Small Width Treatment issue

Hi,

First I will apply the following method:
https://answers.launchpad.net/mg5amcnlo/+faq/2312

and remove all the mass/width for very light particle.
Particle 1 and 4 should have a zero width (and mass). Having a non zero width for such particles is a non sense.
(even a non zero mass for particle 1 is a non sense to my point of view but this might be arguable).

Second what is the mass of h1?
your syntax generate an onshell h1 which decays to two onshell Z.
So if m_h1 < 2*m_Z your syntax is not valid and can lead to issue.

Third,
The warning is to make you aware that we use trick to allow to have meaningfull result in presence of particle with very small width. The trick is mandatory to avoid to have issue with numerical precision of the code.
The warning is very relevant in this case since the trick use the narrow-width-approximation which is likely not valid for particle 1 and 4.
But reducing the threshold will not really help in this case.

Cheers,

Olivier

> On 30 Jun 2021, at 20:10, Aya Ismail Beshr <email address hidden> wrote:
>
> New question #697793 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/697793
>
> Hi,
>
>
> I am working with 2HDM typeI model, the UFO model I am using with MG5 is generated by SARAH. I am running the process:
>
> p p > h2 > h1 h1, h1 > b b~, (h1 > z z, z > mu+ mu-, z > mu+ mu-),
>
> only on 100 events as I need to compare the x-sections for certain benchmark points. An issue arises that some particles have very low width < threshold, so I tried the solution mentioned on launchpad:
>
> 1e-12 = small_width_treatment.
>
> however, I still have some particles with very low width:
>
> WARNING: Particle 1 with small width detected (9.9059485e-27): See https://answers.launchpad.net/mg5amcnlo/+faq/3053 to learn the special handling of that case
> WARNING: Particle 4 with small width detected (3.8036293e-13): See https://answers.launchpad.net/mg5amcnlo/+faq/3053 to learn the special handling of that case
> WARNING: Particle 13 with small width detected (1.2606804e-18): See https://answers.launchpad.net/mg5amcnlo/+faq/3053 to learn the special handling of that case
>
>
> This issue still causes the code to run very slowly, I tried to change the width to the order of -30 as a trial to overcome the least width, but it seems that it is wrong!
>
> Any suggestion with the issue will be much appreciated!
>
>
> Thanks,
> Aya
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

Thanks for reply!

I will try adding those restrictions.

The h1 mass is ~ 125 GeV, so in this sense mass h1 < mass ZZ, so this will make issue as you kindly mentioned. In this case should I do the decay with madspin instead and change decay mode to none?

Thanks,
Aya

Sorry, changing the spinmode to none.

Well,

Thanks,
Aya

Thanks Olivier Mattelaer, that solved my question.