Free Flight, Free Speech

Thinking about an implementation of stall protection, i find the following:

The critical angle of attack is a constant directly related to the form of the wing. If you would look at airspeed for stall, you have to take into account all sorts of other factors, like weight. But angle of attack is a constant. Of course when the wing changes you get a different critical angle of attack, so you have as many critical angle of attacks as there are flap settings. For an Airbus that would be 4 i believe.

A fixed-wing aircraft by definition is stalled at or above the critical angle of attack rather than at or below a particular airspeed. The airspeed at which the aircraft stalls varies with the weight of the aircraft, the load factor, the center of gravity of the aircraft and other factors. However the aircraft always stalls at the same critical angle of attack.

https://en.wikipedia.org/wiki/Angle_of_attack#Critical_angle_of_attack

So for stall protection in flightgear, the only thing needed is to be able to readout the angle of attack from the fdm, and know the 4 (on an Airbus) critical values.

Kind regards, Vincent

Most airliners have a huge amount of washout because of their swept wings, this tend to make their stalls somewhat gentle.

Stalls come in all types of aircraft, from brutally abrupt and liable to send you into an unrecoverable spin, to the generously benign and resembling more of a mush/stall were the plane doesn't quite stop flying but just increases it's descent to forward motion while still managing to stay controllable.

I don't know about Yasim, but JSB actually can tell you anyway when you stall. fdm/jsmbsim/stall-hyst-norm changes from 0 to 1 is the stall condition is reached. It's a double and it changes in between as well. So something like stall_hyst_norm>0.98 or so should do the trick. The advantage is, it should react not only on the AoA but also on additional factors like if you are in a tunr (because AoA alone is only usable if your wings are level, otherwise you lose additional lift by sideslip).
Does that solve the problem at least for JSB?

I will shortly be redoing the lift tables in my Airbus FDMs, witha life due to alpha profile for each flap/slat setting, instead of summing the lift alpha with lift flap and lift slat. This way, I can finally have realistic stall behavior. I then have to carefully monitor it, and find the proper alpha values where she stalls, to program the indicators on PFD. Stall protection on the Airbus happens way before the plane actually stalls, so I cannot use that property for that.

Kind Regards,
Josh

jwocky wrote:The advantage is, it should react not only on the AoA but also on additional factors like if you are in a tunr (because AoA alone is only usable if your wings are level, otherwise you lose additional lift by sideslip).

The interesting thing about angle of attack is that the critical value for a wing is always the same (with the same flap setting). "Additional factors" only influence the actual angle of attack, not the critical angle of attack number. In a turn one wing can have a different actual angle of attack number from the other. But the critical angle of attack number stays the same for both wings.

During a turn in level flight, greater lift is required to offset increased load factor; the critical angle of attack is therefore reached at higher airspeeds. The formula is as follows—normal stalling speed times the square root of the load factor equals banked stall speed; accordingly, an aircraft with a stall speed of 50 KTS in a 60°-banked turn (load factor of 2.0) will stall at 71 KTS.

During a climbing turn, the inner wing has a smaller angle of attack than the outer wing; the outer wing will therefore stall first. The reverse is the case for descending turn, where the inner wing has a larger angle of attack and will therefore stall first.

http://www.langleyflyingschool.com/Pages/CPGS+4+Aerodynamics+and+Theory+of+Flight+Part+1.html

So the readout of angle of attack from the fdm must be possible for both wings separately, to be able to evaluate the one with the highest angle of attack.

Btw. Loosing lift (e.g. due to slip) indicates a decrease of angle of attack. Nothing to worry about, unless you went over the critical angle of attack, and stalled the wing.

Kind regards, Vincent

That is exactly the problem. The AoA stays the same but ... the AoA at which the plane stalls is different. So, for a stall protection system, you can't rely entirely on AoA. Which is my point in the earlier post.

No, i am saying exactly the opposite. A wing stalls at the critical angle of attack. Every article i read about it states: The critical angle of attack is a constant related to the form of the wing. So angle of attack is a perfect way to make stall protection. If you can readout the actual angle of attack from the fdm, you only need to compare it with the critical angle of attack of the wing.

Kind regards, Vincent

But you realize, you have two wings plus a tail?

As i said before, ideally you should be able to readout the actual AoA of the wings separately, because they can have different actual AoA's under certain circumstances.

The tail has not much to to do with it. It can only influence the attitude of the plane, and thereby the actual AoA of the wings. That actual AoA is what you measure and is what you need to respond to.

Kind regard, Vincent

Angle of attack..... mmmmmmm let's see..

With a normal dihedral with a -5 devs beta the left wing has an increase of aoa and the right a decrease.

With a normal wing sweep and a -5 beta the left wing has an increase in effective dynamic pressure and the right wing a decrease.

With a -1 rad sec R rotation the left wing has a decrease in U velocity and the right an increase

With a +1 rad sec P rotation the left wing has a decrease in W velocity and the right an increase.

And decrease in incidence angle from root to tip means a decrease in aoafrom root to tip.

With an increase in UV velocity to W velocity the aoa decreases.

That's a good start for you.... you've a long way to go.

Simon.