6-disk FGT optimization for charge reconstruction of high energy forward electrons
Updated: Jan 19 09:50:05 EST 2007

All plots on this page are produced for a single electron events with pT of 30 GeV/c.

  1. Input: one-electron events.
    Generated electrons have fixed pT and zVertex of -30, 0, or +30 cm, and uniform eta and phi distribution. Sample of high pT electrons thrown from different zVertex location over eta range [1,2] are shown here: Z=-30cm , Z=0cm , Z=+30cm, or combined 3Z locations.

    The plot shows 9 electrons with fixed pT=40 GeV/c, Z={-30,0,+30}cm, and eta={1.0,1.5,2.0}

  2. Assumptions about resolution of detectors.
    * Use: Vertex, IST1+2, SSD, new FGT, truncated TPC(nHit>=5), ESMD (also try FST)
    detector assumed resolution weight of the point Remarks
    vertex 200 mu m in X,Y,Z W=1/(200 mu m)^2 added as a hit
    IST1 20 mum in r*phi
    0.5 mm in Z
    W=1/(20 mu m)^2 -
    IST2 0.5 mm in r*phi
    20 mum in Z
    W=1/(0.5 mm)^2 -
    SSD 20 mum in r*phi
    1 mm in Z
    W=1/(20 mu m)^2 -
    (FST) 20 or 60 mu m in X,Y
    10 mu in Z
    W=1/(20 or 60 mu m)^2 2 disks: 1st & 4th,
    used only in config B & C
    mock hit
    60 mu m in X,Y
    10 mu in Z
    W=1/(60 mu m)^2 6 disks *), see plot ,
    config=D uses 4 disks
    TPC 1 mm along arc
    1 mm in Z
    W=1/(1mm)^2 * drop padrow #1 and #13
    * drop hits at |Z|>197 cm
    * drop all hits if below 5
    Endcap SMD
    mock hit **)
    1.5 mm in X,Y,Z W=1/(1.5 mm)^2 at xPoint of Geant helix
    w/ SMD plane
    *) Location of the FGT disks spans the z-range of 60-150 cm,
    (i) 6 disks: zStep=18 cm, Rin(z)/cm=3+ (z-60)/4, Rout=43cm=const
    (ii) 4 disks: zStep=30 cm, Rin(z)/cm=3+ (z-60)/4, Rout=43cm=const
    **) Based on SMD response study by Jim

  3. Hit distribution (Rxy vs.eta) for 3 vertex locations. Shown for config=B,C. Config A looks the same except FST is off. Config D has only 4 FGT disks (the first and last are the same) and no FST.
    Z vertex= - 30 cm
    Z vertex= 0 cm
    Z vertex= + 30 cm

  4. The following detector configuartions have been considered
    Config FST FGT ( 60 mum ) Vertex,IST,SSD,TPC,ESMD
    A no 6 disks all ON = default
    B 2 disks, 60 mum 6 disks all ON
    C 2 disks, 20 mum 6 disks all ON
    D no 4 disks over the same z-range all ON
    E no no all ON
    E1 no no ESMD sigX,Y=15 cm

    More configurations requested by Bernd, Will --> plots are here.
    Config FST FGT 6 disks
    A1 no degrade to 80 mum
    A2 no degrade to100 mum
    A3 no degrade to 120 mum
    A4 no degrade vertex sigX,Y,Z from 0.2 --> 0.5mm
    A5 no degrade vertex sigX,Y,Z from 0.2 --> 1.0mm
    A6 no degrade only vertex sigZ from 0.2 --> 30 cm,
    assume only beamLin constrian is used. BBC could give us 30 cm vertex-Z resolution.
    A7 no degrade FGT disk efficinfy to 90%
    A8 no degrade FGT disk efficinfy to 85%

  5. Quality of the reco 3D tracks (fit of circle+line) is defined based on a difference of the reco direction of primary track at the vertex vs. Geant track direction.

    Track reco efficiency is defined as the ratio of# of reco tracks (N1) w/
    * nFitP>=5, including vertex as a hit
    * delPhi<3 mrad
    * delTheta <3 mrad
    to the # of generated electrons (N0).

    Charge reco efficiency requiers additinal
    * the sign of the reco charge is correct. Track counter is N2.
    No cut on reco pT is imposed.

  6. Resulting track, charge reco efficiency and contamination for electrons w/ pt=30 GeV/c.
    Note, we do not want to be in the magenta sqare.
    track Eff= N1/N0 charg eff = N2/N0 wrong charge (N1-N2)/N1
    Config A : 6 FGT disks, no FST
    Config B : 6 FGT disks + 2 FST disks (60 mum)
    Config C : 6 FGT disks + 2 FST disks (20 mum)
    Config D : 4 FGT disks over same z range, no FST
    Config E : no FST, no FGT
    Config E1 : no FST, no FGT, ESMDsig=15cm

  7. Conclusions
    1. 6 FGT disks provides charge reco efficiency above 80% and contamination below 20% for eta<2.0 and zVertex in [-30,+30]cm.
    2. FGT Rout=43 cm is perhaps to high, repositioning of the disk #4 to a smaller Z will allow to reduce its Rout (I hope).
    3. FST does not improve charge reco efficiency since it always overlaps with many FGT or TPC layers.
    4. 4 FGT disks reduces charge reco for eta>1.5

    PR plots:
    * comparison of charge efficiency for proposed configuration A, vs. if FGT is dropped (Conf=E), vs. if also ESMD point is degraded (conf=E1), PS version

    * Radial hits distribution vs. eta for 3 z-Vertex, config=A, PS version

    * 3D view of 6 GEM disks