Proton ID Meeting

1
PID FPD Software Gilvan Alves (Lafex/CBPF)

• Overview
• Software Tools

Aug 28, 2003
2
The Charge

• Definition of the object class - proton
• Tools for estimating rates and efficiencies of
  the objects (p/pbar)
• Algorithms for offline reconstruction relevant
  to the object ID
• Algorithms for L1/2/3 Triggers
• Development of the analysis tools that define
  the quality of the object
• Calibration of the objects - understand
  resolutions
• Online and Offline software

3

• FPD objects
• FPD Tracks
• HIT Coincidence between U and V planes
• confirmed by a X plane
• Track Two HITs in a spectrometer (2 pots)
• constrained to the DØ
  IP.
• Halo Track
• Early-time hits in trigger scintillator (60-90
  ns)
• before crossing
• Possibly in  coincidence  with in-time hits
• in opposing spectrometer.

4
FPD Detector Hit
A fiber PLANE (U, V, X) has Two layers of
parallel fibers (U, U V, V X, X) Offset by
2/3 fiber spacing A HIT is defined as
Combination of fibers in the two LAYERS. 20
20 fibers gives 79 possible HIT values with
0.27/?12 mm resolution.
Bottom view of the detector
Scattered proton
5
FPD Detector Position
The U and V planes are oriented at 45.º A
U-HIT V-HIT determines position of a
track (There are 4,128 U-HIT V-HIT
combinations.) Signals in X and X layers used
for confirmation, ghost hit elimination, and to
measure fiber hit efficiency.
V-plane
U-plane
6
FPD Detector Track

• A TRACK is determined from the positions in two
  FPD
• detectors (straight line, with small
  correction for
• electrostatic separators and quadrupoles)
• There are (4,128)2 17 M track combinations
  (equations)
• Only some are consistent with origin at the
  IP.
• Adding DØ IP as constraint
• Determines momentum and scattering angle
• Reduces number of equations to 500 k
• Very large number of equations
• Use ranges of (?, t ) to reduce number
• One FPGA (Virtex 2000 ?) per spectrometer
• Three double-wide DFE
• We already have a full set of equations for L1
  trigger

7
(No Transcript)
8
LM and Multiple Interaction Rejection
Multiplicity more efficient than Time
9
L3 L3Unpack Tool Preliminary version
available and being tested Tracking
Tool Fast implementation of the offline
tracking Test/Prod released
Improvements in ghost tracks rejection Single
Interaction Vertex information implemented
Already in CVS Including Calorimeter
information on Elong Released for production
10
PID Certification Document
11
(No Transcript)
12
Selection Cuts for Elastic Data
13
Backgrounds
PID Criteria need to revisited for new data
14
We have the following Tools

• Generate and Process MC full chain (need to be
  tested)
• Diffractive Generators(pompyt, pomwig,
  phojet, scipyt) in the release
• FPD geometry beamline in d0gstard0sim gt
  FPDDigiChunk gt FPD_RECO
• Filter events at L3 (given L1 L2 bits are set)
• Tracking, Unpacking Single Int. Tools in the
  release (not tested on raw data)
• Process events offline (standalone now - needs
  d0reco integration)
• Unpacking - released - tested on raw data
• FPD_Reco - released - needs raw data testing
• Geometry - pre-release version
• Database access (offline)
• Control/Monitor Pot movement, HV
• Roman Pot movement - Python - stable
  version, more automation in progress
• Online Examine
• pre-release version

15
MS
CM ?
Trigsim
W V
R A
WV Wanda, Vladimir CM
Carley RA Rafael, Antonio
MS Mike Strang
16
What is (still) missing

• FPD_Reco Integration ltgt D0Reco
• Trigger Simulator ltgt Raw Data Trigger Info
• Forward Vertexing ltgt L3 Offline
• L3 Geometry Offline??
• DSTTMB info ltgt D0reco
• FPD_Analyze ltgt Offline (DSTTMB)
• Alignment Tool (offline?) ltgt Elastic stream
• Calibration Tool (offline) ltgt LMBData