Go-around after engine failure in light twin
 

Sorry if this topic has been covered, but a recent accident in my area is bothering me, and I have some questions.

To summarize, the aircraft was a Piper PA-34-220T Seneca, with optional 3 blade props. It departed a North Texas airfield in route to Missouri. An hour out, the left engine failed, and the pilot requested a landing at an Oklahoma airfield, but did not declare an emergency. A witness noted the left prop not turning when the aircraft made what seemed to him to be a fast approach. The witness then noted that about halfway down the 5600 ft. runway at an altitude of 5-10 feet, the pilot applied power to the right engine for a go-around. This aircraft has a max landing roll of 1400ft.

The witness observed the pilot retract the gear and start a shallow climb to about 200agl, but he observed the flaps were still extended. Then the pilot banked and turned left to start a go-around circuit. The witness said that almost immediately after the pilot started the left turn, the aircraft rolled completely to the left, and nose dived right into the ground behind a line of trees, killing all 5 aboard including the pilot's entire family. All 3 props of the left engine were found in the fully feathered position by the accident investigators. The main landing gear was found fully down and locked, and the flaps were found to be fully retracted (???).

So here are my questions. Why did the pilot try a go around at this runway's midspan at 5 ft altitude, when the rollout distance of his aircraft was only 1400ft max? Why did the pilot not declare an emergency? Why did the pilot try a left hand circuit with a dead left engine, instead of a right hand circuit?

There seems to be several things wrong here, but what's bothering me the most is the low and slow left hand turn into a dead left engine. Won't this kill you in a light piston twin? Wouldn't you rather want to turn right instead? Also, what should you do, if you absolutely have to perform a go-around under these circumstances?

What do you guys think?

PS - I forget, the weather was clear afternoon, 10 miles vis, winds 3kts and variable. The temp was 6C, dew point -6C, altimeter 30.35, airfield altitude of 770ft.

(edited to add additional weather data)

I am struggling to think of a situation that would cause me to try a go around, if I were in the flare on a long runway after an engine failure.

Even in the case of suspected gear failure, with a dead donkey surely it's better to be on the deck with a bent aircraft than struggling to go around and clean up from 5 feet on a single engine.

single engine commital height
 

A light twin will not climb unless all the conditions are right , the aircraft has to have the prop on the dead engine feathered ,the gear UP, 5 degrees of bank into the live engine , flaps up , cooling flaps in the correct position and the WAT conditions also have to be met ,above all blue line speed must be maintained to climb or reduce the rate of descent to a mimimum.

After a go-around as been decided apon it will take time before the aircraft can be re-configered for the climb and during this time the aircraft will be going DOWN.

It is therefore important that a single engine commital height be decided apon taking into account all the above factors and the objects around the airfield that may have to be avoided.

A typical commital height for a low time pilot would be 600ft AGL ,the aircraft should be flown at blue line speed untill this height is reached , from that point on the aircraft MUST land should be slowed to Vref and the landing flap deployed.

It is quite simple once below the commital height the aircraft has NOT got the performance to go around and so must land even if this means going off the far end of the runway at a slow speed or putting the aircraft on a part of the airfield not normaly used for landing aircraft.

To go around below the commital height will mean that the aircraft will be unable to climb clear of some close in object and hit it or the pilot will pull the nose up to try to avoid the object , the speed will bleed off untill the aircraft reaches Vmca and yaw control is lost the aircraft will then yaw and roll towards the dead engine very rapidly and will usualy hit the ground inverted.

I can only speculate that the left turn in this case was the pilot starting to lose control in yaw as to try to turn would have further reduced the allready very small ,if any rate of climb.

Thank you foghorn and A&C, very sobering stuff. This pilot had 10 years experience, but I'm not sure of his accumulated hours, he was a businessman pilot.

It seems to me a pilot who buys a light twin has to be aware of the performance limitations of his/her aircraft on one engine, including the parameters A&C discussed. Buying a light twin is not a decision to be made lightly, especially by a low time pilot. I, like many, also feel that light twins were certified to an inferior standard.

The really sad thing, is this pilot seems to have done everything right (except for not declaring an emergency), with a good showing of piloting skills, up until the point where he decided to perform a go-around. If he had just flown the a/c to touchdown and stood on the brakes, everything would have been just fine, to the joy off all.

At certain times you are just one bad decision away...

A turn towards the dead engine isn't a problem in itself, as long as the usual caveats are observed such as above Vmc etc.

Sounds like a fast approach, speed not bleeding off in the flare as fast as the end of the strip seems to be approaching so a go-around started.

As has been said, without the correct configuration the only altitude gain will be by trading off airspeed. If persisted then a Vmc departure is the likely outcome.

And having said all of the above, an overrun off the far end of the runway at 20 kts is far preferable to a stall/spin fatality on a failed go-around!!

During my ME training my instructor took me through some go around on a single engine, I was not aloud to use trim. The alt where I was told that the r/w was blocked was about 400’, with one engine shut and the other at full power it was very hard on the leg, after one circuit and one go around the leg started to shake. Hard to say why the pilot did what he did but he made one bad mistake, he has tried to go around.

AC-DC:

When you were given the single engine go around were you using simulated zero thrust on the "failed" engine?

The engine wasen't actually feathered was it?

Cat Driver:

Would it not make more sense to make a very shallow climb,gain alot of airspeed (more aerodynamic stability=easier to control) and get out of the circuit..trim out..them go and land your plane?Then I`d make sure I have the altitude (when you eventually reach the airfield again),come in fast with the throttle closed an the live engine and it will handle alot better will it not??

ME training scares me as they seem teach you howto fly the aircraft like a turboprop,and drum it into you that you must chase the blue line speed all the time..when there are better options?

Always easy to speculate and criticise a decision after the event. When things go wrong it is often difficult to make the right decisions and this is why minor problems turn into catastrophic disasters. Despite having loads of runway ahead I guess he just felt he had got it wrong and wanted to give it another go. In the cool clear air of sitting in front of a computer screen I would opt to land and take an overrun unless there was a sheer cliff ahead of me (or a major highway).

Only flown Cessna 310 and 340 and Seneca. Had one engine failure to date and flew on to next airfield to land declaring dead engine but not calling mayday.

The Cessna 310 would want me to put it on the ground ASAP in the case of failure of one power unit. In feathered condition the climb is very S L O W.

Did an airtest on it once and the feathering linkage stuck so had to fly around for an hour trying to get it right (assumed it was iced up). In the end went for feathered engine approach into our 800 mtr strip. Not much room to get it wrong and I don't think a low go around would have been good news.

I would rather take the certainty of almost right and safe than maybe better or maybe worse but the voice inside your head says "I can do it better next time".

I was astonished recently to learn that a light twin loses something like 80% of its performance with a single engine failure. But I was even more surprised to learn that there is apparently no published POH engine out allowance for the Seneca even at sea level and ISA. Is that really true? If so, I wouldn't go near one with a barge pole.

Why is there no twin PT6 centreline thrust aeroplane with contra rotating, independent props? Darn sight safer than these 'conventional' twins, I'd have thought?

I would agree with EtOPS that a shallow climb all the way upto a safe altitude would be a must prior to turning. However, being that I fly a senecca all the time I would not recommend a single engine go around feathered por not, especially with 5 people on board. That aircraft was probably pretty close to gross wt. What is a huge factor missing is what was the density altitude. If it was a hot day then we can expect relly bad performance.

As to why would he go around. That would be second guessing, but I would rather belly up down a runway into the grass and even knock into a fence/trees at the end of the runway at 10kts than try and climb in one of them. Yes you can, and I simulate feather for my students during a Single engine ILS and make them Go-around from the DH. If they don't clean up quickly that rate of decent doesn't reduce much. I think the left turn if intentional may have been a bad idea as you further lose climb performance due to increased Drag. What I see a lot of here in the US is that people own there own plane and fly around 10years, but when do you think was the last time they practiced a lot of engine failures. I see this when I give people the required two year flight review.

Beagle, most light twins lose about 80% of its climb performance when they lose an engine. The senecca is 83%. The FAA only requires aircrafts over 6000lbs to have a single engine climb performance. It is published in the POH on seperate tables what the multi and single engine performance should be at different altitudes.

At least with a single you know you're always going down when the engine quits. With a light twin, the possibility of remaining in the air on one is definitely a double-edged sword. And, of course, you have twice the chance of an engine failure.

We had a good thread on this a while back in which I vented similar prejudices. (Jeez, I've turned into a pprune old bore.)

QDM

S/ ENG. APPS
 

N4S23,
I assume you are using an artificially high DH (or simulated committal height) for your s/e ILS,as you are putting yourself right into the scenario of the subject crash. Your DH on an ILS must reflect the fact that you may have to go-around ,and most normal DH`s on ILS`S are around 2-300 agl; you must descend whilst cleaning -up and getting power on and controlling the a/c first. That should never be taught at 2-300 ft -never..

Formationfoto,
Why did you try to get into an 800m strip when there are lots of civil/mil airfields around the UK, AND not declaring a Mayday either? I would hope that you would not repeat that again..:rolleyes: :rolleyes:

N4S23, the temp was 6C, dew point -6C, altimeter 30.35, airfield altitude of 770ft.

I can't help thinking that a little knowledge is a dangerous thijng, (and a multi-engine rating on a PPL is certainly 'a little knowledge').

A few FACTS about flying light twins:

1. Most of the flying done in light twins is at low AUWs - even an old Seneca 1 will climb on one engine with only 2POB and a light fuel load. In this condition, they're quite safe.

2. If you've any brains at all, you will declare a single-engine committal height prior to any approach, below which you will land or crash, but not attempt a go-around. How low you set that height is largely down to the aircraft. It should be about 500-600ft for a competent pilot in a good aircraft.

3. What sort of airfield you will land at with an engine out is down to you - I wouldn't go into an 800m strip with an engine out, but then I've only got half-a-dozen thousand hours or so.

4. The comments above about 'not declaring a Mayday' demonstrate a weak understanding of the big picture. What do you think this will achieve? Calling Mayday doesn't get you any more thrust, height, or speed - and those are the things you need! Yes, the little red land rover might come out to meet you at a small airfield. At a larger airport, the fact that you're on one engine (assuming that you bother to mention it to ATC at all - and you should, because you're not going to be able to taxi far after landing!) will prompt the controller to get the AFS out.

The main point, I think, is that a Performance E aircraft is not a performance A aircraft. that's why you don't buy a ticket from an airline and find yourself on a Seneca. I believe that pilots of light twins know far too little about the real performance of their aircraft. Yes, you can go-around on one engine in a light twin, provided you start at the right height and do the drills correctly whilst flying accurately. No, you can't do it lower than that, or if you get those things wrong.

Rumbo,

By calling a Mayday or Pan gives you free airspace and therefore extra brain cells to fly the aircraft.

Its called load shedding. I would rather tell air tragic so that I know I will have the runway for me and everyone will get out my way. Also the R/t drops dramatically.

This has to be a good thing for low houred pilots in particular. They need all the brain power to fly the aircraft.

Something I see very frequently in Multi Training (Senecas) is that pilots come in high and fast on a simulated single engine approach. They are frightened to get "low" and "slow" on the approach (a good thing) but fail to achieve the correct approach angle. They then find themselves over the grass strip going at warp factor ten. If they had made a definate decision at commital they would have the option of go-around. But they very rarely do. They are hell bent on salvaging a balls-up approach.

etops773
 

I dont think that you understand the lack of performance that you can expect from a light twin with only one engine working.

The climb will be very shallow when the aircraft is flown at "blue line " speed , above that speed and the aircraft will not climb and below that speed the aircraft will not climb and you run the risk of the speed decaying towards Vmca , on a good day dont expect more than 2-300ft/min ROC.

Light twins require far more in the way of stick and rudder skills than larger aircraft and a deep understanding of aircraft performance (or rather the lack of ) it is clear from some of the posts above that people expect far more in the terms of performance from a light twin than it can give.

The rule of thumb I was given, by a very wise and experienced instructor, was that in a light twin the failure of one engine means that the other engine will take you all the way to the crash.

Some will climb on one engine, some won't, he said. Persuaded me to go for high-peformance singles rather than twins. I have no regrets on that score.

I've had my engine failure (are we only allowed one each under JAA?) in mid-channel - made it to land. That was in a single.

Airprox,

You're absolutely right, if we're talking about CAS. Outside CAS, calling Mayday may help, but it's certainly not a priority. How often, when instructing, do you see the stude get the RTF call out in near-perfect style whilst forgetting to fly the aircraft? This is something that really used to worry me - studes forget that we survive by 'aviating, navigating, and communicating', in that order, whatever aircraft we're in.

Flight Safety,

This aircraft has a max landing roll of 1400ft.

Don't know what this means. Did you mean MIN landing roll?

Engine failure is a mayday situation.

Sounds like he had time to make the call (failure whilst en-route) - and this may or may not have changed the outcome, depending on what balked him into trying to go-around from such a low height

sycamore,

Your DH on an ILS must reflect the fact that you may have to go-around ,and most normal DH`s on ILS`S are around 2-300 agl; you must descend whilst cleaning -up and getting power on and controlling the a/c first. That should never be taught at 2-300 ft -never..

The CAAFU will expect you to do an ILS to height 250' simulated assymetric (assuming a normal cat1 ILS with 200' DH and having explained to them that ACH would be 350')

You learn that way, and you're examined that way for IR.

CaptAirProx,

Something I see very frequently in Multi Training (Senecas) is that pilots come in high and fast on a simulated single engine approach

Do you have an ILS they could use for the approach (even in VMC) so they can see that engine-out 3 degrees is still do-able?

Keef,

...light twin the failure of one engine means that the other engine will take you all the way to the crash.

Surely the first part of that sentence is: "An improperly handled..."?

Handled appropriately could mean accepting a forced-landing cf an SEP, descending to engine-out ceiling (and planning that's above MSA) - or it may not mean either of these things...

Hi Rustle,

The CAAFU will expect you to do an ILS to height 250' simulated assymetric (assuming a normal cat3 ILS with 200' DH and having explained to them that ACH would be 350')

You learn that way, and you're examined that way for IR.

In the perfect world everybody would learn this in a decent simulator. Problem is, it's expensive, but then so is flying a twin, and so are funerals :( When flying on a moonless night over mountains or the desert in a single, or being low and slow in a twin with only one engine producing power - you are really betting on engines being as reliable as they are.

You are correct in saying that if properly handled a twin should be safe. Recurrent sim training is the only real way to ensure you can properly handle a twin, you cannot practice dangerous situations in a real light twin because they are too dangerous!! Especially if the person training only has a few more hours than the student which was very common when airlines would hire instructors with only 100 hours multi time.

If you are engine out in a twin and ceilings are low at your destination, a properly handled engine out trip to your alternate might be better airmanship.

regards

Again, I'm worried by all the talk about ILS DHs and one-engine-out flying. The two are completely separate!

There is nothing wrong with flying the ILS to DH with one engine simulated failed. There is a lot wrong with thinking you can go around with an engine genuinely failed at 200ft. I'm not a CAAFU man, so I won't comment on their procedures, but when I do ILS training and testing in light twins I always have in the back of my mind that if an engine fails for real below about 600ft, then it's straight into the single-engine approach configuration, attitude, and power setting, and with the intention of landing on whatever is available when we see the runway. It's no trouble provided you've got Cat 1 conditions or better.

Similarly, there's nothing to stop you carrying on to your destination if the weather is good enough to carry out an approach and landing. You simply fly the approach with the decision to land already made. This is where a Mayday will get you a sterile runway etc. (at most places).

It's the big picture, guys and girls!

Rumbo de Pista

Don't worry, no confusion between DH and ACH here :)

Merely highlighting the expectation from CAAFU - simulated (read zero thrust) assymetric, go-around at "normal" DH.

and thanks for not correcting my deliberate cat 3 error...
(now edited and corrected :))
I meant cat 1 :rolleyes: DOH!

Rumbo
Most of the people here just have not got the pictue yet and it is simple......A light twin will NOT go around from ILS minimums on one engine if you try you will crash if you are lucky it will be a soft crash onto the flat ground near the runway if not it will be an object that you cant clear because you cant clean up the aircraft and pick up speed or as you try to climb the speed drops off and you go below Vmca roll on your back and hit the ground inverted.

So this having been said you have to pick a height at which you can clean the aircraft up while trading height for speed and then climb away at blue line speed this height is not likley to be much below 600ft AGL (for pilots new to twins) now that is about 350 ft above an ILS DH so if you have a cloud base below 600ft then you are landing at the wrong airfield.

Think about this hard guys with an engine not working you lose about 80% of the aircrafts performance.
I am sure that the CAA dont expect you to go around from ILS DH with one engine (simulated ) out because they know the aircraft wont do it and they value there lives to much ,the CAA may seem daft some times but they are not stupid !.

From my IR test I remember that the engine was failed on the go around from an ILS and once the aircraft was clean ,the single engine approach was an NDB with a MDH of about 450 ft ,not a lot of room to play with but workable for those of you who are in practice.

Chuck Ellsworth

That is correct.

A and C,

Trust me on this one.

IRT wasn't that long ago that I have forgotten, nor was the renewal :)

Simulated EFATO after NDB approach - handle it, then back to two engines...

Then, simulated assymetric before RV for the ILS, assymetric ILS to DH (having explained to examiner that ACH was 350'), go around from 250' (plus 50 minus nothing - usual tolerance) into the visual circuit for assymetric bad-weather circuit to land.

This is daft as a blanket statement.

There are many light twins that, at some or all weights, will go around quite capably from ILS DH on one engine -- some do so with rather more alacrity than a single doing a go around. To rule out the idea of a go around in all circumstances may turn a perfectly manageable situation into a an accident.

It's also undoubtedly the case that many light twins at or close to max weight cannot safely carry out a go around from ILS DH.

The pilot needs to know the performance that can be expected from the aircraft in the prevailing conditions and at the actual weight. Ignorance of that can lead to some poor risk management in either direction. Perhaps it did so in Oklahoma.

it depends on the a/c load, wx, pilot´s skills etc.
I often fly on C303 and C421. The C421 makes a se-go around, the C303 not. I beleive with a Seneca it is the same problem..!?
Even our KingAir C90 doesn´t like ga´s at full load... ;-)

What about the rudder trim?

As the engine failure happened a long time before the approach could he have trimmed out the rudder force, and then not removed it – combine this (maybe) with a high approach speed and stress maybe resulting in confusion just above the runway (what’s it doing?) “forcing” him into applying full power?

Or

If he didn’t trim out the rudder holding it could cause tiredness and associated judgement error

Or

He may have trimmed it then removed the trim before starting the approach

I know my leg killed me after doing a join, hold and NDB approach on one engine (IRT)

The problem is not just whether it will/will not go around.

The available climb gradient capability will often not be sufficient to meet the missed approach min. obstacle clearance gradient.

You may be climbing, but not steeply enough to follow the charted missed approach.

Ok at some airfields eg Sumburgh & others that have lots of ocean over which to climb. Not so good at others that have nasty mountains around.

Hi Beagle

Not sure what you mean by your statement:

May be you can elaborate a bit and I will have a look and see if I can find what you are looking for.

Formationfoto I can only echo Sycamore's sentiments:

If things would have gone wrong that would have made interesting reading in the AAIB report.

'The experienced pilot encountered a problem with the feathering mechanism which he tried to sort out by flying on one engine for an hour.

Version 1. He then suffered an engine failure on the live engine and had to make a forced landing in a field the aircraft was a write off in the subsequent landing.

Version 2. After he established that the problem could not be sorted out in the air he flew back to the home base a 800 meter grass strip............................

BTW what is the accelerate-stop distance for a 310?

MHO of course

FD

Flyin'Dutch'

BTW what is the accelerate-stop distance for a 310?

There's a page of gubbins about this in the POH depending upon weight and temp, but here's one example of a 310R Turbo (with assumptions):

2700rpm/32" MP
Full power before brakes released
Nil headwind/tailwind
Flaps up
Level hard surface
Engine fails at Vr
Live to Idle and heavy braking at Vr ;)

5500lbs AUW (= MAUW)
10 degrees C ambient
Sea level
Vr (this weight) 92KIAS

Accel/Stop distance is 3150 feet.

Not a list of variables you want to figure "on the fly" so to speak...

Hi Rustle,

Thanks for your reply.

IIRC the CAA advice a factor of 1.3 for short dry grass which takes this number to 4200ft which if my maths is correct is only 502 meters or 1600 more than available at that strip. :eek:

Interesting to see how people are happy to paint themselves in corners.

FD

I concur with rustle regarding the IRT - I know that at least certain examiners in the Leeds CAAFU require the assymetric approach to be down to minima for the respective approach followed by an assymetric go around from DA/MDA.

Ditto on the renewal test with an IRRE.

Daft aha ?
 

So Bookworm a daft statment by me ?.

I see little point in writing about what a lightly loaded aircraft will do on a day when the pressue is high and the temp is low.

This is a thread about aircraft performance at the legal maximum weight at an airfield that has the temp and pressue at the WAT limit.

I think that most of us have worked out that if an aircraft is light it will climb better than if it is at its MTOW.

The fact is that a light twin at a with a typical load on an average english summers day with a low time pilot is unlikly to make a safe go around from less that about 600ft.

Yes bookworm you can pick perdantic holes in some of my statments but you look at the performance graphs and you will see that these aircraft will make 150-200ft/min while crossing the ground at 95ish KT when the aircraft is clean , its not a lot of up for a lot of distance across the ground.

Now will you please tell me how you clean the aircraft up and increase the speed from the final approach speed to blue line from an ILS DH of aprox 250 ft without going down and trading height for speed ?.

I have had to do a single engine go around in a light twin for real fortunatly I had nominated a very conservative committal height of 500ft AGL and it was not to much of a drama but if I had nominated 250ft AGL I would be dead it is as simple as that .

A and C may well be daft, but stupid ? no , and very much alive !.

The SE rate of climb of a PA34-200T (don't have the data for a 220T) at sea level is in the region of 200' per minute, at 4570 lbs (max T/O weight), gear up, flaps up, inop engine feathered. At 10,000' this is in the region of 140' per min at standard temp. SE ceiling is around 14,000'.

So its not always the case that a light twin won't climb on one engine, though the POH for the 200T states that a SE go-around "should be avoided if at all possible" and that under "some conditions of loading and density altitude a go-around may be impossible"....

I was always taught to commit myself to landing when at the FAF of a precision approach and the gear goes down, whether it is on a runway, taxyway or midfield, rather than go around. There is nothing wong with practicing asymetric landings all the way down so long as the 'simulated' engine failure is carried out with the throttle, and that if a go-around becomes nescessary then both throttles are used.

Cheers
EA

Sycamore and Dutch
The 800 mtr strip is my home base and one which I have regularly practiced single engine approach to land, and on singles engine off to land (simulated). So I knew what to expect as I had practiced it many times. The time spent trying to free the feathering mechanism (to get to unfeather and to restart the engine) was all spent at 2000 ft overhead the field from which a full engine out glide could be achieved.

Oh yes forgot to mention that I also had a test pilot with me so my confidence in achieving the landing safely (which incidentally I did with around 200 mtrs of paved runway to spare - an a day with a helpful 15 kts of head wind) was not entirely the result of blind faith in my own abilities.

Had I regarded this as beyond my experience or close to the margin for the conditions / aircraft I would indeed have elected for a lengthier runway and more options.

One of the advantages of a relatively short runway for regular operations is that you get used to handling the aircrast in those conditions.

A and C

Now will you please tell me how you clean the aircraft up and increase the speed from the final approach speed to blue line from an ILS DH of aprox 250 ft without going down and trading height for speed ?.

The approach to DH was flown at (min) Blue Line.

Normally you're at blue line to ACH, so during the IRT/renewal you're at blue line to DH.

Remember that Vat in a GA7 is 75KIAS and Blue Line is 85KIAS, losing 10 knots in a Cougar in 200-odd feet (or 3/4 mile) isn't hard ;)

PS, check your PMs

A and C

I apologise for my use of the word "daft". It was unnecessarily emotive.

Nevertheless, I disagree with you.

But it's not. Neither in the thread leading up to your statement nor in the statement itself do you qualify your "a light twin will NOT go around from ILS minimums on one engine" line. You make it sound as if it applies at any weight.

To take an example, my own aircraft manages a lamentable 150 fpm at MTWA, even at ISA sea level. I wouldn't dream of attempting a go around below 600 ft. With just me on board at the end of a journey, I get at least three times that. To "crash" in preference to a go around in an aircraft capable of more than 400 fpm may be your call, but it's not mine.

In answer to your question, I aim to reach ILS DH at no less than blue line speed (91 kt) anyway. It's not hard to lose the extra few knots required in the half mile from there to the threshold. Flying a single engine ILS, I'd probably fly at 100 kt to DH anyway. Though it's not ideal, most ILS-equipped runways are long enough for that to be fine.