MuonGeoModel and MuonSurfaces

1
MuonGeoModel and MuonSurfaces

• MuonGeoModel-00-02-80
• For TGC, RPC and CSC 2 (one for the phi view,
  one for the eta view) plane surfaces are
  associated to each gas gap Rectangular bounds
  are used for RPC, Trapezoidal Bounds (Rotated
  Trapezoidal Bounds) for phi (eta) surfaces in TGC
  and CSC
• Transforms are defined in such a way that
• for a phi strip TrklocalPosition is (xvarying
  with strip , 0const Local Eta)
• for a eta strip TrklocalPosition is (yvarying
  with strip , 0const Local Phi)
• This (MuonGeoModel-00-02-80) is a follow up of
  the most urgent requirements of generic track
  fitters discussed on the common tracking meeting
  of Oct 17, 2006 slides from Ed (pages 11-13)
  two surfaces approach
• http//indico.cern.ch/conferenceDisplay.py?confId
  7400
• In the previous discussion the possibility to
  benefit of the use of LocalParameters (where the
  parameter definition is strictly necessary, no
  need to rely on conventions like phi is first or
  the measurement is first) was raised for
  several reasons, it was felt the proposal was not
  convincing enough !
• SinStereo a somewhat less urgent business
  discussed in the same meeting
• TGC phi strips are not parallel to each other
  but almost pointing
• The various proposals are collected in
  http//indico.cern.ch/conferenceDisplay.py?confId
  7400 (pages 8-10)
• sinStereo as a methid of TrkDetElementBase
• Rotated error matrix
• Disc surface with trapezoidal bounds for TGC phi
  strip position represented by phi and 1D error
  matrix (this was considered the most elegant
  solution)

2
MuonGeoModel and MuonSurfaces

• Ive been trying to move for TGC from plane
  surfaces with trapezoidal bounds to Disc Surfaces
  with trapezoidal bounds
• The difficulties are
• Disc surfaces have never really been used as
  surfaces associated to detector elements
  (however, they are used a lot in the description
  of cylinders in the Muon Tracking Geometry by
  Sharka sw needed for holes-on-track, for
  example)
• Some extra development is needed in DiscSurface
  to allow the Disc reference point (center of the
  cylindrical symmetry) to be different that the
  center of the surface
• Some extra discussion is needed because the
  LocalPosition of a phi strip will be given by
  (Rconstant, phivarying with strip ) i.e. the
  measurent will again come in second slot a
  preliminary collection of options and
  considerations (by Ed, Andy, Markus, Wolfgang)
• modifying the order of local coordinates for
  discs (R,phi)-gt(phi,R) looks un-natural and it
  would impact on a lot of central tracking
  software
• Working with LocalPositions for TGC in the form
  (R, phi) will need detector dependent new code
  for ROT creators and local-pattern-recognition
• A new kind of surface trapezoidal surface,
  meeting just from the design the requirements and
  constraints coming from really-trapezoidal
  surfaces ???