next up previous
Next: Number of contributors Up: Performance Previous: Performance

Hitfinding performance

As mentioned in the introduction, the ability of the cluster/hitfinder to identify as many real padrow crossings as psossible (while finding as few false crossings as possible), will affect all TPC and global physics analyses that follow it. Here, some of the main aspects of this ability are discussed.

The HIJING event generator was used to generate p-p and Au-Au events, which were passed through GEANT (and the MCT analysis module) and then through the TPC slow simulator TSS [1]. P10 gas settings with 512 time buckets were used in all simulations, with all other TSS switches set to default values.

For the p-p events, typically 11000 of the 69.9 million (0.16%) pixels were occupied (i.e. above threshold), arising from about 800 padrow crossings of about 40 charged particle tracks.

For the Au-Au events, pixel occupancy was about 7-8%, with about 5300 padrow crossings per sector; this would lead to about 128000 crossings for the full TPC, but we often simulated and analysed only a few sectors at a time, due to machine memory constraints. About 2000-3000 charged particles passed through the active volume of the TPC.

Michael A. Lisa
Tue Feb 6 15:49:35 EST 1996