IR Blocking & Blackbody Flux Calculator

A static, browser-only LCARS-themed calculator for the blackbody photon flux landing on a detector pixel through a three-stage cryogenic filter stack. A warm source (default 280 K) radiates through three cold stages (default 65 K / 3.5 K / 800 mK) onto the pixel; the tool plots counts/nm vs wavelength and reports the integrated counts/sec.

Features

• Reactive optical-bench schematic with the controls laid out per stage (temperature, aperture diameter, axial distance, substrate + thickness, coating).
• Full radiative model: the warm source and each cold stage emit as graybodies, each attenuated by the colder downstream filters; cold-stop geometry sets the étendue onto the pixel.
• Thickness-scaled substrate materials computed from first principles: N-BK7, fused silica (Suprasil / Infrasil), sapphire, MgF₂, CaF₂.
• Coatings from a measured-curve library (DARKNESS, PICTURE-C, MEC Prime, ASAHI YSC1100, ASAHI YSC0750, ITO, M254C cold mirror), plus CSV import, a Custom coating (ideal top-hat or a designed Ta₂O₅/SiO₂ multilayer), and Custom 2 — a vendor spec transcribed as Tavg ≥/≤ x% bounds per wavelength band.
• Coating designer: synthesize a real Ta₂O₅/SiO₂ stack from a passband target (transfer-matrix forward model + bounded refinement), edit the layer list by hand, and see the designed filter's throughput overlaid on the flux plot on a log right-hand axis.

Run locally

The app must be served over HTTP — ES-module imports and fetch() of the filter JSON do not work from file://.

python3 -m http.server 8000   # or: npm run serve
# open http://localhost:8000/

Tests

npm test        # node --test over test/*.test.js

The calculation

Photon spectral radiance L_ph(λ,T) = (2c/λ⁴)/(exp(hc/λk_BT) − 1); each emitter k contributes A_pix · QE · Ω_k · ε_k(λ) · (∏ downstream T_j) · L_ph(λ,T_k), where Ω_k is the projected solid angle of the limiting (cold-stop) aperture between plane k and the pixel. The per-nm spectrum is integrated (trapezoid) to counts/sec. Units and assumptions are documented inline in js/physics.js.

Optical model

Coatings and substrates share one swappable OpticalElement interface (transmission/reflection over a wavelength grid), so the flux engine never cares how an element's curve was produced:

• Substrates (js/materials.js) are computed from physical thickness: T = (1−R)²·τ(λ,d), ε = (1−R)(1−τ), with R from the Sellmeier refractive index and bulk τ from either a tabulated absorption coefficient (fused silica, sapphire, MgF₂, CaF₂) or the measured N-BK7 curve scaled from its 10 mm reference (Beer–Lambert). Each stage has a Thickness (mm) input. Outside a material's validity range (rangeNm, bounded by its UV and multiphonon IR absorption edges) the substrate is opaque — τ = 0, ε = 1 − R — rather than holding the edge value, which would turn a falling absorption edge into an infinite transmission plateau.
• Coatings are measured T(λ) curves (js/filters.js), the analytic Custom top-hat (CustomCoating), a designed multilayer (MultilayerCoating), or Custom 2 (SpecCoating) — piecewise Tavg bounds per band, where each row's stated bound is the transmission used (so the result is the worst case the spec permits), later rows override earlier ones on overlap, and wavelengths no band covers transmit fully. The editor keeps adjacent bands contiguous — editing either side of a boundary drags its partner, and removing a band closes over its span — so an open region has to be stated as a band at 100% rather than left implicit. Only λ outside the outermost bands falls through to the uncovered default.
• Multilayer designer (js/tmm.js, js/synthesis.js): a transfer-matrix forward model for Ta₂O₅/SiO₂ stacks, plus a synthesizer that builds a chirped long-wave-blocking edge filter from the Custom passband and refines the layer thicknesses (asymmetric objective: maximize in-band, suppress long-wave leakage out to the ~3.4 µm glass cutoff, leave the short side free). Layer count is adjustable; the result is demonstration-grade (see js/synthesis.js for the limits).

Filter & material library

The bundled coating curves in data/filters/*.json are checked in and are all the app needs. They were converted from raw manufacturer/measured source curves (kept outside this repo) with the py313

• numpy scripts in tools/; tools/notebook_parity_reference.py regenerates the physics parity fixture from the bundled curves. Substrate material data (Sellmeier coefficients + absorption tables, with citations) lives inline in js/materials.js. You can also import your own coating/substrate curve at runtime from a 2-column file (fraction / percent / OD).

Deploy (GitHub Pages)

Settings → Pages → Deploy from branch → main, / (root). All files are static at the repo root; no build step.