GDML external geometry loading warnings (expected?)

Dear all,
I am trying , after i debugged the GDML file i am using with the ROOT GeomManager Loader to include the full LHCb detector passive material into the AllPix2 simulation.

I get a lot of warning messages when i do so, which however are not seen if i use the ROOT geoManager to Import the gdml file.

Warnings are of the type :

|23:00:16.394| (WARNING) [I:GeometryBuilderGeant4] Caught Geant4 exception GeomVol1002: Overlap with volume already placed !
                                                             Overlap is detected for volume wrapper_detector12_phys:0 (G4Box) with _dd_Structure_LHCb_AfterMagnetRegion_T_FT0x191249b0:1514 (G4SubtractionSolid)
                                                             overlap at local point (34.4405,924.551,-326.92) by 58.258 cm  (max of 996 cases)
                                                   NOTE: Reached maximum fixed number -1- of overlaps reports for this volume !
|23:00:16.394|   (ERROR) [I:GeometryBuilderGeant4] Overlapping volumes detected.

Which back in the log gives other warnings

|23:00:10.831| (WARNING) [I:GeometryBuilderGeant4] Caught Geant4 exception GeomSolids1001: Solid - lvRich2SphMirrorHexSeg52Sub0x19287d40
                                                   All 100k attempts to generate a point on the surface have failed!
                                                   The solid created may be an invalid Boolean construct!
|23:00:03.518| (WARNING) [I:GeometryBuilderGeant4] Caught Geant4 exception GeomVol1002: Overlap with mother volume !
                                                             Overlap is detected for volume _dd_Geometry_BeforeMagnetRegion_Rich1_Rich1PmtEcrLogList_lvRichPmtStdEC0InModule119_pvRichPmtShieldingLongPlateInEC0InModule1190x18b631b0:1004 (G4Box) with its mother volume _dd_Geometry_BeforeMagnetRegion_Rich1_Rich1PmtEcrLogList_lvRichPmtStdEC0InModule1190x18b61610 (G4Box)
                                                             protrusion at mother local point (5.68314,-0.352239,-15.5) by 500 um  (max of 78 cases)
                                                   NOTE: Reached maximum fixed number -1- of overlaps reports for this volume !

I can send the full log in case needed, Apparently the simulation can run regardless as can be seen in the visualization (bad quality) picture i attach

I was wondering what are the limitations the gdml interface has compared to the TGeoManager from ROOT, if any ( like perform the subtractions of volumes etc…)

Thanks in advance

Hi @rquaglia

to be honest, I’m not much of a GDML expert. We use the GDML parser provided by Gean4, and I am not entirely sure how feature-par this is with TGeoManager. The code from our side is extremely simple and can be found here:

There is one thing that might point to our own code, the first collision you have there is with wrapper_detector12_phys - which is an object we generate. It wraps the sensor, chip and possible support layers - and sometimes is a bit too large, since we just calculate the maximum size the stack of elements has and then extend that box in all directions by that amount. Code here:

…which uses DetectorModel::getSize().

Maybe that is a good hint? Otherwise I’d need your geometry (and the GDML file) to hunt things down in more detail.

All the best,