So the plan in January 2004 was that I was going to do a quick multiengine rating while in Ithaca. The tale of my first multi-engine flight is here. Well, after four years and over 27 hours in the plane, I finally took and passed the checkride. I was planning to write up how the training progressed, but it was so spotty, with literally 6 months or a year between flying the Duchess, that it wasn't worth it, since progress was so slow. But now that it's finished, I can at least write up a description of the checkride.
Click on the thumbnail to see a full-screen version (640x480). Most of these photos were taken by David St. George.
First off, I should mention that the planes heater has long since been fixed and works quite nicely. And in general, all of the systems in the plane were working pretty well. The GPS database was way out of date (from 2002 or so), but that's not a big deal as we weren't using that for approaches.
We headed out to do airwork, and practice the basic maneuvers. This included steep turns (for some reason, I tend to have trouble holding it to the required +/- 100 feet), slow flight, power on and power off stall, and some Vmc work. In general, things went pretty good. We did an instrument approach back to Ithaca.
After some landings, we had lunch and reviewed for the oral portion of the test.
After reviewing the definition of Vmc and things like that over lunch, we went back and did some airwork and pattern work. The weather was a bit crappy, but not totally horrible. Probably 4 miles visibility, so we stayed in the pattern and did various landings, including spot (precision) landings. It was a matter of aim a little sort of the mark, come in at the proper airspeed, get into the flair, throttle the engines back to idle (i.e., "chop the power" but not abruptly) and the plane will land. At idle power, the two "windmilling" propellers generate a lot of drag, so it's not that hard to land pretty close to a specified point.
We did a short field takeoff, and a soft field landing (applying just a little power right before touchdown). I would occasionally forget to do the carbureator heat check, and David stressed that and suggested adding it to the general flow when I do the (gas) pumps on, as part of the GUMPS mnemonic (Gas, Undercarriage, Mixture, Propellers, Seatbelts) for the final approach checklist.
At one point while doing the U (undercarriage) part, I noticed nothing happened, the gear didn't come down. A quick check showed that the circuit breaker (on the far right side of the cockpit, right in front of David) was pulled. David said to pretend the breaker wasn't tripped. The manual gear extension in the Duchess is extremely easy. Lift the flap on the floor by the pilot's feet and turn the vale level 90 degrees. There's a little tool in the pocket in the pilot's side door that makes it easy to give the valve a twist, and about 5 seconds later, the gears are down and locked ("three green lights"). This was while we were in the downwind leg, in the kind of marginal weather. David was looking through the book to check if the value should be reset or not. Note that turning the valve releases the hydraulic pressure that keeps the gears up. They then extend and lock. Putting the valve back allows the system to be able to retract the gear, but as long as the gear lever is in the down position, and the circuit breaker is pulled, nothing will happen. I figured I would fly the plane and David could read the manual to double check. Once he had done that, I put the valve back where it was and we landed. On the roll again, we made sure the gear would come up and it worked fine.
I'm not sure if it was that time or not, but David pulled one engine when we were around 600' up. I did the requisite cleanup and (simulated) single engine landing.
We did a bunch of different takeoffs and landings. The only thing we didn't really didn't try was the short-field landing.
On Sunday morning, we continued the training with more emergency proceedures. Because there were some clouds, we filed IFR to Syracuse and back, as the weather looked better there. We took off and when we broke out on top, we requested a block altitude for maneuvering. When talking to Syracuse approach, we asked for an area in which we could maneuver. The approach controller seemed to be unhappy about all the "paperwork" we would generate for them by filing IFR but not actually flying up there. We told them we had not decided yet and needed to do some single engine maneuvering. Oh, I should mention at this point David had failed an engine and this time had me shut it down for real, including fully feathering the propeller. I was also under the hood at the time.
Syracuse approach told us we'd have to go turn south and talk to Elmira if we wanted a place to maneuver. So we did. To give us somewhere to go, David had me set Dansville into the GPS, so I'd have a course to follow.
Elmira was equally "eager to help." The controller said he was not aware of any Letter of Understanding between the flying club and Elmira Approach about our practice area, but he gave us an area, defined by the Ithaca VOR radial and DME in which we could maneuver. We restarted the engine and let it warm up. We did various stuff including an emergency descent (pretty steep) and then asked Elmira for vectors for the VOR 14 approach into Ithaca. Once again, we were running on one engine with the other one set for zero-thurst to simulate a failed engine so I could practice a single-engine instrument approach.
He gave us vectors and then handed us off to Ithaca Tower. I was on course and called up the tower. Since we were still far out, I opted to wait on lowering the landing gear, as that's a point when you really commit to the approach (no missed approach on one engine). The tower called back and said Elmira called them up and reported that were were a few miles north of the course and that we should climb to 3000 feet and switch back to Elmira. So, we were now doing a single engine missed approach. Fortunately, we were at 2700 feet, so our 300 feet/minute climb rate was adequate to get us back.
It was odd as the HSI indicated we were right on course. We told them our #1 VOR showed we were OK, so we'd be using our #2 nav radio and wanted to try the approach again. There were clouds and I was under the hood, but there were enough breaks that David could see where we were. At some point later, after we had things set up for the approach, David a small blue light just above and to the right of the HSI indicated that it was getting its course from the GPS (which I had set to Dansville a while back), rather than the VOR. Ooops. Once I realized that, I said to David, "That's a stupid way to die." The Elmira controller must have been paying a lot more attention since we had problems with the course the first time. Not another word griping about anything. He gave us good vectors quite a ways out to make sure we were OK.
After the landing, I decided it was OK and we could call it a day. The next day would be the checkride in Dunkirk.
I had to drop off my car to get its yearly state inspection. David and I arranged to meet at the dealership at 6am. Due to some confused communication and bad signs in the lot, we eventually got together a half hour late (we were waiting for each other on opposite sides of the lot), and headed to the airport.
David said he'd get the Duchess pre-flighted and out of the hangar, so I could get a weather briefing, file a flight plan, and get ready to go. I got a weather briefing, filed the flight plan to Dunkirk, and did the takeoff calculations (normal takeoff, accelerate-stop, and accelerate-go distances; single engine climb rate; single engine service ceiling). I got to the hangar just as David was pulling the plane out.
He went to get some things and told me to get in and get ready. While he was gone, I used my handheld radio to pick up our instrument clearance. David came back and said that we would be "expediting" things as much as possible. We were running behind schedule, but we could still make it by 8:30am; it was around 7:15am. We would do a two-crew operation, including call-outs.
I started the engines (having done it 3 times in the past 2 days, I was feeling more comfortable about the checklist and flows). We called for a taxi clearance and continued the checklists as we did so. David requested an on-course heading after departure from ground control.
I must admit, I was being pushed to a level that stretched my abilities to do everything quickly and correctly. But David was there to catch my mistakes. This was not a simulated check-ride. There was no time for that. This was trying not to waste time, and having two pilots and 4 sets of eyes watching to make sure things are OK. We got to the end of the runway, and I did the engine runup. I believe I forgot to do the prop check. David caught that. I didn't have time to beat myself up about missing stupid things nor psyche myself out.
We took off and were cleared to our on-course heading direct to Dunkirk. We climbed through a 2000' cloud layer and broke out at 5000'. The sun was rising behind us at 7 o'clock (position, not time) and a crescent moon was visible at 11 o'clock. Very pretty. We climbed to 7000'.
After a few minutes, at David's suggestion, I turned on the autopilot (2 axis plus altitude hold), had David watch for traffic, and got the POH (Pilot's Operating Handbook) for the Duchess out and worked out all of the takeoff calculations for Dunkirk, estimating the examiner's weight, and full fuel. It took 5-10 minutes. The radio was also silent. No other flights were talking to ATC in that area.
A minor "issue" was that VOR at Dunkirk was NOTAM'ed (NOtice To AirMen) as out-of-service. So we couldn't fly any VOR approaches to the airport. There were also GPS approaches, but the database was around 5 years out of date, and so couldn't be used for any of those approaches. Our IFR options were...well, there were none, at least at Dunkirk. We could go to some other airport nearby in a pinch. But we were hoping that the field would be VFR so we wouldn't need to do an approach.
ATC reported that the ceilings were quite a bit above VFR minimums, and that we should be able to get a visual approach. Yay. We were given a descent through the cloud layer, and broke out probably around 3000'. It's always a tricky planing situation, deciding when to slow down and reconfigure the plane for landing. Around 10 miles out I started retarding the throttles and reducing the propeller speed. With the airport in sight, we cancelled our IFR clearance and ATC told us to switch to the common traffic advisory frequency (CTAF). As we called a 5 mile final, a snow plow at Dunkirk told us he had just plowed the runway, the breaking action was good and the runway was clear. Nice to know.
GUMPS final approach checklist. Gas on. Undercarriage down and locked. Mixture full rich. Propeller high RPM setting. Seat belts on. Final approach checklist done. The runway was white, but not hard to spot. Flair and land and then taxi in. We called to get the plane refueled as we were heading in.
It was now 8:25am. We made it on time.
And then began the challenge of dealing with the FAA's new automated online web-based mess, er, I mean system. The night before, I discovered that it is impossible to actually register for a commerical, instrument multi-engine add-on rating. The system permits a person to sign up for a VFR-only multi-engine checkride, or an instrument add-on to a multi-engine rating, but doing both at once...inconceivable. And needless to say, there was no documentation or help on how to register for such a rating. Just to clarify, this is a very common multi-checkride, and the only difference between a VFR-only multi-rating and a VFR/IFR multi-rating is the requirement to fly 2 instrument approaches, one with both engines operating, and one simulating a failed engine.
After poking around the web the night before the test, I discovered that the secret, undocumented way to proceed is to register for a VFR-only test (the system would not let me register for an IFR test, since I didn't have a VFR-multi rating required to take the IFR add-on), take the VFR/IFR test, and then have the examiner override the system and remove the VFR-only restriction when he issues the rating. Simple, eh?
Turns out the examiner was not aware of this, so we needed to educate him. Not a big deal. What was a big deal was that in order to begin the test, I must digitally "sign" my application, the CFI who was recommending me for the test must then "sign" the application, and finally the examiner looks it over and starts things going. The only problem was that David was unable to sign it. He called the help line. He called them 6 times. At one point the advice was "just keep trying to submit the form and eventually it should work." We had to call them so I could submit the form, so David could sign it, and so Lou could sign it. It took a half hour. It was remarkable how difficult and poorly design their system was. It was as if no one had ever actually used it, or rather no one who needed to use it used it. I'm sure it was tested, but not in the way it would be used. But enough of a rant about modern software development and testing. We were finally able to begin the test.
He told of a plane that lost an engine and had to divert to Dunkirk. There were clouds, but not too low and the pilot did everything right and brought the plane in safely on one engine, even managing to taxi to the ramp (not always an easy task on one engine). The pilot had had problems before in other multi-engine planes and bought a new plane to avoid having to do a single-engine maneuver again. This was his new plane. And while he had to once again do a single-engine landing, his experience and skill served him well.
The next day a different pilot in a different plane, a Baron, was diverting to Dunkirk because of problems in one of his engines. The weather was lower at Dunkirk, and better elsewhere. The pilot stayed high and kept his speed up. He broke out of the clouds over the airport, way too high to land and was forced to go-around, and so throttled up and climbed back into the clouds. Very shortly after that, perhaps near the departure end of the runway, the plane fell out of the clouds, level with the horizon, but spinning like a top, falling. It was in a flat spin, and impacted the ground, killing the pilot.
The question was simple: What happened, what when wrong?
There were many easy answers, like the pilot could have diverted to a nearby airport with better weather, a longer runway (e.g., Buffalo International), maintained more diligence to stay on all of the altitude and airspeed targets. But the bigger question remained: Why the flat spin?
Obviously, the pilot lost control while doing the go-around, and in a multi that means letting the airspeed get below Vmc. But a Vmc rollover tends to be a nose low spin, not a flat one. Lou and David shared their theories, which were similar: the pilot had been attempting to control the plane with aileron, not rudder. And when the plane got below Vmc, he tried to use the aileron to stop the spin. It brought the nose up, but did not stop the spin, instead bring the plane into a flat spin.
The lesson is that rudder is essential, and easy to forget about if not properly trained and reviewd.
Because of the time lost from the FAA web site and other appointments, we then moved onto the checkride portion.
After a few questions, Lou said that he'd get settled into the plane and once I had finished I could join him. "In other words," he said, "I'm cold."
I finished up the inspection and then got in.
I started the engines and then taxied out to the runway. Lou mentioned that they had just plowed the taxiway, but there wasn't enough snow to build up mini-walls on each side. The walls help stop the snow from drifting onto the taxiway.
I did the takeoff briefing: Full power, accelerate to and rotate at Vr (71 mph) and set the pitch to climb at "blue line" (85 mph). Once we have established a positive rate of climb, the gear comes up. After passing 500 feet above ground level, I lower the nose and accelerate to Vy (101) and reduce the throttle to 25 inches manifold pressure, and the propellers to 2500 rpms. At 1000 feet the fuel pumps come off one at a time.
Following that is the safety briefing: If there are any problems on the runway, the throttles come back to idle and we stop. Any problems in the air while the gear is down, we're landing straight ahead (no matter what). Once the landing gear comes up, any problems and we're a go (i.e., continue flying).
We took off and did the initial climb-out. There were some clouds a couple thousand feet up. He pulled an engine on me and I did the initial "identify, verify, and feather" procedure, though I only partially pulled the level back until he said "stop" and then set the "dead" engine to be "zero thrust" (instead of feathering it and shutting it down, it's set to a low power level to just overcome the drag it would produce).
I was using the rudder and a slight bank. Apparently I wasn't using enough rudder. Lou said to me, "Are you a Baron pilot? Are you a Baron pilot?!? ARE YOU A BARON PILOT?" I was kind of caught off guard. My response was, "No, I'm a Duchess pilot." He said, "THEN USE MORE RUDDER AND DON'T BE A BARON PILOT." He then throttled the "dead" engine back to normal power.
We steered around them a little, and did some initial maneuvering and then decided to climb on top and head north over Lake Ontario. Lou called Buffalo Approach to get radar flight following.
After climbing on top, we did a steep turn. It was OK, but not great. I don't know why, but that's been one of my weaker maneuvers in the Duchess. I think we then did some slow flight and ended it with a power off stall. While recovering from the stall, I think he pulled the engine again. I used a lot more rudder this time. I think he was happier with that.
He said the next thing is to demonstrate the complete shut down of an engine. I did the procedure. Really pull the prop level to the "feather" seting, and watch the propeller completely stop turning. Then pull the mixture to idle/cut-off, close the cowl flap, set the fuel tank on the good engine to cross-feed from the tank of the dead one, turn off the magnetos, and then run through the checklist to make sure nothing was missed.
Then we did a restart. Since this is not part of the test and it was his plane, he could help me on that. But the procedure is simple, make sure the plane is going 110 mph (lower the nose as needed), turn on the magnetos, push the propeller controller to high-RPM setting, and watch the propeller start to spin. The set the mixture to rich, and the throttle just a little open until the engine starts. Then adjust the throttle to maintain a low power until the engine warms up.
When I pushed the propeller controller forward, it spun about 180 degrees and then stopped. And then it spun another 180 degrees and then stopped. We were right at 110 and really the plane needed to be going a little faster to get the propeller to start spinning from just the airstream. So a quick touch of the starter helped, and soon the engine was coming back to life.
We did some other maneuvers and by this time the clouds had all lifted. Then he asked me what I'd do if there was an engine fire. I said I'd do an emergency descent. He said to demonstrate one. The procedure is simple, pull the throttles to idle, drop the landing gear once the plane is below 140 mph, and then lower the nose and descend at 140 mph.
I pulled back the throttles and then pushed way forward on the yoke. We were in a very nose low attitude. I looked over at the airspeed indicator and it said 120. I pushed harder on the yoke. At this point, we're descending at about 1500-2000 feet per minute, I'm looking straight out the window staring down onto Lake Ontario. We're doing 130. I push forward a little more. I was seeing nothing but water. Lou says that's fine and now I can recover to a level flight. I do.
After that it's a demonstration of Vmc. I throttle back an engine (I forget which one it was) and slowly raise the nose until I can no longer maintain directional control. At that point, I announce we're at Vmc, throttle back the good engine, lower the nose to "blue line" and then bring power back into the good engine and maintain a climb at Vmc. I think I did a decent job maintaining the airspeed.
Lou said that since the VOR was inoperative at Dunkirk that we would just do a GPS approach and that he would set things up (since I really wasn't all that familiar with all the deatils of that unit) and he'd give me vectors to final. I put on the hood. This is the two engine approach. It goes fine, and I try to be mindful of the altitudes. I also did not want to drop the gear too early. So once established on final, I lower the gear control lever. Nothing happens. I announce that the gear hasn't come down which is a "problem." He asks what I would do. I said I'd check the circuit breaker panel (the gear breaker was pulled) and he said, "everythings fine there, so then what would you do." I briefly describe the emergency gear extension procedure, which is to open the little door in the floor by my feet and take the tool that's sitting in the pocket in the door, and turn it 90 degrees and the gear would drop and lock. He says that's fine, and lets me push the circuit breaker back in and lower the gear.
As we get to the minimum descent altitude, he tells me that I can't see the field. So I perform a go-around. As we climb out he fails an engine. This time I'm better about it, "identify, verify, feather" (with the word "stop" coming before I complete the last step). I establish a small bank, proper rudder and get a climb going. He then gives me the engine back and says that he'll give me vectors on two engines to "fast forward" to the point where we start the approach and then we'll do a single engine approach.
He also sets up the GPS for an approach from the opposite direction. Out where the approach would begin, he said I can pick whichever engine I want to "fail." I do (I forget which one I chose). The approach goes decently, and we "break out" early enough. I do the simulated single engine landing. We clear the runway and taxi back to the end. He says we'll do a visual landing in the traffic pattern.
On the runway, I apply full power and just as we start to roll, he pulls the fuel lever on one engine to idle/cut-off (he covered the fuel levers with a paper so I couldn't see which one). I immediately bring both throttles back and keep us on the runway. He says that's fine, pushes the fuel lever back to full rich and tells me to continue the takeoff. We do. As we're turning left downwind, he says that we'll do a short-field landing.
As a minor point, I had never done a short field landing in that plane before. He tells me that for a short field approach, instead of 1.3Vs we'll use 1.2Vs and asks what that would be. The normal final approach speed is 85, so I say (guess) 80, which turns out to be correct (enough). So we do that. It's also a precision landing but the runway is still covered with a layer of snow, so I can't really see the 200 foot long white 1000-foot markers (1000' past the start of the runway), which makes it easier. I land, and he says we can taxi in.
I do the shutdown procedure and that's it. I passed the checkride.
David flew us home in the Mooney, as I was pretty burned out at that point.