Change majority of rock to vac. Add back remaining rock column in front of detectors only

[jdkio]

Removes all rocks except those right in front of the detectors, and a little to the sides. It's a bit tall in Y to capture the beam upstream, which is coming down at a 3 degree angle (see below). Working on tech note right now

Here is a comparison of before and after. Green is the rock layer. Note that the shade of green is different because after is using the density change from PR #46, but that isn't present in this PR. Left image is X vs Z, and right is Y vs Z.

Before

After

Where do neutrinos hit?

We can use the MC to see which interactions make particles which ultimately hit our detectors. To be safe, I defined a region encompases the entire cave (without the prism offshoot or open ceiling):

const TVector3 ND_CAVE_START(-5.0, -6.0, 0.0);
const TVector3 ND_CAVE_END(5.0, 4.0, 30.0);

If any particle (except muons) from a vertex reached that point, the neutrino vertex was included in the distributions below. Muons or secondary particles which include muon parents weren't included because they can travel much further. They are covered by the rockbox technique. This was more about establishing what the size of the anti-fiducial needed to be, including the amount of rocks on the sides. I don't think rockbox expands sideways, just in front

One interesting thing is that we get a small amount of events from the ceiling and from the opening left by prism. This makes sense as there's nothing blocking those particles. We also get some from as deep as about 10m underground. The rate is small, but I'm still working out how small. We likely will want to change our flux window

Why not rotate it 3 degrees?

The geometry is setup in a strange way. If we want to rotate and/or move the rock volume then that also rotates the ND geometry because they are nested inside. So I purposely didn't rotate anything once I figured this out

[jdkio]

The tech note is here. It's a work in progress

[ast0815]

Please don't leave any commented-out lines in the code, unless they are there for a good reason. If so, please state that reason as part of the comment. Otherwise remove them. If it is just to keep old configs for posterity, that is what the git history is for.

[alexbooth92]

This is looking great Jeffrey, excellent that there is a technote in progress as well.

Unrelated to the code change itself - when you are defining your region "encompassing the entire cave", these number are rough right, even discounting the PRISM offshoot and the funky roof shape as you have noted? I got cavern starting at ~ -30 cm rather than 0 cm in z for example.

[alexbooth92]

One more thought I just had - will this change propagate to both what we call fiducial and antificudial geometries?

[jdkio]

One more thought I just had - will this change propagate to both what we call fiducial and antificudial geometries?

Yes, WORLDggd.cfg and ND_Hall_Rock.cfg are also used by antifiducial so the change is picked up there too

Unrelated to the code change itself - when you are defining your region "encompassing the entire cave", these number are rough right, even discounting the PRISM offshoot and the funky roof shape as you have noted? I got cavern starting at ~ -30 cm rather than 0 cm in z for example.

Yes, I eyeballed the numbers from some low resolution vtx plots. I probably couldn've done that better. I can update the numbers in the code if you have a better set for X and Y as well

Please don't leave any commented-out lines in the code, unless they are there for a good reason. If so, please state that reason as part of the comment. Otherwise remove them. If it is just to keep old configs for posterity, that is what the git history is for.

Fair point, I removed all comments that didn't have a clear comment explaining them

[ast0815]

@jdkio @alexbooth92 What is this about the hall not starting at z=0 ? I always heard that the origin of the coordinate system is defined by the position where the beam axis enters the hall. Is that not true? Is this a bug in this PR or in the geometry before?

[jdkio]

What is this about the hall not starting at z=0 ? I always heard that the origin of the coordinate system is defined by the position where the beam axis enters the hall. Is that not true? Is this a bug in this PR or in the geometry before?

I heard the same thing. This PR doesn't move the location of anything, so the z position at the start is the same. It's hard to translate gdml coordinates to real coordinates so it's possible that z=0 is only roughly true. But since that's unrelated, I assume I'm okay to merge @ast0815?

[alexbooth92]

@jdkio @alexbooth92 What is this about the hall not starting at z=0 ? I always heard that the origin of the coordinate system is defined by the position where the beam axis enters the hall. Is that not true? Is this a bug in this PR or in the geometry before?

I did the working out on pen and paper to get the corners of volDetEnclosure and found the front face of that volume to be at -30.54 cm. This is in the gdml we've been using for all the recent work. Someone who knows their way around gdml better than me should definitely check though.

[ast0815]

I think those 30 cm might just be the door:

I'm happy to merge.

[alexbooth92]

I think those 30 cm might just be the door:

I'm happy to merge.

Excellent, glad there is an explanation! This is good to know moving forward as we think about the point beyond which we are using the rockbox assumption.