Mt Erebus accident.

[DozyWannabe]

Originally Posted by chris lz

But in a sense, isn't that precisely why INS is not as reliable as ground aides? They can be inadvertently misprogrammed. The crew themselves can mistakenly punch in a wrong coordinate. If you've never been to destination before, wouldn't it also be prudent to check the programmed coordinates you've just entered and make sure they match the actual coordinates for the McMurdo Sound route?

This is where things get murky, thanks to the actions of certain contemporary ANZ management people and the US radar station staff. If Collins was true to form he would have checked and re-checked the co-ordinates on his ops sheet against what was in the AINS system before takeoff and found no discrepancy. The rest of the story would be found in his little A6 ring binder. This was recovered from the site by police and sealed in an evidence bag by Inspector Gilpin, noting that it contained "numerous pages of legible technical writing and figures that indicated they related to the flying of aircraft. We recognised that this could be of importance to any investigation into the crash, and I sealed and secured it in a bag before it was returned to McMurdo". The NZCAA officials were not there yet, so it wound up in the hands of Chief Pilot Gemmell, and was never seen again in its intact state publicly, instead turning up at the 1981 inquiry with all its pages removed.

True to his meticulous nature, Collins performed all the cross-checking at home the previous night, which can only lead to the conclusion that contained in that book was proof that he was certain they were going down McMurdo Sound and not over Erebus. This would have rendered ANZ's case at the inquiry null and void, because it relied entirely on the idea that the nav track had *always* run over Erebus, and it was the McMurdo track that was an aberration. Hewitt's navigation change was input to the computer at approximately 1:40AM the morning of the flight, and nobody told Collins or any of his crew. At his preflight briefing he had every right to believe that the co-ordinates he had fastidiously plotted before retiring the night before and the ones he now held in his hand, and would later cross-check as he input the co-ordinates into AIMS, were one and the same.

I'm aware the Lewis Bay track and the expected route down McMurdo Sound would have appeared similar. But before the letdown commenced, could the crew really be 100% certain of their position by visual means alone?

Up until the crash flight, it was all they had. The southernmost waypoint in all previous INS-navigated Antarctic flights was to a point just west of the Dailey Islands and had no navaid (because it was, as it turns out, a typo rather than an actual fix), and the NDB at McMurdo, which was the secondary navaid fix, had been withdrawn in 1978. And forget "similar" - unless you were intimately familiar with the area, in certain conditions they would have appeared nigh-on identical. Prospector likes to talk about Beaufort Island being on the wrong side, but looking at the let-down tracks, it's possible they would have emerged from cloud south of the island - photos exist of the island taken from within the passenger cabin, but there is no proof that it would have been visible from the flight deck. In any case they were taking a visual fix from Cape Tennyson, thinking it was Cape Royds, and Cape Bird, thinking it was Cape Bernacchi.

Hewitt selected the new navaid fix in 1978, which would be the TACAN south of Ross Island, but because of the typo, none of the crews ever knew that was the intent - they simply entered the co-ordinates and assumed that the new route was designed to follow the military track. To my mind it looks like Hewitt simply entered the co-ordinates into the computer from his data sheet - did not re-check them against a map as he was entering them, and forgot about it entirely until Captain Simpson reported the 26 mile discrepancy three weeks before the Collins flight. After making the correction, Hewitt made no attempt to contact the computer section to confirm that it had been updated, and likewise the computer section did not confirm with nav section that the work had been done. Again, to my mind very unprofessional behaviour.

The only major concern from the confirmed voices on the CVR seems to be that Mac Central was out of radio contact in the last few minutes of the flight - however the transponder was coding, indicating that the crew believed that they were being tracked on radar. The radar tapes that covered the last four minutes and change leading up to the crash had been erased, and Mahon got a very frosty reception from the US base when he visited and made it known he was aware of this.

I want to make it clear that even though I think Vette and Mahon's investigations were more in-depth and the conclusions more correct than those of Chippindale, this does not invalidate entirely the work that Chippindale did, and I don't think he was any less than scrupulously honest about what he believed. The problem as I see it is that the information Chippindale had to work from was tightly controlled by ANZ, and that most of what he had to work with was only the material that ANZ wanted him to see. The only major fault I see in Chippindale's methodology was inviting Captain Gemmell to participate in his re-write of the CVR transcript, and indeed, revising the content of the CVR transcript from what was agreed in Washington in the first place. I'm sure that had he known that Captains Gemmell and Crosby were engaged in the obtaining and destruction of evidence from Collins' and Cassin's files - even clearing material from their homes without the knowledge or permission of their families, he would not have been so sanguine about their involvement in his investigation.

[IGh]

Investigator-err, bias, mis-directed, evidence excluded

From Dozy, just above, commenting on the INVESTIGATION PROCESS:

"... I think Vette and Mahon's investigations were more in-depth and the conclusions more correct than those of Chippindale ... scrupulously honest ... The problem ... the information Chippindale had ... was tightly controlled by ANZ ... was only the material that ANZ wanted him to see. The only major fault I see in Chippindale's methodology was inviting Captain Gemmell to participate..."

Dozy -- NICE, there should be a forum for such critique of the mishap-"investigator", constructive additions to too-limited investigations. [Maybe I'll go back over the years and tack these investigator-errs into an organized compilation: investigating-the-investigator, or expanding the focus of an investigation.]

This mishap- INVESTIGATION(s) [NZCAA & Royal Commission's] turned out much better than some others of that same year 1979: The Royal Commission PROCESS seems much better than the UNCHALLENGED product from the USA-system, where the USA's NTSB is legally the SOLE judge of their past investigative-product (never any critical oversight from any Royal Commission).

The NTSB might consider a PETITION from the French BEA (eg, ATR72 at Roselawn), but will mostly ignore any domestic-petition that contradicts the USA's big manufacturer (eg, the NTSB's endorsement of the "Boeing Scenario" in AAR81-8 while the direct evidence and the surviving crew told the Board otherwise). Boeing created data, analysis, simulator-runs, insisting that the NTSB agree that the direct evidence ranked lower than Boeing's created "evidence". [That investigation described in AAR81-8 was the longest most expensive investigation in the then history of the NTSB, & the B727 was then the best selling product of the USA's biggest exporter.] http://www.pprune.org/safety-crm-qa-...-accident.html

Dozy -- thanks for those comments on the investigative-process: the USA's "independent" Safety Board concept, over the past forty years, seems mostly to produce an "investigative product" of lesser quality. Where as the public's option of the NZ- Royal Commission might force the investigator [NZCAA] toward a more complete, better quality product; & reduce subtle bias in evidence-rejection (eg, Chippendale's exclusion of Vette's HF-components).

[chris lz]

Wiggy:

Any error in the above processes, (e.g. in measuring the accelerations) will result in a mismatch between where the INS "thinks" you are and where you really are in the world.

In practical terms, in 1979 on a DC-10, is it reasonable then to suppose that this type of error was far more likely than a failure of a ground based navigation aide? I'm thinking the answer is "yes." But then I have heard from one pilot that AINS is "much more accurate" than what a typical ground based aide would predict for a plane's exact position. Maybe this is true also?

[DozyWannabe]

Originally Posted by chris lz

In practical terms, in 1979 on a DC-10, is it reasonable then to suppose that this type of error was far more likely than a failure of a ground based navigation aide? I'm thinking the answer is "yes." But then I have heard from one pilot that AINS is "much more accurate" than what a typical ground based aide would predict for a plane's exact position. Maybe this is true also?

The problem as this applies to the Erebus crash is that on each Antarctic flight using it the AINS had been completely accurate every single time. The reason ANZ tried so hard to muddy the waters was because their navigation section had mis-keyed the fix in the first place and left it that way for over a year, even to the extent that the charts presented to the pilots at briefing showed the "incorrect" route - this left them vulnerable to accusations of incompetence in the first instance. Correcting the route to the TACAN in the middle of the night without notifying the crew who had been briefed on the previous route and were due to leave in the morning could, in the hands of a lawyer, open them to accusations of gross negligence in trying to cover up their earlier incompetence - hence the all-out effort to destroy all evidence that showed the McMurdo Sound track had been the accepted route for a year.

Regarding ground navaids, the McMurdo NDB had been gone since 1978 - all there was in the area was the TACAN, which - while it had been the intended waypoint as far as nav section's seniors were concerned since the NDB was decommissioned - was never on the route supplied to the pilots between 1978 and 1979, nor was it's existence briefed. If crews had known about it, the only purpose it could serve would be to provide a fix after they had made the left turn south of Ross Island, because to use it as a direct fix during their south-westerly track would have taken them straight over, or indeed into the side of, Erebus - which at low level would have been a major safety hazard, and at high level would have contradicted the whole purpose of the flight, which was sightseeing.

@IGh - Ironically it was the global response to the Mahon report that made accident investigation what it is today. I disagree that the NTSB's independence harms the quality of it's output, however, and I know the "'Hoot' Gibson" incident, of which you are talking, fairly well. It wasn't just Boeing who advanced that particular theory though, the truth is that a lot of line pilots knew about the use of CB's to pop the slats in cruise - my suspicion is that the NTSB and the FAA had been looking for an opportunity to end the practice for a long time and for better or worse used that incident to do so. To discuss it further here would be taking the thread off-topic however.

[henry crun]

DozyWannabe: Referring to the Tacan you said "the only purpose it could serve would be to provide a fix after they had made the left turn south of Ross Island".

My understanding was that civil aircraft could interrogate the DME part of Tacan but not the direction indicating part.

If I am correct they could not have obtained a fix from it.

[IGh]

DozyW -- re'your's on that 1979 B727 case "... line pilots knew about the use of CB's to pop the slats in cruise ... the NTSB and the FAA ... looking for an opportunity to end the practice..."

From inside, there never were any B727pilots who had done that "pop the slats" at Crz FL390: On Oct'2 1980, in Seattle, aboard the company's ship E209, TBC did the purported C/B drill -- it didn't work; TBC tried various methods, using various Circuit Breakers, eventually finding other combinations of C/Bs. Similarly, at the operator TWA, no pilots had done it -- never heard of it (but those second-hand rumors live-on). The IIC [Dean Kampshore] admitted this later, during deposition.

Re' INS into McMurdo, I had flown into McMurdo the during the Oct'78 re-supply, landing on the "ice". Just as we had done, ANZ also forced their INS's into "Grid" mode while south of 60S. These ANZ-guys had been given some extra training on INS & its weaknesses [the err-rate curve after 5-hours shoots up exponentially].

Re' the NDB, MacJob's _AD_ [Vol2, pg73, mid], contrary to the ANZ briefing, the McMurdo NDB was actually still operative during Nov'79 mishap.

[DozyWannabe]

Originally Posted by IGh

Re' the NDB, MacJob's _AD_ [Vol2, pg73, mid], contrary to the ANZ briefing, the McMurdo NDB was actually still operative during Nov'79 mishap.

That's interesting. I wonder why they'd have told ANZ that it had been withdrawn - although it might be that it had been officially withdrawn while still technically operative. I read that the US base weren't particularly happy about the sightseeing flights.

[Tagron]

Chris lz

The accuracy of a triple-INS installation may be very good, but it is not perfect. Over the course of a six-hour flight it would not be unusual to see an error in the region of 1 - 3 nautical miles, and greater on exceptional occasions. The error is likely to be cumulative, i.e. the longer the flight the larger the error. It can only be identified by comparison with a known fix point, so poor fixing performance (as opposed to a system malfunction) may not be immediately apparent to the crew. For these reasons Flight Management Systems navigational computations used radio navaid fixing, principally DME/DME, to update the INS-generated position.

In the case of the Antarctic flight no such updating would have been available, so the crew were relying on raw INS data.. For the en route navigation this would have been entirely satisfactory - the ANZ experience of operating DC10s over the Pacific and North Atlantic would have told them that. As wiggy has suggested most if not all companies would prohibit descent below MSA based solely on INS position, for the above reasons plus the risk of inaccurate data entry. But that is for a descent in IMC.

The principal requirement for a VMC descent below MSA is that safe terrain clearance must be maintained by visual means throughout the descent. Provided that requirement is met then there can be no objection to the use of INS. Whether to fly the direct INS track or to use the INS data to assist situational awareness while following a different track would depend on the circumstances at the time. Note these are general comments about VMC descents as I do not know how the requirements were specified in the ANZ SOPs of the time.

The visual misidentification was a fatal error. That was an opportunity to verify INS accuracy beyond doubt. Had the visual position been correctly identified the crew would surely have realised there was something seriously amiss with their INS position and maintained altitude to await a corroborative fix. As it was, confirmation bias seemed to set in. The INS could not be so wrong. A 27 nm error in their INS performance was quite contrary to their experience

henry crun

I agree with you. To the best of my knowledge civilian aircraft can only use the DME function of TACAN, not the bearing information.

[chris lz]

Tagron:

The principal requirement for a VMC descent below MSA is that safe terrain clearance must be maintained by visual means throughout the descent.

But isn't it also a basic requirement than in VMC, a visual letdown below MSA first requires a positive visual confirmation of one's exact position? I was under the impression the crew was only able to form a visual confirmation (in their minds at least) after they had descended below the cloud layer and could see the terrain to the left of their track, most of which (including Erebus) was not visible at 16,000 ft. (My memory could be wrong here.)

[DozyWannabe]

Originally Posted by Tagron

The visual misidentification was a fatal error. That was an opportunity to verify INS accuracy beyond doubt. Had the visual position been correctly identified the crew would surely have realised there was something seriously amiss with their INS position and maintained altitude to await a corroborative fix. As it was, confirmation bias seemed to set in. The INS could not be so wrong. A 27 nm error in their INS performance was quite contrary to their experience.

Regarding visual misidentification - I highly recommend watching the complete "Impact Erebus" video if you have not already seen it, as it demonstrates the optical illusions involved in a very useful and understandable way. The phenomenon of whiteout as it was briefed to the ANZ pilots was actually the wrong kind of whiteout - the type caused by refractions from powdery snow low to the ground, as opposed to the type which trapped Captain Collins and his crew, which can affect almost any altitude below overcast cloud. It was an error, no question about it, but it was an error capable of affecting even pilots who know the area well - with the fog bank (well below them) obscuring the end of the bay and the foot of Erebus, the only visible difference between the cliffs of Lewis Bay and those of McMurdo Sound would be the way the cliffs in the latter fan out in the distance - at 1,500ft that would have been extremely difficult to spot.

Originally Posted by chris lz

But isn't it also a basic requirement than in VMC, a visual letdown below MSA first requires a positive visual confirmation of one's exact position? I was under the impression the crew was only able to form a visual confirmation (in their minds at least) after they had descended below the cloud layer and could see the terrain to the left of their track, most of which (including Erebus) was not visible at 16,000 ft. (My memory could be wrong here.)

The full overcast affected only the immediate area of Ross Island itself. North of Ross Island, the cloud was patchy and there were more than enough gaps to make a visual let-down viable (as you can see from the passenger photos). The problem was not that the crew could only form visual confirmation once under the cloud at the point of let-down, it was that once down, Lewis Bay and McMurdo Sound were almost indistinguishable from one another from the point-of-view of the flight deck in the conditions they encountered. Once they had reached the level Mac Central had invited them to descend to, the only clue that something was wrong was that they could no longer communicate with Mac Central via radio.

The question that bothers me is that if the radar let-down was completely above board from Mac Central's end, why did they erase the radar tapes?

[chris lz]

Originally Posted by DozyWannabe

The full overcast affected only the immediate area of Ross Island itself. North of Ross Island, the cloud was patchy and there were more than enough gaps to make a visual let-down viable. . .

But again, while in VMC, isn't confirmation of position required before one passes below MSA?

[henry crun]

After reading Mahon's book I got the strong impression he believed that because the DC10 was within the theoretical coverage of McMurdo radar it would have been displayed.

Did anyone else feel the same way ?

[DozyWannabe]

The reason they thought they were being tracked on radar was because the transponder was coding at the time.

@chris lz - I'll have to defer on that one, but Vette and Mahon both concluded that no rules were broken as far as the crew were concerned.

[Dream Land]

Leave the tours, to tour pilots!

[stepwilk]

To the best of my knowledge civilian aircraft can only use the DME function of TACAN, not the bearing information.

Aren't TACANs typically co-located with VORs?

[DozyWannabe]

Originally Posted by Dream Land

Leave the tours, to tour pilots!

Well, as I said in my earlier post, when ANZ started the Antarctic flights they did their homework - researched the limits of INS (as IGh says, enforcing GRID mode) - which is partially why the course was manually navigated until 1978, and included in the protocol two directives - that each flight should carry two Captains, a First Officer and two Flight Engineers, and that in order to command an Antarctic flight, a prior familiarisation trip was mandatory. Between 1977 and 1979 both those directives were dropped, in favour of a Captain, two F/Os and two F/Es, and losing the familiarisation requirement. Whether this was down to the flights becoming routine, cost-cutting or other factors will never be known. I have my suspicions however - at the time ANZ appeared to be a successful multi-million dollar flag carrier, but under the surface it was almost entirely reliant on government subsidy following the National acquisition. Qantas was about to start offering their own Antarctic flights, which would have eaten into ANZ's market share considerably.

Originally Posted by stepwilk

Aren't TACANs typically co-located with VORs?

Typically? If you say so. In the vicinity of the southernmost active volcano in the world? It certainly doesn't look like it at the time. Remember that the only flights down there prior to the mid-70s were military flights supplying McMurdo and Scott Base.

[Dream Land]

Well I think it's a very complicated accident, I don't want to criticize anyone based on my YouTube knowledge of the accident, but with my knowledge of scud running, Alaska and the Grand Canyon, I am well aware of what it involves.

Anytime a pilot drops through a hole and descends to 1500 foot above the ground, should know exactly what he's doing, doesn't matter whether it's a Cessna 206 or a DC-10. And seeing lights on your transponder blinking shouldn't lead anyone to assume they are within radar control, especially an experienced aviator. Tragic.

[DozyWannabe]

Originally Posted by Dream Land

Anytime a pilot drops through a hole and descends to 1500 foot above the ground, should know exactly what he's doing, doesn't matter whether it's a Cessna 206 or a DC-10.

That would appear to be the crux of the problem in that they thought they did, thanks to a combination of the INS programming not matching their briefing, and weather conditions that were little-understood at the time making where they were headed look to all intents and purposes identical to where they thought they were headed.

And seeing lights on your transponder blinking shouldn't lead anyone to assume they are within radar control, especially an experienced aviator. Tragic.

It was the blinking transponder in combination with the invitation to descend which gave them that impression. Again, if Mac Central didn't make a c*ck-up, why erase the tapes?

[chris lz]

Originally Posted by DozyWannabe

all intents and purposes identical to where they thought they were headed

That's what Vette claims in his book, but I am not necessarily convinced. They see land where they expect to see land, according to Vette, and that in his mind translates to a "positive fix."

Consider some interesting comments from the crew which I would guess you have already read:

Mulgrew to passengers 4 min before impact: "I still can't see very much at the moment. I'll keep you informed as soon as I can see something that gives me a clue to where we are [emphasis added]."

FE 3 min before impact "Where's Erebus in relation to us at the moment?"

Mulgrew 3 min before impact: "Yes- - - no, no, I really don't know." [(this is not a response to the FE's question quoted above)].

Mulgrew 1 min 40 sec before impact "I reckon Bird's through here and Ross Island there. Erebus should be here."

Mulgrew 50 sec before impact "Looks like the edge of Ross Island there."

These remarks no doubt are subject to a wide degree of interpretation, and I would agree with Vette none of them shows anything so blatant as that the crew feels lost or confused. But I can't help thinking they don't convey the sense of "positive fixes" either. Vette calls the view from the aircraft cockpit a "visual counterfeit" of the McMurdo Sound track. But how much so? I would think that any track where there is terrain to your left and to your right would have more than a minute posibility of being taken by a flight crew for what they think should be there. Vette describes walls of clouds rising above them on either side, and clouds and fog below them, obscuring much of the landscape. Given this, I'm wondering if "doesn't contradict what they expect to see" is really the equivalent of a positive identification.

[IGh]

Big skepticism of McMurdo Radar

Observation from several slots earlier, re' ANZ's perception:

"... they thought they were being tracked on radar was because the transponder was coding at the time...."

The crews flying into McMurdo (annual resupply) were TRAINED to mostly disregard any position information offered by the local Navy-guys staffing McMurdo_radar. Also, due to repeated arrival-CFIT mishaps [turbojets while in radar-contact] during the mid-1970's, the TERRAIN-threat was a BIG training-lesson for any crew flying with a North American operator. An old-fashioned Pro-Navigator was aboard each heavy-jet landing at McMurdo, charts-out, constant challenging of any position-fix [mostly disregarding any info' from McMurdo-radar].

I recall flying downwind at McMurdo, with the ICE Rwy insight, and the only visual "horizon" were those ICE-CLIFFs at some un- certain distance, with the higher terrain obscured by low cloud.