Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Thermodynamics at µ = 0 on the lattice
Y. Aoki
U. Wuppertal
RHIC Physics in the Context of the Standard Model
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 1 / 19
Staggered thermodynamics
Equation of State : YA, Z. Fodor, S.D. Katz, K.K. SzabóJHEP 0601:089,2006 [hep-lat/0510084]
Order of the transition : YA, G. Endrodi, Z. Fodor, S.D. Katz, K.K. Szabó
◮ Control most of the systematic errors in staggered thermodynamics◮ We cannot address locality problem of staggered
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 2 / 19
Staggered thermodynamics
Equation of State : YA, Z. Fodor, S.D. Katz, K.K. SzabóJHEP 0601:089,2006 [hep-lat/0510084]
Order of the transition : YA, G. Endrodi, Z. Fodor, S.D. Katz, K.K. Szabó
◮ Control most of the systematic errors in staggered thermodynamics◮ We cannot address locality problem of staggered
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 2 / 19
Systematic errors of the EoS
0.0
1.0
2.0
3.0
4.0
5.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
T/Tc
p/T4
pSB/T4
3 flavour2+1 flavour
2 flavour
Bielefeld 2000.
finite MD step size ∆τ
a dependence is notinvestigated
ma fixed while changing a(T = 1/(Nta))
Solution
Use exact algorithm: RHMC
Do several a with improvedaction
Tune ma to reproduce theT = 0 experimental spectrum:LCP
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 3 / 19
Systematic errors of the EoS
0.0
1.0
2.0
3.0
4.0
5.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
T/Tc
p/T4
pSB/T4
3 flavour2+1 flavour
2 flavour
Bielefeld 2000.
finite MD step size ∆τ
a dependence is notinvestigated
ma fixed while changing a(T = 1/(Nta))
Solution
Use exact algorithm: RHMC
Do several a with improvedaction
Tune ma to reproduce theT = 0 experimental spectrum:LCP
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 3 / 19
Discritization error
Rorational symmetry breaking
Taste symmetry breaking
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 4 / 19
Improving discritization error
Rotational symmetry breaking◮ Symanzik (tree level) gauge action→ improves gauge sector◮ Single (stout) link action→ does not improve f at T =∞
⋆ Large deviation from continuum SB⋆ a → 0 behavior is very good
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
4 6 8 10 12 14 16
Nτ
fF(0)(Nτ)/fcont, F
(0)STANDARD
NAIKP4
Heller, Karsch, Sturm
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 5 / 19
Improving discritization error
Rotational symmetry breaking◮ Symanzik (tree level) gauge action→ improves gauge sector◮ Single (stout) link action→ does not improve f at T =∞
⋆ Large deviation from continuum SB⋆ a → 0 behavior is very good
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
4 6 8 10 12 14 16
Nτ
fF(0)(Nτ)/fcont, F
(0)STANDARD
NAIKP4
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
1/Nt2 (∝ a
2)
0
0.5
1
1.5
2
2.5
3
f(N
t) / f
cont
Nt=6 4
Standard (stout, ours)Naik (Asqtad, MILC)p4 (Bielefeld)
Heller, Karsch, SturmY. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 5 / 19
Improving discritization error
Rotational symmetry breaking◮ Symanzik (tree level) gauge action→ improves gauge sector◮ Single (stout) link action→ does not improve f at T =∞
⋆ Large deviation from continuum SB⋆ a → 0 behavior is very good
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
4 6 8 10 12 14 16
Nτ
fF(0)(Nτ)/fcont, F
(0)STANDARD
NAIKP4
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
1/Nt2 (∝ a
2)
0
0.5
1
1.5
2
2.5
3
f(N
t) / f
cont
Nt=6 4
Standard (stout, ours)Naik (Asqtad, MILC)p4 (Bielefeld)
Heller, Karsch, SturmY. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 5 / 19
Improving discritization error
Rotational symmetry breaking◮ Symanzik (tree level) gauge action→ improves gauge sector◮ Single (stout) link action→ does not improve f at T =∞
⋆ Large deviation from continuum SB⋆ a → 0 behavior is very good
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
1/Nt2 (∝ a
2)
0
0.5
1
1.5
2
2.5
3
f(N
t) / f
cont
Nt=8 6
Standard (stout, ours)Naik (Asqtad, MILC)p4 (Bielefeld)
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
1/Nt2 (∝ a
2)
0
0.5
1
1.5
2
2.5
3
f(N
t) / f
cont
Nt=6 4
Standard (stout, ours)Naik (Asqtad, MILC)p4 (Bielefeld)
Heller, Karsch, SturmY. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 5 / 19
Improving discritization error
Taste symmetry breaking,local fluctuation of the gaugefield is responsible for
Reduced by the stout linksmearing(Morningstar & Peardon).
0 0.1 0.2 0.3 0.4 0.5
mG2
0
0.2
0.4
0.6
0.8
1
mπ’
2
Naive
a-1
=1 GeV
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 6 / 19
Improving discritization error
Taste symmetry breaking,local fluctuation of the gaugefield is responsible for
Reduced by the stout linksmearing(Morningstar & Peardon).
0 0.1 0.2 0.3 0.4 0.5
mG2
0
0.2
0.4
0.6
0.8
1
mπ’
2
NaiveStout Link
a-1
=1 GeV
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 6 / 19
LCP: to be always on the physical pointThe line of constant physics
mud(β), ms(β).
Tune ms using Nf = 3 degeneratesimulations and LO ChPT
m2PS
m2V
∣
∣
∣
∣
Nf =3
(m = ms) =m2
ηs
m2φ
=2m2
K −m2π
m2φ
.
obtain ms(β)
mud = ms/25
should be checked in Nf = 2 + 1simulation.
LCP1
3 3.2 3.4 3.6 3.8 4β
0
0.05
0.1
0.15
0.2
ms
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 7 / 19
How close to the physical and continuum limit ?
0 5 10 15mπ
2/Tc
2
0
5
10
15
(mπ’
2 -mπ2 )/
Tc2
MILC, standard
Bielefeld, p4
This work, stout
Nf=2, Nt=6
Nf=3, Nt=4
MILC, AsqtadNf=2+1, Nt=6
Nf=2+1, Nt=6
∆′
π =m2
π′ −m2π
T 2c
= (m2π′ −m2
π)N2t
(T = 0 masses, measured at βc(Nt ))
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 8 / 19
EoS procedureintegral along LCP
pT 4
∣
∣
∣
(β,m)
(β0,m0)= −
fT 4
∣
∣
∣
∣
(β,m)
(β0,m0)
= −N4t
∫ (β,m)
(β0,m0)
d(β,mud ,ms)
〈 − Sg/β〉〈ψψud〉〈ψψs〉
.
pT 4
∣
∣
∣
(β,m)
=p
T 4
∣
∣
∣
(β,m)
(β0,m0)(T 6= 0)−
pT 4
∣
∣
∣
(β,m)
(β0,m0)(T = 0).
RHMC algorithm does not require the extrapolation in step size
T=0: no msimud = mphys
ud data.
pT 4
∣
∣
∣
(β,mphysud )
=p
T 4
∣
∣
∣
(β,msimud )− N4
t
∫ mphysud
msimud
dmud 〈ψψud 〉.
Extrapolation only needed for 〈ψψud〉(mphysud )← msim
ud = {3, 5, 7, 9} ×mphysud .
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 9 / 19
EoS procedureintegral along LCP
pT 4
∣
∣
∣
(β,m)
(β0,m0)= −
fT 4
∣
∣
∣
∣
(β,m)
(β0,m0)
= −N4t
∫ (β,m)
(β0,m0)
d(β,mud ,ms)
〈−Sg/β〉〈ψψud 〉〈ψψs〉
.
pT 4
∣
∣
∣
(β,m)
=p
T 4
∣
∣
∣
(β,m)
(β0,m0)(T 6= 0)−
pT 4
∣
∣
∣
(β,m)
(β0,m0)(T = 0).
RHMC algorithm does not require the extrapolation in step size
T=0: no msimud = mphys
ud data.
pT 4
∣
∣
∣
(β,mphysud )
=p
T 4
∣
∣
∣
(β,msimud )− N4
t
∫ mphysud
msimud
dmud 〈ψψud 〉.
Extrapolation only needed for 〈ψψud〉(mphysud )← msim
ud = {3, 5, 7, 9} ×mphysud .
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 9 / 19
EoS procedureintegral along LCP
pT 4
∣
∣
∣
(β,m)
(β0,m0)= −
fT 4
∣
∣
∣
∣
(β,m)
(β0,m0)
= −N4t
∫ (β,m)
(β0,m0)
d(β,mud ,ms)
〈−Sg/β〉〈ψψud 〉〈ψψs〉
.
pT 4
∣
∣
∣
(β,m)
=p
T 4
∣
∣
∣
(β,m)
(β0,m0)(T 6= 0)−
pT 4
∣
∣
∣
(β,m)
(β0,m0)(T = 0).
RHMC algorithm does not require the extrapolation in step size
T=0: no msimud = mphys
ud data.
pT 4
∣
∣
∣
(β,mphysud )
=p
T 4
∣
∣
∣
(β,msimud )− N4
t
∫ mphysud
msimud
dmud 〈ψψud 〉.
Extrapolation only needed for 〈ψψud〉(mphysud )← msim
ud = {3, 5, 7, 9} ×mphysud .
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 9 / 19
Simulation Procedure
many β’s (16 pts for Nt = 4, 14 pts for Nt = 6)
given β, msims = mphys
s (β) fixed. → mphysud = mphys
s /25.T 6= 0:
◮ msimud = mphys
ud .◮ Ns = 3Nt .
T = 0:◮ msim
ud = {3, 5, 7, 9} ×mphysud
◮ keeping Lsmπ > 3
finite size effect?◮ less than stat. error for several β’s checked
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 10 / 19
EoS
1 1.5 2 2.5 3T/Tc
0
2
4
6
8
10
p/T
4
SB limit
Nt=4
Nt=6
Nt=∞Nt=4
Nt=6
1 1.5 2 2.5 3T/Tc
0
5
10
15
20
25
30
ε/T
4
SB limit
Nt=4
Nt=6
Nt=∞
Nt=4
Nt=6
Multiplied with cSB/cNt (cSB = limNt→∞cNt ), Nt = 4(red), 6(blue).
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 11 / 19
EoS
Multiplied with cSB/cNt (cSB = limNt→∞cNt ), Nt = 4(red), 6(blue).
Tc: inflection point of χI/T 2.
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 12 / 19
ms −mud phase diagram
Columbia 1990.
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 13 / 19
ms −mud phase diagram
Non-lattice approachms = mud →∞: global Z3 symmetry
◮ 3d, Z3 symmetric spin model:◮ low T : symmetric phase: confined◮ high T : broken phase: deconfined◮ 1st order transition
( Yaffe & Svetitsky )
1/m = 0+◮ small perturbation by external field
h ∝ e−βm (Banks & Ukawa)◮ 1st order transition
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 13 / 19
ms −mud phase diagram
Non-lattice approach
chiral symmetry:(Pisarski & Wilczek)
◮ ms →∞, mud = 0: Nf = 2:2nd order with O(4)
◮ ms = mud → 0: Nf = 3:1st order
◮ m = 0+: 1st order
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 13 / 19
ms −mud phase diagram
Non-lattice approach
chiral symmetry:(Pisarski & Wilczek)
◮ ms →∞, mud = 0: Nf = 2:2nd order with O(4)
◮ ms = mud → 0: Nf = 3:1st order
◮ m = 0+: 1st order
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 13 / 19
ms −mud phase diagram
Nt = 4Columbia 1990
Lattice approach
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 14 / 19
ms −mud phase diagram
Lattice approachms = mud →∞:
◮ 1st order confirmed by lattice⋆ 1st order (Columbia)⋆ 2nd order (Ape)⋆ 1st order by Finite Size Scaling
(Fukugita et al)
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 14 / 19
ms −mud phase diagram
Lattice approachms = mud →∞:
◮ 1st order confirmed by lattice⋆ 1st order (Columbia)⋆ 2nd order (Ape)⋆ 1st order by Finite Size Scaling
(Fukugita et al)
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 14 / 19
ms −mud phase diagram
Lattice approachms →∞, mud = 0: Nf = 2:
◮ Wilson fermion confirmed O(4)scaling
◮ Staggered: no O(4), O(2) scalingobserved.1st order ? (Pisa)
◮ possible ∆τ artifact(Philipsen Lattice2005)
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 14 / 19
ms −mud phase diagram
Lattice approachms = mud : Nf = 3: end point(Bielefeld-Swansea)
◮ mπ,c = 290 MeV [standard, Nt = 4]◮ mπ,c = 67 MeV [p4, Nt = 4]◮ Large discretization error!→ needs investigation on adependence
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 14 / 19
ms −mud phase diagram
Very important to determine theorder at the physical point: all µ 6= 0physics depend on this.
Position of the physical point:open question in the continuum limit.
Very demanding to explore all themass range.
Perhaps doable to determine theorder at the physical point.
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 14 / 19
Strategy
Exact algorithm (RHMC)
Finite size scalingCorrect quark masses
◮ tune ms and mu so thatratios of mπ, mK , fK takephysical values.
◮ LCP2: mu = ms/27.3, ms →
Continuum limit
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 15 / 19
Strategy
Exact algorithm (RHMC)
Finite size scalingCorrect quark masses
◮ tune ms and mu so thatratios of mπ, mK , fK takephysical values.
◮ LCP2: mu = ms/27.3, ms →
Continuum limit
LCP2
3 3.2 3.4 3.6 3.8 4β
0
0.05
0.1
0.15
0.2
0.25
0.3
ms
LCP1
LCP2
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 15 / 19
Finite Size Scaling of chiral susceptibility χud
Ns/Nt = 3− 5(6).No volume dependence found.How about in the continuum limit ?
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 16 / 19
Finite Size Scaling of chiral susceptibility χud
Ns/Nt = 3− 5(6).No volume dependence found.How about in the continuum limit ?
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 16 / 19
Continuum limit of the peak hightχ has power divergence, which should be subtracted.
∆χ = χ(T 6= 0)− χ(T = 0).
To make RG invariant quantity, mass squared is multiplied
m2ud∆χ
T 4c /(m
2ud∆χ).
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 17 / 19
Finite Size Scaling after continuum extrapolation
0 0.01 0.02 0.03 0.04
1/(VTc3)
0
10
20
30
T4 /(
m2 ∆χ
)
1st order
χ2 /dof=87/2
Susceptibility does not diverge as V →∞
→ Crossover
Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 18 / 19
Summary
Stout-link smearing improvement was used to reduce the tasteviolation. Works quite well.
EoS
The equation of state was calculated with Nt = 4 and 6 lattices,using LCP1 (approximation).
For the reliable continuum extrapolation, Nt = 8 simulation isneeded.
Order of the transition
Fine tuned LCP2 was used.
Continuum limit of chiral susceptibility was obtained usingNt = 4, 6, 8, 10.
Finite Size Scaling was applied.
Cross-over was found for the physical point.
Tc ?Y. Aoki (U. Wuppertal) Thermodynamics at µ = 0 on the lattice RBRC workshop 19 / 19