Modifying couplings in loop processses

Asked by David Englert on 2017-03-20

Dear MadGraph Developers,

I would like to study the gg->Zh process considering both the triangle and box quark loop contributions with the following couplings modified:
- g_{hZZ}
- g_{htt}
- g_{Ztt}
The only information I'd like to extract is the cross section for these loop processes for different partonic energies.

I understand you can use the 'loop_sm' UFO model to study this process in SM. Tracing back from 'vertices.py' I have identified
GC_32, GC_37, GC_22 and GC_23 as the relevant coupling parameters.

My naive idea for tackling this problem is to modify the UFO file by:
- introducing three new coupling modifier parameters in the 'parameters.py'
- modify the there couplings in 'couplings.py' accordingly by introducing these couplings modifiers

Question at this point:
- Are these modifications enough, or should I also alter the couplings in 'CT_couplings?'

If the answer to the above yes, I'd need to change them; then how should I proceed modifying the couplings counter terms? Do I need go back a level to generate UFO model with modified couplings with FeynRules? Or is there a simpler approach?

Thank you for the help in advance.

Best,
David

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
David Englert
Solved:
2017-04-22
Last query:
2017-04-22
Last reply:
2017-03-21

This question was reopened

Dear David,

> Question at this point:
> - Are these modifications enough, or should I also alter the couplings in 'CT_couplings?’

You indeed also need to modify the ‘CT_couplings’ since your modification will impact the R2 counter-term that we need in order to evaluate the loop.

> If the answer to the above yes, I'd need to change them; then how should I proceed modifying the couplings counter terms?

You can either do it by hand by identifying which R2 is modified and modify the associated coupling accordingly.
It might indeed be safer and easier to use FeynRules.

Cheers,

Olivier

> On 20 Mar 2017, at 16:13, David Englert <email address hidden> wrote:
>
> New question #576549 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/576549
>
> Dear MadGraph Developers,
>
> I would like to study the gg->Zh process considering both the triangle and box quark loop contributions with the following couplings modified:
> - g_{hZZ}
> - g_{htt}
> - g_{Ztt}
> The only information I'd like to extract is the cross section for these loop processes for different partonic energies.
>
> I understand you can use the 'loop_sm' UFO model to study this process in SM. Tracing back from 'vertices.py' I have identified
> GC_32, GC_37, GC_22 and GC_23 as the relevant coupling parameters.
>
> My naive idea for tackling this problem is to modify the UFO file by:
> - introducing three new coupling modifier parameters in the 'parameters.py'
> - modify the there couplings in 'couplings.py' accordingly by introducing these couplings modifiers
>
> Question at this point:
> - Are these modifications enough, or should I also alter the couplings in 'CT_couplings?'
>
> If the answer to the above yes, I'd need to change them; then how should I proceed modifying the couplings counter terms? Do I need go back a level to generate UFO model with modified couplings with FeynRules? Or is there a simpler approach?
>
> Thank you for the help in advance.
>
> Best,
> David
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

David Englert (ganzilion) said : #2

Thanks Olivier Mattelaer, that solved my question.

David Englert (ganzilion) said : #3

Sorry, the way I phrased the question might have been a bit confusing, misleading you to believe that this is an NLO process.
Let me emphasize that I would like to study gg -> Zh at leading order(!), which should be a loop-induced process including one quark loop only (box or triangle), which doesn't have a tree level equivalent. This should be convergent, and if I am not mistaken you don't need UV or R2 terms in this scenario since it is not an NLO calculation, and doesn't include any divergencies.
Reflecting on this I'd anticipate that the original proposal is correct meaning that there is no need to modify the 'CT_couplings' or other files apart from 'couplings.py' within UFO in this specific scenario. Could you please confirm this Olivier?

Also correct me if I am wrong but you cannot go to NLO for a loop induced process in madgraph, as that would mean that two loop processes are considered which are not automatized (yet).

Cheers,
David

Hi David,

> Sorry, the way I phrased the question might have been a bit confusing, misleading you to believe that this is an NLO process.

No I was well aware that you were speaking of a loop-induced process.

> This should be convergent, and if I am not mistaken you don't need UV or R2 terms in this scenario since it is not an NLO calculation, and doesn't include any divergencies.

Well this is not fully correct.
First the R2 are a finite part and are needed for any loop. So you certainly need it in this case.
Second, in general you are correct that you do not need UV, but some counter-example actually exists (typically involving mixing of particle)

> Also correct me if I am wrong but you cannot go to NLO for a loop
> induced process in madgraph, as that would mean that two loop processes
> are considered which are not automatized (yet).

Correct.

Cheers,

Olivier

> On 21 Mar 2017, at 16:22, David Englert <email address hidden> wrote:
>
> Question #576549 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/576549
>
> Status: Solved => Open
>
> David Englert is still having a problem:
> Sorry, the way I phrased the question might have been a bit confusing, misleading you to believe that this is an NLO process.
> Let me emphasize that I would like to study gg -> Zh at leading order(!), which should be a loop-induced process including one quark loop only (box or triangle), which doesn't have a tree level equivalent. This should be convergent, and if I am not mistaken you don't need UV or R2 terms in this scenario since it is not an NLO calculation, and doesn't include any divergencies.
> Reflecting on this I'd anticipate that the original proposal is correct meaning that there is no need to modify the 'CT_couplings' or other files apart from 'couplings.py' within UFO in this specific scenario. Could you please confirm this Olivier?
>
> Also correct me if I am wrong but you cannot go to NLO for a loop
> induced process in madgraph, as that would mean that two loop processes
> are considered which are not automatized (yet).
>
> Cheers,
> David
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

David Englert (ganzilion) said : #5

Hi Olivier,

Just a small update on this topic which might be of use to someone else. I checked whether introducing the coupling modifiers while NOT changing any of the UV or R2 counterterms. and it seems to do the trick: the cross sections scale with k^2 just as you'd expect, please have a look here:
https://www.dropbox.com/s/6jwipivpvnzbspj/CHM_kappa_study.html?dl=1

Cheers,
David