Aerospace Blog Blog

Aerospace Blog

The Aerospace Blog is the place for conversation and discussion about aeronautics, astronautics, fixed-wing aircraft, future space travel, satellites, NASA, and much more.

Previous in Blog: Stealth for Sale: Boeing vs. Lockheed-Martin   Next in Blog: Space Age Meets Internet Age
Close
Close
Close
10 comments
Rate Comments: Nested

FedEx MD-11 Crash

Posted March 27, 2009 11:30 AM by masu

A few days back a McDonald Douglas MD-11 operating as FedEx Flight 80 crashed at Tokyo's Narita International Airport and resulted in the tragic deaths of both pilots.

So, what went wrong?

In order to understand what went wrong we need to cover how an aircraft is landed and surprise, surprise, it's a little more complex than one might think. First off is a list of the relevant technical terms and a diagram of the approach profile of an aircraft shown as the green dashed line and the runway from above that shows the markings.

  • VS or Velocity Stall This is the speed at which the wings stall or cease creating lift. It usually results in a sudden drop of the nose and often results in the aircraft going into a spin that can be difficult to recover from.
  • Threashold This is the beginning of the landing strip and is usually designated by parallel lines running across the runway.
  • Aiming Point This is the point that the pilot aims the aircraft at during the approach. It is often marked on the runway by two large rectangles on either side of the runway. It can also be designated by a series of lights on either side of the runway called a Visual Approach System or Precision Approach System.

The approach and landing are broken up into five segments:

Glide Slope: Initially the aircraft is set up on the Glide Slope which is a straight line that terminates in what is called the aiming point that is usually designated by the wide set of lines on either side of the runway. The way you maintain the glide slope varies between aircraft types but surprisingly large jets and gliders do this in very much the same way. With the jet the engines are set to a specific glide slope power output and then as with the glider the angle of the glide slope is maintained using air breaks. Before the word police jump in the spelling of break is deliberate. Air breaks don't actually slow the aircraft down, they disrupt or break up the flow of air over the wing and therefore reduce the lift the wings are creating. The resultant reduction in lift increases the rate of descent and therefore increases the angle of the glide slope. From the pilots point of view maintaining a constant glide slope is fairly easy to do and is achieved by keeping the aiming point in the same relative position throughout the entire approach procedure.

Round Out: This is where things start to get complex as if the aircraft were to follow the glide slope all the way to the ground the impact due to the vertical speed would be too great for the landing gear and airframe to tolerate. This would result in the wheels and undercarriage being forced up through the wings and fuselage. Instead when the aircraft is about a wing span above the ground the pilot pulls back on the controls to arrest the vertical speed and set up the aircraft so it is flying horizontally. The exact timing varies from aircraft to aircraft but it is about this point that the engine power is reduced to idle.

Flare of Hold Off: This is probably the most important and difficult part of the whole procedure. At this stage the aircraft should be flying horizontally or parallel to the runway with the wheels a meter or so off the ground. Since the throttles have been reduced to idle the airspeed will start to wash off. Inorder to maintain the horizontal flight path the pilot gently pulls back on the controls. This gently raises the nose and increases the angle of attack on the wings which increases the lift and enables the pilot to maintain the horizontal flight path. This takes some practice to acchieve and needs to be a smoot increase in the attidude of the nose but every pilot alive has practiced this hundreds of times and should be able to do it without thinking.

Touchdown: Eventually the reducing airspeed and increasing angle of attack reach a point called VS where the wings can no longer create lift ant the wings will stall. This results in a sudden and dramatic reduction in the lift they are creating and the aircraft drops the last metre or so and the wheels contact the runway.

Roll Out: The aircraft's weight is now completely supported by the wheels and the aircraft is no longer flying. The pilots then use various braking systems like spoilers, wheel brakes, reverse thrust to slow the aircraft to a speed that it can safely be manoeuvred using the undercarriage rather than flight controls.

So, why such a convoluted technique? Why allow the aircraft to drop a metre or two onto the runway, surely a smoother approach would be to gently fly the aircraft onto the runway?

There's an very good reason for this procedure and it has to do with getting the aircraft to change from being an aircraft to a land craft in as little time as possible.

If you were to fly the aircraft onto the runway it would still be going fast enough to fly or faster than VS. If for any reason the nose wheel where to rise due to bouncing, hitting a bump, centre line marker etcetera the resultant raising of the nose would increase the angle of attack on the wings. Since the aircraft is still travelling faster than VS this would increase the amount of lift the wings create and as a result the aircraft would start to fly again and gain altitude. Since the engines are at idle this sudden increase in altitude washes off airspeed until it drops to VS at which time the wings stop creating lift. The nose of the aircraft then drops and the aircraft impacts the ground nose first with a vertical speed that is far greater than the airframe is capable of sustaining. The end result is a less than desirable landing and usually involves the destruction of the aircraft and those onboard.

Now that we have a basic working knowledge of how aircraft are landed we can apply it to the video of the FedEx Flight 80 crash. The three images below were taken from the video of the accident and show the aircraft slightly before, during and slightly after the aircraft first touched down.

If you look at the angle the fuselage makes to the ground the images show that it is increasing across the three images. This is exactly what we would expect to see during the round out, however, as is clearly shown by the puff of whit smoke in the middle and cloud of smoke in the right image the aircraft has touched down during the round out.

This is where things start to go dramatically wrong as it means the aircraft has touched down while still travelling faster than VS. This means that it will start flying again if the nose of the aircraft rises.

The next set of images show the aircraft shortly after it first touched down.

Unfortunately due to the tail of an aircraft in the foreground you cant see why but the sequence shows that for some reason the nose of the aircraft starts to rise. This means that since the aircraft landed and is still travelling faster than VS the increased angle of attack on the wings will create more lift and cause the aircraft to become airborne again.

The final series covers the time slightly before the second and major impact with the runway.

The first image of this series shows how the aircraft has become airborne and is now close to the angle at which the wings will stall and corresponds with VS. The second image is slightly later and shows how the nose has dropped dramatically due to the airspeed falling below VS and the wings stalling. The aircraft is now travelling too slow for the wings to fly and as a result the normal flight controls are ineffective. Since the aircraft is airborne it means that the normal ground handling controls are also useless and as a result the aircraft is completely out of control. The final image shows the second time the aircraft impacts the runway, but this time nose wheel first and with a vertical speed way in excess of the capacity of the landing gear and airframe.

We can now see what has happened, but like many answers it raises the following questions:

1. Why did the aircraft touch down with excess speed?

2. How could experienced pilots make such a mistake or misjudgement?

3. Since Narita International Airport is equipped with automated landing facilities why were the pilots flying the aircraft manually?

4. Was there some sort of mechanical failure in the aircraft that caused the aircraft to touch down prematurely?

5. Was there some sort of weather phenomenon that caused the aircraft to touchdown prematurely?

These are the questions the accident investigators must answer and it's beyond what can be achieved by just watching the video.

However, the meteorological conditions could have been a contributing factor and the pilots from aircraft that landed immediately prior to the accident reported encountering wind shear at around 2,000 feet.

The aircraft landed on runway 34L which means that the runway was aligned with a magnetic heading of 340°. The wind at the ground was coming from 320° so it was almost straight down the runway, but it was fairly strong at 26 knots (48 km/hr 30 m/hr) with gusts to 40 knots (74 km/hr 46 m/hr). This could definitely be a contributing factor as it could have caused a sudden drop in their airspeed of 14 knots which at approach speed could be devastating. However, pilots should allow for this when preparing to land by increasing the speed that they fly along the glide slope. This does make the round out and hold of more difficult to achieve, but atain, pilots are trained to do this and it shouldn't be a problem.

A final point that's worth mentioning is that FedEx flight 14 had a very similar accident at Anchorage International Airport on 31st July 1996, but fortunately all those on board escaped with only minor injuries.

This article is in no way meant to be a definitive answer to why the aircraft crashed but rather give the readers the background needed to understand what is happening in the video and stimulate discussion.

Reply

Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Guru

Join Date: Mar 2007
Location: Etherville
Posts: 12362
Good Answers: 115
#1

Re: FedEx MD-11 Crash

03/27/2009 4:33 PM

That's a very detailed and informative post, masu.

Narita seems to be known for difficult landing conditions. I don't have the knowledge to comment on this topic, but I did find some footage illustrating the type of problems pilots can encounter at Narita. Would you outline the circumstances when the crew would take over and land manually - specifically, I'm wondering that given the conditions at Narita, is it almost routine for pilot's to take control when unexpected wind shear is frequently encountered. Or does it work the other way, and automated landing systems are better at safely landing the aircraft when unexpected wind conditions occur ?

Completely out of my depth,

Kris

__________________
For sale - Signature space. Apply on self addressed postcard..
Reply
Guru

Join Date: Oct 2007
Location: Transcendia
Posts: 2963
Good Answers: 93
#4
In reply to #1

Re: FedEx MD-11 Crash

03/27/2009 11:04 PM

Thanks a lot for jumping in.

Hopefully a "Heavy" pilot will weigh in at sometime. In such cases we are glad to accept anonymous or "Guest" postings.

I myself if I was an investigator would be looking at CG, or center of gravity loading.

Typically from what I know landings are done by hand unless no other option is left, such as in out of fuel, and no alternate available, and no visibility.

I feel sympathy for the friends and families of the lost pilots.

__________________
You don't get wise because you got old, you get old because you were wise.
Reply
Guru

Join Date: Mar 2007
Location: Etherville
Posts: 12362
Good Answers: 115
#7
In reply to #4

Re: FedEx MD-11 Crash

03/28/2009 4:02 AM

The MD-11 autopilot has been found to be faulty before. Flying aircraft with know faults is nothing new. For example,this happened with Nimrod. Currently, taking a ride in a 777 is a bit of a gamble unless you know what type of engine is fitted - some have Rolls Royce engines with a known fault, and others don't (General Electric, I think). The type of engine on any particular flight is not advertised on the ticket.

In this particular incident, masu raises the question of why were the pilots in control. That seems to imply autopilot is the norm for landing. Although the incident I linked above was not related to landing, maybe it hints at a wider problem of flying with known faults that may be relevant. The failure mode of the autopilot (fail passive/ fail operational) may be relevant, but again, I'm out of my depth !

__________________
For sale - Signature space. Apply on self addressed postcard..
Reply
Guru
United States - Member - USA! Hobbies - Musician - Sound Man Engineering Fields - Mechanical Engineering - More than a Hobby Technical Fields - Technical Writing - New Member

Join Date: Oct 2008
Location: City of Roses.
Posts: 2056
Good Answers: 101
#2

Re: FedEx MD-11 Crash

03/27/2009 6:36 PM

Enjoyed the post. Thanks!

__________________
Don't believe everything you read on the Internet!
Reply
Anonymous Poster
#3

Re: FedEx MD-11 Crash

03/27/2009 10:47 PM

Looks like the fairly well known phenomena of Pilot Induced Oscillation, or PIO, to me. This situation occurs when the pilot reactions are out of phase with the aircraft motion, and thus the aircraft movements are exaggerated by the pilot control inputs. Even experienced pilots could be subject to this when fatigue sets in, along with an added complication of excess airspeed to compensate for the windshear.

Reply
Guru
Popular Science - Cosmology - New Member

Join Date: Feb 2009
Location: Halcottsville, NY
Posts: 665
Good Answers: 16
#5

Re: FedEx MD-11 Crash

03/28/2009 2:10 AM

It never ceases to amaze me when I can walk away from an aircraft after a landing. As a sailor, I have watched while the laminer flow has been interupted over an air foil, and the subsiquent collapse of the sail has led to some pretty hairy results, especially in a 'round-up', where a few degrees of wind shift can cause loss of control. Pilot error is often blamed in place of all the myriad variables of CE, CB, CLR, etc.. Since the winds are known to be gusty, it wouldn't surprise me that the pilot did exactly the right thing at the moment, but got blind sided by another shift, while momentum carried the aircraft beyond the point of control. It happens to everyone who has ever taken the helm.

Damn glad I don't have to do that for a living!!

__________________
De gustibus non est dispudandum.
Reply
Guru

Join Date: Dec 2006
Location: Germany 49° 26' N, 7° 46' O
Posts: 1950
Good Answers: 109
#6

Re: FedEx MD-11 Crash

03/28/2009 3:46 AM

Good Post

RHABE

Reply
Guru
Popular Science - Weaponology - New Member Netherlands - Member - New Member Fans of Old Computers - Commodore 64 - New Member

Join Date: Sep 2007
Location: Japan
Posts: 2703
Good Answers: 38
#8

Re: FedEx MD-11 Crash

03/29/2009 9:34 PM

It was extremely windy that day all over the country i guess that has also to do something to with, especially the way the plane went back up again looked unnatural

__________________
From the Movie "The Big Lebowski" Don't pee on the carpet man!
Reply
Guru
Australia - Member - New Member Fans of Old Computers - H316 - New Member Hobbies - Model Rocketry - New Member

Join Date: Jun 2006
Location: Port Noarlunga, South Australia, AUSTRALIA (South of Adelaide)
Posts: 3048
Good Answers: 75
#9

Re: FedEx MD-11 Crash

03/30/2009 8:25 AM

G'day fellow frustrated ground dwelling pilots,

I had a look at the Aviation Safety Network's records for Narita and according to that this is the only incident/accident that may be a direct result of the local meteorological or airfield conditions. Mind you, the record only covers the most serious events so it's possible that the conditions may be causing a whole stack of minor incidents that to date the pilots/aircraft have been able to dig themselves out of prior to digging themselves into a big hole in the ground.

I also read the write up from the link Kris supplied in post #1 which pretty much says what I have. However, I must admit that I don't totally agree with everything, particularly the following statement:

  • There are no circumstances under which a pilot of any type should deliberately select a nose-down attitude at that point - if, indeed, pilot selection of the nose-down attitude is what actually happened. During ballooning following touchdown the nose MUST be held up (if the elevator authority allows it) and appropriate power applied, whether the crew are trying for a successful second touchdown or for a go-around.

While you should never let the aircraft get into a nose down attitude during the final stages of the landing you may very well need to give the controls a considerable amount of forward movement. This is particularly so if the aircraft balloons and approaches a stall as it appears to have happened in this case. The only way to prevent a stall is to move the control column forward and reduce the angle of attack. Also the use of throttle to control the situation is unlikely to help particularly with jet engines which are notorious for throttle lag.

At a fundamental level jet engines are like a really big double or triple stage turbocharger that give you so much compression that the pistons are not longer necessary. Unfortunately and like turbocharged engines there can be a considerable turbo lag or delay in the response of the engine to throttle variations. As a result it's very unlikely that an increase in the engine power would be available before the aircraft had impacted with the ground and started to break apart.

Throttle lag can be a real nightmare, especially for pilots trying to land on aircraft carriers. In the event they miss the arrestor cables there would be no time for the engines to respond to an increase in the power so the pilots hit full go round power between 5 and 10 seconds before they reach the flight deck. It means that the arrestor system had to work harder as it not only needs to stop the aircraft but also overcome the thrust the engine/s are producing. However, it's a darn side better than loosing a very expensive aircraft and it's crew every tenth landing or so.

Also in post #1 Kris asked

  • I'm wondering that given the conditions at Narita, is it almost routine for pilot's to take control when unexpected wind shear is frequently encountered. Or does it work the other way, and automated landing systems are better at safely landing the aircraft when unexpected wind conditions occur ?

Personally I have never flown a large multi engine airliner, but from what I have been told the automatic landing system is perfectly capable of and possibly more capable than a human pilot at coping with out of the ordinary conditions. I did see a video of the system in operation some time back and it was so accurate at controlling the aircraft that the nose wheels were making thud sounds as they ran over the slightly protruding runway centreline lights. The only time I have ever been able to do that myself was flying a Piper PA28-180 with perfect conditions and absolutely no wind whatsoever. Even then I only managed to hit a couple of the little bugger not all of them down the length of the runway.

In post #4 Transcendian wrote:

  • I myself if I was an investigator would be looking at CG, or center (sic) of gravity loading.

Centre of gravity could definitely give results similar to this, however, such mistakes are extremely rare and would more than likely show up on takeoff rather than landing. There is also an inbuilt system that uses the load on each of the undercarriage legs to calculate if the CG is within specs and aggress with the manual loading calculations. As a result it would be very unlikely that they would have a problem. But then again, stranger things have happened.

  • Typically from what I know landings are done by hand unless no other option is left, such as in out of fuel, and no alternate available, and no visibility.

I managed to find a couple of YouTube videos that show the landing of MD-11 that are worth looking at:

  • MD-11 Crosswind Landing Video: This is a CG rendition from the pilots seat (the command seat in aircraft is usually the left seat except for helicopters where the command seat is on the right) showing an automated landing with a 36-52 knot wind at 324°. When this is combined with the runway's alignment of 235° gives a relative angle of 89° or pretty much directly across the runway. In the simulation the autopilot flies the aircraft right to the ground with the only input from the pilots being the application of reverse thrust and is only disconnected at the very end when they need to turn off the runway.
  • MD-11 Autoland 19R Schipol: This is a real landing with insets that show the captain, first officer, aircraft from outside and central console. This is almost an exact reproduction of the CG rendition with the exception of the cross wind component.

Landing an aircraft with a 36-52 knot cross wind is something that pilots would only ever do if there was absolutely no other option available. However, it shows that at least in theory the aircraft is capable of landing the aircraft even with over the top conditions.

Getting slightly off topic for a moment, but there are a few things worth looking at in the videos:

  • Aiming Point: The aiming point is designated by the large white rectangles painted on either side of the runway centre line. Since the glide slope has a constant angle the aiming point appears to be stationary relative to the pilots point of view. It just gets bigger as you get closer while staying in pretty much the same place.
  • Round Out: When the aircraft crosses the threshold the autopilot raises the nose slightly in order to arrest the sink rate or vertical speed of the aircraft.
  • Holt Off or Flair: The aircraft's slightly nose high attitude increases slightly during the hold off. As the aircraft reaches the next set of white runway markers the wings cease creating lift and the aircraft's main wheels contact the runway.
  • TVAS Indicators: To the left of the aiming point you can see a series of lights four white lights that form a line perpendicular to the runway. This is a visual aid that pilots use to ensure that they maintain the correct glide slope. The image below shows how the TVAS system works.

While they are on the glide slope the lights will appear as they do in the videos with all the lights forming a single line of white lights. (bottom frame)

The system has a further six lights that form either a standard or inverted T dependant on whether you are above or below the glide slope. The remaining frames in the image show how the system appears depending on how far above or below the glide slope the aircraft is.

For those that are interested in aviation and flying the image on the right came from a publication called the Visual Flight Guide. The VFG contains everything a pilot needs to know about the operation and tasks pilots need when operating aircraft in what is referred to as visual meteorological conditions.

You can download a copy of the VFG from the Australian Civil Aviation Safety Authority's web site.

Regards, masu

__________________
An elephant is a mouse built to government specifications.
Reply
Anonymous Poster
#10
In reply to #9

Re: FedEx MD-11 Crash

03/30/2009 10:16 AM

Many thanks, masu.

That's a fantastically good post (content & style), and is much appreciated.

Cheers,

Kris

Reply
Reply to Blog Entry 10 comments
Copy to Clipboard

Users who posted comments:

Anonymous Poster (2); Epke (1); Kris (2); masu (1); RHABE (1); RVZ717 (1); Tippycanoe (1); Transcendian (1)

Previous in Blog: Stealth for Sale: Boeing vs. Lockheed-Martin   Next in Blog: Space Age Meets Internet Age

Advertisement