Verification of Tracking in G4MICE
description
Transcript of Verification of Tracking in G4MICE
1
Chris Rogers
MICE VC, 15th Dec 2004
MICE Simulation Meeting 19th Jan 05
Verification of Tracking in G4MICE
2
Position - comparison with ICOOL
• Downstream x against upstream px
• Absolute value goes linearly, as we would expect• Difference between ICOOL and G4MICE also linear up to px(z = -5711) = 30 MeV
<x> = [x(ICOOL) + x(G4MICE)]/2 at z = 5711dx = [x(G4MICE) - x(ICOOL)] at z = 5711
1 mm step sizeInitially, x = (0,0,-5711); p = (px, 0, 200)Initial px plotted on y-axis
3
Position - comparison with ICOOL (redux)
• Downstream x against upstream px
• Absolute value goes linearly, as we would expect• Difference between ICOOL and G4MICE also linear up to px(z = -5711) = 30 MeV
<x> = [x(ICOOL) + x(G4MICE)]/2 at z = 5711dx = [x(G4MICE) - x(ICOOL)] at z = 5711
1 mm step sizeInitially, x = (0,0,-5711); p = (px, 0, 200)Initial px plotted on y-axis
4
Position - along z
• Again we have mean x-position and x between ICOOL and G4MICE• Note these are the absolute values• Initial px = 30 MeV (left), 100 MeV (right)• where particle crosses axis, relative difference • “Typical” difference ~1% - 10% (grows along channel)• Implications for emittance? - significant
5
Position - along z (redux)
• Again we have mean x-position and x between ICOOL and G4MICE• Note these are the absolute values• Initial px = 30 MeV (left), 100 MeV (right)• where particle crosses axis, relative difference • “Typical” difference ~0.1% at downstream end of the channel• Implications for emittance? - significant
6
Momentum - comparison with ICOOL
• Downstream px against upstream px
• Absolute value again looks quite linear• Difference between ICOOL and G4MICE also linear up to px (z = -5711) = 30 MeV
<px> = [px(ICOOL) + px(G4MICE)]/ 2 at z = 5711dpx = [px(G4MICE) - px(ICOOL)] at z = 5711
1 mm step sizeInitially, x = (0,0,-5711); p = (px, 0, 200)px plotted on y-axis
7
Momentum - comparison with ICOOL (redux)
• Downstream px against upstream px
• Absolute value again looks quite linear• Difference between ICOOL and G4MICE also linear up to px (z = -5711) = 30 MeV
<px> = [px(ICOOL) + px(G4MICE)]/ 2 at z = 5711dpx = [px(G4MICE) - px(ICOOL)] at z = 5711
1 mm step sizeInitially, x = (0,0,-5711); p = (px, 0, 200)px plotted on y-axis
8
Momentum - along z
• Here we have mean x-position and x between ICOOL and G4MICE
• Again relative difference when the muon crosses the axis
• “Typical” difference ~ 1% at 30 MeV, 10% at 100 MeV
• High transverse momentum particles are of interest for physics analysis
• Should seek to understand the discrepancy
9
Momentum - along z (redux)
• Here we have mean x-position and x between ICOOL and G4MICE
• Again relative difference when the muon crosses the axis
• “Typical” difference ~ 0.1%
• High transverse momentum particles are of interest for physics analysis
• Should seek to understand the discrepancy
10
Emittance Performance - magnets only (1 mm)
ICOOLG4MICE
• Systematic difference ~ 0.2 %• Not yet good enough; we would like this to be << 0.1 %• 1 mm step size = processor heavy
1 mm step size, ~ 1000 events
11
Emittance Offset - a bug in ecalc9?
• G4MICE emittance plots come out offset from ecalc9 plots by a small amount.– Calculate emittance in ecalc9 for some small set of
particles– Calculate emittance in G4MICE for the same set of
particles– They are different
• Take out the reference particle and the first event from the G4MICE calculation then the emittance is the same in ecalc9 and G4MICE– I think this has to be a bug in ecalc9
12
Proposed Fixes/Improvements
• Fix B-field interpolation algorithm plot– Plot decent contours
• Use the same field map
• Check emittance offset
• Fix/improve VirtualPlanes interpolation algorithm
• Plot error(step size)
• Absorbers/RF/Cooling plots