Low pT F reconstruction in PHOBOS 2004

1
Low pT F reconstruction in PHOBOS 2004

• Based on
• Analysis of data from simulation of 40mln low pT
  (lt0.5GeV) F
• Corey presentation
• Papers
• Star - F production in AuAu at 130 GeV, Phys.
  Rev. C65 (2002)
• Star - F production in AuAu at 200 GeV,
  nucl-ex/030601
• Phenix - F production in AuAu at 200 GeV,
  nucl-ex/0209028
• Presentations of preliminary results by Phenix
  and Star

Andrzej Olszewski
2
Statistics Needed

• Rate measurements
• inclusive gt 100 Fs
• in bins of centrality gt 50 Fs in most central
  bin
• ? means gt 1000 Fs in inclusive sample
• in bins of transverse momentum gt 100 Fs in
  lowest and the highest pT bins ? means also
  1000s of Fs needed
• Shape measurements
• Both mass peak position and width (shape) require
  more than 5000 Fs for a single point

3
AuAu 200 GeV Fs / event

• Phenix 200 GeV inclusive dN/d?(F) 2.01
• STAR 130 GeV 10 central dN/d?(F) 5.7
• STAR 200 GeV 10 central dN/d?(F) 6 6.5
• F rates in 200 GeV AuAu, over full rapidity
• 45 per 10 central event
• 38 per 25 central event
• 14 per inclusive event

4
Low pT Fs / event

• STAR 130 GeV mT slope
• 0.378 MeV (10 central)
• and statement of increasing slope with energy
• STAR 200 GeV mT slope 0.340 MeV on average
• Phenix 200 GeV using 0.320 MeV in
  simulations
• Whats the difference?
• pT lt 0.5 GeV 20 - 24
• pT lt 0.15 GeV 2 - 2.5

5
Low pT F - geometrical acceptance

• Properties of simulated Fs
• Exponential mT with T 378 MeV, pT (0, 0.5)
• Corey 20106 experimental VZ -20, 20
• Andrzej 40106 flat VZ -40, 0
• Geometrical acceptance depends on the Track
  Quality criteria(for pairs of Kaon tracks from
  F?KK- decays)
• Trackable1 gt 6 spectrometer layers hit
  1872 / 40 106
• Trackable2 gt all curved track layers hit
  1150 / 40 106
• Trackable3 all possible layers with hits
  435 / 40 106
• AccHough matching Hough tracking
  1 / 40 106

6
Estimation of usefull statistics - events

• Written
• 120 mln events on tape
• Triggering on collision efficiency
• 60 collision events
• Triggering on vertex
• 80 in usefull vertex range (-40,10)
• Vertexing efficiency
• 80 RMSSelVertex
• This makes 46 mln usefull events and 644 mln Fs
  total

7
Estimation of usefull statistics - Fs

• After geometrical acceptance cuts we are left
  with
• pT lt 0.5 GeV
• 6000 Fs decaying to trackable1 Kaons
• 1400 Fs decaying to trackable3 Kaons
• pT lt 0.15 GeV
• 600 Fs decaying to trackable1 Kaons
• 150 Fs decaying to trackable3 Kaons
• If we can reconstruct and identify trackable3
  Kaons efficiently we are already close to measure
  production rates of even the lowest pT Fs

8
Kaon identification efficiency
Even with Kaon tracks found F may be not
reconstructed if the particle is not identified.
Fortunately for these low pT Fs most Kaon decay
products have momenta lower then 0.7 GeV
9
Tracking low pT Fs

• Hough tracking
• after strong AccHough geometrical cut
  efficiencies of reconstruction higher pT
  (gt0.5GeV) Fs close to 80.
• lower pT Fs not reconstructed so far
• Banana tracking
• just started looking at this tracking
• found it as easy to use as Hough tracking, and
  fast!
• not ready to reconstruct Fs, missing templates
• first results indicate possible capability of
  reconstruction of low pT Fs, efficiency unknown.

10
First results on Banana efficiency
Efficiency of tracking on all findable tracks in
MC event vertex Z (-1,0), single arm SpecN
11
First results on Banana efficiency
This is best performance. Vertex Z (-1,0),
single arm SpecN I have seen lower efficiency at
large negative vertex Z!
12
How much more statistics do we need?

• Uncertainty in tracking efficiency prevents from
  giving a good estimate of how much more events we
  need.
• To be safe we would need to increase statistics
  by a factor of 10 or more. With good tracking
  efficiency this would make possible more precise
  measurements and with worse tracking this would
  enable us at least to estimate rates at low pT
  end.
• With the increase by a factor of 2-3 the rate
  measurement is still kept only possible.

13
How could we increase statistics?

• By using centrality trigger
• but in this way we loose some physics goals
• by going to 25 central the rate of Fs is
  increased only by a factor of 2.5 since it scales
  with the number of produced particles

Geometrical acceptance

• By moving vertex Z range and making it more
  narrow around Z -30cm value. Gives maximum
  factor of 2 gain.
• This is not true for good tracks with hits in
  all spectrometer layers.
• Possible to use only if tracking very good in
  finding those difficult tracks.

14
Summary

• If we want to make high statistics F measurements
  at low pT we need both a very efficient tracking
  of partially registered Kaons from F decays and
  we would have to make an extra effort to increase
  statistics in every possible way. This would be
  risky.
• With reasonable tracking which seems possible we
  should be able to measure rate of F production at
  low pT and some high statistics measurements at
  higher pT. This however leaves only option of
  doing central triggering in order to increase
  statistics.

15
Backup

• BACKUP SLIDES

16
Banana tracking on Data
17
Geometrical acceptance for good decays
18
Acceptance
19
Acceptance

20
Summary

1. Results are limited only to yields, the mass peak
   position and width are only compatible with PDG
   values
2. dN/dy of f scales linearly with number of charged
   particles but the ratio grows with increasing
   energy of the collision
3. dN/dy of f scaling with both number of charged
   particles and collision energy follows closely
   Kaon production scaling
4. dN/dy and inverse slope measurement shows boost
   invariance for f within ylt1.0
5. The slope of f transverse mass distribution
   increases for AA collisions with energy
6. In AuAu collisions, within the measured
   centrality range, the shape of the transverse
   momentum distribution weakly depends on the
   centrality absence of significant hadronic
   re-scatterings for f?

21
dN/dy vs. nCharge
STAR preliminary

• dN/dy increases
• linearly with nCharge
• Statistical errors only.

22
f Mt Spectra
130 GeV 200 GeV
PRELIMINARY
PRC, 65 041901(R) (2002)
----- Mt exponential Fits
23
ltpTgt Centrality Dependence
1) p, K, p mean transverse momentum ltpTgt
increase in more central collisions 2) Heavier
mass particle ltpTgt increase faster than lighter
ones as expected from hydro type collective
flow
1) p, K, p mean transverse momentum ltpTgt
increase in more central collisions 2) Heavier
mass particle ltpTgt increase faster than lighter
ones as expected from hydro type collective
flow 3) f-meson seems flow differently.
Star preliminary