← Engineering Notes

Stall speed and angle of attack: what they mean for drone design

By Keayvan Keramati · ~7 min read

Stall speed already shows up as one line in wing loading — it's the number that sets how big the wing has to be. But it's worth pulling apart on its own, because "stall speed" is a slightly misleading name: the real cause of a stall is angle of attack, not speed. Understanding that distinction is what lets you reason about turns, gusts, and — for a VTOL — the riskiest seconds of the whole flight: transition.

What is stall speed?

A wing's lift coefficient C_L can only go so high before the airflow separates from the upper surface and lift collapses — that ceiling is C_Lmax. Stall speed is the airspeed at which the wing is already working at C_Lmax and is just holding the aircraft's weight:

Fly any slower than this in level flight and there is no C_L high enough to keep lift equal to weight — the wing stalls and starts losing altitude no matter what the pilot or autopilot commands.

Angle of attack is the real cause — not speed

Lift coefficient is a function of angle of attack (AoA, α), not airspeed:

C_L climbs roughly linearly with α up to a critical (stall) angle — typically 12–18° for a simple UAV airfoil — and then drops sharply as the flow separates. Speed only enters the picture indirectly: at a given weight, flying slower means the dynamic pressure q = ½ρV² is lower, so the wing needs a higher α to make up the same lift (L = q·S·C_L(α)). Stall speed is simply the speed at which the AoA needed to hold weight equals the critical angle. An aircraft can stall at any speed — including a fast one — if it's pulled to a high enough AoA (a hard pull-up or an aggressive turn); it cannot stall at any AoA below critical, no matter how slow.

What changes stall speed

The turn case is worth spelling out, because it's the one that catches people out: in a coordinated turn at bank angle φ, lift must support both weight and the extra centripetal load, so the wing effectively carries n·W, where the load factor is:

A 45° bank already means n ≈ 1.41 and a ~19% higher stall speed than straight-and-level flight; a 60° bank means n = 2 and a ~41% higher stall speed. This is why turning stall speed — not the 1g number — is the one that should set your safety margin for any maneuvering flight.

Does a VTOL or tail-sitter have a stall speed?

The vehicle can hover, so it feels like the question shouldn't apply — but the wing still has the same C_Lmax and the same formula. What changes is how much of the weight the wing is asked to carry. Using the tail-sitter convention (pitch angle γ, 90° = hover, 0° = full forward cruise), the rotors carry the whole weight in hover and the wing carries a share that grows as the vehicle tilts over:

At γ = 90° (hover), cosγ = 0 and the wing needs no lift at all — rotors do everything. As γ → 0° (cruise), cosγ → 1 and the wing must carry the full weight, exactly like a fixed-wing aircraft at its own stall speed. The danger is doing that hand-off too early: if the airframe has tilted most of the way to cruise attitude before it has built up enough forward speed, the required AoA blows past the critical angle and the wing stalls — mid-transition, with the rotors already partly unloaded. That's why a minimum transition speed, set with a safety margin above the wing's stall speed, is a hard requirement for any winged VTOL:

Worked example

Reusing the 3 kg UAV from the wing loading article: W = 29.4 N, S = 0.28 m², C_Lmax = 1.2, ρ = 1.225 kg/m³.

In a 45° banked turn (n ≈ 1.41):

If this airframe were a tail-sitter VTOL instead, a sensible minimum transition speed (k_margin = 1.2) would be ≈ 14.3 m/s — the forward speed it needs before finishing the tilt to full cruise attitude.

This is a point-mass, linear lift-curve model. Confirm C_Lmax, the lift-curve slope and the real critical angle with airfoil data, wind-tunnel or CFD before freezing a transition schedule.