Hi all,
Please use the branch name "pionmc-release-2-1" for the latest version
of BlastLib2 to crunch the pionmc Monte Carlo simulation. Also use the
branch name, "pionmc-3-0" for the latest version of Blast_Params, as
well. (The pionmc, itself, does not need a tag.) In order to get that,
please type,
% cvs -r pionmc-release-2-1 BlastLib2
% cvs -r pionmc-3-0 Blast_Params
(% indicates the shell prompt)
Once you do that for the first time, all you need to update the source
code thereafter, is
% cvs up BlastLib2
% cvs up Blast_Params
% cvs up pionmc
The latest pionmc modifications are on the coda recording, which have
been exact copy from the blastmc. However, there is a crucial problem
found on these code, which affected the event rate on our Monte Carlo.
Both blastmc and pionmc coda stream occasionally give multiple tdc and
adc data set for a single pmt, (or a single wire), which never happens
in the experimental data. As a result, the cruncher code (namely, lrn)
simply use the last entry of the tdc/adc value and ignores the valid hit
on the earlier record.
The latest version of pionmc only records the earliest tdc hits to the
coda file. As a result, the simulation gives 5 times more event rate
than before.
Because the tracking efficiency cannot exceed 1, the data event rate
*cannot* exceed the simulation event rate when simulation assumes the
perfect detection efficiencies. Currently, all of pionmc simulation
assumes the target thickness = 7x10-14 nuclei/milli-barn (Genya's
value). The findings are that this target thickness is consistent with
the simulation (simulation rate > data rate). If Genya's value is correct,
the tracking efficiency is about 67 % in the electron left and 90 % in
the electron right. This excludes the possibility that the target
thickness is less than 6.3x10-14 nuclei/milli-barn.
Because I have another tracking simulation by myself, findings from the
comparison with the geant simulation is that the most inefficiencies of
the tracking come from the low energy proton as well as the trajectories
hitting some detector insensitive regions. Simple comparison of the
event rates gives the overall tracking efficiencies of 48 % (trigger 1
ep coincidence). This value could be still too small, but I do not have
a definitive findings so far that there are still some problems either
on the simulation or on the crunch code. In other words, this may be a
reasonable value from the fact that my transportation simulation code
may be too idealistic in comparson with geant on pionmc.
Thanks so much for taking time to read my message,
Aki
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:33 EST