ATSB Report on Angel44 asymmetric accident now published.

[Centaurus]

ATSB report just published on the accident to VH-AIZ, an Angel 44 twin engine'pusher' aircraft.

https://www.atsb.gov.au/publications/investigation_reports/2019/aair/ao-2019-072/

The aircraft crashed during simulated engine failure after lift-off at Mareeba on 14 December 2019. The report is concise, thorough and IMHO required reading by all Examiners and flying instructors. The instructor simulated an engine failure shortly (100 feet) after liftoff from a touch and go landing. Directional control was lost almost immediately.

The reasons why the loss of directional control occurred combined with other allied factors, are comprehensively covered. The report examined previous accidents of this nature of which three were over the period 2008 to 2017. It would have been more useful to have gone back a few more years to 2003. That year a Duchess VH-JWX crashed at Camden following a touch and go landing. In that accident, the highly experienced instructor cut the mixture control on the right hand engine at 50 feet. The student (a current airline pilot) was able to maintain directional control but the aircraft failed to climb on one engine because of the high drag of the windmilling propellor on the simulated failed engine. One wing clipped a tree and the Duchess belly landed under control. It then caught fire. The occupants were uninjured by the impact but received severe burns while exiting he aircraft. The instructor subsequently died.

The circumstances of both accidents (simulated engine failure following a touch and go landing) were similar. Thus the ATSB report to the Angel 44 event would have done well to include reference to the Camden accident. Both accidents prove that simulated engine failures at extremely low altitudes in light multi-engine aircraft are simply not worth the risks involved.

See ATSB report on the Camden Duchess accident.
https://www.atsb.gov.au/publications/investigation_reports/2003/aair/aair200300224/

I

[machtuk]

Tragically all unesesary -(

[Checklist Charlie]

It's not just simulated engine failures at extremely low altitudes in light multi-engine aircraft

One just has to remember Tamair at Tamworth, a 707 off East Sale and a Viscount at Mangalore.

CC

[KRviator]

VH-ANB at Darwin (https://www.atsb.gov.au/media/3546615/ao-2010-019.pdf) as well, where they went to flight idle and simulated a simultaneous failure of the engine and autofeather system. Which itself was almost a carbon copy of the B1900 incident at Williamtown (https://www.atsb.gov.au/media/24342/aair200000492_001.pdf) only a couple of years earlier...

EDIT: I hadn't seen the ATSB animation from the -ANB crash. It's available for download HERE (https://www.atsb.gov.au/media/5772638/ao-2010-019-embraer-120-loss-of-control-annimation.mp4). Doesn't take long for things to go awry...

[aroa]

As a close witness to this accident I had made an unrecorded statement to a young ATSB visitor . I heard it coming, looked to the strip 100 mts away to see it and watched it rotate and climb poorly only about 200mtrs from runways end. At approximately 100 ft it banked right with the gear down, and it was all down hill from there....over the corn field. When it banked right again, my initial thought was it was to avoiding the upcomming power line and /or buildings ahead....and sh*t that is a LOW level circuit Distance between those difficult to asssess from my perspective,
But sadly it wasnt a turn at all it was the right wing letting go airflow -wise and it plunged into the corn, after about 1/4 rotation. I wasnt aware of a lot of noise after the initial turn off the strip ...it was like watching from behind a glider on finals ...coming into land ...wrongly ! Its a bugger if you dont get it right.
RIP them 2 gents.

[Hamley]

Your examiner/instructor/trainer should include in their brief ‘engine failures will not be simulated below XX AGL’

If they don’t brief it, ask them about it before takeoff.

If they don’t put a number on it that sounds reasonable, you are getting closer to being one of these stories.

Thank you to all my mentors who have imposed limits during training.

[Centaurus]

Some months after the Camden Duchess accident I talked to the survivor who had returned to Canada after an extensive stay in hospital recovering from burns. He was an Air Canada pilot. He was full of praise for the help of the AFAP.who quickly organised administrative help for him while he was in hospital.

Before the flight, which was an instrument rating test in the Duchess, the instructor gave him a comprehensive briefing. Among other things, the plan was for an instrument approach under the hood at Camden followed by a circling approach This was to be followed by a touch and go and simulated engine failure after takeoff and a return to Bankstown

Because night was approaching the student made it clear he would not accept an engine failure after takeoff due to the rules saying no engine failures at night except in the cruise. The instructor accepted that but nevertheless cut a mixture just as the pilot had selected gear up shortly after lift off. The student was caught by surprise and urgently asked the instructor to reintroduce power to the 'failed' engine because the aircraft wasn't climbing due to the drag from windmilling prop. The instructor was attempting to re-introduce power when one wingtip clipped a tree. The aircraft hit and slid up sloping ground beyond the end of the runway before going through a fence. It had almost stopped when one wing hit a metal girder in long grass and the fuel tank split and caught fire. Both pilots escaped through their respective doors but the instructor received severe burn injuries.

The student was able to reach the instructor and carry him away from the burning aircraft. Fortuitously, a police van on a nearby road saw the accident. The student still carrying the instructor called for the police to get them to the nearest hospital. There was a delay while the police radiod for instructions and finally the two pilots were taken to hospital for treatment.

The simulated engine failure after liftoff took place after the Duchess was already well down the runway on the touch and go. It was therefore closer to rising terrain and a line of trees.than for a normal takeoff. It could be argued that one lesson from both the Mareeba and Camden accidents is that instructors should think twice before failing an engine following a touch and go since any obstacles to be cleared in the takeoff splay could be a lot closer than normal. A steeper climb gradient may be needed to clear those obstacles. This may prove difficult to achieve - especially with drag from a windmilling propeller on the 'failed'engine.

[KRviator]

Your examiner/instructor/trainer should include in their brief ‘engine failures will not be simulated below XX AGL’

If they don’t brief it, ask them about it before takeoff.

If they don’t put a number on it that sounds reasonable, you are getting closer to being one of these stories.In the Camden Duchess case, the ATO agreed not to pull an engine below 500AGL and then went and did it anyway almost immediately after rotation! The phrase "with friends like these..." comes to mind.

Most everyone knows your average light twin is not going to be able to accelerate to Vyse with gear & flaps out and the 'failed' engine windmilling. Hell, even some turbine twins can't do it, which is why they have autofeather (and incidentally, why going to Flight Idle instead of Zero Thrust simulates dual independent failures on autofeather-equipped engines...)

Has anyone had an instructor, or briefed to an instructor, words to the effect of "in the event of an engine failure before clean configuration and 200/500AGL I'll close the good throttle and land straight ahead"? If so, what was their reaction?

[B2N2]

From the report:

The instructor had limited experience in multi-engine aeroplanes with retractable landing gear and only one short flight in the Angel 44 aircraft several years earlier.

I am sorry to say the instructor had no business being there and should have declined the flight.
How can you conduct a proficiency check/ flight review in an aircraft that you are not familiar with.
:suspect:

[havick]

From the report:



I am sorry to say the instructor had no business being there and should have declined the flight.
How can you conduct a proficiency check/ flight review in an aircraft that you are not familiar with.
:suspect:

Sadly CASA examiners themselves are the worse examples of this behavior.

[Slippery_Pete]

I have complete confidence that until the day I die, there will continue to be more GA deaths in Australia from engine failure training than from actual engine failures.

The lack of progress in this area is beyond belief.

[Centaurus]

I have complete confidence that until the day I die, there will continue to be more GA deaths in Australia from engine failure training than from actual engine failures.

Astute observation. One problem is the standard of instruction. A second problem is instructor personality. Some are over-confident cowboys and these sort of people rarely change for the better..buyer beware

[CptSauce]

Has anyone had an instructor, or briefed to an instructor, words to the effect of "in the event of an engine failure before clean configuration and 200/500AGL I'll close the good throttle and land straight ahead"? If so, what was their reaction?

During my ME training it was SOP to accelerate to VYse then retract gear. Any engine failure before that would be rectified by closing both throttles and landing ahead. This was practiced over and over in the sim until it became second nature. My instructor was very much in favour of it. In a training environment on a light twin I personally think it's the way to go.

As has been said above it's notoriously difficult in many twin aircraft to climb and accelerate in TO config with a windmilling prop. Throw in poor pilot technique with an instructor who takes too long to take control and you've got a real problem. An engine failure that early on has a risk of loss of directional control, which I dare say is far more of a fatal issue than a twin glider with its gear down.

I suspect different skill levels can handle different procedures, but as a training ME pilot I felt most comfortable with this approach.

[runway16]

In the Angel 44 accident the name of Camden, NSW, continues to rise.

Over the years Camden has had a continuing list of twin engine aircraft accidents. The usual theme is an engine failure, practise or for real at low level just after take-off. The aircraft then comes down off the northern end of the runway on the upslope. Some fatal, some serious injuries, some walk away. The aircraft are usually destroyed.
Off the cuff I can recall Hudson, Dove, Dutchess, Twin Comanche. There are others.

Camden would appear to have had more twin accidents than any other airport that I can recall. Please correct me if I am wrong.

Please add a few more and dates. These accidents go all the way back to the end of WWII. Finger trouble seems to be the main cause.

[Pinky the pilot]

I have complete confidence that until the day I die, there will continue to be more GA deaths in Australia from engine failure training than from actual engine failures.


Concur. Some of the stories I have heard over the years still send a shivver up my (repaired) spine!:ooh:

I did most of my CPL and all initial MEIFR training in a Seneca 1, and anyone who has flown this type of aircraft will know about the handling characteristics seemingly peculiar to that particular aircraft.:uhoh::eek::yuk:

I possibly had more simulated EFATO's in that a/c than some people have had (insert your favourite whatever) but one thing that my Instructor, the late and much missed Tony Kingham never ever did was do a simulated EFATO below at least 400'.:=

He made it clear to me on numerous occasions that any failure below that altitude would be genuine!:eek:

[Capt Fathom]

I don’t think there was anything peculiar to the Seneca 1 that doesn’t apply to most light GA piston twins! The remaining engine merely took you to the site of the crash.
It was up to you whether that crash was under control or not!

[Pinky the pilot]

I don’t think there was anything peculiar to the Seneca 1 that doesn’t apply to most light GA piston twins!

A reasonable observation Capt Fathom, but please name me another light piston twin that had such heavy elevators.:ooh:

Also, AFAIK most other light piston twins had a better assymetric ROC than the Seneca 1. (190fpm @ 91kt at ISA from memory. Can tell a story about a simulated EFATO at Waikerie on a 40C day:E:D)

The remaining engine merely took you to the site of the crash.
It was up to you whether that crash was under control or not!

Yup!:ok:

[Capt Fathom]

name me another light piston twin that had such heavy elevators


The Seneca 2,3,4 and 5? Just kidding.

The heavy elevator was only evident during landing IIRC. The fix was not to use full flap. That gave a much better elevator response.

[Centaurus]

Please add a few more and dates. These accidents go all the way back to the end of WWII. Finger trouble seems to be the main cause.

1 January 1950. Lockheed Hudson VH-SMK. Herald Flying Services carrying newspapers as freight. Night takeoff on Camden 06. One witness who saw the landing lights during initial climb then saw the lights of the aircraft descending steeply. Aircraft reached 150-200 ft then stalled and crashed and caught fire on impact. Both pilots killed. . Investigation reported starboard engine at low power on impact while port engine at high power. Speculation that engine failure of starboad engine may have occurred but no solid proof. Carby icing conditions existed
runway16 is offline Report Post (https://www.pprune.org/report.php?p=10909967)

[Centaurus]

I did most of my CPL and all initial MEIFR training in a Seneca 1, and anyone who has flown this type of aircraft will know about the handling characteristics seemingly peculiar to that particular aircraft

With two pilots and no payload the C of G was well forward meaning for landing it was significantly nose heavy. And I mean real heavy. With no payload the fix was to have sufficient ballast down the back to move the C of G back. If my memory serves me correctly the handling characteristics were normal

[Slippery_Pete]

As has been said above it's notoriously difficult in many twin aircraft to climb and accelerate in TO config with a windmilling prop.

It’s not notoriously difficult - it’s impossible. I’m yet to fly a light twin (and I’ve instructed on many) that will climb on one engine with a propeller windmilling - at any altitude or temperature or weight.

Airline category turboprops are the same. At max weight they go up at 6-700fpm OEI with autofeather. Fail the auto feather, and they descend at a minimum of 300fpm even when empty.

As sure as the night follows day,
more Aussie pilot’s lives will be lost in the future in these bull**** training accidents.

[Checkboard]

Multi-engine aircraft are divided by FAR 23 (http://www.access.gpo.gov/nara/cfr/cfrhtml_00/Title_14/14cfr23_00.html) into two weight classes - above and below 6000lbs, and those below 6000lbs are divided into two classes depending on Vso (stall speed in the landing configuration) above and below 61 knots CAS.

Only those twins that weigh more than 6000lbs or have a Vso higher than 61 knots need to demonstrate any single-engine climb performance at all for certificaion, and the requirements are pretty meager. Basically the regulation says that these aircraft must demonstrate a single engine capability at 5000' (ISA) with the inop engine feathered and in a clean configaration.

The only requirement for an aircraft less than 6000lbs, and with a Vso less than 61kts (like the Aztec) is that its climb performance (positive or negative) be determined.

There is nothing in the FAR which says a light twin aircraft must fly while in the take-off configuration with one engine inop - and so attempting to train for that is a nonsence applied by instructors who have a half-arsed understanding of airline operations, which are on aircraft in a totally different category.

[aroa]

Further to the downed Angel. Two things stick out...to me anyway..
1. the first circuit (which I didnt see) was said to be airborne after 300 mtrs Angel being STOl, there must have been good grunt available then. Which didnt last because rough running was heard on downwind and approach. After run ups > they tried again.
2..On the secong take -off roll, why nobody chopped the power and jumped on the brakes when there was pracitaly none left...Que? 3. Did the rough running engine actually fail.?....adding to the cut back from the "failed" engine thus giving leading to the impression of a silent glder final.
Perhaps we'll never know. The aircraft had been sitting around for a couple of years, old fuel, mud wasps up the tank vents ...no air in, no fuel out
BIQ the Aztec a year or so before ,was likewise left for years. That got parted out in the bush shortly afte a "hairy goat" departure which I also witnessed.

[john_tullamarine]

Some thoughts, if I may be indulged a tad.

The instructor had limited experience in multi-engine aeroplanes with retractable landing gear and only one short flight in the Angel 44 aircraft several years earlier.

However, that doesn't always apply. The Camden Duchess mishap instructor was at the other end of the experience spectrum - military, GA, corporate, airline experience. Both Centaurus and I knew him well and I had done several GA renewals with him in the past. Experience doesn't count for all that much if/when you stick your neck out too far once too often and put yourself in an unwinnable situation ....

I have complete confidence that until the day I die, there will continue to be more GA deaths in Australia from engine failure training than from actual engine failures.

A predictable consequence of a lack of technical knowledge. One of my increasing concerns, over the past several decades, has been the active reduction in operator/regulator interest/emphasis on crew technical knowledge. There is no need for the line pilot to become a design or flight test engineer but a very healthy level of technical knowledge usually helps us to keep our noses clean whilst we enjoy our careers.

Your examiner/instructor/trainer should include in their brief ‘engine failures will not be simulated below XX AGL’ If they don’t brief it, ask them about it before takeoff.

Or, even, tell them rather than ask. After a few unnecessary frights as a sprog, it became a standard briefing point on endorsements/renewals (unless I knew the instructor/testing officer well enough to know that he would not do so) that any failure below (my choice of a nominated height) would result in my closing both throttles and landing more or less ahead - including my initial Class One issue with a well-known DCA examiner many years ago. Never had any problems as none ever saw it necessary to test my resolve on the point.

Another anecdote comes to mind: A close, and very experienced, colleague, post 89, decided that he should renew a GA rating. This necessitated his undertaking an endorsement on whichever lightie was to be the requisite aircraft. The endorsing instructor made a point of having to have a look at Vmca (why ? for routine civil operations, there is absolutely no reason to play near Vmca except, perhaps, for one's initial multi endorsement and, then, only for a brief exposure to yaw handling problems for a degree of familiarisation). Anyway, my colleague, knowing a bit about such things, saw fit not to play as decreed and applied a part rudder bootful of foot to limit the speed excursion to which the aircraft would go prior to yawing uncontrollably - apparently the instructor was quite perplexed - "but it always goes slower than that". However, foolish pride was satisfied and they continued with the rest of the endorsement program.

Overall, the underlying problem is a dreadful lack of pilot technical knowledge and this is not helped by the "blind leading the blind" syndrome we so often see both with theory and flight training of pilots. Hopefully, observations by greybeards, sprinked throughout PPRuNe, will assist in education of those coming on, however limited the success may be.

[Ascend Charlie]

Sadly CASA examiners themselves are the worse examples of this behavior.

I was to do a PVT IFR instrument "review" in the Boss's chopper, which was the only one of its model in Oz - plenty of similar types, but none of this model. The trick was to find an instructor able to do the renewal. Sadly, the experienced instructor who flew my previous review in this aircraft was the one mentioned above in Camden, and was no longer available. There was another civil ATO with great experience on multis and instrument flying, but not endorsed on type. We had known each other for many years with some mutual respect, and being a "review" it was not a test as such, but to fly the sequences to the required standard, and refresh any that were a bit lacking - there is no "FAIL" in a review. You do it until all is good. I was happy to have him do the flight, as he would not be required to operate the systems. But CA$A insisted that one of their FOIs do the job.

They named one lad in Darwin. I would have to pay for his time from when he left the Darwin office, fly him to Sydney, pick him up, do the test, pay for his overnight, and send him back to Darwin. Plus pay a Test Fee. This lad had a bare endorsement on type and hadn't flown one for 5 years, and of course was not familiar with this model, which had FADEC .

CA$A also suggested another one, in Melbourne, and who I had known from RAAF days. Same situation, pay pay pay and bare endorsement on type and never flown the model. But CA$A reckoned that a type-inexperienced uncurrent desk jockey was safer than a current experienced but non-type-rated instructor. So we used the Melbourne man.

The pre-flight briefing included me telling him :"You will NOT be touching the controls. If you desire to simulate an engine failure, you will say so, and I will work the throttles."

After the review was over, we had to fly back from Richmond to Sydney, via Long Reef and the chopper lanes down the beaches. He asked gingerly "May I have a fly?" so I let him have a go. The grin on his face lit up the cockpit, and probably doubled his time on type.

[havick]

I was to do a PVT IFR instrument "review" in the Boss's chopper, which was the only one of its model in Oz - plenty of similar types, but none of this model. The trick was to find an instructor able to do the renewal. Sadly, the experienced instructor who flew my previous review in this aircraft was the one mentioned above in Camden, and was no longer available. There was another civil ATO with great experience on multis and instrument flying, but not endorsed on type. We had known each other for many years with some mutual respect, and being a "review" it was not a test as such, but to fly the sequences to the required standard, and refresh any that were a bit lacking - there is no "FAIL" in a review. You do it until all is good. I was happy to have him do the flight, as he would not be required to operate the systems. But CA$A insisted that one of their FOIs do the job.

They named one lad in Darwin. I would have to pay for his time from when he left the Darwin office, fly him to Sydney, pick him up, do the test, pay for his overnight, and send him back to Darwin. Plus pay a Test Fee. This lad had a bare endorsement on type and hadn't flown one for 5 years, and of course was not familiar with this model, which had FADEC .

CA$A also suggested another one, in Melbourne, and who I had known from RAAF days. Same situation, pay pay pay and bare endorsement on type and never flown the model. But CA$A reckoned that a type-inexperienced uncurrent desk jockey was safer than a current experienced but non-type-rated instructor. So we used the Melbourne man.

The pre-flight briefing included me telling him :"You will NOT be touching the controls. If you desire to simulate an engine failure, you will say so, and I will work the throttles."

After the review was over, we had to fly back from Richmond to Sydney, via Long Reef and the chopper lanes down the beaches. He asked gingerly "May I have a fly?" so I let him have a go. The grin on his face lit up the cockpit, and probably doubled his time on type.

Sounds like a broken record, particularly in the helicopter world.

[PA39]

Sounds like a broken record, particularly in the helicopter world.
i have around 14k hrs in multi engine training. My contribution is that VYSE or blue line was determined on a new aircraft fliown by a test pilot under ideal conditions. The 50 yr old hacks that most fly are med to high af time, certainly ageing in engine time, flakey paintwork, a few dings on the leading edges and flown by non test pilots. BEWARE.

[Capt Fathom]

The 50 yr old hacks that most fly are med to high af time, certainly ageing in engine time, flakey paintwork, a few dings on the leading edges and flown by non test pilots

And that was the point made by the Botany Bay DC3 pilot when his aircraft failed to climb away on one engine.