4. KT 相图，O(3) 相图
-
O(3) - 非线性Sigma model
-
- O(3)自由场 - 归一化序参量：
F = 1 2 ρ ∫ d d x ( ∇ m ⃗ ) 2 Z = ∫ D [ m ⃗ ] δ ( m 2 − 1 ) \begin{aligned} & F = \frac{1}{2\rho} \int d^d x (\nabla \vec m)^2 \\ & Z = \int D[\vec m] ~ \delta( m^2 - 1 ) \end{aligned} F=2ρ1∫ddx(∇m)2Z=∫D[m] δ(m2−1)
因此，对 m -3 分量矢量函数具有归一化约束，这令整个自由场理论呈现出非线性
归约成2个变量，采用柱坐标，限定 r = 1：耦合参数 g，设定时空最小截断尺度a
∇ m x , y = ∇ ( 1 − m z 2 c o s ( θ ) ) , ∇ ( 1 − m z 2 s i n ( θ ) ) → ( ∇ m ) 2 = ( ∇ m z ) 2 + ( 1 − m z 2 ) ( ∇ θ ) 2 + ( m z ∇ m z ) 2 1 − m z 2 → m z = ( g a d − 2 ) 1 2 ϕ ; g = ρ a 2 − d < < 1 → F = ∫ d d x ( 1 2 ( ∇ ϕ ) 2 + 1 − g a d − 2 ϕ 2 2 g a d − 2 ( ∇ θ ) 2 + g a d − 2 2 ( 1 − g a d − 2 ϕ 2 ) ( ϕ ∇ ϕ ) 2 ) = ∫ d d x ( 1 2 ( ∇ ϕ ) 2 + 1 2 g d − 2 ( ∇ θ ) 2 − 1 2 ( ϕ ∇ θ ) 2 ) \begin{aligned} & \nabla m_{x,y} = \nabla (\sqrt {1-m^2_z} cos(\theta)), \nabla (\sqrt {1-m^2_z} sin(\theta)) \\ ~~\rightarrow ~~ & (\nabla m)^2 = (\nabla m_z)^2 + (1-m^2_z)(\nabla \theta)^2 + \frac{(m_z \nabla m_z)^2}{1-m^2_z} \\ \rightarrow ~~ & ~ m_z = (g a^{d-2})^{\frac{1}{2}} \phi ~ ; g = \rho a^{2-d} << 1 \\ \rightarrow ~~ & ~ F = \int d^dx (\frac{1}{2}(\nabla\phi)^2 + \frac{1-ga^{d-2}\phi^2}{2ga^{d-2}} (\nabla\theta)^2 + \frac{ga^{d-2}}{2(1-ga^{d-2}\phi^2)} (\phi \nabla \phi)^2 )\\ & ~~= \int d^dx (\frac{1}{2}(\nabla\phi)^2 + \frac{1}{2g^{d-2}} (\nabla \theta)^2 -\frac{1}{2}(\phi\nabla\theta)^2) \end{aligned} → → → ∇mx,y=∇(1−mz2cos(θ)),∇(1−mz2sin(θ))(∇m)2=(∇mz)2+(1−mz2)(∇θ)2+1−mz2(mz∇mz)2 mz=(gad−2)21ϕ ;g=ρa2−d<<1 F=∫ddx(21(∇ϕ)2+2gad−21−gad−2ϕ2(∇θ)2+2(1−gad−2ϕ2)gad−2(ϕ∇ϕ)2) =∫ddx(21(∇ϕ)2+2gd−21(∇θ)2−21(ϕ∇θ)2)
定义配分函数，并进行近似（ \epsilon = 2+d ）
Z = ∫ D [ ϕ ] D [ θ ] e − F = ∫ D [ θ ] e − ∫ d d x 1 2 g a d − 2 ( ∇ θ ) 2 ∫ D [ ϕ ] e − ∫ d d x ( 1 2 ( ∇ ϕ ) 2 − 1 2 ϕ 2 ( ∇ θ ) 2 ) = ∫ D [ θ ] e − ∫ d d x 1 2 g a d − 2 ( ∇ θ ) 2 e 1 2 ∑ k ( ln ( 2 π k 2 − ln ( 1 − ( ∇ θ ) ∣ k = 0 k 2 ) ) ) \begin{aligned} Z = & \int D[\phi]D[\theta]e^{-F} \\ = & \int D[\theta]e^{-\int d^dx \frac{1}{2ga^{d-2}} (\nabla \theta)^2} \int D[\phi] e^{-\int d^dx (\frac{1}{2}(\nabla \phi)^2 - \frac{1}{2}\phi^2 (\nabla \theta)^2)}\\ = & \int D[\theta]e^{-\int d^dx \frac{1}{2ga^{d-2}} (\nabla \theta)^2} e^{\frac{1}{2}\sum_k(\ln(\frac{2\pi}{k^2} - \ln(1-\frac{(\nabla \theta)|_{k=0}}{k^2})) )} \end{aligned} Z===∫D[ϕ]D[θ]e−F∫D[θ]e−∫ddx2gad−21(∇θ)2∫D[ϕ]e−∫ddx(21(∇ϕ)2−21ϕ2(∇θ)2)∫D[θ]e−∫ddx2gad−21(∇θ)2e21∑k(ln(k22π−ln(1−k2(∇θ)∣k=0)))
仍然用连续极限，代入Z-指数项，删掉自由项积分，近似 ∇θ = const
∑ k ( . . . ) = 1 ( 2 π ) 2 ∫ d d k ( . . . ) ; 1 2 ln ( 1 − ( ∇ θ ) 2 ∣ k = 0 k 2 ) ≈ − 1 2 ( ∇ θ ) 2 ∣ k = 0 k 2 → 1 ( 2 π ) d ∫ d d k ( − 1 2 ( ∇ θ ) 2 ∣ k = 0 k 2 ) = − 1 ( 2 π ) d ( ∇ θ ) 2 ∣ k = 0 2 k 2 ∫ Λ / b Λ d d k k 2 = − 1 ( 2 π ) d ( ∇ θ ) 2 ∣ k = 0 2 k 2 S d Λ d − 2 ( d − 2 ) ( 1 − b d − 2 ) = − 1 ( 2 π ) d 1 2 k 2 S d Λ d − 2 ( d − 2 ) ( 1 − b d − 2 ) ∫ d d x ( ∇ θ ) 2 \begin{aligned} & \sum_k (...) = \frac{1}{(2\pi)^2}\int d^dk(...) ~;~~ \frac{1}{2} \ln(1-\frac{(\nabla\theta)^2|_{k=0}}{k^2})\approx -\frac{1}{2}\frac{(\nabla\theta)^2|_{k=0}}{k^2} \\ & \rightarrow \frac{1}{(2\pi)^d} \int d^dk (-\frac{1}{2}\frac{(\nabla\theta)^2|_{k=0}}{k^2})\\ & = - \frac{1}{(2\pi)^d} \frac{(\nabla\theta)^2|_{k=0}}{2k^2} \int_{\Lambda/b}^{\Lambda} \frac{d^dk}{k^2} \\ & = - \frac{1}{(2\pi)^d} \frac{(\nabla\theta)^2|_{k=0}}{2k^2} \frac{S_d\Lambda^{d-2}}{(d-2)}(1-b^{d-2}) \\ & = - \frac{1}{(2\pi)^d} \frac{1}{2k^2} \frac{S_d\Lambda^{d-2}}{(d-2)}(1-b^{d-2}) \int d^dx (\nabla\theta)^2 \end{aligned} k∑(...)=(2π)21∫ddk(...) ; 21ln(1−k2(∇θ)2∣k=0)≈−21k2(∇θ)2∣k=0→(2π)d1∫ddk(−21k2(∇θ)2∣k=0)=−(2π)d12k2(∇θ)2∣k=0∫Λ/bΛk2ddk=−(2π)d12k2(∇θ)2∣k=0(d−2)SdΛd−2(1−bd−2)=−(2π)d12k21(d−2)SdΛd−2(1−bd−2)∫ddx(∇θ)2
归纳为 θ 变量的连续极限场论，ϕ 自由度被率先积掉 (Intergrate out)
Z = ∫ D [ θ ] e − ∫ d d x 1 2 g a d − 2 ( ∇ θ ) 2 + ( − 1 2 ) ( S d Λ d − 2 ( 2 π ) d ( 1 − b 2 − d ) ∫ d d x ( ∇ θ ) 2 Z = \int D[θ] e^ {-\int d^dx \frac{1}{2ga^{d-2}}(\nabla\theta)^2 + (-\frac{1}{2})(\frac{S_d \Lambda^{d-2}}{(2\pi)^d}(1-b^{2-d})\int d^dx(\nabla\theta)^2 } Z=∫D[θ]e−∫ddx2gad−21(∇θ)2+(−21)((2π)dSdΛd−2(1−b2−d)∫ddx(∇θ)2
因此，归纳为对 θ 变量积分，并进行修正
Z = ∫ D [ θ ] e − ∫ d d x 1 2 g ′ a ′ d − 2 ( ∇ θ ) 2 1 g ′ a ′ d − 2 = 1 g a d − 2 − S d Λ d − 2 ( 2 π ) d ( d − 2 ) ( 1 − b 2 − d ) \begin{aligned} & Z = \int D[\theta] e^{-\int d^dx \frac{1}{2g'a'^{d-2}}(\nabla\theta)^2} \\ & \frac{1}{g'a'^{d-2}} = \frac{1}{ga^{d-2}} - \frac{S_d\Lambda^{d-2}}{(2\pi)^d(d-2)}(1-b^{2-d}) \\ \end{aligned} Z=∫D[θ]e−∫ddx2g′a′d−21(∇θ)2g′a′d−21=gad−21−(2π)d(d−2)SdΛd−2(1−b2−d)
- O(3)自由场 - 归一化序参量：
-
- 耦合参数 - RG Equation
β ( g ) = ( 2 − d ) g + S d ( 2 π ) d g 2 g ∗ = 0 → β ( g ) ∣ g ∗ = 0 ≈ ( 2 − d ) g g ∗ = ( 2 π ) d S d ( 2 − d ) → β ( g ) ∣ g ∗ = g 0 ≈ ( d − 2 ) ( g − g 0 ) \begin{aligned} & \beta(g) = (2-d)g + \frac{S_d}{(2\pi)^d}g^2 \\ & g^* = 0 ~ \rightarrow \beta(g)|_{g*=0} \approx (2-d)g \\ & g^* = \frac{(2\pi)^d}{S_d}(2-d) ~ \rightarrow \beta(g)|_{g* = g_0} \approx (d-2)(g-g_0) \end{aligned} β(g)=(2−d)g+(2π)dSdg2g∗=0 →β(g)∣g∗=0≈(2−d)gg∗=Sd(2π)d(2−d) →β(g)∣g∗=g0≈(d−2)(g−g0)
d > 2：g = 0 稳定不动点（有序态，∇θ<<1），g = g0 不稳定不动点 （相变，\nu = 1/2-d）
d < 2：g = 0 不稳定不动点（相变，\nu = 1/d-2），g = g0 稳定不动点（非物理，< 0 ）
- 耦合参数 - RG Equation
-
- 渐近自由 ~ d = 2
d g d ln b = g 2 2 π > 0 → ∫ g 0 g d g g 2 = 1 2 π ∫ 0 ln b d ln b → 1 g 0 − 1 g = 1 2 π ln ( b ) \begin{aligned} & \frac{dg}{d \ln b} = \frac{g^2}{2\pi}>0 \\ \rightarrow & \int ^ g_{g_0} \frac{dg}{g^2} = \frac{1}{2\pi} \int^{\ln b}_0 d\ln b \\ \rightarrow& \frac{1}{g_0} - \frac{1}{g} = \frac{1}{2\pi} \ln(b) \\ \end{aligned} →→dlnbdg=2πg2>0∫g0gg2dg=2π1∫0lnbdlnbg01−g1=2π1ln(b)
这意味着随着b增大，g流向强耦合：夸克禁闭同 Kondo 问题
`
- 渐近自由 ~ d = 2
-
XY model - KT 相变
-
- XY model - Fxy 的连续极限
定义 S 在平面内旋转，因此有分量 Sx = Scos(θ) ,Sy = Ssin(θ)
H X Y = ∑ < j , k > J j , k S ⃗ j ‘ S ⃗ k S ⃗ j ‘ S ⃗ k = S ( c o s ( θ j ) , s i n ( θ j ) ) ‘ S ( c o s ( θ k ) , s i n ( θ k ) ) = S 2 ( c o s ( θ i − θ j ) ) \begin{aligned} & H_{XY} = \sum_{<j,k>}J_{j,k} \vec S_{j}`\vec S_{k} \\ & \vec S_j ` \vec S_k = S(cos(\theta_j),sin(\theta_j)) ~`S(cos(\theta_k),sin(\theta_k)) \\ &~~~~~ = S^2 (cos(\theta_i - \theta_j)) \end{aligned} HXY=<j,k>∑Jj,kSj‘SkSj‘Sk=S(cos(θj),sin(θj)) ‘S(cos(θk),sin(θk)) =S2(cos(θi−θj))
考虑最近邻作用，并丢掉常数项，取连续极限，鞍点展开，有
H X Y = ∑ < j , k > ( − J ) c o s ( θ j − θ k ) = ∑ j ∑ δ = x ^ , y ^ ( − J ) c o s ( θ j − θ j + δ ) = ∑ j ( − J ) ( c o s ( θ j − θ j + x ^ ) + c o s ( θ j − θ j + j ^ ) ) = ∑ j ( − J ) ∑ δ = x ^ , y ^ ( 1 − 1 2 ( θ j − θ j + δ ) 2 ) = ∑ j ( − J ) ( 2 − 1 2 ( θ j − θ j + x ^ ) 2 − 1 2 ( θ j − θ j + y ^ ) 2 ) \begin{aligned} & H_{XY} = \sum_{<j,k>}(-J) ~ cos(\theta_j -\theta_k) \\ & = \sum_j \sum_{\delta = \hat x , \hat y }(-J)~cos(\theta_j - \theta_{j+\delta}) \\ & = \sum_j (-J) (cos(\theta_j - \theta_{j+\hat x}) + cos(\theta_j - \theta_{j+\hat j})) \\ & = \sum_j (-J)\sum_{\delta = \hat x,\hat y} (1 -\frac{1}{2}(\theta_j -\theta_{j+\delta})^2) \\ & = \sum_j (-J) (2-\frac{1}{2}(\theta_j -\theta_{j+\hat x})^2 -\frac{1}{2}(\theta_j -\theta_{j+\hat y})^2 ) \end{aligned} HXY=<j,k>∑(−J) cos(θj−θk)=j∑δ=x^,y^∑(−J) cos(θj−θj+δ)=j∑(−J)(cos(θj−θj+x^)+cos(θj−θj+j^))=j∑(−J)δ=x^,y^∑(1−21(θj−θj+δ)2)=j∑(−J)(2−21(θj−θj+x^)2−21(θj−θj+y^)2)
以下是取连续极限的符号规则，最终以Lattice为对象写出配分函数
1 a 2 ∑ j ∑ k → 1 a 2 ∫ d 2 r ( θ j − θ j + x ^ ) 2 → a 2 ( ∂ x θ ) 2 F X Y = − K 2 ∫ d 2 r ( ( ∂ x θ ) 2 + ( ∂ y θ ) 2 ) θ ( r ⃗ ) = ∫ d 2 k ( 2 π ) 2 e i k r θ k Z X Y = ∏ r ∫ d θ r e − F X Y = ∏ k > 0 ∫ d θ k ∗ d θ k e − K k 2 ∣ θ k ∣ 2 \begin{aligned} & \frac{1}{a^2} \sum_j \sum_k ~ \rightarrow \frac{1}{a^2} \int d^2r \\ & (\theta_j - \theta_{j+\hat x})^2 \rightarrow a^2(\partial_x\theta)^2 \\ & F_{XY} = -\frac{K}{2} \int d^2r ((\partial_x \theta)^2+(\partial_y \theta)^2) \\ & \theta(\vec r) = \int \frac{d^2k}{(2\pi)^2} e^{ikr} \theta_k \\ & Z_{XY} = \prod_r \int d\theta_r e^{-F_{XY}} = \prod_{k>0} \int d\theta^*_k d\theta_k e^{-K k^2|\theta_k|^2} \end{aligned} a21j∑k∑ →a21∫d2r(θj−θj+x^)2→a2(∂xθ)2FXY=−2K∫d2r((∂xθ)2+(∂yθ)2)θ(r)=∫(2π)2d2keikrθkZXY=r∏∫dθre−FXY=k>0∏∫dθk∗dθke−Kk2∣θk∣2
在高温与低温展开下，关联函数有呈现从指数律衰减到幂律衰减的相变过程
T / J > > 1 : e − ∣ r i − r j ∣ ln ( T J ) T / J < < 1 : 1 ∣ r i − r j ∣ T 2 π J \begin{aligned} & T/J>>1: e^{-|r_i-r_j| \ln(\frac{T}{J})} \\ & T/J<<1: \frac{1}{|r_i - r_j|^{\frac{T}{2\pi J}}}\\ \end{aligned} T/J>>1:e−∣ri−rj∣ln(JT)T/J<<1:∣ri−rj∣2πJT1
这种相变被称为，K-T 相变，是涡旋导致的相变，它令低温长程关联失效。
- XY model - Fxy 的连续极限
-
- Vortex to S-G model
我们定义涡旋场由若干涡荷为权重的单位涡度的叠加如：
θ ( r ) = ∑ j θ j = ∑ j q j arctan ( y − y j x − x j ) \theta(r) = \sum_j \theta_j = \sum_j q_j \arctan(\frac{y-y_j}{x-x_j}) θ(r)=j∑θj=j∑qjarctan(x−xjy−yj)
显然，该叠加符合 Laplace - Equation，以及作为涡旋的物理量：
I = ∮ d l ⃗ ∇ ( θ ( r ⃗ ) ) = 2 π E v = ∫ d 2 r J 2 ( ∇ ( θ ( r ⃗ ) ) ) 2 = J π ln L R S v = k ln ( Ω ) = 2 k ln ( L R ) F V = E V − T S V = ( J π − 2 k ) ln ( L R ) \begin{aligned} & I = \oint d \vec l~ \nabla(\theta(\vec r)) = 2\pi \\ & E_v = \int d^2 r ~ \frac{J}{2} (\nabla(\theta(\vec r)))^2 = J \pi \ln{\frac{L}{R}} \\ & S_v = k \ln(\Omega) = 2k\ln(\frac{L}{R}) \\ & F_V = E_V - TS_V = (J \pi-2k) \ln(\frac{L}{R}) \end{aligned} I=∮dl ∇(θ(r))=2πEv=∫d2r 2J(∇(θ(r)))2=JπlnRLSv=kln(Ω)=2kln(RL)FV=EV−TSV=(Jπ−2k)ln(RL)
可以计算出，T ~ T_KT 存在一个相变温度，当 T>T_KT，F<0，会倾向于形成单涡旋
这将在下面导出与 Lattice-Spin 对偶的 Vortex-Charge 模型：S-G model，用于解释涡旋致高温态
- Vortex to S-G model
-
- S-G. Scaling
对XY-model进行 Villian 近似与 Hubbard Stractonovich 变换，作为描述 XY model 的流方程
S-G 模型 取连续极限 可以写成：
Z S G = ∫ D [ ϕ ( x ⃗ ) ] e − F S G F S G = ∫ d x ⃗ ( 1 2 K ( ∇ ϕ ( x ⃗ ) ) 2 − 2 y a − 2 cos ( 2 π ϕ ( x ⃗ ) ) ) \begin{aligned} & Z_{SG} = \int D[\phi(\vec x)] e^{-F_{SG}} \\ & F_{SG} = \int d\vec x ~(\frac{1}{2K}(\nabla \phi(\vec x))^2 - 2ya^{-2} \cos(2\pi \phi(\vec x))) \end{aligned} ZSG=∫D[ϕ(x)]e−FSGFSG=∫dx (2K1(∇ϕ(x))2−2ya−2cos(2πϕ(x)))
在鞍点展开，并在动量空间上观察到能隙：
= ∫ d x ⃗ ( 1 2 K ( ∇ δ ϕ ( x ⃗ ) ) 2 + 2 y a − 2 ( 2 π ) 2 1 2 ( δ ϕ ( x ⃗ ) ) 2 ) = ∫ d k ⃗ ( 1 2 K k 2 + 4 π a − 2 y ) ∣ δ ϕ k ∣ 2 \begin{aligned} & = \int d \vec x ~(\frac{1}{2K}(\nabla \delta \phi(\vec x))^2 + 2ya^{-2} (2\pi)^2 \frac{1}{2}(\delta \phi(\vec x))^2) \\ & = \int d \vec k ~(\frac{1}{2K}k^2 + 4\pi a^{-2}y) ~|\delta \phi_k|^2 \end{aligned} =∫dx (2K1(∇δϕ(x))2+2ya−2(2π)221(δϕ(x))2)=∫dk (2K1k2+4πa−2y) ∣δϕk∣2
划分完高低能，对模型重标，积掉ϕ(x) >：
cos(ϕ(x)< + ϕ(x) >)，出现高低能混杂的情形，这里需要含K的积分对y参数进行修正
F S G = ∫ d x ⃗ ( 1 2 K ( ∇ ϕ < ) 2 + 1 2 K ( ∇ ϕ > ) 2 − 2 y ( cos ( 2 π ϕ < ) cos ( 2 π ϕ > ) − sin ( 2 π ϕ < ) sin ( 2 π ϕ > ) ) ) Z = ∫ D [ ϕ < ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ < ) 2 ∫ D [ ϕ > ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ > ) 2 e ∫ d x ⃗ 2 y cos ( 2 π ϕ < ) cos ( 2 π ϕ > ) − sin ( 2 π ϕ < ) sin ( 2 π ϕ > ) \begin{aligned} & F_{SG} = \int d\vec x (\frac{1}{2K}(\nabla \phi_<)^2+\frac{1}{2K}(\nabla \phi_>)^2 -2y (\cos(2\pi\phi_<)\cos (2\pi\phi_>) -\sin(2\pi\phi_<) \sin(2\pi\phi_>))) \\ & Z = \int D[\phi_<] e^{-\int d \vec x \frac{1}{2K} (\nabla \phi_<)^2 } \int D[\phi_>]e^{-\int d \vec x \frac{1}{2K} (\nabla \phi_>)^2}e^{\int d\vec x~ 2y \cos(2\pi\phi_<)\cos (2\pi\phi_>) -\sin(2\pi\phi_<) \sin(2\pi\phi_>) } \end{aligned} FSG=∫dx(2K1(∇ϕ<)2+2K1(∇ϕ>)2−2y(cos(2πϕ<)cos(2πϕ>)−sin(2πϕ<)sin(2πϕ>)))Z=∫D[ϕ<]e−∫dx2K1(∇ϕ<)2∫D[ϕ>]e−∫dx2K1(∇ϕ>)2e∫dx 2ycos(2πϕ<)cos(2πϕ>)−sin(2πϕ<)sin(2πϕ>)
经典操作了，展开一阶项，用期望值取出混杂的高能部分以表示积掉【Intergrate out】，
好处就是，能够留出高能标自由场的积分独立表示，直接扯回到低能积分式的表达里
Z ≈ ∫ D [ ϕ < ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ < ) 2 ∫ D [ ϕ > ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ > ) 2 ( 1 + 2 y ∫ d x ⃗ cos ( 2 π ϕ < ) ⟨ cos ( 2 π ϕ > ) ⟩ 0 − sin ( 2 π ϕ < ) ⟨ sin ( 2 π ϕ > ) ⟩ 0 ) = ∫ D [ ϕ < ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ < ) 2 ∫ D [ ϕ > ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ > ) 2 ( 1 + 2 y ∫ d x ⃗ cos ( 2 π ϕ < ) ⟨ cos ( 2 π ϕ > ) ⟩ 0 ) = ∫ D [ ϕ < ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ < ) 2 ∫ D [ ϕ > ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ > ) 2 ( 1 + 2 y ∫ d x ⃗ cos ( 2 π ϕ < ) ⟨ 1 2 ( e i 2 π ϕ > ) + ( e − i 2 π ϕ > ) ⟩ 0 ) \begin{aligned} & Z \approx \int D[\phi_<] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_<)^2} \int D [\phi_>] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_>)^2}(1 + 2y \int d \vec x \cos(2\pi\phi_<) \left \langle \cos (2\pi\phi_>)\right \rangle_0 - \sin(2\pi\phi_<) \left \langle \sin(2\pi\phi_>) \right\rangle_0 ) \\ & = \int D[\phi_<] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_<)^2} \int D [\phi_>] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_>)^2} (1+2y \int d \vec x \cos(2\pi\phi_<) \left \langle \cos (2\pi\phi_>)\right \rangle_0 ) \\ & = \int D[\phi_<] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_<)^2} \int D [\phi_>] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_>)^2} (1+ 2y \int d \vec x \cos(2\pi\phi_<) \left \langle \frac{1}{2} (e^{i2\pi\phi_>}) +(e^{-i2\pi\phi_>}) \right \rangle_0 ) \end{aligned} Z≈∫D[ϕ<]e−∫dx2K1(∇ϕ<)2∫D[ϕ>]e−∫dx2K1(∇ϕ>)2(1+2y∫dxcos(2πϕ<)⟨cos(2πϕ>)⟩0−sin(2πϕ<)⟨sin(2πϕ>)⟩0)=∫D[ϕ<]e−∫dx2K1(∇ϕ<)2∫D[ϕ>]e−∫dx2K1(∇ϕ>)2(1+2y∫dxcos(2πϕ<)⟨cos(2πϕ>)⟩0)=∫D[ϕ<]e−∫dx2K1(∇ϕ<)2∫D[ϕ>]e−∫dx2K1(∇ϕ>)2(1+2y∫dxcos(2πϕ<)⟨21(ei2πϕ>)+(e−i2πϕ>)⟩0)
注：动量关联到位型关联
⟨ ϕ ( k ) ϕ ( − k ) ⟩ = K k 2 ⟨ ϕ ( x ) ϕ ( x ) ⟩ = ∫ d 2 k ( 2 π ) 2 K k 2 = ∫ Λ / b Λ d k 2 π K k = K 2 π ln b \begin{aligned} & \left \langle \phi(k)\phi(-k) \right \rangle = \frac{K}{k^2} \\ & \left \langle \phi(x)\phi(x) \right \rangle = \int \frac{d^2k}{(2\pi)^2} \frac{K}{k^2} = \int^{\Lambda}_{\Lambda/b} \frac{dk}{2\pi}\frac{K}{k} = \frac{K}{2\pi}\ln b \end{aligned} ⟨ϕ(k)ϕ(−k)⟩=k2K⟨ϕ(x)ϕ(x)⟩=∫(2π)2d2kk2K=∫Λ/bΛ2πdkkK=2πKlnb
作为 exp 展开项当中的第一项 k-space 积分修正，映射回指数，并对照地修正y参数
⟨ e − i 2 π ϕ > ⟩ 0 + ⟨ e i 2 π ϕ > ⟩ 0 = e − 1 2 ( 2 π ) 2 ⟨ ϕ > 2 ⟩ = e − 1 2 ( 2 π ) 2 ( K 2 π ln b ) = b − π K Z = ∫ D [ ϕ < ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ < ) 2 e ∫ d x ⃗ 2 y b − π K cos ( 2 π ϕ < ) ∫ D [ ϕ > ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ > ) 2 Z = ∫ D [ ϕ < ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ < ) 2 − 2 y b − π K cos ( 2 π ϕ < ) ∫ D [ ϕ > ] e − ∫ d x ⃗ 1 2 K ( ∇ ϕ > ) 2 \begin{aligned} & \left \langle e^{-i2\pi\phi_>} \right \rangle_0 + \left \langle e^{i2\pi\phi_>} \right \rangle_0 = e^{-\frac{1}{2}(2\pi)^2 \left \langle \phi^2_> \right\rangle } = e^{-\frac{1}{2}(2\pi)^2 (\frac{K}{2\pi} \ln b)} = b^{ - \pi K} \\ & Z = \int D[\phi_<] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_<)^2} e^{\int d \vec x ~2yb^{-\pi K} \cos(2\pi \phi_<) } \int D [\phi_>] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_>)^2} \\ & Z = \int D[\phi_<] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_<)^2 -{ ~2yb^{-\pi K} \cos(2\pi \phi_<) } } \int D [\phi_>] e^{ - \int d \vec x\frac{1}{2K} (\nabla \phi_>)^2} \end{aligned} ⟨e−i2πϕ>⟩0+⟨ei2πϕ>⟩0=e−21(2π)2⟨ϕ>2⟩=e−21(2π)2(2πKlnb)=b−πKZ=∫D[ϕ<]e−∫dx2K1(∇ϕ<)2e∫dx 2yb−πKcos(2πϕ<)∫D[ϕ>]e−∫dx2K1(∇ϕ>)2Z=∫D[ϕ<]e−∫dx2K1(∇ϕ<)2− 2yb−πKcos(2πϕ<)∫D[ϕ>]e−∫dx2K1(∇ϕ>)2 作为对y参数的高能修正：与原位型空间的 F_SG 进行对照
F S G = ∫ d x ⃗ ( 1 2 K ( ∇ ϕ ( x ⃗ ) ) 2 − 2 y cos ( 2 π ϕ ( x ⃗ ) ) ) F G L ′ = ∫ d x ⃗ ( 1 2 K ( ∇ ϕ ( x ⃗ ) < ) 2 − 2 y b − π K cos ( 2 π ϕ ( x ⃗ ) < ) \begin{aligned} & F_{SG} = \int d\vec x ~(\frac{1}{2K}(\nabla \phi(\vec x))^2 - 2y \cos(2\pi \phi(\vec x))) \\ & F'_{GL} = \int d\vec x ~(\frac{1}{2K}(\nabla \phi(\vec x)_<)^2 -{ ~2yb^{-\pi K} \cos(2\pi \phi(\vec x)_<) } \\ \end{aligned} FSG=∫dx (2K1(∇ϕ(x))2−2ycos(2πϕ(x)))FGL′=∫dx (2K1(∇ϕ(x)<)2− 2yb−πKcos(2πϕ(x)<)
展开二阶项，可以获得对 K 的修正，暂时不想写那么多，找机会再更新完吧
- S-G. Scaling
-
- S-G. Flow
因此获得 XY-RG Flow Equation
d K d ln b = − y ^ 2 2 K 3 d y ^ d ln b = ( 2 − π K ) y ^ y ^ = ( 4 π ) 2 Λ 4 y \begin{aligned} & \frac{d K}{d \ln b} = -\frac{\hat y^2}{2} K^3 \\ & \frac{d \hat y} {d \ln b} = (2-\pi K)\hat y \\ & ~ \hat y = \frac{(4\pi)^2}{\Lambda^4} y \\ \end{aligned} dlnbdK=−2y^2K3dlnbdy^=(2−πK)y^ y^=Λ4(4π)2y
- S-G. Flow
扫一扫
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
