madspin decaymode onshell (a bit urgent..)

Asked by Sihyun Jeon on 2017-12-09

Hi MadGraph team,

I again have questions about the use of spinmode onshell.

We are using Richard Ruiz's card in [1], "Heavy Neutrino"

With this we do the following command lines :
define mu = mu+ mu-
define e = e+ e-
define w = w+ w-
generate p p > n1 mu [QCD]
output HeavyNeutrinoToMuMu_Schannel_M20_NLO
set spinmode onshell
decay n1 > mu j j

The last 2 lines are done in madspin.

With this, we expected to get same-sign and opposite sign dimuon events with 2 jets where the diagram looks like [2], since we generatlised muon charges by defining mu as mu+ and mu-.

However, it looks like this makes only OS dimuon events (if the muon from p p > is + charged, then muon from n1 > is - charged..).

Is it an attempt to make the lepton number conserved?

I would like to know if this issue could be solved somehow.

Thanks, Sihyun Jeon.


Question information

English Edit question
MadGraph5_aMC@NLO Edit question
No assignee Edit question
Solved by:
Sihyun Jeon
Last query:
Last reply:


If I run the the command that you refer to, the computation returns a zero cross-section.
Please provide to me the command to change the benchmark point to match yours.



Sihyun Jeon (shjeon) said : #2

Sorry forgot to mention about setting the n1 mass

Doing "set mn1 20" will make it work I think.


Still not working... Still returning zero cross-section.
This is related to some vanishing coupling, indeed if I ask MG5aMC to remove all interaction with zero coupling before the diagram generation. then your process does not correspond to any Feynman Diagram.



Sihyun Jeon (shjeon) said : #4

Hi Olivier

Ahh I'm deeply sorry.

You should also modify parameters card. I believe muon and n1 coupling is set to 0. as default

VmuN1 = Parameter(name = 'VmuN1',
                  nature = 'external',
                  type = 'real',
                  value = 0.1,
                  texname = 'V_{\\text{muN1}}',
                  lhablock = 'NUMIXING',
                  lhacode = [ 4 ])

Sihyun Jeon (shjeon) said : #5

I checked

 generate p p > n1 mu [QCD]


 decay n1 > e j j

makes both SS and OS.
My first guess to this problem was that this was due to lepton number conservation (maybe a problem related to model cards) but I think it's not...


I found the origin of the problem.
The problem is that you technically break quantum mechanics since your two same sign muon are not interchangeable any more.
indeed flipping the momenta of the two muons will not give the same matrix-element evaluation.

The problem is that madspin does not know which muon is the first and which one is the second and assign them wrongly.
I guess that one dirty solution would be to do the generation like this:
 generate p p > mu n1 [QCD]
(i.e. flipping the ordering at generation time) but I did not test it.



Sihyun Jeon (shjeon) said : #7

Hi Olivier, it seems to be working. Thanks alot