Transcript heavyFlavorTriggerProp

Heavy Flavor
Flavor in
in ‘04:
‘04:
Heavy
..
Prospects for
for J/Y,
J/Y,
Prospects
Manuel Calderón de la Barca Sánchez
Heavy Flavor Working Group
Analysis Meeting,
November 9-11, 2003
Status of Studies for 2004 Run
 Full document on the web
 http://www.star.bnl.gov/protected/heavy/www/proposals/onium.pdf
 Estimated Production Rates
 Trigger
 Implementation, Efficiency and Rejection Power
 Simulations and Tests in d+Au and p+p
 Prospects for Measurements in Au+Au
 Plans and Beam Use
2
Production Rates
 Rates are
 per event, into e+e-, per unit rapidity at y=0
Minbias
Top 5%
Top 10%
Top 20%
J/Y
1.35 x 10-4
6.21 x 10-4
5.54 x 10-4
4.46 x 10-4
Y’
4.96 x 10-6
2.28 x 10-5
2.03 x 10-5
1.64 x 10-5
U
1.18 x 10-7
5.43 x 10-7
4.84 x 10-7
3.89 x 10-7
U’+U’’
4.31 x 10-8
1.98 x 10-7
1.77 x 10-7
1.42 x 10-7
 Ok, how many do we see?
 Acceptance, centrality selection, offline cuts…
 J/Y: Requiring Eelectron>1.2 reduces the signal
3
Production Rates II
J/y Yield:
24 J/y  e+e-, p>1.2 GeV
per 1M min-bias events
U Yield:
0.046 U  e+e-, p>3 GeV
per 1M min-bias events
98 J/y  e+e-, p>1.2 GeV
per 1M 10% central events
0.19 U  e+e-, p>3 GeV
per 1M 10% central events
(Triggered event ~ central)
Include:
- fAB (hard fraction)
- 1 unit  4p
- Acceptance (3/4 Barrel)
- momentum cut
4
Simulations for Au+Au @ 200 GeV
 Background
 Hijing Minbias
 Signal
 Pythia; settings tuned for J/Y, modifications for U
 Processed Background and signal in GSTAR
 Mixed 1 quarkonium event with 1 Hijing event
 Done at ZEBRA bank level (M. Potekhin)
 Run algorithms
 Hijing dataset : Background Rejection Power
 Hijing + quarkonium : Trigger Efficiency
5
EMC Topology Trigger
 Divide f into 6
sections
 Find a tower above a
threshold
 Look in the 3
opposite sections in f
 If another tower
above threshold,
trigger.
6
Rejection
Efficiency
J/y Trigger: L0 Efficiency and Rejection
Cutting at E=1.2-1.4 GeV is the region where we can work.
Less than that:
•Tower calibration
•No rejection power
More than that:
•Little signal left. Efficiency is 10% at 1.6 GeV
7
J/y Trigger: L0 Efficiency and Rejection
No free lunch:
High rejection only for
peripheral events.
Most signal in central
events.
98% of the yield is in
top 60% central.
8
U Trigger: L0 Efficiency and Rejection
Compare 2 approaches at L0, the topology trigger and
the single High-Tower trigger.
9
U Trigger: 1/Rejection vs Efficiency
High-Tower trigger at L0 gives better efficiency for the
same rejection power. Decision: use High-Tower at L0
for the upsilon trigger.
10
J/y: L0 & L2 Efficiency and Rejection
Most of the work is done by the topology trigger.
Again, high rejection only for peripheral events.
11
J/y: L0 & L2 Efficiency and Rejection
For central
collisions, every
event fires
Trigger works for low multiplicity, where yield is small.
Not much to gain…
12
U Trigger: L2 Efficiency and Rejection
Using High-Tower at level 0 and L2 fast invariant mass:
Sample the parameter space of the triggers in tandem.
Minimum Bias
Central
Can achieve rejections of 103 keeping the efficiency above 80%.
13
U Trigger Performance
e
R
Cuts
L0
L2
HighTower
ADC
Cluster
Energy
Cos q
Mass
Efficiency
0.95
300
10
4.5
0
7
Balanced
0.90
833
10
5.0
-0.2
6
Rejection
0.85
2174
10
5.0
0
8
Trigger works beautifully. Large rejection power and good
Efficiency.
The problem is that there is not much signal!
Need Luminosity!!
14
Putting it all together: J/Y Rates
Rejection is good only in
peripheral trigger.
}
Yield is small in
peripheral. Trigger
enhances by only a
factor ~2 there.
Killer: Would require 14%
of STAR’s bandwidth
Simply take minbias and
central triggers, obtain a
good sample.
15
U Rates
U Trigger enhances
yield by more than a
factor of 10.
Trigger works!
Even at the highest L,
L0 Rate is ~100 Hz
L2 Rate is ~ 3 Hz
Low rate.
L2 Rate can be
reduced by ~3
keeping efficiency high
Room to move.
16
Heavy Quarks in Run IV: Charm
 J/Y Trigger works only in peripheral
 Very well suited for UPC (F. Meissner)
 Yield is small (2% in 100-60% bin)
 Background rate is high
 Would require 14% of STAR’s bandwidth
 Given this, enhancement of only a factor of 2.
 We will not request a dedicated trigger.
 J/Y Analysis will rely on large minbias and central
data sets.
 50 x 106 minbas + 50 x 106 central
 3 s signal + 5 s signal in central
 Based on very conservative estimates on
background rejection (significance ~ e/h)
17
Heavy Quarks in Run IV: Beauty

 Trigger looks promising
 Efficiency can be kept above 85%
 Rates are low!!
 Take advantage of full luminosity
 Same applies to ALL rare triggers (High-Tower)
 Shift crews need to be FAST.
 Will rely on L3 Express stream for quick
turnaround
 Conservative estimate: ~100 Upsilons in 2004
 U has never been measured in heavy ions before
 Next step, implementation and testing online.
(Hank, Falk, Tonko, Chris)
18
Effect of Cos q Cut
L2 cut on cos q < -0.2 is
still quite safe for U.
Topology trigger, limit by p/3,
Limits J/Y p^ to ~ 5 GeV/c.
19