Suggestion regarding stall prevention logic without airspeed sensors

[danieliorgaaeromodelism]

Bug Report

Issue details

Description

"STALL_PREVENTION" behavior appears unclear when no airspeed sensor is installed ("ARSPD_USE = 0").

The documentation states that stall prevention relies on measured or synthetic airspeed. Since aerodynamic stall is fundamentally related to angle of attack rather than airspeed itself, it is difficult to determine how accurately stall prevention can assess stall risk when neither a Pitot tube nor an AoA sensor is available.

Concern

Consider a fixed-wing aircraft in a steep descent with:

• High true airspeed
• High aerodynamic energy
• Significant downward flight-path angle
• Relatively low GPS ground-speed component in the horizontal plane

In this situation, it is unclear:

1. Which variables are used by "STALL_PREVENTION" to estimate proximity to stall when no airspeed sensor is present.
2. Whether synthetic airspeed estimation could incorrectly classify the aircraft as being in a low-energy condition.
3. Whether stall prevention could unnecessarily limit pilot authority during a recovery maneuver.

Additional Observation

Pilots often recognize an approaching stall through reduced control effectiveness, such as:

• Reduced elevator authority
• Reduced aileron authority
• Increasing control input resulting in diminishing aircraft response

ArduPilot already has access to:

• Commanded rates
• Achieved rates
• PID controller outputs
• Control surface commands

Has control-effectiveness degradation ever been evaluated as an additional indicator of stall proximity?

Version

ArduPlane 4.6.1

Platform

[x] Plane

Airframe type

Fixed-wing aircraft

Hardware type

STM32F405-based flight controller

Logs

No logs available.

This report is primarily a request for clarification regarding the implementation and behavior of "STALL_PREVENTION" when operating without an airspeed sensor.

Questions

1. What measurements and estimations are used by "STALL_PREVENTION" when no airspeed sensor is available?
2. Can "STALL_PREVENTION" limit pilot authority based solely on synthetic airspeed estimation?
3. Are there known limitations in steep-descent scenarios?
4. Has control effectiveness (commanded response versus achieved response) ever been considered as a stall-proximity metric?

Expected behavior

The documentation should clearly describe:

• The complete decision chain used by "STALL_PREVENTION" when no airspeed sensor is available.
• The limitations of synthetic airspeed estimation.
• Any situations where pilot authority may be limited based on estimated rather than measured aerodynamic conditions.

[LupusTheCanine]

We do not implement any stall prevention other than limiting roll based on airspeed and not intentionally slowing down below AIRSPEED_MIN, both functions require airspeed estimate or measurement to be in use and preferably accurate 😅.

The only way to be limited by estimated airspeed-based limits is to enable stall prevention when using synthetic airspeed.

Use of airspeed sensor is recommended.