Centrality, Npart & Ncollision at BRAHMS

H. ItoUniversity of Kansas

For the BRAHMS Collaboration

Why do we want Npart?

• Comparison of P-P, P-A, and A-A?• Comparison of Different Centrality?

Can we live without Npart?•For comparison of P-P,P-A & A-A, we can compare the most central collisions.

What is the concern about using Npart?•No Direct Measurement•Is there any indirect measurement (ZDC,etc)?•If not measured, are there any model dependencies?

What Is Measured?

Ncharge in -4.7< <4.7

We can measure charged particle multiplicities.

What we have done with Npart up to now?

Using experimental centrality and comparing with the model calculation, we got Npart for our event selections.

How we calculated Npart.Measure charged

particle multiplicitiesby Silicon and Tile • Correct for missing fraction of

the cross section and background events with small multiplicity by Tile only.

• Correct anomalous background events by Silicon and Tile.

Assuming centrality and multiplicity have a linear

correlation, select events by slicing the corrected multiplicity

distribution vertically.

By knowing the centrality, use Monte Carlo simulation to get

corresponding Npart.

Event Selection for Minimum Bias

• ZDC is very close to100 % efficient.– Not efficient at the most central events.

• Beam Beam (BB) is only about 75 % efficient.– Not efficient at the peripheral events.

Although the correlated events by BB and ZDC can be used to select only real collision events, it is not used for minimum bias selection.

TPC vertex BB vertex ZDC vertex

Multiplicity and Centrality Cut

Event selection is done by slicing multiplicity distribution vertically.

5%50%

Efficiency for Minimum Bias Events

Efficiency of the Tile getting hit in minimum bias event is about 99 % in the Monte Carlo simulation.

How to get the fraction of missing cross section?

Measured multiplicity distribution by Tile

Throw out

Using Hijing with GEANT, estimate how many events we should have for very small multiplicity events.

%101.200...0

4...0

M

M

Event

Event

Centrality Selection in Monte CarloMultiplicity measured by

Silicon and Tile is use to

select centrality.

Wounded Nucleon Calculation

NN= 40mb, a = .54 fm

40-50

30-40

20-30

10-20

5-10

0-5

Npart (Hijing)

Ncollision (WN)

Npart(WN)Centrallity

346

293

228

164

114

76

352

299

235

165

114

75

873

708

516

333

206

118

Measured Neutron by ZDC

50 % CentralVery Few Neutrons ???

Correlation of Tile and Silicon multiplicity

Correlation of BB and Silicon-Tile multiplicity

Npart, Ncollision and Impact Parameter

NN= 40mb, a = .54 fm

Ncollision

Npart