Caravan engine failures

[lilflyboy262]

I have posted this in the maintenance section as well, but I would like to get some answers and need the combined knowledge of all those out there and feel more people will see it here.

In the last 18 months that I have been in Botswana, we have had two caravans go down shortly after take off from seperate companies. Before you automatically think africa = poor maintenance, in this part of the world it is not so.

One was just out of a heavily contaminated airstrip, where shortly after getting airborne, a loud bang was heard and the plane carried out a forced landing, where it overturned in the soft field and the aircraft was damaged beyond repair. Thankfully with no fire nor fatalities.

After pulling the engine apart, nothing was found to be wrong with the engine and the thought was perhaps a water induced flame out.
There was no rain at the time and the aircraft took off with the inertial seperator was open. After it was closed the "bang occured"

The second accident occured a few days ago where shortly after getting airborne, there was a loud bang heard and the plane went down. Unfortunately this time the plane clipped trees and came to a halt against some other trees approximately 50m futher on. There was a fire post impact. This time we were not so lucky and 8 people died, including the pilot, 4 luckily escaped.
This time there was no weather at all. Temperature was around 25-30 degrees (same as the first accident).
Company proceedure at this company is to also close the seperator shortly after takeoff.

The accident hasn't been throughly investigated as it only just happened and I can only speculate as to the cause of the accident. But meanwhile my friends and co-workers continue to fly while this is investigated and I'm looking for answers while I am on leave.

My concerns are that is there a possibility that at a certain airspeed, temperature, and angle of attack, whereby closing the seperator could lead to a surge/stall situation with disrupted airflow into the engine.
Could this be a situation unique to operating these aircraft here? Could there be other accidents out there like this one?

[square]

Do you know if the continuous ignition was turned on in these accidents? I know of an incident where a pilot had a flame-out in a Caravan and turned the continuous ignition on, and apparantly it fired right back up without losing too much % NG. I always use it on arrival and departure when below 1000' AGL in the van.

[MUSKEG]

If I understand this correctly the seperator is off for T/O? presumably to get the torque and cooler temps required for T/O? Then after T/O it is engaged. I have not operated a caravan so not sure as to why this is being done. Our off strip stuff (Beech,DHC-6) was done with the seperators engaged for ground ops, T/O and landing, and icing. Why do these companies engage the seperators once away from high FOD area?

[MUSKEG]

It might be we are talking the same language using different terminology. Open= engaged, closed=disengaged? Funny how this language can work in different settings and application.

[lilflyboy262]

Not entirely sure, unless they already had the ignitors on, on the take off roll.
In both cases they happened very low. The fatal one was near MAUW so would have come down like a stone. I would have been more concerned about shutting everything off than trying a relight at that altitude.

Most of the time the seperators are opened sometime on the approach to land, open for the hot turn arounds, and then closed shortly after takeoff. Our company SOP's are to close at 40kts on the take off roll before getting airborne. By then most FOD isn't an issue other than large birds.
Common sense to apply when operating around heavy rains and heavily contaminated runways (like what will be happening in the next few months)

The seperators are designed for use in flight so whenever you close them should not be an issue. But so far it is the only thing I can see that is in common with these two accidents that have a very similar sequence of events.

These engines are bullet proof. I have seen compressor blades that have had FoD take huge hunks of blades out. The engine continued to run for over a week before a pilot finally decided that something was wrong and took it in to get checked. Have also seen the attachment to the seperator come loose and the door being able to open and close at will.
Its why I think that there might be something more unusual at play.

[Mr. North]

Just to be clear, in the Caravan the Inertial Separator is set to BYPASS to engage. It is turned off by returning it to the NORMAL position.

One was just out of a heavily contaminated airstrip, where shortly after getting airborne, a loud bang was heard and the plane carried out a forced landing, where it overturned in the soft field and the aircraft was damaged beyond repair. Thankfully with no fire nor fatalities.

What was this runway contaminated with? Gravel? Water? Wildlife? I ask because I've seen a number of bird-strikes with the Caravan and every so often one goes right up the plenum. With the Inertial Separator in BYPASS mode, most birds pass right through as pink mist but on one occasion I've seen one get stuck inside near the gate. The aircraft landed with the separator in BYPASS without incident. Ever since I've made a habit of keeping the separator in BYPASS whenever I think the a/c has ingested something substantial. ie: Ice, bird, FOD... I will only return the separator to NORMAL after I shut it down on the ground. For all you know that object could still be lodged in the plenum and returning the separator back to NORMAL could dislodge it and possibly cause further damage or even suffocate the turbine. This isn't a hard rule and occasionally I need the extra power or cooler temps but it's a good habit to keep I think. In any case you said they inspected the engine and couldn't find anything so it doesn't sound like they hit any birds or debris?

At what point in the climb out are people returning the separator to NORMAL? I don't know anyone who can physically return the inertial separator to NORMAL without first reducing power to 1000 ft.lbs/tq first. Almost goes without saying but such a sharp reduction in power during the climb out while you're low, heavy, and hot is never a good idea.

[lilflyboy262]

I guess I should have been a little more clearer in the description of the seperator. Open = Bypass, Closed = Normal. Just being a little lazy in that description.

Not really sure what height they return it to Normal. Whenever they feel they are stable enough to do it I guess. Probably before they get above 100kts as that becomes pretty difficult. Anything above 120 starts to take a decent amount of strength.
As I said before, my company is 40kts on the ground roll.

The first runway was contaminated with large areas of surface water. Usually this is mixed with sand and a surface called calcrete.
Another theory was along the lines of what you said. Mud built up inside and once the seperator was returned to Normal, the mud dislodged and went through the engine. BUT no trace was found.

As for the second runway. It was in perfect condition. A little dusty but nothing untoward, certainly nothing to trouble the engine. Birdstrike is very much a possibility in this region. Most of the time though, it is more of an occurance on approach or on the climb out due to them circling on thermals at higher altitudes at that time of day.
I was in there perhaps 45 minutes before the accident and saw perhaps 2 or 3 vultures circling on a base leg but nothing around the runway.

[Carrier]

This topic has been explored before here and on other forums. Take a look at:

http://www.flyafrica.info/forums/showth ... ne+failure

http://www.pprune.org/african-aviation/ ... zania.html

http://www.pprune.org/african-aviation/ ... -tanz.html

http://www.pprune.org/caribbean-latin-a ... dents.html

http://www.pprune.org/biz-jets-ag-flyin ... c-12s.html

http://www.pprune.org/african-aviation/ ... bique.html

http://www.flyafrica.info/forums/showth ... outh+sudan

http://www.pprune.org/african-aviation/ ... ngala.html

http://www.pprune.org/dg-p-general-avia ... e-tsv.html

http://www.pprune.org/dg-p-general-avia ... -down.html

http://www.flyafrica.info/forums/showth ... ne+failure

The PC12 thread referred to by Panama Jack seems to have gone missing from Avcanada.

I stand by my comments in some of the threads. If the 12 persons had been in a twin such as a DHC6 then an engine failure after take off should have been no more than an inconvenience (& major expense to the owner) and eight more people would still be alive. If you use single engine aircraft at night, over water beyond gliding distance of land, in serious IMC or over inhospitable terrain then the chances of surviving an engine failure are very low. The above threads should give some idea of what the odds are.

[lilflyboy262]

A majority of the area is wide open salt pans, large areas of savannah or lagoons large enough to ditch in. There is no terrain to speak of. We only operate in VFR Day.
If you cannot operate single engine aircraft here, then they should be banned from aviation for good.
The reason why this second one was fatal was because it was probably the worst airstrip in the area where it could happen. The strip itself is completely surrounded by trees with the exception of one escape path out to the flood plain.

Crap runways here rule out most low wing planes limiting us to twin engine aircraft like the Twotter or the Islander. The Islander's second engine will only take you to the crash site in the event of an engine failure and the passenger loads cannot really justify a plane as large as the twotter.

Im not entirely sure of how many have crashed in total. I know there have been a few. But there is over 2000 of them out there, and have done something like 13 million flight hours. Its not entirely without its merits.

[KK7]

Most of my flying in the C208B has been with the inertial separator in BYPASS for the entire flight. It was open for take off, climb out, cruise and landing, due to low level operations and the very high likelihood of ingesting birds. The engine performs quite well with the inertial separator in either position. I will note that most of this flying has also been in African but on Canadian registered aircraft.

Due to the high probability of bird strikes and sand and dirt particles at our altitude, we inspected the inertial separator before and after flight as well as the inlet screen on the engine by using a mirror on a stick, and a flashlight. On more than one occasion I have found a bird stuck on one of the separator vanes. I am certain that if we had returned the inertial separator to NORMAL the bird would have been dislodged and either continued through out the back, or would have gone towards the engine inlet.

Now that being said, there is an inlet screen and I have seen quite a few things on that screen from feathers to multiple entire birds. I have also had small birds injected into the engine and we had very little indication that this had happened. Personally I would not touch the inertial separator during a take off roll, it should either be left in the BYPASS or NORMAL position. I would prefer to be at a safe altitude before changing it. You risk overtorquing the engine at take off power, dislodging something on the intake vane, and having it in BYPASS has its use in the lower levels too at least up to 500 feet. I think in your area where flocks of birds are prevalent, this is a wise precaution IMO.

I hope the cause is found in good time and I look forward to hearing what caused these two accidents.

[Caracrane]

Bypass is closed and Normal is open as far as i'm concerned. But Just be intelligent enough and not taking up to 14 people or so by their rules, you might have more room in a event of a failure. Have the same failure with the same amount of paxs on board a KA-100 and tell me about the result

[Redneck_pilot86]

Open and closed is a poor phrase for the separator. There is normal position (air being directed into the engine) and bypass (air being directed out the left rear of the cowl)

One thing that came to mind when reading this, specifically the first scenario, was a bleed valve failure. Thats what happened in this CADOR:

http://wwwapps.tc.gc.ca/Saf-Sec-Sur/2/c ... pe=0&narr=

[PanEuropean]

...One thing that came to mind when reading this, specifically the first scenario, was a bleed valve failure.

I don't think that a compressor bleed valve failure would cause an engine failure. At high power settings, the bleed valve is normally closed. If it were to open at high power, the engine would be unstable for a few moments (temperatures would spike up), then it would continue operating at a lower power output. Although that lower power output might not be enough to sustain level flight in a heavily loaded single, the result would probably be more of a controlled descent than a sudden forced landing.

Most catastrophic failures of PT6A series engines have been caused by compressor turbine failures. A CT failure would result in a total loss of power... and would also produce a loud bang.

Michael

[Doc]

Don't know how to explain the "bang", but otherwise they sound like they had their fuel supply somehow interrupted? Perhaps fuel contamination? I'm not buying into the bleed valve or intake contamination. I know deep grass can be a problem, but I can't see it where you operate, with the descriptions of "standing" water.
The "bang" sounds like a compressor stall...but not twice in that period of time. As is usually the case, witnesses tend to unintentionally exercise "artistic" license when describing aircraft accidents.
Methinks the fuel just stopped flowing for some reason.

[Go Juice]

I don't have a lot of turbine experience but from what I have heard, most PT6 failure are caused by the FCU going on "strike"


any truth to that?

[Cat Driver]

any truth to that?

I've had two.

[Go Juice]

Cat: from what you've whitnessed, before the fcu goes, does it show signs that it is going to or it's a total surprise?

Any specific parameters to look for? Fflow, torque ...


Thx

[Cat Driver]

When the fuel flow quits it is like shutting off a light switch....it just stops when the fire goes out.

There was no indication whatsoever that it was about to quit.

[Go Juice]

Yeah i figure but the hours before were there any fluel flow fluctuation? Weird reaction when you changed the power levers?

[Cat Driver]

The airplanes were Twin Otters operated by an airline on sked runs and at the time I was chief pilot of the airline and thus flew the airplanes on a random basis, I flew a lot as it was my way of ensuring the crews were flying to the standards that were required of them.

So to answer your question, no there were no indications that either airplane showed any indications of anything wrong with the way they ran.

Had any crew noticed anything wrong they would have snagged it because they knew that was what I demanded of them as pilots, and by doing so they would have improved their chances for advancement if I had the opportunity to do so.

Maybe that is part of the reason I am still around to irritate you guys here...because I always tried to be the best I could be and avoided grey areas in decision making.

It was a long time ago and my method of SMS was to fly with the crews and gain their trust by being one of them.....