Touch and go - my worst landing ever


I recently returned from Lijiang (in Yunnan, China). One of my legs was a flight from Chengdu to Beijing, and it is certainly a flight I will never forget.

After the plane took off in rather dense fog, and then leveled off, the pilot said that we will be cruising at 20,000 meters. Of course, this sounded rather strange, given that the highest I have ever been to was 46000ft (or 14km). It wasn't a good sign that the pilot didn't know at which height we're flying. He probably meant 12,000 meters or 20,000 ft, one of which is on the high side for this 2hr flight and the other on the low side.

During the flight, I received one of the most tasteless breakfasts I ever had, but this is not part of the story.

As we approached Beijing, the sky was clear, perhaps somewhat hazy. The plane started to descend. About 5 min before the landing (ehem... the first landing), there was just a little turbulence, but it quieted down. As we approached the runway, everything seemed fine.

The touchdown was, however, far from perfect. Upon touching the runway, the plane bounced perhaps 10 meters into the air. As it did so, it developed a significant roll to the right (so I could see the ground better...), and then touched again rather hard, with what appeared to be just one set of wheels. Immediately, the pilot increased to maximum thrust and ascended. During these 15 seconds or so, I was sure the plane had it, and that it would flip sideways (at least, that is what would happen to me if I were flying it on a flight simulator, but hey, I am not being payed to fly this thing!).

After taking off again (having touched the ground twice already...), the pilot circled the runway for a second approach, and for the 20min until the next landing, I was thinking on whether the landing gear could withstand such a shock without losing its structural integrity, and of course, what would happen to the plane if it did fall apart upon the 3rd touchdown.

Given that the pilot didn't prepare us for crash landing, nor that there were fire engines ready along the runaway, I guess that the pilot had full confidence that the Airbus engineers did a good job with the design of the landing gear, and indeed they did. I hate to think, perhaps the pilot knows from previous experience that the gear can withstand stronger shocks without a problem...

My suspicion was that the pilot approached the runway too fast, so he couldn't get rid of the plane's lift, which is why he bounced. And because of his roll, the bounce was asymmetrical, thus increasing the sideways roll even more.

As luck would have it, the first person I talked to afterward (on the shuttle between the terminals), was a french pilot trainer, training Chinese commercial pilots. He told me that it is very rare to have such landings, especially considering the very calm weather we had in Beijing. He said that sometimes, if the wind shifts while landing, this could also happen. When I told him that we landed on the same runway as the first attempt, he confirmed that it is was simply a very bad landing. But he also said that putting full thrust and taking off under such conditions is standard practice, since otherwise the plane can stall and crash (as I am well aware from my own flight simulator experience).

Incidentally, the Chinese use foreigner pilots as trainers to force their pilot students to practice their English. They also do that in inner Mongolia where there is little traffic and a lot of clear weather.

Another interesting aspect of this Lijiang trip I had, is that Lijiang airport is at about 2400m. The pressure at this height is lower than the normal cabin pressure in commercial flights (which is usually set to about 2000m). Thus, as we started our decent to Lijiang, the pilot had to reduce the cabin pressure, and not increase it! There was no need of course to equilibrate my middle ear pressure with the outside. (When the outside pressure drops, air leaves the middle ear quite easily, it is when the outside pressure increases that we have to push air in to it, or at least swallow a lot). This effect would be more pronounced if flying, for example, to Lhasa at over 4km.

And now a lesson for my pilot who obviously needs to work on his landing speeds.

How much larger should the landing speed be at Lhasa airport, as compared with sea level? (Actually, the pilot doesn't even know the speed at sea level).

The first thing we need is the air density at 4km. From meteorology, many know that the height of the p1=850mb level is about h1=1500m. Thus, at h2=4000m, the air density is roughly:
$$ {p_2 \over p_0} = \left(p_1 \over p_0\right)^{h_2/h_1} \sim 0.65 $$

This assumes of course that the atmosphere is isothermal with an exponential decay with a scale height h0: $ p = p_0 \exp (-h/h_0) $.

The second thing to know is that at low velocities (such that the mach number is unimportant), and low viscosities (such that the viscosity is not important), the pressure, and therefore the force, only enter the different equations through the combination of ρv2 (think of the Bernoulli equation for example).

This means that in order to get the same lift, one needs a velocity which is $ \sqrt{1/0.65} \sim 1.25\  $ times faster.

Furthermore, since the deceleration distance is x = v2/(2a), if the plane can decelerate with the same force (i.e., the same deceleration rate), then it will need a runway which is about 50% longer (or, a lighter load on the plane!). At Lijiang, it is only about 12% faster and 25% longer.

Now that my starbucks coffee at Beijing airport is finished, so is this post.

Addendum: During the subsequent flight I saw a nice gravity wave in the clouds. I'll probably write about it some day, but here is a picture of it anyway.


A gravity wave somewhere above the Ural mountains. Picture taken in the flight following the landing incident.


Type:

blog topic:

Comments

I may be reading too much into your story, but it seems likely to me that the underqualified, inexperienced co-pilot tried to land at the same speed that he had previously used at Lijiang and that the captain took command after the second bounce, grabbing the yoke and advancing the throttles. The most dangerous moment was the transition as the left seat said, "my airplane", and started to fly, not having flown into that situation. It may be just the reason you brought this up because it strikes me as a perfect metaphor for the circumstances surrounding the AGW debate. Underqualified, inexperienced but avid climatologists (co-pilot) use wrong analysis (Lijiang landing speed) to lead politicians to the brink of catastrophic policy (near crash), when better qualified scientists (captain) step in with correct expanation (recovery). Again, the point of greatest danger is the transition of control, which still lies in our future!

I learned to fly as an undergrad Physics major at Harvey Mudd College back in the 70's when they had an active extracurricular aeronautics program, and I remain an active instrument rated private pilot. As far as the pilot's instruments are concerned, the landing and takeoff speeds are always the same regardless of altitude. The reason is that the airspeed indicator measures pressure in an open tube (called the pitot tube) facing forward, a differential (the pitot pressure vs a static pressure source) Bourden tube arrangement calibrated into the desired unit, generally knots,. *Most* all of the aerodynamics affecting flight are the same at the same indicated airspeed (IAS).

At higher altitudes and higher than standard temperatures the true airpeed (TAS) can be much higher than the IAS but if one is on short final approach at 30% above the stall speed in the landing configuration (a common target), that IAS will be the same irrespective of altitude.

The problem is often the view out the window while planning an approach. If the pilot makes a visual approach without consciously adjusting the *distances* of the pattern flown based on density altitude and surface winds, the pilot will end up with more energy when they wanted with no good way to dissipate it and still land at a proper speed. During takeoffs, the most dangerous part of the flight, a pilot flying not by instrumentation but rather the sights and sensations of the takeoff roll may find themselves trying to will the plane into the air before speed sufficient for flight is gained, slowing the needed acceleration in the process, making the takeoff roll even longer.

Finally, the problem might not even have been initially caused by the pilot. Air traffic controllers sometimes, in their zeal to keep aircraft separated, make requests to pilots which interfere with an orderly slowdown and descent. Dealing with that, and when to decide the approach cannot be salvaged and a go around must be made, is something all pilots have to struggle with.

The equipment gives distorted values, but those which ensure that the place flights faster at higher altitudes!

As for your last point, I don't think that an air controller behavior is the reason for this landing. If it was, then the plane would not have rolled as it did...

Really not a nice thing to feel!

I had 3 landings since then, and I must admit that "flashbacks" passed through my mind each time.

OK, you did loose me a bit with the calculations but I am right there with you bouncing and turning in the plane.

I had a similar experience in an airport in Australia although the biggest difference being I guess is that it was really windy and while we were already involved in a bumpy ride the landing became an experience not to remember. I am not sure how I managed to get through the whole process without a new pair of trousers but I did. However I would have to say that just reading your account I think your experience would be far more frightening since everything was going well to just suddenly bounce at the runway would have probably enforced the extra clothing requirement.
I am not sure I want to read any more stories like this from you!

Just happened to me too today.. And not even all the way around the world but a short flight from Montreal to Toronto, landing at Toronto City Center (Island) airport.

Porter Air flies Q400 turbo props (nice airline and nice airport otherwise) but we had an uneven touch-down that however didn't seem overly bad, didn't seem overly heavy and didn't seem overly un-even. (Just a little, but I've had worse). Still the pilot decided to take off again and go around, so that's after touching down.

I was watching the first landing and I don't think we touched down too far up the runway.. maybe it was just that the landing was uneven, that the runway wasn't that long to begin with, and that there's icy water at the end of it. ;) It was also snowing a little bit, so probably a good call on the part of the pilot to take it back into the air and do it again.

Second attempt was perhaps a bit harder than usual but perhas the pilot just wanted to make sure it stayed down. :)

Conditions were light snow but not overly windy.