update to use wave port

[FilipeFcp]

Hi all. I just updated it to use WavePort, which simplifies the workflow. @mingzhou-flex, it would be great if you could check the RF part.

---

Note

Low Risk
Documentation/example-only change in a Jupyter notebook; no production code paths, though RF numerics should be validated by an RF reviewer as noted in the PR.

Overview
The TFLN electro-optic modulator example (TFLNTidy3d.ipynb) replaces the manual RF CPW mode-solver setup with tidy3d.rf.WavePort, aligning the notebook with the microwave RF API.

get_sim() now defines a cross-section WavePort with MicrowaveModeSpec and AutoImpedanceSpec, so voltage/current integration paths and characteristic impedance are inferred from conductors on the port plane instead of being configured by hand. The RF ModeSimulation is built via wave_port.to_mode_simulation(), and the function returns sim_vpi, mzm_solver, and wave_port for downstream use and visualization.

Narrative and plotting are updated to document WavePort and overlay the port plane on the CPW cross-section; parameter sweeps still key off the mode solver produced from the port.

^{Reviewed by Cursor Bugbot for commit 1eaf163. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

Spell Check Report

TFLNTidy3d.ipynb:

Cell 3, Line 13: 'staircasing'
  > # Use permittivity staircasing for local epsilon queries
Cell 15, Line 73: 'AutoImpedanceSpec'
  > # The WavePort solves the CPW mode and, with an AutoImpedanceSpec, automatically
Cell 38, Line 1: 'voltage-dependent'
  > Next, we run the voltage-dependent mode simulations with `web.run`.

---

Checked 1 notebook(s). Found spelling errors in 1 file(s).
Generated by GitHub Action run: https://github.com/flexcompute/tidy3d-notebooks/actions/runs/28389259268

[alec-flexcompute]

Looks good @FilipeFcp - however I'm not too experienced with Waveport, so I'm interested to hear @mingzhou-flex's feedback

[mingzhou-flex]

Looks good to me. I only have one minor comment for the in-line comments above the definition of WavePort.

In general, the size of the port only needs to cover both the signal and ground electrodes plus a bit of space around them where the fields die off, and the boundary has to intersect with some conductors for AutoImpedance to work. Given that, I think it is better to explain the choise of port size, rather than directly set it to sim_vpi.size[1/2]. Setting it to sim_vpi.size[1/2] works here becasue the groun electrodes extend all the way to the boundary of the simulation.

[FilipeFcp]

Thanks @mingzhou-flex.
Can you clarify if this is what you mean:

[mingzhou-flex]

Yes.