QCD 4b generation problem

[setesami@cern.ch]

I was trying to generate 93 93-> b bbar b bar final state in the pp collision using the sherpa. when i run the code it works to some extend and calulate the cross section up to few itteration and afterward the reported cross section is Nan due to the below error. I was wondering if this process is exist in the sherpa and in this case did i miss to set some parameters. the Run.dat is in the attachment.

Cheers, Mohsen

full optimization: ( 1m 44s elapsed / 1h 3m 41s left ) [19:26:03] 1.0548e+11 pb +- ( 7.02162e+10 pb = 66.568 % ) 30300 ( 30779 -> 97.5 % ) full optimization: ( 2m 11s elapsed / 1h 3m 32s left ) [19:26:30] 1.27585e+11 pb +- ( 6.84872e+10 pb = 53.6797 % ) 36360 ( 36979 -> 97.7 % ) full optimization: ( 2m 38s elapsed / 1h 3m 23s left ) [19:26:57] TChannelWeight: bad momenta!!!! -1 - 1 (1) 1: (43.2375,0,0,43.2375) 2: (43.2375,0,0,-43.2375) 3: (64.5984,-1.97008e-05,-1.22223e-05,21.8766) 4: (21.8766,1.97008e-05,1.22223e-05,-21.8766) TChannelWeight: bad momenta!!!! -1 - 1 (1) . . . . . . Process_Group::Differential(): Cross section is 'nan'. Phase_Space_Integrator::AddPoint(): value = -nan. Skip.

[Frank Siegert]

Hi Mohsen,

You are generating 4 massless b quarks and do not impose any jet requirements on them (in the selector section). This means the cross section has to diverge due to QCD infrared singularities.

I would advise you to study our documentation[*] for how to technically implement parton level cuts, and in general maybe consult with colleagues locally (or at an MC school) who are familiar with Monte Carlo event generation to get an idea of the basic conceptual issues involved with it.

Cheers, Frank

[*] http://sherpa.hepforge.org/doc/Sherpa.html#Selectors