Zero three-body decay width for stop

Asked by Daniele Bertolini on 2012-03-27

Hello,
I'm using MG5 with the MSSM in UFO format, which I got from FeynRules. I have replaced the param_card with one containing my spectrum/mixings/widths. The latter has been generated using SoftSusy+SDecay. The lowest states in my spectrum include the two stops (which are comparable admixtures of Left and Right states), one sbottom (mostly Left) and the lightest neutralino. I would like to study the decay t1 or t2 -> b1 w+. It cant proceed with both final states on-shell, so either b1 or w+ must be off-shell with a three body final state. I'm generating the processes with MadGraph and caluclating the decay width with MadEvent. My problems are:
1) while t2 > b1 w+, w+ > e+ ve has a non zero width i get zero for t2 > b1 w+, b1 > n1 b
(the process b1 > n1 b alone has a non-vanishing decay width). Why does the diagram with off-shell b1 give zero width?
2) as I said, according to the mixing matrix in the param_card, t1 and t2 are almost equal admixtures of Left and Right so the above decay with t2 replaced by t1 should exist as well. Again, the process is generated by MadGraph but the decay width that I get is zero, in this case both with a w+ or a b1 off-shell.

Do you have any clue about how I could fix those problems? Your help would be greatly appreciated.
Thanks a lot.
Daniele

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
2012-03-28
Last query:
2012-03-28
Last reply:
2012-03-27

Hi Daniele,

Thanks for using MG5 and to keep us updated with your trouble.

By running with the default param_card.dat, I successfully computed
non zero decay width for both channel.
So this is (at least in part) due to the spectrum of masses (or
couplings) that you input.

In particular, I suspect a problem with the cut on the invariant mass
that you include.
Let me explain this point. When you use a decay chain like:
> t2 > b1 w+, w+ > e+ ve
you force the w+ to be onshell (and NOT offshell). onshell meaning
that it should be close to it's invariant mass.

the following parameter in the run_card.dat defines this concept of
"close to it's invariant masss"
#**********************************
# BW cutoff (M+/-bwcutoff*Gamma)
#**********************************
   15 = bwcutoff ! (M+/-bwcutoff*Gamma)

so this might be responsible of your zero result.

I would strongly suggest to use instead

> generate t2 > b1 e+ ve
> add process t2 > b1 n1 b

Then you are not sensitive to this cut.

Cheers,

Olivier

On 27-mars-12, at 18:55, Daniele Bertolini wrote:

> New question #191860 on MadGraph5:
> https://answers.launchpad.net/madgraph5/+question/191860
>
> Hello,
> I'm using MG5 with the MSSM in UFO format, which I got from
> FeynRules. I have replaced the param_card with one containing my
> spectrum/mixings/widths. The latter has been generated using SoftSusy
> +SDecay. The lowest states in my spectrum include the two stops
> (which are comparable admixtures of Left and Right states), one
> sbottom (mostly Left) and the lightest neutralino. I would like to
> study the decay t1 or t2 -> b1 w+. It cant proceed with both final
> states on-shell, so either b1 or w+ must be off-shell with a three
> body final state. I'm generating the processes with MadGraph and
> caluclating the decay width with MadEvent. My problems are:
> 1) while t2 > b1 w+, w+ > e+ ve has a non zero width i get zero for
> t2 > b1 w+, b1 > n1 b
> (the process b1 > n1 b alone has a non-vanishing decay width). Why
> does the diagram with off-shell b1 give zero width?
> 2) as I said, according to the mixing matrix in the param_card, t1
> and t2 are almost equal admixtures of Left and Right so the above
> decay with t2 replaced by t1 should exist as well. Again, the
> process is generated by MadGraph but the decay width that I get is
> zero, in this case both with a w+ or a b1 off-shell.
>
> Do you have any clue about how I could fix those problems? Your help
> would be greatly appreciated.
> Thanks a lot.
> Daniele
>
>
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Daniele Bertolini (danbert) said : #2

Hi Olivier,
thanks a lot for your quick and detailed answer.
From the fact that the width for t2 -> b1 w+ was zero, while for t2 -> b1 w+, w+ > e+ ve was non-vanishing my naive (and wrong) understanding was that the second case was the full diagram with an off-shell w+. I was not aware about this BW cutoff, thanks for letting me know. I will try the calculation with the full set of diagrams as you suggested.

Daniele Bertolini (danbert) said : #3

Now it seems to work! Thanks!

Daniele Bertolini (danbert) said : #4

Thanks Olivier Mattelaer, that solved my question.