MadGraph5_aMC@NLO

VBF slepton production

Asked by Michael Holzbock

Dear MG5 authors,

I am a member of the ATLAS experiment where we want to generate sleptons produced in a VBF topology. We assume quite simple phenomenology with the sleptons being slightly (~1-50 GeV) heavier than the neutralino LSP and all other SUSY particles are decoupled. We are using the MSSM_SLHA2 model, with only the SUSY masses being changed by hand (in case more technical details are needed, I put our run-, proc- and param cards here:https://cernbox.cern.ch/index.php/s/rUaRH7G8pKcF1jS ). Our naive approach is to do (using MG 2.8.1)

>>> define susystrong = go ul ur dl dr cl cr sl sr t1 t2 b1 b2 ul~ ur~ dl~ dr~ cl~ cr~ sl~ sr~ t1~ t2~ b1~ b2~
>>> define slep = el- el+ er- er+ mul- mul+ mur- mur+
>>> generate p p > slep slep j j / susystrong QED=99 QCD=99 @1

which seems to give us already reasonable looking distributions. However, we observed that the production of right-handed sleptons is heavily suppressed (left- to right-handed sleptons appear in a ratio of roughly 150:1 for the model we tested with a slepton mass of 200 GeV).

This is quite contrarily to claims in some papers we found on this matter, e.g. arXiv:hep-ph/0304192, that state both production modes should be nearly on equal footing as photon-mediated diagrams dominate. It should be noted that the predictions of the total cross section we get from MG roughly agrees with what is stated in the paper (and in particular we observe the same feature that the VBF surpass the DY cross sections at around 300 GeV in slepton mass).

Any hints if the suppression of right-handed slepton is expected or can be resolved somehow in our generation setup would be highly appreciated from our side!

Thanks!
Michael

Question information

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

Hi,

My guess is that you are mixing two production channel here you have VBF which should be
>>> generate p p > slep slep j j / susystrong QED=99 QCD=0 @1

and QCD production
>>> generate p p > slep slep j j / susystrong QED=0 QCD=99 @1

Your syntax add both (and other type of production actually)
and this is likely why you do not agree with literature where they typically consider VBF alone

Cheers,

Olivier

Revision history for this message
Jeff Shahinian (jshahinian) said : 		#2

Hi Olivier,

Many thanks for the feedback!

Indeed, we've been including other production modes, as an event selection targeting a VBF-like topology can in principle still have some efficiency for QCD processes.

In any case, we tried turning off QCD production as you suggested. Additionally, we separated the right-handed and left-handed processes to help debug, so our MG calls look like this:

>>> define susystrong = go ul ur dl dr cl cr sl sr t1 t2 b1 b2 ul~ ur~ dl~ dr~ cl~ cr~ sl~ sr~ t1~ t2~ b1~ b2~
>>> define slep_L = el- el+ mul- mul+
>>> define slep_R = er- er+ mur- mur+
>>> generate p p > slep_L slep_L j j / susystrong QED=99 QCD=0 @1
>>> add process p p > slep_R slep_R j j / susystrong QED=99 QCD=0 @2

The cross-sections (after Pythia8) still come out with the large discrepancy between left-handed and right-handed slepton production:

Processes, with strategy-dependent cross section info
number xsec (pb) xerr (pb) xmax (pb)
     1 1.2768e-02 3.3581e-05 1.2854e-02
     2 1.0856e-05 9.4391e-08 1.0362e-02

With respect to using QED=99 QCD=99, the left-handed cross-section essentially remained the same (surprising to me, at least), while the right-handed cross-section was reduced by a factor of ~7. So the discrepancy actually got worse, compared to what we see in arXiv:hep-ph/0304192.

So I think we can cross off QCD vertices from the list of possible culprits. Happy to run any other tests and/or provide logs that you think might be helpful. Thanks in advance!

Best,
Jeff

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

I'm not a MSSM expert to know how correct your reference paper is.
Looks like they have performed the computation by hand (since they mention using HELAS) Which makes sense since this is a very old paper. From the benchmark/model that you use, they are difference in term of coupling and Feynman Diagram between left and right diagram related to the coupling to the W boson.

My suggestion here is that you double check out those numbers with another code that support UFO model (like Sherpa or whizard).
This being said you have some unconventional cut in your file which were not implemented in your reference paper
and your param_card throws a lot of warning about missing input but I guess that those are irrelevant here.

Cheers,

Olivier

Revision history for this message
gdse ser (olicas3333) said : 		#4

Condor cluster are quite specific of their implementation examine to different cluster, unfortunatly I do no longer have such kind of cluster to be had to check anymore. Since loop triggered are pretty abnormal task due to such wave machine, it's far possible that the two aren't well suited. See more here https://waterfilterinsight.com/best-faucet-water-filter-reviews/

Can you help with this problem?

Provide an answer of your own, or ask Michael Holzbock for more information if necessary.

To post a message you must log in.