MadGraph5_aMC@NLO

Madgraph 5 and Quarkonium

Asked by Ameir Shaa

Hi Madgraph Team,

I have a Lagrangian describing BSM physics (milli-charged particles or mCPs). I used FeynRules to generate a MadGraph model based on that Lagrangian to generate DY produced mCPs.

Now, I'd like to include Quarkonium decays to mCPs in my analysis. However, as I understand it, MG5 does not include provisions for Quarknonia states. MG4 used to have a tool called MadOnia which could do it but it is not available in MG5.

As I understand it, I could use HELAC-Onia to study Quarkonia decays. However, these decays would be only to SM particles (muons, taus, etc) and not the mCPs I would like to study (which can only be generated in MG5 thanks to FeynRules).

My question is then this: how can I reconcile the MG5 analysis of mCPs with HELAC-Onia? I would like to decay, for example, J/Psi to two mCPs. The decay can only be done in HELAC-Onia while the mCPs can only be generated in MG5. Is there a way to reconcile the two so I am able to study the decay of J/Psi to mCPs?

Many thanks in advance for your answer.

Many thanks,

Ameir

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Hua-Sheng Shao Edit question
Solved by:
Hua-Sheng Shao
Solved:
Last query:
Last reply:
Revision history for this message
Hua-Sheng Shao (erdissshaw) said : 		#1

Dear Ameir,

You are right. There is no direct way to do what you asked at this stage. However, depending on the nature of the mCPs you are considering, there might be alternative working ways for you. For instance, if you can take the mCPs as some fake SM particles (like muons, taus etc), in principle you can generate the J/Psi samples with MadOnia/HELAC-Onia or even just Pythia8 alone and decay the J/Psi particles with the SM decay modes. After that, you can replace the decay products with your mCPs and renormalise your cross sections. Of course, the partial width/decay branching ratio of J/psi -> mCPs should rely on your calculations.

Cheers,

Hua-Sheng

Revision history for this message
Ameir Shaa (ameirshaa) said : 		#2

Hi Hua-Sheng,

Many thanks for your reply. This was extremely helpful.

Wrt "you can replace the decay products with your mCPs and renormalise your cross sections" - how exactly would one go about doing the replacement? The kinematics of muons/taus are different from the kinematics of my mCPs. Would that be a problem in this solution?

Many thanks in advance for your answer.

Many thanks,

Ameir

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

Do you have a langrangian with the quarkonia as a fundamental particles?
Such that you can simulate the decay on its own? If so you could use madspin (in spinmode=none) to attach the decay.

Revision history for this message
Ameir Shaa (ameirshaa) said : 		#4

Hi Olivier,

I do not have such a Lagrangian at the moment. Suppose I did have such a Lagrangian; what would be the appropriate steps I need to take?

Would it be: Lagrangian -> FeynRules -> MG5 with MadSpin on?

Would I simply "generate p p > psi psi~" in MG5 after importing the custom model file from FeynRules (where psi represents the MCP) or would there be other steps. I have never used MadSpin before..

Many thanks in advance for your answer.

Many thanks,

Ameir

Revision history for this message
Hua-Sheng Shao (erdissshaw) said : 		#5

Hi Ameir,

You can generate the J/psi sample with J/psi decaying into two muons (which you can certainly change the mass of the muons to the mass of your mCP). In the LHE file, you simply replace the muon PDG to the PDG of mCP. If your mCPs are also fermions and coupled to the photon as the same as muons, in principle you get all spin correlations correct too. Otherwise, if you do not care about the spin correlations, you simply decay the unpolarised J/psi into your mCPs, which is uniform for the angular distribution in the rest frame of the J/psi particle. So whether such tricks may result into any problem. I do not know. It depends on the observables you are interested in.

If you just take J/psi as a elementary particle as suggested by Olivier, it might be not easy. Even you managed to find a phenomenological Lagrangian, it may require you to tune to the real data. Again, it would depend on the problem you are considering/facing.

The last point is for J/psi, the first leading process you should consider at the LHC is "generate p p > psi j" (not two J/psi !).

Hua-Sheng

Revision history for this message
Ameir Shaa (ameirshaa) said : 		#6

Hi Hua-Sheng,

Many thanks for your advice - I simply cannot thank you enough for being so helpful!

I followed your advice and generated a sample of J/Psi's in HELAC-Onia. I then decayed the J/Psi's to muons in HELAC-Onia itself and output the results to an LHE file. Since this was just a test, I did not change the mass of the muons so they would have their original mass of ~105MeV.

However, I noticed in the LHE file generated by HELAC-Onia, the mass (reflected in the mass column of the LHE file) of the muon (and electron) is zero. But the mass of the J/Psi is set at 3GeV.

Furthermore, I used E^2 - P^2 = m^2 to calculate the mass of the muons produced from the decay of the J/Psi and they are not 105MeV which means that they must be off-shell? But that cannot be as they are final state real particles and not virtual..

Would you happen to know why the muon mass in the LHE file is zero and the muons seem off-shell? Did I perhaps make a mistake somewhere?

Many thanks in advance for your answer.

Many thanks,

Ameir

Revision history for this message
Best Hua-Sheng Shao (erdissshaw) said : 		#7

Dear Ameir, indeed, in HELAC-Onia, the muon mass has been set to be zero from the J/Psi's decay products. I have no idea whether MadOnia allows you to lift such a limitation. I therefore propose the following way.

1. You just generate J/Psi LHE sample (without decay).

2. You write a small code/script to let J/Psi decay into mCPs by reading the event information from the LHE file. If it is just a 1 -> 2 decay, the kinematics are essentially trivial in the rest frame of J/Psi (see for instance eqs.(5.33-5.40) of https://arxiv.org/pdf/1907.04898.pdf). If it is a 1 -> n (n>= 3) decay, you may need to implement the matrix element information. I guess yours is just 1 -> 2 decay, right ? Of course, you loose the spin information of the mother particle.

3. Then you output a new LHE file and interface to Pythia8 (or others MCs).

Revision history for this message
Ameir Shaa (ameirshaa) said : 		#8

Hi Hua-Sheng,

Many thanks for your solution!

Indeed, it is a 1 -> 2 decay which makes it easier. However, I need to boost the daughter mCPs in the direction of the parent J/Psi. If the J/Psi is travelling along the z-axis only, then the problem is somewhat trivial. However, since the J/Psi has momentum along all 3 axes, the problem is somewhat complicated. Would you happen to have any insights on how this can be solved? Also I am assuming I need to randomise the angle at which the mCP is emitted in the rest frame of the J/Psi?

Many thanks in advance for your answer.

Many thanks,

Ameir

Revision history for this message
Ameir Shaa (ameirshaa) said : 		#9

Hi Hua-Sheng,

Many thanks for all your help and advice! I have managed to decay the J/Psi particles using a script of my own and then boosting the mCPs.

Please allow me to buy you a cup of coffee the next time I am at CERN!

Many thanks,

Ameir

Revision history for this message
Ameir Shaa (ameirshaa) said : 		#10

Thanks Hua-Sheng Shao, that solved my question.

Revision history for this message
Hua-Sheng Shao (erdissshaw) said : 		#11

Hi Ameir, ah… sorry, I am busy with a workshop these days. I forget to answer you. It is good that you manage to solve your problem.

> On 12 Jan 2022, at 05:15, Ameir Shaa <email address hidden> wrote:
>
> Question #700070 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/700070
>
> Ameir Shaa confirmed that the question is solved:
> Thanks Hua-Sheng Shao, that solved my question.
>
> --
> You received this question notification because you are subscribed to
> the question.