PHI Crash in Louisiana Jan 2009 - 8 Dead, 1 Injured

[twisted wrench]

Wow this is getting interesting.

We have either a dual FADEC shut down or/and a dual Auto-pilot failiure.

Both the Engine FADEC and Sperry SPZ-7600 auto-pilots are totaly independent systems, the likely hood of either of those systems to have a dual failure at the same time is very unlikely if to happen at all.

I think the speculation is getting out of hand!!!

[SASless]

I can think of several reasons why the "zero torque" situation can happen....starts with a huge "BANG" and a loosely joined set of extremely spare parts. At some point in the break up of an aircraft...assuming the recorders are still doing their thing....would not that be an indication we would expect to see?

As what led to that break-up.....is anyone's guess.

The engines could still be running...but not connected to anything...or connected to something but the rest of the dynamic components have departed for other places...again...no torque needed.

The 76 is very slippery at high speeds...as is the 109....even the venerable 412 can be a hand full when the AFCS kicks off at high torque and airspeeds.

Airline pilots undergo "upset" training....do helicopter pilots? Or is that another one of those "verbal brief only" items on the check ride?

[Phil77]

Easy Mr. twisted wrench! I did not say or speculate that my story is the likely cause in combination with a dual flameout, although reading SASless post that was posted while I was responding, I think as well that the timeline is important. We do know that usually a chain of events lead to an accident.

I did however mention at the bottom of my previous post, that it sounded like a possibility to somebody who did not read the statements of CycColl.

My comment was merely intended to underline the benefits that can come out of a discussion ("speculation" sounds so negative).

[HeliComparator]

SAS

I never said the argument was won on Pprune (and neither did I say who the winner was) however it has been won on the street (well in the sky I suppose)

F1100 / twisted

You should not find the suggestion of both engines going offline simultaneously that impossible. Both engines are controlled by the same design of FADEC with the same hardware and the same software inside. Something affecting one engine that is software related (a bug) or from an external source (RF interference, icing etc) is going to have just the same effect on the other engine. If the probability of having a single engine failure is 1 in 10,000 hrs then the probability of a double engine failure is most definitely not 1 in 100,000,000 hrs - its much more probable than that.

On the EC225 fleet we have had 2 cases of double complete AHRS failure, we have had no cases of single complete AHRS failure. The 2 cases were caused by a software bug that didn't agree with the Montrose A platform. Fortunately the third AHRS system uses different hardware and software and remained OK. Had we had 3 AHRS systems of the same design, it might have been an accident (btw the bug was fixed a year or so ago). This is why fly-by-wire fixed-wings have different flight control computers manufacturerd by different companies using different hardware and software (though trying to achieve the same behaviour), rather than just installing 3 of the same kind of control system.

Also bear in mind the 777 that crash-landed at Heathrow the other year - both engines failed to produce power within a few seconds of each other, in that case I think it was fuel icing. The equipment (engines, AFCS computers etc) may seem to be separate, but there are always common elements.

HC

[Rusty Bifilar]

I’m not a C++ guy so you people that know tell me if this scenario is possible. A false main rotor overspeed signal causes both DECU’s to rapidly reduce power. N1 does not go below idle so both generators stay online. Would the AFCS decouple with an overspeed or would it try to maintain speed and/or altitude depending on which mode was engaged?

[twisted wrench]

The C++ monitors N2 input signals to the DECU and not NR speed, so the likely hood again of both engines showing a overspeed (real or false) again very unlikely.

The common link is the collective lever postion anticipator that will sense increase or decrease in collective position to adjust power settings.

I noticed in Sasless comments the Torque went to zero , in CYCCOLL report only stated a significant reduction in torque.

I also am very interested to see what happened to this C++, I have many of them on the base I work at.

[SASless]

I probably used "zero torque" when thinking along the lines of what Cylcol had reported....but was talking about a generic event and not the one that happened to the PHI aircraft in particular.

My conjecture also would have to allow for the differences between shaft driven transmissions vice direct input drives. My thinking was more along the lines of a mechanical failure vice a digiital brain unit running amok. The main discussion was more along the lines of how the various digital brains interact.

[HOSS 1]

Think About It

77-

So, let me get this straight. You actually believe that if the AFCS/SAS were to immediately go offline while in high speed cruise, the t/r thrust will somehow cause the aircraft to immediately roll? Kind of like a loaded spring that is released? That the AFCS is normally counteracting?

The AFCS TRIMS the aircraft controls to the current flight condition. That is, the t/r is providing the correct amount of anti-torque for trim coordinated flight. (Which by the way, is not all that much due to the sizable aerodynamic anti-torque.) So while the the aircraft will quickly be marginally stable and very "loose", it certainly is not a wound up spring waiting to flip you inverted. The inverted flight comes later, from the pilot's inability to fly the unstable dutch roll mode.

Do you not practice non-augmented (sas off) flight? I suggest you do so. How do you know your skills will allow you to fly the aircraft after a dual SAS failure? Remember, the aircraft is certified VFR SAS OFF up to Vne.

Hoss

Also, The -76C++ DECUs do indeed use Nr in the control laws, but close the loop around N2 in normal operation.

[FH1100 Pilot]

Mumbo Jumbo?

With respect to this specific accident, barring catastrophic structural or engine failure, I think we can safely - let's be honest - rule out: 1) The pilots deliberately pulled both engines off in high cruise; 2) Fuel contamination/exhaustion.

But helicopters don't just fall out of the sky. "Something" happened and the ship hit the ground hard - really hard. It's difficult to tell from the pics, but it doesn't look like they hit nose first. So, no "lawn-dart." Whatever "it" was, it happened fast, however.

twisted wrench, who at first thought that the speculation about this accident was getting "out of hand" wrote:

The common link is the collective lever position anticipator that will sense increase or decrease in collective position to adjust power settings.

Aha! Is this the mysterious "potentiometer" I keep hearing about? If this collective lever position anticipator fails, could it "fool" the DECU's into a mode where they reduce power on both engines simultaneously? Is the CLPA a single unit, or are there two of them for redundancy?

Mumbo jumbo, eh?

To me, it doesn't sound like an autopilot failure. The autopilot was probably doing what it was designed to do. The crew was probably cruising in "three cue" which, as others have noted is not such a great idea.

[If all else fails]

Quotes

77 - "due to the elevated T/R, the aircraft will snap into a roll;"

Hoss - "So while the the aircraft will quickly be marginally stable and very "loose""

So, agreement? that some level of instability will likely ensue following an unexpected loss of AFCS/SAS. Surely then, response/reaction time becomes the key as to whether the secondary and further effects (e.g. of yaw - induced by sizable aerodynamic anti-torque) have placed the helicopter in a hazardous attitude. The immediacy of any divergence being a factor of airspeed and the amount of divergence, a factor of the time before recovery/intervention.

But this is the bit that surprises me.

Quote 77

"My boss said, that he failed the Heliflight e.g. on a few very seasoned guys and due to the fact that it was never taught in the sim almost all of them did not see the T/R as the cause for the roll and would have rolled it straight into the ground."

Why would this not be taught in the sim?

[ramen noodles]

Having done a vast amount of stability flight test in several helicopters, including the S76 series, I find this entire line of speculation factless and groundless.

1) Switching the dual AP off in high speed flight will not cause any awful response. Could the Charles Lindburghs who think so just try it to see what happens?

2) A simultaneous (within 1 sec) dual engine failure or a dual stability failure is virtually impossible, unless you do it on purpose, even if a self-appointed expert thinks otherwise. Unless you deliberately turned them off together, they will not fail at the same time.

3) The Tail Rotor does not cause any wild roll in an unstabilized S76, all you have to do is let go of the stick to see that.

My Lord, we have entered that black hole that accident investigators dread, where the "could have happened" starts to dominate the scene, in the absence of real knowledge and real understanding.

Give pprune a D- for this thread.

[Phil77]

Why I would give some posters a D- here on pprune, is the fact that if they have a different opinion on a theory, they can't help but chastise any good discussion... oh, but not without elaborating at length on the topic themselves to prove their impressive knowledge!


I haven't tried it myself no, but also no reason to question the chief pilots reports what happened on several checkrides - why would he pull that out of his a**? Anyway, I will inquire about the mechanical reasoning for the upset I was talking about.

Hoss: As If-all-else-fails said, I think we're not that far apart; except maybe that I could have used a better word than "snap".
You said "The inverted flight comes later, from the pilot's inability to fly the unstable dutch roll mode", that is exactly what I was getting at: the pilots inability to correct the situation properly. Apparently there are a few people out here that think that the S-76 becomes pretty unstable without all her helpers. Again, I will try to quote the expert a little more detailed.

As far as the not practicing that behavior in the sim, I stand to be corrected.

Phil

[tottigol]

Pardon my ignorance, but does the C++ have any sort of Manual or Mechanical reversion mode for its FADEC?

On a different subject, I know for a fact that one of our C++ had an inadvertent (pilot induced) dual AP disconnect and the ride very quickly turned into a very exciting one, on par with some of the world's most famous roller coasters.

[Hippolite]

In the old days, I remember doing a short SNS Leman shuttle (about 6 stops) for Amoco using an S76A BJVZ (vice an unserviceable Bell 212) in which both lanes of the Hamilton Standard Phase 2 were U/S.

Stabilization was not required in the Ops Manual at that time for day VFR so we just flew the trip from start to finish raw without any stabilization.

Its no big deal and if you don't, you should be practicing it regularly. Granted, the Sperry system of actuators does give the aircraft a slightly different feel than the Hamilton Standard.

[High Nr]

1) Switching the dual AP off in high speed flight will not cause any awful response. Could the Charles Lindburghs who think so just try it to see what happens?

Suggest you inquire about a C++ operated by a large US based company in Australia a year ago.
Crews were PBI trained and current, but when the AFCS failed, the aerobatic airframe was "Overstressed" in the recovery efforts.

Took two weeks of inspections to get the airframe flying again.

[HOSS 1]

77- Yes. My point is there is no tail rotor "pre-load" waiting to roll you over when SAS is removed. But, I believe most pilots have little appreciation for just how demanding an unaugmented -76 is above 145kts. This is why I suggest all have a look at it first hand.


Also, the C++ certainly has a manual reversion mode. Controlled by a small DC motor on the fuel controller, commanded by a incr/dinker thumb switch on the collective. Again, something everybody should be practicing, especially in the C++ since you can't hurt the machine. The DECU will kick you out of manual mode and resume engine control if you go outside any limit.


HO5S

[ramen noodles]

High Nr said, Took two weeks of inspections to get the airframe flying again.

Slow lookers.

[cpt]

A bit like an F117 then ?

[High Nr]

No.

I understand that the Factory Aeronautical Engineers took a while to formulate the inspection criteria, as the loads measured via the onboard HOMP were very scary and not covered by the normal inspection manuals.

Then NDT and the eyeball took up the rest of the time

The bottom line is that a fast S76 with no AFCS and no Attitude Force Trim is a real handful, even for experienced crews.

Now throw in some IMC in the wee hours, and that crew will be working very very hard.

Not sure how all this relates to a Dual Reduction in Tq and the consequent drop in Nr.

Now back to the main topic.

[Flapwing]

No AP's you have a handfull..... if you are a ham-fisted pilot!