This page contains links to plots showing the performance of the MC simulation and reconstruction of muon data.
Production version p15.04.00
This version has (1) the Run II Wamus geometry rather than Run I geometry, and (2) the muon system (detector and toroid) rotated, to be consistent with the reference system defined by the central tracking.
Single muons, pt=40 GeV
Tests on a sample of 2,000 single muons at pt=40 GeV (generated by Sarah Eno). Efficiencies vs eta, phi, and the quality of Muon - CD track matching are fine (compare with the p13.08 plots below).
Production version p13.08.00
The status of muon simulations in p13.08.00 is the same as in p13.03.00.
Below are links to plots for p13.08 certification samples:
5,000 "zmumu" events and 20,000 single muon events at pt=50 GeV.
(1) Note that the "zmumu" simulation
includes the continuum DY dimuon production, in addition to
the Z resonance. The mean pt of generared muons is 14 GeV.
(2) Note that the single muons were generated with eta>0, with a distribution steeply falling with eta.
zmumu
Hits in B and C layers, associated with tracks
Tracks with tr_group>0 are selected. The number of scintillator hits is plotted for tracks with the number of wire BC hits greater or equal 5.
Central track matching - test of the global geometry
Distance in the drift direction (z in Wamus, x or y in Famus) between the c.o.g. of a local muon track at the A layer and the matching ChargedParticle propagated to the A layer, in cm. In both Wamus and Famus, "deldft" is centered at 0 with the accuarcy of 1mm, and the width of the distribution is consistent with MCS for pt>15 GeV muons. The results are consistent with the perfect alignment of the muon system wrt to the central tracking system. The RMS spread due to MCS is about 50/p cm (p in GeV).
Overall performance
pt, eta and phi distribution of MC muons, within abs(eta)<2 (yellow) and of the Muonid objects with a successful global fit (Idnseg=3 and Idchisq<100). The best match is selected in case of ambiguity (Idcentralrank=1).
Single muons, pt=50 GeV
Production version p13.03.00
Results of tests on 1,300 zmumu events from the MC certification sample produced at Lancaster. Note that the "zmumu" simulation includes the continuum DY dimuon production, in addition to the Z resonance. The mean pt of generared muons is 14 GeV.
MC vs real muon detector
The previous MC releases had the muon MC geometry "as designed".
This version includes modifications in the Famus detector positions and rotations
necessary to reflect the detector "as built": the
Famus Pixels and MDT chambers have survey-based positions and small but non-zero rotation
angles wrt to the ideal values (90, 180, 270 deg).
The Wamus detector still remains as designed. The Wamus chamber positions remain as in Run I, and rotations are kept at 90 deg, etc. The four PDT chambers: 115, 116, 135 and 136 (bottom B layer) have a more complicated history. They were mounted upside down wrt the original design. The software required a simultaneous change in the simulation and in the geometry rcp used in reconstruction. Version p11 remained unchanged, version p13.03 is fully revised. Both are self consistent (i.e. OK from the user's point of view). Versions p12 and p13 prior to p13.03 are not self consistent - which means that bottom B layer PDT chambers don't contribute to reconstructed tracks. (Note: geometry used for real data has been correct since p10.13).
The Geant and the MC version of reco geometry are mutually consistent. The consistency between the geometry used in MC and in reconstruction has been checked through the detailed comparison of the position of the Geant hits and reconstructed hits with "Geantino" (Martijn Mulders).
Tests using simulated raw data
The plots below show the reconstruction results for
the MC-generated raw data. The plots show
that there are no missing muon detector elements,
hits are distributed evenly withing detector elements, and tracks
traversing the live detector are reconstructed successfully.
Also, the matching with the central track is checked to be
consistent with the ideal
cd-muon matching (no unexpected shifts) to 1 mm in the drift direction
(z in Wamus, y in Famus top/bottom and x in Famus sides).
The final results - muon reconstruction efficiency including cd matching -
in pt, eta, phi, are found to be consistent with the design.
Wire and scintillator hits - standard plots
Hits in B and C layers, associated with tracks
Tracks with tr_group>0 are selected. The number of scintillator hits is plotted for tracks with the number of wire BC hits greater or equal 5.
Distance in the drift direction (z in Wamus, x or y in Famus) between the c.o.g. of a local muon track at the A layer and the matching ChargedParticle propagated to the A layer, in cm. In both Wamus and Famus, "deldft" is centered at 0 with the accuarcy of 1mm, and the width of the distribution is consistent with MCS for pt>15 GeV muons. The results are consistent with the perfect alignment of the muon system wrt to the central tracking system. The 1 RMS spread due to MCS is about 50/p cm (p in GeV).
Overall performancept, eta and phi distribution of MC muons, within abs(eta)<2 (yellow) and of the Muonid objects with a successful global fit (Idnseg=3 and Idchisq<100). The best match is selected in case of ambiguity (Idcentralrank=1).The ratio of the reconstructed to generated muons is a measure of the combined detection/reconstruction efficiency in the muon system and in the central tracker (including the geometric acceptance).