[simonge]

Briefly, what does this PR introduce?

Makes the changes required to match the current EIC accelerator beamline setup in the Far Backward region. The vacuum pipe spanning the tagger drift volume has needed some significant revisions, simplifying the implementation but removing some functionality to allow the "antechamber" design.

Incorporates the volumes reviewed in #978, placed correctly.

Introduces a new script which automatically updates the xml description of the magnets, strengths and positions based on a provided twiss and survey file. (The current files have been added to the calibrations but could be removed as don't do anything other than act as an input to the script.)

Some extra tinkering of positions may be required. In particular the luminosity pipe along the z-axis doesn't extend into the B2AeR magnet. There is a small vacuum/air boundary between the tagger exit windows and the beampipe cylinder. These could be fixed in this PR or a future PR as they won't have the most significant impacts.

The Luminosity components will need their positions updated to match the new layout.

Not the cleanest PR.

What kind of change does this PR introduce?

• Bug fix (issue Update the FarBackward beamline, magnets and taggers #963)
• New feature (issue #__)
• Documentation update
• Other: __

Please check if this PR fulfills the following:

• Tests for the changes have been added
• Documentation has been added / updated
• Changes have been communicated to collaborators

Does this PR introduce breaking changes? What changes might users need to make to their code?

Hopefully nothing.

Does this PR change default behavior?

There is likely to be significant changes to the acceptance and reconstruction of the taggers.

[simonge]

Visual comparison between the CAD and simulation. Looks like definitely some positions to adjust still.

[veprbl]

Is the problem with generate-prim-file here that my hideous boolean volume has too many nested layers?

Could be. Perhaps we can use another geometry configuration to generate prim files. cc @wdconinc

[wdconinc]

Is the problem with generate-prim-file here that my hideous boolean volume has too many nested layers?

Could be. Perhaps we can use another geometry configuration to generate prim files. cc @wdconinc

Just change it to central only, maybe? But "hideous boolean volume has too many nested layers" can be a problem in its own right... regardless of whether dawn is what barfs on it first.

[veprbl]

Agreed with @wdconinc, but I think it's also fair to say, despite how much of a technical marvel dawn is, we are not looking to support it indefinitely. So I would not sink time into debugging geometries for it when you can descope.

[simonge]

Agreed with @wdconinc, but I think it's also fair to say, despite how much of a technical marvel dawn is, we are not looking to support it indefinitely. So I would not sink time into debugging geometries for it when you can descope.

Thanks, I've temporarily removed it while I debug other issues.

[simonge]

Looks like the BENCHMARKS_TAG has changed so that some benchmark jobs aren't being triggered as it doesn't contain nightly anymore.
https://github.com/eic/detector_benchmarks/blob/master/benchmarks/lowq2_reconstruction/config.yml#L31

Is there a safer approach would stop this happening again?

[veprbl]

(Was resolved in eic/detector_benchmarks#269)

We don't run extra nightly to conserve CI resource.

[simonge]

Is onnxruntime not available for python 3.13? Or am I missing something?
https://eicweb.phy.anl.gov/EIC/benchmarks/detector_benchmarks/-/jobs/7504460#L101

From looking in the containers repository it appeared the requirements for the job might already be in the container?
https://github.com/eic/detector_benchmarks/blob/master/benchmarks/lowq2_reconstruction/requirements.txt
https://github.com/eic/containers/blob/master/spack-environment/ci/spack.yaml#L63

[veprbl]

Maybe you need to update version constraints in the requirements.txt?

[github-actions]

Capybara summary for PR 1004

• rec_dis_10x100_minQ2=1000_epic_craterlake (Δ: 4577)
• rec_dis_10x100_minQ2=1_epic_craterlake (Δ: 4934)
• rec_dis_18x275_minQ2=1000_epic_craterlake (Δ: 4954)
• rec_dis_18x275_minQ2=1_epic_craterlake (Δ: 4969)
• rec_dis_5x41_minQ2=1_epic_craterlake (Δ: 4014)
• rec_e_1GeV_20GeV_epic_craterlake (Δ: 1709)
• rec_pi_1GeV_20GeV_epic_craterlake (Δ: 1527)
• sim_chargedgeantino_1GeV_20GeV_epic_craterlake (Δ: 6)
• sim_dis_10x100_minQ2=1000_epic_craterlake (Δ: 403)
• sim_dis_10x100_minQ2=1_epic_craterlake (Δ: 419)
• sim_dis_18x275_minQ2=1000_epic_craterlake (Δ: 421)
• sim_dis_18x275_minQ2=1_epic_craterlake (Δ: 421)
• sim_dis_5x41_minQ2=1_epic_craterlake (Δ: 395)
• sim_e_1GeV_20GeV_epic_craterlake (Δ: 305)
• sim_geantino_1GeV_20GeV_epic_craterlake (Δ: 3)
• sim_pi_1GeV_20GeV_epic_craterlake (Δ: 227)

Material map validation

• Validation plots
  ^{_{Last updated 2026-10-26T07:07-04:00 cb909b7}}