CADMesh overlap warning, maybe mesh problem

Please fill out the following information to help in answering your question, and also see tips for posting code snippets. If you don’t provide this information it will take more time to help with your problem!

Geant4 Version:11.4.0
Operating System:windows 10 (64-bit)
Compiler/Version: MSVC(Visual Studio Build Tools 2022, MSBuild 18.0)
CMake Version: 4.1.1


Hello,

I am working on importing a TCAD-generated nanosheet FET structure into Geant4.

The device geometry is exported from TCAD region by region as SAT files. Each region is then converted to STL and imported into Geant4 using CADMesh, so that different materials can be assigned to each region.

The problem is that Geant4 reports overlap warnings between some of the imported regions. The maximum overlap is around 3.9 nm. Since the device dimensions are on the nanoscale, I am concerned that even nanometer-scale overlaps may affect the simulation results.

I suspect that the overlaps are mainly caused by the CAD-to-STL conversion and the resulting tessellated mesh. I have tried using SALOME to regenerate/remesh the geometry. This reduced the number and magnitude of the overlap warnings, but did not completely remove them.

I would like to ask for advice on the best way to handle this type of geometry import.

Specifically:

  • Is STL import via CADMesh reliable enough for nanoscale semiconductor device structures?
  • Are there recommended mesh-cleaning or remeshing procedures to prevent overlaps between adjacent regions?
  • Would it be better to use another workflow, such as GDML, STEP-to-Geant4 conversion, or reconstructing the geometry directly in Geant4?
  • Are there Geant4 tolerance settings that should be adjusted when dealing with very small geometries, or is that generally not recommended?

I would appreciate any guidance from users who have experience importing complex nanoscale CAD geometries into Geant4.

Thank you.

You should not use nanometers in the CAD that is suited for usual objects like screws. I usually enlarge everything by a factor of 10 or 100 and dilute the density to keep the results sane.