steve2267 wrote:In another thread, JohnWill I think it was, commented that at high AoA (e.g. 45°), left/right stick commands roll around the aircraft flight or velocity vector. Such a roll will be a combination of body-axis roll and yaw. Trying to visualize this some more, it is not clear to me if stomping on the left / right rudder pedals at 45° AoA will be a pure body-axis roll as at 0° AoA, but I think it will be. (For example, at 0° AoA, pure body-axis yaw results in no change in angle-of-attack. At 90°, pure body-axis yaw also does not change alpha. That is, (body-axis) yaw 90° left or right at 90° AoA, and you are still at 90°alpha. But I may need to think about this some more.)

Usually, roll axis control variable LCV is stability-axis roll rate at normal and high dynamic pressures. Stability axis roll rate is

Ps = Pb * cos(α) + Rb * sin(α)

where Pb and Rb are body-axis roll and yaw rate, respectively. The reason Ps is used as LCV instead of Pb is that at AoA rolling about the body axis converts α to β. E.g., if I roll 90° around the body x-axis at 20° α, I now have 20° β and 0° α. Since we want to maintain the current angle-of-attack and coordinated flight (i.e., β small) we roll about the stability-axis.

At low dynamic pressures, LCV is Pb because at low speeds because it is less desirable to roll around the velocity vector and because at low qbar the incidence angles (α and β) become difficult to measure and can also become undefined.

The yaw control variable NCV is Rb at low qbar and is a mix of Rs = Rb * cos(α) - Pb * sin(α), β feedback, and decoupling terms at high qbar. The goals are to

- maintain zero steady-state lateral acceleration with no pedal input (i.e. coordinated flight)
- achieve a satisfactory response of sideslip and Rs with pedal input.

Based on AIAA 2002-6020, NCV for the X-35B in non-hover mode is

NCV = Rb * cos(α) - Pb * sin(α) - g/V * sin(φ) * cos(θ).