Sun position uses the textbook equations with Cooper's declination:
dec = 23.45 deg * sin(360/365 * (284 + n))
sin(alt) = sin(lat) sin(dec) + cos(lat) cos(dec) cos(H), H = 15 deg * (t - 12)
cos(A) = (sin(alt) sin(lat) - sin(dec)) / (cos(alt) cos(lat))
The direct beam is attenuated with Kasten-Young air mass and Meinel's law, DNI = 1361 * 0.7^(AM^0.678) W/m2, at about 100 lm/W. Each desk, and each point of the whiteboard, is tested by casting a ray toward the sun through the real window opening (including the 0.25 m wall reveal). The blind is a stack of slats of width w = 25 mm at pitch s = 21.5 mm; for profile angle p and slat tilt f the open fraction of the beam is
T_blind = max(0, 1 - w |sin(p + f)| / (s cos p)), tan p = tan(alt)/cos(gamma)
Diffuse sky light uses the BRE average daylight-factor formula DF = T Aw theta / (A (1 - R^2)) with an exponential fall-off with depth, so the back row is always dimmer than the front row.
Simplifications: clear-sky model only (no clouds), no inter-reflections for the direct beam, the room is empty of people, glass transmission 0.78 at normal incidence with a Fresnel-like angle factor, no trees or buildings outside. Real design work uses Radiance or EnergyPlus; this lab shows the geometry honestly, it is not a compliance calculator.
Real-world targets: about 500 lx on desks (EN 12464-1) and no direct sun on a whiteboard.
Independent checks run in your browser: