MadGraph5_aMC@NLO

Momentum conservation at vertex

Asked by andlessa

Hi,

I've have implemented some momenta-dependent vertices following the instructions here:

https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/FormFactors

In particular, I have used the Fortran implementation and defined something like:

FFVNP4 = Lorentz(name = 'FFVNP4',
                                     spins = [ 2, 2, 3 ],
                                     structure = 'Gamma(3,2,-1)*ProjP(-1,1)*MyFunction( P(-1,1)**2, P(-1,3)**2 ,P(-1,2)**2 )'
                                     )

since the form factor depends on the invariants p1^2, p3^2 and p2^2.

As far as I know the UFO format assumes all momenta to be incoming, so we should have p1+p2+p3 =0.

However, if I make the replacement:

   P(-1,3)**2 -> (P(-1,1)**2) + (P(-2,2)**2) + 2*(P(-3,1)*P(-3,2))

I get different results. I've also tried: P(-1,3)**2 -> (P(-1,1)**2) + (P(-2,2)**2) - 2*(P(-3,1)*P(-3,2))
and once again the results do not agree.

Could momentum conservation be violated at the vertex? Am I doing something wrong?

Cheers,

-Andre

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#1

Hi,

Can it be a problem of numerical accuracy?

Cheers,

Olivier

Revision history for this message
andlessa (lessa-a-p) said (last edit ): 		#2

I don't think so, because the cross-sections differ by a factor of 2.
One of the momenta going into the vertex is off-shell.
But that shouldn't be a problem, right?

Is there a way to print out the momentum vectors, so I can explicitly check?

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said : 		#3

You can add the print statement within the various routine associated to your routine which are located within the following directory:
Source/DHELAS/
If you are not unlucky, you will find multiple file starting with
FFVNP4
(like FFVNP4_0.f , FFVNP4_1.f , FFVNP4_2.f, ...)

Where you can print the momenta directly (note that the convention in that file is NOT the same as the one within UFO model).

To see those print statement, you need to check the log present in directories like
SubProcesses/P1......./G....

Cheers,

Olivier

Revision history for this message
andlessa (lessa-a-p) said : 		#4

Thanks a lot!

That allowed me to check what was going on and indeed I had some numerical instabilities due to tiny unphysical momenta caused by the machine precision!

Best

Revision history for this message
andlessa (lessa-a-p) said : 		#5

Thanks Olivier Mattelaer, that solved my question.