Lost Erebus tape holds vital clues
Join Date: Jun 2001
Location: UTC +8
Posts: 2,626
Likes: 0
Received 0 Likes
on
0 Posts
Desert Dingo:
The setting of "go-around" thrust (94% N1...?) should never be a concern nor a procedure during a GPWS response in IMC. The only correct response is to "firewall" the throttles.
Collins did not do that and he did not get the benefit of the significant overboost capability of the CF6 engines, ...in a very cold climate...probably up to 15%.
...This indicates to me that the power increase had already been made and the call was to confirm that the power was set correctly. Anyone got SOPs for a ground proximity warning that is not along the lines of:
1. Apply power and rotate to climb attitude.
2. Check configuration (gear flaps speedbrakes etc)
3. Confirm thrust set.
1. Apply power and rotate to climb attitude.
2. Check configuration (gear flaps speedbrakes etc)
3. Confirm thrust set.
Collins did not do that and he did not get the benefit of the significant overboost capability of the CF6 engines, ...in a very cold climate...probably up to 15%.
Last edited by GlueBall; 13th Dec 2004 at 19:43.
Join Date: Apr 1999
Posts: 326
Likes: 0
Received 0 Likes
on
0 Posts
There have been several questions/statements on their altitude and letdown procedures so wanted to post this info which I was given by one of the DC10 pilots who testified and was on the transcript team.
Letdown procedures were given verbally at the briefing.
Approval was given for 1500ft
Approval was given to use own initiative on letdown procedure and approach
Testimony at trial was given by briefing pilot John Wilson and AirNZ Operations Manager Doug Keating
-------------------------------------------------------------------
Rananim
When I read the above post from SA, I agree with you - the Kiwi CAA surveillance was awfully inadequate, especially with an operator that had lost two a/c (DC8 & L188) in crew training operations. CAA, ANZ, weather and the crew all contributed to this disaster and I suspect that the crew provided the smallest portion.
Letdown procedures were given verbally at the briefing.
Approval was given for 1500ft
Approval was given to use own initiative on letdown procedure and approach
Testimony at trial was given by briefing pilot John Wilson and AirNZ Operations Manager Doug Keating
-------------------------------------------------------------------
Rananim
When I read the above post from SA, I agree with you - the Kiwi CAA surveillance was awfully inadequate, especially with an operator that had lost two a/c (DC8 & L188) in crew training operations. CAA, ANZ, weather and the crew all contributed to this disaster and I suspect that the crew provided the smallest portion.
Join Date: May 2000
Location: Here. Over here.
Posts: 189
Likes: 0
Received 0 Likes
on
0 Posts
Glueball:
I think that we can agree now that Capt Collins did not "hesitate in his response" or "only *call* for max power".
We are now debating the difference between using go-around power or firewalling the thrust levers. Here I agree with you that the correct response would be to firewall the thust levers. Did Collins do this? We do not know. Probably not, based on his call to check "go around power". The only facts we do know are that just before impact the power was increasing and the aircraft was rotating through 10 degrees nose up. However, the point is that whatever thrust setting was used, it would not have made any difference to the end result.
From Chippindale's report:
And in more detail:
One thing discovered in the simulator was:
However, I recall reading somewhere that this required a radical turn away from the slope, and pre-supposed you could see the ground to work out which way to turn.
Collins had six seconds and no chance to get it right. The rest of us have had 25 years to think about how we could have done it better.
I have no problem admitting that if I had been in that situation I would have hit the ground too, and it annoys me when all the armchair experts come up with invalid reasons why it would never have happened to them.
I think that we can agree now that Capt Collins did not "hesitate in his response" or "only *call* for max power".
We are now debating the difference between using go-around power or firewalling the thrust levers. Here I agree with you that the correct response would be to firewall the thust levers. Did Collins do this? We do not know. Probably not, based on his call to check "go around power". The only facts we do know are that just before impact the power was increasing and the aircraft was rotating through 10 degrees nose up. However, the point is that whatever thrust setting was used, it would not have made any difference to the end result.
From Chippindale's report:
The crew responded expeditiously in the circumstances to the GPWS warning. Simulator trials
proved conclusively that with an unexpected warning such as this, it would have been impossible
to avoid the accident with a normal pilot’s response allowing reasonable identification and reaction
times to the GPWS warning.
proved conclusively that with an unexpected warning such as this, it would have been impossible
to avoid the accident with a normal pilot’s response allowing reasonable identification and reaction
times to the GPWS warning.
1.16.3 The performance of the GPWS was evaluated and it was assessed that the warning was in
accordance with the expected performance in the “terrain closure” and “flight below 500 feet
without flaps and undercarriage extended” modes of the equipment (modes 2A and 4
respectively). The profile of the terrain prior to the impact was reconstructed in Air New
Zealand’s DC 10 simulator and the performance of the aircraft was evaluated to determine if the
collision could have been avoided in response to the warning and that the warning was in fact
given at the maximum time before impact that could be expected.
1.16.4 The flights in the simulator indicated that experienced pilots would not have avoided a collision
and that the warning given was in accordance with the design specifications of the GPWS.
accordance with the expected performance in the “terrain closure” and “flight below 500 feet
without flaps and undercarriage extended” modes of the equipment (modes 2A and 4
respectively). The profile of the terrain prior to the impact was reconstructed in Air New
Zealand’s DC 10 simulator and the performance of the aircraft was evaluated to determine if the
collision could have been avoided in response to the warning and that the warning was in fact
given at the maximum time before impact that could be expected.
1.16.4 The flights in the simulator indicated that experienced pilots would not have avoided a collision
and that the warning given was in accordance with the design specifications of the GPWS.
With sufficient rehearsal it was possible to fly the aircraft away from the approaching slope
when an extreme manoeuvre was initiated in response to the onset of the GPWS warning.
when an extreme manoeuvre was initiated in response to the onset of the GPWS warning.
Collins had six seconds and no chance to get it right. The rest of us have had 25 years to think about how we could have done it better.
I have no problem admitting that if I had been in that situation I would have hit the ground too, and it annoys me when all the armchair experts come up with invalid reasons why it would never have happened to them.
Join Date: Mar 2000
Location: Arizona USA
Posts: 8,571
Likes: 0
Received 0 Likes
on
0 Posts
Well, I gotta admit, it sure was a set of very tragic circumstances.
One thing that I cannot quite fathom is...what the hell could they have expected to see at fifteen hunderd feet that they could not have just as well seen at 15,000 or ten thousand, or even five thousand?
Can being that low (fifteen hunderd) have really been all that important?
I sure wouldn't think so.
OTOH, I haven't done this rather low sightseeing...except at the outer marker, on an ILS approach.
One thing that I cannot quite fathom is...what the hell could they have expected to see at fifteen hunderd feet that they could not have just as well seen at 15,000 or ten thousand, or even five thousand?
Can being that low (fifteen hunderd) have really been all that important?
I sure wouldn't think so.
OTOH, I haven't done this rather low sightseeing...except at the outer marker, on an ILS approach.
JW411: Twenty years ago I was invited to fly an ex-ANZ (KSSU)DC-10-30. (For those of you DC-10 pilots out there who are unfamiliar with such animals, even the flightdeck door was hinged on the wrong side).
From my point of view the most significant difference to all other DC-10s that I had ever flown was the Collins AINS 70 navigation system. This basically consisted of three Collins INS platforms which fed two navigation computers located behind the captain. They, in turn, fed the automatics and were controlled by two large alpha-numeric keyboards.
The nav computers could have all of the waypoints for the proposed flight loaded by a rather large (by modern standards) "Jep Tape" (although we distrusted the system so much that we did it the old way).
We were so bothered about this bit of kit that we spent a day with Collins in Los Angeles talking about it. The guy who gave us the course told us that after the Mt Erebus disaster the US Navy picked up everything they could find at the accident site. All the bits were put in plastic bags and flown back to Long Beach.
They were then laid out on benches in a hangar for the manufacturers to identify. He, to his total astonishment, found the carbon discs from inside the nav computers lying there on a bench and they were able to replay the last part of the flight.
I believe that it was as a result of this that the "missing waypoint" that took them through the mountain was identified.
JW411: As I stated in my earlier post (re the KSSU DC-10's having the the most advanced Area Nav System of their time) as I heard it, there were two 8K core modules that were recovered from the crash site - these were non-volatile memory - and thus the flightpath (with the W/P changes) were able to be reconsructed.
Absolutely fascinating.
Cheers!
From my point of view the most significant difference to all other DC-10s that I had ever flown was the Collins AINS 70 navigation system. This basically consisted of three Collins INS platforms which fed two navigation computers located behind the captain. They, in turn, fed the automatics and were controlled by two large alpha-numeric keyboards.
The nav computers could have all of the waypoints for the proposed flight loaded by a rather large (by modern standards) "Jep Tape" (although we distrusted the system so much that we did it the old way).
We were so bothered about this bit of kit that we spent a day with Collins in Los Angeles talking about it. The guy who gave us the course told us that after the Mt Erebus disaster the US Navy picked up everything they could find at the accident site. All the bits were put in plastic bags and flown back to Long Beach.
They were then laid out on benches in a hangar for the manufacturers to identify. He, to his total astonishment, found the carbon discs from inside the nav computers lying there on a bench and they were able to replay the last part of the flight.
I believe that it was as a result of this that the "missing waypoint" that took them through the mountain was identified.
JW411: As I stated in my earlier post (re the KSSU DC-10's having the the most advanced Area Nav System of their time) as I heard it, there were two 8K core modules that were recovered from the crash site - these were non-volatile memory - and thus the flightpath (with the W/P changes) were able to be reconsructed.
Absolutely fascinating.
Cheers!
Last edited by flash8; 14th Dec 2004 at 11:42.
Guest
Posts: n/a
I recall reading an article some time after the Erebus accident where Boeing and other manufactures changed their GPWS pull-up procedures from a recommended 15 degree pull up to a 20 degree pull up manoeuvre.
The article stated that Sundstrand (it was their GPWS system?) did some research on CFIT accidents and came out with figures that indicated the "average" CFIT hit the hill at 300 ft from the top at 230 knots and the "average hill" which is hit had a 18 degree slope. My figures may be a bit inaccurate but it was close to that.
15 degrees was the usual body angle for go-around in most early jets and by default was the body angle which pilots aimed at during a GPWS pull-up. By going aggressively (meaning real fast) to a minimum of 20 degrees - as well as firewalling the throttles, you played the law of averages and hoped you won on that occasion.
A vital simulator exercise to be practiced until competent, in my opinion.
The article stated that Sundstrand (it was their GPWS system?) did some research on CFIT accidents and came out with figures that indicated the "average" CFIT hit the hill at 300 ft from the top at 230 knots and the "average hill" which is hit had a 18 degree slope. My figures may be a bit inaccurate but it was close to that.
15 degrees was the usual body angle for go-around in most early jets and by default was the body angle which pilots aimed at during a GPWS pull-up. By going aggressively (meaning real fast) to a minimum of 20 degrees - as well as firewalling the throttles, you played the law of averages and hoped you won on that occasion.
A vital simulator exercise to be practiced until competent, in my opinion.
Join Date: Sep 1999
Posts: 541
Likes: 0
Received 0 Likes
on
0 Posts
Would agree that GPWS escape maneuvers should be practiced more and particularly following descent with speedbrakes extended as already mentioned.The cali crash demonstrates the need for this very clearly.Response must be instinctive,immediate and aggressive.
Join Date: May 2000
Location: Here. Over here.
Posts: 189
Likes: 0
Received 0 Likes
on
0 Posts
I think he has a good point though.
Just how close to the snow do you have to get to see that it is.... errr.....white and featureless ??
Bugger....never thought I would ever be supporting 411A
Just how close to the snow do you have to get to see that it is.... errr.....white and featureless ??
Bugger....never thought I would ever be supporting 411A
There's been the odd winter whiteout incident where a skiplane in flight begins losing airspeed in spite of full power and comes to an eventual halt in ascending deep powder
Don't try this in a jet
Don't try this in a jet
Last edited by RatherBeFlying; 15th Dec 2004 at 01:47.
Join Date: Sep 2000
Location: New Zealand
Posts: 15
Likes: 0
Received 0 Likes
on
0 Posts
For anyone interested, Otago University and the Hocken Archives in Dunedin have a large amount of Erebus crash related material, including books, video recordings, inquiry transcripts, legal reports, etc.
Join Date: Apr 2003
Location: Germany
Age: 76
Posts: 1,561
Likes: 0
Received 0 Likes
on
0 Posts
Training video refers...
I recall seeing a training video, one of those made using actors speaking lines from the crash CVR transcript, about this crash.
The main point made in the training video was that there were landscape features that looked exactly like what the crew expected to see. But due to poor visibility what they saw were features on a different scale in a different location, one that put them on a collision course with terrain. The crew was held to have reacted to the expected visual cues without checking very exactly their actual lat/long position. Presumably a look at a chart would have showed them what they were actually looking at rather than what they thought they were looking at.
I don't know if this is absolutely correct in the light of all the facts, of course. I saw the video about 4 years ago so that I may well have forgotten many of the points made. That said, it was a very interesting safety presentation about visual illusions and the need to cross-check critical nav data even when what is seen is what one expects to see. I assume most airline crews spend very little time at low level navigating by visual charts and landscape features.
The main point made in the training video was that there were landscape features that looked exactly like what the crew expected to see. But due to poor visibility what they saw were features on a different scale in a different location, one that put them on a collision course with terrain. The crew was held to have reacted to the expected visual cues without checking very exactly their actual lat/long position. Presumably a look at a chart would have showed them what they were actually looking at rather than what they thought they were looking at.
I don't know if this is absolutely correct in the light of all the facts, of course. I saw the video about 4 years ago so that I may well have forgotten many of the points made. That said, it was a very interesting safety presentation about visual illusions and the need to cross-check critical nav data even when what is seen is what one expects to see. I assume most airline crews spend very little time at low level navigating by visual charts and landscape features.
Moderator
Join Date: Oct 2004
Location: England
Posts: 2,519
Likes: 0
Received 0 Likes
on
0 Posts
Desert Dingo
If you had actually seen Mt Erebus you would realise that it is not just flat snow but, on a good day, is most spectactular active volcano. On a bad day it is solid cloud
The surrounding area is also magnificent with the low sun showing fantastic relief as you can see from the photo taken from a QANTAS flight when sightseeing resumed.
Spent two and a half years flying in Antarctica and it is not a place to make a mistake. - it bites back.
HWB
If you had actually seen Mt Erebus you would realise that it is not just flat snow but, on a good day, is most spectactular active volcano. On a bad day it is solid cloud
The surrounding area is also magnificent with the low sun showing fantastic relief as you can see from the photo taken from a QANTAS flight when sightseeing resumed.
Spent two and a half years flying in Antarctica and it is not a place to make a mistake. - it bites back.
HWB
Last edited by Halfwayback; 15th Dec 2004 at 16:28.
Keeping it simple`
Even though it may not be, but paring back the extraneous issues.
1. Pilots have preflight briefing of route
2. Route is changed. Pilots unaware.
3. On the day pilots think they are flying into an area where they know they can descend to xxx ft.
4. They look out the window. They are clear of cloud, in VMC.
5. There is cloud above that stretches to the horizon. It causes whiteout (NOT IMC), obscuring any surface definition that would otherwise alert them to their being in the wrong spot.
6. The accident.
Yes it was a CFIT.
They could not (through inadequate training perhaps) recognise the dangers and were lulled into a sense of security by being VMC and where they thought they were.
1. Pilots have preflight briefing of route
2. Route is changed. Pilots unaware.
3. On the day pilots think they are flying into an area where they know they can descend to xxx ft.
4. They look out the window. They are clear of cloud, in VMC.
5. There is cloud above that stretches to the horizon. It causes whiteout (NOT IMC), obscuring any surface definition that would otherwise alert them to their being in the wrong spot.
6. The accident.
Yes it was a CFIT.
They could not (through inadequate training perhaps) recognise the dangers and were lulled into a sense of security by being VMC and where they thought they were.
