Note
Go to the end to download the full example code.
Sweep wavelengths with a fixed sensor geometry#
Assign tabulated spectra to two regions, keep source and detector geometry fixed, and compare collected power at five wavelengths. All coefficients are illustrative. The sweep runs independent monochromatic P1 solves; it is not a broadband source integration or a validated tissue spectrum.
import os
import numpy as np
import MomentRT
from MomentRT import (
P1,
CartesianMesh,
RobinBoundary,
Simulation,
SpectralMaterial,
StructuredMedium,
SurfaceDetector,
SurfaceSource,
ureg,
)
show_figures = os.environ.get("MOMENTRT_SPHINX_GALLERY") != "1"
Define tabulated optical properties#
Tables are linearly interpolated. Out-of-range wavelengths are rejected. The refractive index is constant across the internal interface at every wavelength, as required by the current continuous-fluence solver.
table_wavelengths = np.array([650, 750, 850, 950]) * ureg.nm
background = SpectralMaterial(
phase_function=MomentRT.phase_function.HenyeyGreenstein(),
wavelengths=table_wavelengths,
absorption=np.array([0.025, 0.014, 0.018, 0.035]) / ureg.mm,
reduced_scattering=np.array([1.5, 1.3, 1.1, 0.9]) / ureg.mm,
refractive_index=1.4,
)
upper_material = SpectralMaterial(
phase_function=MomentRT.phase_function.HenyeyGreenstein(),
wavelengths=table_wavelengths,
absorption=np.array([0.06, 0.03, 0.025, 0.05]) / ureg.mm,
reduced_scattering=np.array([1.8, 1.6, 1.4, 1.2]) / ureg.mm,
refractive_index=1.4,
)
sample = StructuredMedium(
origin=(-4, -3, 0) * ureg.mm,
size=(8, 6, 4) * ureg.mm,
)
sample.add_background(material=background)
sample.add_layer(
name="Superficial region",
thickness=1 * ureg.mm,
material=upper_material,
)
Configure the source, mesh, boundaries, and detectors once#
Power is already coupled into the sample. The sweep retains that same optical power at every wavelength unless a power vector is supplied. Boundary effective reflectance and detector acceptance also stay fixed.
source = SurfaceSource(
face="z_min",
position=np.array([-1.5, 0, 0]) * ureg.mm,
radius=0.8 * ureg.mm,
wavelength=800 * ureg.nm,
optical_power=1 * ureg.mW,
)
mesh = CartesianMesh.from_sample(
sample=sample,
shape=(17, 13, 13),
)
boundary = RobinBoundary(effective_reflectance=0.4)
solver = P1(
boundaries={
face: boundary
for face in ("x_min", "x_max", "y_min", "y_max", "z_min", "z_max")
},
)
reflection_detector = SurfaceDetector(
face="z_min",
position=np.array([1.5, 0, 0]) * ureg.mm,
radius=0.75 * ureg.mm,
collection_axis=np.array([0, 0, -1]),
acceptance_angle=35 * ureg.degree,
)
transmission_detector = SurfaceDetector(
face="z_max",
position=np.array([0, 0, 4]) * ureg.mm,
radius=0.9 * ureg.mm,
collection_axis=np.array([0, 0, 1]),
acceptance_angle=60 * ureg.degree,
)
simulation = Simulation(
medium=sample,
source=source,
mesh=mesh,
solver=solver,
detectors={
"reflection": reflection_detector,
"transmission": transmission_detector,
},
)
system_figure = simulation.plot_system(
show=show_figures,
)
Run the ordered wavelength sweep#
Each result keeps its own fields, wavelength, and detector measurements. The original sample, source wavelength, and detector geometry are retained.
wavelengths = np.array([700, 750, 800, 850, 900]) * ureg.nm
spectrum = simulation.sweep(
wavelengths=wavelengths,
)
print(f"Geometry preparations: {spectrum.diagnostics['geometry_preparations']}")
print(f"Preparation time: {spectrum.diagnostics['preparation_time'].to('s'):~P}")
print(f"Per-wavelength run times: {spectrum.diagnostics['solve_times'].to('s'):~P}")
spectral_figure = spectrum.plot(
wavelength_unit="nm",
power_unit="mW",
show=show_figures,
)
for wavelength, power in zip(wavelengths, spectrum.detector_power["reflection"]):
print(f"Reflection at {wavelength:~P}: {power.to('mW'):~P}")
Geometry preparations: 1
Preparation time: 0.0901987259999828 s
Per-wavelength run times: [0.043688597000027585 0.0417160460000332 0.041540698000005705 0.042602724999994734 0.04193266100003257] s
Reflection at 700 nm: 0.0019590985219869353 mW
Reflection at 750 nm: 0.002272609112024002 mW
Reflection at 800 nm: 0.002367707455554829 mW
Reflection at 850 nm: 0.0024643911018476516 mW
Reflection at 900 nm: 0.002317102091676802 mW
Inspect one monochromatic fluence volume#
The third solve corresponds to 800 nm. Its geometry is the same as in the other solves; its optical coefficients are interpolated at this wavelength. Overlays place the instrument on the spatial field; the spectrum above keeps its wavelength and power axes. Named overlays give the two apertures distinct legend entries.
fluence_figure = spectrum.results[2].fluence.plot(
view="volume",
scale="log",
overlays=[source, reflection_detector, transmission_detector],
show=show_figures,
)
Inspect the modeled geometry#
This view shows the configured geometry and instrument footprints.
gallery_system_figure = simulation.plot_system(show=False)
Total running time of the script: (0 minutes 0.438 seconds)