Monte Carlo Simulation

In analysing the performance of the CTD SLT we have made extensive use of MC simulated data. Using the LEPTO event generator, 1000 NC and 1000 CC events were generated. Default values were taken for the LEPTO parameters but a greater than 100 GeV was required and the parton cascade model for the QCD evolution was used. In addition, BG background was simulated by generating 2000 pp interactions using the FRITIOF generator.

To simulate the effects of materials and the responses of the CTD we have tracked the generated events using the ZEUS trigger MC [4]. All ZEUS components in the MC were turned on and the shower terminators were used. To simulate the BG background, pp events were generated uniformly between -9 m and +1 m along the beam line. This beam-line region gave a rate for at least one hit in the CTD of about 60 kHz [2].

The simulation of the CTD responses uses a planar drift approximation. Planar drift lines (PDL) are lines through each sense wire parallel to the tangent to the circle which is centred on the axis of the CTD and through the cell centre. Tracks crossing these lines are considered to give hits on the corresponding sense wire in the planar drift approximation.

The information recorded for each hit is as follows.

• The number of the superlayer, cell, and sense wire containing the hit.
• The spatial co-ordinates of the intersection of the PDL with the track trajectory.
• The distance from the sense wire to the intersection of the PDL and the track, along the PDL.
• The mean energy loss of the particle in the region nearest the hit sense wire.
• A pointer to the track bank giving information about the track that generated the hit.
• A sign representing the truth of which side of the wire the track was on.
• A flag indicating the usefulness of the planar drift approximation.
• The slew time (the particle time-of-flight minus a reference time-of-flight).

The spatial co-ordinates, track pointer, and drift sign are used as truth for efficiency studies. The track pointer can be used to form a link back to the generated track, vertex, and event kinematics. The final drift distance is obtained by calculating the drift time from the drift distance along the PDL and the drift velocity of the gas, adding the particle time-of-flight, and subtracting the reference time-of-flight. This total time is then converted back to a drift distance using the drift velocity of the gas. The energy loss was not used in the studies presented in this note.

The output of the trigger MC is written to a disk file in ASCII format for further processing. This data includes the response information from all ZEUS components in the MC and the above responses from the CTD. We will refer to the above information as hit data. The hit data does not include a simulation of the CTD electronics, inefficiencies, noise, or resolutions.

The CTD hits are further processed by the ZGANA package [5]. The processing steps include

1. adding noise hits,
2. smearing hit information with nominal design resolutions,
3. adding systematic uncertainties,
4. deleting hits due to inefficiencies,
5. parameterising the CTD and calorimeter first level triggers, and
6. converting the ASCII format of the CTD information to BOS banks.

The first four steps are described below, while the last two steps are described in the next section.

The detector hits and tracks are read into common-blocks using the ZGANA-reading package. Noise hits from the uranium surrounding the CTD are distributed in the CTD according to the predicted distribution and rate [6]. No synchrotron-radiation noise is simulated. The drift distances are smeared with a gaussian distribution of 130  m width and signal propagation delays along the wires are added to the drift times. The z-by-timing and energy loss are also smeared but are not of importance here. Next, 1% of the hits are deleted representing dead or inefficient channels. All inefficiencies, noise, and resolutions are parameters of the simulation and can be varied. Also, the CTD layer structure of the hits are mapped onto the readout channels in the FADC crates, and dead boards or crates can be simulated. This was not done in the studies presented here. All the above information is placed in an additional CTD common-block and is referred to as the CTD digitisation data.