Hi,
I would probably try to cook up some replacement rule, like
FCI[LC[a, b][p, q] LC[a][r, s, t] LC[b][r, s, t]] /. {
Eps[a_LorentzIndex, b_LorentzIndex, rest__] Eps[a_LorentzIndex,
rest1__] Eps[b_LorentzIndex, rest2__] :> 0
}
although I of course understand that with many terms this might be slow.
The issue
is that FeynCalc is not a tensor algebra system, so it cannot put
certain expressions to
zero judging purely from their symmetry/antisymmetry properties. One
would need something like xAct, Cadabra, RedBerry etc. to handle that
automatically.
Cheers,
Vladyslav
Am 30.11.18 um 22:54 schrieb Pablo Sanchez Puertas:
> Hello and congratulations for the program and its maintenance,
>
> For some reasons, in my calculations it is way easier to keep the LC tensors uncontracted up to the very end, so I prefer using simplifications via EpsEvaluate[] and Contract[,Rename->True,EpsContract->False].
> I found however that the following input
>
> EpsEvaluate[LC[a, b][p, q] LC[a][r, s, t] LC[b][r, s, t]]
>
> does not produce a vanishing result (but same output as input) despite this should be obvious from antisymmetry reasons. When using
>
> Contract[LC[a, b][p, q] LC[a][r, s, t] LC[b][r, s, t]]
>
> of course leads to 0, but then it would contract other LC tensors in my expression, that I prefer to avoid.
>
> Since the reasons for which is zero is only antisymmetry, I was wondering if this could be easily implemented into my expressions.
>
>
> Best regards,
> Pablo
>
This archive was generated by hypermail 2b29 : 09/04/20-12:55:05 AM Z CEST