(B) Find N part for d+Au collisions? 0-10%10-20%40-60%100-80% Aneta Iordanova University of Illinois at Chicago N part Determination and Systematic Studies for √s NN = 200GeV d+Au Collisions in (A) What is N part ? IMPORTANCE: N part connects experiments with theoretical models UNFORTUNATELY: cannot be measured directly! Deuteron (d) Gold (Au) Parameters of Nuclear Collisions: Impact parameter b Number of participants N part Number of collisions N coll others … Participants Spectators b 2003 Run Detector Additions: 2 Time Zero Counters, 2 Spectrometer Trigger Scintillator Detectors (45 0 and 90 0 from beam pipe), Time of Flight Counters moved, 2 Proton Calorimeters How can we Measure N part ? Detector EOct [arb.units] N part Peripheral → Central Npart is unknown Assumptions: Some measured quantities in the DATA correctly reflect the collision geometry ! Chose that Quantity carefully → For Example: Quantity proportional to the Minimum Ionizing ‘Energy’ Signal in -3 <  < 3 (Octagon detector), which we call simply EOct Experimental approach Use Theoretical Models N part is calculated! In Monte Carlo: Map the simulated quantity (EOct) to the calculated N part for a given model Associate equivalent MC/Data (EOct value) with the same Connect Data - MC DATA measured cross section MC distribution with trigger and vertex biases Data and MC (biased) distributions match well Data cut = MC cut X scale factor Extract for each bin in biased MC distribution Scale Normalize Unbiased EOct distribution– represents the full geometrical cross-section Found final /, RMS/ % Cross-Section Systematic errors on Trigger Bias : Accounts for the missed part of the cross-section Is our Trigger simulated properly? Vertex Bias : Provides clean event sample - Compared MC (Trigger +Vertex Bias) with MC (no bias) - only peripheral bins affected Overall Efficiency from EOct Distribution Shape matching - from Data/Hijing MC (~82%) - from Data/Glauber MC (upper limit on how wrong we could be) 10% error Different Triggered Events → Introduce “Trigger Bias” Minimum Bias = Paddle Trigger Counters, Single Arms This Trigger sees the most of the collision cross-section “Single arm” At least one hit in either Positive or Negative Paddle Counter in coincidence with the Crossing Clock (+CC) - Positive Paddle Single arm – PPSingle - Negative Paddle Single arm – PNSingle “Two arm” At least one hit in each Paddle Counter +CC - narrow Paddle coincidence (removes the background) – PP&PNn 100% 97% 91% 51% Ratio (2hit/1hit) in Paddle Counters Trigger PPSingle Trigger PNSingle Data MC Counts Efficiency EOct [arb.units] Counts All MC Events 100% PN&PP Trigger+Vertex 82% from (MC) EOct distribution from N part distribution % Cross-Section Hulthen – Woods-Saxon Woods-Saxon % Cross-Section % Difference Glauber MC Studies (inelastic cross-section  NN = 42mb) Gives us the upper limits in our systematic errors on Only N part available (no detector simulated) Very difficult to introduce smearing which looks like EOct All Studies Follow Steps: Match Hijing and Glauber distributions Use cut positions from Hijing Find from Glauber MC Apply error to from Hijing Distributions do not match well! for the most central and peripheral bins is different compared to Hijing Not Realistic → Must have Smearing Glauber MC (N part -2) HIJING EOct Case1: No smearing (use N part ) Glauber MC (N part -2) + pG HIJING EOct Introduce some smearing G ~ √N part *Gaus(0,1) Case2: Scan for different parameters p Smearing closer to Hijing EOct for p~1 still deviates from Hijing for the peripheral bin. Should include the Trigger Bias Glauber MC p 1 N part + p 2 G + p 3 N part 4/3 HIJING EOct Add additional scaling term ~N coll Case3: Affects low centrality HIJING with bias / HIJING Peripheral bins with bias have larger (+ 5%, 3%, 1%) compared to the unbiased cases Case4: N part Centrality Bin Errors Glauber Errors Efficiency Final Central → Peripheral Example for final systematic errors on - combined studies from Hijing and Glauber MC Monotonic dependence between N part and EOct Slice this distribution into percentile bins of cross-section (6 bins, for example) Peripheral Collisions: Impact parameter, b, is large Central Collisions: Impact parameter, b, is small Peripheral → Central Hijing Monte Carlo % Cross-Section Bins Peripheral → Central (C) Are we sure? Trigger Configuration Simulated: All MC Events (difficult to see!) PP or PN Single (1 Hit or more in PP OR 1 Hit or more in PN ) PN&PP (1 Hit or more in PP AND 1 Hit or more in PN ) T0N&T0P (1 Hit or more in T0P AND 1 Hit or more in T0N ) Trigger Bias for this analysis EOct [arb.units] “Trigger Bias” + “Vertex Bias” Efficiency (Overall Event Selection) Hijing Monte Carlo (E) Hijing Monte Carlo Errors due to smearing: Simulated EOct has effects from electronic noise and spatial (vertex) distribution Slice Monte Carlo N part distribution in %-le bins ( “true” result) (smearing affects most central and peripheral bins) Peripheral → Central Errors due to different Hijing versions: We used 2 Hijing versions, with different nuclear density profiles for the deuteron Compare the N part values for the two Hijing versions - Woods-Saxon (1.381) - Hulthen (1.383) (small difference between these versions) (F) Glauber Monte Carlo Hijing uses Monte Carlo similar to Glauber multiple scattering model to calculate N part What if Hijing N part calculations are wrong? Scan for different parameters ( many more scaling and smearing function tested ) Add Trigger + Vertex Bias Counts N part, the number of nucleons participating in the collision Octagon Detector Z, η Schematic Plot not to scale E Oct Counts EOct Birger Back, Mark Baker, Maarten Ballintijn, Donald Barton, Russell Betts, Abigail Bickley, Richard Bindel, Wit Busza (Spokesperson), Alan Carroll, Zhengwei Chai, Patrick Decowski, Edmundo García, Tomasz Gburek, Nigel George, Kristjan Gulbrandsen, Stephen Gushue, Clive Halliwell, Joshua Hamblen, Adam Harrington, Conor Henderson, David Hofman, Richard Hollis, Roman Hołyński, Burt Holzman, Aneta Iordanova, Erik Johnson, Jay Kane, Nazim Khan, Piotr Kulinich, Chia Ming Kuo, Willis Lin, Steven Manly, Alice Mignerey, Gerrit van Nieuwenhuizen, Rachid Nouicer, Andrzej Olszewski, Robert Pak, Inkyu Park, Heinz Pernegger, Corey Reed, Michael Ricci, Christof Roland, Gunther Roland, Joe Sagerer, Iouri Sedykh, Wojtek Skulski, Chadd Smith, Peter Steinberg, George Stephans, Andrei Sukhanov, Marguerite Belt Tonjes, Adam Trzupek, Carla Vale, Siarhei Vaurynovich, Robin Verdier, Gábor Veres, Edward Wenger, Frank Wolfs, Barbara Wosiek, Krzysztof Woźniak, Alan Wuosmaa, Bolek Wysłouch, Jinlong Zhang ARGONNE NATIONAL LABORATORYBROOKHAVEN NATIONAL LABORATORY INSTITUTE OF NUCLEAR PHYSICS, KRAKOWMASSACHUSETTS INSTITUTE OF TECHNOLOGY NATIONAL CENTRAL UNIVERSITY, TAIWANUNIVERSITY OF ILLINOIS AT CHICAGO UNIVERSITY OF MARYLANDUNIVERSITY OF ROCHESTER Hijing Monte Carlo Errors due to Efficiency: Over-estimating or under- estimating the overall event selection efficiency from MC with comparison to Data has an effect on the cut positions when slicing EOct distribution in percentile cross-section bins Compare the N part values when efficiency varied with ± 5% (small effect on the central bins and larger, accumulative effect on the peripheral bins) Overestimated efficiency, cut position direction move Counts Unbiased EOct Distribution Trigger+Vertex Biased EOct Distribution Hijing Monte Carlo EOct (D) Bias and Efficiency d Au ZDC N ZDC P dAu x z PP PN CN CP T0NT0P Positive Paddle Counter 16 scintillator slats  t[ns]  z[m] Negative Zero Degree Calorimeter Positive Čerenkov Counter Negative T0 (TimeZero) Counter 10 Čerenkov counters Central → Peripheral