Name: Aliaksandr Dubrouski (email_not_shown)
Date: 04/09/15-12:32:54 PM Z

2 Stanislav

Look at the input form of the \[Epsilon]^(K2*P1*P2*K1), because what you
see is just a format used to display the contraction of LeviCivita tensor
with four 4-momenta.
Substitution rules work for the InputForm expressions. This is basically
not a FeynCalc question but a Mathematica programming question. You can
specify the custom tensor with Tensorfunction as follows
>>> Tensorfunction[t, mu, nu, ...]
See for
Additionally I would suggest substituting the contractions of your tensor
with four-vectors after the FeynCalc will contract the lorentz indices for
you with the vector components.

I.e. you do not need to specify the tensor in components as you mentioned
phi_mu_nu = diag{0,1,1,0} but specify it as a substitution rules for tensor
contracted with two arbitrary 4-vectors and its trace.

2015-04-09 0:30 GMT+03:00 Stanislav <>:

> Okay, several questions have arised...
> 1. Is there a way to specify my own tensor? Metric and LeviCivita are not
> enough for me. For example, i'd like this one with lower indices to be
> specified: phi_mu_nu = diag{0,1,1,0}
> Does FeynCalc contract this kind of predefined tensors?
> 2. When contracting 4 vectors with LeviCivita symbol i get an object like
> \[Epsilon]^(K2*P1*P2*K1)
> I would like either to make this contraction explicit or to replace it
> with already known expression.
> But things like
> >>> M = M /. \[Epsilon]^(K2*P1*P2*K1) -> my_expression
> do not work. What's going on?

            Aliaksandr Dubrouski

This archive was generated by hypermail 2b29 : 05/27/17-03:40:02 PM Z CEST