SU(2) Pure YM on R^4: Derivation and Solutions

  • Thread starter Thread starter Charles_Henry
  • Start date Start date
  • Tags Tags
    Pure Su(2)
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
4 replies · 3K views
Charles_Henry
Messages
14
Reaction score
0

Homework Statement



Derive the pure SU(2) YM theory on [itex]\mathbb{R}^4[/itex] from the action. Let [itex]A_{\mu} (x)[/itex] be a solution to these equations. Show:

[itex]\tilde{A}_{\mu} (cx)[/itex] is also a solution (with the same action).Background

The Euclidean YM action

[itex]\mathbb{S} = - \int_{\mathbb{R^4}} Tr (F \wedge \ast F)[/itex]

yields

[itex]D \ast F = 0[/itex]

Let [itex]\ast: \wedge^{p} \rightarrow \wedge^{D+1-p}[/itex] be a linear map, such that

[itex]\ast (dx^{\mu_{1}} \wedge ... \wedge dx^{\mu_{p}} = \frac{\sqrt{|det(n)!|}}{(D+1-p)!} \epsilon^{\mu_{1}...\mu_{p}}_{{\mu}_{p+1}...{\mu}_{D+1}} dx^{\mu}_{p+1} \wedge ... \wedge dx^{{\mu}_{p+1}}[/itex]

if [itex]G = SU(2)[/itex] we choose a basis [itex]T_{a}, a = 1, 2, 3[/itex] for an Anti-Hermitian 2 x 2 matrix

[itex]T_{a}, T_{b}[/itex] = [itex]- \epsilon_{abc} T_{c}, T_{a} = \frac{1}{2} i \sigma_{a}[/itex]

where [itex]Tr(T_{a}, T_{b}) = - \frac {1}{2} \delta_{ab}[/itex]

where [itex]\sigma_{a}[/itex] are pauli matrices, and a general group element [itex]g = exp (\alpha^{a} T_{a} )[/itex] with [itex]\alpha^a[/itex] real. Whence,

[itex](D_{\mu} \phi)^{a} = \partial_{\mu} \phi^{a} - \epsilon^{abc} A^{b}_{\mu} \phi^{c}[/itex] and [itex]F^{a}_{{\mu} v} = \partial_{\mu} A^{a}_{v} - \partial_{v}A^{a}_{\mu} - \epsilon^{abc}A^{b}_{\mu}A^{c}_{v}[/itex]

when [itex]D+1=4[/itex] is a gauge theory in Minkowski space [itex]M[/itex], and [itex]A[/itex] is the gauge potential,

[itex]( \ast F)_{{\mu}v} = \frac{1}{2} \epsilon_{{\mu}v{\alpha}{\beta}}F^{{\alpha}{\beta}}[/itex]

A two form [itex]F= \frac{1}{2} F_{{\mu}v}dx^{\mu} \wedge dx^{v}[/itex] is self dual or ASD when [itex]\ast F = F[/itex] and [itex]\ast F = - F[/itex] respectively

[itex]-Tr ( F \wedge \ast F) = - \frac{1}{2} Tr (F_{{\mu}v}F^{{\mu}v}) d^{4}x = \frac{1}{4}F^{a}_{{\mu}v}F^{{\mu}va} d^{a}x[/itex]

[itex]d^{4}x = \frac{1}{24} \epsilon_{{\mu}v{\alpha}{\beta}}dx^{\mu} \wedge dx^{v} \wedge dx^{\alpha} \wedge dx^{\beta}[/itex]

with identities

[itex]\epsilon_{{\mu}v{\alpha}{\beta}}dx^{4} = - dx^{4} \wedge dx^{v} \wedge dx^{\alpha} \wedge dx^{\beta}[/itex]

and [itex]\epsilon_{{\alpha}{\beta}{\rho}{\sigma}} \epsilon^{{\mu}{v}{\rho}{\sigma}} = - 4 (\delta^{\mu}_{\alpha} \delta^{v}_{\beta})[/itex]

Instantons are non-singular solutions of classical equations of motion in Euclidean space whose Action is finite.

[itex]F_{{\mu}v} (x)[/itex] ~ [itex]O (\frac{1}{r^3})[/itex]

[itex]A_{\mu}[/itex] ~ [itex]\partial_{\mu} gg^{-1} + O \frac{1}{r^2}[/itex] as [itex]r \rightarrow \infty[/itex]

note: I understand that the gauge transformations g(x) needs to be defined only asymptotically, so [itex]g: \mathbb{S}^{3}_{\infty} \rightarrow SU(2)[/itex] is extended to [itex]\mathbb{R}^4[/itex] if its degree vanishes:

For example:

If [itex]M_{1}[/itex] and [itex]M_{2}[/itex] are oriented, compact, D-dimensional manifolds without boundary, and [itex]w[/itex] is a volume-form on [itex]M_{2}[/itex]. where [itex]deg (f)[/itex] of a smooth map [itex]f: M_{1} \rightarrow M_{2}[/itex] is given by

[itex]\int_{M_{1}} f\ast w = [ deg(f) ] \int_{M_{2}} w[/itex]

let [itex]y \in M_{2}[/itex] when [itex]f^{-1}(y) = {x; f(x) = y}[/itex] is finite, and the Jacobian [itex]J(f)[/itex] is not zero (if [itex]x \in U[/itex] with local coordinates [itex]x^{i}[/itex] and [itex]y \in f(u)[/itex] with local co-ordinates [itex]y^{i}[/itex], then we can assume:

[itex]\mathbb{J} = det \frac{\partial y^{i}}{\partial x^{J}}[/itex] if [itex]y^{i} (x^{1}, ... x^{D})[/itex]

deg (f) is an integer given by

[itex]deg (f) = \Sigma_{x \in f^{-1} (y)} sign [ \mathbb{J} (x) ][/itex]

(proof withheld)

therefore:

[itex]f: X \rightarrow SU(2) = S^{3}[/itex] where X is closed.

[itex]deg(f) = \frac{1}{24\pi^2} \int_{X} Tr[(f^{-1} df)^3][/itex]

the boundary conditions are understood in terms of one-point compactifications [itex]S^4 = \mathbb{R}^4 \cup {\infty}[/itex] which has a conformally equivalent metric to that of a flat metric in [itex]\mathbb{R}^4[/itex]

A solution of YM equations on [itex]S^4[/itex] project stereographically to a connection on [itex]\mathbb{R}^4[/itex] with a curvature which vanishes at infinity.

Scaling Argument:

A Field(s) [itex](A, \phi)[/itex] given by a potential one-form and a scalar higgs-field:

[itex]E = \int_{\mathbb{R}} d^{D}x [|F|^{2} + |D \phi |^{2} + U(\phi)[/itex]

[itex]= E_{F} + E_{D_{\phi}} + E_{U}[/itex]

if [itex]A(x)[/itex] and [itex]\phi (x)[/itex] are critical points:

[itex]\phi_{c} (x) = \phi (cx)[/itex]

[itex]A_{c} (x) = cA(cx)[/itex]

[itex]F_{c} = C^{2} F(cx)[/itex]

[itex]D_{c} \phi_{c} = c D\phi (cx)[/itex]

which leads to

[itex]E_{(c)} = \frac{1}{C^{D-4}}E_{F} + \frac{1}{C^{D+2}}E_{D_{\phi}} + \frac{1}{C^{D}}E_{U}[/itex]

[itex](D-4) E_{F} + (D-2) E_{D_{\phi}} + DE_{U} = 0[/itex]

note: I believe I am looking for a solution where [itex]E_{D_{\phi}} = E_{U} = 0[/itex] in D=4

The Attempt at a Solution



A YM action S within a given topological sector

[itex]c_{2} = \frac{1}{8 \pi^2} \int_{\mathbb{R}} Tr( F \wedge F) > 0[/itex]

bounded from below by [itex]8\pi^2c_{2}[/itex]

[itex]F \wedge F = \ast F \wedge \ast F[/itex]

[itex]\mathbb{S} = - \frac{1}{2} \int_{\mathbb{R}^4} Tr[(F + \ast F) \wedge (F + \ast F)] + \int_{\mathbb{R}^4} Tr (F \wedge F) = - \frac{1}{2} \int_{\mathbb{R}^4} Tr [(F + \ast F) \wedge \ast (F + \ast F) + 8 \pi^2c_{2} \geq 8 \pi^2c_{2}[/itex]

when [itex]F = - \ast F[/itex] hold

some bib: Atiyah, M.F and Ward, R.S (1977) Instantons and Algebraic Geometry, Commun. Math. Phy. 55, 117-124

Sacks, L. and uhlenbeck, K (1981) The existence of minimal immersions of 2-spheres, Ann. Math 113, 1-24
 
Last edited:
Physics news on Phys.org
nonsense
 
Last edited:
more nonsense.
 
Last edited:
ok, i see it now...derive the euler-lagrange equations from the action that leads to gauge potential. Assume we could derive a solution of pure YM IN R^4 from the Vector Potential by defining invariance along any coordinate of our choosing.
 
Last edited:
This is meant to be much simpler than you're making it out to be.

Charles_Henry said:
Derive the pure SU(2) YM theory on [itex]\mathbb{R}^4[/itex] from the action.

Literally use the Euler-Lagrange equations to obtain the equation of motion for the gauge potential.

Let [itex]A_{\mu} (x)[/itex] be a solution to these equations. Show:

[itex]\tilde{A}_{\mu} (cx)[/itex] is also a solution (with the same action).

This is meant to follow from the scale-invariance of pure YM. Assume [itex]A_{\mu} (x)[/itex] is a solution. Consider [itex]A_{\mu} (cx)[/itex] and make a change of coordinates, taking into account that [itex]A_\mu[/itex] scales like a tensor of the appropriate degree.