Shy take a look at how the thing works. It was not the input that failed but lets say the "output" to cancel the input via the feedback mechanism.

It's pretty simple - when all 3 points attached to the lever line up with the servo valve in the neutral position nothing happens as the control shaft moves and cancels the input request.

As the feed back end became detached - nothing to cancel the input - servo motors to the stop and in this case full right pedal?

BTW - Sikorsky has been fitting centreing as the result of accidents or incidents not unlike this one - not pro-actively.

The design and acceptance of the AW models may be under review shortly. If not I will be very surprised.

Shy,

At the risk of telling you how to suck eggs -

https://cimg5.ibsrv.net/gimg/pprune....5797265e2b.png

Setpoint = TR pedal control

Load = TR Pitch change or duplex bearing

Feedback lever disconnected at the hydraulic cylinder (servo)

Disconnect the feedback lever after the hydraulic valve has moved and there is nothing for the input side to work against to shift it again. Moving the pedals would more than likely just move the free end of the feedback lever not the pilot valve.

Rainy cold day here....second cup of Coffee in hand and some time on my hands while I anticipate the Army-Navy Football game this afternoon at which the World pauses for a few hours (for those like me anyway)!

After reading all of the posts I have garnered some thoughts on this tragedy....mostly relating to the design of the Tail Rotor Control system on the Mishap Aircraft.

I have only the knowledge gathered by means of the posts here and some articles in the media.

I assume the 169 and 189 share a common design ( or identical systems common to the two types) thus the AD's apply only to the 169/189.

The design must not be as "simple" as suggested due to the various discussions upon what role all the various components played in that event.

The conversation about which Nut loosened/tighten and so forth....was indicative of that.

Some questions:

Why did the 169/189 wind up with this particular design of Tail Rotor Control system?

Is the design "new" and "different" to all previous such systems used by AW?

If so....if a new and unique design for the 169/189.....why?

Did the Design Team ever consider such a failure mode during the design and testing process?

If so....how did they resolve any issues that arose?

Is this design "failure tolerant" to the minimum degree necessary assuming inspection and servicing intervals and procedures were properly carried out at the factory and by the operator ?

Is this design an actual improvement over past designs or is it too complicated/complex in design?

Are there design flaws that set the stage for just such a failure as this to occur?

Is there a method to shut off the hydraulic pressure to the Tail Rotor Servo(s) and if so....would the Tail Rotor assume a somewhat neutral position?

(As violent and rapidly as the aircraft reacted to this actual failure.....I am thinking there would have been scant time for analyzing the problem by any Pilot.)

Anyone else have these or similar questions about the AW-169/189 Tail Rotor Control System?

The TR pitch control is hydraulically operated (dual system), not hydraulically assisted. There is no manually 'selectable' TR hydraulic system shut off or manual reversion. This isn't a unique design particular to this helicopter type.

Yes, I have done. If you can be bothered to read back my previous posts on this topic (they date back some 17 years), you will see that I don't disagree with you.The servo ran away to full travel because the servo valve remained open; there was nothing to centralise it once a common part became detached. The common factor / weak point in this design is that the pilot control inputs route via the same lever that nulls the servo; unfortunately that's what came adrift because a single nut disconnected. On other designs, this isn't the case.

But surely the purpose of airworthiness legislation is to ensure that everyone designing aircraft learns from past history; irrespective of which type first had an accident....

Am I correct in assuming the Tail Rotor cannot be controlled manually in the event of a dual hydraulic failure?

Loss of feedback.
Remember as a little boy seeing the remains of a mill boiler house after the steam engine flywheel exploded -because the speed governor was BELT driven from the power output shaft, and the belt had broken!!!!
No loss of life on that occasion, but more by luck than judgement.
Lessons were learned, not least by this budding engineer.

Undoubtedly but the main focus should be as to why a relatively common and foreseeable problem - a bearing binding, would allow a primary control to disconnect itself and be driven to the pitch limits.

It is a single point of failure and, worryingly, one that is designed to fail at some point.

Hi SASless,
Unfortunately, I cannot tell if the T/R system offers any level of T/R manual-only control or, the response of the T/R blade pitch to total loss of Hyd power. I will have to take issue with the writers of the report here. Under the paragraph "Tail rotor control operation" there is scant detail. Particularly, the "servo actuator" component is arrowed twice on the illustrations but not referred to in the para. Additionally, this part of the para "The lever pivots around the connection at the control shaft end and creates a demand on the hydraulic system via the SOLENOID VALVE, which moves the hydraulic piston and control shaft of the actuator" seems to be in error as a solenoid valve is an electrically operated valve. Should one read this to understand that movement of the lever is switching a solenoid valve? Hmmmm

OAP

OAP,part of the yaw channel AFCS...?

Hi sycamore,
No idea I am afraid. My reference to the para "The lever pivots around the connection at the control shaft end and creates a demand on the hydraulic system via the solenoid valve, which moves the hydraulic piston and control shaft of the actuator" is to highlight words that seem to me to be either; badly written, incorrect or confusing.

OAP

Among my Laundry List of questions was this one......

Are we seeing in the various posts before and after my question showing up....confirmation that understanding how this 168/189 Tail Rotor Control system operates is far more difficult than at first glance?

I appreciate all the posts as they come from knowledgeable people and that quality of discussion is very informative.

This is where having a Maintenance Manual for the 169 would be very useful....to read what the Manufacturer has to say about it all.


Oh....and by the way....Army BEAT Navy yesterday! Go Army-Beat Navy! :D:ok:

Not current on any of this, far too old! But proportional solenoid valves are available, so it is possible but not shown on the drawing posted. Maybe a dual operated valve?

Forget the fact that I have type knowledge, from a maintenance engineer's point of view I found the report poorly written, with poor terminology, and confusing. This has been proven by the majority of the posts and questions in this thread since the report was published.
The use of the word 'solenoid', as quoted above, was incorrect and misleading. As has been pointed out, a solenoid is an electrical-mechanical component. I would refer to the item that the report is referring to as a servo valve, or spool valve. These are standard hydraulic servo component terms.
The AW169 tail rotor servo and pitch control system really is pretty straight forward and not particularly unique, compared to some I've seen over the years. It's not always easy though to visualise these things unless you have one in front of you, or a series of detailed photos. I still think that the report could have done a better job of clearly relaying the technical information to the industry and public.

nodrama, perhaps you could elaborate on the `cable` arrangement to the tail ,from the pedals..? ..Is this a new terminology for a `teleflex` control,as cables are usually `pullers,not pushers`..?Why not carbon fibre rods ?
If one had a hyd failure,does the servo have a `bypass`loop,..?to allow manual inputs from the pedals..?
Can the controls be checked on the ground for `full,free and correct` operation prior to start,or only with an ext,hyd rig...?
Could you also confirm the spider pitch links go to the rear of the t/r blades,ie behind the feathering axis....?
thanks in advance...

Having served in both I come out a winner either way SAS.

https://cimg5.ibsrv.net/gimg/pprune....143d872388.png

A ‘flexiball cable’ is what it is called. They’ve been used successfully by several helicopter manufacturers for the last 20+ years (e.g Eurocopter/ Airbus, MD, Agusta), though a flexiball control was attributed to the cause of an EC135 crash in Japan in 2007. The last statement with reference to no maintenance and lubrication isn’t strictly true. They are subject to periodic friction force checks and the eye-end sliders sometimes get greased (depends on aircraft type).

1. See previous posts
2. With one engine running in 'Acc Drive'
3. Yes

nodrama. I would be interested in your thoughts on lubrication of the failed bearing, with the burned 'black grease' on the rod nearby. In my experience with classic cars, black graphite grease can 'dry out' if not used regularly - and this a/c was apparently not used all that frequently?

Nodrama,thanks for that; assume `acc drive` is similar to WX,S-K,Lynx...
Leonardo (the Engineer),said `Friction keeps the World together,Lubrication (grease,oil, money ), allows it to spin``...pity his namesake didn`t take note......

With my 28 yrs of aircraft maintenance involvement I can only agree with nodrama, that report is surprisingly ambiguous - AAIB was always a first-class source of information. Using term "solenoid valve" for a hydraulic system component is softly said, misleading. (except in case there is truly an electromagnetic valve involved - in that case I will have to bury myself )

nodrama, can you comment or confirm that torque applied to castellated nut, that holds duplex bearing in place on control rod, has no effect on the bearing itself i.e. that overtorquing it will not "squeeze" balls between inner and outer race, but "only" overload the rod-threads?
Is duplex bearing actually made of two completelly separate bearings, turned one against the other, or is it one component? From the Fig. 4 in AAIB S2/2018 it seems there are two inner races, but one outer race. Such design implies that nut torque could have an effect on bearing friction? (if there is a gap between inner races-depends on design) Please help clarifying this, if you can!

hoistop

hoistop, I think this is a good question. The AAIB report on page 5 states that this nut was "found to have a torque load significantly higher than the required assembly value".

In addition, on page 7, paragraph 2 "the increased torque load on the castellated nut that remained on the spider end of the shaft is consistent with rotation of the tail rotor actuator control shaft"

Are they saying the jammed bearing would tighten the nut? If the rotor is turning anti-clockwise, the bearing jams and the thread is clockwise it does not seem to make sense. Or does it mean it could have tightened on impact?

on an aircraft the size of a 169 there is no way you are going to move the TR without hydraulic assistance.

Aircraft like the Squirrel or Gazelle are enough of a handful with hyds switched off in yaw.

In the nineties, at BHL Alister Gordon, introduced an amendment to the 212 flight test schedule to include a double hydraulic failure. His reasoning was that if an aircraft had a single failure on a rig, it could still be flown back, on the assumption that a second failure would be controllable. It wasn't, the aircraft each time went into a right hand dive and turn, and the pilot urgently said turn it back on, or words to that effect.
The amendment was withdrawn after Bell said ,"stop being so stupid, your going to kill yourselves".
With regard to the S76 tail control aft quadrant spring centring system, this was fitted to put the system to neutral in the event of a cable brake. It would have no effect on a servo runaway.

malabo - which models? Ones with 2 bladed TRs? I think TR power in these modern aircraft is considerably higher - the 139 for example can deal with a 40 Kts crosswind - I stand by my assertion that manual control isn't possible.

ShyTorque, thanks for your explanation and that's how I initially read the report as well. However if you pick up a bolt and a couple of nuts it would seem that in order to loosen the bolt which came off the torque you need to apply via the inside race of a bearing would tend to loosen the nut holding that bearing on also. I am lost as to how this can happen the other way around. Do the actual directions make sense to you? If so I must be misunderstanding the way it is assembled.

It wasn't you. :rolleyes:

I don't know if 2-bladed TR systems take less control input force than three bladed, I'm just expecting some correlation between weight of helicopter and the tail rotor thrust, or control input force. AW169 gross (increased with a kit) is 10582 lbs. Bell mediums with a single hydraulic to the tail-rotor were the 430 - 9300, 205 - 11,200, 212 - 11,200, 412 - 11,900. All these are flyable and landable with no hydraulics to the tail-rotor servos. I had to do it in training for those types and I trained others to do it later.

I'm hesitating on the AW169 because I'm unsure of the mechanics of the Teleflex system and whether the leverage and control strength is there. Agusta/Leonardo, like Bell, can build some pretty stout pedals though, judging from the abuse they can take on the AW139 just setting the parking brake. It would be a major surprise to me if the 169 tail rotor wasn't controllable without hydraulics Gotta eat crow and agree with crab on this one, given no drama's patient explanation, but it has two systems and the main rotor isn't controllable without at least one, so a somewhat moot point that only serves to pad the thread while we wait for answers.

Not thread drift, but just to correct any misunderstandings, the 212 main rotor is controlled through a stab bar and can be flown with both hydraulics shut off, of course, otherwise Bell could not have certified it with that switch logic. Flight Check Procedures require shutting both off in flight at 70 knots - look it up. In my early days in the industry we expected pilots to be able to land it with both systems off, and we all did, and when we became instructors we taught it. Even spaghetti-armed Brit pilots that found it impossible at BHL, magically found the strength on this side of the Atlantic where failure meant no job.

It isn't possible to even move the flying controls on an AW169 without at least one hydraulic system pressurised. That's why I made the point of saying earlier that the MR/ TR pitch control is hydraulically operated and NOT hydraulically assisted. There is a major flying control design difference. The pilot input isn't direct to the rotor pitch control, but to the hydraulic servo input valves. If the pressurised servos aren't commanded to move, nothing is moving. That's why there are two systems, for redundancy. How much clearer does it have to be?

I think we all understand.

M25,

The tail rotor hub and blades on a 169 are on the right hand side of the tail boom and rotate from nose to tail at the top (i.e anti-clockwise as you look at them from the right hand side of the aircraft). The control shaft sits inside the outer drive shaft and hub. When the duplex bearing began to seize, the control shaft would have also tended to rotate in the same direction as the hub, i.e. anti-clockwise. *Any drag on the nut on that end would tend to tighten it because it has a right hand thread; it would be the same as tightening any normal right hand threaded nut and bolt.

Edit: *As the hub rotates in its entirety, this cannot actually have been the case, my error!

If you now move to the left side of the aircraft, the tail rotor and "errant" control shaft shaft appear to be moving in a relatively clockwise direction. The castellated nut on that side also has a right hand thread. With the shaft rotating clockwise, any drag on the nut (i.e. from contact with the stationary pin carrier to which it was bolted) would tend to cause the two to be unscrewed.

If instead the control shaft and its nut on that end had a left handed thread, the relative motion would have tended to tighten them up, as is the case on the right hand side of the assembly. Whether that would have helped prevent the catastrophic failure, I really don't know. If the design of the pin carrier had allowed the control shaft and nut to spin freely together, they would have presumably stayed together, even in rotation.

Regarding the TR control, I think there are several slightly different angles at play here: 1) In the control system on the 169 physically connected so that it's possible, given enough force to change the pitch (apparently no). 2) Would it be possible, considering the forces at play, to design a control system with a manual fallback.

If 2) is true, I guess one could discuss if 1) was a smart design choice. More than the TR control system in isolation should probably be considered if such, if the aircraft won't fly without hydraulics for other reasons, manual fallback for the TR might be completely pointless.

Generally I always prefer manual fallbacks, but they even make cars where the steering wheel and brakes has no such fallback these days. I guess it's part of a trend of over-confidence in system designs that leads to arrogance. They can probably also save some money designing a system without such fallback in many cases. I have no idea what considerations are behind the 169 control system were though, so this isn't meant as a speculation for their reasoning.

I suspect that as has been stated here before, the forces required to move the tail rotor blades on such a big aircraft would be too large for the pilot to manage without assistance from a hydraulic servo. I know that from personal "feet on" experience on rather smaller aircraft where the hydraulics become no longer available (as we know, many smaller types of helicopter such as the Squirrel, A109 etc can revert to manual flight control, but even then it's very hard work to control them).

What would be better would be a design arrangement that "self centres" the pitch angle of the blades to a pre-determined, neutral setting if the pilot's normal yaw control system is lost. This would allow flight to be continued, to some sort of a controlled running landing, as on other helicopters. Pilots can be trained how to handle this lesser, although still serious, type of emergency.

Interestingly, all the 'bearings' in the AW169 tail rotor (flap/ pitch/ lead-lag) are elastomeric. I'm wondering whether, with no hydraulic pressure to the TR control servo, the TR blades would return to their 'neutral' position by themselves under the force of the elastomeric's wanting to return to their normal unloaded position? I could try it with a hydraulic rig sometime, but it wouldn't take into consideration any dynamic loads that would be there in flight.

Where they naturally revert to depends on their CTM / ATM ratio... but the aerodynamic loads are very large and would easily overcome any tendency to "self centre" at rest via the elastomerics.

When the shaft unwound itself from the nut, presumably the shaft was then free to float in and out, so it was the buoyancy of the blades that pulled it out. These blade grips could have had bob weights fitted to keep them neutral if control went limp.

Might have been different if he had done a yaw check after initial lift-off...
Todays helicopters are designed with much shorter and more blades giving a higher `solidity ratio`,and shorter moment arms to the tail rotor,with almost as much `area` in front of the rotor mast as behind,than `classics like the WX,S-K,etc,and giving reduced directional stability...
In that case in the event of a t/r failure,it should be possible to design the CTM/ATM to position to a net thrust against power bias,for control failures,and `fuses` in the hydraulic system in the t/r system in the case of a t/r hyd leak/failure there.