Created attachment 572 [details]
GDML File

This file is being imported by following code:

  G4GDMLParser parser;
  G4String filename,GDMLsolid;
  filename = "/homes/sanchitsharma/Geant4/CNEC/Thermal_CAD/source/src/final_cnec_tool.gdml";
    // Create the run manager (MT or non-MT) and make it a bit verbose.";
  parser.Read(filename);GDMLsolid = "final_cnec_tool";
  auto DummyLV2 = parser.GetVolume(GDMLsolid);
  auto tool_sol = DummyLV2->GetSolid();

  G4RotationMatrix* rottool  = new  G4RotationMatrix;
  rottool -> rotateZ(270 * degree); 

  G4ThreeVector Trantool(0. * m ,0. * m, 0. * m);
  
  G4Transform3D shift(*rottool, Trantool);

  auto tool = new G4DisplacedSolid("tool_box",tool_sol, shift);
  auto tool_log = new G4LogicalVolume(tool, Aluminum, "tool_log");
  // Place the trenches

  G4VisAttributes* greytl = new G4VisAttributes(G4Colour::Grey());
  greytl->SetVisibility(true);
  greytl->SetForceSolid(true);
  tool_log->SetVisAttributes(greytl);

  new G4PVPlacement(0, {-1.2447 * m ,0.5334 * m, -0.167 * m}, tool_log, "tool_phys", worldLog, false, 0, checkOverlaps);

Sorry, but your report is fairly incomplete and hard for us to investigate.
- Can you please clarify what you mean by "are doing just fine without the CAD model i.e. only C++ based Boolean geometry" ?  Do you have exactly the same geometry represented as C++ code, rather than the import of this rather massive (11.6 MB file) GDML tessellated geometry?  Do you expect us to debug such geometry?
- There is no information on how the world volume is being set up.
- There is no information on how to reproduce the problem.
- The way the tessellated model is placed in the world is rather peculiar. Why are you creating a displaced-solid, further transforming it, rather than simply placing it through a normal transformation (rotation+translation)?

The run-time warnings you get may mean different things, a real overlap somewhere or a badly defined tessellated structure. How did you generate the tessellated structure?

Created attachment 573 [details]
Geometry with CAD import

Created attachment 574 [details]
Geometry without tool

Created attachment 575 [details]
Geometry without CAD import

(In reply to Gabriele Cosmo from comment #2)
> Sorry, but your report is fairly incomplete and hard for us to investigate.
> - Can you please clarify what you mean by "are doing just fine without the
> CAD model i.e. only C++ based Boolean geometry" ?  Do you have exactly the
> same geometry represented as C++ code, rather than the import of this rather
> massive (11.6 MB file) GDML tessellated geometry?  Do you expect us to debug
> such geometry?
> - There is no information on how the world volume is being set up.
> - There is no information on how to reproduce the problem.
> - The way the tessellated model is placed in the world is rather peculiar.
> Why are you creating a displaced-solid, further transforming it, rather than
> simply placing it through a normal transformation (rotation+translation)?
> 
> The run-time warnings you get may mean different things, a real overlap
> somewhere or a badly defined tessellated structure. How did you generate the
> tessellated structure?

Humble apologies for the confusion caused from my end. Here are my explanations:

1) I have made geometry of an entire experimental facility in C++ code. The experimental setup has a tool for holding detectors. Only that tool is imported in CAD. If I import the tool the Geant4 output shows no Overlaps but gives the error mentioned 'track stuck' in the output. If I do not import the tool and model my detectors as if they are floating in the air, then every thing works just fine. I have attached pictures of my geometry "Geometry with CAD Import" and "Geometry without CAD Import" in this thread.

2) World volume definition:

  G4double world_hx = 26 * m;
  G4double world_hy = 26 * m;
  G4double world_hz = 26 * m;
  G4VSolid* worldBox = new G4Box("world", world_hx, world_hy, world_hz);

  G4LogicalVolume* worldLog = new G4LogicalVolume(worldBox, nist->FindOrBuildMaterial("G4_AIR"), "worldLV");
  G4VisAttributes* visAttr = new G4VisAttributes();
  visAttr->SetVisibility(true);
  worldLog->SetVisAttributes(visAttr);
 
  G4VPhysicalVolume* worldPhys = new G4PVPlacement(0, {}, worldLog, "world", 0, false, 0);

3) I have attached the GDML file and will email you the whole source file to your email so that the error can be reproduced.

4) To import the CAD solid I am using an example in which a displacement solid was used, I was just following it. 

5) I generated the .CTL file from SALOME and then converted it to GDML using an online python code "CAD to GDML converter".

I appreciate any help in this regard.

Thanks!

The geometry is wrongly setup with clear overlaps!

Please look more carefully to what gets reported:

   :
-------- WWWW ------- G4Exception-START -------- WWWW -------
** G4Exception : GeomVol1002
      issued by : G4PVPlacement::CheckOverlaps()
Overlap with volume already placed !
          Overlap is detected for volume tool_phys:0
          with Tube:0 volume's
          local point (-99.2713,22.956,1302.01), overlapping by at least: 290.97 um 
NOTE: Reached maximum fixed number -1- of overlaps reports for this volume !
*** This is just a warning message. ***
-------- WWWW -------- G4Exception-END --------- WWWW -------
Checking overlaps for volume Tube8 ... OK!
Checking overlaps for volume BshieldP ... OK!
Checking overlaps for volume BshieldP1 ... OK!
Checking overlaps for volume BshieldP2 ... OK!
Checking overlaps for volume BshieldP3 ... OK!
Checking overlaps for volume PE_1 ... OK!
Checking overlaps for volume PE_2 ... OK!
Checking overlaps for volume RoofBPE_1 ... OK!
Checking overlaps for volume RoofBPE_2 ... OK!
Checking overlaps for volume RoofBPE_3 ... OK!
Checking overlaps for volume doorBPE_1 ... OK!
Checking overlaps for volume doorBPE_2 ... OK!
Checking overlaps for volume TubeCap ... 
-------- WWWW ------- G4Exception-START -------- WWWW -------
*** G4Exception : GeomVol1002
      issued by : G4PVPlacement::CheckOverlaps()
Overlap with volume already placed !
          Overlap is detected for volume TubeCap:0
          with tool_phys:0 volume's
          local point (2658.51,-114.3,20.0837), overlapping by at least: 6.31646 cm 
NOTE: Reached maximum fixed number -1- of overlaps reports for this volume !
*** This is just a warning message. ***
-------- WWWW -------- G4Exception-END --------- WWWW -------
Checking overlaps for volume CapCover ... 
-------- WWWW ------- G4Exception-START -------- WWWW -------
*** G4Exception : GeomVol1002
      issued by : G4PVPlacement::CheckOverlaps()
Overlap with volume already placed !
          Overlap is detected for volume CapCover:0
          with tool_phys:0 volume's
          local point (2794,1.66024,-51.9401), overlapping by at least: 180.454 um 
NOTE: Reached maximum fixed number -1- of overlaps reports for this volume !
*** This is just a warning message. ***
   :

I suggest you to resolve all reported overlaps before issuing a bug-report (this case was already analysed on Hypernews with similar diagnostics...).

Few considerations in addition:
- your code loads twice the GDML model (once in the main() and once in the DetectorConstrution::construct()... why??);
- you are placing every volume in a flat way inside the world with no one single mother-daughter relationship; this, besides being inefficient can also easily expose to generation of overlaps if shells/walls are not accurately described;
- the use of a displaced-solid for placing the CAD structure is not necessary and just causes inefficiencies, as you compute the transformation (translation/rotation) on the fly at every intersection...

Created attachment 576 [details]
Qt view of the setup, after disabling few nodes...

The CAD structure appears to be totally off centre...

Created attachment 577 [details]
This gdml file does not gives any overlaps error

Created attachment 578 [details]
Output of with the previous GDML file

Respected Dr. Cosmo,

Many thanks for looking into this. I am very grateful. The tool is not supposed to be in the center due to gravity, I am confident that my placement is correct. 

Thanks!

I humbly request you to please reconsider this. I look forward for your solution. I will follow your advice regarding mother and daughter volume for this point onward.

sincere regards,
Sanchit Sharma

Please resolve the overlaps reported. The GDML you attached again is the same as the previous you provided and is the one I tested.
There is a clear overlap with the tube and cap, as reported !!
Sorry, I cannot help more and I'm closing this ticket.

Dr. Cosmo,

Thanks very much for your time.

Regards,
Sanchit Sharma