Survey return zero cross section for pp->ad

The problem is likely that you are cut via the dynamical scale choice.
If our default dynamical scale is lower than 4 GeV^2 we automatically discard the phase-space point since this means that the PDF are not reliable since you are in a non perturbative regime.

To check if this is indeed the case, you can run at fix scale (not too small obviously otherwise the running might lead to crash).
If this is indeed the issue, I do not know how you can reliably generate this at LHC collider for such light mass since you break the factorization theorem and perturbativity.

Cheers,

Olivier

> On 25 Feb 2019, at 20:17, mehdirahmani <email address hidden> wrote:
>
> New question #678792 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/678792
>
> Hello,
>
> I used Feynrule to construct a new model, which is similar to dark SUSY model already available on the website. The difference is that it has dark photon instead of higgs as mediator/
>
> pp -> ad->n1 n1->...
> ad: in this model is a massive (0.25 GeV) dark photon, a vector portal.
> n1: lightest visible neutralino (10GeV)
> n1 decays to dark photon and low mass dark neutralino (nD: 1GeV)
> dark photon decays to mu mu
>
> Survey return zero cross section.
> Typical reasons are the following:
> 1) A massive s-channel particle has a width set to zero.
> 2) The pdf are zero for at least one of the initial state particles
> or you are using maxjetflavor=4 for initial state b:s.
> 3) The cuts are too strong.
> Please check/correct your param_card and/or your run_card.
>
> it also gives these massages:
>
> Gtk-Message: Failed to load module "pk-gtk-module"
> Gtk-Message: Failed to load module "canberra-gtk-module"
> Gtk-Message: Failed to load module "pk-gtk-module"
> Gtk-Message: Failed to load module "canberra-gtk-module"
>
>
> How can I resolve this?
>
> Thank you,
>
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thank you for your response. This partially solved my problem. Changing the dark photon mass (ad) to 10 GeV lets them to be produced in pp collisions and they are present in .lhe files. However when I use the madspin for decays I get this error:

INFO: generate 10000 decay event for particle ad
Command "generate_events " interrupted with error:
NoDiagramException : No amplitudes generated from process Process: ad > n1 n1 NP=6
   Decay: n1 > nd ad
     Decay: ad > mu+ mu-. Please enter a valid process
Please report this bug on https://bugs.launchpad.net/mg5amcnlo
More information is found in '/afs/cern.ch/user/m/merahman/MadGraph5_10GeV/MG5_aMC_v2_4_2/MSSMD/run_01_tag_1_debug.log'.
Please attach this file to your report.

Do you have a suggestion on how to fix this?

Thanks,

Hi,

Do I understand that you try to do the following decay:
decay ad > n1 n1, (n1 >nd ad, ad > mu+ mu-)

If that's the case, MadSpin is not the correct tools for this since you can not be in narrow width approximation for such decay.
if you use the following syntax:
decay ad > n1 n1, n1 >nd mu+ mu-
This would already make more sense, but this include a three-body decay that the normal mode of MadSpin can not handle.

Cheers,

Olivier

Thank you for your suggestion.
The decays is decay : ad > n1 n1, (n1 > ad nD, ad > mu+ mu-)
Would decay ad > n1 n1, n1 >nd mu+ mu mean the same thing?
If MadSpin can not handel decays like this, what would be your suggestion to get around this? what other direction should I look into?
Thanks again for helping,

The syntax
ad > n1 n1, (n1 > ad nD, ad > mu+ mu-)

means that you have one ad onshell which decays in two ad onshell plus two nD (via two n1 onshell as well_
This is not possible by conservation of energy impulsion.

The syntax:
decay ad > n1 n1, n1 >nd mu+ mu -

request the same final state but do not request the two ad to be onshell. (it still requires the two n1 to be onshell).
Now I do not know your model, so I do not know if the associated Feynman diagram is the same for both syntax.
It is possible that the second syntax has more Feynman diagram compare to your original syntax.

If MadSpin can not handel decays like this, what would be your suggestion to get around this? what other direction should I look into?

What kind of precision do you need for such decay? If you are fine with no off-shell effects and loosing spin correlation between production and decay, then you can use madspin and include the line set spinmode none in the Cards. If you need something more accurate maybe Herwig can solution.

Cheers,

Olivier

Thank you for providing this info.

I just tried your proposed syntax " decay ad > n1 n1, n1 >nd mu+ mu - " and this the error that I get when I try to generate events:

INFO: generate 100 decay event for particle ad

Command "generate_events " interrupted with error:
InvalidCmd : No particle >nd in model
Please report this bug on https://bugs.launchpad.net/mg5amcnlo
More information is found in '/afs/cern.ch/user/m/merahman/MadGraph5_largewidth/MG5_aMC_v2_4_2/MSSMD/run_01_tag_1_debug.log'.
Please attach this file to your report.

quit

But I think I do have that vertex in my model.

Here are the vertices that I added in the mode file (.fr)

(* Add new physics n1-nD-ad *)
(* LN1NDAD := gd*(neu1bar.Ga[mu].Ga[5].neuD)*AD[mu]; *)
(* Option2: use imaginary number i *)
LN1NDAD := I*gd*(neu1bar.Ga[mu].neuD)*AD[mu];
(* Add new vertex ad-mu-mu *)
LADMuMu := gd*(mbar.Ga[mu].m)*AD[mu];
(* Add new vertex qq-ad *)
LdqdqAD := gd*(dqbar.Ga[mu].dq)*AD[mu];
LuquqAD := gd*(uqbar.Ga[mu].uq)*AD[mu];
(* Add new vertex ad-n1-n1 *)
LADn1n1 := gd*(neu1bar.Ga[mu].neu1)*AD[mu];

(* Collecting all the pieces together *)
Lag := ToDirac[SolveEqMotionF[SolveEqMotionD[LVector+LChiral+LSuperW+LSoft]]] + LFeynmanGFix + LN1NDAD + LADMuMu LdqdqAD + LuquqAD + LADn1n1;

Thanks,

This is just a space missing.
 decay ad > n1 n1, n1 > nd mu+ mu-

Cheers,

Olivier

Thank you for pointing that out :) I corrected that.

I do have the line " set spinmode" in my card but I still get this error:

INFO: generate 100 decay event for particle ad
Command "generate_events " interrupted with error:
NoDiagramException : No amplitudes generated from process Process: ad > n1 n1 NP=6
   Decay: n1 > nd mu+ mu-. Please enter a valid process
Please report this bug on https://bugs.launchpad.net/mg5amcnlo
More information is found in '/afs/cern.ch/user/m/merahman/MadGraph5_largewidth/MG5_aMC_v2_4_2/MSSMD/run_03_tag_1_debug.log'.
Please attach this file to your report.
quit

does this mean madspin is unable to handle ad > n1 n1 ?

Thanks,

The line to add is
"set spinmode none"

Did you put that one? looks like you did but I want to be sure.

NoDiagramException : No amplitudes generated from process Process: ad > n1 n1 NP=6
   Decay: n1 > nd mu+ mu-. Please enter a valid process
does this mean madspin is unable to handle ad > n1 n1 ?

Please check directly in madgraph the following commands:
import model MYMODEL
generate ad > n1 n1 NP=6, n1 > nd mu+ mu-

If it fails with the same error as before, generate each part separatly
generate ad > n1 n1 NP=6
and
generate n1 > nd mu+ mu-

to see which one of the two is not implemented in your model.
MG5aMC is able to handle such decay ... if such decays is allowed in your model (i.e. if the associated vertex exists in your model)
If you want to check if a vertex exists you can do
display interactions xd n1 n1
to see all vertex containing those three particles

Cheers,

Olivier

Yes the line I used is "set spinmode none".

If by checking directly in madgrah you mean editing the proc_card.dat, I have this:

set group_subprocesses Auto
set ignore_six_quark_processes False
set loop_optimized_output True
set loop_color_flows False
set gauge unitary
set complex_mass_scheme False
set max_npoint_for_channel 0
import model MSSMD_UFO ../models/MSSMD_UFO/param_card.dat --bypass_check
define p = u c d s u~ c~ d~ s~
define j = u c d s u~ c~ d~ s~
define l+ = e+ mu+
define l- = e- mu-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
generate p p > ad
generate ad > n1 n1 NP=6, n1 > nd mu+ mu-
# Output processes to MadEvent directory
output MSSMD -nojpeg

when executed, I get this error again:

INFO: Trying process: ad > n1 n1 NP<=6 @1
Command "import /afs/cern.ch/user/m/merahman/MadGraph5_largewidth/MG5_aMC_v2_4_2/proc_card.dat" interrupted in sub-command:
"generate ad > n1 n1 NP=6, n1 > nd mu+ mu-" with error:
NoDiagramException : No amplitudes generated from process Process: ad > n1 n1 NP=6 @1
   Decay: n1 > nd mu+ mu-. Please enter a valid process
quit

When I execute the following:

set group_subprocesses Auto
set ignore_six_quark_processes False
set loop_optimized_output True
set loop_color_flows False
set gauge unitary
set complex_mass_scheme False
set max_npoint_for_channel 0
import model MSSMD_UFO ../models/MSSMD_UFO/param_card.dat --bypass_check
define p = u c d s u~ c~ d~ s~
define j = u c d s u~ c~ d~ s~
define l+ = e+ mu+
define l- = e- mu-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
generate p p > ad
generate ad > n1 n1 NP=6
# Output processes to MadEvent directory
output MSSMD -nojpeg

I get this error:

generate ad > n1 n1 NP=6
Interpreting 'NP=6' as 'NP<=6'
INFO: Trying process: ad > n1 n1 NP<=6 @1
Command "import /afs/cern.ch/user/m/merahman/MadGraph5_largewidth/MG5_aMC_v2_4_2/proc_card.dat" interrupted in sub-command:
"generate ad > n1 n1 NP=6" with error:
NoDiagramException : No amplitudes generated from process Process: ad > n1 n1 NP=6 @1. Please enter a valid process
quit

and when I do:

set group_subprocesses Auto
set ignore_six_quark_processes False
set loop_optimized_output True
set loop_color_flows False
set gauge unitary
set complex_mass_scheme False
set max_npoint_for_channel 0
import model MSSMD_UFO ../models/MSSMD_UFO/param_card.dat --bypass_check
define p = u c d s u~ c~ d~ s~
define j = u c d s u~ c~ d~ s~
define l+ = e+ mu+
define l- = e- mu-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
generate p p > ad
generate n1 > nd mu+ mu-
# Output processes to MadEvent directory
output MSSMD -nojpeg

I get:

generate n1 > nd mu+ mu-
INFO: Trying process: n1 > nd mu+ mu- NP<=6 @1
Command "import /afs/cern.ch/user/m/merahman/MadGraph5_largewidth/MG5_aMC_v2_4_2/proc_card.dat" interrupted in sub-command:
"generate n1 > nd mu+ mu-" with error:
NoDiagramException : No amplitudes generated from process Process: n1 > nd mu+ mu- NP=6 @1. Please enter a valid process
quit

Please kindly let know if I understood what you were suggesting.

Thanks,

So looks like that your "ad" particles does not couple to "n1".

Actually I also notice that even if such vertex is include in your model
ad > n1 n1 is not possible since you need n1 to be lighter than half of ad mass for that to be meaningful.

Cheers,

Olivier

Hi,

I have double check with this model:
http://feynrules.irmp.ucl.ac.be/raw-attachment/wiki/MSSMD/MSSMD_UFO.zip
and it does not contain the ad n1 n1 vertex ...
But I guess that you are not using that model.

Cheers,

Olivier

Hi Oliver,

I mentioned in one #6, maybe you missed it here are the vertices that I added to MSSMD:

Here are the vertices that I added in the mode file (.fr)

(* Add new physics n1-nD-ad *)
(* LN1NDAD := gd*(neu1bar.Ga[mu].Ga[5].neuD)*AD[mu]; *)
(* Option2: use imaginary number i *)
LN1NDAD := I*gd*(neu1bar.Ga[mu].neuD)*AD[mu];
(* Add new vertex ad-mu-mu *)
LADMuMu := gd*(mbar.Ga[mu].m)*AD[mu];
(* Add new vertex qq-ad *)
LdqdqAD := gd*(dqbar.Ga[mu].dq)*AD[mu];
LuquqAD := gd*(uqbar.Ga[mu].uq)*AD[mu];
(* Add new vertex ad-n1-n1 *)
LADn1n1 := gd*(neu1bar.Ga[mu].neu1)*AD[mu];

(* Collecting all the pieces together *)
Lag := ToDirac[SolveEqMotionF[SolveEqMotionD[LVector+LChiral+LSuperW+LSoft]]] + LFeynmanGFix + LN1NDAD + LADMuMu LdqdqAD + LuquqAD + LADn1n1;

Are these vertices look good to you?

But you point about ad mass is valid I think. I will adjust n1 mass and try again. Will let you know the result.

Thanks,

I have seen that but this is the input of FeynRules, and I do not have experience in that code.

Looks like you still do not have that interaction in your UFO model.

Cheers,

Olivier

> On 28 Feb 2019, at 23:47, mehdirahmani <email address hidden> wrote:
>
> Question #678792 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/678792
>
> Status: Answered => Open
>
> mehdirahmani is still having a problem:
> Hi Oliver,
>
> I mentioned in one #6, maybe you missed it here are the vertices that I
> added to MSSMD:
>
> Here are the vertices that I added in the mode file (.fr)
>
> (* Add new physics n1-nD-ad *)
> (* LN1NDAD := gd*(neu1bar.Ga[mu].Ga[5].neuD)*AD[mu]; *)
> (* Option2: use imaginary number i *)
> LN1NDAD := I*gd*(neu1bar.Ga[mu].neuD)*AD[mu];
> (* Add new vertex ad-mu-mu *)
> LADMuMu := gd*(mbar.Ga[mu].m)*AD[mu];
> (* Add new vertex qq-ad *)
> LdqdqAD := gd*(dqbar.Ga[mu].dq)*AD[mu];
> LuquqAD := gd*(uqbar.Ga[mu].uq)*AD[mu];
> (* Add new vertex ad-n1-n1 *)
> LADn1n1 := gd*(neu1bar.Ga[mu].neu1)*AD[mu];
>
> (* Collecting all the pieces together *)
> Lag := ToDirac[SolveEqMotionF[SolveEqMotionD[LVector+LChiral+LSuperW+LSoft]]] + LFeynmanGFix + LN1NDAD + LADMuMu LdqdqAD + LuquqAD + LADn1n1;
>
> Are these vertices look good to you?
>
>
> But you point about ad mass is valid I think. I will adjust n1 mass and try again. Will let you know the result.
>
> Thanks,
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thanks for checking the my lines. My assumption was the lines:

(* Add new vertex ad-n1-n1 *)
LADn1n1 := gd*(neu1bar.Ga[mu].neu1)*AD[mu];

adds the ad n1 n1 vertex. I don't know why it not operating as it should. Can you introduce me to someone who can help me on this issue?

Thanks,

Hi,

You should contact one FeynRules author (by email) in order to have an answer on that.

Cheers,

Olivier

> On 1 Mar 2019, at 19:23, mehdirahmani <email address hidden> wrote:
>
> Question #678792 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/678792
>
> Status: Answered => Open
>
> mehdirahmani is still having a problem:
> Thanks for checking the my lines. My assumption was the lines:
>
> (* Add new vertex ad-n1-n1 *)
> LADn1n1 := gd*(neu1bar.Ga[mu].neu1)*AD[mu];
>
> adds the ad n1 n1 vertex. I don't know why it not operating as it
> should. Can you introduce me to someone who can help me on this issue?
>
> Thanks,
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.