Wildly inconsistent cross section in process specification

Asked by Ferdinand Schenck on 2018-07-30

Hi,

So we are trying to properly study interference of our BSM signal with SM processes. As part of that we have been using squared coupling orders to specify the interfering diagrams (no problem so far).

For a start I'm just going to focus on our signal process.

We basically have two ways of specifying our signal process:
Fully specify the decay chain:
1) p p > j VLQ / p t t~ tp tp~ bp bp~ x x~ z h a, (VLQ > WW bb)
or restricting the decays with coupling orders:
2) p p > j w+ b~ / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4
     p p > j w- b / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4

Unsurprisingly, the cross-sections are basically equivalent, no problem so far.

But the above processes have an un-decayed W in the final state, and as we have to pass it off to Pythia this is not great.

So to fix that, we can specify the final state to be fermions, for our two cases:
Fully specified decay chain:
3) p p > j VLQ / p t t~ tp tp~ bp bp~ x x~ z h a, (VLQ > WW bb, WW > ferm ferm)
restricting the decays with coupling orders:
4) p p > j ferm+ ferm+ b~ / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4
      p p > j ferm- ferm- b / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4

The cross-section of 3 corresponds pretty well with 1 and 2 (the small difference can probably attributed to the W being forced to be on shell in the first case)

but the cross-section of 4 is less than half of what is expected, even though the feyman diagrams seem to be equivalent.
Also, in the case of 4 the ratio of hadronic to leptonic final states is changed.

Either I'm misunderstanding some part of how to specify the process correctly, or something is wrong in the way MadGraph calculates cross sections in this case.

To reproduce, the model, plus the cards for cases 1-4 can be found here:

Model: https://cernbox.cern.ch/index.php/s/lEqCDDW94nhIvpQ
Cards: https://cernbox.cern.ch/index.php/s/cPTR1XvBpMn4FqX

I tested these on MG 2.6.2, 2.6.3 and 3.0 (beta)

Cheers,

Ferdi

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Last query:
2018-07-30
Last reply:
2018-07-31

Did you have cut_decays set to True in the run_card?
Otherwise you will not have the same cut on 3 and 4 on the final state.

Cheers,

Olivier

> On 30 Jul 2018, at 18:37, Ferdinand Schenck <email address hidden> wrote:
>
> New question #671232 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/671232
>
> Hi,
>
> So we are trying to properly study interference of our BSM signal with SM processes. As part of that we have been using squared coupling orders to specify the interfering diagrams (no problem so far).
>
> For a start I'm just going to focus on our signal process.
>
> We basically have two ways of specifying our signal process:
> Fully specify the decay chain:
> 1) p p > j VLQ / p t t~ tp tp~ bp bp~ x x~ z h a, (VLQ > WW bb)
> or restricting the decays with coupling orders:
> 2) p p > j w+ b~ / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4
> p p > j w- b / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4
>
> Unsurprisingly, the cross-sections are basically equivalent, no problem so far.
>
> But the above processes have an un-decayed W in the final state, and as we have to pass it off to Pythia this is not great.
>
> So to fix that, we can specify the final state to be fermions, for our two cases:
> Fully specified decay chain:
> 3) p p > j VLQ / p t t~ tp tp~ bp bp~ x x~ z h a, (VLQ > WW bb, WW > ferm ferm)
> restricting the decays with coupling orders:
> 4) p p > j ferm+ ferm+ b~ / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4
> p p > j ferm- ferm- b / p t t~ tp tp~ bp bp~ x x~ z h a YWB^2==4
>
> The cross-section of 3 corresponds pretty well with 1 and 2 (the small difference can probably attributed to the W being forced to be on shell in the first case)
>
> but the cross-section of 4 is less than half of what is expected, even though the feyman diagrams seem to be equivalent.
> Also, in the case of 4 the ratio of hadronic to leptonic final states is changed.
>
> Either I'm misunderstanding some part of how to specify the process correctly, or something is wrong in the way MadGraph calculates cross sections in this case.
>
> To reproduce, the model, plus the cards for cases 1-4 can be found here:
>
> Model: https://cernbox.cern.ch/index.php/s/lEqCDDW94nhIvpQ
> Cards: https://cernbox.cern.ch/index.php/s/cPTR1XvBpMn4FqX
>
> I tested these on MG 2.6.2, 2.6.3 and 3.0 (beta)
>
> Cheers,
>
> Ferdi
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Ferdinand Schenck (fschenck) said : #2

Hi Oliver,

Thanks for the quick response.

Changing cut_decays to True does bring 3 in line with 4, which is odd because I would have hoped for the opposite effect (i.e. made 4 act like 3)

What I find so odd is that we basically have 1 Feynman diagram with a non-trivial cross-section (2 if you count it's conjugate),
and this diagram has a W decay, so looking at the final states I would expect 67% hadronic, and 33% leptonic final states by cross section.

By using command 3 (and with cut_decays=False) this is exactly what I see.

But by using 4 (or 3 with cur_decays set to True) I get 70% leptonic, and 30% hadronic, which is not what I expect.

Cheers,

Ferdi

Hi,

Do you understand what cut_decays = True means?
I would suggest to think about it (or read the comment in the run_card associated to it if the name is not clear enough). After that it should be clear why 3 became 4 (and how to make 4 to be 3 --if this is possible for your process--)

Cheers,

Olivier

> On 30 Jul 2018, at 21:52, Ferdinand Schenck <email address hidden> wrote:
>
> Question #671232 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/671232
>
> Status: Answered => Open
>
> Ferdinand Schenck is still having a problem:
> Hi Oliver,
>
> Thanks for the quick response.
>
> Changing cut_decays to True does bring 3 in line with 4, which is odd
> because I would have hoped for the opposite effect (i.e. made 4 act like
> 3)
>
> What I find so odd is that we basically have 1 Feynman diagram with a non-trivial cross-section (2 if you count it's conjugate),
> and this diagram has a W decay, so looking at the final states I would expect 67% hadronic, and 33% leptonic final states by cross section.
>
> By using command 3 (and with cut_decays=False) this is exactly what I
> see.
>
> But by using 4 (or 3 with cur_decays set to True) I get 70% leptonic,
> and 30% hadronic, which is not what I expect.
>
> Cheers,
>
> Ferdi
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Can you help with this problem?

Provide an answer of your own, or ask Ferdinand Schenck for more information if necessary.

To post a message you must log in.