Friday, August 5, 2016

A Hard Landing in Truckee

Last Sunday (July 31) I had a very hard landing while flying the NCSA Grob 103 (N3981C) at KTRK. Several members of the club have asked me what happened, and since I think my experience can help others avoid a similar situation, here is a description of what happened.

First, let's be clear:  cause of the event was pilot error.  I'd love to say otherwise, but I cannot.  I blew it.

The short version is that, despite being aware of the hazards and the weather conditions, and actively thinking about them throughout setting up the landing and the downwind leg, somewhere between turning base and landing I managed to let the airspeed bleed off so that my final approach was made without sufficient airspeed margin. The more detailed description is as follows:

The landing took place on runway 20 at approximately 2:20 pm, at the end of a flight of approximately 38 minutes duration. I was flying from the back seat; I had a passenger on her first flight in a small aircraft in the front seat.

The general conditions of the day were that the valley was filled with smoke from nearby fires, there was a southwesterly wind varying from 10 to 20 knots on the ground, and the air was somewhat turbulent aloft. The smoke affected the visibility, so that nearby mountains were visible in outline, but not detail, and from 10,000 feet directly over the airport the town of Truckee was easily located but hard to make out in detail; locally, however, (within several miles) visibility was fine and was no factor.

My passenger became somewhat nauseous as we climbed through 10,000 feet in bumpy thermals, and asked me to fly straight instead of turning; I accommodated her wish, flying a long back-and-forth pattern through the house thermal over the frog pond, losing a bit of altitude on each pass until I was down to 8300 feet or so, and informed her we would have to thermal or land. She asked me to land, so I headed back to the initial point.

The area around the North Star parking lot was solid lift, and I had to use full spoilers to drop to pattern altitude for entry on the 45. I closed the spoilers before I entered the pattern; realizing I wanted to preserve my altitude. While maneuvering toward the pattern entry I tuned in AWOS. Wind conditions were reported as “240 at 10 knots, gusting 21 knots”.  The wind was at 40 degrees to the runway heading, and I realized this meant the crosswind and parallel components would be similar in strength and would likely be somewhat less than 10 knots.  In a gust, though, the crosswind component could be noticeably higher, even as much as 16 kts or so.

I consciously applied the adage of adding “1/2 the wind speed” to the book approach speed of 55 knots.  I split the difference between wind and wind+gust and decided to add 7 knots to the approach speed, making my approach speed 62 knots. I also reminded myself to make a high approach and turn downwind-to-base “inside the spillway” (Martis Lake Reservoir has a spillway northeast of the approach end of runway 20; the plan was to make the turn between the end of the runway and the spillway.)

There was no traffic in the pattern; a tow plane and glider departed from runway 20 as I entered the downwind leg and was no factor, and a general aviation aircraft was at the threshold of runway 29; they radioed to me that they would hold their takeoff roll until I either crossed runway 29 or rolled to a stop on runway 20; again, they were no factor.

I made several radio calls: that I was about to enter the pattern, upon entering the 45, upon turning downwind, upon turning base, and upon turning final. It was just after I called the turn to base that the general aviation aircraft holding on 29 called me and we had our brief conversation.

I made my turn to base leg inside the spillway, as planned**. I opened my spoilers on the base leg, but as I turned final I felt I was getting low too fast and closed them. I felt my speed was ok*, but I do not recall actually checking it at that time.

Then things got dramatic. As I crossed the cliff at the end of runway 20 I encountered either a sudden lull in the wind or a rotor over the cliff’s edge; whichever it was, the practical effect was sudden loss of airspeed and, without sufficient airspeed margin, a loss of lift. The aircraft dropped “like a stone.” It was a sickening feeling to sense it dropping so fast and realizing I had essentially no control over what it would do. 

But I knew what was obviously what was going to happen- a hard landing was in my very near future!

We landed very hard on the main wheel, at or just barely beyond chevrons marking the displaced threshold. I rolled to a stop on the apron where the tow planes normally park. The landing was hard, but I did not feel the tail wheel hit, and I do not think it did. Nor do I think the nose wheel hit; it felt like I came down squarely on the main wheel. I do remember being a bit surprised that we did not bounce back into the air noticeably.

So how did this happen? I believe the cause of the hard landing is that I allowed my airspeed to bleed off during the base leg and the turn to final, and hence did not have sufficient margin on the airspeed to compensate for any gusts, lulls, or wind shear that I might encounter (and that, obviously, I did encounter). This was exacerbated because, despite my best intentions, I did not come in on a high, steep approach. I allowed myself to come in too slowly, and lower than I had planned.

As I said when I started: pilot error.

=======

I wrote all the words above Monday, while the incident was still fresh in my mind. Tuesday morning I carefully examined my IGC file using OLC, skylines.aero, and Doarama, and discovered, to my surprise, that:

* 1) I was flying even more slowly than I had recalled on my final approach. The GPS ground speed on my downwind leg was 70kts, on base leg 62 kts, and on the final approach 45 knots. If I assume 10 kts of wind at 240 degrees the airspeed values become more like 63 kts on downwind (essentially as I had planned) but only 52 knots through most of the final leg. I also saw that I allowed the speed to further bleed off to 38 kts (GPS ground speed) just at the threshold. If the wind were still at 10 kts that would have been 45 kts of air speed, barely above stall speed, as I crossed the threshold about 50 feet above the ground. If I hit a lull (or the wind was blocked by the trees, which is what I think actually happened) I would have been at or below stall speed! That is what I believe occurred; I stalled the aircraft while still well above the ground, and down we went. I still had forward momentum, of course, so still rolled out along the runway after the landing.

**2) moreover, although I started my turn to base "inside the spillway," I did not crab and I let the wind blow me into a rounded base leg rather than a square one, drifting downwind so the base leg was actually outside, rather than inside the spillway, making my final leg somewhat longer than I planned.

3) The analysis confirmed something I had realized just after the incident, namely that I should have flown the entire pattern higher, and that I had lost sufficient altitude on the downwind leg that I should have made my base turn considerably earlier, probably when I was even with the numbers. I had actually made two turns between NorthStar and the Initial Point (the so-called Lone Pine) specifically to burn off altitude with the goal of entering the pattern around 7400 feet, 1500 feet above ground. But I made the turns too far out from the IP and was in sink from where I made the turns all the way to the IP, so that when I reached the turn to downwind I was already down to 6800 feet, just 900 above terrain! I did observe during the downwind leg that I was lower than I wanted to be; however, for some reason, I did not correct for it appropriately with an earlier turn to base.

In short, I was too low and too slow and I did not recognize and correct for those facts.

= = =

So, what is my corrective plan for the future?

First, whenever I fly in future I will strive to be more attentive to my landing plan, making sure I am following it and also making sure that it is working; I’ll be much quicker to make an adjustment, such as making the base turn sooner.

That's in general.  A more specific corrective plan involves what I will change about flying at Truckee. I talked with Mike Mayo shortly after the incident – when I told him my airspeed plan had been 62 kts (approach plus half the wind), he observed that the "approach speed" in the POH was selected with nice, smooth-air, open field German landing strips primarily in mind, and not for the kind of roiling conditions commonly found in Truckee (and other gliderports in the Western USA) and that his approach at Truckee is to always come in at least at 70 kts, if not 80 kts, precisely because of the rotor and tree-line effect on runway 20. As he pointed out, runway 20 is a long runway, and coming in with plenty of airspeed at worst means the ground crew has to drive the golf cart a long way to get you. His comments make sense to me, so I will elevate my basic approach speed for KTRK from 55 to 65 in a G103 and then add for the wind speed from there. This would have made my target speed for this landing 72 kts. I will also follow my plan of entering the pattern at a higher altitude than I did. In conversation with Larry this afternoon, he observed that he likes to hit the Lone Pine IP at least 1500 feet AGL; in fact, that had been my plan, but I actually was 600 feet lower than that when arrived at the IP. I think, with the long runway to use, a higher, steeper, faster approach has now become my normal plan for Truckee, with modifications for specific conditions as necessary.

====

Finally, there is a very frustrating side note to all this. I was well aware of Bill Gawthrop’s crash at Truckee two years ago. Like many pilots, I have studied his trace and I have read with interest not only his description of the event but also the lengthy discussions and analyses of his accident on rec.aviation.soaring. I was very much aware of the mistakes Bill made, and I actually thought about them as I was setting up for the landing pattern.

And yet, I still came very, very close to replicating his experience.

Having a plan is only useful if you follow it while it is working, recognize when it isn't, and take decisive remedial action when necessary.

V/R,
Van Henson

You can see the trace of the flight at Skylines.aero, at:
https://skylines.aero/flights/66931/

I've also put together a Doarama visualization of the last few minutes of the flight only, which may be seen at:
http://doarama.com/view/872321

I also posted it on OLC, so another trace can be seen at:
http://www.onlinecontest.org/olc-2.0/gliding/flightinfo.html?flightId=1234794473

No comments: