e+ e- > e+ e- ta+ ta- ?

Hi,

The UCL Webpage does not allow anymore any web generation/run.
This is still possible on the UIUC web page.

Now I would strongly advice to install the code on your laptop (especially for the type of process that you are looking for).
Installing our program is really easy (technically you just have to unpack it) and the user interface on a laptop is actually nicer than the one provided by the web.

Cheers,

Olivier

Hi Olivier,

thanks! It might be a good idea to tell users of the UCL weg page that the service was discontunued.

I got the code and went through the tutorial with e+ e- > e+ e- ta+ ta- but it looks like MG5 won't do the right thing out-of-the-box. The run card has undefined beam particles (0 = lpp1/2), ridicously large beam energy (500 GeV) and all kinds of cuts on particles which will not be produced.

If e+ e- two-photon diagrams are possible at all I guess I need a custom setup to generate them safely. Would you have a pointer for specific documentation for such a case, perhaps a paper or similar?

The physics aim is to reproduce the DELPHI study of e+ e- > e+ e- ta+ ta- at LEP2 which gave ~400 pb XS at sqrt(s) = 180-200 GeV, and then to get a prediction for the same process at sqrt(s)= 10.5 GeV, i.e. for B-factories.

Cheers, Stefan

Hi Stefan,

I'm sorry that MG5aMC does not fit your needs.

> The run card has undefined beam particles (0 = lpp1/2),
> ridicously large beam
> energy (500 GeV)

I do not consider that energy as stupid, this just depend of the accelerator that you have in mind.
In any case, this is a default value that you can change as you want.
For beam profile, we do not have that much support for non delta-beam, some people made a plugin to support ISR profile, you can check: 1804.00125.

Now let me stress that the run_card is that a file that you can( have to) edit to fit your need.
So do not hesitate to change the energy of the beam by editing the file.

> and all kinds of cuts on particles which will not be
> produced.

The latest version 2.7.0 should hide automatically all those cuts that are irrelevant for your process. I guess that you are not using the latest version (are you?)

> If e+ e- two-photon diagrams are possible at all I guess I need a custom
> setup to generate them safely. Would you have a pointer for specific
> documentation for such a case, perhaps a paper or similar?

I do not understand your question here.
If you want to have a four point interaction vertex e+ e- a a
then you need to import an equivalent BSM model.
You can load your favorite model from a long list
(you can see the list of model that we can automatically download from
display model_list).
If needed you can check CMS/ATLAS who have their own UFO model, and if needed you can create your own model (either manually or via a tool like FeynRules)

Cheers,

Olivier

> and all kinds of cuts on particles which will not be
> produced.

> On 24 Jan 2020, at 11:05, Stefan Kluth <email address hidden> wrote:
>
> Question #688217 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/688217
>
> Stefan Kluth posted a new comment:
> Hi Olivier,
>
> thanks! It might be a good idea to tell users of the UCL weg page that
> the service was discontunued.
>
> I got the code and went through the tutorial with e+ e- > e+ e- ta+ ta-
> but it looks like MG5 won't do the right thing out-of-the-box. The run
> card has undefined beam particles (0 = lpp1/2), ridicously large beam
> energy (500 GeV) and all kinds of cuts on particles which will not be
> produced.
>
> If e+ e- two-photon diagrams are possible at all I guess I need a custom
> setup to generate them safely. Would you have a pointer for specific
> documentation for such a case, perhaps a paper or similar?
>
> The physics aim is to reproduce the DELPHI study of e+ e- > e+ e- ta+
> ta- at LEP2 which gave ~400 pb XS at sqrt(s) = 180-200 GeV, and then to
> get a prediction for the same process at sqrt(s)= 10.5 GeV, i.e. for
> B-factories.
>
> Cheers, Stefan
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

I made progress, I found a talk from on MG5 from the recent FCC-ee workshop, and from that I learned that for e+e- one needs to install the model "loop_qcd_qed_sm" and to do

generate e+ e- > e+ e- ta+ ta- [noborn=QED]

MG5 downloaded a few libraries and is now happily reducing the results of >5000 diagrams. Should be ready soon.

I'll contact the author of the talk (Antonio Costantini) for more details about this, it should be possible to work in lower orders and perhaps QED only.

Thanks for your input! Cheers, Stefan

Hi,

I'm aware of that

Note that the syntax [noborn=QED] is (by syntax) assuming that you do not have tree-level diagram. Since this is not the case here, it is possible that the loop-computed are NOT finite.
If that's the case the result will be meaningless

The way to test if you can use that syntax safely is to test if the loop returns a finite number for a random phase-space point. This can be done via:
generate e+ e- > e+ e- ta+ ta- [sqrvirt=QED]
output
launch

And here is the result:

|================================================================================================
|| Results for process epem > epemtaptam (Loop-induced)
|================================================================================================
|| Phase-Space point specification (E,px,py,pz)
|
| 5.0000000000000000e+02 0.0000000000000000e+00 0.0000000000000000e+00 5.0000000000000000e+02
| 5.0000000000000000e+02 0.0000000000000000e+00 0.0000000000000000e+00 -5.0000000000000000e+02
| 8.8551333054502976e+01 -2.2100690287689979e+01 4.0080353191685333e+01 -7.5805430956936632e+01
| 3.2832941922709853e+02 -1.0384961188345630e+02 -3.0193375538954012e+02 7.6494921387165888e+01
| 1.5235810946743061e+02 -1.0588095966659220e+02 -9.7709638326975707e+01 4.9548385226792817e+01
| 4.3076113825096758e+02 2.3183126183773851e+02 3.5956304052483051e+02 -5.0237875657022109e+01
|
| Unknown numerical stability because MadLoop is in the initialization stage.
|
|| Loop amplitude squared, must be finite:
| Finite = 1.0249816951450901e-15
|(| Pole residues, indicated only for checking purposes: )
|( Single pole = 1.0987815745683220e-16 )
|( Double pole = 1.0418226269192490e-17 )

So you can see that each loop has a single pole and double pole. (i.e. the contribution is not finite). So you can not use the syntax noborn= for this process.

Cheers,

Olivier

> On 24 Jan 2020, at 13:44, Stefan Kluth <email address hidden> wrote:
>
> generate e+ e- > e+ e- ta+ ta- [noborn=QED]

Hi Olivier,

thanks for pointing this out!

Do you know how to generate if I want only the QED (mostly two-photon interaction) diagrams, at LO would be sufficient?

Cheers, Stefan

Hi Stefan,

I guess what you are looking for is
generate e+ e- > e+ e- ta+ ta-

This would include all QED diagram at LO.
Is that what you are looking for?

Cheers,

Olivier

Hi Olivier,

actually yes, I ran "generate e+ e- > e+ e- ta+ ta-" and obtained 48 graphs including the two-photon interactions. That turns out to be more simple than I thought.

However, now I have to learn how to apply cuts properly on the final state particles. Since I want to calculate untagged two-photon interactions, I want no cuts on the final state e+e-, i.e. the beam particles can have scattering angles ~=0. The taus should be in the acceptance of a detector, so I would have to place a cut on the angle between the final state taus and the beam direction. I am reading the Madgraph paper arxiv.org/pdf/1405.0301.pdf now, but if you would perhaps have a pointer to a specific example for this application I would be grateful.

Cheers, Stefan

Hi,

In most of the cases, you can/should use the pre-defined cuts present in the run_card.dat file.
Note that however you need to be careful that such cut are done at parton-level.
This might not have a huge impact for the lepton, but it will make a difference for your tau obviously (which is only observe via it's decay mode in the detector).

In general, cut in our tool should be quite weak and done for two reasons:
1) avoid (N)LO singularities. For your process you have to put some cut on the electron in order to avoid the LO singularity of the matrix-element (both the soft and the collinear one).
2) setting cuts to avoid too much inefficiency when you apply hard cut (i.e. detector cut) later in your analysis work flow.

Cheers,

Olivier

> On 27 Jan 2020, at 10:22, Stefan Kluth <email address hidden> wrote:
>
> Question #688217 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/688217
>
> Stefan Kluth posted a new comment:
> Hi Olivier,
>
> actually yes, I ran "generate e+ e- > e+ e- ta+ ta-" and obtained 48
> graphs including the two-photon interactions. That turns out to be more
> simple than I thought.
>
> However, now I have to learn how to apply cuts properly on the final
> state particles. Since I want to calculate untagged two-photon
> interactions, I want no cuts on the final state e+e-, i.e. the beam
> particles can have scattering angles ~=0. The taus should be in the
> acceptance of a detector, so I would have to place a cut on the angle
> between the final state taus and the beam direction. I am reading the
> Madgraph paper arxiv.org/pdf/1405.0301.pdf now, but if you would
> perhaps have a pointer to a specific example for this application I
> would be grateful.
>
> Cheers, Stefan
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.