Celeritas is a new Monte Carlo transport code designed to accelerate scientific discovery in high energy physics by improving detector simulation throughput and energy efficiency using GPUs.
The ATLAS tile calorimeter raytrace is having some problems. The first one I identified by simplifying a detector until it was a couple of cylinder segments. Previously this failed to raytrace while crossing the inner cylinder's surface (from left to right):
This is the original CSG tree graph, where "16" is "outside TileTBEnv" and set to false:
tilecal-bug-0.pdf
Originally this simplified to:
tilecal-bug-2.pdf
which did not eliminate the inner cylinder surface 3 (which can be proven 'true' by being inside tiletbenv) due to early replacement of a node. Now it's fixed:
tilecal-bug-1-fixed.pdf
Unfortunately although the raytrace for tilecal has fewer errors, there's still a problem:
The ATLAS tile calorimeter raytrace is having some problems. The first one I identified by simplifying a detector until it was a couple of cylinder segments. Previously this failed to raytrace while crossing the inner cylinder's surface (from left to right):![image](https://github.com/celeritas-project/celeritas/assets/741229/4eb6256a-8fef-4153-9b28-fab1c08cc06e)
This is the original CSG tree graph, where "16" is "outside
TileTBEnv
" and set to false: tilecal-bug-0.pdf Originally this simplified to: tilecal-bug-2.pdf which did not eliminate the inner cylinder surface 3 (which can be proven 'true' by being inside tiletbenv) due to early replacement of a node. Now it's fixed: tilecal-bug-1-fixed.pdfUnfortunately although the raytrace for tilecal has fewer errors, there's still a problem: