What's Wrong with this Picture (MSP "Hint of the Month") ?

I often marvel at the fact that this hobby has now kept me interested for almost 12 years. The reason is simple: I can make it harder faster than I can learn how to do it right, so I am never satsified with my performance.

The flight I did this afternoon is a case in point. I have been "working" on a new Cat V schedule for the MSP Hub for a couple of years now. As part of this exercise, I have convinced myself that I need to fly all the new flights I intend to add to the schedule. I do this by giving them "experiment" flight numbers, like 15X27. The 1 designates the MSP Hub, the 5 designates a Cat V flight and the X means developmental. So today I did number 27 in this series, a flight from Gate H7 at KMSP to Dallas Fort Worth (KDFW). I started work on this flight several weeks ago, but for various reasons, I wasn't able to conplete it end to end until today.

As always, I planned the flight carefully. I started with a real world flight plan for a B738 obtained from www.flightaware.com. I then obtained the current weather for KMSP and the forecast weather for KDFW and built a flight plan incorporating the appropriate runways and the (associated) SIDs and STARs. So, since the wind at KDFW was forecast for the time of my arrival to be at 4 kts from the north, I selected a northbound runway, and since I like a short taxi to the gate, I selected Rwy 36R which promised to bring me off the runway close to Terminal B, where I choose gate B11. I put all of these selections in my ATC program (Radar Contact) so that I wouldn't have to get into a tussle with the controllers to get just what I wanted. I then "filed" a flight plan which included the BYP6.TUL STAR for Runway 36R, including a few additional waypoints that would take be from the DIRKK waypoint south of the airport and guide me smoothly into the initial fix on the 36R IAP.

What could go wrong you might ask. Well nothing...until ATC started me on my descent into DFW. As always, they started me down well before the 737 flight computer would have me start down, but I'm used to that. It means that I will burn more fuel than forecast, but I haven't heard a peep from SPA Operations about that, so I figured: what the heck. I go along with this because I'm trying to keep Radar Contact on board for the whole flight, no matter how inappropriate some of their instructions might be. I'm not sure why I do this, but it's probably because I just love those "attaboys" the program gives me at the end of the flight when I've done everything right.
Anyhow, I figured, they may be planning a "near side" approach for me (i.e. from the north), which would require me to get low and slow sooner than the "far side" approach I had so carefully planned.

Sure enough, when I got close enough to get the DFW ATIS broadcast, I learned that the wind was from the south and the airport was landing and departing on Rwy 17L. Drat, the forecast must have been bad. So, I set about reprogramming the FMC (flight computer) for a17L approach. This took some time, as it turns out that I was past the point of no return for the BYP6.TUL approach to 17L, so I had to do a good deal of cutting and pasting to build a profile to get me into 17L. While I was feverishly punching data into the FMC, we were descending into DFW and I was above the vertical profile and too hot (fast). Right, ATC tells me I have missed my descent point and vectors me off to lose altitude. OK, that gives me more time to reset the ILS frequencies in both recievers and reset the course setting on both sides of the cockpit for the "FAT" of 17L (176 deg.). Doing this will allow me to engage both autopilots below 1500 AGL and employ an "autoland" wherein the computer flares the aircraft, deploys the spoilers and applies the brakes.

About this time I start hearing on the radio that the wind is now coming from the north (as forecast) and aircraft are being told to land on Rwys 35 and 36. Drat again, but too late to change, so I request Rwy 17L and get permission to land southbound. This should work, because the wind is not too great.

Everything looks great for 17L. I'm down to the vertical profile and about to intersect the ILS. I go to my traffic program and clear the approach of AI aircraft so that I won't get a waveoff from ATC and I start telling my FS2Crew copilot to lower the flaps, while the computer steps down the airspeed automatically using the autothrottle.

We make the turn onto the final approach course for the ILS well below the glideslope and I engage the VOR/LOC and then the GS. After that I engage the second autopilot and look to see if it sticks. It does, so we're all set up for the autoland and this drama is going to end happily.

That brings us to the cockpit picture I posted in the Forum, with the title: "Whats Wrong with this Picture?". You can see that we're all set to go. I have just commanded flaps 15 and put the gear down. We are on the localizer and glideslope in the "LAND 3" mode with flare and rollout both armed for our landing. The spoilers are armed and everything is cool. Except - take a look at the speed tape. It says we're at flaps 5 and 172 kts. As you can see from the green numbers on the bottom of the speed tape, we are supposed to be landing at Flaps 30 and 146 kts. What's happening?

Now look up on the MCP (mode control panel) and note that 169 kts is illuminated in the IAS window. This is the commanded speed for the autothrottle. Ordinarily, when the FMC is controlling the autothrottle, this display is blank. It's lit now because the FMC has relinquished control of the autopilot to the ILS beacon and relinquished conrol of the autothrottle to the pilot. Whoops!.. this is why we're still at flaps 5 and 172 kts. In this mode the pilot (that's me) has to change the autothrottle setting to slow the plane and deploy the flaps sequentially.

This is what happens when you get "behind the plane". All that time I spent reprogramming the FMC left me trying to catch up with events in the cockpit. At this point I should have awakened to what was happening and called for a missed approach. Instead, I thought the aircraft was slowing and elected to continue the landing. The autoland worked, but we touched down at about 168 kts. That might have been OK save for two things. First of all, the autobrakes were set at 2 for a nice gradual stop, easy on the paying passengers. In this situation you want the autobrakes at maximum, or better yet, full manual braking. I didn't do either. We still would have been OK if we were landing on 36R, as originally planned. That runway is 13,400' long. Unfortunately, 17R is merely 8500' long.

What happened? I got the aircraft stopped using reverse thrust and full manual braking (after I understood the gravity of the problem), but not before I overran the runway and went a couple of hundred feet into the grassy overrun. Not good. Fortunately those sarcastic guys at FSDO (Radar Contact) didn't hear about it and I got a "good except for..." rating. Not what I wanted, but better than the sarcastic tone they use when I really screw up.

So, I'm still learning and there's a long way to go. Lest you think that these lessons don't apply to the "RW" environment, read up on the Asiana 214 crash on July 6, 2013 at KSFO or the AF447 crash into the South Atlantic on May 27, 2011. In both cases the accidents occurred because the pilots failed to understand that their automated flight system was in a different "mode" than they assumed. In the case of the Asiana crash, the pilots failed to recognize that the autothrottle was disengaged.<br /><br /><!-- editby --><br /><br /><em>edited by: Westcoast, Aug 30, 2014 - 08:26 PM</em><!-- end editby -->

Well Yoland you're very wise to deal with unexpected changes by flying the airplane, rather than by reprogramming the computer. These days I think there is a renewed emphasis in pilot training on knowing how the aircraft feels under various flight conditions and on flying the aircraft by using the basic instruments and looking out the windscreen. I suppose it's my interest in the automated systems that leads me to approach every flght by trying to automate the whole sequence. I need to remember to practice my "hand flying" skills more frequently so I can deal with surprises when they come up.