Error occurs when calculating sub-GeV dark matter

Asked by ZH Cheng

when I calculate the process chi chi~ > e+ e- for the effective operator " Gf chibarchi fbarf" , when I set the dark matter mass to 100MeV, the result of maddm is:
___________________________________________________________________________________________________________________
################################################
# MadDM v. 3.2 #
################################################

################################################
# Relic Density #
################################################

Omegah2 = 6.57e+00
Omegah_Planck = 1.20e-01
xsi = 1.00e+00 # xsi = (Omega/Omega_Planck)
x_f = 1.00e+01
sigmav_xf = 3.71e-28 # cm^3 s^-1
# % of the relic density channels
%_xchixchix_ssx = 0.00 %
%_xchixchix_emep = 100.00 %
%_xchixchix_uux = 0.00 %
%_xchixchix_vtvtx = 0.00 %
%_xchixchix_ccx = 0.00 %
%_xchixchix_ttx = 0.00 %
%_xchixchix_mummup = 0.00 %
%_xchixchix_vevex = 0.00 %
%_xchixchix_ddx = 0.00 %
%_xchixchix_bbx = 0.00 %
%_xchixchix_vmvmx = 0.00 %
%_xchixchix_tamtap = 0.00 %

################################################
# Indirect Detection #
################################################
# Results in brackets display [prediction, upper limit]

# Annihilation cross section computed with the method: madevent
# <sigma v>[cm^3 s^-1] of continuum spectrum final states and Fermi dSph limits (if available, else -1)
xchixchix_ddx = [0.00e+00,-1.00e+00]
xchixchix_ttx = [0.00e+00,-1.00e+00]
xchixchix_vlvl = [0.00e+00,-1.00e+00]
xchixchix_ll = [3.21e-36,-1.00e+00]
xchixchix_bbx = [0.00e+00,-1.00e+00]
xchixchix_ccx = [0.00e+00,-1.00e+00]
xchixchix_ssx = [0.00e+00,-1.00e+00]
xchixchix_tamtap = [0.00e+00,-1.00e+00]
xchixchix_uux = [0.00e+00,-1.00e+00]

# Global Fermi dSph Limit computed with pythia8 spectra
Total_xsec = [3.21e-36,-1.00e+00]
Fermi_Likelihood = -1.76e+01
Fermi_pvalue = 0.00e+00

#################################################
# CR Flux at Earth [particles cm^-2 s^-1 sr^-1) #
#################################################

# Fluxes calculated using the spectra from dragon

Flux_neutrinos_e = nan
Flux_neutrinos_mu = nan
Flux_neutrinos_tau = nan
Flux_gammas = nan
_____________________________________________________________________________________________________________________
It can be seen that there is a problem with the output energy spectrum, showing "nan".

At the same time, madevent also shows an error:
__________________________________________________________________________________________________________
<MGGenerationInfo>
# Number of Events : 100000
# Integrated weight (pb) : 69.298
</MGGenerationInfo>
<MGGenerationInfo>
# Number of Events : 100000
# Integrated weight (pb) : 3.0968e-06
</MGGenerationInfo>
</header>
<init>
52 52 5.000000e+02 5.000000e+02 0 0 0 0 -4 1
3.096800e-06 1.325700e-09 3.096800e-06 1995
<generator name='MadGraph5_aMC@NLO' version='3.3.2'>please cite 1405.0301 </generator>
</init>
<event>
 4 1995 +3.0968000e-06 2.00000000e-01 7.81860800e-03 nan
  9100001 -1 0 0 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.9999991286e-06 1.0000000002e-01 1.0000000000e-01 0.0000e+00 -1.0000e+00
 -9100001 -1 0 0 0 0 -0.0000000000e+00 -0.0000000000e+00 -1.9999991286e-06 1.0000000002e-01 1.0000000000e-01 0.0000e+00 -1.0000e+00
       11 1 1 2 0 0 -8.5974541474e-02 +2.8951572333e-02 -4.2073562742e-02 1.0000000002e-01 0.0000000000e+00 0.0000e+00 1.0000e+00
      -11 1 1 2 0 0 +8.5974541474e-02 -2.8951572333e-02 +4.2073562742e-02 1.0000000002e-01 0.0000000000e+00 0.0000e+00 1.0000e+00
</event>
<event>
 4 1995 +3.0968000e-06 2.00000000e-01 7.81860800e-03 nan
  9100001 -1 0 0 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.9999991286e-06 1.0000000002e-01 1.0000000000e-01 0.0000e+00 -1.0000e+00
 -9100001 -1 0 0 0 0 -0.0000000000e+00 -0.0000000000e+00 -1.9999991286e-06 1.0000000002e-01 1.0000000000e-01 0.0000e+00 -1.0000e+00
       11 1 1 2 0 0 -8.7280464371e-02 +3.3613981069e-02 +3.5386732257e-02 1.0000000002e-01 0.0000000000e+00 0.0000e+00 -1.0000e+00
      -11 1 1 2 0 0 +8.7280464371e-02 -3.3613981069e-02 -3.5386732257e-02 1.0000000002e-01 0.0000000000e+00 0.0000e+00 -1.0000e+00
</event>
......
___________________________________________________________________________________________________________________
You can see that "nan" appears in the event.

So of course pythia8 will also get errors:
——————————————————————————————————————————————————————————
 *------- PYTHIA Error and Warning Messages Statistics ----------------------------------------------------------*
 | |
 | times message |
 | |
 | 10 Abort from Pythia::next: reached end of Les Houches Events File |
 | 1 Info from SLHAinterface::initSLHA: No MODSEL found, keeping internal SUSY switched off |
 | 1 Info from SLHAinterface::initSLHA: importing MASS entries |
 | 1 Warning in SLHAinterface::initSLHA: ignoring DECAY tables |
 | 1 Warning in SLHAinterface::initSLHA: ignoring MASS entries |
 | 1 Warning in SLHAinterface::initSLHA: ignoring empty DECAY tables |
 | 1 Warning in SLHAinterface::initSLHA: massless particle forced stable |
 | |
 *------- End PYTHIA Error and Warning Messages Statistics ------------------------------------------------------*
     Histogram not shown since it is empty

     Histogram not shown since it is empty

     Histogram not shown since it is empty

     Histogram not shown since it is empty

     Histogram not shown since it is empty

     Histogram not shown since it is empty

     Histogram not shown since it is empty
____________________________________________________________________________________________________________________

I would like to know how to solve this problem, or whether MadDM is suitable for sub-GeV dark matter mass calculations.

Question information

Language:
English Edit question
Status:
Solved
For:
MadDM Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said :
#1

Hi,

I'm more from the madgraph side, but this nan is not really surprising since at those scale QCD is not perturbative and therefore you do indeed expect numerical issue when computing alpha_s at such scale.

So when your process produces lepton, then you obviously have no problem and you could avoid the nan by hardcoding the scale to be larger (not sure if maddm allows it or not --do you have access to the run_card of madevent?--). Off course all the QCD part (but seems to be 0 in your case) is quite problematic to model correctly in this regime.

Cheers,

Olivier

Revision history for this message
ZH Cheng (chengo) said :
#2

Thanks Olivier Mattelaer, that solved my question.