Helicopter down outside Leicester City Football Club
Regarding duplex bearing side nut.
Yes - as long as God was flying alongside holding the nut:-)
Thing is that the nut is not held by anything. Instead the shaft is prevented from rotating because it is clamped to the feedback arm on the far side. If the inner race began to turn on the shaft because of forces transmitted through the failing bearing the inner race would rotate anti-clockwise relative to the shaft. This might tend to UN-screw the nut on the shaft.
Since the split pin on that nut remained intact the nut was not rotated very much if at all. The AAIB say that the release torque on the nut was higher than expected - however no cause or explanation for this is actually stated. Tightening was not caused by the nut being rotated by forces transmitted through the failed duplex bearing.
Two earlier posts mentioned this aspect of the geometry also, I apologise that I have not credited them but at present I don't want to take the time to look for the posts.
Thing is that the nut is not held by anything. Instead the shaft is prevented from rotating because it is clamped to the feedback arm on the far side. If the inner race began to turn on the shaft because of forces transmitted through the failing bearing the inner race would rotate anti-clockwise relative to the shaft. This might tend to UN-screw the nut on the shaft.
Since the split pin on that nut remained intact the nut was not rotated very much if at all. The AAIB say that the release torque on the nut was higher than expected - however no cause or explanation for this is actually stated. Tightening was not caused by the nut being rotated by forces transmitted through the failed duplex bearing.
Two earlier posts mentioned this aspect of the geometry also, I apologise that I have not credited them but at present I don't want to take the time to look for the posts.
Join Date: May 2010
Location: Boston
Age: 73
Posts: 443
Likes: 0
Received 0 Likes
on
0 Posts
The shaft was still attached (well actually part of if I read the diagram correctly) to the hydraulic actuator that was hard over due to lack of feedback.
BTW: As another example of very bad outcome from failed feedback the preliminary report on the Lawrence MA gas explosions reveals that the overpressure (75psi in a low pressure 1/2psi line ) was caused by switching to a newly installed pipe section with out moving the pressure sensing lines from the decommissioned ancient iron pipe. This caused the regulators to go full open, why there was no local override path on the regulators is a different question.
Last edited by MurphyWasRight; 11th Dec 2018 at 20:46. Reason: typos
Avoid imitations
Join Date: Nov 2000
Location: Wandering the FIR and cyberspace often at highly unsociable times
Posts: 14,572
Received 412 Likes
on
217 Posts
Regarding duplex bearing side nut.
Yes - as long as God was flying alongside holding the nut:-)
Thing is that the nut is not held by anything. Instead the shaft is prevented from rotating because it is clamped to the feedback arm on the far side. If the inner race began to turn on the shaft because of forces transmitted through the failing bearing the inner race would rotate anti-clockwise relative to the shaft. This might tend to UN-screw the nut on the shaft.
Since the split pin on that nut remained intact the nut was not rotated very much if at all. The AAIB say that the release torque on the nut was higher than expected - however no cause or explanation for this is actually stated. Tightening was not caused by the nut being rotated by forces transmitted through the failed duplex bearing.
Two earlier posts mentioned this aspect of the geometry also, I apologise that I have not credited them but at present I don't want to take the time to look for the posts.
Yes - as long as God was flying alongside holding the nut:-)
Thing is that the nut is not held by anything. Instead the shaft is prevented from rotating because it is clamped to the feedback arm on the far side. If the inner race began to turn on the shaft because of forces transmitted through the failing bearing the inner race would rotate anti-clockwise relative to the shaft. This might tend to UN-screw the nut on the shaft.
Since the split pin on that nut remained intact the nut was not rotated very much if at all. The AAIB say that the release torque on the nut was higher than expected - however no cause or explanation for this is actually stated. Tightening was not caused by the nut being rotated by forces transmitted through the failed duplex bearing.
Two earlier posts mentioned this aspect of the geometry also, I apologise that I have not credited them but at present I don't want to take the time to look for the posts.
I too assumed your (and others) proposed mechanism of tightening as I read the report. I then noticed that the mechanism was wrong and re-read the report carefully. The AAIB do not make any statement as to the mechanism of excessive tightening. Perhaps it would have been better if they had stated that explicitly as quite a few people here (and presumably other readers too) have made that incorrect (I believe) connection in the context presented.
Maybe it was tightened too much last time it was assembled? How about some fretting during the failure process creating rough surfaces or loose material within the joints? I just made the last one up, I have no idea if such a thing might actually occur.
I can't see that it is important since I doubt very, very much that the tight nut had anything to do with the crash. Well unless the bearing was designed to have it's preload set by the nut torque. Excessive preload would certainly explain the failure of the bearing. My experience of such bearing configurations (not aviation) is that they are always done up hard and the dual race bearing has a small clearance within it.
We'll likely find our next year.
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.
Try googling servo hardover.....
Am with chop on this one instead of looking down on him with your expert opinions would someone please kindly explain how that control shaft was still connected to the actuator when It was disconnected from the lever mechanism I can only see two fixing points on that shaft top and bottom?????
For those confused by the reported "significantly higher" disassembly torque of the nut at the duplex bearing end. One would hope that the final report will explain this observation. However, I believe that the initial report referred to here makes no observation about the split pin locking of that nut and so, presumably, the split pin and that nut were in their position, as fitted.
OAP
OAP
Avoid imitations
Join Date: Nov 2000
Location: Wandering the FIR and cyberspace often at highly unsociable times
Posts: 14,572
Received 412 Likes
on
217 Posts
I can understand what Chopjock is asking.
However, the hydraulic servo is designed to control the pitch angle of the blades, not the other way round. The self centreing effect of the blades (if any) isn't enough to overcome the power of the servo. The servo is obviously designed to be more than powerful enough to push / pull them into the required position. If the servo control valve remains open, because there is nothing to close it, the servo will drive the blade pitch angle to full travel, in this case it must have gone to full negative pitch, or close to it.
However, the hydraulic servo is designed to control the pitch angle of the blades, not the other way round. The self centreing effect of the blades (if any) isn't enough to overcome the power of the servo. The servo is obviously designed to be more than powerful enough to push / pull them into the required position. If the servo control valve remains open, because there is nothing to close it, the servo will drive the blade pitch angle to full travel, in this case it must have gone to full negative pitch, or close to it.
Join Date: Jan 2011
Location: on the cusp
Age: 52
Posts: 217
Likes: 0
Received 0 Likes
on
0 Posts
Am with chop on this one instead of looking down on him with your expert opinions would someone please kindly explain how that control shaft was still connected to the actuator when It was disconnected from the lever mechanism I can only see two fixing points on that shaft top and bottom?????
jimjim1 You are right regarding castellated nut holding duplex bearing. It is obvious that if failing bearing caused inner race to start rotating on shaft, this action would put UNSCREWING force to the nut if it has a normal, right tread - I believe it is. But nut stayed in place and only increased torque was reported. Will come back to that later.
Hovewer, torque transmitted to control shaft was strong enough to overcome the clamping force of castellated nut on the other side-at control lever and its pin carrier. Control shaft and nut (plus still intact but strained cotter pin) started rotating in pin carrier. I would speculate that nut immediatelly moved a little in UNSCREWING direction, thus slightly decreasing clamping force, (due to slack and/or initial deformation of cotter pin) . Enormous heat and clamping pressure then caused the nut to friction-weld to carrier pin, thus resisting rotation. Now cotter pin could not withstand the torque between control shaft and firmly welded nut and it snapped. Control shaft started unscrewing out of nut. Once shaft threads disengaged completelly, (it was very fast-just the length of the threads) control was lost. I suspect that while unthreading, rod moved (quickly) to the right for the length of the tread and changed pitch of blades, causing initial right yaw (?)
It remains unclear to me what exactly happened then , as in S2-2018 report there is not enough information how exactly hydraulic power cylinder engages with control rod and I do not have type knowledge-wether control shaft was hydraulically moved to the physical limit of actuator, or it was just free to move, thus leaving blades to rotate in pitch axis until equilibrium of all forces was established or pushed by runaway actuator.
Report says that T/R actuator continued changing the pitch of the blades until they reached physical limit. Sentence above says that rod was disconnected from lever.
Duplex bearing - if its castellated nut would unscrew or at least allow rotation of inner race on control shaft, the whole arrangement would still function until heating would weld inner race to the shaft. (with some lubricant, it might not happen very fast) Hovewer, evidence suggest that OUTER race rotated in its place. Wether inner race also rotated, report doesn`t say. I hope AAIB CT scanned also (duplex bearing side) castellated nut before they removed the cotter pin and found (abnormally high) nut torque. Any tell-tale deformation of the cotter pin would be altered-destroyed with removal.
I am not that surprised that nut torque on duplex bearing nut was found abnormally high, considering significant overheating, possible movement between to-be-clamped parts.. .
It often happens that correctly torqued assembly needs significantly higher unscrewing torque when disassembled again after a while, especially if exposed to elements, heat and dynamic loading.
Of course, if it is left-treaded, than torque increase is somewhat expected.
Bearing seizure obviously caused the accident, but what caused bearing failure, remains to be explained. Hovewer, I feel a bit uneasy that such failure dooms the helicopter.
I am trying to figure out if a feature, allowing rotation and thus integrity of flight controls, could (should?) be designed into the system, like alternative way of rotation between slider and control rod if bearing seizes-up and providing tell-tale sign (vibration) that would give time to pilot to recognize the problem and act before everything goes South. And I have a very personal interest in that.
hoistop
P.S. Could anyone provide a diagram of AW169 T/R ctrl sys that explains this description from S2/2018 report:
The lever pivots around the connection at the control shaft end and creates a demand on the hydraulic system vie the solenoid valve, which moves the hydraulic piston and control shaft of the actuator.
Hovewer, torque transmitted to control shaft was strong enough to overcome the clamping force of castellated nut on the other side-at control lever and its pin carrier. Control shaft and nut (plus still intact but strained cotter pin) started rotating in pin carrier. I would speculate that nut immediatelly moved a little in UNSCREWING direction, thus slightly decreasing clamping force, (due to slack and/or initial deformation of cotter pin) . Enormous heat and clamping pressure then caused the nut to friction-weld to carrier pin, thus resisting rotation. Now cotter pin could not withstand the torque between control shaft and firmly welded nut and it snapped. Control shaft started unscrewing out of nut. Once shaft threads disengaged completelly, (it was very fast-just the length of the threads) control was lost. I suspect that while unthreading, rod moved (quickly) to the right for the length of the tread and changed pitch of blades, causing initial right yaw (?)
It remains unclear to me what exactly happened then , as in S2-2018 report there is not enough information how exactly hydraulic power cylinder engages with control rod and I do not have type knowledge-wether control shaft was hydraulically moved to the physical limit of actuator, or it was just free to move, thus leaving blades to rotate in pitch axis until equilibrium of all forces was established or pushed by runaway actuator.
Report says that T/R actuator continued changing the pitch of the blades until they reached physical limit. Sentence above says that rod was disconnected from lever.
Duplex bearing - if its castellated nut would unscrew or at least allow rotation of inner race on control shaft, the whole arrangement would still function until heating would weld inner race to the shaft. (with some lubricant, it might not happen very fast) Hovewer, evidence suggest that OUTER race rotated in its place. Wether inner race also rotated, report doesn`t say. I hope AAIB CT scanned also (duplex bearing side) castellated nut before they removed the cotter pin and found (abnormally high) nut torque. Any tell-tale deformation of the cotter pin would be altered-destroyed with removal.
I am not that surprised that nut torque on duplex bearing nut was found abnormally high, considering significant overheating, possible movement between to-be-clamped parts.. .
It often happens that correctly torqued assembly needs significantly higher unscrewing torque when disassembled again after a while, especially if exposed to elements, heat and dynamic loading.
Of course, if it is left-treaded, than torque increase is somewhat expected.
Bearing seizure obviously caused the accident, but what caused bearing failure, remains to be explained. Hovewer, I feel a bit uneasy that such failure dooms the helicopter.
I am trying to figure out if a feature, allowing rotation and thus integrity of flight controls, could (should?) be designed into the system, like alternative way of rotation between slider and control rod if bearing seizes-up and providing tell-tale sign (vibration) that would give time to pilot to recognize the problem and act before everything goes South. And I have a very personal interest in that.
hoistop
P.S. Could anyone provide a diagram of AW169 T/R ctrl sys that explains this description from S2/2018 report:
The lever pivots around the connection at the control shaft end and creates a demand on the hydraulic system vie the solenoid valve, which moves the hydraulic piston and control shaft of the actuator.
Still don't understand how a servo hardover could full pitch them blades all the way when the control shaft (inner) was found inside the drive shaft (outer) some engineering that for a hydraulic system to still be able to push/pull on that shaft why its inside there surely it was aerodynamics that pitched the blades after the failures. Someone please make me eat humble pie
Avoid imitations
Join Date: Nov 2000
Location: Wandering the FIR and cyberspace often at highly unsociable times
Posts: 14,572
Received 412 Likes
on
217 Posts
Bearing seizure obviously caused the accident, but what caused bearing failure, remains to be explained. Hovewer, I feel a bit uneasy that such failure dooms the helicopter.
I am trying to figure out if a feature, allowing rotation and thus integrity of flight controls, could (should?) be designed into the system, like alternative way of rotation between slider and control rod if bearing seizes-up and providing tell-tale sign (vibration) that would give time to pilot to recognize the problem and act before everything goes South. And I have a very personal interest in that.
I am trying to figure out if a feature, allowing rotation and thus integrity of flight controls, could (should?) be designed into the system, like alternative way of rotation between slider and control rod if bearing seizes-up and providing tell-tale sign (vibration) that would give time to pilot to recognize the problem and act before everything goes South. And I have a very personal interest in that.
If the pin carrier had been designed to contain a simple bearing allowing free shaft rotation of the control end of the shaft, instead of clamping it, the nut would not have come undone.
Last edited by ShyTorque; 12th Dec 2018 at 17:53. Reason: grammar!
Many thanks nodrama for your effort!
I understand this system is a classic follow-up mechanism.
eee... I am still confused a bit. Would that mean that control shaft itself is actually a hydraulic piston with attachments on both sides?? It means when it started rotating O rings/packings/seals became runners for fast spinning control shaft??? I really wonder how this control shaft looks like when it`s put on the table...
ShyTorque my slight uneasiness on this design feature is with pun intended... I agree with you totally.
Hovewer, there are many many individual parts in each helicopter, that will make it stop flying immediatelly in case of failure/breakage. For instance, there are over 40 bolts in Bell 412 flight control system that are replaced every 1000 hours, b`couse - any of these breaks - bye, bye.... plus many other things. Truly, these are simple items and thus very reliable, (until someone fails to put in a cotter pin. and no one notices...) A failed bearing is another story and is not so uncommon.
hoistop
I understand this system is a classic follow-up mechanism.
eee... I am still confused a bit. Would that mean that control shaft itself is actually a hydraulic piston with attachments on both sides?? It means when it started rotating O rings/packings/seals became runners for fast spinning control shaft??? I really wonder how this control shaft looks like when it`s put on the table...
ShyTorque my slight uneasiness on this design feature is with pun intended... I agree with you totally.
Hovewer, there are many many individual parts in each helicopter, that will make it stop flying immediatelly in case of failure/breakage. For instance, there are over 40 bolts in Bell 412 flight control system that are replaced every 1000 hours, b`couse - any of these breaks - bye, bye.... plus many other things. Truly, these are simple items and thus very reliable, (until someone fails to put in a cotter pin. and no one notices...) A failed bearing is another story and is not so uncommon.
hoistop
Join Date: Mar 2007
Location: UK
Posts: 321
Likes: 0
Received 0 Likes
on
0 Posts
I think so. Never seen the inside of a servo that’s been stripped down. I do know that the pitch control rod is one part though, from input end to output end (where the nuts attach).
Last edited by nodrama; 12th Dec 2018 at 15:20.
Join Date: Dec 2018
Location: UK
Posts: 4
Likes: 0
Received 0 Likes
on
0 Posts
jimjim1 You are right regarding castellated nut holding duplex bearing. It is obvious that if failing bearing caused inner race to start rotating on shaft, this action would put UNSCREWING force to the nut if it has a normal, right tread - I believe it is. But nut stayed in place and only increased torque was reported.
In any case I'm sure the issue of whether any possible range of torque values could contribute to a failure of this particular design of duplex bearing will be looked at.
If the pin carrier had been designed to contain a simple bearing allowing rotation of the at the control end of the shaft, instead of clamping it, the nut would not have come undone.
Many thanks to nodrama for the explanations here - without the pictures it's very hard to understand what is going on here. Have been lurking here since the beginning, with increasing horror.
Join Date: Jan 2013
Location: Midlands
Posts: 136
Likes: 0
Received 0 Likes
on
0 Posts
To muddy the waters, the same figure also indicates a mechanical link from the control cable to the spider, but a number of posts have said it can't move without the hydraulics.
Last edited by Pozidrive; 12th Dec 2018 at 16:42. Reason: quoted wrong figure no.
