hotfix: remove unneeded warning with Medium2D and mode solver

[caseyflex]

Whenever a Medium2D and a ModeSolver are present, we get warnings Evaluating the permittivity of a Medium2D is unphysical. We should be checking the volumetric equivalent structure instead.

Greptile Overview

Updated On: 2025-11-06 16:49:01 UTC

Greptile Summary

Fixed spurious warnings when using Medium2D structures with ModeSolver by switching from self.simulation.structures to self.simulation.volumetric_structures in the _intersecting_media method.

• The volumetric_structures property automatically converts Medium2D structures to their 3D AnisotropicMedium equivalents
• This prevents the unphysical permittivity warning that was triggered when the mode solver accessed Medium2D.eps_diagonal() directly
• The fix is minimal (one-line change) and uses the existing infrastructure designed specifically for this conversion
• No test changes needed as this fixes a warning issue rather than functional behavior

Confidence Score: 4/5

• This PR is safe to merge with low risk, pending a changelog entry
• The change is well-targeted and uses existing infrastructure (volumetric_structures property) that was designed specifically for converting Medium2D to 3D equivalents. The fix is minimal (one line) and directly addresses the root cause of spurious warnings. However, the PR is missing a changelog entry as required by project conventions for bug fixes, which is the only reason for not scoring 5/5.
• No files require special attention - the change is straightforward and correct

Important Files Changed

File Analysis

Filename	Score	Overview
tidy3d/components/mode/mode_solver.py	5/5	Changed from self.simulation.structures to self.simulation.volumetric_structures in _intersecting_media method to use converted 3D equivalents of Medium2D structures, avoiding unphysical permittivity warnings

Sequence Diagram

sequenceDiagram
    participant MS as ModeSolver
    participant Sim as Simulation
    participant Scene
    participant Medium2D
    participant VolumetricMedium as AnisotropicMedium

    Note over MS: _intersecting_media() called
    MS->>Sim: Get background_structure
    MS->>Sim: Get volumetric_structures
    Note over Sim: Returns structures with<br/>Medium2D converted to<br/>3D volumetric equivalents
    Sim->>Sim: _volumetric_structures_grid()
    loop For each structure
        alt Structure has Medium2D
            Sim->>Medium2D: volumetric_equivalent()
            Medium2D->>VolumetricMedium: Create AnisotropicMedium
            VolumetricMedium-->>Sim: Return 3D medium
            Note over Sim: Replace Medium2D structure<br/>with volumetric equivalent
        else Structure has 3D medium
            Note over Sim: Keep structure as is
        end
    end
    Sim-->>MS: Return volumetric_structures
    MS->>Scene: intersecting_media(plane, structures)
    Scene->>VolumetricMedium: Access permittivity
    Note over Scene: No warning since<br/>using physical 3D medium
    Scene-->>MS: Return intersecting media list

Loading

[greptile-apps]

_{1 file reviewed, 1 comment}

_{Edit Code Review Agent Settings | Greptile}

[github-actions]

Diff Coverage

Diff: origin/develop...HEAD, staged and unstaged changes

• tidy3d/components/mode/mode_solver.py (0.0%): Missing lines 1917

Summary

• Total: 1 line
• Missing: 1 line
• Coverage: 0%

tidy3d/components/mode/mode_solver.py

  1913     @cached_property
  1914     def _intersecting_media(self) -> list:
  1915         """List of media (including simulation background) intersecting the mode plane."""
  1916         total_structures = [self.simulation.scene.background_structure]
! 1917         total_structures += list(self.simulation.volumetric_structures)
  1918         return self.simulation.scene.intersecting_media(self.plane, total_structures)
  1919 
  1920     @cached_property
  1921     def _has_fully_anisotropic_media(self) -> bool:

---

[momchil-flex]

I think it makes sense. But is the 2D medium included in the mode solve as well through the volumetric representation, or is it ignored?

[caseyflex]

I think it makes sense. But is the 2D medium included in the mode solve as well through the volumetric representation, or is it ignored?

It should be included in the mode solver, either via simulation.epsilon_on_grid (which explicitly uses volumetric_structures) or via the subpixel processing