Diagramatic perturbative expansion of QCD

[Sunset]

Hi!

Has anybody seen the perturbative expansion of the generating functional of QCD Z[J,\xi,\xi*,\eta,\eta*] expressed with Feynman diagrams? I mean, there should be an expansion, containing external sources denoted by something like
-------o abbreviation for i \int d^4 x J
-------# abbreviation for i \int d^4 x \eta
and so on...

I haven't found any book showing this.

Is it maybe simply the sum of all possible graphs with their combinatorical prefactors that can be constructed from the Feynman rules?

Best regards Martin

 

[Avodyne]

Is it maybe simply the sum of all possible graphs with their combinatorical prefactors that can be constructed from the Feynman rules?

Yep. When sources are attached, there is an extra combinatoric factor of 1/n! for each type of source, where n is the number of that type of source that appears in the diagram.

 

[Sunset]

Hi Avodyne!

This would be really cool... but I want to make sure, that we mean the same thing:


\eta is the quark source
J the gauge field (gluon) source
\xi the ghost source

quark line is a straight line
ghost line is dottet
gluon line is twidled

i \int d^4 x \eta I draw as a dot
i \int d^4 x \xi as triangle
i \int d^4 x J as box

So Z is equal the little bitmap I attached (up to prefactors and understood that there are infinite many more diagrams i.e. all possible ones ) ? (I didn't take care of colors and flavors, just assume there is only one flavor and one color, if one takes into account more colors then more diagrams...) One would have to draw ALL possible graphs, that means disconnected graphs as the last one, too.

If that is true , is the generating functional W=lnZ also exactly the sum of ALL possible connected diagrams?

 

[Sunset]

Oh wait, I did a mistake. Understood, each external point should have a source (forgot to draw them)

 

[Sunset]

I corrected that one

 

[Avodyne]

For some reason I'm not able to view your 2nd picture. In the first, the quark loop with a single gluon attached is zero. And one has to get the combinatoric factors right. But then, yes, Z is just the sum of all possible diagrams (connected and disconnected), and W=log(Z) is the sum of just the connected diagrams.

This is all explained pretty well for phi^3 theory in the book by Srednicki (google to find a free draft copy online).

 

[Sunset]

ok, great!

I wasn't sure if really everything goes the same way as in phi^n theory.

Thanks a lot,

Martin