AS332L2 Ditching off Shetland: 23rd August 2013
Yes, although note its in a section entitled "manufacturers review of recorded data" so once AAIB realise its wrong, they will be able to blame the French!
Join Date: Mar 2001
Location: UK
Posts: 206
Likes: 0
Received 0 Likes
on
0 Posts
Anyway "it" wasn't caused by any of the above, "it" was caused by prolonged inattention to the airspeed having previously intentionally destabilised the approach by substantially reducing power to slow down.
Anyway, as said, doesn't really matter as the real cause of the accident was failure to monitor the basics and perhaps an over reliance of AP.
Last edited by obnoxio f*ckwit; 18th Oct 2013 at 20:48. Reason: To add last comment
The "grapevine" informs me that the captain had fairly recently come back onto the L2 after being on the EC225.
I leave it to HC to derive the various possibilities and potential for this to exacerbate the risk of human error.
If my information is wrong, I apologise now.
I leave it to HC to derive the various possibilities and potential for this to exacerbate the risk of human error.
If my information is wrong, I apologise now.
HC,
The second time I went to the linked report some additional pages showed up.....which answers most of my questions which came after the first go.
Perhaps part of it did not load or something.
The second time I went to the linked report some additional pages showed up.....which answers most of my questions which came after the first go.
Perhaps part of it did not load or something.
Colibri
When asked in an oil and gas forum both EC and CHC said that this was not the case.
The "grapevine" informs me that the captain had fairly recently come back onto the L2 after being on the EC225.
I leave it to HC to derive the various possibilities and potential for this to exacerbate the risk of human error.
If my information is wrong, I apologise now.
I leave it to HC to derive the various possibilities and potential for this to exacerbate the risk of human error.
If my information is wrong, I apologise now.
Join Date: Jul 2013
Location: UK
Posts: 193
Likes: 0
Received 0 Likes
on
0 Posts
The exact rate of descent at impact is not known, but the impact was survivable
followed by a "100 FEET" automated call two seconds before impact with the surface of the sea
Guest
Posts: n/a
failure to monitor the basics and perhaps an over reliance of AP.
Join Date: Feb 2001
Location: Europe
Posts: 535
Likes: 0
Received 0 Likes
on
0 Posts
GipsyMagpie
I can't believe I'm reading your last post - maybe it's time you reconsidered your career. You're not really going to tell me that an ATPL/CPL needs to be warned that low collective settings can result in excessive rates of descent when the AFCS is coupled to VS are you?? If so you just should not be flying - your understanding of basic physics, let alone absorption of training, is simply inadequate.
With regard to the hazards of AFCSs, just how much nannying do you want? So you want 4 axis? And who or what is going to ensure you put the correct parameters into the system? The frequency for the ILS even? Maybe you'd like the whole lot automated so you can sit there all day doing nothing?
The more automation you have the more scope for boredom, over-reliance and degradation of monitoring and even hand flying skills. Of course there is a balance - no one is going to advocate SP hand flown ILSs. But given the number of instrument approaches that are flown quite acceptably with 3 axis AFCS systems, the current balance is well within the scope of two well trained pilots, concentrating on the job.
Maybe in this incident use of AS hold rather than VS would have been better - at least then the consequence of inadequate power set would have been a higher ROD from the outset and when the radalt warning sounded, assuming neither crew picked up the ROD cue, pitching up into a climb with the heave of collective would have avoided the VRS that took them into the sea.
I can't believe I'm reading your last post - maybe it's time you reconsidered your career. You're not really going to tell me that an ATPL/CPL needs to be warned that low collective settings can result in excessive rates of descent when the AFCS is coupled to VS are you?? If so you just should not be flying - your understanding of basic physics, let alone absorption of training, is simply inadequate.
With regard to the hazards of AFCSs, just how much nannying do you want? So you want 4 axis? And who or what is going to ensure you put the correct parameters into the system? The frequency for the ILS even? Maybe you'd like the whole lot automated so you can sit there all day doing nothing?
The more automation you have the more scope for boredom, over-reliance and degradation of monitoring and even hand flying skills. Of course there is a balance - no one is going to advocate SP hand flown ILSs. But given the number of instrument approaches that are flown quite acceptably with 3 axis AFCS systems, the current balance is well within the scope of two well trained pilots, concentrating on the job.
Maybe in this incident use of AS hold rather than VS would have been better - at least then the consequence of inadequate power set would have been a higher ROD from the outset and when the radalt warning sounded, assuming neither crew picked up the ROD cue, pitching up into a climb with the heave of collective would have avoided the VRS that took them into the sea.
There wasn't much automation in the S61 that BAH flew into the water in 1983! I don't believe this is an automation issue - it's a monitoring and awareness issue.
Try switching off the cylic trim then letting go of the cyclic - that's pretty horric too.........
Just engage ALT at height and put the lever down. Watch the airspeed wash off and the aircraft fall out of the sky. Horrific.
Join Date: Aug 2012
Location: Hassocks, Mid-Sussex
Age: 67
Posts: 278
Likes: 0
Received 0 Likes
on
0 Posts
A HELICOPTER boss urged people not to jump to conclusions about pilots involved in a fatal crash after investigators revealed they found no technical fault with the aircraft.
Vice-president of quality and safety for CHC Helicopters Duncan Trapp urged people to realise that the report was not conclusive and defended the training of the pilots.
He said: “It is important to remember that the investigation is ongoing and nothing has been concluded as to why this accident happened.”
Vice-president of quality and safety for CHC Helicopters Duncan Trapp urged people to realise that the report was not conclusive and defended the training of the pilots.
He said: “It is important to remember that the investigation is ongoing and nothing has been concluded as to why this accident happened.”
Maybe in this incident use of AS hold rather than VS would have been better - at least then the consequence of inadequate power set would have been a higher ROD from the outset and when the radalt warning sounded, assuming neither crew picked up the ROD cue, pitching up into a climb with the heave of collective would have avoided the VRS that took them into the sea.
Ask yourself which Flight Control was most crucial in this whole thing....and which parameter is the one that is required to change to carry out the maneuver?
In this case....altitude change was the primary focus and airspeed could have been left alone from the start.
I know it sounds a bit elementary to suggest a well trained crew can not adjust two parameters simultaneously.....and that is not what I am saying.
If an Approach Airspeed had been set first.....then the descent initiated.....then an incorrect power setting would have been the very first thing to show up and be corrected as the Collective would have been an "Active" control. One Pilot would have had his grubby mitt busy working the Lever to maintain the ROD while the AFCS was maintaining the Airspeed.
Be the Approach Speed been 110 Kts or 80 Kts......it would have made very slight difference. It is not like the aircraft were lined up by the dozens to land at Sumbrugh that day.
Unfortunately for Duncan, people are going to jump to conclusions. An outsider, reviewing CHC's strategy of selling off their training assets and outsourcing a good chunk of their training for no reason other than increasing capital and profit, would conclude that they were not really interested in quality of training, merely ticking a box to satisfy the minimum standard.
Personally, having done my L2 conversion at CHC I am aware that the standard was similar to Bristow at the time (10 years ago), although I do wonder if that is still the case, despite the best efforts of the folk at the sharp end. Not many outside observers are in my position though.
Personally, having done my L2 conversion at CHC I am aware that the standard was similar to Bristow at the time (10 years ago), although I do wonder if that is still the case, despite the best efforts of the folk at the sharp end. Not many outside observers are in my position though.
Join Date: Dec 2003
Location: Scotland
Posts: 261
Likes: 0
Received 0 Likes
on
0 Posts
I don't know about the L2 training, but having completed a 332L conversion with CAE in Stavanger, I can only comment on my experiences. The standard of simulator training was outstanding! The pre-session briefings were the best I have every had in 35 years of military and civil flying. Once in the sim, the pace and content were again spot on, perfectly tailored to the crew's experience levels. Having spent 11 years at Bristow, I can honestly say that the CAE training was as good if not better.
I don't think the standard of training per se is the reason this tragic accident occurred. It was, as always, a combination of human errors, from the design of the AFCS system, through company management culture to the pilots themselves. And not forgetting the oversight, or lack of, by the CAA?
bondu
I don't think the standard of training per se is the reason this tragic accident occurred. It was, as always, a combination of human errors, from the design of the AFCS system, through company management culture to the pilots themselves. And not forgetting the oversight, or lack of, by the CAA?
bondu
26500-done
Bondu - not saying CAE training is bad necessarily, just that outsourcing your training as a policy looks bad for a large/mainstream operator. Although I think there will be an element of pot luck as to whether you get an instructor with any experience on type or in role.
The other issue is the flexibility that being in control of your own simulator brings. Not an issue when things go to plan, but when you have a candidate who is marginal and really needs more training, there is a strong temptation to just pass them because you know that there is no more sim time available at short notice, and anything else is too difficult. This is something I found a problem at Helisim, but not once we got our own ones.
Of course its not "required" to have your own sim, but when you do, to sell it is a big backwards step that says a lot to an outsider about your attitudes.
Bondu - not saying CAE training is bad necessarily, just that outsourcing your training as a policy looks bad for a large/mainstream operator. Although I think there will be an element of pot luck as to whether you get an instructor with any experience on type or in role.
The other issue is the flexibility that being in control of your own simulator brings. Not an issue when things go to plan, but when you have a candidate who is marginal and really needs more training, there is a strong temptation to just pass them because you know that there is no more sim time available at short notice, and anything else is too difficult. This is something I found a problem at Helisim, but not once we got our own ones.
Of course its not "required" to have your own sim, but when you do, to sell it is a big backwards step that says a lot to an outsider about your attitudes.
I agree with Bondu. Having been a sim instructor (on the 332s in fact), I know the guys who instruct full time provide better instruction than pulling guys off the line to instruct part time, which was my role. My experience with CAE instructors has been first rate.
Last edited by Bladestrike; 19th Oct 2013 at 14:37.
However is it also not true that you can get a job with CAE, do a type rating course in the Sim, and then instruct on it having never flown it, and maybe never even flown offshore? This doesn't mean that your instructional technique will be poor, but I suggest you will lack an insight into the real world of operating that type in that role.
Time History
The time-history included in the report raises a few questions.
A 40 degree heading change at the end, yet the pedal position isn't on the trace.
The derived radio altitude rate does not square with the rate one derives by taking the slope of the proportional traces of baro and radio altitude at the end. Radio altititude rate signals have been suspect in other applications.
VRS, at least in the SA models I've been exposed to, typically results in a significant increase in cockpit N/rev excitation. Would be good to see the accelerometer traces along the same timeline as the Figure 1 data. On the same plot, would be ideal.
Seems a rush to judgement and perhaps a too convenient action to just hang it all on VRS, at least based on that one Figure as presented.
Would be helpful to see the AFCS/Autopilot commands and actuator out puts along the same time base.
A 40 degree heading change at the end, yet the pedal position isn't on the trace.
The derived radio altitude rate does not square with the rate one derives by taking the slope of the proportional traces of baro and radio altitude at the end. Radio altititude rate signals have been suspect in other applications.
VRS, at least in the SA models I've been exposed to, typically results in a significant increase in cockpit N/rev excitation. Would be good to see the accelerometer traces along the same timeline as the Figure 1 data. On the same plot, would be ideal.
Seems a rush to judgement and perhaps a too convenient action to just hang it all on VRS, at least based on that one Figure as presented.
Would be helpful to see the AFCS/Autopilot commands and actuator out puts along the same time base.
Hi john
Yes, I noticed a bit of a "disconnect" between the radalt and the derived vertical speed at the end. You have to remember that the L2 was designed in the late 80s and certified in the early 90s and as such, has limited "extra" parameters on the FDR and mostly at the minimum sample rates. I think the lat and long accelerations are only at 4Hz and might even be 2 Hz, therefore well below the Nyquist frequency for 1-per-rev vibration. Also I'm pretty sure that only the pilot's control positions are in the data frame, not the AFCS series actuator outputs. The low sample rate might also explain the radalt / vertical speed disconnect - maybe they hybridised their VS parameter with some integrated normal acceleration data - that parameter being the one with the highest sample rate.
Yes, I noticed a bit of a "disconnect" between the radalt and the derived vertical speed at the end. You have to remember that the L2 was designed in the late 80s and certified in the early 90s and as such, has limited "extra" parameters on the FDR and mostly at the minimum sample rates. I think the lat and long accelerations are only at 4Hz and might even be 2 Hz, therefore well below the Nyquist frequency for 1-per-rev vibration. Also I'm pretty sure that only the pilot's control positions are in the data frame, not the AFCS series actuator outputs. The low sample rate might also explain the radalt / vertical speed disconnect - maybe they hybridised their VS parameter with some integrated normal acceleration data - that parameter being the one with the highest sample rate.
Last edited by HeliComparator; 19th Oct 2013 at 16:49.
Time History Data
Thanks, HC.
I was referring to the 4/rev excitation, not the 1/rev, but as you wrote, perhaps the sample rate didn't support that range ( I didn't work it out, but am sure its 16-21 Hz or so ).
Is it possible that the AFCS/Autopilot Computer has some memory built in? If so, I'd bet that it would have the frequency range to assist.
Think they didn't put up the directional control trace because there is none? Most surprising if true.
All in all, one would hope that there is a lot more applicable data to review than that published in the single figure. I have to believe that the pilots office at EC is into this 100% and can assist. Just guessing, but would not be surprised if they took the VRS sedative with some skepticism as well.
I was referring to the 4/rev excitation, not the 1/rev, but as you wrote, perhaps the sample rate didn't support that range ( I didn't work it out, but am sure its 16-21 Hz or so ).
Is it possible that the AFCS/Autopilot Computer has some memory built in? If so, I'd bet that it would have the frequency range to assist.
Think they didn't put up the directional control trace because there is none? Most surprising if true.
All in all, one would hope that there is a lot more applicable data to review than that published in the single figure. I have to believe that the pilots office at EC is into this 100% and can assist. Just guessing, but would not be surprised if they took the VRS sedative with some skepticism as well.