Fix geometry overlaps found by geometry-doctor - #15760
Conversation
This fixes an overlap between the FT0 A-side cable container and the beam pipe. - 0CAA is a 40 x 40 x 0.5 cm air box centred on the beam axis, so it crosses the pipe wall and fills the pipe bore. - The box is now a composite shape with a 3 cm hole. The pipe outer radius there is 2.5 cm and the nearest cable sits at r = 5.68 cm. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This fixes a problem in the A-side beam pipe, whose vacuum volume occupied no space. - aluSideAVac was a daughter of aluSideA, a hollow Pcon that does not contain its own bore, so the navigator never descended into it. - The beam line from z = 44.5 cm to z = 426 cm was CAVE_Air, not PIPE_VACUUM. - The vacuum is now a daughter of the barrel. - Its bore stays at the third section's inner radius up to kZ36. The wider flange bore beyond it is occupied by the RB24 vacuum modules. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This fixes a problem in the station 1 quadrant frame, where two Inox layers carried no material. - SQ03 and SQ05toSQ07 were placed at +kHzOuterFrameInox, their own half thickness, so each sat entirely inside the Epoxy layer it is stacked on. - They are now placed at +kHzOuterFrameEpoxy, the convention SQ17to23 and SQ18to24 already use. - This affects both quadrant chambers, SQM1 and SQM2. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This gives each HMPID absorber plate its own volume name and copy number.
- createAbsorber was called twice, so both plates became a volume named Habs
placed with copy number 0.
- GetVolume("Habs") returned only the 40 mm plate, and both node paths were
/cave_1/barrel_1/Habs_0, so the 80 mm plate could not be addressed at all.
- The GDML export had to invent Habs0x2, Habs0x3 and Habs_00x1 for it.
- The plates are now Habs2 and Habs4, with copy numbers 2 and 4.
- The geometry is unchanged: same shapes, same media, same placements.
https://its.cern.ch/jira/browse/O2-4136
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
|
Very good catch. |
|
For the Alu PIPE I can understand that CheckOverlaps does not find the extrusion, because the two volumes share only a boundary. But why does it not find the complete overlap between the ft0 cable box and the PIPE ?? |
|
Sorry!, I only see now what happened. I placed the vacuum inside the alu mothervolume when restructuring the geometry and did not realise that I should have changed the inner radius of the mother volume and that way it became air. |
No worries ! The overlap checker appears to be partially blind to these defects. It pays off to have some complementary checks in the geometry-doctor now. |
3 participants
Fixes hidden material defects found by geometry-doctor