Benchmark channels for CGEM performance studies Cristina Morales - - PowerPoint PPT Presentation

Benchmark channels for CGEM performance studies

Cristina Morales for the CGEM software group 13.03.2017 1/31

2.396 and 2.9 GeV

2/31

Close to ppbar-threshold

Cristina Morales

• We have studied the channel at three CMS energies

Ecms= 2.0, 1.975 and 1.95 GeV. At lower energies antiprotons annihilate before reaching tracker

• This channel allows to compare tracker perfomances at low

proton/antiproton momenta

• No background remaining after selection (permill level)
• Phokhara v9.1 used to generate e+e- → pp with NNLO ISR under boss665p01

3/31

To study the track's vertex resolution, the cuts corresponding to point of closest distance of approach are not applied. The procedure is done for proton and for anti-proton and for ECM= 2.0, 1.975 and 1.95 GeV

POAxy,z resolution of the tracks

The xy and z vertex resolution get better with increasing momenta The xy vertex resolution is better than the z vertex resolution The vertex resolution for proton and anti-proton is comparable Red: proton Black: aproton Red: proton Black: aproton 5%

Ecm = 2.0 GeV proton POAxy proton (cm)

POAxy resolution (cm) POAz resolution (cm) Track momentum (GeV/c) Track momentum (GeV/c) Can CGEM achieve better vertex resolutions at low momenta 150% 50% 4/31

P and Pt resolution of the tracks

To study the track's momentum resolution, the cuts corresponding to momentum window are not applied. We look at the transverse momentum and at the longitudinal momentum and we compare with the MC truth, Ecms= 2.0, 1.975 and 1.95 GeV

proton 2.0 GeV Proton p reco – proton p true (GeV/c)

red: proton black: aproton red: proton black: aproton 2% 30% P resolution (MeV/c) Pt resolution (MeV/c) The p and pt resolution get better with increasing momentum. The resolution is almost the same for proton and antiproton The pt resolution gets worse by 30% as we go to lower track momenta. Can CGEM do better? 5/31

De/dx resolution

To measure the resolution at each energy, all cuts are applied except for de/dx;

proton ECMS= 2.0 GeV Normalized de/dx pulse height positive track

The resolution in dedx is MUCH worse as the track momentum decreases: multiple scattering? There are other PID possibilities, still it would be interesting to see dE/dx for CGEM 7% Red: proton Black: aproton Normalized DE/DX pulse height resolution (MeV/c) 114% 6/31

Better resolutions in the track variables could help to enhance the efficiency also at low momenta. (GeV/c) Tracking effs Track momentum (GeV/c)

red: proton black: aproton

Slight decrease in tracking effciency at lower

• momenta. Antiproton trackin eff worse than for
• proton. Can CGEM bring it up?

Proton tracking eff costheta_p

Efficiencies flat except for at low/high theta_p. Can CGEM recover high/low polar angles?

Tracking efficiencies and total selection efficiency

Selection efficiency as a function

• f the track momentum

7/31

Martin Ripka

at ECM = 3.773 GeV

• Since pions and muons behavekinematically and for the detector in the same way,

both ee--> pipi and ee--> mumu are treated as signal

• This channel allows to compare tracker perfomances at low

proton/antiproton momenta

• Expected remaining background coming from Bhabha 1%
• Phokhara v9.1 used to generate e+e- → mu+mu- with NNLO ISR under boss665p01

8/31

(only events within geometrical acceptance) Efficiency = number of selected tagged events/ generated events (taged and untagged) 9/31

10/31

We will provide the tracking effs vs pt 11/31

Andreas Pitka

at ECM = 4.42 GeV

12/31

After selection (bacup slides) remaining background below 1% level!!. Composition:

• pen charm: D*0 D0*bar, D*+ D*-, pi0 D0 D0*bar

13/31

17/30 ?

It has been checked that the low energetic MC tracks do not come from pion decays

Event navigator used to match true particles with the recons. It is checked whether PID and mother particle are correct Not full selection applied

14/31

17/30 Fundamental to include it in CGEMBoss for this channel!! 15/31

17/30 16/31

17/30

true true true

17/31

Stuart Fegan

at ECM = 3.097 GeV

22/30 18/31

23/30 Selection efficiency: ~45%; Background contamination: below 1%

Selection efficiency: ~45%; Background contamination: below 1%

19/31

20/31

21/31

Peter Weidenkaff

at ECM = 3.773 GeV

22/31

23/31

K- tracking efficiency 24/31

25/31

Flight distance obtained after secondary vertex fit

26/31

e+e- -> Lambda Lambdabar

27/31 Cui Li

28/31

29/31 27%

56% 30/31

distribution

• f Rxy

31/31 Other variables studied: POAxy and POAz of all tracks at 2.396 and 2.9 GeV (see backup slides)

Summary & Outlook

1) We have done most of the studies we wanted to perform for MDC as IT. 2) We are working on Bhabha (+ISR) 3) We will compare the obtained results with the data quality group for the mumu, bhabha and J/Psi to 3pi channels 4) We will prepare a short description of the analysis and the obtention of the results shown here. 5) Soon our J/Psi analyser will leave Mainz. We need to decide whether or not to pursue his studies. 6) Waiting impatiently for CGEMBoss to start tests.

To do

backup

The selection is based on the HIM's selection for low energy scan data collected in 2015 No background remaining after selection (permill level) 2.0 GeV 1.975 GeV 1.95 GeV Two good charged tracks of opposite charge (Rxy, Rz and good MDC tracks required)

Selection criteria

If tracks leave a TOF signal then the difference between then < 4ns Vertex fit to improve resolution of the track's momenta All cut limits addapted to the ECMS Back to back signature between tracks after boost to CMS: 170°, 160°, 160° Momentum window cut: pexpected -4σmeasured < pp,pbar < pexpected + 3σmeasured Particle identification using directly deposited energy in MDC: dedx

Cristina Morales, 14.02.2017

To study the track's vertex resolution, the cuts corresponding to point of closest distance of approach are not applied. The procedure is done for proton and for anti-proton and for ECM= 2.0, 1.975 and 1.95 GeV

POAxy,z resolution of the tracks

Ecm = 2.0 GeV proton Ecm = 2.0 GeV proton The xy and z vertex resolution get better with increasing momenta The xy vertex resolution is better than the z vertex resolution The vertex resolution for proton and anti-proton is comparable Red: proton Black: aproton Red: proton Black: aproton 5% POAxy proton (cm) POAz proton (cm) POAxy resolution (cm) POAz resolution (cm) Track momentum (GeV/c) Track momentum (GeV/c) Can CGEM achieve better vertex resolutions at low momenta 150% 50%

P and Pt resolution of the tracks

To study the track's momentum resolution, the cuts corresponding to momentum window are not applied. We look at the transverse momentum and at the longitudinal momentum and we compare with the MC truth. Ecms= 2.0, 1.975 and 1.95 GeV proton 2.0 GeV 2.0 GeV proton red: proton black: aproton red: proton black: aproton The p and pt resolution get better with increasing momentum. It is almost the same for proton and antiproton The pt resolution gets worse by 30% as we go to lower track momenta. Can CGEM do better? 2% Proton pt reco – proton pt true (GeV/c) Proton p reco – proton p true (GeV/c) P resolution (MeV/c) Pt resolution (MeV/c) 30%

De/dx resolution

proton To measure the resolution at each energy,all cuts are applied except for de/dx; ECMS= 2.0 GeV

Normalized de/dx pulse height positive track Normalized de/dx pulse height negative track

aproton ECMS= 2.0 GeV 7% The resolution in dedx is MUCH worse as the track momentum decreases: multiple scattering? There are other PID possibilities, still it would be interesting to see if CGEM can do better Red: proton Black: aproton Normalized DE/DX pulse height resolution (MeV/c) 114%

(GeV/c)

0.8%

Total selection efficiency and geometrical acceptance

Selection efficiency as a function

• f the track momentum

Geometrical acceptance Better resolutions in the track variables could help to enhance the efficiency also at low momenta. (GeV/c) The angular distribution of the events is almost inependent of the geometrical acceptance (form factors model almost constant for this momentum range)

Proton and anti-proton tracking efficiencies

ed to Eff_proton = number of events with 2 good tracks / number of events with 1 (anti-proton) or 2 tracks Since only one track can be identified, a new, harder signal selection is used: Tracking effs Track momentum (GeV/c) red: proton black: aproton Proton tracking efficiency costheta_p Efficiencies flat except for at low/high theta_p. Black: 2GeV; Red: 1.975 GeV; Blue: 1.95 GeV; Can CGEM recover high/low polar angles? Slight decrease in tracking effciency at lower

• momenta. Can CGEM bring it up?

Expected remaining background coming from Bhabha 1%. ee--> pi+pi- treated as signal (similar kin behavior as ee--> mumu)

Remaining background below 1% level!!

17/30

17/30

23/30

24/30

30/30