Hawarden Nov 2013 fatal crash - AAIB report

[Shaggy Sheep Driver]

Air Accidents Investigation: Cessna 310Q, G-BXUY (http://www.aaib.gov.uk/publications/bulletins/november_2014/cessna_310q__g_bxuy.cfm)

Very sad. Failure of the critical engine due fuel mismanagement. What I see hiding between the lines in this report is an apparently competent and conscientious pilot who did things correctly, but perhaps had a bit of an obsession about saving a few bob on fuel?

[Chilli Monster]

A little bit unfair I would say. He arrived with 45 minutes of fuel on board - just a shame it was in the wrong tanks. Had he done the sensible thing - got airborne on the mains, ran down the auxiliaries, then switched back to the mains - no problems.

[Shaggy Sheep Driver]

Indeed, you may be perfectly correct.

He had the fuel on board, and must have known how low the main tanks were getting. One wonders why he didn't use the aux tanks, or if he did (that seems a possibility in the report) why he didn't use them earlier and therefore for longer? he seems a switched-on sort of guy who would be aware of what was going on.

Logically of course he wouldn't save fuel by not using the aux tanks for longer, simply increase the risk of the mains running dry. Yet the report makes quite a point of his keenness to save fuel costs wherever he could, and the accident was the result of fuel starvation, even though he had it on board but in the wrong tanks. There may be absolutely no connection there, of course. Or there may have been some sort of subconscious reluctance to use what was in the aux tanks.

[charliegolf]

From the gps plot in the report he seemed very close to the runway.....

Had he chopped the throttles, would he have made the runway?

I ask not to second guess him, but just to know.

CG

[Shaggy Sheep Driver]

From the report, I'd imagine he was unaware the left engine was failing as the CSU was masking its surging to some extent. He'd be aware 'something' wasn't right as the aircraft departed heading, hence the go-around being initiated. Then I think what followed would happen too quickly for him to do much about it.

Incidentally did I read the report correctly re the anti-feather latches? If the engine fails in that the oil pressure to the CSU drops, the prop will feather, but if the engine is stopped on the ground in the normal way, the latches hold the blades at positive pitch? And in this case, where the engine ran down due lack of fuel the prop would behave as if in a ground shut shut-down; the latches would engage and the prop would not therefore have feathered?

If so, that seems an unwise design flaw, doesn't it? Surely one always wants the prop on a twin to feather if its engine fails?

[Jhieminga]

You are correct, but the crux is in the speed of the prop running down and the reaction of the pilot. If the prop rpm drops enough before you get the lever in feather, you may be too late. I think there was a discussion about this not too long ago.

Edited to add: and you need the latches as you wouldn't be able to get the engine running with the prop in feather. At least that's how it's always explained.

[Barcli]

Most are built that way SSD, Senecas, I know intimately, you have to physically feather the prop using the lever before it ceases to rotate, I belief only turbo props will " auto feather" when engine power is lost

[Shaggy Sheep Driver]

OK, so it will never feather unless the pilot puts the prop lever manually into 'feather' while he still has enough RPM for the ant-feather latches not to have yet engaged?

[Jhieminga]

Have a look at the article in post no.1 of this thread: http://www.pprune.org/pacific-general-aviation-questions/549618-propellor-feathering-light-twin-engine-aircraft.html

[Shaggy Sheep Driver]

Thanks Jhieminga. Scary! Wouldn't it be better to arrange things so a prop on a failed (or shut down) engine on a twin goes to fully coarse, instead of fully fine?

Going to full fine is like an airbrake! (At least it is on a Yak52!).

I'd go further and say the prop should fail to 'auto feather' position at all times. For normal shutdown after flight there should be a facility for the pilot to engage the anti-feather latches while the prop is still turning, so subsequent start is easily achieved.

In other words, the failsafe position is 'autofeather', with a pilot-operated over-ride for use during normal, non-emergency, shutdown to prevent auto-feather.

I realise that we have strayed from the original thread subject. If the mods want to move this, that's OK by me.

[Sir George Cayley]

I noted both throttles were found fully forward which suggests to me the pilot thought he had full power available on both engines.

With the runway length available keeping the speed up above the blue line wouldn't be a problem would it?

SGC

[Shaggy Sheep Driver]

I think the report alludes to the left engine CSU masking the failing engine, so he probably thought he did have full power available for the go around. The pitch up is a bit of a mystery, but the report talks of it being perhaps power and trim induced. Perhaps the left engine was developing power intermittently, and happened to respond initially to full power being selected, leading to the pitch up, only to fade almost immediately leading the almost instant loss of control.

[9 lives]

A 310Q has an excellent fuel system, but it does require pilot management, and there are a few "gotchas", but experienced or well trained 310 pilots know them. There are system reasons to use certain tank's contents in a certain order. There are also pumps which circulate the fuel within the tip tanks. I suppose of one of those pumps quit, the fuel in a tip tank might not be as available as one might hope.

GA opposed piston engines do not have auto feather systems, so the pilot is going to have to command the feather, if it is desired. However, if the engine is allowed to stop, or seizes to a stop, the prop will not be featherable, as it feathers only with open blade latches, and on centrifugal force, which goes to zero with no rotation. This is warned in some POHs, but not all. I have to convince to have it added to the Lycoming powered DA-42 Flight Manual, as it was not a tribal knowledge item at Diamond. Thus, I make it a part of training whenever I can.

You would not want a prop going to coarse pitch after a failure, as there is still lots of drag from a coarse prop, and then that drag would increase once you went to feather. The plane will be hard to handle in this configuration, so anything that prolongs that is worse.

A 310 is a capable plane on one engine, but it gets bad fast if you let it out of control - you just may not get it back.

For those with more interest, this article is really good:

http://www.avhf.com/html/library/Leave_Yourself_An_Out.pdf

[Mach Jump]

I'd go further and say the prop should fail to 'auto feather' position at all times.

At first glance this might seem desirable, but the problem with this is that if there was a short power loss for a few seconds, such as running a tank dry in the cruise, the prop would feather, and stop rotating. This would leave you with a stopped, rather than windmilling enging, which may be impossible to re-start after changing tanks.

You would not want a prop going to coarse pitch after a failure, as there is still lots of drag from a coarse prop, and then that drag would increase once you went to feather. The plane will be hard to handle in this configuration, so anything that prolongs that is worse.
:confused:

The drag of a windmilling prop reduces as the pitch becomes coarser, reducing to min. drag as the prop reaches the max. coarse position, Ie. feathered.


MJ:ok:

[Miserlou]

Step Turn said,
"A 310Q has an excellent fuel system, but it does require pilot management, and there are a few "gotchas"".

So it doesn't have an excellent fuel system then!
An excellent fuel system would have 'on' and 'off', not require any further management and there would be no gotchas.

[piperboy84]

You would not want a prop going to coarse pitch after a failure, as there is still lots of drag from a coarse prop, and then that drag would increase once you went to feather.

I know SFA about this, but I thought that statement was correct as the purpose of feathering was to get the prop to stop, and that being at course pitch would create more drag than fine because it was windmilling faster on an engine not producing power. Is that not correct?

[27/09]

A prop at fine pitch creates more drag than a prop at coarse pitch. Pulling a non feathering constant speed prop into full coarse will extend your glide in an aircraft like an Arrow or C182 etc.

Having a prop go to coarse pitch when power is reduced or lost means you no longer have a constant speed prop. A constant speed prop reduces blade angle, i.e. goes to a finer pitch when power is reduced in order to maintain the set RPM. Likewise when power is increased the prop goes to a coarser pitch to absorb more power and maintain the set RPM.

The latches on a feathering prop only drop into place at RPM below about 1300 to1400 RPM, which on an engine with a constant speed prop is a pretty low power setting. Even at idle power and any airspeed above around 80 knots you'll find the RPM will be at least 1400 RPM.

Remember RPM hardly even changes when an engines fails, the constant speed prop reduces blade angle and RPM pretty well stays the same as before the engine failure. Even manifold pressure remains pretty well constant after a failure. Try "failing" an engine sometime at a safe altitude by pulling the mixture to Idle Cut Off, you will see what I mean. I'm primarily talking about a twin here as the speed remains relatively high and the forward speed keeps the failed engine turning. I would expect a similar situation for a single too at a normal glide speed.

Under pretty well all conditions with an engine failure there is no rush to feather the prop for fear of it not feathering due to the latches operating, at least with a metal prop, some of the lighter composite props may act differently.

[Above The Clouds]

Piperboy84

You are confusing rotational drag with drag created by a disking prop due to forward flight.
Drag created from a windmilling prop in fine pitch during forward flight far exceeds that of rotational drag created by a prop in course pitch, hence the sense of urgency to feather the propellor to avoid disking.

[Shaggy Sheep Driver]

Yes, my eyebrows rose when I read the post above that a course pitch prop would produce more drag than a fine pitch one. I'm not a multi pilot, but put that massive paddle on the front of a Yak52 into fine and then, as you hang in the straps due the nose-low pitch angle to keep the aeroplane flying, tell me it's not producing as much drag as when in course!

[9 lives]

A coarse pitch prop will have less flight drag than a fine pitch prop, but not a lot less, compared to being feathered. Definitely, coarsing out a CS prop makes a difference when you're gliding a single, but no where near as much difference as feathering it. In a twin, if the engine stops running, (and you're not going to get it running again) you must feather it. If you select coarse pitch without feathering it, yes, you'll get a little less drag, but an increased likelihood that the prop will stop turning, and then you won't get it feathered at all. At that point, stuck in coarse pitch, you're in trouble.

I have spent hours flight testing this on the Lycoming DA-42, Twin Comanche, and Navajo during propeller change approval work. In many cases I found that an engine shut down with the mixture and the plane around Vmca, if you did not feather promptly, the engine would stop completely, and could not be feathered after that. This was my common experience. A starter restart was my only option at that point, because it would not windmill start, if it had stopped. With lots of altitude, I could sometimes dive it, but this is a very risky waste of altitude.

One of the few single engined aircraft I know of, which can be feathered in flight is the Caravan. I have feathered and glided them may times during testing (to eliminate "P" turbulence under the belly for a minute). The difference between coarse pitch glide, and feathered drag is incredible. So having had that experience, I would never want to be stuck in coarse pitch in a twin, if I could get to feather.

For those who read the article, twins on one engine can be very marginal - and that's with one feathered. If it's not, that really marginal performance is very compromised, so may be negative.

I flew both the C 310R, and the C 303. The 310R had six fuel tanks, which certainly must be carefully managed (so yes, you could run out on two of six, if you don't reselect, and you could even pump fuel overboard, if you go ti wrong). The C 303 had only two tanks - left and right - On or off. In the 310, you could manage the fuel on a long flight so as to safely run four of those six tanks dry. The result was that the uncertain fuel quantity in the remaining two tanks was a much lower percentage of the total fuel, so your fuel quantity planning could be much more precise, and less worrisome. Or, in the C 303, you could watch the two fuel quantity indicators bounce under 1/4, knowing the accuracy of some Cessna fuel quantity indicators, and knowing that you're not seeing the coast of Iceland yet - that made me nervous. There are also excellent structural advantages to the 310 multi tank system. I'd rather learn to, and then manage the fuel to gain those benefits, than to get ot be lazy, and suffer a system with less utility.

[27/09]

In many cases I found that an engine shut down with the mixture and the plane around Vmca, if you did not feather promptly, the engine would stop completely, and could not be feathered after that.

This may well be the case in flight testing scenarios where you operate to the edge of the envelope and sometimes beyond. As a matter of interest what type of props where you testing?

However, When is an aircraft going to be at Vmca and feathering is going to be needed very promptly?

The way I was taught to fly a light twin the procedure for any engine failure below blue line was to close both throttles and land ahead. This was due to the fact that at blue line, performance was marginal and worse than marginal below blue line, and speed was likely to decrease further before you got things sorted and attempted to accelerate to blue line ,plus you would be losing height. In the intervening time you were most likely to have contacted the ground or lost control at Vmca.

You were much better off to treat the light twin like a single engine aircraft below blue line and land ahead while maintaining control rather than trying fly away and lose control.

At blue line and above I think you'll find the prop will windmill at speeds that allow you to feather the prop in a considered and timely fashion.

I've seen the photos of botched shut downs, wrong engine feathered, and aircraft forced landed with both engines shut down. These incidents may may have been caused by the pilot thinking they had to rush to feather the prop. Under normal circumstances you have adequate time to do your drills properly. I'm not saying you have forever but you don't need to rush.

[Big Pistons Forever]

The way I was taught to fly a light twin the procedure for any engine failure below blue line was to close both throttles and land ahead. This was due to the fact that at blue line, performance was marginal and worse than marginal below blue line, and speed was likely to decrease further before you got things sorted and attempted to accelerate to blue line ,plus you would be losing height. In the intervening time you were most likely to have contacted the ground or lost control at Vmca.

.

Excellent advice from 27/09.

This is exactly what I also teach. 2 other points

1) My end of flight checklist for a 310 (and similar complex pistons) i sdevided into 3 short set of checks. It goes descent/approach. prelanding and final.
The first item on the descent check is "fuel selectors & quantity". It is designed to prevent exactly this kind of accident.

2) System knowledge is important. The C 310 POH specifies that the aux tanks be used in level flight because they have no slosh box or other anti unporting features. However this does not mean that the engines will stop if they are used in other than straight and level flight. In a case like this accident where for what ever reason, main tank fuel is critically low, aux fuel can be safely used as long as pitch and bank angles are kept low.

[9 lives]

I certainly agree that an engine failure in most light twins is well treated by a throttles closed, land ahead, if that is what the pilot is comfortable with. The 310R will climb away on one, with great care. That was a part of my multi endorsement, which was done in the 310, back in the day when single engined work was compete shutdowns. The 310 fuel system makes it near impossible to be running on the aux's with empty mains, as the excess fuel from the aux's will be returned to the mains, so when you go back to them, you'll always have more than was there when you left them. This is a part of getting to know the Cessna twin fuel systems.

I have tested MT props on several twins, as well as a number of other external mods, which required complete retesting of all of the single engined performance. This was always done with the critical engine shut down and feathered - so lots of restarts, and a couple of single engined landings. My DA-42 Lycoming testing was also for feather/unfeather with MT's and McCauleys, if I recall, and in flight restarting, which I did dozens of times. We did have a snag in the unfeathering accumulator, which necessitated some starter restarts from feathered - lumpy. I never did any of this close to the ground, and a lot of the critical stuff within a glide to the runway.

[UV]

perhaps had a bit of an obsession about saving a few bob on fuel?

He arrived with 45 minutes of fuel on board - just a shame it was in the wrong tanks

Did anyone else notice that he appears to have arrived at Seo de Urgel with 26 litres on board? The report says he departed with 448 litres in the tanks but that he uploaded 422 litres... (See Table 3 on page 18)

And thats not counting the 7 litres of unusable fuel (page 14)...