MadGraph5_aMC@NLO

Completely separate particle and ISR generation from particle decay

Asked by Siyuan Sun

Hi

I'm trying to generate and decay stops

I want to do:
generate p p > t1 t1~ $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~, (t1 > n1 w+ b / x1+ x2+ b1 b2), (t1~ > n1 w- b~ / x1+ x2+ b1 b2) @1
add process p p > t1 t1~ j $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~, (t1 > n1 w+ b / x1+ x2+ b1 b2), (t1~ > n1 w- b~ / x1+ x2+ b1 b2)@2
add process p p > t1 t1~ j j $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~, (t1 > n1 w+ b / x1+ x2+ b1 b2), (t1~ > n1 w- b~ / x1+ x2+ b1 b2) @3

however, this is prohibitively slow.

Is there a way to separate the stop generation and stop decay completely into two separate steps.

For example:

Can I generate via madgraph:

generate p p > t1 t1~ $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~ @1
add process p p > t1 t1~ j $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~ @2
add process p p > t1 t1~ j j $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~ @3

and then feed the output back into madgraph and do the decay
(t1 > n1 w+ b / x1+ x2+ b1 b2), (t1~ > n1 w- b~ / x1+ x2+ b1 b2)

in a completely separate step?
The decay is 1->3 body because the t1 mass is less then the n1 + top mass.

Or is there some syntax that allows you to completely separate the stop decay from the stop generation. I dont think the current syntax does this
generate p p > t1 t1~ j j $ go susylq susylq~ b1 b2 t2 b1~ b2~ t2~, (t1 > n1 w+ b / x1+ x2+ b1 b2), (t1~ > n1 w- b~ / x1+ x2+ b1 b2)
as I think madgraph still has to calculate the interference between the stop decay products with the additional j j's

I want to ignore all interference between ISR and stop decay products in the interest of time.

Thank you very much,

-Sun

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Valentin Hirschi
Solved:
Last query:
Last reply:
Revision history for this message
Best Valentin Hirschi (valentin-hirschi) said : 		#1

As you may know the usual solution for these kind of problems is MadSpin.
However, since your decay is 1>3, the default MadSpin implementation will not work.

As indicated in MadSpin's wiki page (https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/MadSpin), you can put the following option in madspin_card.dat:

set spinmode none

This will change the operational mode of MadSpin and generate decay events independently and "attach them" to the production process.
This means that you will neglect not only the interference between stop decay products and ISR, but also *all* spin correlations. Also, there will be no Breit-Wigner smearing of the stop particles originally produced onshell, so that their decay products will always reconstruct the stop mass exactly.

So this approximation is rather crude, but event generation will be significantly faster this way. You can always compare this approximation with the full simulation (i.e. using decay chains) for one benchmark points and then decide for yourself if it is accurate enough for your purpose.

As of now, I believe the full decay chains and the MadSpin in 'spinmode none' mode are your only to options in MG5aMC.

Revision history for this message
Siyuan Sun (ssun) said : 		#2

Hi Valentin,

I know this is really late but I just wanted to say thank you for the answer. Madspin worked perfectly.

Best,

-Sun

Revision history for this message
Siyuan Sun (ssun) said : 		#3

Thanks Valentin Hirschi, that solved my question.