J/y efficiency and background rate
Mass plot of two tracks identified as e+e- inside J/y mass window
J/y  m+ m- efficiency vs. background rate
J/y  m+ m-
J/y  e+e-
Efficiency
68 – 77%
1st m pT > 6 GeV 2nd m pT > 3 GeV
68%
both e pT > 5 GeV
Output rate (Hz)
bbm(6)X
260 – 380*
170
* output background rate, does not include rate of J/y from non-b events
16 April 2017

18. Level2: Hadronic final state selection using Jet RoI
How to be efficient for Bs  Ds ( f ( K+K- ) p ) p ?
LVL1: 1m RoI pT (m) > 6GeV
≥1 Jet RoI ET > 5GeV
~ 2 Jet RoI per event
Efficiency to find Bs decay products in
the region around Jet RoI
Df = Dh = 1.
Df = Dh = 0.75
Df = Dh = 0.5
LVL2:
Confirm m RoI
Confirm Jet RoI
Open the region (Dh x Df =1.5 × 1.5) around jet in order to find Bs decay products
Find ID tracks in selected region (IdScan )
M (K+K-)
EF: Refit ID tracks in Level-2 RoI
Vertex reconstruction
Transverse Decay length cut
16 April 2017

19. Level2: f and Ds selection
Invariant mass of particle identified as Dsfp
signal
Invariant mass of particle identified as f  K K
Using: 1008MeV < M ( f → K K ) < 1033 MeV
1880MeV < M ( Ds → f p ) < 2024 MeV
Signal Efficiency: Including (excluding) LVL1 Jet RoI efficiency
Rate
50% (58%)
310 Hz
Study was done for events with
pT ( Bs ) >10 GeV
16 April 2017

20. Level2: rare radiative B-decays
LVL1: 1m RoI pT (m) > 6GeV
≥1 EM RoI ET > 5GeV
Rate: 1075Hz
Selection of:
Bd  K*0 g
and
Bs  f g
LVL2:
Confirm m RoI
Confirm g RoI using:
Calorimeter
Open the region around g
Find ID tracks in selected region
Reconstruct K*0K+p- or f K+ K-
Opening angle cuts, Impact parameter cut
Rate: 23 Hz
For 30fb-1 we have:
15000 Bd  K*0 g
4800 Bs  f g
EF: Refit ID tracks in Level-2 RoI
Vertex reconstruction
Mass cuts
Rate: 0.6 Hz Bd  K*0 g
0.5 Hz Bs  f g
16 April 2017

21. EF: Rates of Rare decays in higher luminosity
Di-m Mass, (MeV)
B K* m+ m-
Bm+ m-
Not normalized
LVL1: 2m RoI pT (m) > 6GeV
LVL2:
Confirm each m RoI
m stand alone reconstruction (mFast)
combine m with Inner Detector track (mComb)
Mass cut
EF: Refit ID tracks in Level-2 RoI
Decay vertex reconstruction
Transverse Decay length cut:
Lxy > 500mm
Angular Distribution cut
Efficiency estimation:
70% of B m+m-
60% of B K* m+ m-
Output rate < 10 Hz
bbm+m- both m pT>6 GeV
16 April 2017

22. Estimates of Overall Efficiencies
Make Rough Estimates of expected efficiencies for some example channels
A lot of unknowns: e.g. efficiency for LVL2 combined m + ID & m efficiency in end-cap
LVL1+LVL2
Event
Filter
Overall Efficiency
B  m+ m-
~ 75%
95%
~70%
Events with m+ and m- pT > 6 GeV
B m+ m- K*
85%
~65%
Bd J/y (m+m-) Ks
First m: ~ 85%
J/y ( m+ m- ): 77%
Overall: ~ 65%
~100%
Events with one m pT>6 GeV and other m pT > 3 GeV
Bd  J/y (e+e-) Ks
Muon: ~ 85%
J/y ( e+ e- ): 72%
Overall: ~ 60%
~60% Ev. with e+ and e- pT > 5 GeV
~40% 1st e pT>5 GeV, 2nd e pT>2 GeV
Bs  Ds p
Ds  f (K+K-) p
Ds (f p): 50%
Overall: ~ 40%
~40%
Events with Bs pT>10GeV
and K, K, p pT > 1.5 GeV
16 April 2017

23. Conclusions
Studies of the benefits of a RoI based B-trigger are still being investigated
Have demonstrated a flexible strategy for B-physics studies from initial running to final luminosity
We are looking towards first collisions in 2007 when we hope to record B-physics data from ATLAS.
16 April 2017