Dear CST user hello 😊

I am working on simulating a complex 3D Fabry-Pérot micro-resonator. As a first step, I am reconstructing the cubic resonator from the CST tutorial file, while adapting it to considerations for my future model. This model consists of a vacuum cube enclosed in a thin metallic layer, similar to the setup in the CST tutorial, but modified to better align with our design goals. Since I am already encountering several challenges at this stage, I would appreciate your guidance in understanding the correct approach to achieving this.

In this first phase, I aim to reproduce the cubic model described in the CST tutorial file—a vacuum/dielectric cube covered with a thin metallic layer. However, instead of solving it using the eigenmode solver (which only provides information on the mode frequencies and field enhancement inside the structure), I would like to solve it using TLM (As advised by the CST consultant) and Planewave. This approach was previously recommended for our problem due to the high aspect ratio of the structure, its resonant nature, and its complexity, as well as the need for a broadband solution.

Future Model Expansion (Next Steps) - In the next phase, I will:

• Replace the dielectric material inside the structure.
• Add an external dielectric coating.
• Introduce a thin 2D spherical dielectric panel inside the structure.

Ultimately, my goals are to -

• Compute the power/field enhancement on the inner spherical dielectric layer.
• Analyze reflection as a function of different wavelengths.
• Design a device that enhances light locally on the inner dielectric layer, dependent on the excitation frequency.

Initial Model Setup and Challenges -

For constructing the initial model, I followed the CST tutorial file for a cubic resonator solved with eigenmode analysis. In this model, a vacuum cube is placed inside a PEC background, I chanced it to be:

• 3D solid metallic layer around the vacuum cube and defined it as a Lossy Metal, as recommended in the CST guides and here before and theoretical references. For an optimal resonator, the metal thickness should be 1-3 times the skin depth.
• Additionally, since the eigenmode solver cannot handle 2D thin panels, if such a thin layer were defined as bulk metal, it would require an extremely fine mesh, significantly increasing simulation complexity.)
• I changed the background to Normal (epsilon = 1) and kept the electric boundary conditions on all sides.

Eigenmode Solver Results

• I obtained the same results as in the CST tutorial model, confirming that my setup was correct.
• However, since eigenmode solvers do not handle external excitation, I switched to TLM solver using a Planewave excitation* along the -z direction.

TLM Solver Setup- To configure the TLM solver, I made the following adjustments:

1. Boundary conditions:

o Open (Add) along the wave propagation direction, and due to the dielectric nature of the external material.

o Open (Add) along lateral directions to allow field expansion and the dielectric nature of the external material.

o No symmetry was applied at this stage

2. Field Monitors: Defined at relevant frequencies based on eigenmode results.

3. Hexahedral Mesh:

o Increased to 20 Mesh Cells per wavelength near the structure.

o Mesh cell reduction for any edge - by Disabled the mesh cell reduction only for Lossy Metal edge – aiming to refine it for every edge.

4. TLM Solver - Special SettingsL:

o Ensured Lossy Metal is treated as Translucent.

o PBA Boundary reflections by disabling the TLM absorbing feature, to prevent excessive reflections from the boundaries.

5. Problem Observed in TLM Solver

• The simulation runs, but the field inside the structure is extremely weak.

• The field enhancement inside the resonator is much lower than expected.

• Switching to Waveguide excitation caused additional issues:

o TLM solver only excites the first mode.

o The wave did not penetrate the structure, even though it should theoretically allow easy calculation of SP.

o Additionally, CST reports that the metallic layer is not properly meshed, despite being correctly defined as Lossy Metal.

Key Questions

For any help I can attache 4 model files for reference:

I would greatly appreciate any guidance on how to properly configure this simulation for accurate field enhancement and broadband response.

Looking forward to your insights! Best regards, Gal