FCCee MDI overlaps of 2cm detected

[neskongarr]

Check duplicate issues.

• Checked for duplicates

Goal

Just ran the overlap checker on the MDI_01_ShapeBased_v01 standalone .xml, and there are numerous overlaps of up to 2cm, see attached file.

test_overlap_v01.txt

Operating System and Version

Alma 9 (lxplus)

compiler

gcc/16.1.0-43042

The version of the key4hep stack

latest nightly (04/09/2026)

Package Version

Reproducer

The command I ran was
ddsim --compactFile $K4GEO/FCCee/MDI/compact/MDI_o1_ShapeBased_v01/MDI_o1_ShapeBased_v01_standalone.xml --runType run --macroFile /eos/user/n/nkontaxa/FCC_internship/k4geo/overlap.mac --part.userParticleHandler='' > test_overlap_v01.txt

using the macro (couln't attach the file):

/geometry/test/run
exit

Additional context

No response

[neskongarr]

While investigating the issue further, I realised the overlap warnings are caused by shield objects defined in k4geo/FCCee/MDI/compact/MDI_o1_ShapeBased_v01/Beampipe_o4_v05.xml, more particularly, objects whose detector type (i.e. which cpp driver they call) is DD4hep_Mask_o1_v01.

Implementing v02 of the MDI SR Mask inside the upstream beampipes also yielded such overlap warnings, even when no obvious overlap was made. It wasn't until I defined the SRMask (which calls the DD4hep_Mask_o1_v01 driver) as a daughter volume of the vacuum which constitutes the MDI beampipe (which calls the DD4hep_Beampipe_o1_v01), did the overlap warnings disappear.

There is thus reason to believe that some of the remaining overlaps are also caused by an ill-defined geometry hierarchy of the MDI, driven by the fact that components which should be hierarchically related to each other are instantiated using different cpp drivers.

[aciarma]

ShapeBased_v01 is a legacy version from pre-CDR studies, both the shieldings and HOMabsorbers which cause the issue are not present anymore.