Setup
  (layers from the top down: the bottom one, the substrate, is the first laid down)
  the rib at D = 0.1 µm (each D alike):
    layers:
      cladding       half-space   air, n 1.000000
      film           1 µm         film, n 3.440000 in its shapes; air, n 1.000000 elsewhere
      substrate      half-space   substrate, n 3.400000
    shapes:
      on film: rectangle centred at (5, 0) µm, 10 µm along x by 3 µm along y
    rib waveguide: width 3 µm, thickness 1 µm, etch_depth 0.9 µm, sidewall 90 °, length 10 µm, loss 2.7 dB/cm, wavelength 1.15 µm, span 0.1 µm
  every D:
    slab: D = 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7 µm, the etch 1 µm − D deep; at D = 0 the strip device
    wavelength: 1.15 µm
    window: 10 µm across by 9 µm up, its half behind a vertical mirror plane
    grid: the half's 100 × 180 cells: 50 nm by 50 nm
    solver: the full-vector mode solver by Hadley's equations, TE; the device's 2D model (the effective index method) beside it
    tolerance: 2e-6 on Table I's 6 decimals
  materials:
    air: a constant index, n 1; valid 1 to 1.3 µm
      source: G. R. Hadley, J. Lightwave Technol. 20, 1219 (2002), Fig. 13, doi:10.1109/JLT.2002.800371
      n 1.000000 at 1.15 µm
    film: a constant index, n 3.44; valid 1 to 1.3 µm
      source: G. R. Hadley, J. Lightwave Technol. 20, 1219 (2002), Fig. 13, doi:10.1109/JLT.2002.800371
      n 3.440000 at 1.15 µm
    substrate: a constant index, n 3.4; valid 1 to 1.3 µm
      source: G. R. Hadley, J. Lightwave Technol. 20, 1219 (2002), Fig. 13, doi:10.1109/JLT.2002.800371
      n 3.400000 at 1.15 µm

Hadley's standard rib (3 um wide, a 1 um film of 3.44 on 3.40 under air, 1.15 um, TE) through the rib device, by Hadley's equations on a 50 nm grid (the half, 100 x 180 cells); the device's 2D model beside it
  D (um)   Table I     the device   difference   2D (EIM)    its error
  0.0      3.412022    3.412022     +2.3e-7      (a strip)
  n_eff, D = 0.0 um                      3.412022   paper 3.412022   ± 2e-6      ok
  0.1      3.412126    3.412126     +3.6e-7      3.414262    +2.1e-3
  n_eff, D = 0.1 um                      3.412126   paper 3.412126   ± 2e-6      ok
  0.2      3.412279    3.412279     +5.0e-7      3.414262    +2.0e-3
  n_eff, D = 0.2 um                      3.412279   paper 3.412279   ± 2e-6      ok
  0.3      3.412492    3.412493     +9.4e-7      3.414262    +1.8e-3
  n_eff, D = 0.3 um                      3.412493   paper 3.412492   ± 2e-6      ok
  0.4      3.412774    3.412775     +1.4e-6      3.414262    +1.5e-3
  n_eff, D = 0.4 um                      3.412775   paper 3.412774   ± 2e-6      ok
  0.5      3.413132    3.413133     +1.2e-6      3.414262    +1.1e-3
  n_eff, D = 0.5 um                      3.413133   paper 3.413132   ± 2e-6      ok
  0.6      3.413571    3.413572     +1.1e-6      3.414386    +8.1e-4
  n_eff, D = 0.6 um                      3.413572   paper 3.413571   ± 2e-6      ok
  0.7      3.414100    3.414101     +9.9e-7      3.414653    +5.5e-4
  n_eff, D = 0.7 um                      3.414101   paper 3.4141     ± 2e-6      ok
the effective index method's largest error here: 2.1e-3 (2D, printed, not checked)
all 8 within tolerance of the paper
