Issue with Assembly volume and g4-copynumbers

[lopezzot]

Check duplicate issues.

• Checked for duplicates

Goal

The issue is the reproducibility of g4-copynumbers for volumes placed inside an Assembly volume.

I am placing volumes inside an Assembly volume and assigning g4-copynumbers (j) to them (together with volIDs).

Assembly AssemblyEndcap("DREndcapTubes");
 for (std::size_t j = 0; j < static_cast<std::size_t>(NbOfZRot); j++) {
    // Placement of right endcap (positive z-coordinate)
    RotationZ rotz1(M_PI / 2.); 
    RotationZ rotz2(j * phi_unit); 
    RotationX rotx(M_PI);
    RotationY roty(M_PI);
    Transform3D slice_trnsform(rotz1 * rotz2 * rotx * roty,
                               Position(0, 0, (innerR)*tan(thetaB) + length / 2.));
    PlacedVolume phiERPlaced = AssemblyEndcap.placeVolume(phiERLog, j, slice_trnsform);
    // ID this volume with one number, the rotation
    phiERPlaced.addPhysVolID("stave", j);

    // Placement of left endcap (negative z-coordinate)
    RotationY rotyleft(0.);
    Transform3D slice_trnsformleft(rotz1 * rotz2 * rotx * rotyleft,
                                   Position(0, 0, -1. * ((innerR)*tan(thetaB) + length / 2.)));
    PlacedVolume phiELPlaced = AssemblyEndcap.placeVolume(phiERLog, j+NbOfZRot, slice_trnsformleft);
    phiELPlaced.addPhysVolID("stave", j+NbOfZRot);
  }  // end of slice/stave placement

when I try to access the g4-copynumber in the SDAction code with

auto StaveID = aStep->GetPreStepPoint()->GetTouchable()->GetCopyNumber(4);

this cpno is not the same I set in the geometry code. I believe g4-copynumbers for these volumes are changed by DD4hep in the conversion from TGeo to G4 geometry.

This seems similar to #1271 fixed in #1277.

Thank you.

Operating System and Version

Any

compiler

Any

ROOT Version

Any

DD4hep Version

Any

Reproducer

std::cout<<aStep->GetPreStepPoint()->GetTouchable()->GetCopyNumber()<<std::endl;

this would be different w.r.t. the g4-cpno set in the geometry construction code for volumes placed inside an Assembly.

Additional context

This was spotted while reviewing this k4geo PR key4hep/k4geo#411

[andresailer]

Hi @lopezzot
Since your reproducer is a bit in the short side...

What copy number do you see, and which copy number do you assign?

[andresailer]

I am running ctest -V -R t_ClientTests_sim_Assemblies with

index c3b734e5..33c96393 100644
--- a/DDG4/plugins/Geant4SDActions.cpp
+++ b/DDG4/plugins/Geant4SDActions.cpp
@@ -657,6 +657,8 @@ namespace dd4hep {
         //       << " edep: " << h.deposit() << std::endl ;
         void *prePV = h.volume(h.pre), *postPV = h.volume(h.post);
 
+        std::cout << "Copy Number " <<  step->GetPreStepPoint()->GetTouchable()->GetCopyNumber()<<std::endl;
+
         Geant4HitCollection* coll = sensitive->collection(0);
         /// If we are handling a new track, then store the content of the previous one.
         if ( mustSaveTrack(h.track) )  {

And that copy number comes from here

DD4hep/DDG4/src/Geant4AssemblyVolume.cpp

Line 144 in 960cfa0

numberOfDaughters + i,

[lopezzot]

I confirm #1362 fixes this issue. Thank you so much!