Add CLD_o2_v07

[Zehvogel]

Improves load time from 33s to 1s on my machine. :)

I am also planning to add layer assemblies for the endcaps, but that is less performance relevant. Additionally, I also want to add a new tracking region definition to exclude the LumiCal now that that is possible.

BEGINRELEASENOTES

• Added TrackerBarrel_o1_v06 using a stave assembly instead of directly placing the sensors into the layers
• Added CLD_o2_v07 using the new TrackerBarrel_o1_v06

ENDRELEASENOTES

TODO:

• Check if surfaces are still pointing in the right directions
• cleanup, add back the ZPlanarData and NeighbourSurfacesData
• Physics validation
• Document better that a module in the xml is not the same as what is called a module in the encoding/structure
• Add new tracking region
• Update README
• Run overlap check
• Repair surfaces

Evaluation results:

• Only unrelated overlaps in the LumiCal
• Tracking surface origin, u and v vector differences between old and new TrackerBarrel less than 1 fm
• Able to run reconstruction on files produced with the previous version and reproduce exactly the same digitised hits and tracks

[BrieucF]

Hi Leonhard, nice! Can you also update this README when ready?

[Zehvogel]

I added a polycone for the tracking volume to exclude the LumiCal. It should still contain all of the beam pipe seen by prompt particles. Alternatively we could use a boolean shape where just the LumiCal envelope is removed from the cylindrical tracking volume, but I guess that is slightly slower?

I am also not sure yet what the correct setting for connected would be here:

    <!--
       If the volume should be connected to the world:   connected="true"
       else if the volume is part of the parallelworld:  connected="false"
       The volume is always connected to the top level!
    -->
    <parallelworld_volume name="tracking_volume" anchor="/world" material="Air" connected="true"

[Zehvogel]

Some references for the people interested in the tracking region business (@atolosadelgado):

AIDASoft/DD4hep#1303
AIDASoft/DD4hep#384

[Zehvogel]

Size of the MCParticles collection when using Geant4TCUserParticleHandler (old) vs. Geant4TVUserParticleHandler (new) for a 100 electrons at 10-100 GeV and theta of 50-100 mrad, i.e. directly into the LumiCal. All the shower particles are no longer in the MCParticle collection :)

[andresailer]

I have most of the overlaps related to the LumiCal fixed (#376) andresailer@6e48fc5

There is one between the lumical and the innertracker envelope that I still need to work out.

[andresailer]

G4 macro

/vis/open OGLIX 1000x1000+600+400
/vis/viewer/reset
/vis/viewer/set/upVector 1 0 0 
/vis/viewer/set/viewpointVector 0 1 0  
/vis/viewer/set/sectionPlane on 0 0 0 m 0 1 0
/vis/viewer/set/style w

# focus on the lumical vs innnertracker issue
/vis/viewer/zoomTo 30 
/vis/viewer/panTo 1.08 0.16
/vis/geometry/set/forceWireframe all -1 true
/vis/drawVolume

[andresailer]

No more overlaps andresailer@df24589

[andresailer]

@Zehvogel Let's merge this, and I make a PR with my commits?

[Zehvogel]

ok