Trying It Out

Sometimes for my keeping-my-hand-in sim sessions I fly approaches that I know to be tough, or ones that I happen to have plates for, and sometimes I just fly what's in the news. After reading about and discussing the AA runway excursion at Jackson Hole, I decided to sim it.

I set up at and departed from an airport to the west of Jackson Hole. I wasn't playing with a B757, just a King Air, thus setting myself a much easier task than the one the accident crew faced. Also to my knowledge flight sim doesn't have an option to set the CRFI -- that stands for Canadian Runway Friction Index. What's it called in other countries? I set the weather up to match reported conditions at Jackson Hole at the time of the accident: 400' ceiling, light snow, zero Celsius and visibility of 1 mile, with wind 240 at 10 knots. And I sought out the approach plates. I found the ILS or LOC Y RWY 19, printed it off, dug out some LO charts and started playing the game.

My plan for the approach is that on station passage at JAC I turn left to intercept the I-JAC localizer outbound and descend to 14,600', (that's the 14,100' permitted outside QUIRT on the procedure turn, corrected for temperature). I have to get down to 9900' (e.g. FAF 9700' with temperature correction) fairly swiftly after the course reversal in order not to be intercepting the localizer from above, so I check to see how far out I can go on the procedure turn. The plate says "Remain within 10 NM." It doesn't say 10 nm of what. A Canadian plate would give the fix there. It can't be of JAC, because QUIRT is 17.3 DME from JAC, so maybe the rule on US plates is from the depicted FAF at FAPMO. I've decides to try it with that. The published DH is 7063, or 612' above the threshhold. That's unusual, it's usually only 200', but this is a mountainous area so perhaps it doesn't have a good enough missed to allow descent that low. The cold weather correction adds thirty-five feet for a DH of 7100 for me. I can't possibly make the field that way with a 400' ceiling, so AA must have their own approach. That's not uncommon. The missed is easy, climb pretty much straight out tracking a radial for 26 miles.

So go. The JAC VOR doesn't identify on the published frequency. It exists in the game database on a separate frequency, so I switch to that. Twenty-five miles back from the JAC VOR I started descent to 15,000', as depicted for my sector in the MSA (minimum safe altitude) circle. I didn't calculate the correction for that, but by the time I was close to 15,000' I knew I'd be almost outbound anyway. I put the JAC VOR up on NAV2 and the ILS on NAV1, so I was ready to track. The ILS frequency works. The AA crew would have had contact with ATC and vectors, of course, but if my flight sim game has an ATC feature sophisticated enough to give vectors, I don't use it.

I started the course reversal at QUIRT and barely kept it within the 10 nm, chopping throttles and diving as soon as I had intercepted inbound, but the glideslope was already alive below me. I messily dived for it, then realized that I could see the field, ten miles away, with no cloud layer below. This happens a lot in MSFS: somehow the weather I set goes away. I have to miss anyway, because I'm way too high.

I ask a US pilot friend about the vague "within 10 nm" instruction and he produces a company Jeppesen (much better, but more expensive) chart for the same approach, on which it is explicit that the 10 nm is measured from QUIRT. He doesn't remember what the trick is to know what fix to use on the NOAA plates that I have. I then find another version of the approach, this one the ILS or LOC Z RWY 19. I don't know what that Y and Z are, or why the depicted ILS should be different on a plate where the LOC approach is different. It may just be that the approach has been improved since 2009. Maybe they blasted some rock so you can now follow a normal ILS.

I go back to the sim, make another attempt to set the weather, set the rate of change slider for the weather to zero and restart from a saved position approaching JAC VOR descending through 17,000'. This time I go out past QUIRT, do my course reversal damnit the glideslope is still below me. It's not too far down so I dive for it, then do the whole thing again with more aggressive descent through the procedure turn and I intercept the glideslope and localizer together in descent. Practice session done. It's as much practice of working out the details of the approach as it is of the actual handling.

I don't believe the US regs require me to make cold weather corrections, so perhaps the coldest seasonal temperatures are already incorporated in the plate. It's kind of bizarre that an instrument rating in one country qualifies me to fly IFR in any country despite the fact that small regional differences can confuse or even kill.

This isn't intended as any kind of reproduction of the incident, just a sim scenario inspired by it. Tomorrow I may fly out of Cairo, taking my inspiration from that news headline.

Meanwhile I have some more information on the AA overrun that inspired this sim session. On the 757, the aircraft determines if it is flying or on the ground based on the indications from tilt sensors on the main landing gear. My friend says, "Since the "air" input is only 1/2 second, my assumption is that it's not a bounce (I doubt both tilt sensors could show air for only 1/2 second during any kind of bounce). If anything, it was a very smooth landing that momentarily triggered no-tilt, then back to tilt, before showing on the ground, or a momentary malfunction of the air/ground sensing." He confirms that 132 knots is a normal touchdown speed for the 757 for a moderately heavy airplane, which would have applied here, due to a full passenger load and alternate fuel.

He also pointed out an error in my original blog entry. The thrust reversers are not armed during the approach, they are selected manually after landing (the spoilers are armed). I was getting mixed up with the autobrakes.

Meanwhile, I have an HR telephone interview in a couple of days.