Download presentation
Presentation is loading. Please wait.
Published byAustin Walsh Modified over 9 years ago
1
Marcel Stanitzki 1 More results... SiD PFA Meeting 04.06.2008 M. Stanitzki
2
Marcel Stanitzki 2 Segmentation studies First approach keep the total HCAL thickness constant vary steel thickness and number of scintillator layers Detector tags SIDish_v2_hcalXX (XX= number of layers) Second approach keep λ Ironl constant at n λ vary steel thickness and number of scintillator layers Detector tags SIDish_v2_hcalXX_lYY XX= number of layers) YY= number of lambda
3
Marcel Stanitzki 3 The variants Some Comments different Mokka compared to all other studies SIDish_v2_hcal40 is the “standard” SiDish ! λ Iron λ done with λ Iron =168 mm and λ Scint =795 mm note: there is some more material between HCAL and ECAL
4
Marcel Stanitzki 4 The results
5
Marcel Stanitzki 5 λ Iron Fixed 4.5 λ Iron SIDish_v2_hcalXX_l45
6
Marcel Stanitzki 6 Number of layers
7
Marcel Stanitzki 7 λ Total
8
Marcel Stanitzki 8 λ Iron
9
Marcel Stanitzki 9 n Layers / λ Iron
10
Marcel Stanitzki 10 Fixed total thickness SIDish_v2_hcalXX
11
Marcel Stanitzki 11 Number of layers
12
Marcel Stanitzki 12 λ Total
13
Marcel Stanitzki 13 λ Iron
14
Marcel Stanitzki 14 n Layers / λ Iron
15
Marcel Stanitzki 15 λ Iron Fixed 5.0 λ Iron SIDish_v2_hcalXX_l45
16
Marcel Stanitzki 16
17
Marcel Stanitzki 17 Playing with the ECAL Point raised by Harry, is the ECAL optimal ? we see a benefit going from 20+10 to 30+10 layers better segmentation helps ? or just pure thickness ? Effect is ~ 2 % Made a SiDish_ecal_q37 SiDish with 37 layers but same overall thickness Make a SIDish_ecal25_50 20+10 layers 2.5 mm /5.0 mm tungsten thickness and smaller gaps (1 mm) will change global radius (very small effect)
18
Marcel Stanitzki 18 Some results
19
Marcel Stanitzki 19 some plots
20
Marcel Stanitzki 20 Z dependence Taking the standard samples and looking in the forward... 0.9< cosθ Thrust < 1.0, so integrating everything in that region This is way less statistics plus there are two jets and not one well defined u-quark !
21
Marcel Stanitzki 21 Some help... cosθ=0.71 cosθ=0.92 cosθ=0.9 LDC00Sc SIDish_r125_z19 SIDish
22
Marcel Stanitzki 22 at 91 GeV
23
Marcel Stanitzki 23 at 200 GeV
24
Marcel Stanitzki 24 Energy Dependence
25
Marcel Stanitzki 25 Let's Play Fit the z and B dependence for the forward region Proposed function : For 91 GeV Fit wants no B-Field Dependence: For 200 GeV there is very weak B-Field Dependence (ignored) suggests increase with May require a few more points
26
Marcel Stanitzki 26 First result
27
Marcel Stanitzki 27 Fitting z and Energy Ignoring B... Fitting again need more points and statistics for B field Calorimeter impact These Effects are in the noise so far
28
Marcel Stanitzki 28 Z dependence (II) Due to popular request by a single gentleman Norman kindly generated u jets going at cos(θ)=0.92 for three energies: 50, 100, 250 250 GeV done for LDC00Sc and SIDish something funny, which needs cross-checking it looks like they are all over the place not only at cos(θ)=0.92
29
Marcel Stanitzki 29 Some Plots LDC00Sc SiDish
30
Marcel Stanitzki 30 250 GeV LDC00Sc SiDish
31
Marcel Stanitzki 31 Results
32
Marcel Stanitzki 32 Some plots
33
Marcel Stanitzki 33 Some plots (II)
34
Marcel Stanitzki 34 Using the other fit model Doesn't really work so well, but didn't really expect it to either as we are having 1 jet vs. two jets etc....
35
Marcel Stanitzki 35 Start from scratch Use same model as before B dependent term set to 0 For 50 GeV Jet Fit For 100 GeV Jet Fit:
36
Marcel Stanitzki 36 Results
37
Marcel Stanitzki 37 Results (II)
38
Marcel Stanitzki 38 Z dependence There is a linear dependence between energy resolution and z Both studies tell the same story a longer SiD is better For physics with two jets effect is more pronounced B field has little impact one wouldn't expect Ron's comment from Monday segmentation, radius and B field all add up here Don;t really understand the small differences between 50 and 100 GeV jets...
39
Marcel Stanitzki 39 Plans for 4.5/5.0 λ generate a 60 layer version to see a turnover effect (like for the fixed total thickness) Generate another set 5.5 λ Iron 30,40,50,60 layers 3.5 λ Iron 30,40,50,60 layers That should cover it Run a few points using 180 GeV Jets...
40
Marcel Stanitzki 40 Conclusions HCAL seems to say λ Iron is important layers/λ Iron is important need more samples ECAL prefers fine segmentation in the first layers Depth is a good thing A longer detector is better...
Similar presentations
© 2024 SlidePlayer.com Inc.
All rights reserved.