Shadowing Update 24 June 2019 Kevin Ewart and Thomas Rainbolt - - PowerPoint PPT Presentation

Shadowing Update 24 June 2019

Kevin Ewart and Thomas Rainbolt Indiana University

φ

 The photodetecting bars are located

behind the three TPC wire planes, as well as a conducting mesh

 Currently, the shadowing effects are

implemented as a position-independent 70% transmission

 This is an approximation that is known

to be inaccurate, especially near the APA plane Background

Bruce Howard (wire spacing exaggerated 10x)

 Two basic approaches to improve on the current method, both using geometric

• ptics:

− Approximation-heavy method using a modified version of the simple shadowing

formula (“approximate method”)

 Advantages: fast, easy to understand  Disadvantages: inaccurate in certain regimes

− Nearly exact (in the limit of geometric optics) method (“robust method”)

 Advantages: detailed geometric optics effects, high accuracy in most regimes  Disadvantages: slow

Background

Approximate method- diagram and methodology

Photodetector θ Light source Wires

pitch gauge

• Based loosely on Bruce Baller's

method

− lbne docdb-4134

 For each wire that goes over the

detector, use the simple shadowing formula:

 Return the average of all the

resulting transmission coefficients

θ

pitch * cos(θ)

Robust method- diagram and methodology

Photodetector Light source Wires

• For each wire that goes over the

detector, project its shadow onto the photodetector

• Then, calculate the solid angle

subtended by that shadow

–This is the same as the solid angle

subtended by the wire that intersects the projection of the photodetector

• Then return

T = 1 – (sum of wire solid angles ) / (solid angle subtended by the photodetector)

Robust method- overlapping shadows

Photodetector Light source Wires

• In extreme cases, wire shadows may

start to overlap

• If this happens, the rest of the

photodetector in that direction will be totally shadowed

• Instead of calculating per-wire

shadowing, this entire block of shadow is lumped together

Overlapping shadows Block shadow

Both methods- getting the final transmission

• First, the APA frame's shadow is projected onto the photodetector. Due to the central

support bar, this means there are two sub-bars. It is also possible in extreme cases that

• ne or both sub-bars is entirely shadowed by the frame.
• For each sub-bar that is not totally shadowed:

–Perform the chosen method for all three wire planes, as well as the horizontal and

vertical wires of the mesh (total of five planes)

–Multiply the resulting transmission coefficients to get the transmission for that sub-

bar

–Convert this final coefficient into an effective shadowed solid angle

• Then add all effective shadowed solid angles (including the shadows from the APA

frame), and divide by the total solid angle of the full photodetector.

Approximations

• We assume an infinite wire plane

–This is almost entirely accurate in the full code, due to APA frame shadowing

• We assume the change in shadowing due to the exact positions of the wires is a

negligible effect

• If the one-plane wire shadows do not overlap, two plane shadowing is EXACT
• We neglect all reflections
• We assume geometric optics is an adequate description of shadowing
• NOT currently accounting overlapping wire shadows giving non-exact two-plane

shadowing (known to be non-negligible)

• We assume triple shadowing effects are negligible

– Effect is less than 1.4%

Results

• For the following plots, we will be using the “local” coordinates of the photodetector.

The local origin is positioned on the surface of the bar, at the lower left-hand corner of the bar as defined by the figure below

Wires MUST be sloped in this direction, z-axis is mirrored if necessary to make this true. x-axis potentially mirrored so that the light source is out of the page with positive x This results in a left-handed local coordinate system (oops), which should not matter. Origin z y

Results

• Light source positioned at x=10cm,

y=variable, z=110cm (center of bar)

• φ=90° from the vertical, i.e. wires

parallel to the photodetector

• y-coordinate on the horizontal axis,

transmission coefficient for a single wire plane on the vertical axis

• Photodetector extends from y=0 to

y=8.6cm

Results

• Light source positioned at x=50cm,

y=4.3cm (center of photodetector), z=variable

• φ=0° from the vertical, i.e. wires

perpendicular to the photodetector

• z-coordinate on the horizontal axis,

transmission coefficient for a single wire plane on the vertical axis

• Photodetector extends from z=0 to

z=220cm

Results

• Light source positioned at x=10cm,

y=4.3cm (center of photodetector), z=variable

• φ=0° from the vertical, i.e. wires

parallel to the photodetector

• z-coordinate on the horizontal axis,

transmission coefficient for a single wire plane on the vertical axis

• Photodetector extends from z=0 to

z=220cm

Results

• Light source positioned at x=variable,

y=4.3cm, z=110cm (center of photodetector)

• φ=0° from the vertical, i.e. wires

parallel to the photodetector

• x-coordinate on the horizontal axis,

transmission coefficient for a single wire plane on the vertical axis

Results

• Light source positioned at x=50cm
• z-coordinate on the horizontal axis, y-

coordinate on the vertical axis, contoured by transmission coefficient

• Full simulation of 3 wire planes +

mesh + APA frame

Results

• Light source positioned at x=10cm
• z-coordinate on the horizontal axis, y-

coordinate on the vertical axis, contoured by transmission coefficient

• Full simulation of 3 wire planes +

mesh + APA frame

Results

• Deviates from currently implemented T=0.7 by 10%

–Approximate method: for x < 220cm –Robust method: for x < 210cm

• For one wire plane the two methods deviate by

– 10% from each other for x < 40cm – 1% for x < 100cm

• For 3+mesh+APA the two methods deviate by

– 10% for x < 25cm – 1% for x < 30cm

Conclusions

• A position-dependent shadowing method has been implemented
• There are significant deviations from the current implementation of position-

independent T=0.7 for most of the detector

• This is especially important for understanding Argon 39 backgrounds
• Robust method more sensitive to the shape of the photodetector

Moving forward- implementation

• Several possible methods:

–Build shadowing into the photon library

• If we are moving away from the library, not this one

–Add shadowing methods to PhotonVisibilityServiceS2 –Add shadowing as its own service, edit the relevant photon visibility services to call

it

• Unless there are other opinions, I will implement both the approximate and robust

method, and add a fcl parameter as to which, if either, to use

• Adding support for the ARAPUCAs is easy, I just have not done it yet.

Questions?