1. B. Meadows University of Cincinnati
Towards FastSim fDirc
B. Meadows
University of Cincinnati
Brian Meadows
June 17, 2009

2. U. Cincinnati Simulation Studies Manpower:
Include fDirc into FastSim (in stages)
Introduce construction algorithm
 replace C by 1/ in PID selectors
Manpower:
Rolf Andreassen (postdoc 50% time)
w/ B. Meadows, M. Sokoloff
Rolf will defend his thesis by mid August.
Brian Meadows,
June 17, 2009

3. FastSim/Dirc  FastSim/fDirc Staged upgrade
Continue to use the Ring Dictionary
It generates known distribution photons for the bars we have, and will use in any reincarnation of the Dirc
Replace the SOB
Different QE
Add timing information (in addition to qC, fC)
Use a stand-alone simulation of bar geometry to understand relationship between track dip angle (l) and photon TOF that affects resolution in both qC and charged particle TOF.
Reconstruction:
Integrate timing and (qC, fC) information into b (actually, 1/b is more Gaussian)
Follow GEANT4 simulations closely for QE, resolutions in time and (qC, fC)
Brian Meadows,
June 17, 2009

4. TOF in Low Momentum Region
Possible improvement in / separation ?
BaBar Dirc
(BAD 1500)
BaBar DCH dE/dx
TOF does better than Dirc at low momentum where:
Dirc efficiency falls off AND
Difference between  and  depends more critically on measurement of |p|
Brian Meadows,
June 17, 2009

5. Rudiments of the fDirc Simplified Layout
PMT array (Hamamatsu H-8500 Flat-panel MaPMT
 resolution ~ 140 ps
Oil (or quartz ??) SOB
Focussing mirror (f~0.5 m) maps a direction onto a “point” in the detector plane.
Conventional PMT’s
 resolution ~1.6 ns
Water stand-off box
No focussing
Size of PMT window and cross-section of bar introduce uncertainties in photon direction
Brian Meadows,
June 17, 2009

6. Rudiments of the fDirc Two main accomplishments:
Reduces chromatic error (dispersion in bar)
Reduces geometric error arising from finite cross-section of bar and diameter of PMT photo-cathode.
Beam tests at ESA show that the idea works.
Projected performance
Brian Meadows,
June 17, 2009

7. Beam Tests at ESA
From JJV
Brian Meadows,
June 17, 2009

8. Chromatic Dispersion
Cherenkov angle depends on photon wavelength  because n does
 Gives ~ 4 mrad spread in c
Photons propagate distance Ldrift in bar at group velocity vgroup:
Gives spread in drift time tdrift of several 100’s of ps.
Precise measurements of tdrift improve resolution of c
the right ones.
See: NIM A533 (2005)
and: NIMA595 (2008)
Brian Meadows,
June 17, 2009

9. More Precisely … TOF resolution is not simply ps
(May well be < 50 ps though)
We do not simply measure c either.
TOF Constraint:
 same for all photons for track.
What we really measure is:
TTOT = T0 + TOF() + Tdrift ()
A fit, with constraints, can solve for 1/ and for T0.
T0 constraint:
Same value
for all tracks
This leads to an interesting interplay between TOF and c !
Brian Meadows,
June 17, 2009

10. Forward Tracks (TOF Better, c worse)
Detector
Mirror
Short photon drift
Lots of photons, BUT
 Poor chromatic correction in c
Long track:
Also, lots of photons
TOF (uncertainties in TOF from origin of photons along path in the Dirc bar) anti-correlate with those of photons drift times drift
Brian Meadows,
June 17, 2009

11. Tracks with no Dip (TOF Poorest)
Detector
Mirror
Long photon drift
 Good chromatic correction in c
Relatively poor TOF measurement:
Lots of photons and long tracks
Drift~0 while TOF is same fraction as at other dip angles.
Brian Meadows,
June 17, 2009

12. The “Open Ring Problem”
Detector
Mirror
“Open ring”:
Forward tracks have most
Photons on one side of track.
“Problem”:
If a forward track scatters,
The scatter affects c for
All photons in the same direction.
Measurement of  from TOF
Almost unaffected by the scatter.
Is relatively good wrt c for such tracks anyway.
Brian Meadows,
June 17, 2009