Flight Plan for the Ride

I described a flight test recently and what I didn't include was a measure of the artificiality of the flight test scenario compared to real life IFR. While there are some pilots who are making very short hop flights in busy airspace, and you might follow a missed approach with a different approach to the same facility, typically real life gives you quite a length of time between after takeoff checks and setting up for the approach. The flight test is an artificial situation where the cruise portion of a typical flight is pretty much eliminated. This, combined with the stress of a test, makes it much more intense.

Most examiners have a little routine that they follow. They aren't secret routines: they'll typically tell you what they are going to ask for at the time you book the ride. They want you to do well, and knowing the parameters of the artificial situation gives you the opportunity to prepare. Flying a trip like that "cold" can be very challenging. A friend gave me one such trip to fly in the simulator. It had a missed approach followed by an approach to an airport almost in line to the missed and I was laughing at how frantically I had to work to get up and down and on the glideslope of the second runway.

Pilots flying out of a particular airport have local knowledge tricks that can help. For example, maybe tracking inbound on this NDB there is a shoreline effect that means if you seem to be right on, you will lose the track crossing the shoreline about two miles before the beacon. Maybe there is no DME to this VOR, but if you have DME hold on for this other VOR, not involved in the procedure, station passage will be at 20.7 DME. Maybe that yellow silo is exactly 2.2 nm from the threshold, so stay inside it during your circling approach. It helps to know the typical holds you'll be given, even when they aren't published holds.

I now know the routines of a few different examiners in a few different provinces, but I thought it would be fun to know more. I have a request. Send me (e-mail, please) the typical "IFR Flight Test" route assigned where you fly, or where you did your last renewal. People in other countries can play too, but I might have outdated plates and not know the local regulations. I think in the USA private pilots don't have to renew their IFR ratings, so they might not have done a flight test in a long time.

I'm going to fly these routes in Microsoft Flight Simulator with minimal preparation, tell you all the ways I screwed up, then try them again until I do them perfectly and come up with hints. I'll have a good way to test my skills on down days at work, and we'll eventually develop a database of hints for likely flight tests.

Museum Pieces

After my PPC ride I had some time to kill before going home, so I went to see the air museum that was visible from the run up area. When I arrived they were only open for another hour and a quarter, so once I had extricated myself from the friendly volunteer at the entrance I went straight to the static display in the yard that was open for the public to go on board. It was a Boeing 737-200, hardly an antique, but it's always fun to have a close up look at an airplane.

"Would you like to go on board?" asked a volunteer at the foot of the gangplank.

"I think I'll do a walkaround first," I said. The engine cowlings were open, and I've never had a chance to walk under the belly of a B737. The volunteer turned out to be a retired mechanic for this type, able to name and describe every component and line on the entire aircraft. I wish I could tell you what I learned, but I'm afraid it was just such a barrage of information entering my flight test tired brain that I just let him talk. A few details that come to mind ... the logo lights are in the wingtips, shining back on the logo painted on the tail ... the hydraulic pressure in the lines automatically applies the brakes during gear retraction to stop the spinning of the mainwheels ... there are big buffer pads in the nose gear bay to stop its spinning ... there was a very interesting system under the nose that automatically repositioned a splash guard with nosewheel retraction to protect it from dirt and water, but I guess my brain was full by then, because I've forgotten what it was the intake for and how it worked.

The walkaround complete, I went up the boarding stairs. There was a crowd of children talking to a pilot in the cockpit, asking the standard questions like, "how do you remember what all the switches do?" I hung back and inspected the galley until the kids moved on, then I slid into the right seat and turned to talk to the seated pilot, and another who was standing behind him. "Hey!" I said, "We know each other."

"I know," he said, and remembered my name. I met him when I was a student pilot. He's retired now from the airlines and this was his first day volunteering here. The mechanic came up from outside too and the four of us spent the next forty-five minutes telling outlandish pilot stories. There's the one about the pilot who's coming in to land in a Stearman, while ATC is waiting for a B737 to start rolling on the runway. "Do a 360 for spacing," says the air traffic controller to the pilot of the Stearman, not specifying left or right, because it doesn't matter. The pilot does a 360. He pulls back, flies straight up, over and back level again on the original heading. A 360 degree vertical loop. It's too good a story to look into too closely.

They have lots of stories about a time in aviation that I hadn't considered before: the dawn of mandatory radio communication. There was a whole breed of aviator who paralleled the silent movie star. They could fly airplanes, but didn't necessarily have any verbal communication skills. Or even usable English. The generation after them had to switch mid-career to computerized cockpits. I wonder what new paradigm I will have to join in my career.

At the end of the day they closed up the airplane. I learned how to stow the electrically powered gangway on a B737 (it runs off the hot bus, like my airstairs and cargo area lights). As they hummed away into place I shook my head at the idea of a B737 being in a museum already. "It doesn't feel like a museum piece," I said.

"Neither do we!" chorused the retired pilots. And they didn't, either. They felt like friends.

On the bus on the way home I laughed, catching myself wondering why so many towns around here were named after NDBs. Of course the NDBs were all named after the towns, it's just that I'd spent the last few days tuning the NDBs and have never been to the towns.

PPCed Again

The flight test is early the next morning. I don't feel ready for the in-air portion. I'm still getting in finger trouble with the autopilot, but I've become used to minimal training.

I prepare all the paperwork for the exam flight: operational flight plan, weight and balance, Nav Canada flight plan, weather briefing, performance calculations, company exams and my licence and medical. I put the last two into the same folder with the paperwork, so they're all ready and I don't have to go through the little ritual of digging them out of the licence wallet when he asks for them.

I haven't met the examiner yet, but I'm told he's easy going, won't kill me for details, just wants to see me in control of the airplane and knowing where to go. I'm there at seven to make sure the airplane is preflighted and ready to go. He arrives early and as we sit down to do the paperwork he asks me "What kind of day are we going to have?" I brief him quickly on forecast and actual ceilings, visibility, winds and precipitation. It's likely to be VMC throughout the flight, except perhaps if they give me a hold at 7000'. He's satisfied with that, and doesn't ask any questions. I wonder for a moment if that was supposed to be an ice-breaking question and not a request for a weather briefing. Ah well, my friends and relatives know better than to ask me what the weather is like, if they don't want to hear that it's 13 degrees and there's a 5000' stratus layer with at least ten miles visibility in light rain. The examiner tells me to go and file the flight plan now so it's ready in the system when we want to go.

When I return, he appears to have already looked through the folder of paperwork and satisfied himself with my preparation. I've found that preparing for a renewal flight test is a matter of showing that you're prepared. Overpreparing guarantees minimal ground questions. Which kind of makes sense because a pilot who already holds the rating obviously knows how to do it, and the mistake they are most likely to make is not preparing. The airplane is fuelled and preflighted so he runs through what he's expecting on the test, and we start up.

I take my time going through all the preflight checks: clearance delivery, ground control, engines, propellers, avionics, autopilot, flight instruments. There's an air museum on the field, I position for the run up so the exhibits are visible from the airplane. I learned on my initial multi-engine flight test the advantage of having a distraction available for the examiner during the run up. It takes the fumble out of your fingers when the examiner is reminiscing.

I call ground "ready" from the run up area, as they requested, and then call tower ready from the hold short area. Tower tells me I'm cleared for takeoff, contact departure airborne. I make my calls on the roll. Take-off power set, airspeed alive, gauges green, rotate. With the correct climb speed established, I reset the flight director, engage heading mode on the autopilot, and call departure. Aviate, navigate, communicate. I do the after take-off checks and turn on course as directed. The holding VOR is already tuned; I identify it and determine the direct course for it so I'm ready for my clearance, which is a parallel entry. I brief the examiner on the entry. "It will be a left turn to the outbound heading of 200 degrees, a left turn to track--or fly direct--inbound, and then right turns in the hold." I set up NAV2 for the hold and track to the station on NAV1. I use heading mode to track as opposed to NAV mode, because nav mode is pretty sloppy on this autopilot.

Before we reach the VOR he tells me I have a simulated alternator failure. "That's a checklist item" I say, and reach for the emergency checklist.

"What will it say?" he asks before I open it.

"It will tell me to recycle that side of the master switch, and if that doesn't fix it to turn off the CB for the inoperative alternator, and reduce electrical load."

He's happy with that and I continue to the VOR. That's the second time I've had an examiner ask for a common sense recitation of what to do in a case where the standard calls for rigid checklist following. Perhaps it's something in their notes. I continue to the VOR.

The entry goes as planned. At one minute outbound I turn left towards the station but am nowhere near the inbound track, with full scale deflection on the HSI. That's par for the course on a parallel entry, not something I did wrong, which is why I briefed it as I did. I take up an aggressive intercept heading. Now I'm not doing what I said I would. Well I am, but roughly. "Which of those needles are you navigating by?" the examiner asks me. He's caught me. I twirl NAV2 and show that the heading I have taken is pretty much direct the station. I change heading to omit the "pretty much," and turn back outbound at station passage, starting the clock at the flag flip. After I turn inbound, on track this time, he calls the approach controller to ask for "maneuvers in the hold."

I demonstrate a steep turn, not perfect, but trading off airspeed and bank angle deviations to keep it all within limits. Next an approach to stall in landing configuration, which is good, because that means the gear will be down and I won't be distracted by this airplane's gear horn, which sounds like the other airplane's stall horn. He gives me an engine failure as I recover from the stall, then gives me back the engine and asks for the ILS approach.

I tune the nav aids, brief it and slow to approach speed. I'm late on approach flap, getting it in at the same time as the gear, but the airplane follows the profile at the approved speed. I crosscheck the altitude at beacon inbound. At decision height I call "minima no contact I have control." I'm treating the autopilot like a skilled but stupid and distractable co-pilot, so I talk to it as if it were a crew member. Power up, going up, gear up, flaps up, speak up. I tell tower I'm in the missed approach and they have me call departure, who vector me half way into the next province, and then back for the NDB approach to the same runway.

I'm determined not to get caught high and fast on this approach, so before I reintercept the track, I disengage altitude hold and reduce the power, aiming for the procedure turn altitude. The inbound track comes up quickly, and I intercept it. Within five degrees I can continue my descent all the way to beacon crossing height, but ... huh? The airplane has barely descended. I double check that altitude hold is off. What kind of airplane doesn't descend when you reduce the power without changing the trim? I shove the nose down, disconnect the autopilot altogther, chop the power, (yes, in that inefficient order), add flaps and do a repeat performance of the practice session, plummeting to MDA just in time to give the examiner the requested touch and go. There's another simulated engine failure in the climbout and then a visual return for landing.

At parking, I shut down the left engine, check the hydraulic pump on the right, and then shut the right one down too. "Did you check the other hydraulic pump at start up?" the examiner asks.

"Yes, I push down the gear handle before engine start, and then check its return before starting the second."

He didn't notice me doing that, not surprising as I put it down as part of my prestart flow, not immediately before start, and I had a slightly tricky time getting the left engine rpm to stabilize, so the start was a slow process. He's happy with that, had been going to lecture me on it.

In the debrief I of course do not escape a lecture for my sloppy NDB approach. I am responsible for getting the airplane to descend, regardless of what the autopilot does or doesn't do. Afterwards I asked my trainer and perused the manual looking for what I could have done wrong, but it seems to have been as random as the glideslope deviation during practice.

So an autopilot needs to be treated as a skilled, stupid, distractable and occasionally homicidal copilot. And you can't even make it buy you a beer when it screws up.

Anyway, I got my PPC card signed, my IFR rating renewed and learned some things. I pledge to do better next time.

The Right Approach

A specific instance of tracking is doing an instrument approach. For this, the pilot wants to intercept and track a VOR or localizer signal towards the runway, and, if the approach is an ILS, intercept and descend on the glideslope. It's easy to do. Once the relevant nav aid is tuned and the trackbar turned to the inbound track, the pilot presses the approach button. Typically at this point the autopilot heading and altitude modes are engaged, and the pilot directs the airplane around using the heading bug. When cleared for the approach and on a reasonable intercept heading, the pilot presses the APPR button. The display shows APPR/ARM. The "ARM" is because it's not on the approach yet, it's still intercepting.

When it calculates that it is time to bank for the intercept, the HDG and ARM lights go out and the CPLD light goes on. If this is a VOR or localizer approach, the pilot is in charge of the step down altitudes, but if this is an ILS approach, the pilot need only slow the airplane down to approach speed and lower the gear just before glideslope intercept. The autopilot detects the glideslope, the ALT light goes out and the airplane noses down and follows the glideslope. It's actually way cool, especially in wind, but you have to watch it. At one point during practice the autopilot started to go above slope. We were in VMC without conflicting traffic above. I asked the pilot whose regular airplane this was the classic Airbus question, "why's it doing that?" He didn't know, and by this point the airplane had "busted" being half scale below slope, so I pushed the disengage button and hand flew it. It's actually in some ways easier to hand fly, because you don't have to disengage the autopilot to fly the missed approach.

I didn't get a chance to try it, but this autopilot also has a back course mode, allowing back course localizer approaches. When the BC mode is engaged the HSI track bar shows normal sensing, not reverse sensing, tracking inbound on the back course, or outbound on the front course, (e.g. for a full procedure LOC approach).

Using the autopilot for an NDB approach or hold was the worst, because the autopilot has no ability to do ADF tracking. When I do ADF tracking I make a lot of small banks, tiny corrections I'm not even aware I'm making. The examination standard for ADF approaches is within 5 degrees of track. Two degrees off is barely readable, but once you have four degrees you can have ten degrees like that. It's all about reacting very quickly to very small deviations. Flying on autopilot I'm supposed to duplicate that with the heading bug, but I haven't yet learned to anticipate the delay in response to the bug movement well enough to track accurately. I think I'll just use the autopilot to hold step down altitudes and MDA and do the lateral control myself. When I did the second NDB approach in practice I briefed and set up for one approach and then was suddenly cleared for another, literally as I was intercepting the track, so I just took control from the autopilot to reef it around and down for a very weak approach.

The autopilot has eleven modes all together, but I think I'm done with them for now. The TOGA (take off/go around) mode was unserviceable and I only did three flights and the test.

NAV Mode

While flying in at a constant heading and altitude is just great, pilots often prefer to be on a constant track, that is following a straight line over the ground, even though that means turning the nose back and forth a bit in the air to compensate for changing winds. This can be done by plotting before the flight exactly which bits of the ground you're supposed to be over, then looking diligently out the window throughout the flight, noting and correcting for each deviation as it occurs. That's how you do it on the commercial flight test (using one of two arithmetical methods of correction, not just veering a bit to the left). But this test is not of visual skills. In real life pilots fly long distances in a straight line by following a navigational instrument: a GPS, a VOR or an ADF. This airplane has a GPS, but it isn't an IFR installation, so it's not coupled to the autopilot. This airplane is a proper northern airplane with an ADF, but to my knowledge no one makes an autopilot that can do ADF tracking, because ADF tracking requires semi-magical powers and/or a lot of swearing. Only real pilots are capable of that. So in this airplane the NAV mode is all about the VOR.

I'm sure I've blogged about the VOR, but quick review: a VOR is a station on the ground that creates beams radiating horizontally in all directions. Each beam is called a radial, and the VOR receiver in the airplane is capable of displaying the angular distance of the airplane's position from a selected radial, and whether flying parallel to the radial would take you closer to or further from the station. A pilot's usual tactic is to merge with the radial on an angle, that's called intercepting, and then follow the radial as exactly as possible either towards or away from the station, that's called tracking. Tracking allows you to follow a straight, predetermined line on the ground even if you can't see the ground and the wind is changing.

In autopilot speak, this is the NAV mode. It allows the autopilot to intercept and track a VOR radial just as a pilot would. So to use it, you have to do all the things a pilot has to do before intercepting a radial: tune and identify the station, set the CDI to the desired track, and bug a reasonable intercept heading. Then you activate the HDG function and press the NAV button. That shows on the annunciator panel as NAV/ARM. The airplane flies straight until the airplane approaches the radial and then turns to intercept. At that point the HDG mode switches itself off and the mode switches from NAV/ARM to NAV/CPLD.

The system is totally dependant on the pilot to tune the correct frequency and select the correct radial, but with that done, it tracks pretty well. It even uses internal logic to choose when to start turning, not unlike what an experienced pilot would do, based on the deflection of the CDI and the rate of change. It's not perfect, but maybe they program that in so pilots can feel superior.

It would work the same way if the GPS were in the system. In fact, I don't believe the autopilot itself knows whether its input is the GPS, or a VOR. I think once the pilot selects which navigation input is displayed on the HSI, the autopilot just follows it. The difference between intercepting a GPS track and a VOR radial would be taken care of by the GPS, which is clever enough to give signals anticipating turns.

We Control the Vertical

A two-axis autopilot has capability to control pitch as well as roll, which means that it can be used to hold altitude not just heading. There are different ways this is implemented.

The big difference is in whether your autopilot has the altitude capture feature or not. If it does, you can tell your autopilot to "climb to such and such an altitude and then level off." There is usually a part of the autopilot console where you can set the target altitude. The autopilot knows to reduce its rate of climb or descent approaching that altitude. As the pilot, you need to ensure the correct power setting for the task; it's not an autothrottle and you don't want to be diving at high speeds, nor pitching up for a climb with decaying low speed.

If your autopilot does not have altitude capture, as this one doesn't, you have to wait until you're just about at the target altitude, considering the rate of climb or descent, and hit the ALT mode button at just the right moment. You may coast past and have to fix it afterwards, or you may have a habit like the one I'm trying to break, and just disengage the autopilot, level off manually, and then hit autopilot hold. I have this habit because the autopilot controls by motoring the trim, not by physically moving the yoke and hence elevator the way I do, and the effect of one airplane I fly trying to level the airplane with elevator trim is seriously sick-making. When the plane makes the pilot wonder if she can reach the vomit bags from her seat, it's worth hitting the red switch.

If you're flying level at one altitude and want to descend, there's two ways to do it. One is to leave the ALT mode engaged and to press on the vertical trim switch, not the electric elevator trim on the yoke, but a switch on the autopilot control box, and that will give a climb or descent of up to 600 fpm, while you are holding it. Or if you want a faster descent, you press the autopilot ALT button, disengaging that mode, descend like a normal person, and re-engage the ALT hold when you're at the altitude you want to hold. You can, of course, leave the heading mode on during the descent, so that the autopilot controls straight line flight and you control the climb or descent.

This engaging and disengaging makes for fun times, because to disengage the autopilot I am used to, you can press the big red autopilot disconnect button on the yoke, or just apply pressure to the yoke. As soon as there's a disagreement between what I'm telling the airplane and what the autopilot is telling it, that autopilot surrenders and gives control over to me. This particular autopilot, however, has much more confidence in itself. If I apply pressure to the yoke without disengaging the autopilot, it trims against me. And it will continue to trim against the force I exert until it runs out of trim, or I give up. At that point the control forces exerted by the trim may be more than the autopilot can handle, causing it to disengage, and leaving me having to exert fifty pounds of control force while retrimming. The fifty pounds figure is from the manual, not personal experience. I'm sure I didn't ever get much beyond twenty-five pounds before remembering to hit the disengage switch. From now on, I plan to use the disengage switch every time so I'm ready for either sort.

The way I fly without an autopilot, the descent rate depends on the power setting and the speed remains the same, unless I deliberately trim nose down to get a higher speed in the descent. During the training on one approach I didn't bring the power back enough because I'd started the descent on autopilot, and my speed was too high for my approach timing. I resolved to reduce the power before asking the autopilot to descend, just as I would if I were asking for the descent with the yoke, even though the training pilot did it the other way around.

It seems like quite a lot of bother to avoid having to fly the airplane, doesn't it?

We Control the Horizontal

I'm advised that I'll be doing my much delayed pilot proficiency check this week. That means that if I get the job I'll have two career-critical flight tests right after one another. Lovely. This one can serve as the warm up.

Fortunately regulations require me to have training, not just go straight to a test, even though this is the airplane I normally fly. Well it isn't exactly. I'll be doing the test in an unfamiliar airplane, not even owned by my company, because the company airplanes are busy working and company hasn't been able to schedule them home in in a way that works out. This particular airplane is a different version of the airplane I normally fly for work. I'm not worried about that, because I've flown all the versions of this model and am familiar with the way they vary. I think the examiner will probably cut me some slack if I rattle off the numbers for the more familiar plane. The pilot who flies this airplane every day gives me a sheaf of paper to photocopy, not only the numbers for the airplane itself, but a bunch of helpful notes left over from his own training.

The test will include my ability to use the autopilot, yet this autopilot is not the same as the one I use at work. There is no standard interface for autopilots, and the complement of things they can do isn't a given, either, so you really have to learn how to use one, not just figure it out as you go along. Before the first day of my PPC training I took the autopilot manual home with me, to get acquainted. Curiously, the autopilot has the same name as the company I just interviewed with. I won't try to interpret that as an omen; I'll just leave it as a hint for those in the know. I pretty much gave it away already anyway.

I started out trying to blog about the Bilby autopilot but I figured I'd better explain a little about autopilots in general first, and well my explanations are a bit like my movie synopses. So here's the first part.

In order to tell an autopilot which direction to fly, you have a heading bug a little pointer that sits on the rim of the heading indicator and which can be positioned, with a knob or by keypad inputs. It shows the heading you are asking the airplane to fly. Many non-autopilot-equipped airplanes have heading bugs too. They are useful to remind the pilot where she is supposed to be going, or to help her crab to hold a radial in wind. This autopilot has one, and it is controlled with a knob on the right side of the horizontal situation indicator. There is another knob on the left side, the OBS, which is used to select the radial of a VOR, or the localizer track to the runway. Airplanes are not standardized on which is on the left and which on the right, so a pilot should look at the symbols before twisting one of the knobs. The OBS has an arrow and the heading selector has a pentagon, the same shape as the heading bug. The heading bug can be different colours, too, and you might be surprised how distracting this can be. This one is orange, a pretty standard colour, but I've had them yellow and green for sure, and possibly other colours too.

If the autopilot is on, working, and the heading mode is engaged, and you turn the heading bug to a heading more than a few degrees away from the current one, the airplane will roll into a standard rate (3 degrees per second) turn. and roll out on the selected heading. If you turn it to a heading very close to the current one, it will still roll, but but at a shallower bank angle. You have to be careful if you want a turn of close to or more than 180 degrees to move the heading bug part way and then wait until the airplane has turned far enough that it won't try to take a shortcut and turn in the opposite direction.

I mentioned the heading mode. You put the autopilot in a particular mode or combination of modes to tell it what you want it to be doing. For this particular autopilot to engage heading mode from not having the autopilot active at all, I would push the command wheel steering button (on the yoke) to reset the flight director, then press the HDG button, then move another switch forward to engage the autopilot. The HDG light on the dashboard panel should now be illuminated, and the airplane should continue to follow the heading bug when I turn it.

It's usually pretty good at that, holding the heading to within a few degrees and correcting promptly when turbulence causes a deviation. In this particular aircraft the alignment of the autopilot to the bug is a little bit off, so I have to align the left edge of the bug, not the centre part, with the intended heading. I get used to that soon enough. That alone is a pretty convenient feature. It allows a pilot to look up things in a manual, read a chart, eat lunch, retrieve dropped items all with hands off the yoke. You still have to keep a lookout, but it's a big help. It's also very useful at night or in IMC if you get momentarily disoriented, you can have the airplane fly for a few minutes while you scan the instruments and reassert in your brain which way is up.

Idiot Light

I work closely with airplane mechanics (often called engineers, although they don't actually hold engineering degrees) to keep my airplane working, so this blog entry from an aircraft maintenance engineer amused me. It's the view from the other side, and a message for pilots who don't heed the rule about not messing with the people who maintain your aircraft.

Mr.V, the blogger whose story this is, hasn't set up his blog so I can link to individual posts, so I have quoted the post in question below. His language and spelling are saltier than mine, but I'm not going to edit someone else's post.

Thursday, May 8, 2008

In my line of work I meet a lot of pilots. For the most part they are nice and cordial but then again there are some complete dicks....

The following is an account of a line call this morning.

6:10 am (radio goes off, Aircraft 218 requesting maintenance)

me: Good morning sir, what seems to be the problem

Mr pilot: the AP (auto pilot) indicator light is pointing in the wrong direction

the AP indicator is a dummy light that shows which person has the auto pilot engaged. This one happened to be pointing to the left. It is also a dummy light it has no effect on the AP.

Me: Well it looks like avionics was working on the auto pilot last night and and perhaps put in the wrong bulb. this is not something we carry in the maintenance truck

Mr Pilot: well this is not going to do we have got to get this fixed.

me: well it is not flight critical and it is just a dummy light it is not worth taking a delay over.

Mr (now rude) pilot: Just change the bulb and we will be on our way

Me: I just said we don't carry these in the maintenance van

Mr rude pilot: then get in your van and drive back to the hanger and get one.

Me: Perhaps if you called me earlier than 5 minutes before you where to board I could have taken care of this for you.. Like i said it is not worth taking a delay over

Mr extremely rude dick head pilot: Don't you tell me how to do my job just fix the dam system.

Me: Ok then I got and idea can I see the flight can? (flight can is a log book)

I took the flight can back to the maintenance van called maintenance control and deffered the entire auto pilot system.

The look on the pilots face when I handed him back the flight can and put the inop placard on the auto pilot indicator and pulled and collared the Ap circuit breakers..... Priceless

and the flight left on time

By way of explanation to non-pilot/maintenance readers, the pilot in command does have the right to decide whether equipment is working to his satisfaction. It isn't out of line for him to demand that a piece of equipment he is supposed to use works perfectly, even if the broken portion isn't core to the function of the equipment.

The thing is, for this particular company, the autopilot itself was not a required item for flight, just a nice-to-have item. So an unnecessary light on the autopilot was really not required. The pilot was, by being obstreperous, arguing that the light was a safety concern. And yes, I can see how having the wrong pilot indicated as the one who had engaged the autopilot could cause a problem. Transport Canada or the FAA could certainly (and have many times) asked for entire systems to be replaced because of malfunctions in unnecessary components. In fact, had the FAA been looking over the shoulders of the pilot and mechanic, they would probably have asked for exactly what the mechanic eventually did.

The mechanic gave the pilot every opportunity to see that he would be better off accepting the airplane with the faulty light, but the pilot wouldn't take it. So the mechanic solved his problem. Safely, legally, expeditiously and hilariously the mechanic solved it. Disabling and placarding a faulty but not required system is absolutely by-the-book. No one can touch the mechanic for doing that. And now the pilot isn't allowed to use that system. So as punishment for being a jerk, the pilot now has to hand fly the airplane all day. Brilliant.

Of course, most likely he'll just have the co-pilot hand fly it, and be a jerk to him or her, too.

While I'm linking to other people's blogs, Julien at Making Time for Flying has posted some pictures that perfectly illustrate tow bars and the flat spot I was fearing from the unauthorized tow my airplane received a few posts ago. It's not an uncommon sight if a pilot has accidentally landed with the brakes on or locked up the brakes, but the title of Julien's blog post gives away what caused that pilot to lock the brakes. I'm always amused by the various creatures people manage to hit with their airplanes, around the world.

Wind and Time

Out of Lincoln I planned for a five and a half hour flight to Florida. The forecast showed that I should get pretty much a direct crosswind for the first part of the flight, and then a little bit of a push further on. I didn't count on the tailwind for the flight plan; I was just going to take it as a bonus if it happened.

Almost immediately after takeoff I suspected that that plan was not conservative enough. As soon as I turned en route my groundspeed dropped, and at level-off the ETA calculated by the GPS remained above six hours after the airplane had sped up to cruise. A weather system moves a little further east or west than forecast, doesn't have to be far, and the associated winds aren't going in the forecast direction at the point you are. Oh well. I originally planned this trip on a different route, one that would have given me howling tailwinds most of the way, but other operational concerns (i.e. my boss) dictated the route I took. So this is where I am. I'll keep a good eye on my watch and probably stop somewhere in Alabama. I've never been to Alabama.

Nebraska gives way to Kansas, and I get the song "Everything's Up To Date in Kansas City" stuck in my head. This is doubly irritating because Kansas City isn't even in Kansas. It's in Missouri. And when I say "song" I mean that one line, it's all I know of the song. I don't even know the tune. So I have the words "everything's up to date in Kansas City" running through my head for an entire state.

Eventually I'm distracted by the changing accents of the air traffic controllers. As soon as I crossed the 49th parallel the American accents became apparent. There's a change in the vowel length and something happens to medial Ts, but entering the southeast corner of the state there's another change from merely American to Dukes of Hazzard American. People from around here can probably identify the state someone is from, but I can't.

I'm approaching a giant river. It's so large that I don't need to look at the sectional chart to know that it is the Mississippi. I think this is my first time flying across it in daylight. Maybe I've flown across further north where it is less impressive. I have the autopilot on so I can get out the camera, but before I can take any pictures for you there is an air traffic controller yelling at me. Oh firetruck. The autopilot has silently disconnected and the airplane has descended 350 feet. I grovel an apology, regain the lost altitude, and suffer contritely the air traffic controller's diatribe against me. Now I lose any trust I had in that autopilot. I breathe on it and it disconnects. I think I exude a hormone or an electro-magnetic field that messes with autopilots. My aviation career is one long history of autopilots not doing what they were supposed to. They dive and climb and roll and fail to follow the bug.

The Mississippi was big and wide, curvy and muddy, snaking through flat green country as far as I can see in either direction. The cities here tend to be small and flat. I haven't seen any clusters of glass and steel towers, but I know there are some big cities around here.

The headwind has subsided and the ETE is hovering just below six hours. I'm carrying six and a half hours of fuel. A half hour is the minimum legal fuel I may plan to land with, but that only applies to day, and the day is getting on. Plus there could be thunderstorms at destination. I need forty-five minutes reserve fuel for night, and I'm still going both south and east, both bringing night closer, this time of year. The other issue is my duty day: if I can't reasonably expect to shut down at destination within 14 hours of the time I caught an airport cab this morning, then I can't take off. I do duty day math. I have enough time for a fuel stop if I need one.

I call flight services, looking for any CBs, forecast thunderstorms or new NOTAMs at my destination. It looks good. I ask for official night, something I should have found in my original flight planning, but I'm not sure where to find it on the US weather site. Neither is the briefer. He eventually comes back and tells me nightfall is at 20:45. I look at my watch, not the one on my wrist, which is still set to Mountain Time, but the Zulu watch on my clipboard. That was two hours ago. I ask him to confirm. "That doesn't sound right. It's already dark there?" The sun behind me is at about a 40 degree angle to the horizon. He double checks and realizes he has given me local time. He translates to Zulu. I'm golden. I'll be there an hour before dark. My groundspeed is still increasing. It looks like I'll land in daylight with 40 minutes of fuel remaining.

US briefers often do that: translate times to local, often without telling me. They are trying to be helpful, but it's dangerous and confusing. I think they have a tool that allows them to redisplay the weather products in local. I don't mind comments like "convective activity not really expected until mid afternoon" or "the rain will probably stop by midnight" but when you give the time in numbers, those numbers should be UTC.

I land at my destination airport in Florida, and, cleared to taxi, I take a few moments to orient myself. I've been here before. The best part is that the staff at the FBO remember me and welcome me back. I know recognizing customers and being nice to them is part of their job, but it's a really good feeling. I'm not sure which I like better in my job: going to new places and meeting new people, or going back to places I've been before and getting to see them again.

Oh and I'm going to start adding the tags "airplanes," "commercial aviation" and "flying" to all the posts they suit, even though those are the default topics for the blog, because I looked at my tag cloud and realized it made this look like a weather and avionics blog.

Counting to Ten

My airplane is in the hangar for scheduled maintenance. I flew it here in the dark, with a slight tailwind that brought me in 0.2 h before the maintenance was due, rather than 0.1 before as I had planned. (The 0.1 early was of course planned in case the wind or ATC situation went the other way. I've been here before, and because this is a hundred hour (performed every hundred hours, not taking a hundred hours, but longer than the regular oil-and-filter-change) inspection.

When I arrived I parked outside the maintenance hangar and phoned the mechanic for a ride. Company has rented hangar space and flown one of our mechanics down to do the work, so we have someone familiar working on the airplane, and we're not paying the local shop rate for a simple oil change. I hauled my stuff up to the airfield gate and discovered it was locked, even from the airside. The plate where the combination to get back in would usually be written instructs me to call airfield security to get through, from either side. Both they and my ride arrived quickly and I was allowed out.

Next morning at the hangar they're laughing at my list of snags. Most of it is label-related: the little plastic cover over the emergency exit handle has become brittle from time and sun exposure and cracked into many pieces. The cover isn't essential, but I suspect the exit operation instructions on it are legally required. There are several circuit breakers with illegible labels, and one of the tanks has lost the label declaring its fuel type and capacity. They point me at an electronic label maker and tell me to go for it.

This appears to be one of the areas in which I am mentally arrested at the age of ten. I love making labels. The label machine allows me to change the height and width of the letters independently, print up to four lines of text on one strip of label tape, and choose black or white letters on black, white or clear label tape. I label everything. Now I don't have to squint through scratches to identify my CBs, and I no longer have two radios labelled "Com 1."

I also remove the pilot's seat from the airplane. This is much like one of those twisted wire puzzles, at first appearing completely impossible, without turning off the physical existence of one of the wooden bulkheads. When finally I twist it the right way and get it out of the cockpit, I wonder if I'll ever be able to get it back in. The co-pilot's seat comes out more easily and then three cents (two Canadian, one U.S.), several pistachio shells (definitely my chief pilot's), and assorted small fasteners are no problem to remove. I vacuum the interior of the plane, and am contemplating washing the bits that still have cowls on when I'm called to do another job. They explain my task.

An oil drum has been rolled over next to the airplane on casters, and it is fitted with a hand pump. The outlet for the pump is inserted into the crankcase. To refill the crankcase with clean oil, I should turn the crank handle on the pump backwards until it stops, then turn it forward until it stops again, from stop to stop is about two and a half turns of the handle and equals one quart of oil. I have to add eleven quarts of oil to each engine.

"Wow," I tease our mechanic. "I am going to tell all my friends about the time a mechanic trusted a pilot with a task that involved counting past ten!"

We're just about through the inspection checklist when everything grinds to a halt. An emergency airworthiness directive came out a few weeks ago for this airplane. It passed on initial inspection, but somehow in the last fifty hours, the plug has become out of tolerance. The airplane fails the test and is now grounded until the part can be replaced. It looks fine to the naked eye: the threads seem undamaged and there are only a few nicks and scratches. But a micrometer shows a different story. I could experience a power loss if I go on flying with this one.

While the replacement part is couriered in, we get an avionics tech to come and have a look at a transponder that ATC reports as intermittent, after several hours of flight. It's a KT76, exactly the same model that died on me in Florida last year. I hope it won't be as expensive as that one was. The tech goes into my avionics bay and sets up test equipment that looks like something Mr. Spock would use to analyze zeta radiation density for Captain Kirk. It tells the tech that the transponder is putting out the right power, and has some frequency drift, but likely not enough to produce the problem. He traces the problem to an antenna the length of my finger and the diameter of a chopstick. It needs replacing, and--this is amazing--he has one, and it only costs $160. In aviation maintenance budget terms, that's equivalent to the pennies I found wedged in the seat rails.

Even more amazing, the antenna replacement solves the problem, and ATC is one again happy with my blips.

Parts Exchange Dance

Ok, now back to my life, and the uncooperative autopilot. If that US Airways A319 had had a bullet through the autopilot, it might start looking deliberate rather than accidental, but there would likely be grounds to consider it to be justifiable gizmocide.

Airplane repairs ought to be a lot like automobile repairs, and I suppose they mostly are, but the process seems different. I'm going to characterize airplane repair as more honest. Everyone admits from the outset that they have no clue what is wrong. The automotive shop pays a slight amount of attention to what the customer says, then tests it for themselves and says "You need a new X." Then when the new X doesn't fix it, they say, "you also need a new Y, and a Z, too." The aircraft maintenance engineer listens to the pilot's description of the symptoms and says, "You might need a new X, or maybe a Y. Could be the Z." So they get an X, and the airplane still doesn't work, so then they try with a Y. Because the stuff is so expensive, the customer gets to be more involved in the process. You'll probably all tell me I just have had bad experience with auto mechanics.

Rarely are airplane parts just available at the location the airplane is being repaired, and you can't get generic parts at Canadian Tire, so you have to order them from the manufacturer. Sometimes there is a loaner X available or another airplane in the fleet that can be cannibalized for parts, so you don't have to wait for everything to arrive. We spent the day alternating between watching the tech tweak things and then getting in the plane to test the tweak. Sometimes the malfunction was evident with the electrics on but the engines off. Sometimes it worked with the engines off, but malfunctioned after start

One of the possibilities is just to fly the whole thing down to some fabled Autopilots R Us facility in Oklahoma or Kansas or somewhere where they actually know how these things work. No one here seems to really know.

We never got as far as another test flight before my part in the exercise ended so I could go back to work.

Autopilot Autonomy

The airplane is finally ready for us to do our practice run. In this cold weather, the maintenance hangar has curtains across it in sections, so that the doors can be opened to admit and release aircraft without emptying the whole hangar of warm air. But our airplane is at the back, so this is a mass evacuation. All the airplanes are towed outside, one by one. There is some consternation when ours leaks a LOT of fluid as it is towed over the threshold of the hangar, but we then realize that the heater exhaust muffs must have been packed with snow when it was towed in, and the meltwater didn't drain out until now, when the airplane was tipped up slightly to go over the lip of the hangar. Last time I had fluid pour out of an airplane like that I had lost a fuel pump seal, but this is just water.

We install the tire valve cap and everything else checks out okay for the flight. I'll be in the back while my co-worker who hasn't flown for the longest will fly from the left seat, with the chief pilot in the right.

There's no cockpit door, but there are dividers behind the front seats, so I can't see what's going on in the front very well, just follow along through the checklist sequence and the taxi clearance. There's an extra headset plug back here, quite common in airplanes where there is not a third, observer seat in the cockpit, so I can hear the discussion up front, too. All equipment ground checks okay.

The chief pilot is familiar with local airspace, so gives directions to the training area as we depart VFR. We climb above the official training area, as it's a little too small for our speed, and then left seat engages the autopilot. The guy in the left is flying single pilot, so the conversation from the front lacks the crisp clarity of two crew SOPs so I have to guess a little what is going on. The airplane banks slightly left, and I see hands go to the autopilot roll control on the centre console and watch a steep left bank develop, before we roll level again.

"I'm in suspense back here," I say. "Please tell me you were turning the roll knob."

"We were," comes the response. "To the right."

Full right roll input on the autopilot produced a steep left bank. That would not be the mark of a properly-functioning autopilot.

"Does the altitude hold at least work?" I ask. When the airplane went in for autopilot repairs it could hold a heading slightly left of the selected one, and could keep its wings level, but would pitch up violently when the altitude hold was engaged.

They endeavour to keep the airplane right side up long enough to see if it can hold altitude. Are you familiar with the term 'homesick angel?' If this autopilot is emulating any specific pilot, it's an airsick student on his second or third lesson. (On the first lesson they usually don't get this violent, because they are afraid to break anything or make the airplane fall down).

They pull the circuit breaker on the autopilot and practice some airwork before returning for an ILS approach. The air was calm, the approach well-flown, and we got to look out the window at the chief pilot's home as we went by. On very short final the controller said calmly, "just to confirm you are cleared to land runway ..." and I forget which runway number it was, but I knew we already had a landing clearance, so I wasn't concerned. The people in the front suddenly realize, however, why the controller has made this assertion. They've been cleared for the practice ILS that is aligned with one runway, but cleared to land on another, diverging runway that starts with the same digit.

A little low level maneuvering later and we've landed on the correct runway, to taxi back to the maintenance hangar for more autopilot work. It's the airplane equivalent of the last moment lane change to make your exit. As we taxi in I text base: "Landed. Airplane still broken."

The return text comes after shut down, as we discover the gear doors don't stay closed after power is removed. "Old problems or new problems?"

"Both."

I leave to the chief pilot the task of itemizing them.

The weather, on the other hand, is absolutely glorious for flying. It's warmed up to barely below freezing and it's clear with very light wind.

Second Try

I get a phone call informing me that the autopilot works, this time, and to book a flight. We reassemble in the big city of ... oh I'll let you figure it out, to once again try to renew our PPCs in a functional airplane.

We're staying across the street from the airport, in a big hotel with one elevator that appears to be afraid of the lobby. When you get in the elevator and select lobby, sometimes it goes almost to the lobby, then reverses direction and goes back up to some higher floor. Sometimes it stops on the first or second floor. (That is, sometimes it stops on one and sometimes it stops on the other, not a concession to the fact that different countries call the same floor "first" and "second." In that debate Canada comes down firmly on the fence. Consequently the only way to know what floor something is on here is to look at the labels.) And sometimes our frightened elevator gets almost to the lobby then freezes in terror and you have to press the door open button to persuade it to let you out.

We head out to the airplane, which is still in the back of a maintenance hangar. Apparently they won't be taking it out today because they are still waiting for a part. When we climb into the cockpit to do drills, there is an empty space in the console where the autopilot controls normally are. I wonder to myself how well the autopilot works, but we shall see. We take the aircraft manuals back to the hotel to study.

Next morning our airplane isn't ready to go, but our little inspection revealed a missing valve cap for the nosewheel tire inflation valve and a missing wingnut on the emergency gear extension compartment cover. We take the gear cover and one of the other valve caps and task ourselves with finding the missing parts. The valve cap is easy, and only costs $3. That's only triple what we could get it for at Canadian Tire. But before we can get the wingnut, we need a part number. We bring our quest to a company that operates the same type. "What is the aircraft serial number?" we're asked. I recite it cheerfully from memory. My co-workers turn and gape at me. This is the chick who made it a priority to get business cards because she could never remember her own cellphone number. I just smile. I had to check the serial number in order to determine which set of performance charts to use, and it happened to be a concatenation of a couple of numbers that had more significance to me than my cellphone number. And my airplane is of greater importance to me than my cellphone, anyway.

We find the part number. It turns out we're missing two parts: the one that we came in search of, and the one that was supposed to prevent it falling off in the first place. Each has its own alphanumeric identifier, but is not available today.

At the end of the day we've done a scavenger hunt all over the airport, but not gone flying. I try to salvage the day with an accomplishment, "At least snow won't get into the valve stem now."

"I forgot to put it on!" says my co-worker, producing the valve cap from his pocket. Good grief. At least tomorrow we'll go flying.

Your identify-the-airport clue for today is that there is an eating establishment here named after a runway that no longer exists.

Or Maybe Not

The PPC plan was for a few pilots from my company to meet in one city, where the airplane was undergoing regular maintenance, then practice up on emergency maneuvers for a few hours, before flying together to another city to do the ride.

When I arrived I was informed that the test flight had revealed an autopilot problem and one engine not producing rated power, so we reviewed the paperwork portion of the exercise, discussed the approaches at the airport where the ride would be, and checked into the hotel. The plan was to wait for the engine to be adjusted, then get in the airplane and fly to another city where we could get good autopilot repairs done.

Overnight an unforecast cold front swept through, blanketing the city in freezing fog, deadly droplets of water that not only obscure vision like regular fog, but are so cold that they freeze into ice on any surface they touched. Maintenance fixed the engine problem, and the airplane returned from the run-up looking like a winning entry in the Québec winter Carnival sculpture competition. All they had done was taxi outside and run the engines. It would have looked like an uncarved block if we'd actually tried to fly it. Plans in aviation always change. The weather was forecast to linger over much of our route. We all hung around in the hangar for a bit, read our e-mail, and checked the airline schedules for flights home again.

One pilot was nominated to stick around until an engine test flight could be completed, then fly it to another city where autopilot repair could be effected. We will get a different examiner there to do the rides, instead.

So I went home to wait for that to happen. So Disney, isn't it? Facing your fears, only to have the monster evaporate.