Resonant beamsplitting solver limit

[StackEnjoyer]

In order to avoid infinite calculation loops under resonant conditions, the r_max kwarg limits the maximum number of traced beams per ray segment for the solve_system! interface.

However, this condition is applied to each new split beam in the beam tree, leading to no abort condition if e.g. a Fabry Perot interferometer is traced. This bug effects all subtypes of AbstractBeam.

Solution

Fix the solve_system! interface to track the depth of the split beams. Add e.g. a max_depth kwarg.

[TacHawkes]

MWE:

using GLMakie, BeamletOptics

GLMakie.activate!(; ssao=true)
const BMO = BeamletOptics

const mm = 1e-3

mirror_diameter = 25mm
cavity_length = 12mm
reflectance = 0.92
detector_offset = 70mm

input_wavelength = 632.8e-9
input_waist = 0.02mm

mirror1 = RoundThinBeamsplitter(mirror_diameter; reflectance)
mirror2 = RoundThinBeamsplitter(mirror_diameter; reflectance)

translate3d!(mirror2, [0, cavity_length, 0])

detector = Detector(40mm)
translate3d!(detector, [0, cavity_length + detector_offset, 0])

system = StaticSystem([mirror1, mirror2, detector])

beam = GaussianBeamlet([0, -detector_offset, 0], [0, 1, 0.001], input_wavelength, input_waist)

empty!(detector)
solve_system!(system, beam)

##
if length(detector.hits) > 100
    detector.hits = detector.hits[1:100];
end

x, z, I = intensity(detector; n=400, crop_factor=0.3)

##
fig = Figure(size=(720, 540))
ax = Axis(fig[1, 1]; aspect=DataAspect(), xlabel="x (mm)", ylabel="z (mm)")
hm = heatmap!(ax, x ./ mm, z ./ mm, I)
colorbar = Colorbar(fig[1, 2], hm, label="Intensity (a.u.)")

spots = spot_diagram(detector)
scatter!(ax, spots ./ mm, color=:red)

fig

[TacHawkes]

Solved by #41