Paragliding

Vario- Altitude- Airspeed- Glide- Angle of attack- Bank- Clock- Task- Next zone-

Cross-country task

You hang on lines below a flexible wing, so pitch is a pendulum: brake and speed bar change the angle of attack through the line geometry, and the wing pitches, surges and recovers on its own. The air is a lattice of rising, growing, drifting thermal columns; the plan view marks the ones at your altitude. Get round the course, and round the speed section, faster than the fleet.

kg

Known limits of this model — read before you judge a number
  • No manufacturer publishes a sink-rate polar. Ozone, Advance, Nova, BGD, Gin and Skywalk were all checked, including Nova's full manual and product presentation: they publish planform and certification, never sink against airspeed. The only published speeds for this wing are its minimum 25–30 km/h (EN 926-2 flight test report, item 3, at 115 kg), its 12 km/h brake range and its 30 km/h total range with accessories (certification sticker). The polar is fitted to exactly those, so every sink rate on it is a prediction: best glide 10.27 : 1 at 39.5 km/h, minimum sink 1.04 m/s at 37.5 km/h, sink at trim 1.06 m/s, and 2.00 m/s at the 57.5 km/h top speed for a glide of 7.91. None of those four numbers was measured on a real wing.
  • A held turn input does not settle — it spirals. Full weight shift and nothing else reaches 11.3° of bank after 10 s, 26.4° after 30 s and 33.8° after 50 s, and keeps going: the spiral mode is divergent (linearised time constant 120 s at trim). A thermalling bank has to be held actively, which is what a pilot does, but the certified wing exits a gentle spiral spontaneously (item 8, grade A) and this one only recovers from a developed one, inside the 720° it was fitted to.
  • Full-brake sink is optimistic. 1.37 m/s at the 27.5 km/h minimum speed, where a real wing flown near its minimum sinks more like 1.5 to 2.0 m/s. There is no separation-drag rise before the stall — drag jumps only once the flow is modelled as separated.
  • One cause for the two above. The canopy is four equal-area panels at one chord station, laid on the published arc. That has no spanwise lift distribution, so it over-weights the lightly loaded tips; a rigid arc with fully tilted panel normals came out about ten times too stiff in roll (full weight shift gave 3.6° of bank and a 302 m turn radius after 40 s, against 30° and 29 m). The arc's dihedral effect is therefore carried by a single roll-due-to-sideslip derivative, and one derivative cannot make the roll authority and the spiral stability both right. Roll authority was fitted to the published 20 s for a 180° turn (item 22) and spiral recovery to the published 720° (item 9); the gentle-spiral behaviour is what gives.
  • The pitch pendulum is overdamped, not oscillatory. Measured from the eigenvalues of the 14-state Jacobian at trim: a real root at 52.2 rad/s (time constant 19 ms), which is also the fastest mode in the system and therefore what fixes the timestep — RK4 needs dt below 0.053 s and the game runs at 0.01 s. The 4 to 9 s oscillation a pilot feels is the phugoid, period 8.3 s, damping ratio 0.036 — and it is only just damped, so the wing surges for a long time after an input.
  • Flat earth for the task. Distances are Euclidean in a local tangent plane, not WGS84 geodesics as S7F §4.2 requires. Over a 60 km task in a 50 km box the difference is below the ±5 m cylinder tolerance the same document sets, but it is a departure and it is not modelled.
  • The whole atmosphere is reconstructed. No governing body or manufacturer publishes a thermal model. Core strength, radius, spacing, lifetime, rise rate and the trigger lattice are all invented; only their form is principled — superposed Hill spherical vortices, which are exactly divergence-free, so the sink around a core is a consequence of the field rather than an added term.

What to trust: the task and its scoring, which are the FAI Sporting Code S7F 2026 V1.0 formulae clause by clause, and the speed range, which is the certification sticker's. Everything about how fast the wing sinks, how steeply it banks and how the air moves is a model. The per-item list of what is sourced and what is reconstructed is in CREDITS.txt and under Help.