Transcript information

Experiment NA60
Bose/Einstein Correlations
Advisors:
->Atila Neves: Lisbon, Portugal
->Sanja Damjanovic: University of Heidelberg
->Hans J. Specht: University of Heidelberg
->Hiroaki Ohnishi: Riken, Japan
11-08-2005
David J Hover
Quark Gluon Plasma (QGP)
•
•
•
QGP describes a new state of matter where quarks and gluons no longer
exist in bound states.
QGP exists as a function of temperature and density.
Want to map out details of phase transition:
– Freeze out temperature? Fireball radius?
•
Significance evidence shown in past SPS and offsite experiments but
details still need to be confirmed..
What is the
Threshold???
11-08-2005
David J Hover
Bose-Einstein Correlations (HBT)
• “…intensity interferometry uses correlations between like particles
induced by Fermi or Bose statistics to determine the space and time
dimensions of the particle source.”
– This analysis used to measure the hadronic interaction regions
• Essentially a continuous double slit experiment using boson intensity
instead of amplitude.
Source
Detectors
• Pions are created in the same event. There are two ways to obtain
the same information at the two detectors: X1 and X2. Since the two
particles are indistinguishable, we must add together the two path
amplitudes in order to calculate the probability.
11-08-2005
David J Hover
Parameterize the Correlation
• The correlation is observable in terms of relative
momentum
{ nP1 n p2 }
S
C (q ) 
  F (q )
{ n p1  n p2 } B
– S=Source, or mixed particles in same event (correlated).
– B=Background, or mixed particles of different events
(uncorrelated)
– F(q) = Functional form described by theory
• Different extracted variables (the Q coordinates):
– Qs,Ql,Qo and Qinv
11-08-2005
David J Hover
Correlated Data!!!
• 50K dimuon triggered events binned by centrality, pair charge and
target (made 2520 plots to analyze):
BACKGROUND
SOURCE
More pairs the better:
decreases error but with
increase in computing time.
Or
Correlated
Need more statistics at
low relative momentum
11-08-2005
David J Hover
Correlated Data
C=S/B: MM is
Correlated?
11-08-2005
David J Hover
Corrections (1)
• Momentum Resolution:
– Qinv( MCgen – MCreco)
– Fit with Gaussian
– Extract sigma = momentum resolution = 0.023 GeV
• Not Complete:
– Need more MC events (more than 50K) binned in same way as data.
– Once know, rerun Data with correct bin size.
11-08-2005
David J Hover
Corrections (2)
• Coulomb Corrections:
– Correlation fit assumes particles do not correlate after
production, but they do!!
– Coulomb interaction reduces correlations at small relative
momentums.
• Solve by multiplying with Gamow factor:
– Only source/same event
– Both data and MCreco
• Not entirely correct:
– Different result at small relative momentums (overlap)
– Need to check by adding into MCgen
11-08-2005
David J Hover
Fit and Extract
C2 (Qinv)  1  e
2
2
 Rinv
Qinv
•λ=correlation factor, Rinv = Invariant Radius
•I tried to minimize (Chi-Square) / ndof through grid and iterate method.
•Found Rinv = 9.75 fm; λ= 0.14; Chi-Square / ndof = 3.01
•Unacceptable!!
11-08-2005
David J Hover
So Much More to Do
•
•
•
•
•
Fix the Gamow factor.
More events.
Generate correlation and compare.
Include momentum resolution in data.
Other variables.
11-08-2005
David J Hover
Many Thanks to…
•
•
•
•
•
•
•
•
•
•
NSF
Herr Jeremey
Jean Krisch
Ford
Homer Neal
My advisors
Europe
Google
French doctors
Trains that do not go through France
11-08-2005
David J Hover