• 検索結果がありません。

RG flow of 2D O(N ) Spin Models and Absence of Phase Transitions

N/A
N/A
Protected

Academic year: 2022

シェア "RG flow of 2D O(N ) Spin Models and Absence of Phase Transitions"

Copied!
2
0
0

読み込み中.... (全文を見る)

全文

(1)

RG flow of 2D O(N ) Spin Models and Absence of Phase Transitions

Keiichi R.Ito, Setsunan Univ.

September, 2011

The 2DO(N) spin model with large N is studied by the renormalization group (RG) method triggered by Wilson and founded mathematically by Gawedzki and Kupiainen [6]. Let ϕ(x) RN be the N 1 dimensional sphere of radius (N β)1/2. Using δ(ϕ2 −N β) =

exp[−iψ(ϕ2 −N β)]dψ/2π [4],

ZΛ =

· · ·

exp[−W0(ϕ, ψ)]∏jj

2π (1)

W0(ϕ, ψ) = 1

2 < ϕ,(m2∆)ϕ >+ i

√N < ϕ2−N β, ψ > (2) where (∆)xy = 4δxy −δ|xy|,1 is the 2D lattice laplacian, ϕ(x) RN is dimensionless boson field and ψ(x) R is the auxiliary field which has the dimension (length)2. Let G0 = (∆ +m2)1. The mass parameter m ∼e2πβ is chosen so thatG0(0) =β. Define Gn and C :RΛn →RΛn+1 by

Gn+1(x, y) = (CGnC+)(x, y), (Cf)(x) = 1 L2

z∈20

f(Lx+z) (3) where Lis a positive integer and 20 is the box of size L×L centered at the origin. We apply two types of block transformations C and C = L2C to Wnn, ψn). One is the block spin transformation C : ϕn(x) ϕn+1(x) = (Cϕn)(x), and the other is the block spin transformation C : ψn(x) ψn+1(x) = (Cψn)(x). The main part of Wnn, ψn) is given by

Wnn, ψn) = 1

2⟨ϕn, Gn1ϕn+1

2γn⟨∂µϕn,n⊗ϕn)∂µϕn +⟨ψn, Hn1ψn+ i

√N⟨2n−un), ψn (4)

1

(2)

and the flow of Wn is parametrized by the mass parameters m2n∼L2nm20 in Gn1 (∆ +m2n) and slowly changing parameters γn ∼n/N and un. Here Hn1 =O(1)>0 is a local Hamiltonian of ψn and

n⊗ϕn)(x,i),(y,j)=δx,yϕi(x)ϕj(x) (5) un=N βn, βn=β−const. n+o(n) (6) un is the position of the double-well of the potential. The most surprising term γn⟨∂µφn,n ⊗φn)∂µφn means that the fluctuation field ξn µϕn is almost perpendicular to the block spin field ϕn since γn 0 increases as n → ∞. This term is a reminiscence of 2k−uk), ψk,k ≤n and they sum up to yield γn. Fortunately, however, this term does not disturb the main stream of the RG flow. (This term was found in [2, 3]where it was not serious because of the hierarchical approximation. This was re-encountered in [5].)

These steps can be iterated [7] excluding large field regions. We expect [7] that this leads us to the conclusion given in the title of the talk. This idea may be applied to the study of the non-abelian lattice gauge theory [1].

References

[1] K.R.Ito, Permanent Quark Confinement in 4D Hierarchical LGT of Migdal-Kadanoff Type, Phys. Rev. Letters 55: 558-561 (1985).

[2] K.R.Ito, Origin of Asymptotic Freedom in Non-Abelian Field Theories, Phys.Rev.Letters, 58: 439 (1987)

[3] K.R.Ito, Renormalization Group Flow of 2D Hierarchical Heisenberg Model of Dyson-Wilson Type, Commun. Math.Phys., 137: 45 (1991) [4] D. Brydges, J. Fr¨ohlich and T. Spencer, The Random Walk Represen-

tation of Classical Spin Systems and Correlation Inequalities, Commun.

Math. Phys.83: 123 (1982).

[5] D.Brydges, J.Dimock and P.Mitter, Notes on O(N) ϕ4 models, unpub- lished paper (2010, private communication through D.Brydges.)

[6] K.Gawedzki and A.Kupiainen, Commun.Math.Phys. 99 (1985) 197;

ibid. 106 (1986) 535

[7] K.R.Ito, Paper in Preparation (2011).

2

参照

関連したドキュメント

Thus, parties can note what the other party is measuring, by comparing the results sent to them on the qubits which are in the corresponding basis, when there should be a

Standard domino tableaux have already been considered by many authors [33], [6], [34], [8], [1], but, to the best of our knowledge, the expression of the

In this paper, we classify large P´olya-Eggenberger urns with regard to their asymptotics, give some generic example of each case and some other new results about particular families

For a positive integer n, we write P (n) for the largest prime factor of n, ω(n) for the number of distinct prime divisors of n, and τ(n) for the total number of positive

Our method of constructing intertwining first-order differential operators between generalized principal series representations for G has its origin in the generalized gradients of

We establish why expected value is insensitive to catastrophic risks see the study by Chichilnisky 1996, and use another criterion to evaluate risk based on axioms for choice

— Holonomic and semi-holonomic geometries modelled on a homogeneous space G/P are introduced as reductions of the holonomic or semi-holonomic frame bundles respectively satisfying

The answer, I think, must be, the principle or law, called usually the Law of Least Action; suggested by questionable views, but established on the widest induction, and embracing