Sunday, July 31, 2005

Goodbye Mike

Someone I know died this week, flying. Searchers found his body today.

I don't have much to say about that, but I don't have much to say about funny radio conversations or IFR regulations today either.

Saturday, July 30, 2005

Weather to Approach

While looking for something else in my ancient FAR/AIM, I discovered that CFR 135.225 prohibits American part 135 aircraft from conducting an instrument approach to an airport unless "that airport has a weather reporting facility." That made me do a big double take, because Part 135 is commuter, on-demand, and mail delivery, the sort of thing that goes everywhere at any time in Canada.

I don't think that difference is a bad thing. As someone commented, the United States has a lot of instrument approaches, far more than Canada. Canada has more uncontrolled IFR flight. I cross-checked two pages of the index of a Canada Air Pilot (approach plate book) with the online aviation weather and found thirteen that have no reported weather. There's nothing prohibiting on demand flights from operating into those airports.

CFR 135.225 goes on to forbid an approach unless "the latest weather report issued by that weather reporting facility indicates that weather conditions are at or above the authorized IFR landing minima for that airport."

There is a condition that forbids Canadians to attempt an approach, but it only applies if the airport is equipped with an RVR (runway visual range) transmissiometer, a thingy that reports how far along the runway the pilots can expect to see in the touchdown zone. David has already blogged on the conditions under which the approach ban applies, so I'll just say if the touchdown zone RVR is below 1200 feet, and you have not declared in advance the intention to conduct a missed approach (for training purposes), you can't proceed to the FAF. Not a lot of airports have RVR.

Twelve hundred feet is a quarter mile, half of the half-mile advisory visibility that applies to most ILS approaches. I like that better than banning an approach as soon as the reported weather is at all below minima. Weather observations are always an approximation. If it's just below minima I'd like to have a look. Sometimes localized phenomena make a big difference.

I can remember approaching an airport that is served with no weather, but it is controlled and has an ATIS (does that count for CFR 135.225?) It was early morning, and the tower had just opened. The sky was clear, with calm winds and the runways were clearly visible. The ATIS indicated a visibility of an eighth mile in fog. I looked again and saw that the control tower was enshrouded in a little bubble of fog. It literally was the only fog on the airport. A VFR aircraft would be forbidden to enter the control zone. An IFR aircraft couldn't conduct a visual approach (because of the low reported ground visibility), but could request a contact approach, as it is governed only by flight visibility and a reasonable expectation of reaching the airport by visual reference to the ground.

Once on the ground, with that reported eighth mile visibility, you'd be forbidden to depart, because take-offs are governed by visibility. Fortunately it was a temporary phenomenon. Maybe they turned off the fog machines in the tower.

Friday, July 29, 2005

Clam Chowder

Someone has tired of IFR rule dissections, and sent me an e-mail request for funny things pilots say on the radio. You have to realize that funny things pilots say on the radio are about as funny as funny things people yell in bars. That is, generally not that funny except that they were unexpected. So if you're reading this expecting it to be funny, you're going to be disappointed.

I had just departed an airport on a nice day and was still on the departure/arrival frequency. Another aircraft, male voice, called up for a clearance into another airport in the area. We'll call his flight "Wolverine 53" as I have no recollection what it was, and wolverines are probably as amusing as the story. Unless they attack.

ATC: Wolverine 53 cleared left visual runway 33 Lake Airport.
Male Pilot: Left visual 33, thank you, that gets me a bowl of clam chowder.
Female Pilot: Are you sure you didn't want to give us the right visual?

Apparently the crew had a clam chowder-based bet on which downwind they'd be assigned. Another:

ATC: Panda 12 keep the speed up, faster traffic in trail.
Pilot: [fake Scottish accent] We're givin' it all she's got.
ATC: Aye aye Scotty.
ATC: Panda 12 you can reduce the warp speed now.

Star Trek references are pretty common in aviation.

Thursday, July 28, 2005

Contradictory Information

Comments on my last entry suggest that I have misinterpreted the IFR rules of my nearest neighbour. I don't feel too badly about this, as I know my own country's rules well, don't hold a US instrument rating, and don't currently operate IFR in US airspace, but I'm getting some contradictory information from some experienced US pilots, and I want to straighten it out.

Old Blind Dog, who has been flying aircraft of all sort for a lot longer than me, contradicts what I wrote in my last post, saying "the restriction is applicable to part 135 or 121 operations but not to part 91 corporate or private operations" and "RVR is only controlling prior to commencement of approach" and "For operations under part 91 I can commence the approach no matter what the reported weather."

I'm looking at Part 91 of the Federal Aviation Regulations. (As I admitted earlier my copy of the FAR/AIM is from 1998, so I welcome corrections of fact as well as interpretation.) According to 91.101, Section B applies to "aircraft within the United States and within 12 nautical miles of the coast." No mention of specific commercial operations.

91.175 (c) states that "no pilot may operate an aircraft, except a military aircraft of the United States, at any airport below the authorized MDA or continue an approach below the authorized DH unless --"

Three necessary conditions follow: stability of the approach, visibility, and visual reference. The visibility condition reads, "(2) the flight visibility is not less than the visibility prescribed in the standard instrument approach being used."

The next bit is "91.175 (d) Landing No pilot operating an aircraft, except a military aircraft of the United States, may land that aircraft when the flight visibility is less than the visibility prescribed in the standard instrument approach being used."

91.175 (f) Civil airport takeoff minimums appears to confirm David's comment that part 91 aircraft have no legal take off limitations at all.

I'm glad that what the experienced guy is telling me corresponds more closely to my idea of safety than my interpretation of the FARs, but I want to know how to get what he is saying out of what I'm reading. And yes, I know the joke about what happens if pilots trying to understand regulations, but I'm willing to risk it.

Wednesday, July 27, 2005

American Visibility Limits

In response to my posting about minima, John commented in part:

In the US, if you can see the runway environment or the approach lights, you can descend down to 100' above the touchdown zone (lower under certain conditions). Visibility is what determines whether or not you can land, so visibility is much more than advisory - it's the controlling factor.

Maybe I misunderstood your comment or things are different in Canada?

I looked up the American regulation John cited (14 CFR 91.175 - albeit in my 1998 FAR/AIM) to discover that things are different in Canada.

Canada forbids descent below decision height before "the required visual reference has been established and maintained in order to complete a safe landing". That part isn't different. Both countries require "at least one of the following visual references for the intended runway [to be] distinctly visible and identifiable to the pilot" and follow that sentence with almost identical lists of ten things that constitute the runway environment.

Canada United States
(a) the runway or runway markings(ix) the runway or runway markings
(b) the runway threshold or threshold markings(ii) the threshold
(iii) the threshold markings
(c) the touchdown zone or touchdown zone markings(vii) the touchdown zone or touchdown zone markings
(d) the approach lights(i) the approach light system [...]*
(e) the approach slope indicator system(vi) the visual approach slope indicator
(f) the runway identification lights(v) the runway end identifier lights
(g) the threshold and runway end lights(iv) the threshold lights
(h) the touchdown zone light(viii) the touchdown zone lights
(i) the parallel runway edge lights(x) the runway lights
(j) the runway centre line lights

The order and grouping of the items differs between countries so in the table above I have rearranged the American list (but kept its original numbering) to make it easier to see that it matches the Canadian one. The differences like "runway identification lights" and "runway end identifier lights" are trivial. The major difference is in what I elided on the starred item.

The American reference adds "except that the pilot may not descend below 100 feet above the touchdown zone elevation using the approach lights as a reference unless the red terminating bars or the red side row bars are also distinctly visible and identifiable." This makes the American regulation more restrictive in theory, because it adds a second decision height. What are the calls in a two crew environment in this situation? "Minima, approach lights only" then "100 feet, in sight" or "100 feet, no red bars"?

That was surprising, but not nearly as surprising as the rest of the section. Ther is a fundamental difference in our instrument approach rules.

In Canada, that's it. You're set up to land, you reach decision height, you see the runway, you are allowed to land. The approach plates give both a decision height and a visibility, but as the AIP says, "published landing visibilities are advisory only. These values are indicative of visibilities which, if prevailing at the time of approach, shouold result in required visual reference being established. They are not limiting and are intended to be used by pilots only to judge the probability of a successful landing when compared against available visibility reports at the aerodrome to which an instrument approach is being carried out." They are also used in determining a legal alternate, but that is at least one posting in itself. (The Americans won't believe our alternate rules.)

American pilots are restricted by visibility and ceiling, all the way to landing. They are not allowed to continue an approach if the flight visibility is less than the one written on the plate. They are, in a separate paragraph, forbidden to land if the visibility is below the prescribed one. I'm looking at the plate for the ILS/DME rwy 27 into Rock Springs, Wyoming. DH is 200' agl and the prescribed visibility is 1/2 mile. Lets say a pilot has completed a difficult approach through visibility of half a mile in snow showers, and is now over the runway approaching the flare. The snow shower intensifies and blowing snow drops his forward visibility such that he can see just four runways lights ahead on each side. He's perfectly aligned, the airplane is under control, but he knows that runway lights are 200' apart so that his visibility is now below a thousand feet, nowhere near the requirement printed on the plate. Is he seriously required to go around? Seems crazy to me. At that point in the flight the pilot should be able to decide what is the safest course of action.

By the way, American official publications actually use the word "minimums." It's not just an artifact of illiterate chief pilots, to go with "it's" being used as the possessive. Canadian pubs always use "minima" but I've just spotted the phrase "Category D circling minima applies." It appears that "minima" is becoming a singular noun for pilots.

I will save the radical difference between the Canadian approach ban and the American stipulations of CFR 135.225 for another post.

Tuesday, July 26, 2005

Good Job

RTO in Ottawa last night. Sounds textbook. My kudos to the crew. I hope Transport doesn't manage to come up with something they did wrong in those fifteen seconds.

I'll bet there's a moment where you think, "huh? this isn't the simulator!" And then after the aircraft has come safely to a stop, and the fire department assured you that you're not on fire, and the passengers confirmed all safe, you get to actually appreciate all those hours of hell in the simulator.

Update: a slightly longer version of the story.

Minima and Missed

When you are approaching an airport through clouds, you aren't allowed to just point the airplane roughly in the direction you figure the airport is and go down until you find it, because of the great risk that you figured wrong and there's a hill there instead. You follow an instrument approach procedure, using cockpit instruments to aligning your airplane with radio signals that indicate a safe path to the runway. You have a bit of paper called an approach plate that tells you what frequency the radio signals are on, what morse code identifier verifies that you have tuned it correctly, where the obstacles are and useful things like that.

You descend, following the instructions on the approach plate, until you either find the airport or you don't. This differs from the rough method described in the first sentence of this post, because the approach plate defines very precisely when you're supposed to give up on finding the airport, and what you must do next, called a missed approach. For a precision approach, the plate lists a decision height. When the airplane reaches that altitude one pilot looks out the window and describes what he sees according to a template laid out in the company SOPs. It might be "Runway in sight, twelve o' clock" or "minima, no contact."

Minima is of course the plural of minimum, and I never really wondered what plural minima were involved before. There's the minimum altitude you're allowed to descend to and ...? There's only one instance of two thousand four hundred twenty-three feet, and you're at it. I might speculate that "minima" is easier to say than "minimum" and people might think it was singular, as data and agenda have come to be. Good speculation, except that I've seen company ops manuals where the standard call is "minimums." So it's factually and grammatically incorrect, and any English speaker can tell it's plural. What is this mysterious plural? The British even call final approach (the straight in towards the runway flying, right before you land) "finals". There's not more than one of them. It's one final approach.

I guess it's just pilot grammar, the way the word "missed" has become a noun. "The missed."

A missed approach isn't just for minima. If at any time during the approach you have messed up badly enough that you won't be configured properly for landing (because of airspeed, flaps, gear, captain's pants, etc.) or cannot be certain that you will remain within protected airspace (in this case protected from trees, radio towers and mountains) you must go missed even before you reach the decision height. Your company SOPs and your company culture (quite possibly diametrically opposed) will detail the conditions that require a missed.

Monday, July 25, 2005

Bikes Don't Fly!

But I do fly. Again. You're all correct. I haven't forgotten how to fly. After half an hour I forgot about my back. After two hours I felt my back again, but it didn't stop me from doing my job.

I encountered some performance increasing windshear on landing today. That means that the wind was blowing more strongly towards me close to the runway than it was at a higher altitude, so that as I descended, I suddenly hit a stronger headwind.

Now, if you were riding a bicycle, working against a headwind would decrease your progress. The same is true of an airplane: the stronger the headwind, the slower the progress over the ground. If an airplane that can fly at 200 miles per hour in still air had a wind from behind at 20 miles per hour, its speed over the ground, called groundspeed, would be 220 miles per hour. The groundspeed is indicative of time to destination: stronger headwind, longer time to get there, stronger tailwind, get there more quickly.

So why did I say that suddenly encountering a stronger headwind was performance increasing? Despite all the encouraging comments attached to the last entry, an airplane is not a bicycle. It doesn't push backwards on the ground as it travels, so as it flies through the air, it doesn't know about the ground going by. Although the steady tailwind described above carries the airplane over the ground at 220 miles an hour, the speed of the air passing by the airplane is still only 200 miles per hour. And for the purpose of how the airplane handles in the air, how it feels to the pilot, and whether it is within its safe speed limits, the speed relative to the air, the airspeed, is what matters.

In preparation for landing, the pilot slows the airplane to the correct approach aispeed, based on weight, flaps and other considerations. If there is a strong headwind, this may give a slower than usual groundspeed, but that's okay. We actually would like to have the slowest possible groundspeed for landing, because it makes it easier to brake to a stop. We can't go too slowly, however, because the lift requireed to hold the airplane up depends on airspeed. The approach speed is therefore a compromise between maintaining a high speed through the air and obtaining a low speed over the ground.

Now back to the bicycle, because I have devised a way to continue talking about bicycles. A bicycle speedometer measures how fast the wheels are going around, i.e. the rate the bike is passing over the ground. Imagine that you rigged a bicycle speedometer worked by having a little propeller on the front, so that it measured not how fast you went over the ground, but how fast you went through the air. And imagine that you came around a corner, and suddenly encountered a blast of headwind. Your propeller would momentarily speed up to the sum of your speed PLUS the speed of the oncoming wind, and your eyes would water with the sudden onslaught of air. The added air resistence would then quickly slow you down, but for that moment, the speed of the bike through the air is increased.

So the same thing happens to the airplane encountering a sudden increase in headwind. Momentarily the air flowing over the surfaces is increased. This increases lift, resulting in a combination of a lower descent rate (or a climb) and increased airspeed. And that's what happened to me. I found myself higher than I wanted to be, and faster than I wanted to be, but with a long runway, so the landing was long and ugly but safe.

I should have realized that the winds aloft were lighter than the winds at the field and expected it, but I didn't come straight in: I'd been turned for that runway at quite short notice. I'll do a better job tomorrow.

Sunday, July 24, 2005

Saturday, July 23, 2005

FOD Detection System

The term FOD, pronounced to rhyme with God, describes any debris present on the airside of an airport. Most people say it stands for "Foreign Object Debris." That abbreviation is a little weird, because most of the debris isn't from objects foreign to the runway, but completely native objects like airplanes and trucks. It consists of rivets and pieces of tires, and wheels off suitcases and garbage from the fuel truck driver's lunch. I notice the British say it stands for "Foreign Objects and Debris", which makes a bit more sense. I believe that FOD originally stood for "Foreign Object Damage" -- the result of debris entering a jet engine, puncturing a tire, or otherwise coming into contact with aircraft. When an engine suffers FOD, the engine is said to have been fodded, and so by back formation the culprit became FOD. Re-expanding the acronym required a slight change, but now FOD more often refers to the debris than the damage. We say "FOD damage" like "ATM machine" or "SIN number." Misnamed or not, the stuff is dangerous.

Debris on the runway from another aircraft caused the tire blowout leading to the Concorde crash. Ice and foam insulation striking the shuttle during launch caused the damage that lead to the Columbia shuttle disaster. FOD shortens the life of aircraft engines, and can cause an engine failure at the worst possible stage of a flight: just before the airplane has enough speed to fly. At best it's expensive. Reportedly it costs the industry $4 billion (I'm not sure if those are Canadian billions or British billions, as the number is on a British press release citing an American study) per year.

I can think of a couple of FOD-related problems that I've had. I landed an airplane, parked it went inside to prepare for the next flight, and came back out to discover a flat tire. Maintenance found a screw in it. I'm glad the tire picked up that screw on the way in and not the way out, or it could have made the landing more exciting. Another time, after my flight I taxiied an airplane to maintenance for a routine inspection. Later maintenance called me back to ask what happened to the propeller. I thought they were kidding but I came to look and saw a chunk the size of a quarter missing from the side of one blade of the propeller. They showed me a corresponding smaller ding on the opposite blade. That's very common, as the propeller spins fast enough to hit the object more than once. In that case they were able to "dress" the propeller, filing out the damage and making the propeller the right shape again while keeping it within tolerances. I remember being able to feel the curve of where it had been filed every time I ran my hand over that propeller on subsequent preflight inspections.

Vancouver International Airport has become the very first airport in the world to purchase a radar-based system for detecting FOD. In one test, a bolt was placed in the groove of a runway. Radar eight hundred metres away detected the bolt and the system output GPS coordinates accurate enough for airport personnel in a pickup truck to locate and remove it, in the dark, in under five minutes. The product is made by a British company called QinetiQ and it has also been tested at Heathrow and JFK in New York. Apparently it can detect all manner of FOD including gravel and animals.

I guess that means the end of conversations like

"Tower, we just saw some debris on the runway as we rolled out."
"Roger that, where was it?"
"Just past the intersection with delta, on the right there, about twenty feet from the edge."
"What did you say it was?"
"Not sure, looked like maybe some glasses."

And then someone can't resist getting on the radio and suggesting "martini glasses?" and all hell breaks loose. And that's before they have to close the runway for the guy in the pick up truck to drive up and down looking for the thing. "Just past delta" could cover half a kilometre. As soon as a crew says, "we saw something on the runway" the tower can reply "got it on Tarsier."

Tarsier is named after a cute mammal that has excellent night vison and can swivel its head 180 degrees. No word on whether it can dig a burrow. I think it's arboreal.

Friday, July 22, 2005

At a Canadian Airport . . .

At nine in the morning, on an otherwise fine day, a Boeing 737 ingested an eagle into its starboard engine on short final, with no time to prepare the passengers for an unusual landing. The pilots momentarily lost control of the airplane and it struck the runway hard, causing numerous passenger injuries, a few fatal. The cabin filled with smoke as air from the burning engine entered the system. Many passengers suffered smoke inhalation and some were burned. The flight crew informed the tower of the situation which was quickly becoming obvious. Flight attendants quickly evacuated those passengers who had survived the impact. Fire, ambulance and airport operations arrived on the scene to extinguish the fire, remove the remaining passngers, and to triage and treat the victims. By eleven am everyone had been removed from the scene and all the participants, including most of the dead ones, were enjoying complimentary doughnuts and sandwiches--because the whole thing was an exercise staged to test the readiness of airport and emergency personnel.

The so-called Boeing 737 was actually a smaller airplane, long out of service and with the engines and seat cushions removed. The eagle existed only in imagination. The exercise started with the airplane parked on one of the runways, closed for the purpose of the exercise. The smoke inside was from a smoke machine at the back, run by a guy crouched behind it in the back galley. The smoke outside was from strategically placed smoke bombs around the chocked wheels. "Fire" was indicated by red and yellow streamers. The flight crew and flight attendants were real, from a locally based airline, but a different one than had provided the disused aircraft. The firetrucks, ambulanes and emergency personnel were real, and the passengers were volunteers, carefully briefed, made up, and doused in fake blood. The only passengers who were denied doughnuts were mannequins, posing as dismembered passengers, and they were cleaned up and returned to their boxes before noon.

The smoke machine quickly dropped visibility inside the aircraft to less than arms length. The flight attendants used well-rehearsed shouted commands to urge the passengers out of their seats and out the exits, but the screams (interspersed with giggles) of the 'frightened' and 'injured' passengers made the commands hard to hear. The scenario provided a sobering insight into the conditions that would exit during a true emergency evacuation. Coughing, choking and staggering the passengers emerged onto the tarmac. One wandered off the runway and 'died' in the grass beside it before emergency responders located him. Others wandered confused and interfered with the firefighters. Paramedics moved the passengers away from the aircraft and into green, yellow and red triage areas, according to their injuries and symptoms. The tarmac was swarming with VIPs and observers in orange vests, and before long the Transportation Safety Board, and even a coroner showed up, to begin the analysis.

All the while, real airport operations continued on the remaining runways. A NOTAM had been issued informing crews that the firetrucks, police cars, ambulances, disabled aircraft and sprawled bodies were an exercise only, and they, I hope, found a reassuring way to convey this information to their passengers. I fear there is no good way to tell a planeload of pax that they are going to see an airplane crash scene out the right window as they land. The sooner you tell them, the longer they have to think about the fact that simulations occur because the real thing might happen. Some of them are going to mishear and think there is an emergency affecting their own aircraft, and some are not going to hear, and then think that what they are seeing is real.

An interesting detail: the popular Livestrong rubber bracelets introduced confusion into the triage process, as they were similar to the ribbons emergency personnel used to classify the injured into categories that determined the urgency of their medical treatment. I think if I had time, I would brief my passengers to remove any coloured wristbands, along with their glasses and other sharp objects, in preparation for an emergency landing. And if you want to get helicopter transport to hospital, keep a red wristband handy for such an occasion. Avoid wearing a black one.

All airports have periodic exercises to test readiness and response. They've had emergency simulations at my local airport, but nothing so elaborate, since nothing as large as a B737 operates into here, and operations aren't as complex. They'll dump a scrapped fuselage on a disused taxiway and call the fire department to come out, testing the system of calling them, communicating with them, and ensuring the guy whose job it is to open the gates does so. We still get to watch the firetrucks, but not to the level of excitement and detail that my friends saw in Operation Eagle.

Thursday, July 21, 2005

Ignition and Fuel Nozzles

I've learned a few more things about the PT6, and caught a few misconceptions, so I'm going to revisit the fuel nozzles and the problems it could dump on my beautiful PT6 engines. This entry will be a little rambling, and will repeat some things I said already, but I'll get the obligatory squirrel reference out of the way in the first paragraph.

I've already said that fuel enters a PT6A-20 engine through fourteen nozzles. When you push the fuel lever ON, a valve opens and fuel sprays out of all of them into the burner can. If the glow plugs (or igniters) are functioning properly this should ignite the fuel. PT6A-20 engines tend to be pretty hot at start, which shortens the time to overhaul of the engine, so for early models of PT6A-27 they tried dividing the nozzles into two banks of seven. The fuel supply splits, so that the first seven primary nozzles spray fuel as soon as the fuel is selected on, and the next seven secondary nozzles come online after about 38% NG. That reduced the start temperatures significantly, so in later PT6A-27 and all PT6A-34 engines (both approved, but flat rated to 620 SHP for this airframe) there are still fourteen nozles, but they are arranged as a group of four primaries and ten secondaries. Fuel comes from the fuel control unit through a flow divider. Two pressure cracking valves allow fuel to the primary nozzles at about 12% NG and to the secondary nozzles at about 38% NG. If the secondary cracking valve sticks closed, the engine will never increase above idle NG. That appears to suggest that if the primary valve sticks closed, fuel would not enter the burner at all, and that the cracking valve protects against a pilot who turns the fuel on without verifying that NG has stabilized at or above 12%. Perhaps they don't want us to know, lest we get cavalier with the fuel levers. (We're supposed to call them fuel levers, not condition levers, because condition levers are the ones with an OFF FLIGHT IDLE and GROUND IDLE setting, while these are just ON and OFF).

By the way I mentioned that you could tell if an airplane had igniters or glow plugs installed by checking to see if it had igniter switches. What I neglected to mention is that if it has engine igniter switches, then it contains glowplugs. If it has no igniter switches then it has igniters. There's even a reason for that, not just trying to make things difficult. Igniters or glowplugs are used at start and in turbulence or heavy precipitation, but they are used a little differently.

At start, any delay igniting the fuel could lead to a high T5 at start, and a greater power drain on the battery. Although the battery can be charged from the generator between starting one engine and starting the other, it's not recommended because, like charging your laptop every time you use it, it then doesn't hold as much charge as if it is deep cycled. So at start, whether you have glowplugs or igniters, you want to have them all turned on and as hot and ready as the girls in the telephone dating ads on TV. You don't need to think about that, because with the IGNITION (not the IGNITERS) switch set to NORMAL, both igniters or glowplugs are energized when the start switch is held in the LEFT or RIGHT position. (The person who stenciled the labels on the dashboard only had CAPITAL LETTERS).

During flight in turbulence or heavy precipitation the pilot ought to be concerned (while appearing perfectly calm on the exterior) that the engine could flame out. (That means the flame goes out, like a candle being blown out, not that flames shoot out: we refer to the latter as an "engine fire").1 If the engine flames out, we'd very much like it to relight instantly, so we turn on the igniters or one glowplug in each engine during flight in such conditions. We do this by flipping up the cover over the IGNITION switch and moving it to the MANUAL postion. But if we flew along through heavy rain for a long time with both glowplugs energized, they would degrade and not be as hot and sexy next time we wanted to start an engine. That's where the IGNITERS switches come in. They each have three positions: NO.1, BOTH, and NO.2. While using the glowplugs in flight, you select NO.1 for a period of time and then let the number one glowplugs cool down and select NO.2. I don't know what that period of time is.


1. Can someone who is informed, or at least opinionated about such things please advise me on the proper order of those last three punctuation marks. I think there may be national or at least transatlantic differences.

Wednesday, July 20, 2005

Volante One

I'm pleased to report that the insurance company gave me fair value for the remains of my car. I was already resigned to receiving a pittance for it, so was pleasantly surprised to be offered within $100 of my own estimate. While I was wishing out loud that I could replace the smashed out-of-production model car with an identical non-smashed one, someone helpfully suggested that I look beyond it and consider what the old car lacked. I immediately admitted that it had really lousy cupholders, and then after a few seconds of thought realized, "and it couldn't fly." My face lit up.

I remember the big stack of Popular Science magazines in my grandparents' basement. In the 1950s, people believed that the pace of technological advancement fueled by the war and the its cessation would continue unabated, bringing us flying cars, videophones, space travel, monorails, and every other acoutrement of the cartoon future. We do have camera phones, and the Volante One is a flying car that may be available as a homebuild kit in the near future, as opposed to when we all live in geodesic domes. Money (I may get some extra, as the other driver was held "at fault") is a weak recompense for pain and suffering, but a flying car might leave me feeling no pain.

Okay, I'm not likely to really commute to work in a flying car, unless it comes with rear deflectors, but I wrote to inventor K.P. Rice, pointing out the close connection between silver jumpsuits and flying cars, and suggesting with a grin that he sell the jumpsuits on his website to bring in some revenue and to give future Volante One customers something to start with. He has leafed through the same 1950s magazines, and had a surprise for me.

I have always been a Pop Sci fan so your remarks really hit home. Will take the silver suits under advisement. When I was in the Navy experimental squadron VX5 I tested a a silver flight suit as protection from the heat from your own nuclear bomb, post delivery. I used it as a fire suit for my crewman when first testing the Volante at El Mirage.

So there you have it, the first flight of the modern flying car was made by a person wearing a silver suit. Now all I need is a little bubble helmet. Oh, and a new car.

Tuesday, July 19, 2005

Escaped Comment

I started to respond to comments on the previous entry and my comment got so long I've decided it's tonight's blog entry. Back to the PT6 tomorrow.

I have seen the 'won't engage' problem actually caused by missing or broken teeth on the ring gear, just like anoynmous' friend, and also by the sticky starter arm. I have heard of the engage-at-shutdown technique, but not of the screwdriver-in-the-starter technique. I've never been that desperate. So far I've always managed to obtain a start by turning the propeller to a new position, although sometimes it takes a couple of iterations. Flying commercially there is a very strict line between work the pilot can do and work that maintenance must do. Usually you cross that line when you pull out a screwdriver, but again it's all about desperation.

I wouldn't hand prop an airplane if there were a guy or gal around in coveralls to fix it for me, nor would I attempt it on an engine too big for it to be a success. I have had training hand propping small aircraft. But even if you have had none, there are places in this country where the chance of losing limbs or worse to frostbite or polar bears are greater than the chances of losing them to a propeller. And I know the propeller would be quicker than the cold. Not sure about the polar bear. Do they play with their food?

Carb ice right at start only occurs when it's close to freezing. If it's three degrees and 100 per cent humidity (i.e. raining) this morning and you start sucking that air into the carburettor, the temperature only has to drop three degrees for ice to clog the narrow gap in the only-just-open-enough-to-idle carb throat, and there's not yet a big hot engine block to prevent that happening. In fact, the engine is probably still at minus two, from the overnight chill.

Typically with carb ice at start, the engine starts up completely normally, and then the power drops as the pilot continues with the after start checks. In cold wet weather I will apply carb heat immediately after start. The EGT won't be that high either, so sometimes it dies anyway, but that and opening the throttle more if power starts to drop should do the trick.

It can be amusing if someone is taxiing on a cold, wet day and making little power adjustments, not really paying attention as the power drops. By the time they check carburettor heat, the throttle is almost wide open, but the engine is still at idle because ice almost completely blocks it. They put on carb heat and the power roars up to a cruise setting.

Twenty-five thousand hours and no hung starts, eh? Encouraging. If people knew how many hours we spend memorizing, practicing, being tested on, and briefing each other on procedures we may never use, they'd stop worrying about whether we were fit to fly because our ties are on crooked.

Friday, July 15, 2005

Piston Engine Starting Problems

Looking at the various responses to the complexity of the non-automated start procedure for a turbine engine made me think some more about piston engines. These days, or at least a few weeks ago, I have been operating carburettor- equipped horizontally opposed piston engines. That prestart checklist includes ensuring the fuel is turned on in two different places and that the throttle is far enough open for air to enter the carburettor. I engage the starter with a key, just like a car, except that the lighter weight starter motor means that I have to hold the key in the start position for longer before I release it.

Here are some of the abnormal start possibilities for a piston engine.

If you engage the starter and nothing happens, that means there is no electrical power. So you either have a completely dead battery or a broken connection. (Or you forgot to turn the master switch on). You can jumpstart an airplane if it has a dead battery. You can also hand prop it, as you've probably seen in old movies, but if the battery is completely dead, an alternator won't be able to generate any electrical power. A generator will.

If you engage the starter and there is a clunking noise, but the engine doesn't turn over, there's power but it's not energizing the motor. The clunking noise is the solenoid completing the circuit that is supposed to supply energy to the starter motor. If all you hear is a clunk, then the starter motor is inoperative. Get a new starter motor. Hand propping will work in an emergency.

If you engage the starter and there is a whirring, grinding noise, but the propeller doesn't turn around, then the starter motor gear is not engaging properly with the ring gear on the propeller shaft. This can often be remedied by turning off the mags and magnetos then getting out and turning the propeller by hand to a new position before re-attempting the start.

If you engage the starter and the propeller turns around, but not very enthusiastically, and no start is obtained, you likely have a weak battery. It's like trying to start a car on a cold day. Try for 30 seconds or so, but give the starter a rest before trying again. This is best fixed preventatively: minimize electrical use after shutdown and before start. On cold days turn the propeller through by hand a few times to break up congealation (is that a word?) of the oil. Use extra primer, because less of the fuel evaporates in the cold. Use preheating if available. Jumpstarting is an option, and with some battery power remaining, the alternator will function to recharge the battery. Sometimes this is caused by some kind of binding of the crankshaft, which will alos become evident as you turn the propeller by hand.

If the propeller turns, and the engine fires but it coughs and doesn't catch well enough to sustain firing it may not have enough air or fuel for combustion. Or you might have just released the starter too soon. It might just be insufficient primer, or the throttle not open far enough. Or you can mutter "must be vapour lock" and wander around poking randomly at things, hoping it will work next time you try. You may have to beg access to a heated hangar if it's really cold.

If the propeller turns and the engine fires, but weakly, amidst puffs of black smoke, it's probably flooded. You may be able to smell fuel. There is too much fuel present to start the engine. You misdiagnosed the previous problem and overprimed. Pull the mixture to idle, start the engine and then push the mixture knob in after start. If it's really flooded, you may need to wait a while before attempting the start.

If the engine starts fine, then sputters to a halt, I suspect one of two things. The first is carburettor ice, especially if it's a rainy day, with temperatures above freezing but below ten degrees celsius. In such conditions I monitor rpm or manifold pressure as appropriate right after start and am prepared to put the carburettor heat on right after start. If it's not carb ice it's fuel supply, which could mean something missed in the prestart checklist, or an issue for maintenance.

Sometimes the engine starts but the starter doesn't disengage. If you've ever turned the key to "start" a car that was already idling, you know what that sounds like. Ugh.

At engine start it's important to monitor the oil pressure closely. If the oil pressure doesn't come up into the green arc within thirty seconds of starting, you have to shut down the engine and send someone in coveralls to investigate why.

There's also a possibility that while you're attempting to start the engine, it bursts into flames instead of starting. The pilot must surpress the initial impulse to run away and instead keep cranking the engine. If it starts, that will consume the fuel that was feeding the fire, and then after running for a short time, the engine can be shut down and inspected. If it doesn't start, then you keep cranking while you shut off the fuel supply to the engine in multiple places, shut everything down, and evacuate, bringing the fire extinguisher. Now you have the option of running away, or trying to put out the fire, depending on the comparative level of the fire and your bravery/stupidity. If any passengers are on fire, you should expend your firefighting abilities on them, not the aircraft. I've never had nor witnessed a piston engine burst into flames on start up. It must be quite rare, but it was the first emergency procedure I ever learned. Apart from that, the most dramatic way to cook a piston engine is to not notice that the oil pressure has failed to increase. You can write off an engine costing tens of thousnads of dollars by running it without oil for a few minutes. Burning out a starter motor or running a battery flat are minor inconveniences in comparison.

Thursday, July 14, 2005

Not in This Airplane

There have been some hot sunny days lately, but I'm still remembering flying instead of flying. Dave has a great hot weather story today story for you, so here's one of mine.

I was sitting and sweating, last in a queue of airplanes on a hot day. While we waited on the taxiway, the wind was undergoing its midmorning shift to a new direction. The tower decided to use another runway, more into the wind, and began directing incoming traffic, and airplanes just starting up, towards that runway. Those of us already in line to take off on the original runway were cleared to take off, one by one. By the time the line worked through to me, airplanes that had started up after me were already taking off, on a runway conflicting with mine, so I had to wait even longer for a takeoff clearance.

I was piloting an airplane of the same type as many of our fleet, but which mysteriously would not climb well. It is heavier than average, but not the heaviest. It has had a new propeller, an engine overhaul, and I believe a paint job (which includes having the wings taken off and put back on again) but still climbs like a, well I was going to say 'like a dog' but I've seen dogs climb, and they can get up a hill fairly sharply, tongue lolling about. If this airplane were a dog it would be a fat dog, with a bad leg and have its tongue lolling about just standing still. It's just a slow climber.

Non-mysteriously, a high temperature, a heavily loaded airplane and a tailwind all degrade climb performance. I was sitting pretty close to maximum weight. With the wind now favouring the other runway, I had a tailwind. The poor airplane had to be going forward down the runway at twice the speed of the wind, in order to develop only the same lift as an airplane sitting still at the beginning of the into-wind runway. And because it was so hot, it had to be going faster in still air just to achieve the same pressure differential throught he wing as on a cold day. By the time the airplane was airborne it was further down the runway than it would have been with a headwind or no wind, and further down the runway than it would have been in cold temperatures. Even airborne, the tailwind and heat continued to affect its climb angle, meaning that it gained less altitude than normal for each mile it went forward.

This wasn't dangerous, it was a VFR (looking-out-the-window) flight with no obstacles and a very shallow required climb gradient. I'm patient. I held the climb speed and waited. It would get to altitude eventually. Then tower amended my departure instructions and asked me to make a turn that would require me to overfly another control zone. Or rather the controller expected me to overfly it. I had no such illusions. I accepted the clearance, then requested a radio frequency change. I needed to make a radio call before cutting through someone else's airspace.

There had been a broken layer of cloud in the area earlier, and there were still a few clouds at that level, lurking in a layer of haze. The controller thought that I was unable to climb because there were too many clouds above me to see. It's his responsibility to broadcast advisories on the sky condition in the immediate area of his aerodrome, so before approving the frequency change, he asked me to confirm that I was unable to reach that altitude.

I answered, "Not in this airplane." I spent the rest of the day hearing about how I'd made everyone laugh with that line, from other pilots who knew me and the airplane.

And while I'm reporting funny things from the radio, once I heard a pilot leaving a frequency report, "Just for your information, there's a dead whale here." I'm assuming he meant in the water below, and not plummeting through his altitude like in the Hitchhiker's Guide to the Galaxy.

Wednesday, July 13, 2005

Starting the PT6

When I discussed the PT6 engine function, the compressors were already spinning, the fuel burning steadily in the annular combustion chamber, and the combustion products blasting out through the compressor and power turbines. But if the flow of air is required to contain the fuel, the burning fuel is required to turn the compressors, and the compressors are required to create the flow of air, how does all this start?

Early models took advantage of the wheel-like properties of the turbines and employed squirrels to achieve the initial NG. Unfortunately the high T5 immediately after lightup rendered each squirrel a single-use component, leading to rapid squirrel depletion at popular landing sites1. A more sustainable starting technique was called for.

The current technique, rumoured to have been proposed by a member of the squirrel pool2, is common to many turboprops. Each engine is already equipped with a generator, and as a generator and a motor share most of the same parts, it's not especially cumbersome to use the generator as a starter motor. Adding to the efficiency is the fact that electrical power isn't required to be generated during start, and a starter motor is not needed after start. So the airplane is equipped with two starter-generators.

The first step is to spin the compressor up to sufficient speed to maintain airflow. The preferred way to achieve that is using someone else's electricity, connecting the airplane to an external power supply. It should be able to supply 28 V at a minimum of 800 A. It should also fit the plug receptacle in the airplane. There's nothing in the manual about lugging around one of those plug converter things, so perhaps the international standardization of electrical equipment is better for aviation than for toasters. I know some of these airplanes were sold in England, but perhaps, like English toasters, they were delivered to the consumers without the plugs, and the end users had to install their own plugs3.

It's also possible to start the engines using the battery. Either way the procedure is the same, with the exception of the position of one switch, the EXTERNAL/BATTERY switch. Completion of the rest of the prestart checklist leaves the power levers at idle, the propellers selected to full fine (but lack of oil pressure means actual blade angle will be either feathered or latched at one degree, depending on the modification), DC electrical master on, the boost pumps on, but the fuel levers off. The warning light for the generator should be illuminated, because, with the engine not running, the generator is not functioning.

The pilot grabs the start switch and engages it. The generator light goes out, and the engine starts turning.evidenced by rising NG and Np. The oil pressure too should be rising, but will probably not reach 40 psi, because more air flow is required to seal the bearings. While the pilot is monitoring NG, Np and oil pressure, she should also sneak a look at the battery, because if output drops below 17 V, there may not be enough juice to start the second engine without recharging. Oh, and don't let go of the starter switch yet.

Once NG has stabilized at a value of 12% or higher, the fuel lever should be moved to on. If NG stabilizes below 12%, then don't turn the fuel on. If external power is being used, it may stabilize as high as 23%. The fuel is turned on by moving a lever with a yellowish-orange knob on it. It's next to the propeller lever. Don't let go of the starter switch.

Activating the starter switch not only energizes the motor that spins the engine, but provides power (from an auxiliary battery, so there's enough juice for both) to energize the glow plugs. A glow plug is a hot electrical coil, sort of like the cigarette lighter that occupies a car 12 V receptacle when you first get the car, before you take it out and plug in something more useful, like your GPS or your cellphone. There are two glow plugs in each engine, at the eight and four o' clock positions, if that's important to anyone. When the fuel enters the engine and contacts the glow plugs, the fuel should start burning. We call that engine lightup or lightoff. Don't let go of the starter switch.

Whoosh! The fuel starts burning. You can't hear the whoosh, so you have to watch the internal turbine temperature shoot up, the oil pressure increase, and the NG continue to rise. If T5 doesn't increase, then the fuel failed to light. If T5 skyrockets or NG stabilizes around 30-40% you have a hung start and the air isn't flowing through the engine properly. In either case shut the fuel lever off, but don't release the starter switch. Let it run for another 10 seconds.

Assuming NG rises as it should, it should stabilize at idle: 48% if the propeller is feathered and 52% if it's not, higher at higher altitudes. T5 should drop back from a peak value of not more then 1090 degrees celsius for no more than two seconds. Then you may release the starter switch. The generator light should turn on again.

If there's enough battery power to start the second engine, the generator should be left offline, the second engine started immedately. Otherwise, the power lever should be advanced to idle (whatever that was) plus 15% and the generator on the running engine may be selected on to charge the battery. It's sufficiently charged when the battery charge current load is 0.4 or less. Both generators must be offline to start the second engine.

After start the power should be advanced to idle plus 15% before the generators are selected on, and not reduced below that until the generator load is 0.5 or less. I'll blither later about things that could go wrong during start, but that's all I have to say for now. Except for the footnotes.


1. The bits about the squirrels may not be completely accurate.
2. In fact, the bits about the squirrels are total lies.
3. This hypothesis is only slightly more likely than the bits about squirrels.

Tuesday, July 12, 2005

Congratulations Clumper

I've been monitoring some student pilot blogs lately. It's a way of reminding myself how far I've come, whenever I get frustrated by how far I have to go. The first one to pass her checkride is clumpinglitter (forgive her that odd login name: you try finding a good unused name on livejournal). She is now the holder of a brand new private pilot licence, issued by the FAA, and valid for flying American-registered single engine land airplanes anywhere she pleases.

Security Loophole

A month or so ago, I blogged about an odd e-mail from WestJet. What I didn't mention at the time was that in addition to the risk of being discarded as spam, the e-mail invited me, and anyone who can read a URL, to step through a huge security hole into other people's data. I didn't put it on my blog, for obvious reasons, but thanks to Douglas I managed to forward this information to the WestJet IT department, and they've decided to look into it, "because we want people from all backgrounds to have the opportunity to work at WestJet, and wouldn't want some to be disadvantaged."

Ichneumon isn't WestJet, but it's only one step away. When I started this blog, I mentioned that had no illusions that WestJet would call me up to fly for them on account of my witty insights and keen attitude. That's why today it was the WestJet IT department on the phone, not the chief pilot. Just a little surreal. WestJet and I suit each other so well. There won't be that many candidates who can answer the standard, "what could you contribute to WestJet?" question with "I've already contributed by ..."

Blogging isn't magic, but it concentrates a person's energies on what she wants. Concentrate in a positive way on what you dream, and do something about it every day, and seemingly magical things happen.

Monday, July 11, 2005

Masters & Slaves: Three More Things About Propellers

This is probably my final chapter on the propellers. They do have every right to be this complicated as they are one of the most important systems on the airplane. These last three things are autosynchronization, blade latches, and prop ties, which I just thought of so I had to go back and change both this sentence and the title of the post.

The propellers, as previously and repeatedly mentioned, spin around. And they make noise. And noise is actually a wave in the air, not an up and down wave like on the ocean, but a squishy back and forth wave, like a pulse going along a slinky that is stretched out along the floor. If two slinkies interfere you get a tangled mess of coiled metal, and your mom comes and solves the problem with wire cutters, so you end up with more slinkies, only shorter. (I know this from an unsuccessful slinky race with a friend when I was little. The stigma of having a shorter slinky than all the other kids haunts me still.) If different sound waves interfere, you get a pattern of sound called a beat. So if your two propellers are going at exactly the same speed, you hear a lot of normal propeller noise. If they are going at slightly different speeds, the captain hears a beat, and I hear a lot of propeller noise, and the captain complaining. So the correct way to deal with this is to randomly tweak the propeller levers until the captain gets irritated enough that he syncs the propellers himself. But if I'm lucky, there's an autosynchronization package installed and it will do it for me.

In this airplane the autosynchronization package has magnetic speed pickups (like the kind on bicycle speedometers except way more expensive) on each overspeed governor, an actuator on the right engine and a trimmer on the right primary propeller governor. The pilot need only get the propellers within 2.5% of one another, and then the difference between the two speed signals controlls the actuator to trim the right governor to achieve an rpm identical to the left. Thus the left is the master and the right is the slave. If the propellers go significanly out of sync, the system is inhibited, so that an engine failure or overspeed condition on the left engine won't affect the right. In such a case, as well as during take-off and landing and on the ground, the pilot is expected to have the autosynch turned off.

You can understand from the description of the propeller system that when the engine is shut down that the oil pressure will drop and the propeller blades will be pushed into the feathered position. When the airplane is restarted, max rpm will be selected, so the blade angle will reduce, but there will momentarily be a sudden lurch forward, undesireable for some operations. So if you postion the propeller levers opposite the zero thrust markings, corresponding to approximately a one degree blade angle, the latches will engage and hold the blades in that position. After start, if the propeller lever are kept behind the zero thrust line, when the propellers reach about 30% Np the latches disengage.

My friend has a Jeep convertible (are all Jeeps convertible?) and won't park it anywhere without attaching a steering wheel lock. This airplane does not have an antitheft device. Heck it doesn't even have keys. The so-called control "lock" is just a bar that keeps the rudder, ailerons and elevator from flapping about in the wind. It's also important that the propeller not be allowed to spin in the wind, for two reasons. One is that a spinning propeller could whack someone in the head, and that might convert them into a unsatisfied customer, or a non-productive co-worker. The second reason is that there is no oil pressure when the engine is shut down, so a spinning propeller equates to the front half of the engine spinning with no lubrication, shortening its life. So this is where the prop ties come in. There are different variants of this sophisticated system. Some use a loop of rope around a propeller blade, fastened down to something. Some run a bungie cord around a blade and hook it into an opening in the cowling. In nations where slaves are cheaper than rope, you could station a slave on each side of the airplane to hold the propeller still. Whatever works. But then you'll need extra slaves to load baggage. or you could hold the left one yourself, then the left would be the master and the right the slave, just like the autosynch.

Sunday, July 10, 2005

Big City Scam

Little bit of time, little bit of computer access, bit of a blog post.

This story made its way to me, because of course my friends think I know everything about every aspect of airplanes, even though the airplanes are incidental to the story.

It begins on the sidewalk in a busy downtown area of a large Canadian city. A man in his late twenties is standing at the corner, right where people stepping off the crosswalk will pass him, holding a creased and pencil-marked tourist map while speaking with a north England accent into a cellphone. He's saying something like, "no, it's no use, I can't get there."

The scam target stops and looks, says something like "Can I help you at all?"

The man looks up, says, "hang on, there's someone ... I'll call you back."

"Where are you trying to get to?"

The man sighs and thanks the target, and apologizes that he's very tired and irritated but do you know where there's a tourist centre that's open? "I've been all the way down to this one here on the map, but they are closed for plate glass repairs and I went to the community policing station over here but they were the very most unhelpful people I've ever met. I even got the constable's card so I can lodge a complaint. This is such a horrid city, no offence intended, but I really hate it here."

"What kind of tourist information were you looking for?"

"Well you see I've got three hours before my flight and I've made a mistake, it's--we're not all this bloody stupid in England--but I'm very tired driving back from Winnipeg. I knew my flight was at one forty-five but I seem to have got the wrong day what with the time zones and all, so my flight left yesterday. Bloody Air Canada says that I have to pay a date change fee."

"Ah yeah, probably $500 or something?"

He says they're charging twenty-four pounds, and something about having already used all the travellers cheques with the rental car and he's trying to cash a sterling cheque but none of the banks will accept it. You see where this is going, but it's entertaining enough that you stay tuned.

"Which banks have you tried?"

"I've tried all of them."

Much more on the nature of banks, and Air Canada and how tired he is, and if he misses this flight he'll have to pay $1200 to change the ticket, and his grandmother and the rental car. The police told him to take his laptop to a pawn shop, and this city is so big compared to Torquay. He leads up to it as though he's just suddenly got the idea--it must be fun doing scams, if you have no conscience--that maybe, "Oh I can't believe I'm doing this..."

At this point the target decides to test the story a little. "I'll tell you what. I'll call Air Canada and put the change fee on my credit card. You can write me a cheque, and when you get home send me some chocolate."

"Oh that would be marvellous--but they said it has to be cash."

"And you're running a scam. Air Canada takes credit cards. Good bye." And the target walked away.

The scammer actually had the nerve to chase after him and say, "I can't believe you think I am trying to trick you! I wasn't raised that way." He couldn't produce the air ticket--it's "in the glove compartment of the rental car." My friend extricated himself from the renewed conversation, and kept walking.

A phone call to the community policing station revealed that an annoying British man hadn't been there ranting about an air ticket, so the officers were informed where he was, what he looked like, and what he was doing. There's not much they could do. It's not illegal to stand on street corners and pretend to talk on a cellphone. It's not even illegal to tell strangers stories in the hope they will offer you money. Perhaps he could be done for tax evasion: I doubt he's paying GST on his takings. I wonder how much he makes in a day.

Reading his story in a blog, you can see holes you could fly a 747 through. But it's a clever scam. Here's some of the points:

  • The scammer doesn't approach anyone so
    • he can't be arrested for harrassing people
    • people think, 'I approached him, so how can it be a scam?'
    • he speaks only to people who are predisposed to help strangers.
  • He appeals to the person's desire to defend the honour of their city
  • He badmouths organizations that many people dislike (Air Canada and the police)
  • By saying he has been to the police and they have failed to help him, he rules out calling the police for assistance
  • A built-in excuse for mistakes or inconsistencies in the story--because he's so tired from his drive
  • He never asks straight out for the money or promises to repay it, to make it harder to charge him with fraud.
  • Like all scams, the details required to check his story are inconveniently located
  • Any questions the mark asks him become part of his database, so he can refine his technique

So moral of the story: help people if you are so inclined, but never give cash to strangers.

Friday, July 08, 2005

Preventing Speeding

I described the propeller governor, and the way it lets more oil into the hub if the propeller is going too slowly and cuts off supply if the propeller is going too fast. You now know that if the propeller isn't going fast enough, and the governor allows too much oil pressure into the hub, that the beta reverse valve will keep blade angle above eleven degrees. And you know that if the beta reverse valve fails, and the blade angle drops below nine degrees, that the beta backup valve will close to keep the blade angle no lower than about nine degrees. Pilots always ask of every system, "what if it fails?"

If the pilot valve in the governor were to fail open in cruise, the beta system would prevent the blades from going into beta, but the oil pressure would still hold a propeller in the eleven degree blade angle position, even though it was turning too fast. If the propeller speed continued to increase the overspeed could destroy the propeller, and produce vibrations (or shed propeller blades!) that damaged the airframe. One governor is not enough to guard against this highly undesirable situation, so there is also an overspeed governor.

Normally, the valve to the overspeed governor remains closed, and oil bypasses it going straight from the primary governor to the propeller dome. But the overspeed governor is very similar to the primary governor, with flyweights and a pilot valve, but for the overspeed governor, the flyweights will open its pilot valve if the propeller speed is too high, defined as 101.5% Np. (That is, the propeller is spinning at a rate that is 101.5 per cent of an arbitrary number that doesn't mean anything). The open valve dumps oil pressure through the overspeed governor back to the sump, bypassing the hub and allowing the propeller blades to return to a coarser pitch, slowing their rotation.

A test switch for the overspeed governor temporarily resets the overspeed governor limit to 70% Np. The pilot tests both overspeed governors at once by pressing and holding the test switch with a high rpm selected, and the power at idle, then advancing both power levers to a position that supplies enough torque to achieve full rpm. The Np should not rise above 70%.

Think that's enough? Yeah, right. If the backup system doesn't have a backup system it's not much of a backup system, is it? The overspeed governor is backed up by the NF governor which slows the propeller by cutting off the fuel supply if Np exceeds the selected value by more than 6%. As max Np is 96% (I warned you that 100% was an arbitrary, meaningless number), that means that with the propeller levers full forward, the NF governor takes action at 102% Np. Interestingly, during reverse thrust operations, the NF governor limit is reset to 5% below the selected value--always 96% in reverse--so that the NF governor keeps Np down to 91%. This ensures that the primary governor never cuts off the oil supply, leaving control of the blade angle to the beta valve.

And if all that stuff fails, the pilot (not one of the pilot valves, the woman in the seat) can pull a lever to shut off the fuel to the engine. When inflight engine shutdown emergency checklist is complete, she'll turn to her co-pilot, roll her eyes and say, "It's your turn to tell the passengers they're not going to where they thought they were going." I hate doing those PAs.

Runway Numbers

I just watched a movie where Kurt Russell is a mild-mannered intelligence operative who has to land a B747 because the bad guys have killed the pilots. You know, the sort of thing that happens every week in Hollywood aviation.

He's initally cleared to land on "runway one ell north." That's right, not "one left" but "one ell" plus the designation north. If that doesn't strike you as odd, let me explain how real (non-Hollywood) runway numbers work. Under normal circumstances, a runway is named after the direction you would be going while landing on it, rounded to the nearest ten degrees. So at Downsview, the runway runs northwest-southeast, such that aircraft landing on it would be on a heading of 150 degrees or 330 degrees, depending on which end they were facing. They might be landing on runway 15 (pronounced "one-five") or 33 (pronounced "three-three"). The entire piece of pavement is runway 15/33. If there is more than one runway at the same airport, with the same heading, then they are designated as left and right, and centre if necessary. The pavement markings say 26L and the pilots and air traffic controllers say "two-six left." Okay, most of you knew all that, but I'm trying to be accessible here.

Now when a large international airport has too many parallel runways to get by with "left" "right" and "centre" (or "center" for the Americans), they simply pretend that some of them have a slightly different heading. So Toronto has a 24L/06R and a 24R/06L and another runway, exact same heading of 237 degrees magnetic, but instead of renaming 24R to 24C, and giving the name 24R to the new runway, they pretend that the rightmost runway has a slightly different heading, and call it runway 23. That's allowed. I'm even going to allow for the possibility that a large airport could come up with a slightly non-standard scheme to disambiguate half a dozen parallel runways. But "one left north"? Runway one left already runs north-south, if there were two of them, you'd want an east and a west, not a north and a south.

Our intrepid hero doesn't land on runway one ell north. He comes in too high and is then afraid to turn around. He decides to land at another airport--on runway two-six. Apparently he can turn through 110 degrees just fine. He does land on that runway, although you don't see the arrival end as he touches down. The final touch is that as he runs off the other end (the runway is too short) you see that the other end of the runway is marked with a seven. I guess the runway curves.

Oh and one of the odd things about Americans is that they leave the initial zero off of runway designations. The above would be runway "zero seven" in Canada. I always think single digit runway numbers look very rural and hicktown. Maybe it's because middle-of-nowhere airports don't have parallel runways so a three digit designator like 33L is indicative of a big city. By induction, only one digit implies a town so small I expect dogs to be sleeping in the middle of Main Street.

Thursday, July 07, 2005

Beta Backup Valve

I would forgive you for thinking that reverse and backup meant the same thing, but in this case they don't. The previously discussed beta reverse valve prevents the propellers from going into beta or reverse blade angles unless the pilot wants them to. The beta backup valve is a backup system, designed to save the day if the beta reverse valve fails.

The beta backup system consists of two microswitches, one on the propeller feedback ring, tripped when the propeller blade angle drops below 9 degrees, and one on the power levers. The information I have suggests that there is one microswitch shared between the two power levers, and that sounds vaguely familiar, but I can't remember whether it's in the left or the right power lever. Perhaps there's one for each power lever only in the newer models. At any rate, if the power lever handgrip has been rotated, selecting the beta range, then no power is available for the beta back up solenoid, so the signal from the 9 degree microswitch is ignored.

If the power lever has not been rotated, and the feedback ring indicates that the propeller blade angle has dropped below 9 degrees, then 5 A of current flows from the right DC bus, energizing the beta back up solenoid, snapping the beta backup valve closed. No oil reaches the beta reverse valve, so its failure to close doesn't matter. Oil supply to the propeller hub has been cut off. This, as you know by now (repetition is an excellent way to learn things) results in the springs and counterweights in the hub driving the blades to coarser angles. Once the angle is greater than 9 degrees, the beta backup solenoid snaps open again. If the beta reverse valve is still not doing anything, the blades are driven back towards fine and the scenario resembles children continuing to play after a stalemate in chess, with the two kings chasing each other back and forth, endlessly in and out of check.

The microswitch that senses that a propeller angle is less than 9 degrees also illuminates a blue beta light for that propeller, so the indication to the pilots of a beta reverse valve failure is a flashing blue light. If beta has been selected deliberately, and the propeller angle remains at less than nine degrees, the blue lights will remain illuminated and not flash. (Some models have amber lights instead of blue, but I don't care for them).

As part of a normal landing, after touchdown a pilot might twist and pull the throttles into reverse, using the power of the engine to slow the airplane. Immediately afterwards, the pilot would likely advance the throttles to idle in order to taxi but the blades would takes a few moments to regain the 11 degree idle pitch. To prevent the beta backup system from detecting this situation "ooh, grip not twisted, but blade angle less than 9 degrees!" and cycling unnecessarily, there is another relay in the system. This beta disarm relay is activated if a grip is twisted and the propeller blade is less than 9 degrees. It remains activated until the grip is released and the blades return to idle pitch. While activated it inhibits the beta backup solenoid and illuminates an amber BACK UP DISARMED light. Therefore, the pilots can expect that amber light to go on briefly immediately following a landing where reverse thrust is used. I assume that the BACK UP DISARMED light illuminates any time the throttles are not twisted yet power is not available to the beta back up solenoid.

Of course there has to be a way to test this mess, and there is: a BETA RANGE TEST switch for each power lever. Lifting the test switch bypasses the power lever microswitch, so that power to the beta back up solenoid is maintained even when the handgrip is twisted. So lift the switch, twist the handgrip and pull it back. The blue beta lights will flash as the beta backup solenoid keeps the blade angle cycling back and forth through 9 degrees, and the BACK UP DISARMED switch will flash on and off as it thinks power is being removed from the back up solenoid without the power grips having been twisted.

Beta reverse valve failure is very rare, as it's a simple mechanical system. It's more common for the beta backup propeller microswitch to fail, removing the backup protection or sending a false alarm cutting off the oil supply and necessitating an engine shutdown. I heard that my target company has obtained permission to remove the beta backup valve and fly without a net for the beta reverse valve, but I don't know that for sure.

And if you think these propeller lights are complicated, wait until I tell you about the Gulfstream One.

Wednesday, July 06, 2005

Hard Enough Awake

According to a mention at the end this article, there has been a case of someone trying to fly a helicopter while asleep. As Greybeard has been relating, they are tricky enough to fly while awake. I hope the somnabulist in question didn't manage to start it up. I don't think I've ever done anything in my sleep more exciting than mumbling.

They Should Put It on the ATIS

All the air traffic controllers for the different airport frequencies sit together in one room in the control tower. Pilots can't see them, just hear them. Approaching the airport we first listen to the recorded message about which runway is in use and what the winds are and whether one of the runways is closed because the fire department is practicing on it. The message is called ATIS, which stands for Automated Terminal Information Service, or something similar. When we have the information, we call the tower controller who gives us instructions and landing clearance, and then after landing we call the ground controller for permission to move on the ground. Sometimes there is an inner tower and an outer tower, or a south ground and north ground frequency, or a controller dedicated to clearance delivery before taxi. They trade jobs from time to time through the day, so that no one has to work the busy or the dead job all day. We address them by their positions, so "Winnipeg Ground" or "Halifax Tower" and then we quickly abbreviate that to just "Ground" or "Tower." It's not like we don't know what airport we're at.

On this particular occasion, we've called tower and have been cleared to the circuit, to approach the runway for landing. We descend and turn right base, towards the runway, following the controller's instructions for altitude and traffic sequence. The tower controller is so friendly, and he has such a sexy voice, that I and the other (also female) pilot remark on it.

She's flying, and on final I'm giggling.

"What?" she says.

"Oh, I was just thinking of daring you to do something."

She lands, and then I relate my dare. And then she takes over the radio and does it.

"Ground, Alpha Bravo Charlie, request taxi to the terminal--and can you tell us if Tower is single?"

Apparently, he's happily married. I hope he and he wife had a good laugh over that. I know everyone in the tower must have. We figure the ATIS should include the marital status of the controllers.

Tuesday, July 05, 2005

Beta Reverse Valve

I asked what would happen if torque were too low to drive the propeller at the selected rpm. The answer, based on the information I gave you, was that that the governor pilot valve would open fully, maximizing pressure in the propeller hub, driving the piston all the way forward, and the blade angle to -15 degrees. As you can imagine, this would not be desireable. If the pilot reduces the power to land, she wants to descend to the runway going forward, not in reverse. This doesn't happen, and it is the job of the beta system to ensure it doesn't.

Three rods extend from a feedback ring at the rear of the propeller to the outside of the propeller dome. When the propeller dome moves forward past the position corresponding to +21 degrees blade angle, the linkage from the propeller dome begins to pull the feedback ring forward. This movement is transmitted to the beta reverse valve, which begins to close. At less than an 11 degrees blade angle, the beta reverse valve will fully close, not allowing oil pressure through the beta valve. The governor pilot valve will be wide open, but that won't increase oil pressure until the drop in oil pressure allows the springs and counterweights to drive the blades to a coarser pitch, opening the beta reverse valve, and holding the blades at the ideal idle angle of 11 degrees, until torque increases. As a result, when the propeller is in beta range, the power levers control fuel flow and blade angle.

If the pilot really does want to put the propellers into reverse, useful for parallel parking (I am NOT kidding: you really can parallel park these guys), the beta reverse valve can be deactivated. The pilot does this by twisting the handgrips of the power levers, kind of like you may have to push a button or apply downward force on your car gearshift to shift into reverse. Twisting the handgrips allows the beta reverse valve to open, so that blade angles less than 11 degrees can be obtained. As the power levers move aft from that position, they stop controlling the fuel flow and control only the blade angle through a null zone between 11 and -2 degrees blade angle. As the power levers move back past the position that selects -2 degrees, they start increasing fuel flow--normally you push power levers forward for more power--so that pulling the levers further back allows you to increase backwards thrust, either to reverse into your parking space, or to slow the airplane down more rapidly after landing. A blue (or possibly amber, depending on the model number) light turns on to show that the propeller is in beta mode, with a blade angle less than 9 degrees.

The manual says, literally on the first page, even before the table of contents, that you must never ever ever do this in flight under any circumstances. This of course means that loads of people do, but that the company isn't going to be held responsible if bad things happen.

Monday, July 04, 2005

Hoping for the Hollywood Ending

I found out from a pilot who has flown with the company for years that my name is on the Ichneumon flight training schedule already.

I'd been telling myself not to get my hopes up, not to count those unhatched chickens, that quite possibly I'd only been called to the groundschool because they wanted to have a look at the insane lunatic who belonged to the resume I submitted. And all along, or at least since that flight schedule came out, they've been planning to train me on their beautiful airplanes.

I know, you guys expected it all along. But remember, this is real life, not a Hollywood movie. If it were a Hollywood movie I'd be checking my watch against the run time of the movie saying, "okay, she's had the setback because she refused to lie, she's had the resolution of that, she's had the major injury, now is there enough time left in the movie for her to lose this job and get a better one, or just enough time for her to recover from the injury by pure tenacity and then get this job?" In real life sometimes the ending is really stupid and unfulfilling. I hope this is a good movie.

I have had so many disappointments. Someday, from the security of a good job, I'll tell you about them.

And in other news, space shuttle commander Eileen Collins is getting ready for a launch into outer space, but she chickened out and came back down the stairs after standing in line for a waterslide. There's one aviatrix who is not going to do anything that isn't safely under her control, even in the glare of world media while the fate of the US space program hinges on this mission. Good for her.

Sunday, July 03, 2005

Twist and Spin

If you've ever had one of those balsa wood airplanes you wind up with a rubber band, you're familiar with the concept of the propeller. It's a twisted thing that corkscrews through the air to pull the airplane forward. The ones on the toys, and most training airplanes, have fixed pitch propellers, meaning that the twist is permanent, fixing the blades at one blade angle, such that if you advanced it by screwing it into a wall of jello it would always go the same distance forward (its pitch).

This is the one-speed bicycle of propellers. It has the advantage of being relatively cheap, lightweight and easy to operate. You have to push extra hard going up hills, and your feet have to go round and round really fast to power down hills, but you don't need to mess with shifters or worry about the chain falling off because you got it in the wrong gear. But once you've mastered the art of staying upright and can manage to take one hand off the handlebars, it's nice to have the ten-speed.

When I talked about the critical engine I mentioned that the amount of thrust increases with the angle of attack of the propeller to the relative airflow. It should also be understandable that the greater the angle of the propeller, the more torque that is required to push the propeller around. Small blade angles are called fine pitch and large blade angles are coarse pitch. Fine pitch is like a low, hill-climbing gear on a bicycle. You don't have to push as hard to turn the pedals, but you have to turn the pedals around more times to get the same distance. Substitute "propeller" for "pedals" and you have a propeller set to fine pitch. (The Americans call this low pitch, but we'll ignore them for now, so they don't confuse us). There are more influences, such as the fact that the propeller slips in the air, making its effective pitch less than its geometric pitch and that more energy is lost to drag at high rpm and high airspeed. Shake it all down together and we find that at take-off it's best to have a fine pitch propeller turning really fast and in cruise it's better to have a coarse pitch propeller turning at a more moderate rate. Due to the inconvenience of switching propellers midflight, we use constant speed propellers.

At the front of my propeller hub is a piston, attached through mechanical links to the propeller blades. As the piston moves, driven by oil pressure, the mechanical links twist the propeller blades, the way you might turn an oar in an oarlock to change the angle the oar blade met the water. If oil pressure is high, the piston moves full forward, twisting the blades past full fine (flat) to the mechanical stop at -15 degrees (yes, driving the airplane backwards). If oil pressure is lost, accidentally, or by feathering, springs force the piston back to the full aft position, twisting the blades to +87 degrees, the previously described feathered position. By the way, the angle is measured 30 inches out from the hub, about three-quarters of the way to the tips. Propeller diameter is eight feet, six inches.

The oil is the same oil circulating throughout the engine, but it's stepped up in pressure from 85 psi to 385 psi by a pump built into the base of the propeller governor. The oil is constantly circulating through the propeller hub, because it is allowed to leak out and be replenished. Here is the path the oil takes: through the pump, past the relief valve, through the beta back up solonoid valve, through the beta reverse valve, through the constant speed governor, past the overspeed governor, into the propeller hub, out through the transfer sleeve to the reduction gearbox, back to the rear of the engine for filtering and cooling.

I will start with the governor, because under normal circumstances in cruise flight, you can ignore all the beta and overspeed stuff and pretend it's just the governor and the hub. The pilots set the amount of torque the engines produce, based on the best compromise between fuel economy, how fast they want to get there, and not blowing up the engine. Then they set the desired propeller rpm based on altitude and company practices and personal whim. Notice that: they don't select the blade angle, they select the rpm. The governor sets blade angle in order to achieve the rpm the pilots asked for. As the torque is constant, the governor can decrease rpm by increasing the blade angle or increase the rpm by decreasing the blade angle. If the propeller is onspeed, the oil circulates just as described above, through a partly open pilot valve in the governor. The amount of oil allowed through the governor equals the amount allowed to escape from the hub, so the pressure and thus blade angle remains constant.

If the propeller slows down, the governor corrects it. The governor is driven off the propeller reduction gear, so it slows down, too, and as it does, flyweights on it swing inward. Those flyweights are levered such that when they move in, they open the pilot valve further, so the slower the propeller, the more oil pressure is allowed through the pilot valve, to drive the propeller to a finer pitch, so it speeds up. Likewise, if the propeller is overspeed, going too fast, the flyweights swing outward, closing the pilot valve and cutting off oil pressure to the hub so that the propeller moves to a coarser pitch, slowing it down. When the pilots set the selected rpm, they are adjusting a speeder spring on the governor that determines how fast the flyweights have to be going to close the pilot valve. There's a nice diagram of a propeller governor, and some excellent tips on propeller care in this AOPA brochure.

If you've managed to follow this, then you can tell me what the constant speed governor and the propeller hub piston would try to do to the blade angle if the torque selected were insufficient to turn the propeller at the selected rpm, even with the blades set to full fine.

Saturday, July 02, 2005

Red versus Blue

Original Air Canada pilots and ex-Canadian pilots are still divided, and the conflict might lead to a strike this summer. I'm not even going to comment.

The "Original Air Canada" and ex-Canadian factions at Air Canada are also known as "red" and "blue", after the colours of the airline livery before rebranding ran amok. Air Canada still has the red maple leaf, but for a while the tails were blue and red. Now they've gone to a kind of minty green all over, which doesn't show up that well in a photograph.

By the way, when we talk about the colour of an airplane, we talk about the parts that aren't white. Often that's just stripes, the nacelles and the tail. Most airplanes are mostly white. I guess I absorbed that fairly earlier on, as I learned to fly in a nominally orange airplane, that was really a white airplane with a couple of orange stripes. I often confused friends by telling them "I fly the orange one" while I pointed at a predominately white airplane.

Back when Canadian Airlines was Canadian Pacific Airlines they had real orange airplanes, but I think the orange-to-blue transition is mourned by no one.

In other news, Boeing has selected James McNerney Jr. as its new President and CEO, as the old one was having sex with another executive. It isn't immediately clear why this relationship was grounds for dismissal, as she didn't report to him and there has been no implication of intimidation. I think it's just the usual American paranoia about anything remotely related to sex (link not worksafe).


This is supposed to be an aviation blog, not an injury blog, but you fly the trips dispatch gives you, and this week mine is physiotherapy. In a word: wow.

My knowledge of physiotherapy comes from TV where the injured party spends long panful hours on complicated gym equipment, inspired by the Littlest Hobo, a budding love affair with the physiotherapist, or the need to seek vengeance or win the national championship. Plus I know someone who broke his leg badly, and he screamed during physio, said it was worse than the original injury. So I was looking forward to this with some trepidation.

My fears were utterly unfounded. It was like pilates: seemingly meaningless, easy exercises with a lot of emphasis on breathing and how you hold your stomach muscles. Much of what I have been doing instinctively was right: using my diaphram and posture to protect the injury site, while making gentle movements to retain and promote flexibility. Most of the exercises invlve lying on my back, doing "the dead bug". Flip a hard shelled beetle over and you get the idea, only less frenetic.

I feel fantastic right now. Still some pain, but the natural endorphins from the exercise are better than anything the doctors have prescribed. My ultimate aim is to achieve a level that will reduce my suceptibility to further back injury. Oh, and unlike the movies, my physiotherapist is not in the least hot.

Friday, July 01, 2005

Fuel Control Unit

John asked me a question about the fuel control unit:
Just curious, does the PT6 engine you're describing use P3 bleed air to modulate the fuel supplied fuel control unit? Or are the power levers directly connected to the the fuel metering?

Each power lever is mechanically connected to its respective fuel control unit and the primary governor. Bleed air does not participate in fuel metering.

Jimmy Little asked

Can the engine controller on something like the PT6 automatically detect the conditions caused by the stuck-open bleed valve and issue the appropriate warning? It would seem relatively easy to correlate fuel consumption, temperature, and such to detect this, or at least warn that it's an issue.

I need to point out that this is not a computer-controlled engine. It predates the invention of the microprocessor. So yes, there is temperature detection and yes, correlation between high NG, high T5 and torque lower than selected by the power levers might well indicate a stuck open bleed air valve, but this information isn't digitised. The computer that correlates the data is the pilot's brain, hence the need for all this training. (Not to say that the folks flying the computerized airplanes don't need training: they need to know "what it's doing now.")