preliminary Simulation results

preliminary results of HBGEAN simulations


General remarks

If not stated explicitly, ARTE-01-05-r4 is used (with all the packages like Pythia and GEANT built in). The magnet is switched off (as planned for this measurements) and the full detector is used - at least as dead material, in order to provide the as much flexibillity with respect to the detector setup as possible. As soon as the the detector setup is settled, many calculations will be redone for updating the data.

B-field influence on particle distribution

Since the B field is being switched off, the particle distribution will become more narrow with respect ot the x-direction - the bending plane of the Hera-B magnet. The radial distribution and hit distributions, separated in x-direction and y-direction reveal this effect of switching off the magnet. Similar plots with the magnet switched on can be found in the proposal. It should also be stated that the peak structure is an artefact due to the histogram binning and not real physics.


Technically: The magnetic field is set in two files: gufld.F and utfeld.F. This files had to be modified (Thanks to S. Nowak for his help).


Kinematic aspects

J/psi's are produced in proton - fixed target collisions. Without applying cuts, the J/psi has an average momentum of about 100 GeV/c, an average transverse momentum of about 1.4 GeV/c with an average ratio of transverse momentum and momentum of about 0.02.

Electrons, originating in the J/psi decay, have a typical momentum of 50 GeV, a typical transverse momentum of 1.4 GeV and a ratio of transverse momentum and momentum of about 0.2.


Acceptance for J/psi decay products

The acceptance depends of course on the realised setups geomeetry. A basic assumption is that the ECAL is split in two blocks of 445mm x 672 mm each. Their y-position should be symmetrical with respect to the beam pipe and the minimum distance to the proton beam pipe on the +x-side (in Hera-B coordinate system) is (due to the electron beam pipe) 550 mm.
The geometrical single electron acceptance of ECAL is given by the ratio of electrons found in ECAL and number of generated electrons. Since this depends on the ECAL blocks x-position, the correlation between the number of electrons found in one of the ECAL blocks and their position is made. Please note that the x-position is given by the center. This means a x-position of 67 cm corresponds to an ECAL blocks ranging from 45 cm to 89 cm on +x-side as well as on -x-side.

The plot shows a decreasing single electron acceptance for electrons originating in J/psi decays as the ECAL blocks is placed more outside. For the setup with ECAL as close to the beam pipe as possible, the geometrical single electron acceptance is about 8%.
Since the J/psi decays into an electron and positron (the distinction is, from the design point of view, no necessary and will be disregarded further on), the amount of events with both electrons found in ECAL blocks is important. Triggering events with both electrons in ECAL in the favoured position at (x,y) = (750mm,0) gives a double electron acceptance of slightly below 1%

Background and Trigger studies

The FLT for this setup will mainly rely on the final detectors ECAL pretrigger system. This boards task is to identify clusters within ECAL, determine their size, sum up the clusters energy content and estimate the impact point by determining the center of gravity for each cluster. These data will be fet directly to the Track parameter unit for identifying cluster combinations with the J/psi's invariant Mass.
The trigger will presumably apply cuts on the transverse momentum, momentum and the invariant mass. Further cuts are, of course due to the detector acceptance. For estimating the trigger cuts impact on J/psi events, the first step is to look at the min. bias distributions of detected elctrons from J/psi decays: Applying trigger cuts on the transverse momentum (pt > 1 GeV) and the particles energy (E > 5 GeV), 312 out of 398 events are found. Note that there's no cut on the detector acceptance applied. Taking the detector acceptance into account, the acceptance on J/psi events is about 0.67 %.

The other important point is the behaviour of the background when applying these cuts. For these consideration is important to remind that the used simulation takes the full detector into account. This will always implied as long as it's not stated explicitly. This is important, since all subdetectors will produce secondaries and enhance pair production. Since the used setup will contain less material, the background is crucially overestimated. Anyway: If we find to be able to live with this disadvantegous condition, life will be even better with any realised detector setup.


Invariant Mass analysis

Starting with an unknown particle ID, the invariant mass calculated for all combinations of all clusters. Without any cut (neither on (transverse) momentum nor on geometrical acceptance), the invariant mass spectrum The background gets reduced by applying the FLTrigger condition as described above. This plot shows particularly in the J/spi mass region less entries than the plot before. Requiring both particles to be within ECAL, the Invariant Mass histogram using 400 J/psi events gets rather nice (but be careful: up to now I only took the electrons/positrons into account to keep the computing time reasonable. This would correspond to a perfect electron/pion separation as well as a perfect photon identification). The importance of the particle-id gets clear when looking to the particle type distribution This plot shows the number of clusters of ECAL as function of the particles mass causing this cluster using 200 events. The vast majority are photons. The same spectrum for the mother particles shows that photons do not mainly originate from particle decays, since they have the mother particle mass 0.

Doing the same job for inelastic events including the applied trigger cuts and taking the acceptance into account, one also find some entries in the invariant mass spectrum's region of the J/psi. For the 500 events used for this plot, about 50 hit combinations gave entries in this interesting region. Requiring particles with a non-zero mass and with a mother particle fulfilling the same criterion, the spectrum shows a significantly lower peak at low invariant masses, but the entries with an invariant mass near the J/psi mass are remaining. This gives evidence to the fact that these hits doesn't originates in pair conversion electrons but in directly produced, charged particles.


generated data files

For these investigations, HBGEAN was used for generating event samples with
a) events with a directly produced J/psi (1.2 kEvent)
b) minimum bias events (approx. 1.2 KEvent)
for obtaining greater flexibillity in the detector setup, the full detector in it's final design was assumed and all parts being absent are neglected in the event digitisation.
Last update:
August-25-97 by Robert Loke