%CERTIFY%

HeavyIonMultiplictyAnalysis

General

The goal it to measure Heavy Ion Charged Particle Multiplicity in the most central Pb+Pb collisions at 2.75TeV incident nucleon energy. This page is a group exchange board for the latest developments in the analysis of charged particle multiplicity. Please add you graphs, analysis details and comments to this page.

• Co-Authors: Brian Cole (Columbia U.), Sasha Milov (WIS)

• People involved: Yujiao Chen (Columbia U.), Soumya Mohapatra (Stony Brook U.), Peter Steinberg (BNL), please add yourself.

Analysis details

Pixel detector

The analysis is based exclusively on the ATLAS Pixel Detector. There are two main reasons for that:

• lowest momentum cut defined which allows measuring the bulk part of the particle production. the cut is defined by:
  • stopping of the particles, (mainly protons) in the material (<100 MeV) in the absence of the magnetic field,
  • particle not reaching the pixel layers in the presence of the magnetic filed resulting in ~100 MeV cut.

• low detector occupancy not exceeding 0.5% even in the most pessimistic MC simulation of the most central Heavy Ion events.

Methods, summary

At the moment we exploit several method.

Method	Full magnetic field	No magnetic field
2-point tracklets with ordering	eff>90%, fakes	eff > 95%, less fakes
2-point tracklets w/bkg subtraction	eff >90%, fakes subtracted	eff > 95%, fakes subtracted
Pixel tracks	eff >50%, fakes = ?	eff > ?, fakes = ?
Pixel counting	cluster/track ratio extrapolation from p+p, reliability?
Full tracks	cross-check > 0.5 GeV/c, may come later

Methods, details

• 2-point tracklets with ordering is the most advanced method recovering more that 90% of all tracks. It relies on the hit in the B-layer which needs to be confirmed by the best match in layer 1 or 2. Effectively it uses a very narrow corridor in rapidity and azimuth to find a match and therefore is elimination most of random coincidences. However some random combinations may survive. In the absence of the magnetic filed such probability is very low. This method was used for the analogues PHOBOS paper.

• 2-point tracklets with background subtraction As effective as the first method, but takes all possible combinations satisfying selection criteria (pointing back to the event vertex) resulting in an inherit high level random combinatorial candidates. Those candidates can be removed by mixing, e.g.: flipping the B-layer of the detector. In field off configuration should have higher efficiency and less fakes. This method was used for the analogues PHENIX paper. The two first methods are complimentary to each other.

• Pixel tracks inherits form the robustness of ATLAS tracking, but shows significantly lower efficiency and still has some fakes. It performance in B-off runs is supposed to be much better, and it is currently investigated. Comparing this method with the first two would be a powerful cross-check

• Pixel counting relies on the assumption that the total number of clusters and total number of hits in p+p and Pb+Pb runs should not change. This assumption is a bit of stretch, but is a good item to control.

• Full tracks must yield the same result in full be run with appropriate cut. must be checked.

Action items

Correction and systematic uncertainties:

Results:

---

Major updates:
-- AlexanderMilov - 22-Oct-2010

%RESPONSIBLE% AlexanderMilov
%REVIEW% Never reviewed