Land and Hold Short

I’ve written about icing before, both here and here. Like storm clouds and scud running, icing is one of those things that pilots are supposed to avoid but occasionally stumble into anyway. The Canadian AIP contains some advice for pilots who end up in storm clouds (slow down, keep flying straight, and don’t worry about altitude, or something along those lines), and Rick Durden has written a good collection of scud running survival tips, but there is precious little out there to help us with dealing with icing on a day-to-day basis.

Fly high

OK, you’re planning to fly IFR on a day when there is no moderate or severe icing forecast, but there is cloud and possibly precip along the way, and the temperature at higher altitudes might be close to freezing. What do you do: fly low, to try to stay in the warm air, or fly high, where you might pick up ice?

My tip — and more experienced pilots reading this should feel free to correct me — is to fly high, even if you’re at or below the freezing level. Ice tends to accumulate at very specific altitudes: for example, you might pick up some light clear icing at 7,000 feet, but nothing at 9,000 or 5,000. In theory, then, you can either climb or descend to get out of icing. However, if ice should happen to accumulate quickly, and especially, if you should happen to be flying a heavily-loaded and/or weakly-powered plane, climbing might not be an option (for planes with boots, climbing too steeply is also dangerous, because the high angle of attack can allow ice to accumulate on the wing where the boots cannot reach it).

So, if your only choice is to descend, what happens? Assume that MOCA is 3,000 feet — if you’re already at 5,000 feet, the temperature drops a bit, and you pick up ice, you have only 2,000 feet left to descend safely in the hope that the ice will melt off; if you’re at 9,000 feet, you have 6,000 feet to descend to melt off the ice. That leaves you with a lot more choices. It may even be that 5,000 feet is the icing altitude for much of your route, but you’ll overfly it without ever knowing.

Of course, you don’t want to try this without some hope of warmer air underneath — if there’s an inversion causing freezing rain on the ground, descending probably is only going to make things worse. On a typical spring day or summer day, though, with some cumulus cloud and near-freezing temperatures at cruising altitude and a reasonably low MOCA below, I think that flying high makes more sense than flying low — like always, if you run into any trouble, your altitude is like money in the bank.

Via Aviatrix, I’ve found another interesting aviation blog: Randy’s Journal is a weblog maintained by Randy Baseler, VP Marketing for Boeing Commercial Aircraft. Obviously, the blog reflects company orthodoxy, but it at least uses direct language rather than the marketing mumbling we get in press releases. I plan on reading this one. An interesting starting point is an entry gloating about becoming Air Canada’s exclusive supplier, a massive win for Boeing (who sorely needs one right now after many losses to Airbus).

DISCLOSURE: My launch customer as an independent consultant in 1998 was Boeing Commercial Aircraft, and I did a substantial amount of work for the company as an architect of their system for producing aircraft manuals.

[Updating: on guessing -- see below] The National Air Traffic Controllers Association (NATCA) — the non-union union representing U.S. air traffic controllers — attacked plans to privatize the U.S. air traffic control system with the following statement:

Cleveland controllers alone handle more operations annually than Canada’s entire privatized system.

Cleveland airport? The airport serving that little city called the mistake by the lake? Presumably not, though it doesn’t hurt Carr’s cause to let congress think that’s what he meant. He’s almost certainly talking about Cleveland Center, one of the world’s busiest air traffic control facilities, handling traffic across an enormous part of the northeastern U.S. including two major hubs (Pittsburgh and Detroit) as well as most enroute traffic between Chicago and points east; according to their web page, they handle over 3 million operations per year.

That’s a lot, but is it more than all of Nav Canada? Not quite. According to this news backgrounder from their web site, Nav Canada handled 11 million operations in 2004.

Was Carr deliberately lying? Saying that “Cleveland” has more operations than Canada is clearly an attempt to mislead a bit (he’s hoping that members of congress will mistakenly assume he means the city of Cleveland, rather than an air traffic control unit covering much of the northeastern U.S.), but it might be going too far to call that a lie. What about the number? It might be possible to get a number bigger than 11 million for Cleveland Center simply by creatively counting all operations for all facilities under Cleveland’s airspace. For example, let’s look at an IFR flight from Detroit to Pittsburgh: Detroit tower handles the flight during taxi and take off, then passes it to Detroit departure for initial flight, which then hands it off to Cleveland center for enroute, which then hands it off to Pittsburgh arrival to set up the approach, which then hands it off to Pittsburgh tower for landing and taxi. If NATCA simply added up all of the operations for all ATC units underneath Cleveland Center, that single, short flight would count as five operations. I’m guessing that’s what happened. Nav Canada uses a single, electronic slip for flights from taxi to tie-down, so I’m pretty sure that they count each one only once.

I don’t know if it would be best for the U.S. to stick with its current socialized system or to move to a privatized, Nav Canada style system — I get good service from ATC on both sides of the border — but let’s keep the debate honest, and avoid any Enron-style counting on either side.

Update: one interesting point of discussion in the comments to this posting is my use of phrases like “I’m guessing” and “I’m pretty sure.” It’s worth noting that in both cases, the fuzzy stuff was an attempt to give Mr. Carr the benefit of a doubt — the cold facts alone look very bad for Mr. Carr in this case, and my guesses were attempts to try to find some way that his statement could have been an honest misunderstanding rather than a deliberate deception.

Yesterday, during a Hope Air flight, the virtual odometer in my logbook finally ticked around past 400 hours.

It was a day of flying extremes. I flew three legs totalling 8 hours flight time (7.2 air time), of which 3.3 hours was hand-flown in IMC. The trip involved flying IFR through Toronto terminal airspace over Toronto/Pearson, Canada’s busiest airport; flight over water; a slow-moving cold front blanketing the route with heavy rain and hours of light and moderate turbulence; a brief encouter with clear icing (not with passengers in the plane); and at the end of the final leg, an ILS approach in more turbulence with a circling landing, and then back home in time for supper. I don’t think I’ll have trouble remembering what I was doing when I hit 400.

[Updated] The Canadian Press has an unusually detailed and accurate story about floatplane safety in Canada. Unfortunately, with the summer season arriving, we’re going to see another big spike in the fatal accident rate in Canada as the floatplanes take to the skies, simply because they normally fly without the safety net we have at certified airports: clear, monitored approach paths, known surfaces, etc. (of course, that’s the whole point of recreational floatplane flying, to taxi up to your buddy’s dock at some little lake in cottage country). Also, the consequences of a hard landing are worse: instead of an embarrasing bounce on a runway, you end up digging in a float and flipping, then trying to escape from a cabin upside-down and underwater with a door latch that won’t open.

The strangest part of the article is a claim from Transport Canada that — despite repeated pleas from the Transportation Safety Board, who investigates these fatal accidents — TC cannot issue an Airworthiness Directive (AD) to make floatplane doors easier to open for escaping, because most of the planes are not made in Canada, and thus, are outside their jurisdiction.

Perhaps they’ve forgotten about AD CF-90-03R2, by which Transport Canada mandates an annual inspection of the muffler shroud for possible carbon-monoxide leaks from my American-certified and built Lycoming engine through the heater into the cabin of my American-certified and built Piper Warrior II (for the record, since I fly a lot in the winter, I’d have this inspection done with or without an AD).

Floatplane flying matters more in Canada than the U.S., so if there’s going to be an AD, it will have to come from us: while taxiing up to your friend’s dock is nice, commercial floatplanes are also the lifeline to many small northern communities, bringing in food, medicine, doctors, teachers, etc. and flying out people suffering from medical emergencies. Even if the floatplane drowning fatalities don’t add up to a lot in the bigger scheme of things, they leave this small part of the population — the ones who have no choice but to use floatplanes — extremely vulnerable.

[Update: the difference between the TSB's request and AD CF-90-03 is that the latter deals only with an inspection, while the TSB's request deals with a modification -- perhaps that's the place where TC is stuck.]

The annual inspection for my Warrior is approaching.

As I’ve mentioned before, as a Canadian private aircraft owner, I have the final responsibility for determining that the annual inspection has been completed. In the U.S., on the other hand, it’s not the private owner but an Inspection Authority (IA) who makes that determination and signs a logbook entry returning the plane to service after the inspection. In both countries, of course, the pilot is ultimately responsible for airworthiness before each flight, no matter who has written and signed what in the logs.

In Canada, the Aircraft Maintenance Engineer (AME) performing the annual inspection will typically follow CAR 625, appendix B: part 1 lists a set of standard inspection tasks for any small airplane, such as testing cylinder compression or the torquing and safety wiring of propeller attachment bolts. However, most aircraft manufacturers also publish a customized inspection list for each aircraft, that goes into more detail and includes additional tasks (unfortunately, the manufacturers’ inspection lists are rarely, if ever, available free online).

I’m skipping the 100 hour/annual inspection here, since it isn’t all that different than the generic one in the CARs (aside from a bit more detail). What is more interesting is that the checklist for the PA-28-161 Piper Warrior II includes tasks for 50 hours (i.e. at each oil change), 500 hours (~4-5 years), and 1,000+ hours (~8-10+ years), all inspections not specifically covered in the CARs and not necessarily familiar to owners. I don’t agree with all of these — for example, I’m not automatically going to recondition a fixed-pitch propeller at 1,000 hours if it’s still in excellent shape — but overall, these look like intelligent recommendations, and I’m going to try to make space for them in my maintenance planning. At best, I might avoid a forced landing or worse; at a minimum, I’ll be able to pass on a better-maintained plane to the next owner, and will be able to hold my head up during the prepurchase inspection.

The 50 hour inspection

This one looks huge, but most of the tasks are simple ones, and a good number qualify as elementary work (I’ll post on that in the future). Some of these are things any pilot would do before every flight, such as checking the tire pressure, oleo extension, and alternator belt, and others are part of a normal oil change, like draining the sump or cleaning the plugs. Because my Warrior is blessed with a fully-opening cowl (rather than just a tiny oil door), I can also perform a good visual inspection of the engine, hoses, sparkplug leads, and engine mount before every flight. Still, there’s a lot here that is not part of my normal 50-hour oil change. If I’m changing the oil anyway (say, in the late fall), and already have the plane up on jacks to remove the nosewheel fairing, I think it might make sense to pay for 2 hours or so of an AME’s time to perform the other tasks in this list. It looks like a relatively easy way to maintain a safe plane, and a lot less expensive than the questionable fairy-dust-style safety expenditures many owners make, like backup vacuum pumps (for fixed-gear planes) and traffic alerting systems.

• Inspect [propeller] spinner and back plate for cracks.
• Inspect [propeller] blades for nicks and cracks.
• Remove and inspect engine cowling for damage.
• Drain oil sump.
• Clean suction oil strainer at oil change. Inspect strainer for foreign particles.
• Clean pressure oil strainer or change full flow (cartidge type) oil filter element. Inspect strainer or element for foreign particles.
• Inspect oil lines and fittings for leaks, security, chafing, dents, and cracks.
• Fill engine with oil per information on cowl or service manual.
• Inspect spark plug cable leads.
• Inspect rocker box covers for evidence of oil leaks. If found, replace gasket, torque cover screws 50 inch-pounds.
• Inspect condition of carburetor heat air door and box.
• Inspect all air inlet ducts and alternate heat duct.
• Remove, drain, and clean fuel filter bowl and screen. (Drain and clean at least every 90 days.)
• Inspect condition of flexible fuel lines.
• Inspect exhaust stacks, connections and gaskets. Replace gaskets as required.
• Inspect engine mounts for cracks and loose mountings.
• Inspect condition and tension of alternator drive belt.
• Inspect security of alternator mounting.
• Check condition and security of hydraulic line and fluid in brake reservoir; fill as required.
• Reinstall engine cowl.
• Check landing, navigation, cabin and instrument lights.
• Inspect battery, box and cables (Inspect at least every 30 days). Flush box as required and fill battery per instructions on box.
• Remove, drain, and clean fuel strainer bowl and screen. Drain and clean at least every 90 days.
• Lubricate [fuselage and empennage] per lubrication chart.
• Check emergency locator transmitter battery replacement date and transmitter for operation.
• Lubricate [wing] per lubrication chart.
• Inspect oleo struts for proper extension (N-3.25 in./M-4.50 in.). Check for proper fluid level as required.
• Check tire pressure. (N-30 psi, M-24 psi.)
• Lubricate [landing gear] per lubrication chart.
• Inspect all hydraulic lines and attaching parts for security, routing, chafing, deterioration, wear, and correct installation.

The 500 hour inspection

The 500 hour inspection includes only three tasks not already in the 100 hour/annual inspection:

• (400 hours) Remove rocker box cover and perform a valve inspection.
• Remove and flush oil radiator.
• Replace auxiliary vacuum pump.
• If installed, replace central air filter.
• Clean and lubricate stabilator trim drum screw.

I have my stab trim screw cleaned and lubed regularly because it becomes stiff in cold weather, but I can find no record of the oil cooler ever having been removed and flushed — any gunk caught in it can circulate back into the cylinders, wearing them down and forcing an early overhaul; this year, I plan to have this work done as cheap insurance for my engine. I don’t have an auxiliary vacuum pump in the plane, only the main engine-driven one.

The 1,000+ hour inspection

There are only a few 1,000+ hour tasks:

• Recondition propeller.
• Replace magneto.
• Overhaul or replace engine driven and electric fuel pumps.
• (2,000 hours) Complete overhaul of engine or replace with factory rebuilt.
• Replace engine driven vacuum pump.
• Clean and lubricate all exterior needle bearings.

OK, now it’s time to be realistic. I am not going to replace my engine at 2,000 hours if it’s still going strong. In fact, I think that doing so would actually be more dangerous — the only forced landing I’ve ever seen happened right after a new engine was installed, because the shop attached something incorrectly; two other local forced landings I’ve heard about had similar causes. As long as my compressions are good and there’s no metal in the oil filter, I’d rather stick with an engine that has all its bugs shaken out than a new, unproven one. The vacuum pump, on the other hand, is cheap to buy (a few hundred dollars) and easy to install (about 30 minutes of an AME’s time) — I’ve already had one fail on me, and preemptive replacement here might make sense. I’ll have to think about that one.

Just in case anyone has the mistaken idea that I — who, as a teenage boy, preferred reading books to fixing cars — actually know much about nuts-and-bolts (so to speak) of maintenance, I’ll finish with a simple question: What’s a needle bearing?

The Airport Council International (ACI) released its preliminary 2004 list of the top 30 airports yesterday (ranked by number of passenger movements). Toronto/Pearson saw a huge increase in traffic, but is still just barely clinging to the list in spot 29 (a bit busier than Philadelphia, but not quite so busy as Seattle).

Given its low ranking, it’s funny that Pearson (a) has so much trouble handling traffic volume, and (b) has such a — dare I say it — messiah complex about its importance. For example, when I flew into Philadelphia International in fall 2002, I just showed up unannounced (aside from my IFR flight plan), I paid no landing or ramp fees, U.S. customs drove around to the FBO to meet me, the parking fees were reasonable (USD 20.00/night, first night free), and I found the pace of ATC and general traffic relaxed and quiet both arriving and departing. In contrast, Toronto requires IFR slot reservations for certain times of day, the airport charges close to CAD 200.00 landing fee for a light single (depending on time of day), and Frank Eigler was billed over CAD 700.00 for leaving his plane at Pearson for 48 hours. I’d like to try Pearson once, just to say I’ve done it, but it’s hard to ignore the strong message that I’m not welcome.

Maybe they’re just too busy.

U.S. airport diagrams for all airports with instrument approaches have been available online for a while, along with all U.S. instrument procedures (SIDs, STARs, approaches, etc.); recently, official PDF versions became available which do not go fuzzy when you blow them up and print them out, like the old scanned documents did.

Now, finally, Nav Canada has made the same thing available for Canadian pilots here (click on “CAMS”). Previously, to get free Canadian airport diagrams online, you had to go to the U.S. department of defense (seriously!) at their DAFIF site, and even then, only a relatively small number of Canadian airports were represented (note, however, that the DAFIF is still the only free source for actual Canadian approach plates).

So, how do score Nav Canada for this one?

• +10: made the diagrams available for free, even though it meant paying a government department (U.S. government publications are in the public domain; Canadian government publications are cost-recovery, so you have to pay to play)
• -1: invented a confusing new name, CAMS (Canadian Aerodrome Maneuvering Surfaces)
• -5: whole thing in one massive PDF file (this link will expire soon), instead of a separate PDF for each airport

Final score: +4, or “good effort; keep working to improve”. That last point is particularly annoying, not because of the long download (a majority of Canadians have broadband), but because it’s impossible to link to or bookmark individual diagrams — if I’m posting about an airport, I’d like to be able to make a link, but I cannot. This principle — that everything has to have its own, persistent URL — is called REST, and it’s the basic design that makes the web work. But now, I’m letting high-tech stuff from my other weblog spill over into this one, so I had best stop before I lose the few readers I have.

Kris Johnson has a posting on holy wars in aviation, including the two variants of the very dangerous teaching that you control airspeed with pitch and power. The idea is that students learn to look out the window (which is good) and, for any given RPM — assuming a fixed-pitch prop — memorize what pitch attitude will give them what airspeed. I believe that this teaching approach has two significant consequences:

1. it gets student pilots to first solo faster; and
2. it kills some of those pilots (and their friends and families) after they get their licenses.

While we all want to save money on flying lessons, I think this is a lousy tradeoff. The only way to make a given pitch/RPM combination produce a given airspeed is to make sure that gross weight, density altitude, CG, sideslip, and bank are always exactly the same. During training, while these conditions are not always exactly the same, they’re usually pretty close, so instructors can get away with this little cheat, students solo sooner, and everyone’s happy.

Departure Stall

Some day soon, though, the new private pilot is going to want to do more than fly around the same area under the same conditions with zero-to-one passengers on board, and unfortunately, he or she is going to be in for a big surprise. Consider this: after training through the fall, winter, and spring, the pilot (we’ll make it a male) finally has his PPL, just in time for summer vacation. He loads his wife and two children on board a Cessna 172 or Cherokee, secures the baggage in the back, and fills the tanks (making sure to stay withing W&B, of course). It’s a bit of a hot, humid day, but it’s a low elevation airport and a nice long runway, and the pilot must have taken off here 50 times during training, so it shouldn’t be a problem.

It takes the plane a lot longer to get up to rotation speed in the takeoff roll, but the pilot expected that — that’s why they have those takeoff-distance charts in the POH. It’s also much harder to unstick the plane from the runway; in fact, it seems to want to settle right back down again. Finally, though, the plane is climbing … well, sort of. The pilot is used to about 700 fpm with half tanks and an instructor on board, and 1,000 fpm or better when he’s solo; a quick glance at the VSI, though, shows only about 200 fpm. HUH? Well, the pitch is wrong — the nose is too low compared to every other climbout the pilot ever did — so he pulls the nose up a bit to get closer to the normal climb pitch. The VSI jumps up to 500 fpm, so this was obviously the right choice … except that three seconds later, it’s down to less than 100 fpm. The plane isn’t climbing at all, for any practical purpose. OK, pull the nose up a bit more (the tach shows full power, and the nose is still lower than usual), and sure enough, the VSI jumps up a bit more, before it settles down again, this time with no climb at all.

You can see where this is leading — one or two more pullups (still below the normal climb pitch angle from training) and the flight ends up as one of the many, many summer stall-spin accidents on takeoff. According to the Transportation Safety Board of Canada statistics, from 20-33% of accidents take place during takeoff, literally during the first few seconds of flight. In this case, a glance at the ASI should have told the pilot that the nose was too high, not too low, but the problem was that the nose looked too low, because the pitch was too low, and pitch + power = performance. Normally, the pilot could barely see the ground ahead at all in a Vx or even Vy climb, but this time, the ground seemed to fill a third of the windshield. Even if the pilot did look at the ASI, he might have had trouble believing it, since it was outvoted by the pitch (which showed the nose too low), the power (which showed the correct RPM), and the VSI (which showed the climb rate way too low), and all of that was combined with the stress of having the family on board for the first time, etc. etc.

Angle of attack, angle of attack, angle of attack …

Here’s the problem — it’s not pitch and power, but angle of attack that controls airspeed. The pilot studied angle of attack during groundschool, of course, but it never seemed all that practical, and the instructor’s “pitch + power = performance ” mantra seemed simpler and more logical. As long as the plane was flying under similar conditions (density altitude, gross weight, etc.) you could pretty-much map pitch angle to angle of attack during each phase of flight — there was a “climb attitude”, a “cruise attitude”, and an “approach attitude”. In the example above, however, there was a significant change to some of the conditions, so the mapping didn’t work any more.

My Warrior, for example, might climb as fast as at 1,200 fpm on a winter’s day near sea level with just me on board and half fuel; it can barely manage 400 fpm (at Vy) from the same airport with the whole family, dog, baggage, and full fuel on a hot summer day. Since Vy for my Warrior is 80 knots, assuming no wind I’ll fly forward about 8,100 feet every minute — if I’m climbing at 1,200 fpm, I’ll follow an angle of climb of nearly 9 degrees (!!); if I’m climbing at 400 fpm, I’ll follow an angle of climb of about 3 degrees. Now, my angle of attack, which controls my airspeed, goes on top of that. Let’s say that a 4-degree angle of attack gives me Vy (I’m just guessing):

• when I’m flying light on a cold day, my pitch angle will be 4 degrees for angle of attack, minus (say) 2 degrees for the incidence angle of the wings (which are normally tilted up a bit), plus 9 degrees for the angle of climb, giving me a pitch angle of 11 degrees — I’ll see nothing but clouds and sky outside the windshield.
• when I’m flying heavy on a hot day, my pitch angle will be 4 degrees for angle of attack, minus 2 degrees for the incidence angle of the wings, plus 3 degrees for the angle of climb, giving me a pitch angle of 5 degrees — the bottom third of the windshield will be filled with the ground.

In one case, a pitch of 13 degrees + full power = 80 knots; in the other case, a pitch of 5 degrees + full power = 80 knots. Different pitch, but same airspeed and same power. Clearly pitch + power != performance.

Why do some instructors do this?

Of course, there are many instructors who know the dangers of the pitch + power thing and teach it properly: you have to get to the right airspeed first, and then choose the pitch angle that gives you that airspeed at that power setting at that particular moment — it might be different every time. And you have to choose a new pitch angle not only for every flight, but for every power change in that flight (and even for the same power setting over a long trip, as you burn off fuel and get lighter). Taking the shortcut — memorizing preset pitch angles — might get you to first solo faster, but it might also kill you.

Part of the problem with the instructors who get it wrong might be the fact that a lot of them — especially the ones using flight instruction as a career step — have a surprisingly limited exposure to aviation. It’s even possible that some of the ones who are building time to get into the airlines might never have gone on a really long cross country (the one for the CPL is only a few hundred miles), never have flown a low-powered plane near gross weight on a hot day, never have taken off from a short, obstructed grass strip, never have flown an approach in actual low IMC, never have tried to maintain control inside a TCU with lightning flashing around them, never have done a continuous 180 descent from close downwind to landing to avoid holding up three jets on final at a busy airport, etc. etc. Their experience doesn’t really match much that their PPL students will be doing if they choose to remain private pilots, so instead, they pass on platitudes and old chestnuts that they learned from their instructors, and spend a lot of time on relatively pointless exercises designed to get the student past the flight test rather than helping him or her become a safe pilot.