Fatal Logic

[Evan]

Well, only when the AoA sensors are installed backwards. I blame Stolichnaya.

https://avherald.com/h?article=51b0a785&opt=0

[3WE]

I’ve alway had a gut feeling that AOA indicators need to be a secondary instrument.

Airplanes ALMOST never stall at healthy airspeeds and attitudes.

[bstolle]

That's why the unreliable airspeed table is one of the most important items in the abnormal/emergency checklist.
Many pilots know the most important pitch/power values from this table by heart.

[3WE]

Many pilots know the most important pitch/power values from this table by heart.

You’re kidding?

(Footnote: This is directed at Evan, not you)

1800, 70 knots, nose level…

[Gabriel]

It is difficult to keep unreliable-airspeed pitch by the table when you are pulling up and the plane says "Dave, I cannot let you do that, Dave" and proceeds to command nose down.

Although the controls were manually deflected to command a nose up attitude, the elevator deflected nose down due to an active defense against excessive AoA

Descending through 3000 feet GPWS alerts sounded including "Terrain! Terrain!" and "Pull up! Pull up!". The pilots commanded a maximum nose up, however, the elevators did not follow their command.

The FDR last recorded values of -25 degrees pitch. The MAK reports that the SU95 has a number of protections including a defense against excessive angle of attack.

The FBW killed them*. What happened to the guarded "gimme my effing plane" red override emergency button?

* (That, and the apparently incorrect installation of the AoA vanes, and the fact that the AoA vanes apparently were designed without a simple keying to make its incorrect installation impossible)

[Evan]

It is difficult to keep unreliable-airspeed pitch by the table when you are pulling up and the plane says "Dave, I cannot let you do that, Dave" and proceeds to command nose down.



The FBW killed them*. What happened to the guarded "gimme my effing plane" red override emergency button?

* (That, and the apparently incorrect installation of the AoA vanes, and the fact that the AoA vanes apparently were designed without a simple keying to make its incorrect installation impossible)

Definitely a major design failure. Given the importance of AoA sensors to the FBW protections, it must not be possible to install them in the wrong position.

Bur this also happened to an A320. The probes were installed correctly and were not faulty, but a certified maintenance genius hit them directly with a pressure washer and they froze up at altitude. Same lack of that guarded switch for direct law.

[bstolle]

It is difficult to keep unreliable-airspeed pitch by the table when you are pulling up and the plane says "Dave, I cannot let you do that, Dave" and proceeds to command nose down.

I can't imagine that there's no way to disconnect the protections from the FCCs like on the Airbus. Seems to be exactly the same situation when the Instructor on the A320 told us in the sim:
"I'm now showing you how the A320 will kill you" just in the opposite pitch direction,

[Gabriel]

I can't imagine that there's no way to disconnect the protections from the FCCs like on the Airbus. Seems to be exactly the same situation when the Instructor on the A320 told us in the sim:
"I'm now showing you how the A320 will kill you" just in the opposite pitch direction,

I don't know in this Russian plane, but in the Airbus it is possible. Turn off any 2 of the 3 ADIR's and you have direct law.

[Evan]

I don't know in this Russian plane, but in the Airbus it is possible. Turn off any 2 of the 3 ADIR's and you have direct law.

I think that gets you Alternate 2 (roll is direct).

Hopefully, Airbus pilots have the proper 'off the books' procedure installed in their instant-recall part of the brain. For that rarest-of-rare scenario.

The A320 ‘procedure’ I’m familiar with is to switch off both FAC’s and lower the gear, assuming you are not too far above Vle. But I suppose damaging the gear is preferable to diving into a smoking hole.

The SJ-100 does not have any FTCL switches on the overhead, just ADIRS switches. But I would think those would do the trick.

This plane is a 2014 build. The avionics are western components (Thales, Honeywell...). After the war sanctions, Suhkoi is now using Russian avionics instead. In that case, I think you simply have to pull the AX144 vacuum tubes and open the knife switch.

[Gabriel]

I think that gets you Alternate 2

That would suffice, right? Control law remains stick-to-G or stick-to-fpm (depending on the speed) but stall protections are cancelled (just warnings remain) and autotrim is cancelled too. So HAL losses its ability to override the pilot's commands. Correct?

[Evan]

That would suffice, right? Control law remains stick-to-G or stick-to-fpm (depending on the speed) but stall protections are cancelled (just warnings remain) and autotrim is cancelled too. So HAL losses its ability to override the pilot's commands. Correct?

In Alternate law, HAL is still ‘interpreting’ sidestick commands and determining elevator deflection, but without AoA protections. What I’m unclear on is how malfunctioning (or frozen) AoA probes that remain in agreement affect the FBW ‘interpretation’. Maybe there is an accident report that sheds light on this. Direct law, as you know, does not involve HAL interpretations.

[Evan]

One thought came to mind: did this plane have western OEM sensors replaced with Russian (KRAK or whatever) units that lacked a position-specific design?

[3WE]

stick-to-G or stick-to-fpm (depending…

The Hell Not_Engineer says WT Phugoid…Stick-to-elevator isn’t in Gabe’s post.

Whatever the case, you should have a weed-free seedbed at planting time.

[Evan]

I don't know in this Russian plane, but in the Airbus it is possible. Turn off any 2 of the 3 ADIR's and you have direct law.

This is directly from Airbus 'Safety First' lit, describing a SIM session:

The aircraft is in climb. A dual total pitot blockage failure is set on captain and first officer sides.
While the aircraft is climbing, the erroneous captain’s and first officer’s speeds increase and reach VMO/MMO with the triggering of the Overspeed warning. Then the AP disconnects and the High Speed Protection unduly activates, the aircraft pitch-up, the flight crew tries to counteract with full forward sidestick input, without success. To recover, the flight crew is trained to switch off two ADRs to revert to alternate law and then to manage the aircraft trajectory.

FYI: the dual total pitot blockage described in that paragraph is considered by Airbus to be impossible and has never occurred on an actual flight.

But there is also the Abnormal Attitudes Law reversion if both faulty AoA vanes are in agreement and resolving to greater than 30deg. Even in Alternate 2 (Alternate Law Without Protections), that might still mess with your pilot commands.

Lower the gear at this point and I think you are in Direct law.

[Evan]

Gabriel, you are referring to the OEB48/49 procedures published by Airbus in December of 2014 to address an incident caused by a new probe plate design that was quickly recalled. Airbus has since cancelled the OEB48/49 procedures after there were no proper applications of the procedure and six improper applications. The procedure addressed (then) possible undue activations of Normal Law protections by forcing the logic to revert to Alternate Law Without Protections via disconnection of two ADR's.