March 13, 2008

A Few Small Bumps…

We're cruising steadily at 5,000' in Cessna 051 (the club's G1000-equipped C172), on an IFR flight plan to Stockton (KSCK) for some IFR practice. The forecast has some scattered but benign IMC patches along our route. I've set up the G1000 and the autopilot to get us direct as cleared to JOTLI, a convenient IAF for Stockton's GPS 29R approach. The plan is to do that approach full pilot nav with the autopilot and G1000 coupled, as tonight's flight is supposed to be mostly about IFR systems flying rather than just a raw IFR workout (any excuse to watch the G1000 get us around the course reversal hold automatically again :-)). After the initial approach we'll try to get one or two more approaches in before returning to Oakland (KOAK), where we'll probably need a real clearance to get back in anyway, given the approaching cloud bank (I have one pre-filed just in case). I'm under the hood; E., one of John's instrument students, is doing safety pilot duties in the right seat. It's about 19.30 (7:30 pm for all y'all Americans), so it's dark(ish) outside, but it's easy enough to see the stratus a short distance ahead of us at our altitude. It's not particularly thick, extending maybe a few thousand feet (from about 4,500'), and doesn't extend more than a few miles along our route (at least according to the forecast).

We enter the clouds and I take off the hood. I turn off the strobes and the landing lights. Everything looks good to me: on course, at altitude, the approach has been set up, we've got the ATIS for Stockton. The cloud layer feels benign; the outside air temperature (OAT) is well above freezing. There's some very minor turbulence, but nothing very interesting.

Suddenly there's a few small bumps, then an almighty series of crashing thumps as we're battered around by what I'd characterise as severe turbulence. The autopilot disengages automatically, but only milliseconds before I hit the disengage switch myself. My head hits the cockpit ceiling several times (we're wearing those bloody "smart" Cessna seat belts with the airbags); I lose my headset down the back of my neck. Over the next few seconds we gain and lose several hundred feet in each direction and get into some pretty unpleasant pitch and roll excursions. I grab the yoke with both hands and try to regain control, throttling back to about 2300 RPM (I do this by feel and instinct, since I can't actually see the engine instruments on the MFD at this point). It's clearly not going to get better quickly, and while I'm not having too much trouble keeping us more-or-less upright, we're in all sorts of trouble with the altitude excursions, and we're not going to be able to continue like this for long without a great deal of continuous effort on my part (and a lot of discomfort and the possibility of losing control in IMC). I decide we need to get below the clouds now: I know the area, there's nothing below us, and my guess is that even 1000' below us it's smoother. But I'm wondering how the hell I'm going to get the headset back on while keeping both hands on the yoke. At any rate we have little or no control over our altitude at this stage, so I start doing what I can to descend. It's getting slightly smoother, so I risk reaching back and sitting the headset back on my head, albeit a little off-kilter (understatement). It'll do for now. Both hands back on the yoke, I manage to blurt out something like "NorCal, 051's in heavy turbulence in IMC, any chance of 3,000 now?". We're at 4,500 by now anyway, and it's already smoothing out. "051, I can only give you 4,000. Descend and maintain 4,000". "051, We'll take it. Descend maintain 4,000. Thanks…". I level off in relatively smooth air, clear of the clouds, at 4,000, and start to look over the systems and controls. Everything looks fine, but I'm not going to trust it all immediately. The way ahead looks clear of clouds. My thinking at this stage is basically that even if it's not smoother down here, at least it's VMC, and losing control is a lot less likely. I'll hand fly for a while without the hood on just to see what happens. At least we're still probably on course (I'll check in a minute).

I look over to see if E. is still here. Thankfully, she is, but she's still holding onto the side of the cockpit, just in case. I ask if she's OK; she says she's fine. I'm not sure I believe her, but in any case there's not much we can do at the moment except recover the plane and then work out whether to continue on or do something else. It sure looks and feels benign under the clouds and all around us now. I spend a few seconds giving a mini-PIREP and explanation to NorCal, basically just saying we'd hit unforecast moderate-to-severe turbulence in IMC just back there, and it's a lot smoother at 4,000'. I probably sound a bit shaken on the radio at this point; the controller sounds a little concerned, and I'm actually gratified that he's taking it seriously.

E. and I discuss things for a short while as I recover the course (and my approach plates, which are now on the floor), and decide to press on to Stockton as planned. I mention the possibility of hitting the same thing on the way back, but since the return leg is a fair way to the south of where we are now, and somewhat higher, I decide to play that bit by ear (literally in some ways: I'll have my ear (figuratively) glued to the frequency for reports of any turbulence on the way back). There's no shortage of places to land below us on either leg, and it's basically a nice VMC evening below the clouds. I hope E. hasn't been scared off instrument flying (or flying right seat with me) for ever. She sounds O.K., but you never know, so I stress that I'd really be just as happy if we abandoned the flight as if we continue. But she's up for it. We press on.

* * *

The whole episode probably didn't last more than about ten seconds, and in reality the turbulence wasn't quite the worst I've ever experienced in a light plane. I think I'd characterise it as on the low end of "severe" by the FAA's definition (I had little or no control over altitude, and maintaining pitch and roll control was quite a struggle), but I'm also guessing it was fairly localised and would probably have petered out fairly quickly. It was in many ways a more sudden and prolonged IMC version of the encounter with turbulence I blogged about last year. My first thought after we'd leveled off at 4,000' was "wake turbulence!", but it continued on too long for that and there wasn't any likely culprit in the area (on-frequency, at least).

And it was unforecast, and happened in IMC, which always adds another dimension to things. At the time it just didn't seem too hard to right the plane, keep it under some sort of control, and keep flying with my eyes glued to the G1000. But I'd hate to have done it all with the crappy old mechanical AIs in an older 172 — the potential for losing control in that situation, or in partial panel with no PFD or MFD, is sobering. In any case, it's definitely not something to be complacent about, and, according to E., it was the worst turbulence she's experienced so far [and a few days later I still have a bruise on the top of my head from the experience].

* * *

Luckily, the rest of the flight feels pleasantly anticlimactic, and apart from some mild turbulence in actual IMC on the ILS back into Oakland, things go fairly smoothly and as-planned. E.'s a good safety pilot and keeps me on my toes; it's good having an instrument-savvy co-pilot, and apart from a few minor lapses on my part (below), things go about as well as you'd reasonably expect for someone (me) who really doesn't get to fly enough to be in perfect practice all the time.

The G1000 gets a good workout during the flight. Or, rather, I get a good workout in G1000 usage, and — as always with the G1000 — I'm reminded of what I dislike as well as what I really like about it. There's a lot to like, but when — as happened a couple of times during this flight — you get into an incomprehensible situation where no matter what you seem to be doing the G1000 just sits there presenting an unexpected menu or menu item, or won't let you do what you think you've always been able to do before, well, I think I'm slowly becoming an expert in G1000 workarounds.

For example, on the way out to Stockton (before The Bumps), I ask for and get direct JOTLI for the full pilot nav version of the GPS 29R approach. So I reach over and hit "proc" to load and activate the approach with JOTLI as the IAF, which should send us direct JOTLI. But no matter what I do it only gives me approaches back at Oakland. This ignites a few "WTF?" moments and responses from me, and both E. and myself notice that there's actually an extraneous KOAK at the end of the entered flight plan as well as at the beginning. How the hell did that get there? I think. But it doesn't matter, does it? What matters is how to get direct to JOTLI right now; so let's just scroll up to KSCK, hit "direct", then select the approach; this does the job (especially since at this range direct KSCK is roughly the same heading as "direct JOTLI"). Yes, a simple thing, but a few years ago I might have spent way too long sitting there dumbly wondering about the extraneous KOAK and how to clear the flight plan or how to dial in direct JOTLI while debugging the situation. By any means necessary, as I've learned over the years.

Later, on departing Stockton, no matter what I do, I'm absolutely unable to delete the current flight plan from the inset flight plan window so I can input the new one back to Oakland. Luckily, when we'd picked up the departure clearance midway through the previous approach I'd pre-set the VOR 2 receiver and OBS to give me the VOR version of the initial (30 nautical mile) leg along V195 from Manteca VOR, and when it quickly becomes obvious that the G1000 isn't going to cooperate, I just let the autopilot fly us to and then along the relevant VOR 2 radial. At my own leisure I then simply do a "direct KOAK" with the GPS, then get SUNOL (the end point of V195) from loading the KOAK ILS 27R (our intended approach, of which SUNOL is the IAF for arrivals from our direction). When actually on V195 (according to VOR 2), a simple "activate approach" gets the GPS in sync and in control once again. OK, standard instrument stuff, but (as John will confirm), quite a few instrument pilots seem to get complacent with the better GPS units and forget to setup a fall-back based on VORs or whatever. Again, a few years ago I'd probably have spent several increasingly tense minutes debugging the G1000, with only the initial ATC heading setup properly, and with only a hazy idea of where V195 actually was in relation to my current situation. I'd probably have blundered into the correct course eventually and without incident (but with a great deal of silent swearing :-)), but it's pleasing that I can actually keep cool enough (and think far enough ahead) nowadays that the whole Grace Under Pressure thing seems to be on the verge of happening (all this expensive instrument flying must count for something :-)).

About the only other thing of note was a PIREP to NorCal from a light GA plane of carburettor icing somewhere near Sacramento Exec (KSAC) at roughly our altitude. When I hear this I guess I'm not surprised — the OAT's fairly high, the air's moist, and conditions for carb icing seem just about perfect. What is (pleasantly) surprising is that the pilot reported it at all, and that NorCal also took the report seriously. No, I haven't flown a carburetted plane for several years now, but hearing that someone else is experiencing it might help save a few tense moments (or worse) for someone out there.

* * *

Later, back in the impending drizzle at the Oakland Flyers parking lot, I ask E. (again) whether she was OK with the whole experience. She still insists that she enjoyed it and learned a lot; I hope so. I'm guessing that it was definitely a learning experience for both of us.