Disc Cherenkov Counter

-- KlausFoehl - 25 Apr 2006

• Motivation and Kinematics:
  

Particle ID cannot rely on strangeness conservation as strangeness is fed from heavier flavours, in particular charme, and the CKM matrix is non-diagonal, with both cs and cd matrix elements.

Particles from benchmark-decays go up to 1/2 to 1/3 of the momentum reached in binary decays of the p anti-p system, thus PID need not reach up to the full invariant mass (to discuss).

Detector thickness in the active angular range: 10mm (or similar), fused silica with specific weight 2.2g/cm3, about 9% radiation length. The Li_F material in the solenoid shadow requires moderate precautions against radiation, a Japanese supplier website states.

In the latest PANDA drawing (2006-4-24) there seems to be too little space foreseen to accomodate the focussing lightguide plus photon detectors part of this PID device.

The MagneticField might be too strong fro some photon detectors.

• 3.5sigma boundaries for pi-K-separation:
  
• kfpanda12l.eps: eps separation calculation

This calculation is based on an early 2005 design where the disc radius is 50% larger than the 22 degrees limit. No finished calculation yet for the current lightguide, but on the 25 degrees side I expect the black line to drop to 4GeV/c because of the disc edge being closer to the 22/25degrees radius.

Line starting at 5 degrees shows the effect of a circular hole. Green line the entire focal plane is covered with photon detector pixels, black line only part of it, and also using two different pixel sizes, alltogether 3072 pixel. Red line is lower limit for kaon ID, onset of total reflection for kaon Cherenkov light.

• design suggestion for a curved lightguide:
  
• curvedx3m05x4m01c.eps: eps figure lightguide

Fused silica (blue) disc, LiF (magenta) plate and fused silica lightguide typically 100mm thick. Side view, numbers given in mm. This version using a 4th-order polynomial focusses to better than 1mm (max distance) in the plane shown and certainly can still be optimised a wee bit further.

One surface has cylindrical (1-dim) curvature, this should be feasible to machine with CNC, all other surfaces are flat. All light reflections are total internal reflections.

There are three optical joints, two are fused silica-LiF, the third is fused silica-PMT glass. If the disc is cooled to the temperature of the surrounding detectors (EMC in particular), the expansion coefficents do not differ too much and the glue can take the strain without affecting the optical properties, transmission in particular.

The magnetic field in this area is in the order of 0.2Tesla, one has a few options in orienting the phototubes, but then magnetic shielding might be required nevertheless.

See for comparison

ProximityFocussing