Land and Hold Short

The engine finally came off my plane earlier this week, and it should be on its way to Halifax today for a complete teardown and inspection after my lightning strike in July. I snapped a few cell-phone pictures while I was in the shop.

The first photo shows the propellers from a twin that was forced to do a gear-up landing after both the regular and manual gear extension failed — without the gear, the spinning propellers hit the pavement on landing. The pilot was smart enough to land the plane normally rather than trying to slow down and stop the props first — the props aren’t pretty, but the pilot’s unhurt and the plane, despite some damage, will fly again (hopefully before late fall).

The second is a front view of my magic magnetic engine, with the propeller and cowling removed. If you’re not used to looking at airplane engines (because, say, you rent planes that have only tiny oil-filler doors), the first thing you should note are the huge cylinders, compared to what you’d find in a car engine. Personally, I’m a bit concerned with what looks like damage on the prop spinner plate (I hadn’t noticed it when I was taking the picture).

The third is a view of the accessory drive on the back of the engine — there are various attachments there to allow the engine to spin things that, well, are supposed to spin, as well as room for other toys. The white thing in the middle is the oil filter, which gets replaced with every 50-hour oil change. One of the first things a new private owner-pilot learns is how to cut the safety wire, remove an old oil filter, cut the filter open to look for metal, attach a new filter, and safety wire it. On my plane, it’s possible do all of that without even taking off the cowling.

The dark thing just to the right of and slightly lower than the oil filter is the right magneto, which spins around to generate power for the spark plugs — there is a wire from the magneto going to one plug on each of the four cylinders (the left magneto also has a wire to a separate plug in each of the cylinders, for redundancy). Magnetos are maintenance hogs, and sometimes need to be rebuilt every few hundred hours, but fortunately the costs are not too high.

The grey, red, black, and white thing to the right of and slightly above the oil filter is the vacuum pump, which spins around to create suction to drive the gyros in the attitude indicator and heading indicator. Dry pumps like this one last 1,000 hours on average, but can fail at any time; even worse, unlike most parts, they give no warning — they go from 100% to 0% in a split second. Fortunately, I cannot find a single case of a fatal crash due to a vacuum pump failure in IMC on a fixed-gear plane flying IFR (the drag of the gear makes the plane easier to control), but there are many cases for retractables. Note to self: after losing vacuum pump in a retractable-gear plane, step #1 is lower the gear.

Here’s a close-up picture of the engine’s data plate (still can’t read it, though), and the carburetor. You know that my O-320 engine has a carburetor because of the bare “O-” prefix; if it were fuel injected, it would have an “IO-” prefix. Airplane engines almost universally use updraft carburetors, mounted underneath the crankcase — I’m not sure if there’s a good mechanical reason for that, other than avoiding having a hump sticking up in the middle of the top of the cowling. The throttle lever in the cabin pulls a wire around a series of pullies to a control on the other side of the carb (as does the mixture lever, I think). For some reason, the carburetor is much cleaner-looking than the rest of the engine.

That’s it for now. I’ve been up twice in a rental Cessna 172 to keep current, and hope to be flying my own plane again before the end of September.

I’m going to be grounded for a long time — possibly a few months — so I might not be posting much to this blog. After reattaching my overhauled and tested propeller, the shop was unable to complete a compass swing with either of two different compasses. It turns out that my plane has become heavily magnetized from the lightning hit (which I am now fairly certain happened while the plane was parked). A compass will deflect heavily towards the front part of the plane anywhere a few feet of it.

The good news is that Cherokees are mostly aluminum (and the firewall is stainless steel), so the problem is localized. Planes with steel frames, like the Mooney or most rag-and-tube planes, are extremely difficult to degauss, and are sometimes scrapped after become magnetized. In my plane, the main steel structures that could be magnetized are the engine mount, the crankshaft, and the nose strut, so the plane is probably repairable, though the insurance company may still decide to write it off.

More importantly, though, since the plane is unusually heavily magnetized, it’s almost certain that a strong electrical current passed through the engine block. That means that the engine has to be removed from the plane, shipped to Toronto, and completely disassembled and magnafluxed. Depending on how busy the shops are, I might not see my plane for a few months. I’m grateful right now that I have a helpful insurance broker who’s dealing with the adjuster on my behalf, and I’m also grateful that I took the picture of the prop that I published here earlier.

Aviatrix’s recent (and informative) post about alternate minima raises an interesting question: what’s a runway? For example, let’s take a medium-sized airport with a single paved surface, designated 09/27.

Most of us, describing the airport to a friend in the pilot’s lounge, would say that the airport has a single runway 09/27. However, the CAP would publish approach approaches to two different runways, 09 and 27, and the control tower might say “we’re switching the active runway from 09 to 27″. Both runway 09 and 27 will have the same surface type, snow removal, and edge lighting, but they might have different landing and takeoff distances (due to displaced thresholds) and different approach lighting, and most importantly, they’ll be used under different wind conditions. In the middle latitudes of the northern hemisphere, where I live, the wind will generally be from the west in good weather, so runway 27 would be used most often; however, the wind will often be from the east during times of low visibility, so runway 09 is more likely to see instrument approaches in low IMC.

All of this has to do with the Canadian alternate minima Aviatrix mentions in her posting. According to the CAP GEN, you can choose an alternate with forecast 400-1 weather only if there are “two or more usable precision approaches each providing straight-in minima to separate suitable runways.” So if my hypothetical airport had both an ILS 09 and an ILS 27 approach, and the airport were reporting 400 ft ceiling, 1 statute mile visibility, and winds from 180 at 2 knots, would it be a legal alternate? (I grant in advance that flying when your best available alternate is forecast to be 400-1 is pretty questionable regardless) .

Even if 09 and 27 count as two runways, requiring at least two usable straight-in precision approaches gives some insurance: if the wind shifts suddenly to 270@15 knots (but the visibility stays strangely poor), you can still fly the ILS 09. If a small fog bank drifts over the threshold of 09, you can still fly the ILS 27. On the other hand, requiring precision approaches to two separate runway surfaces gives a bit of extra insurance. If someone does a wheels-up landing and closes down 09/27, for example, you can still land on 18/36 (unless, of course, the wheels-up plane came to rest right at the runway intersection).

Because of this second point, it seems that many instructors and examiners have passed on by word of mouth that separate suitable runways means separate suitable runway surfaces — i.e. runway 09/27 and runway 18/36, not runway 09 and runway 27. However, I have not yet succeeded in finding any formal definition, and the AIP and other Canadian publications (not to mention pilots) use the word runway both ways. In real life, this situation is extremely improbable — very few Canadian airports have ILS approaches from both directions to a single runway surface, and those that do are big airports like Toronto/Pearson and Montreal/Trudeau with lots of other precision approaches available.

Still, what’s a runway?

Michael Oxner has a posting about the weather capabilities of Canadian ATC radar, and mentions that many pilots do not have weather radar on board.

That’s true — weather radar is pretty expensive — but a lot of us do have on-board lightning detection through a Strikefinder or a Stormscope. A mature thunderstorm cell will produce a lot of electrical activity and a lot of moisture, so either lightning detection or radar will help you avoid the real killer weather. The differences have more to do with comfort and convenience.

For example, weather radar can detect heavy rain showers without associated electrical activity. Heavy showers typically have moderate turbulence associated with them, and while that’s not very dangerous, it can make a serious difference for passenger comfort (and for yours, if you have to spend an hour cleaning up vomit after the flight). My Stormscope gives me no warning at all of embedded ACC or TCU without electrical activity. While I have no commercial experience, I imagine that if you are flying passengers in a charter or commuter air service, weather radar is a pretty obvious choice. Weather radar also allows you to check different angles — I’m not sure how easy that is to use, but at least in theory, it should be possibly to figure out how high up the weather goes. Lightning detection is strictly two-dimensional.

On the other hand, lightning detection works well even on the ground when the plane is surrounded by buildings, trees, or larger planes, so it’s possible to switch on the masters and take a peak at the weather around the airport before starting the engine. Weather radar can give a lot of false positives, since it cannot distinguish heavy rain from a thunderstorm; with a Stormscope (and a strong stomach), you can avoid some unnecessary diversions. Finally, a Stormscope can detect electrical activity in the early stages of a developing thunderstorm cell, when the cell might not be producing strong returns on a radar — again, this would be a case of avoiding moderate turbulence rather than preventing an in-flight breakup.

In an ideal world, you’d want both. For passenger comfort, I think I’d do better with radar, but the Stormscope is at least enough to let me fly in summer IMC, when I’d be uncomfortable relying solely on ATC (here’s why). Given that it would be fairly straight-forward to overlay Environment Canada’s weather radar on ATC displays, I find it interesting that Michael reports that Nav Canada has decided to show lightning strikes instead. I guess the rain-vs-lightning debate isn’t over yet.

Update #1: preliminary info from the TSB on lightning, touchdown point, and thrust reversers

Update #2: Environment Canada noticed the crosswind as well

Other aviation bloggers have posted sensible things about yesterday’s accident at Pearson, mainly slapping down the media and (even worse) retired pilots and other aviation experts who should know better than to speculate wildly. Like everyone else, I was glued to the TV and Internet, hoping, but not daring to believe that anyone — much less everyone — could have survived.

Still, while speculation is a bad thing, neutral information is not. Since the Canadian TSB does not publish preliminary reports, I thought it would be worth publishing what factual information I can find.

METAR

I downloaded the 19Z, 20Z, and 21Z METAR cycles from ftp://weather.noaa.gov/data/observations/metar/cycles/. The 19Z cycle had a single METAR report for CYYZ:

CYYZ 021900Z 22007KT 4SM +TSRA BKN050TCU BKN080 24/23 A3003 RMK TCU6AC1 CB ASOCTD SLP168

The 20Z cycle had a whole series of METAR reports, mostly within a few minutes of the accident (which occurred at 20:03Z):

CYYZ 022004Z 34024G33KT 1 1/4SM +TSRA SCT015 OVC045TCU RMK RA2SF2TCU5 CB ASOCTD

CYYZ 022000Z 29011KT 4SM +TSRA BKN051TCU BKN140 23/22 A3002 RMK TCU6AC1 CB ASOCTD LTGCC VIS LWR SW-NW 2 SLP164

CYYZ 022004Z CCA 34024G33KT 1 1/4SM +TSRA SCT015 OVC045TCU 23/ RMK RA2SF2TCU5 CB ASOCTD

CYYZ 022004Z 34024G33KT 1 1/4SM +TSRA SCT015 OVC045TCU RMK RA2SF2TCU5 CB ASOCTD

CYYZ 022004Z CCA 34024G33KT 1 1/4SM +TSRA SCT015 OVC045TCU 23/ RMK RA2SF2TCU5 CB ASOCTD

CYYZ 022020Z 34024G33KT 3SM +TSRA FEW015 OVC040TCU 23/ RMK SF2TCU6 CB ASOCTD

All the reports agree that the wind was from 340° true (350° magnetic) at 24 knots gusting to 33 knots. Visibility at the time of the accident was 1 1/4 statute miles, improving to 3 statute miles by 20:20Z. Ceiling was 4,500 feet AGL, lowering to 4,000 feet AGL by 20:20Z. All report severe thunderstorms and rain, with towering cumulus and cumulonimbus cloud associated. The temperature was 23° celsius, and the dewpoint was the same or very close. (Update: Environment Canada has mentioned the crosswind.)

The 21Z cycle had just one METAR, again, for CYYZ:

CYYZ 022100Z 18013KT 8SM -TSRA BKN055 BKN140 22/19 A3004 RERA RMK SC5AC2 CB ASOCTD FU ALF SLP171

By 21Z, the wind had diminished and backed to 180° true and visibility had improved, but there was still thunderstorm activity. Note in the remarks “FU ALF” (smoke aloft), likely referring to the smoke from the burning plane.

Runway and Approach

According to the Canada Air Pilot, the ILS/DME 24L approach at CYYZ has a decision height of 250 feet AGL (797 feet MSL) with an advisory visibility of 1 statute mile or runway visual range (RVR) of 5,000 feet. The runway heading is 237° magnetic, or 227° true. Relying solely on the METAR, which includes observations take some distance from the runway threshold, the flight landed with a tailwind of between 8 and 11 knots and a crosswind of between 23 and 31 knots.

Update #1: Friday 5 August

On Friday 5 August, Real Levasseur, the lead investigator for the accident from the Transportation Safety Board, shared some preliminary information with the press:

• There is no evidence that the plane was struck by lightning.
• The plane touched down further down the runway than usual, at least for that aircraft type.
• The thrust reversers were working properly.

Note that information from the flight data and cockpit voice recorders were not yet available to Lavasseur at the time of the press conference.

I decided to learn WML today (used for web browsing on cell phones and other small wireless devices), and METARs and TAFs seemed like a good, simple project. If you’re using a WAP-enabled device (i.e. your cell phone’s browser), try visiting this URL:

http://www.megginson.com/wap/aviation.wml

Type the ICAO identifier of an airport into the text field, and you should see a METAR and TAF for the airport, if available (not always the latest, unfortunately; I need to find a better source). This is more for fun than real flight planning, but feel free to play around. There are for-pay services that provide a lot more, including weather maps.