Reflector bench — geometric-optics RF ray trace

Click: select · drag: orbit · wheel: zoom · right drag: pan
G grab R rotate S scale, then XYZ to lock an axis, Shift+axis to exclude it
click to confirm · right-click / Esc to cancel · X/Del delete · Shift+D duplicate · Alt+G/R/S clear

Ready

RF

Frequency GHz

Pattern grid

Max angle °

Range dB

Edge taper dB

Weight each ray by 1/path length

Far-field pattern (u, v direction cosines)

horizontal: u (θ in plane x–z)vertical: v (θ in plane y–z)

Principal cuts

v = 0 cutu = 0 cut

Results

—

reaches planeshadowedmisses plane

Ruze analytic comparison

Compute Ruze lateral-displacement pattern Ruze 3D surface (wireframe over DFT)

Diff range ±dB

Ruze: unobstructed aperture integral, Bessel series, reflector frame

Ray-trace DFT minus Ruze. Red: DFT higher, blue: DFT lower.

main lobe (Ruze ≥ −10 dB rel. its peak)elsewhere above floor

—

Ruze’s model (IEEE Trans. AP-13, 1965) is scalar aperture theory, first order in the feed displacement: it has no occluders, no knowledge of where the aperture plane is or how it is oriented, and no missed rays. His paper treats only the lateral offset; the axial (z) part is handled here with the same first-order path expansion as a defocus phase −k·dz·cosψ. Both patterns are normalised to the ideal undisplaced, unobstructed peak, so the difference map shows shadowing loss as well as model disagreement.

Each ray that reaches the aperture plane becomes a point source with phase −k·(path length). The pattern is the coherent sum over those sources with the extra phase k·(x·u + y·v) in the plane's own frame; its +z side is the far-field hemisphere. Keep the reflector sample spacing under λ/2 or you'll see grating lobes.