Sympathies for TransAsia Flight 235 Crash

what caused the stall, bad weather?

 

The first frame in the dashcam video makes me think that they lost the left engine on climb-out, but it could have been a lot of things.
The WX didn't look that bad, but it's hard to tell.

We'll see. They know where the plane is, and there appear to have been some survivors.
We'll know something soon, I'm sure.

Sympathies indeed for those involved!

 

I know that road (and that airport), and it all seems a bit odd if they were taking off from Songshan Airport: the plane looks to be heading back towards the airport (but looks like it would be going across the runway, if I'm reading all the pictures right). If I'm right - and I could be wrong, as it's a while since I've been there - then I'd guess something had already gone wrong and the pilots were attempting to turn back. It's hard to tell, though: a lot of the buildings around there look the same, and the river twists around the airport.

But they're much, much closer to the airport than Sullenberger was, and the plane was much lower. I doubt there'd have been the time to do the things he did. Unless they were at the final stages of a return to the airport, of course.

 

My guess is that the plane quickly lost power over the city, and the pilot headed for the waterway, but tried to bank sharply to avoid the roadway.

 

You could well be right. If so, it would be very difficult: that area is very built-up, and the Keelung River is pretty narrow.

 

Perhaps the most troubling aspect of this crash, for those not directly involved, is the horrifying ‘endgame’ video.


What we have so far is a probable left-engine failure soon after takeoff. The attitude (left wing very low) of the plane may have been from left wing stalling, or pilots’ extreme maneuvering, or a combination. When a plane is very near the ground, such attitudes are probably unrecoverable.


All commercial aircraft, this type included, are subject to rigorous certification. The airframes and powerplants must be good enough. The carrier’s airplane maintenance and pilot training must be good enough. We are far from knowing whether this carrier has dropped the ball, and this is not the point of my post.


Instead I want to talk about large turboprop twins as a category. These carry about 70 passengers, and some up to 100 pax may now exist, or are being planned. This category exists because turboprops have lower operating costs than twin jets. I really don’t know why that is so, as the turboprop powerplant is essentially a jet engine mounted backwards. Perhaps is it because relatively small jet engines make more thrust when attached to propellers, rather than just emitting hot gas to the rear. This is only my guess. But the category does exist, and ATR 72 is not the only example. Sometimes (rarely, I’d say), they get into trouble, and I want to talk about why.


The limit of my pilot training is single engines, and certainly many other pilots know more about this stuff. But I do know that for twins, loss of engine immediately after takeoff is the worst damn thing. Lots of training is directed at handling this situation. An immediate necessity is to identify which engine has failed, because that propeller needs to be ‘feathered’. This means changing the angle of attack of the propeller blades. If you don’t do it, aerodynamic drag on the ‘dead side’ will be high and work against pilots’ recovery efforts. At present there is only the slightest hint that these pilots may have feathered the wrong engine. But I have to say, the prop blades’ angles was the first thing I saw in the video. As in, ‘that looks wrong’.


Twin turboprops as a category handle many short-haul routes, and ATR 72 is one among several. I am in no position to criticize it as a bad bird. But is has characteristics that can be compared to others in the category. I choose to compare in this way. Each plane has a maximum takeoff weight, and each engine used has a maximum shaft horsepower. Without doubt, they do a fine job with both engines running, and have been approved for the dreaded ‘one engine out’ scenario. Yet they are not all the same.


In this category, the Saab 2000 has a max takeoff weight (MTOW) of 22,800 kg and a single-engine shaft horsepower (SHP) of 4,591. Saab SHP/MTOW is thus 0.20. For the Dash 8/400, that ratio is 0.17. For the Fokker 80, 0.11. For the ATR 72, also 0.11. For the Xian MA600 (perhaps not yet is service), it is 0.10.


Overall this ratio ranges by a factor of 2. The reason to go for a low ratio is, obviously, money. When two are engines running (the normal situation), both are drinking, and fuel has a cost.


Designing a plane that meets certification with the lowest possible operating cost is an obvious goal. Airframe designs are getting ‘cleaner’, so the Xian may not represent an endpoint. Save money on the average flight.


But dang. In rare situations when one engine fails after takeoff, the PAX on board depend on prompt application of pilots’ training and tools (power) available to them. In this category, it appears that Saab and Dash have an edge.


Would I make a flight choice based on this analysis? Have not in the past. Presented with choices in the future, I probably would.


In the current incident, both pilots had lots of time ‘in type’. There is no basis for faulting their responses so far. I only want to say that planes in this category differ in the ratio I defined above. If the airline industry continues to be driven by lower operating costs, we may not see fewer such failures. Such failures are rare, OK? Got it? Commercial air-transport fatality rates are lower than when you drive your car. So, don’t be afraid, but when you have choices, how would you make them?

 

Just to illuminate this portion a bit -- and I'm speaking only generically, not as an aircraft propulsion expert ---

Expelling a thin jet of hot high speed fluid is a relatively inefficient way to propel a vessel in a fluid medium. A fatter jet of cold low speed fluid will produce more thrust for a given fuel power.

The temperature part should be obvious, that is just waste energy. The speed part comes from the momentum and kinetic calculations of the jet flow. Thrust is a momentum concept, and is the jet's mass flow rate times its exhaust speed. (I'm ignoring the intake, which a real propulsion engineer can't do.) But the kinetic power needed to create that flow is the mass flow rate times the square of the exhaust speed. Thus, for fuel efficiency, it makes sense to shoot out a lot of mass at as low a speed as practical. Of course, the required speed of the vessel puts a lower bound on the useful exhaust speed.

In the original jet engines, all of the mass flow (air) went through the combustion chamber and came out as hot high speed gas. Later generations started using an oversized first compressor fan, and some of that air was allowed to bypass the rest of the engine. This bypass flow didn't get heated in the combustion chamber, and didn't get expelled at the same speed, but still contributed to thrust and improved efficiency.

In the most recent engines, bypass ratios are now reaching about 10:1. For every pound of air going through the combustion chamber, 10 pounds just go through the first fan, for thrust. That fan is now more of a ducted propeller than a compressor. The air that is burned is also expanded more efficiently, with more energy extracted, so that the hot gas is now jetted out cooler and slower than from the older engines. The extracted energy drives the increased load on the first fan or 'propeller'.

Turboprops fly slower than jets. This means that they can cut the jet exhaust speed even lower, and use a higher and more efficient effective 'bypass ratio'. This also makes the front 'fan' larger compared to the combustion core.

Fan? Propeller? Same idea. Little ones are usually ducted, big ones are generally open.

Actual jet engine design is far more complex than this.

 

i don't buy the 'flying is safer than driving' statistics. passengers on a plane are at the mercy of the pilots, weather and other factors. drivers are at the mercy of others on the road and weather, but these can be mitigated by defensive driving and common sense. aviation is a commercial industry, and the bottom line will always be the bottom line. passenger safety comes second, and reluctantly, at the behest of regulation.

 

I actually LOVE to fly!
Trouble is...
They never let me.
I have to be in the back with the other passengers.

If GOD had intended for humans to fly, we would not have an interstate highway system and rental cars with unlimited mileage.

I've had to fly many times in the past at the behest of my beloved government, but I concur with bisco.
If I'm able to I'll drive there, thank you very much!

 

The more clips I've seen, the more this appears to be the case. That part of Northern Taipei is extremely densely populated. We'll see what happens with the investigation, but I suspect the pilot saved thousands of lives.

 

Stats can't lie.

 

nope.

 

Just because you don't buy it, doesn't mean it isn't true. It's simply a statistic that there are fewer passengers killed per mile while flying than killed while driving.

Now, I will certainly agree that there is far, far less control (almost none) flying than driving, and that would certainly lead to more anxiety. And, on top of that, airplane accidents are certainly more fatal when they occur than car accidents due to the momentum of collision.

Nevertheless, flying is safer than driving.

Now if you want to say that your personal driving skills are such that YOU are personally safer driving than flying, there may be something to that. But you'd have to be one helluva fantastic driver to convince me.

 

that's what i meant. driving fatalities include all the nut jobs on the road. if there were a way to sort them out, i believe that you would find driving to be safer than flying. i agree that the stat is correct as far as it goes, but it only proves that there are less nut job pilots than drivers.

 

There's no control in most fatal auto accidents either. As a pilot, I can tell you that the vast majority of plane crashes are due to human error.

 

And the person driving reckless in and out of traffic at 1xx+ mph who wrecks out and kills himself and or others is not human error???


I can probably safely say most (fatal) car wrecks are due to human error, stupidity, or caused by someone else's human error,stupidity, or not paying attention or (fill in the blank), not so much mechanical issues, which are few and far between…

Sure I guess you can argue road conditions, but then at that point one could say your driving too fast for the conditions..

Of course I understand that there are sadly innocents every now and again (i.e.. driving down the road minding your own and a tree falls on you…etc)




May the Departed RIP, and may the wounded heal quick..

 

I witnessed a fatal accident when I was 17. The nut job didn't die. Just the two people in the car coming the other way. And two of the nut job's passengers.

 

From an irrational consumer point of view....

I'm not a nervous flier at all. But I always feel less at ease in a turboprop than I do in a jet.

Of those, though, I feel far less at ease in Saabs - which I have to fly in fairly often - than in ATR72s or Dash 8s. I have a reason to dislike Dash 8s - one of them almost hit my Dad once when it crashed at Farnborough, and there's some great TV footage of him running away from a flaming tyre. But there's a stretched Dash 8 that Qantas uses that I feel quite comfortable in. But the Saabs always feel like they're going to shake themselves apart. It's the plane that makes me the most nervous when I'm in it. I do try to avoid them whenever possible.

And I fly in ATR 72s a lot. In fact, I've flown in TransAsia ATR 72s quite a few times, as well as a lot of Bangkok Airways ones and a few others. As far as turboprops go, I've always found them the least scary.

With the exception of Taipei-Kaohsiung, pretty much every turboprop flight I take is on routes where, at one end or the other, the runway is too short, or the approach is too steep, for a jet, even a small jet like an Embraer ERJ - places like Dubbo and Luang Prabang and Koh Samui. So my guess would be that this is what drives the use of turboprops: the need to take off and land on short runways. I know this was always the Dash 8's selling point.

I'd be surprised if cost were a factor. Every trip I've taken on a turboprop cost me a lot more per mile than any trip I've taken on a jet. I suppose there are economies of scale involved, and in some cases monopolies on routes, but even so, my understanding is that jets are a fair bit cheaper per passenger mile, at least once you get beyond a certain distance. Maybe cost drives the use of turbojets on very short routes: Kinmen to Taipei isn't far at all.

 

Yes, the nut jobs do skew the traffic fatality rates. I would not be surprised to find a fairly wide gap between 'average' and 'median' fatality rate, with the median driver much safer than average. And the great drivers much better still. But the gap between cars and commercial airliners is so wide that I'd be astounded if even a 3-sigma driver can do better on the road than in the air. While the great majority of incidents are avoidable, a small fraction is not practically avoidable.

Note that the common flying safety claims are only for scheduled commercial airline flights. Charter flights, air taxis, and general aviation are rated separately and are not equally safe.

Here are some death rates per billion miles, derived from a 2013 NHTSA report --

Overall: 11 per billion vehicle miles
Motorcycles: 212 per billion passenger miles
Car and light truck occupants: 7.3 "
Trains, overall: 0.43 "
Trains, passengers only: 0.15 "
Buses: 0.11 " (excluding victims of crimes)
Commercial airlines: 0.07 "