58th Fighter Squadron F-35A crashes during night landing

[outlaw162]

In situations where there is a large divergence between pilot inputs and the anticipated inputs, the flight control system resets in such a way that pilot inputs may have a minimal effect on flight control surfaces for a significant period of time.

The term 'goat rope' comes to mind.

A bit of a logical fallacy when they reference the MP's simulator experiences of having landed, I assume intentionally, at 200 knots previously, as if that experience was in some way misleading and contributed to the (love the term) 'bewilderment' factor for this event. How can one even make a connection of this sort when he didn't even know he was at 200 knots when he landed? ....So the simulation didn't accurately prepare him for something he was unaware of doing wrong in the first place???

There just seems to be a number of extraneous and minimally relevant items beat to death here.

Maybe a 'talking airspeed indicator' like the F-15E had would have come in handy.

(Re Gums: A-7D HUD green/grey, it had a knob that you could select a nite filter as I recall, but don't remember seeing grey symbology as such, just washed out green)

[spazsinbad]

'outlaw162' said: "...'talking airspeed indicator' like the F-15E had would have come in handy..."

The Sea Venom had the 'looker' OBSERVER in the right hand seat for relaying one knot increments in IAS during a carrier approach. His view out front was even worse (for this situation) compared to the pilot view over the bulbous nose. Earlier this classic 805 squadron linebook entry says it all - a curved carrier approach almost mandatory like piston days of old...

[Gums]

Salute!
and ot
Per Spaz wishes......
+++++++++++++++++++++++++++++
Cross checks are the key to survival, regardless of the cosmic avionics and displays. Ask those French guys from AF447 and the Korean crew from the SFO debacle and ......

So I liked the AoA bracket in the HUD and the indexer chevrons on the canopy bow. Didn't need to calculate approach speed using extra fuel, a few bombs I had left over, and the beat goes on. So a rough gouge and compare with AoA, then fly AoA until touchdown. One of my SLUF instructors liked the Navy approach with zero or only a hint of flare. And I flew with that philosophy.

As a nugget, I had my dollar ride at The Beach after the short course in Tucson. Then a night refueling. All went well except the rain storm during the hook up with the tanker and getting a thousand pounds or so. Gotta admit that it was "interesting", as during my checkout in Arizona I only had one hookup in broad daylight. So head back to homeplate and acquire the ILS and slow down, configure, etc.

The SLUF flap handle had a few microswitches on it, and initial movement was inboard and then large movement back or forward depending on where the flaps were. However, if you "beeped" the handle forward or back or inboard/outboard after the initial movement you 1) stopped flaps at what ever angle they were, 2) beeped then up or down, or 3) raised them. The leading edge flaps came down or went up depending on the basic handle position.

O.K., speed coming down after lowering gear and then flap handle inboard and back. Speed maybe 40 or 50 knots above normal approach, but AoA low as it should be. Slow to get the AoA bracket centered and continue on. Landed and had piss poor braking on the wet runway. Decided I was not gonna stop and lowered the hook. Missed main barrier cable and snagged the carrier chain doofer on the overrun.

What happened was I had hit one of the microswitches on the flap handle and the trailing edge flaps did not come down all the way. The leading edge flaps allowed a high AoA but speed was also higher than with trailing edge flaps all the way down. My new habit of flying AoA without a good crosscheck of speed could have been nasty.

So those steam gauges are great to have, despite all the cosmic stuff.

Gums sends...

[outlaw162]

When questioned, neither the MP, AIB Pilot member (F-35 IP), or F-35A test pilots were familiar with some of the details of the control logic

(Nor, colonel, nor....)

How did the guy questioning them get this knowledge? Secret handshake?

Love this approach to training....

"There's no reason for you to know this, you wouldn't understand it anyway....and you'll never run into a situation where you need to know. Trust me."

I also like the report use of the word 'volatile' as far as describing pitch control at lower than optimum AOAs (higher speeds) in the PA mode. Not 'increased pilot attention required', nor 'can be challenging', but (gasp) 'volatile'.

(The SLUF flap beep switches were very useful for BFM. )

[spazsinbad]

Thanks BOTH. What an interestin' machina was da SLUF. Are all the 'caveat excuses' for the MIShap PILOT because he is of high rank or a nice guy or whatever? What will happen to him now? Snakes in Cockpit on Approach - GO AROUND EARLY!

[Gums]

Salute!

Tnx, Outlaw. After I left the SLUF, for my staff tour, they modified the leading edge flap configuration so you could retract them at higher speeds. Ones I flew until 1974 could lower them and beep trailing edge flaps up or down, but if you retracted them at a high speed they gouged the wing.

I agree about some of the terminology used in the report being unusual, even like slang. One phrase had the term "clueing" versus "understanding", "becoming aware", etc. The "volatile" expression also tickled me. I would prefer "sensitive".

@ SPAZ.... The SLUF was a delight, and had more goodies than any fighter/attack jet had until the F-35. 'course, the F-35 doesn't have a doppler like the SLUF, or the aux UHF, or the ADF feature we used on CSAR to find the survivor, or the two thermos bottles behind the headrest, or ......

Gums sends...

[spazsinbad]

Was one of the THERMOs to piss in - t'uther drink? Did you ever get them mixed up? YECH.

[spazsinbad]

Another story/look at the crash report....

Eglin F-35 crash resulted from tired, distracted pilot and unresponsive tail glitch, investigators find
05 Oct 2020

"An investigation has concluded that the May 19 crash of an F-35A Lightning II at Eglin Air Force Base in Florida was caused by the pilot trying to land at an excessive speed, and a previously unknown flight control logic glitch that left its tail unresponsive.

The report, which the Air Force posted online Sept. 30, also listed multiple pilot mistakes or factors that investigators said significantly contributed to the crash on Eglin’s Runway 30. They found the pilot was fatigued and, as a result, “experienced cognitive degradation” was also distracted at a critical point in the flight due to a misaligned helmet-mounted display. The pilot tried to land with the speed hold engaged and used an alternate cross-check method, and lacked some key knowledge about the fighter’s flight control logic....

...The fighter touched down with the speed hold still engaged — which is a prohibited maneuver — and about 50 knots faster than intended. It also came in at an angle of attack of 5.2 degrees, or 8 degrees shallower than it should have. At that speed, the pilot had to nudge the stick forward to touch down. This caused all three landing gear to touch the ground at the same time, instead of the rear gear touching first and then easing the nose gear down.

The three-point attitude landing caused the nose gear to bounce, and the fighter’s nose rose quickly and sharply. The pilot again pushed the stick forward to try to stop the fighter, which started a series of “increasingly violent” pitch oscillations as it bounced on the landing gear.

The pilot continued to try to regain control of the fighter for about five seconds, but the flight control system became overwhelmed by the quick succession of forward and aft stick movements. This flight control system glitch caused the plane’s horizontal stabilizers to default to a trailing edge down direction and stay there, nudging the nose down.

The pilot pulled back on the stick and hit full afterburners to try to abort the landing. But with the stabilizers pointing down, his effort to take off again was unsuccessful and he was forced to bail out....

...The pilot was fixated on the faulty display symbology during the critical phase of landing, to the exclusion of cross-checking his airspeed and angle of attack, the report said..."

Source: https://www.airforcetimes.com/news/your ... tors-find/

[f119doctor]

Just remember that this is the Accident Investigation Board report - Lawyers looking for fault (or coverup). The AIB is releasable to the public, as opposed to the Safety Investigation Board, where the USAF finds the root causes of the mishap to prevent future events, and is not releasable to the public.

I have seen mishaps where the AIB and SIB came to much different conclusions. Not saying this is the case here, but just remember that the two different investigation boards have different rules and motivations.

It is interesting to see the conclusion that they found a situation where the CLAWS was“saturated “ due to conflicting inputs and resulted in a PIO type situation. Somewhat reminds me of the YF-22 event where Tom Morgenfeld put the aircraft belly down onto the runway after an abnormal flight control response during a go-around during the post Dem-Val testing

[Gums]

Salute!


Great post, Doc.


I feel the safety folks will split cause on pilot error and confusing CLAW.

At least in Viper, the change to approach gains and aoa inputs was instant when WOW showed airborne. As I posted earlier, we did not have "bounce" mode.
-----------
P&W help at Hill in early days was super. I got my plaque on the wall.

Gums sends...

[f119doctor]

Salute!


Great post, Doc.


I feel the safety folks will split cause on pilot error and confusing CLAW.

At least in Viper, the change to approach gains and aoa inputs was instant when WOW showed airborne. As I posted earlier, we did not have "bounce" mode.
-----------
P&W help at Hill in early days was super. I got my plaque on the wall.

Gums sends..

Ah, the not so good days of the UFC/EEC/BUC control system on the F100. Back in development, they moved the ground idle nozzle activation (Idle Area Reset) from WOW to landing gear handle down. If you bounced the landing with WOW control, the nozzle would close, giving higher thrust, making the bounce higher and the next landing harder. Taught to me by one of those original Hill FSRs who became the expert on F100 Trim & Troubleshooting. You actually learned what was going on inside the UFC hydro mechanical brain, which was quite amazing for its time.

[quicksilver]

Reading the report, I experienced a little ‘cognitive dissonance‘ about the idea of ‘flawed logic‘ in the CLAW; not that there can’t be such a thing but, in my experience, and off the top of my head, I know of no man-made ‘thing’ in a jet that a pilot can’t (circumstantially) find a way to eff-up and thereby prove it to be ‘flawed.‘

I’m not sure what test regime would imagine such a circumstance as this mishap — before-the-fact — and thereby mitigate the risk of such a thing occurring in futurity.

Flawed logic? How about simple PIO? And how long has that been around in aviation?

[quicksilver]

“I agree with Quick about having the best system for the envisioned flight condition. OTOH, seems to me that a zero-zero system that has been tested with line pilots more than a few times would have satisfied the requirement.“

Considering that they don’t fly it, and that the design of the canopy is significantly different aft of the windscreen, I’m not sure why anyone in the USMC would care about what technical solution was chosen for use in the canopy of the F-35A.

I also expect that ‘frequency of exposure’ in the low and slow regime as well as the consequential statistical probabilities of survival are different for STOVL jets, hence the ‘det cord’ solution. But, once again, what pilot cares what the technical solution might be as long as it satisfies the probability of success? There may be somebody who does, but I don’t know ‘em.

[outlaw162]

How about simple PIO? And how long has that been around in aviation?

I've never heard of a stab defaulting to full aircraft nose down in a PIO in any other aircraft....and then ignoring pilot inputs temporarily whether in a PIO or not....which I don't think technically this was, a PIO that is, although it would have been if the aircraft had actually responded in that 5.5 second interval from 3 point touchdown to ejection. I'll buy that it was a stick PIO however, but without the aircraft response. He have been better off in an aircraft PIO, at least as far as a somewhat recognizable screw-up with a more predictable outcome. If you can saturate the FCS in 5.5 seconds in the PA 'bounce' mode it probably should have been in some 'test regime', and evidently was, as indicated later in the report. Just one of those 'it'll never happen' oversights.

The report even makes reference to the fact that the MP had landed the simulator a couple of times at 200 knots and that this control anomaly was not simulated accurately. Lack of fidelity in the sim compared to the aircraft when dealing with software iterations that can be rehosted is somewhat unusual in my Level D simulator experience. Yet it is said in the report that 'they' already knew about this flight control computer 'saturation' potential, and when questioning F-35A test pilots, they were unaware of it, and they determined it was not incorporated in the sim.

But this was a 'porpoise' from a three-point landing which is not necessarily uncommon in any tricycle gear aircraft, and the standard recovery is to bring the stick aft of neutral, hold it there, and go around....which he tried, but the control logic precluded that, and in his case, go-around in the 'bounce' mode was just the opposite of what hindsight indicates he probably should have done, just planted it.

Yes, he did eff-it up, and badly, but this full nose down logic exacerbated a simple go around. This FCS 'saturation' potential had already been discovered and evidently available in some document not available to line pilots, instructors....or test pilots, but evidently available to 'someone' on the AIB asking questions.

In fairness, the Harrier mafia never had to deal with tricycle gear porpoises.

[quicksilver]

“The report even makes reference to the fact that the MP had landed the simulator a couple of times at 200 knots and that this control anomaly was not simulated accurately.“

Unless the 200kt landing is part of an EP, why is he practicing such a thing? Perhaps it is part of an EP; stab EHA failure in early dev jet (AA-1 iirc) resulted in very high speed landing in FW.

I’d be interested in understanding the assumptions built into the PA bounce mode logic.