20.01.2015 Views

Feynman Path Integral Formulation

Feynman Path Integral Formulation

Feynman Path Integral Formulation

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

7.5 Gravitational Wilson Loop 251<br />

Fig. 7.9 Gravitational analog of the Wilson loop. A vector is parallel-transported along the larger<br />

outer loop. The enclosed minimal surface is tiled with parallel transport polygons, here chosen<br />

to be triangles for illustrative purposes. For each link of the dual lattice, the elementary parallel<br />

transport matrices R(s,s ′ ) are represented by arrows. In spite of the fact that the (Lorentz) matrices<br />

R can fluctuate strongly in accordance with the local geometry, two contiguous, oppositely oriented<br />

arrows always give RR −1 = 1.<br />

through each of the simplices that meet at that hinge one has for the total rotation<br />

matrix R ≡ ∏ s R s,s+1 associated with the given hinge<br />

[ ] μ [<br />

∏ R s,s+1 = e δ(h)U(h)]<br />

s<br />

ν<br />

μν , (7.94)<br />

as in Eq. (6.32). This matrix describes the parallel transport of a vector round the<br />

loop.<br />

More generally one might want to consider a near-planar, but non-infinitesimal,<br />

closed loop C, as shown in Fig. (7.9). Along this closed loop the overall rotation<br />

matrix will still be given by<br />

R μ ν(C) =<br />

[ ] μ<br />

∏ R s,s+1 . (7.95)<br />

s⊂C<br />

ν<br />

In analogy with the infinitesimal loop case, one would like to state that for the overall<br />

rotation matrix one has<br />

R μ ν(C) ≈<br />

[e δ(C)U(C))] μ<br />

, (7.96)<br />

ν<br />

where U μν (C) is now an area bivector perpendicular to the loop, which will work<br />

only if the loop is close to planar so that U μν can be taken to be approximately<br />

constant along the path C. By a near-planar loop around the point P, we mean one<br />

that is constructed by drawing outgoing geodesics, on a plane through P.<br />

If that is true, then one can define an appropriate coordinate scalar by contracting<br />

the above rotation matrix R(C) with the some appropriate bivector, namely<br />

W(C) =ω αβ (C)R αβ (C) , (7.97)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!