#### Hepth

Gold Member

- 437

- 39

I have terms that I need to simplify like :

(2 Pair[Momentum[p1], Momentum[PB]] Pair[Momentum[p2],

Momentum[Polarization[q, -I]]] Pair[Momentum[PB],

Momentum[Polarization[q, I]]])/(

Pair[Momentum[p1], Momentum[q]] Pair[Momentum[p2], Momentum[q]])

Which looks like :

[tex]

\frac{(p1\cdot \varepsilon (q) )( p2\cdot \varepsilon^{*}(q))}{(p1 \cdot q)}

[/tex]

Everything is general 4-vectors, the polarizations also. Now this, by hand, is easily simplified by doing a polarization sum. The top is just -p1.p2, as the sum gives the negative metric tensor. I can't seem to do this in FeynCalc though. If you use "PolarizationSum[m,n]" itll just spit out the negative metric tensor, but the indices of p1,p2 and epsilon are all INTERNAL, so that [m,n], is actually something internal that I cannot specify.

Has anyone done something like this before? Or do you have an alternative program?