Skip to content
REMI Open Lab

Quick Start

This guide uses the high-level Simulation API for a field probe and a lumped/wire S-parameter example.

1. Minimal high-level run

Section titled “1. Minimal high-level run”
from rfx import Simulation, Box, GaussianPulse


sim = Simulation(
    freq_max=5e9,
    domain=(0.14, 0.06, 0.05),
    dx=2e-3,
    boundary="cpml",
    cpml_layers=8,
)


# A small material loss supplies a finite dissipative mechanism.
sim.add_material("slab", eps_r=2.2, sigma=0.01)
sim.add(Box((0.07, 0.0, 0.0), (0.09, 0.06, 0.05)), material="slab")


# Source and probe sit in the interior, clear of the CPML absorber.
sim.add_source(
    (0.03, 0.03, 0.025),
    "ez",
    waveform=GaussianPulse(f0=3e9, bandwidth=0.8),
)
sim.add_probe((0.11, 0.03, 0.025), "ez")


result = sim.run(until_decay=1e-3)
modes = result.find_resonances(freq_range=(1e9, 4e9))

This open-domain pulse example demonstrates the run and resonance-analysis APIs; it does not guarantee a resonant mode in the selected band, so modes may be empty. For a structure that does return modes, treat each Q as a finite-window estimate. Repeat with longer run lengths and finer meshes, and include every physical radiation, dielectric, conductor, and load-loss mechanism needed by the intended Q definition. Use the cavity validation workflow when Q itself is the target observable.

Use add_source() for unloaded resonance studies.
Use add_port() when you actually want S-parameters against a reference impedance. Treat quickstart S-parameters as an API example; the S-parameter support matrix states the validated configurations and remaining limits.

2. Minimal port / S-parameter run

Section titled “2. Minimal port / S-parameter run”
import numpy as np
from rfx import Simulation, Box, GaussianPulse


sim = Simulation(
    freq_max=5e9,
    domain=(0.10, 0.04, 0.02),
    dx=1.5e-3,
    boundary="cpml",
    cpml_layers=8,
)
sim.add_material("dielectric", eps_r=4.0, sigma=0.01)
sim.add(Box((0.03, 0.0, 0.0), (0.05, 0.04, 0.02)), material="dielectric")


sim.add_port(
    position=(0.01, 0.02, 0.01),
    component="ez",
    impedance=50.0,
    waveform=GaussianPulse(f0=3e9, bandwidth=0.8),
)


preflight = sim.preflight()
print(preflight.format())
preflight.raise_for_failure()


result = sim.run(n_steps=800, compute_s_params=True)
s11 = result.s_params[0, 0, :]
s11_db = 20 * np.log10(np.abs(s11) + 1e-12)

3. Auto-configured thin-substrate setup

Section titled “3. Auto-configured thin-substrate setup”
from rfx import Box, auto_configure


geometry = [
    (Box((0, 0, 35e-6), (0.06, 0.06, 1.635e-3)), "substrate"),
    (Box((0, 0, 0), (0.06, 0.06, 35e-6)), "ground"),
]
materials = {
    "substrate": {"eps_r": 4.4, "sigma": 0.025},
    "ground": {"eps_r": 1.0, "sigma": 5.8e7},
}


cfg = auto_configure(
    geometry,
    freq_range=(1e9, 4e9),
    materials=materials,
    accuracy="standard",
)


print(cfg.summary())


mesh_kwargs = cfg.to_sim_kwargs()
# mesh_kwargs is a planning result, not a complete Simulation.

When auto_configure() detects a thin z-feature, it can switch to a non-uniform z mesh via dz_profile while keeping dx/dy coarser. cfg.to_sim_kwargs() transfers mesh and boundary constructor settings only; it does not include the derived cfg.n_steps. Materials, geometry, sources, ports, and probes must be registered on the Simulation. Before registering the final geometry, align thin material boundaries with the actual dz_profile edges as shown in the Non-Uniform Mesh guide. A zero-thickness Box rasterizes no volume. Represent the 35 µm ground shape with the finite bounds above for mesh planning, then register it with sim.add_thin_conductor(..., thickness=35e-6) rather than as a volumetric material. Add the substrate, source, and observables, run preflight, and resolve every relevant issue before run(). When the planned time budget is intended, call sim.run(n_steps=cfg.n_steps) explicitly.

4. Common next steps

Section titled “4. Common next steps”