(AMPT) Model - RHIG - Wayne State University

Report
Update of Initial Conditions
in A Multiple Phase Transport (AMPT) Model
Zi-Wei Lin
Department of Physics
East Carolina University
Greenville, NC
Work still in progress
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
1
Outline
Present status of the AMPT model
Need to update the initial conditions
Optimize parameters/functions by fitting dNch/dη data
Outlook
Summary
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
2
What do we need
for simulations of high energy heavy ion collisions?
Options:
We need:
Initial particle/energy production
Soft+hard model, CGC, pQCD, ...
Parton interactions/
equation-of-state
Parton cascade (ZPC, MPC, BAMPS),
hydrodynamics, dE/dx, ...
Hadronization
/QCD phase transition
String fragmentation, quark coalescence,
independent fragmentation, statistical
hadronization, ...
Hadron interactions
Hadron cascade (ART, RQMD, UrQMD,
...), freezeout temperature, …
The AMPT model includes the components in green.
In particular, it can be used to study
coalescence of partons into hadrons, thermalization and flow,
dynamical chemical freeze-out and kinetic freeze-out
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
3
Structure of AMPT v1.xx (default model)
HIJING (PDFs, nuclear shadowing):
minijet partons,
excited strings, spectators
A+B
Less partonic
interaction
ZPC (Zhang's Parton Cascade)
Partons freeze out
Hadronization (Lund String fragmentation)
Dominated by
hadronic
interactions
(at very high
densities)
ART (A Relativistic Transport model for hadrons)
Hadrons freeze out (at a global cut-off time);
strong-decay all remaining resonances
Final particle spectra
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
4
Structure of AMPT v2.xx (String Melting model)
HIJING (PDFs, nuclear shadowing):
minijet partons,
excited strings, spectators
A+B
Melt to q & qbar via
intermediate hadrons
Partonic
interactions
dominate.
ZPC (Zhang's Parton Cascade)
Partons freeze out
Hadronization (Quark Coalescence)
Better describes
flow & HBT,
but does not
describe well
single particle
spectra
ART (A Relativistic Transport model for hadrons)
Zi-Wei Lin (ECU)
Hadrons freeze out (at a global cut-off time);
strong-decay all remaining resonances
Final particle spectra
28th WWND, Puerto Rico
April 10, 2012
5
AMPT Source Codes
First public release of AMPT codes: ~ April 2004.
Detailed physics descriptions in
Lin, Ko, Li, Zhang & Pal, PRC 72, 064901 (2005).
Versions v1.21/v2.21 (2008) and v1.11/v2.11 (2004)
are available at
https://karman.physics.purdue.edu/OSCAR-old/
http://personal.ecu.edu/linz/ampt/
also contains more recent test versions, including
v1.25t3/v2.25t3
(8/2009)
v1.25t7/v2.25t7
(9/2011)
v1.25t7b/v2.25t7b
(2/2012)
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
6
http://personal.ecu.edu/linz/ampt/
looks like this
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
7
Update Initial Conditions of AMPT
A+B
HIJING (PDFs, nuclear shadowing):
minijet partons,
excited strings, spectators
Final particle spectra
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
8
Similar updates have been done recently:
In HIJING2.0:
Deng, Wang and Xu, PRC 83 (2011) & PLB 701 (2011):
used GRV (Gluck-Reya-Vogt) parton distribution functions,
 parameterized functions p0(√sNN) & σsoft(√sNN);
new parameters for quark and gluon nuclear shadowing functions
are used to reproduce dNch/dη in AA collisions.
In AMPT:
Pal & Bleicher, PLB 709 (2012)
used HIJING2.0 as initial conditions,
smaller value for the gluon shadowing parameter sg
is needed to reproduce dNch/dη in AA collisions at LHC,
since rescatterings considerably reduce hadron yields at mid-rapidity
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
9
Rescatterings considerably reduce hadron yields at mid-rapidity
In AMPT:
Lund model's a & b values used for pp/HIJING
can describe SPS dN/dy
when final state interactions are turned off.
We have to use different a & b values
to describe dN/dy of AA collisions.
Lin, Ko, Li, Zhang & Pal, PRC 72 (2005);
first shown in PRC 64 (2001).
but this agreement is gone
when final state interactions are included.
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
10
1) Need to Use up-to-date parton distribution functions
It is essential to use
up-to-date PDF
for LHC
&
for heavy flavors
(since ~all come from gluons)
Duke-Owens Set 1 (1984)
used in HIJING1.0 & AMPT
significantly under-estimates
the gluon density at small-x
LHC
RHIC
For this study, we have incorporated into AMPT
the CTEQ6M PDF
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
11
2) Need to use up-to-date nuclear shadowing functions RA(x)
PDF in a nucleus
≠ PDF in a nucleon *A
fA (x) = A* fN (x)* RA (x)
Central Pb+Pb collisions
at √sNN=5.5 TeV
from default AMPT v1.11:
shadowing has a large effect
Deng, Wang and Xu, PLB 701 (2011)
For this study, we have incorporated into AMPT
the EPS09 nuclear shadowing
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
12
Optimize parameters by fitting dNch/dη data
with the default AMPT model (+CTEQ6M & EPS09)
2é
- c (b,s) ù
s
=
2
p
db
1e
ò ë
total
û,
The nucleon-nucleon cross section
in the eikonal approximation:
c (b,s) =
TNN (b) é
ës soft + s jet ( p0 )ùû.
2
In AMPT/HIJING1.0:
p0=2.0 GeV/c,
σsoft≈57 mb.
We follow the following strategy
similar to Deng, Wang and Xu, PRC 83 (2011):
At each collision energy √sNN:
• choose different p0 values
• for each p0, determine the σsoft value that reproduces the experimental σpptotal
• run AMPT for each (p0, σsoft) set,
then compare with dNch/dη data to find the best (p0, σsoft) value.
Go through all relevant collision energies:
p0(√sNN) & σsoft(√sNN)
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
13
Example of the (p0, σsoft) sets
√sNN=200 GeV
from PDG
√sNN=7 TeV
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
14
AMPT results for different (p0, σsoft) sets
vs pp inelastic data
At √sNN=19.6 GeV:
p0=1.0, 1.2, 1.4, 1.6, 1.8 GeV/c.
…
At √sNN=200 GeV:
p0=2.0, 2.2, 2.4, 2.6 GeV/c.
…
At √sNN=7 TeV:
p0=3.0, 3.5, 4.0, 4.5,
5.0, 5.5, 6.0 GeV/c.
Higher p0 gives lower dNch/dη
(except for the lowest energy
√sNN=19.6 GeV)
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
15
Fit p0 to inelastic pp data at different energies
dNch/dη (|η|<0.5) INEL
=2.22 ±0.05
√sNN=200 GeV
 p0=2.35 +0.08-0.09 GeV/c
using AMPT results with interpolation
√sNN=2360 GeV
dNch/dη (|η|<0.5) INEL
=3.77 +0.25-0.12
 p0=3.77 +0.18-0.33 GeV/c
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
16
p0 values fitted to inelastic pp data
p0(√sNN)
increases with collision energy,
related to more partons
at small-x in the new PDF
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
17
Fit p0 to central AA data at different energies
after incorporating EPS09 shadowing functions in AMPT
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
18
AMPT results for different (p0, σsoft) sets
vs central AA data
AuAu at √sNN=19.6 GeV:
p0=1.2, 1.4, 1.6, 1.8 GeV/c.
…
AuAu at √sNN=200 GeV:
p0=2.0, 2.2, 2.3, 2.4, 2.6 GeV/c.
PbPb at √sNN=2760 GeV:
p0=3.5, 4.0, 4.5 GeV/c.
Higher p0 gives lower
dNch/dη/(Npart/2)
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
19
Fitted p0 values
combined
p0(√sNN) fitted to pp data and to AA data are not consistent

This test of the PDF+shadowing update cannot systematically describe
dNch/dη of pp & AA collisions throughout this energy range
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
20
Outlook
Search for consistent p0(√sNN)
from fits to pp data and to AA data
Possibilities include:
1- z ) -bm^2 /z
(
f (z) µ
e
z
a
1) a & b parameters in Lund string fragmentation:
a=0.3 & b=0.8/GeV2 are used in this study,
this is one set of the 2009 fit values in PYTHIA 8.1;
we can explore the a-b parameter space:
e.g. PYTHIA have used a=0.76, b=0.58/GeV2 (2007 fit
values),
and used a=0.30, b=0.58/GeV2 before.
2) Alternative nuclear shadowing.
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
21
Summary
• Work is under way to incorporate into AMPT
up-to-date parton distribution functions and nuclear shadowing functions
• CTEQ6M and EPS09 have been tested within AMPT,
but the functions p0(√sNN) & σsoft(√sNN)
from fits to pp dNch/dη data and fits to central AA data are not consistent
• We will investigate
Lund a,b parameters and alternative nuclear shadowing functions
to obtain consistent p0(√sNN) & σsoft(√sNN) from fits to pp and AA data;
that would allow systematic descriptions
of dNch/dη in pp & AA collisions throughout a wide energy range.
• May require more significant developments:
dynamical quark coalescence in phase space
(instead of space/nearest neighbors),
inelastic parton interactions
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
22
Thank you!
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
23
Backup Slides
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
24
Files in the Source Code include
amptsub.f
art1f.f
hijing1.383_ampt.f
hipyset1.35.f
linana.f
main.f
zpc.f
Fortran routines
input.ampt
Input parameter values
Update of initial conditions introduces new input data files:
cteq6m.tbl
CTEQ6M PDF table
EPS09LOR_197
EPS09LOR_208
EPS09 shadowing function table
for Au and Pb nucleus
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
25
Main Ingredients
HIJING
ZPC
soft strings + hard minijets
2↔2 parton cascade:
gg↔gg, gg↔qqbar, gq↔gq, ...
Hadronization Lund string fragmentation
or
quark coalescence
ART
hadron cascade including:
p r w h K K* f
n p D N * (1440) N * (1535) L S X W
deuteron
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
26
Earlier modifications in the AMPT source code
2005 v1.12/v2.12:
Freezeout time of spectator projectile and target nucleons should be ~0
but not correctly updated in 'ampt.dat';
now corrected
10/2008 v1.21/v2.21:
Added option to turn off φ meson decays at the end of hadron cascade
i.e., at NT=NTMAX
10/2008 test version v1.22/v2.22:
Included deuteron(d) interactions in hadron cascade via d+M ↔ B+B
(M or B represents a meson or a baryon),
also included elastic collisions of d+M and d+B;
similar anti-deuteron interactions are also included.
3/2009 test version v1.23/v2.23:
Included a subroutine to enable users to insert user-defined hadrons
before the start of the hadron cascade
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
27
Recent modifications in the AMPT source code
6/2009 test version v1.25t1/v2.25t1:
Added an option of event selection so that each event will have at least 1 minijet parton
above a set Pt value in the initial condition;
Added an option to embed a back-to-back high-Pt q/qbar pair in each event;
Write out Npart information (spatial coordinates and status of each incoming nucleon);
Added option to write complete parton information before and after the parton cascade
and the full parton collision history for the string melting version
7/2009 test version v1.25t2/v2.25t2:
Added an option to enable users to modify nuclear shadowing smoothly
between no-shadowing and the default HIJING shadowing
5/2011 test version v1.25t4/v2.25t5:
Included the finite widths of resonances (K* η ρ ω Φ Δ)
when they are produced from quark coalescence in the string melting version
2/2012 test version v1.25t7b/v2.25t7b:
Added option to enable random orientation of reaction plane
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
28
Files in the Source Code include
amptsub.f
art1f.f
hijing1.383_ampt.f
hipyset1.35.f
linana.f
main.f
zpc.f
README
Fortran routines
Instructions
(including summary of changes)
Makefile
exec
Script to run AMPT
input.ampt
ana/
Input parameter settings
Directory for output data and diagnostics files
update of initial conditions will introduce new input data files:
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
29
Parameters in
input.ampt:
New options in red
Initial Conditions/
HIJING
(e.g. turn on quenching to
mimic inelastic energy loss)
Hadron Cascade
(e.g. NTMAX=2 turns off
hadron cascade
but still has full parton
cascade & hadronization)
Parton Cascade
Hadronization
Output options
Zi-Wei Lin (ECU)
200.
! EFRM (sqrt(S_NN) in GeV)
CMS
! FRAME
A
! PROJ
A
! TARG
197
! IAP (projectile A number)
79
! IZP (projectile Z number)
197
! IAT (target A number)
79
! IZT (target Z number)
2
! NEVNT (total number of events)
0.
! BMIN (mininum impact parameter in fm)
13.
! BMAX (maximum impact parameter in fm, also see below)
1
! ISOFT (D=1): select Default AMPT or String Melting(see below)
150
! NTMAX: number of timesteps (D=150), see below
0.2
! DT: timestep in fm (hadron cascade time= DT*NTMAX) (D=0.2)
2.2
! PARJ(41): parameter a in Lund symmetric splitting function
0.5
! PARJ(42): parameter b in Lund symmetric splitting function
1
! (D=1,yes;0,no) flag for popcorn mechanism(netbaryon stopping)
1.0
! PARJ(5) to control BMBbar vs BBbar in popcorn (D=1.0)
1
! shadowing flag (Default=1,yes; 0,no)
0
! quenching flag (D=0,no; 1,yes)
1.0
! quenching parameter -dE/dx (GeV/fm) in case quenching flag=1
2.0
! p0 cutoff in HIJING for minijet productions (D=2.0)
3.2264d0
! parton screening mass in fm^(-1) (D=3.2264d0), see below
0
! IZPC: (D=0 forward-angle parton scatterings; 100,isotropic)
0.47140452d0
! alpha in parton cascade
1d6
! dpcoal in GeV
1d6
! drcoal in fm
0
! ihjsed: take HIJING seed from below (D=0)or at runtime(11)
53153523
! random seed for HIJING
8
! random seed for parton cascade
0
! flag for Ks0 weak decays (D=0,no; 1,yes)
1
! flag for phi decays at end of hadron cascade (D=1,yes; 0,no)
0
! optional OSCAR output (D=0,no; 1,yes; 2,initial parton info)
28th WWND, Puerto Rico
April 10, 2012
30
Output files
ampt.dat
Event#
1
2112
2112
2212
2212
111
Test#
Particle# b(fm)
Npart2
Npart1
1 4218 8.0000
84 84 2
0.000 0.000 99.996 0.940
0.000 0.000 99.996 0.940
0.000 0.000 99.996 0.940
0.000 0.000 -99.996 0.940
0.071 -0.334 -0.376 0.135
Particle ID
(PYTHIA)
Final momentum
mass
82 1
6.00
6.78
5.53
-9.31
1.81
83
-4.86 0.23
3.61 0.21
1.48 0.26
-2.75 -0.16
-0.96 -1.18
0.00
0.00
0.00
0.00
7.00
Final position & time
(at kinetic freeze-out)
zpc.dat
Final momentum, position & time of all partons (at kinetic freeze-out)
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
31
EPS09 (JHEP 2009) determines the nuclear modification
to the free proton PDF from the CTEQ6.1M set in the MS scheme
The CTEQ6.1M set provides a global fit
that is almost equivalent in every respect to the published CTEQ6M
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
32
Needs to Further Develop AMPT
Need to use up-to-date Parton Distribution Functions in nuclei:
essential for heavy flavors & LHC
Parton coalescence to hadrons:
Currently, a parton can only coalesce after it does not have
further interactions (i.e., after kinetic freezeout);
Average parton density at coalescence, and effective equation of
state of AMPT depends on σp;
Need to improve Parton coalescence for better EoS;
hadronization condition (~HBT)
& parton cross section (~v2) will be decoupled
 Inelastic partonic interactions
 Including color fields in parton phase.
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
33
RRPL2u (19) model predictions of sigma_total for pp collisions
(ecm==sqrt_s>=5GeV):
sigma_total=
35.45 - 33.34/((ecm^2)^0.5453)
+ 42.53/((ecm^2)^0.4581) + 0.3079*(-3.364 + Log[ecm^2])^2
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
34
Fitting ALICE INEL>0 pp data
At √sNN=900 GeV:
p0=2.0, 2.5, 3.0, 3.5,
4.0, 4.5, 5.0 GeV/c
At √sNN=2360 GeV:
p0=2.5, 3.0, 3.5, 4.0,
4.5, 5.0, 5.5 GeV/c
At √sNN=7 TeV:
p0=3.0, 3.5, 4.0, 4.5,
5.0, 5.5, 6.0 GeV/c
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
35
Zi-Wei Lin (ECU)
28th WWND, Puerto Rico
April 10, 2012
36

similar documents