MadGraph5_aMC@NLO

Poles do not cancel for loop_MSSM

Asked by Simon Berlendis

 Dear experts,

 I try to run aMC@NLO to compute the NLO cross section of a susy process :
 MG5_aMC>import model loop_MSSM
 MG5_aMC>generate p p > dr dr [QCD]

 However, I end up with the following error message when launching the generation :

 INFO: Result for check_poles:
Error detected in "launch "
[...]
aMCatNLOError : Poles do not cancel, run cannot continue

 You can find the ME5_debug file and the param_card.dat here :
http://lpsc.in2p3.fr/atlas/berlendis/tmp/param_card.dat
http://lpsc.in2p3.fr/atlas/berlendis/tmp/ME5_debug

 I also tried to set the decay width to zero as it was said in a previous question/answer section but I end up with the same message.

 Am I missing something here ? Do you have any idea where does the problem come from ?

 Cheers, Simon.

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Valentin Hirschi Edit question
Solved by:
Simon Berlendis
Solved:
Last query:
Last reply:
Revision history for this message
Valentin Hirschi (valentin-hirschi) said : 		#1

Dear Simon,

This model 'loop_MSSM' is an old version unvalidated which should not have been made part of the release (I have removed it now).
The UV renormalization done in this model is indeed wrong for many processes, it must *not* be used.

We are working on the generation and validation of a full-fledged loop-MSSM UFO model supporting the computation of QCD corrections, and it will be made public when available.

Sorry for the confusion

Revision history for this message
Valentin Hirschi (valentin-hirschi) said : 		#2

I should also add that you have to be careful with the mass hierarchy, as for example in the param card you specified, the squark is lighter than the gluino, so that one must in principle implement on-shell subtraction to compute NLO QCD corrections to 'p p > dr dr', and this not yet available in the latest public release of MG5aMC (but it is in development).

Finally, notice that you have a list of publicly available and validated NLO UFO model at

https://feynrules.irmp.ucl.ac.be/wiki/NLOModels

which include some subsets of the MSSM.

Revision history for this message
Simon Berlendis (simon-berlendis) said : 		#3

Hi Valentin,

 Thanx for the complete answer !

 I've just tried to use SUSYQCD model and by settting all the width to 0 I don't have the error message anymore.
 However, I can't manage to estimate the cross section. The generation get stuck at the first iteration.

 But anyway, since the on-shell subtraction isn't really taken into account yet in aMC@NLO, I give up :)

 Best regards,
 Simon.

Revision history for this message
Ethan Hunt (ashwink2) said : 		#4

This is truly an exciting and pleasant information for all https://solitaire-klondike.net if anyone wants to play amazing card game.