Is our dE/dx simulated correctly?

Previous: dynamic range   |   Index   |   Next: ??
Zhengyun's plots here and here show a difference between low-pt Kaons and high=pt muons.
What we see the difference in dE/dx as seen in large dE/dx devices: this is from the Particle Data Group:

PDG particle detectors pdf

Also (from Dave) see slides 6,7 from this CMS ppt presentation,
and slide 19 from this Phobos ppt presentation
However, in this paper: 'Silicon detectors and elementary particle physics' on pg 303 is this figure and text 'As expected, no relativistic rise is observed'.

Also see pg 10 of the PDG section on Passage of particles through matter

Question is: what is Geant-3 doing? Section PHYS332 of the manual appears to know about all effects: thin detectors, delta rays, fluctuations and the rest, as long as parameted LOSS is set to 1, 2, or 3.

In our pisa.kumac file, LOSS is set to 2: 

 LOSS  2      ! GEANT command for Landau fluctuations on
However, the default Pisa file also turns off delta rays: 
 DRAY  0      ! GEANT command for Delta no ray (Landau is on)
which likely means that since delta rays are not produced, the energy now deposited in the silicon includes the energy which would otherwise be carried off by the deltas - causing the relativistic rise seen in our simulations so far.

---> redo simulations with DRAY=1     (1 means produce and track them)

This will likely also raise the hit density.

Pisa defaults:

Energy deposited by a 2GeV K and 20 GeV mu. Peaks are separated by ~20%.

default pisa.kumac

Turn delta rays on, and lower thresholds:

Energy deposited by a 2GeV K and 20 GeV mu. Peaks are separated by ~9%.

The energy deposited by the delta-ray electron is not in this plot. Adding it does not shift the peak, but adds a tail on the upper end.

pisa.kumac with lower thresholds and delta rays turned on (LOSS and DRAY).

The two plots superimposed.

Zhengyun will propagate this through his analysis to see how much of the previous factor ~30 improvement is left.

Estimates show that 1mm Silicon is 'not thin' in the sense of Geant simulations, ans therefore should show more of the relativistic rise and Landau tail.

The ratio of the means is still only ~9%.

Previous: dynamic range   |   Index   |   Next: ??