why monte-carlo ?
DESCRIPTION
Why Monte-Carlo ?. To understand and separate influences of parameters (ex: fog d, distorsions,m.i.p grain density) on efficiency To guide the optimization of reconstruction programs To evaluate track efficiency and systematics errors for different topological conditions - PowerPoint PPT PresentationTRANSCRIPT
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Why Monte-Carlo ?• To understand and separate influences of
parameters (ex: fog d, distorsions,m.i.p grain density) on efficiency
• To guide the optimization of reconstruction programs
• To evaluate track efficiency and systematics errors for different topological conditions
• ECC is the basic OPERA detector , but it is the only one escaping Monte-Carlo !
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
• THEN:The lack of data for various conditions ( particle type, angles, energies, fog densities, ambiant radioactivity integration times, etc) and poor statistics make difficult the reconstruction programs optimization.
• The large feed-back time between beam tests and scanning
• It was attracting to reproduce emulsions data by Monte-Carlo, as for the other detectors
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
• Genima GEN : GEANT3 generation of - single track (1 single field 200x200 2 , double layer emulsion set-up) - or neutrino, showers events in a complete brick set-up. - The correct grain density is set by randomly eliminating some
GEANT track hits to obtain the m.i.p density at : 32IMA: single “classico” FORTRAN program: - read GEANT hits - add background fog - constructs the 32 images of grains, fog,etc present in the 2 emulsion
layers NOW 128x128 pixels instead of 1024x1024 : factor 64 in computing time !!!
-Artificial increase of depth of field to approximate reality - clusterises the images pixels NOW Digital Filter - finds micro-track NOW cut >= 8 hits to accept - finds base-track
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
ccd plane
Focused
plane
Grain plane
d=3.254 mm d=200 mm
r=6mm
The systeme objective-ccd (in red) can move along the optical axis: we keep 16
tomographic planes ; is defined by objective characteristics
The virtual microscope
objective
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
IMA : makes pictures
• For each “grain” x,y,z we calculate the intercepted surface by a conical prolongation , isolating the surface S of pixels to “dark”. The cone aperture is determined by the lense N.A (numerical aperture):
= arcsin(N.A/nrefr )
• Pixels are filled with grey in each of the 16 picture planes; grey density varies with 1 / S and e-d/ ( is an effective emulsion attenuation lenghth)
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Fog and background
• Fog is randomly generated as single grain with a density of 8 grains/ (10 )3
• Album of real emulsions can be used also (radioactivity fossiles tracks)
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Emulsion simulation
• 16 images
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Performances of the European Scanning SystemPerformances of the European Scanning System
Efficiency
Efficiency is being estimated by European laboratories by using plates from a common test exposure
8 plates in contact were exposed to a beam
The brick was rotated in 14 different positions in order to test the angular dependency of efficiency
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Performances of the European Scanning SystemPerformances of the European Scanning System
Purity
Purity is being estimated by European laboratories measuring plates from a common reference batch with the same parameters used to measure efficiency
Tracks found by the scanning system in a few cm2 are manually checkedin order to evaluate the density of fake tracks
Density of fake tracks is less than 1/cm2
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Performances of the European Scanning SystemPerformances of the European Scanning System
Sy = 0
Sy = -0.180
Base-track
Microtrack
6/7 plates4 averages
cutBkgnd = 1 fake/cm2
Efficiency
Salerno
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Offline cut: < -6.667 + 0.4167PH
Efficiency
Napoli
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
R&D Meeting, LNGS, 4th October 2004
Base-track efficiency Jun. 04 reference brick with leadsheets 18 – 25
= N6
N5 + N6
volume 1: sheets 20 – 25volume 2: sheets 19 – 24volume 3: sheets 18 – 23
6 samples of tracks( external sheets only )
Bari
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Efficiency
Neuchatel
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Volume tracks Base tracks efficiency vs peaks (averaged over 7 plates)
Error bars = 1 spread of 7 plate efficiencies
Red: y 0 mrad
Black: y 200 mrad
Bologna
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Scanning performanceBasetracks MIC1 (SmartTracker6)
Dry objective
Bern
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Scanning performanceBasetracks MIC2 (SmartTracker7)
Dry objective
Bern
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
• How fog can influence efficiency ?- Fog effects will be higher at large angle where the
sensible volume around the track is higher- infection of a good cluster: the resulting barycentre
becomes out of association: the hits number can go down from 6 to 5
- Deformations of good clusters by a fog one can produce a deterioration of qui2
- Oppositely, on a 5 hits track, one fog cluster can, by luck, occupy the sixth rank, so the track will be accepted
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Base-track efficiency
Fog effect
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Grain
density
Fog=0 (preliminary)
Nhit 6
Nhit 5
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
• Boliden Lead + Ca (NGL washed by Goslar)• CERN refreshed emulsion (FD (6.3 + 0.5)g/(m)• Spider packing• Delrin cover and side protection• 35°C in the oven (1 year)
FD (6.8 + 0.5)g/(10m)3
Long term test using PbCa (11/2004 11/2005)
20 30 40 50 600
50
100
150
200
Track grains/100 m
Emulsion sensitivity
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
• Distortion
It is simulated by random rotation (0 to max) of the
microtracks ,leaving fixed the intercept with the basis; we varied max from 0 to 40 mrad
Emulsion layer 1 basis
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04
Jean Favier LAPP CERN Physics and soft meeting 23/04/04