Principia, Inc.

It had been a seven hour upwind slog from New York to Oklahoma last month. It was reasonable to be tired. But I should have known something was wrong when I gasped for breath after pushing our plane, a Columbia 400, into the hangar.

It wasn’t until over a week later, finally on the way to the doctor after days of delay, that I admitted it might be blood clots in my lungs. Sure enough. But at first, in the days following our return flight to New York in the flight levels, I thought the long hours on oxygen or decompression might explain my symptoms. I researched absorption atelectasis and the chokes.

Absorption atelectasis

Alveoli provide extremely efficient oxygen transfer from inhaled gases into solution in the blood in the surrounding capillaries and then into chemical bond with hemoglobin. If those inhalations contain little else but oxygen, and if their pressure is substantially reduced by high altitude, then there is a tendency for the alveoli to collapse. The positive pressure oxygen systems used by military pilots in unpressurized aircraft at high altitudes counteract this tendency. On our return trip at FL230, we were using the type of continuous flow system that is standard in light aircraft, approved for use up to about 25,000 feet. These masks have rebreather bags that help conserve oxygen by mixing some exhaled carbon dioxide and water vapor into inhalations. The presence of those gases in inhalations would also counter collapse of alveoli by tending to hold the overall pressure higher even when the oxygen partial pressure decreased rapidly due to absorption into the blood.

I learned atelectasis is a risk in such flights, but began to look elsewhere after a few days. The condition was not improving, and I had flown similar profiles before, albeit not as long.

The chokes

Breathing difficulties, “the chokes”, are among the most serious effects of sudden decompression. Divers who surface too quickly experience a rapid decrease in gas pressures which can lead to the release of super-saturated nitrogen in the form of bubbles trapped inside tissue. In the more common “bends”, the bubbles are trapped in joint tissue, causing pain upon use of the joint. When reproduced in experimental animals, the chokes reveal nitrogen bubbles trapped in pulmonary arteries, leukocytes surrounding them, then development of lung edema.

I contacted medical personnel at the closest facility with a hyperbaric chamber to discuss this possible diagnosis and to see if they thought recompression would be appropriate in my case. As we talked through the history of my symptoms, it quickly became clear that it was not the chokes. My decompression from sea level to FL230 had occurred slowly and resulted in no symptoms at altitude. My recompression had already occurred in the descent for landing days before.

it was now over a week since I first showed the symptoms of “shortness of breath” on exertion, pushing the airplane into the hangar in Oklahoma. That was before the high altitude legs. I had been in relatively good condition immediately before the appearance of that symptom. It was getting progressively worse, not better. I was beginning to notice my nailbeds turning pale with mild exertion, such as slowly climbing a gentle hill. I sometimes had to sit down to get my respiration rate to slow down after climbing stairs. Something had changed radically and was continuing to change.

Deep vein thrombosis (DVT)

In DVT, clots form in the veins of the legs, most commonly in the calves, and most commonly in the left leg. One theory for the higher frequency in the left leg cites the routing of the left common iliac vein behind the right common iliac artery, where it can be compressed against the lumbar spine. This phenomena and theory are referred to as the May-Thurner Syndrome, in honor of R. May and J. Thurner, who first described it in 1957.

Eight years previous I had been diagnosed and treated for deep vein thrombosis (DVT) in my left leg. My recovery appeared to have been complete. There had been no recurrence of swelling in the leg. The presumed cause had been determined to have been very long flight legs in a survival suit, while limiting fluid intact; I had subsequently been persuaded to stop doing the transatlantic flights that required those conditions. In addition, in the years immediately following the DVT, I had been careful to keep hydrated on longer flights, to move my legs regularly, and to watch for any signs of swelling. As time passed with no sign of recurring problems, I had grown lax, limiting fluid intake on long legs to avoid the need to urinate and failing to move my legs regularly.

DVT can also lead to complications in the legs referred to as chronic venous insufficiency (also known as post-thrombotic syndrome). This condition is characterized by pooling of blood, chronic leg swelling, increased pressure, increased pigmentation or discoloration of the skin, and leg ulcers known as venous stasis ulcer. Since my 2001 DVT, I have had mild post-thrombotic symptoms — splotchy red discoloration on the skin of my left ankle, and occasional soreness in my left calf. But these mild post-thrombotic symptoms had not changed recently, so initially I did not consider my new shortness of breath symptom related to my earlier DVT.

I walked into my internist’s office without an appointment early on the day after Labor Day 2009, having finally decided it was time to seek medical help. I was there before he was. When he came in and heard the symptom, he immediately examined me and ordered an EKG. Judging by a change in the EKG, added to the account I gave him of the flights to and from Oklahoma, he came up with a diagnosis of pulmonary emboli within minutes and ordered an ambulance to take me to the nearest emergency room, where I was put on oxygen and a chest CT scan revealed extensive clots in the pulmonary arteries in both lungs. I was injected with a form of heparin, Enoxaparin sodium (manufactured and marketed by Sanofi-Aventis under the tradename Lovenox), to counter the formation of further clots and allow my body opportunity to begin to clear existing clots. I was also started on longterm warfarin to reduce clotting factors and lower the likelihood of further clot formation.

Pulmonary emboli

Blood clots that detach from the place where they originate and travel through the circulatory system to lodge elsewhere are termed emboli. If they lodge in the blood supply to the lungs, they are termed pulmonary emboli (singular: embolus). Almost all pulmonary emboli originate in the veins of the legs, where venous pressure is lower and conditions favor clot formation. A doppler scan of the veins in my legs showed the right leg clear of clots, while my left leg showed a new clot behind my knee. An echo cardiogram showed that my right ventricular pressure was 60 mm at rest, instead of a normal 20 mm, revealing the extent to which my heart had been working to overcome the blockage in my pulmonary arteries.

Heparin

Subcutaneous injections of heparin immediately increases the activity of antithrombin III, which inhibits clotting factors Xa and IIa in the contact activation coagulation pathway formerly known as the intrinsic pathway. Factor Xa in particular is the catalyst for the conversion of prothrombin to thrombin. Inhibiting it therefore reduces thrombin which reduced fibrin formation in clots.

Warfarin

Administered orally, it gradually results in the production of undercarboxylated versions of clotting factors II, VII, IX, and X; these relatively inactive factors are collectively referred to as PIVKAs (proteins induced [by] vitamin K absence/antagonism), and individual coagulation factors as PIVKA-number (e.g. PIVKA-II). Specifically, warfarin acts by inhibits the VKORC1 subunit of vitamin K epoxide reductase, responsible for recycling vitamin K epoxide back to vitamin K and vitamin K hydroquinone.

Because it also negatively impacts the regulatory proteins C, S, and Z, it initially has the undesireable effect of increasing clotting rates, which is an added reason heparin is administered with it initially, especially when warfarin concentrations are being rapidly increased.

Prothrombin time

To gauge the correct dose of warfarin, clot-formation time is measured by exposing a blood sample is to animal tissue. INR (International Normalized Ratio) is the standard used in this measurement, endorsed by the World Health Organization. Normally, without anticoagulants, human INR measures in the 0.8 to 1.2 range. Desired range for anticoagulation therapy is 2.0 to 3.0. Specifically, what is being measured is the tissue factor coagulation pathway, formerly known as the extrinsic pathway, which is impacted by warfarin therapy.

Once the desired INR 2.0-3.0 is established, there is little danger of excessive bleeding. A study published in 2007 concluded that in 2004 in the US, there were 30 million warfarin prescriptions and 46 deaths primarily caused by anticoagulation. In the event an antidote is needed to warfarin, in order to increase coagulation rates, Vitamin K and fresh-frozen blood plasma can be administered.

Aspirin

After my bout of DVT in 2001, my internist took me off warfarin after the clots had been resorbed, and suggested a longterm low-dose (81 mg daily) aspirin regimen. I adhered to that for a few years, then stopped. Now that I have suffered a second more serious occurence with emboli, my internist is recommending longterm warfarin therapy. A condensed account of how aspirin interfers with clotting by altering platelets: thromboxane A2 is normally released by platelets to attract other platelets when physiological conditions trigger the clotting reaction; aspirin irreversibly inhibits thromboxane A2 release by inhibiting cyclooxygenase 1 (COX1); the platelets that are inhibited cannot recover normal functioning, therefore clotting will not return to normal until all or most inhibited platelets are removed from circulation and replaced by normal platelets. Other NSIDs also interfer with platelet functioning, but not irreversibly. Therefore they are not useful in longterm hypercoagulation therapy.

Risk factors for DVT and PE

Surgery, hospitalization, advanced age, obesity, infection, dehydration, immobiization, estrogen-containing contraception, tobacco use; prolonged air travel can combine dehydration and immobilization.

Cause of deaths with PE

The immediate mechanism of death is commonly suffocation or heart failure. A common misconception is that pulmonary emboli can cause stroke. The reverse is true. Clots that originate in the brain can sometimes break free and travel to the lungs, where then can result in sufficient blockage to circulation to result in death.

FAA requirements

FAA regulations and requirements related to my second class medical certificate: unless and until my lungs are clear of clots and my warfarin therapy stabilized, I will not be acting as pilot in command or other required pilot crew member, in compliance with Title 14 CFR, section 61.53, Prohibition on Operations During Medical Deficiency.

By the time of my next pilot’s medical examination, the FAA is going to need see information on family history of thrombotic disease. My father died of heart failure in 1980 at age 67, several years after being diagnosed with congestive heart disease. He had had rheumatic fever at age 13, and likely had mitral valve damage as a result. I rarely saw him in shorts, and rarer still sockless. But when I did, I remarked on the shiny red skin on one or both ankles, often with raw ulcers. It is likely these were post-thrombotic symptoms of undiagnosed DVT and/or PE.

The FAA will want to see recent tests on levels of protein S and C, which are essential components in the normal anticoagulation cycle that keeps unwanted blood clots from forming. Activated protein C (APC) in particular inactivates clotting factor Va, slowing the clotting process. If testing shows APC is not functioning normally, then the FAA will request testing for the factor V Leiden gene mutation, which is usually the cause of this abnormality. It is named after the city Leiden in The Netherlands, where it was first identified in 1994. These tests are typically performed several months after a thrombotic event such as the PE I just experienced. I will need to be taken off warfarin temporarily to allow the tests to accurately show concentrations and effectiveness of protein S and C.

Looking ahead

For the time being, I will be taking warfarin orally in sufficient doses to keep my INR in the desired 2.0-3.0 therapeutic range, checked regularly by my internist, and potentially self-tested on a device such as the InRatio2 by HemoSense. In the longer term, testing is been done on a class of drugs called direct thrombin inhibitors that promise improvement in dose-determination, monitoring, and interaction with food and other drugs. They also generally have a lower half-life than warfarin, permitting easier restoration of hemostatis when required.

One such drug, Ximelagatran, initially showed good efficacy compared with warfarin, but recently development was stopped by manufacturer AstraZeneca because of reports of liver enzyme derangements and liver failure. Dabigatran is under development for similar indications. Recent studies have indicated Dabigatran is slightly more effective than warfarin.

I will renew my lost dedication to limiting long flight legs. When necessary, I will be careful to move my legs and contract my calves regularly, getting up and moving around the cabin when practical, and drinking adequate fluids. I plan to experiment with a Texas (condom) catheter as a possible means to easier urination in a cockpit environment.

It may be advisable to do regular doppler scans of my legs to monitor for new clot formation, especially since there were no observable symptoms in my left leg this time, though a new clot had formed.

Warfarin

Heparin

Coagulation

Steve Demy flies a Columbia 300 based near Vancouver BC, and relishes the challenges of long flights. He watches the wind aloft and meticulously plans ideal altitudes as well as fuel and prop settings. He has regularly made non-stop Vancouver-Toronto trips, a distance of about 1800 nm.

Today I departed Bend OR in a new Columbia 400, N1134N, bound for N07 eventually, in Lincoln Park NJ, where the TKS anti-ice system is scheduled to be installed. With tailwinds forecast to be 75 knots or better, and icing airmets below 18,000 ft enroute, I chose FL250 as cruising altitude, both to take maximum advantage of the tailwind, and to stay well above clouds that might contain ice.

It worked out well. Five hours after departing at about 0945 PDT, N1134N touched down in Madison WI, (KMSN), having covered about 1400 nm on 88.5 gallons of 100LL. The wind held up at 75-85 kts on the tail, gradually diminishing and backing around from west to southwest, allowing us to average about 280 kts groundspeed. It was my demi-Demy leg. I suppose I could have stretched it for another hundred miles or so by throttling back. But my aim was different than one of Steve’s record attempts. I just wanted to get well clear of the mountains, and more than half way to my destination.

Weather was as forecast, with a broken to overcast layer below from Oregon to the Minnesota border, smooth to light turbulence, and almost exactly standard ISA temperature at FL250, -36 dC. Onboard weather data showed two quasi-stationary/cold fronts, one roughly paralleling our course to the north, the other to the south. East of the South Dakota/ Minnesota border, the undercast disappeared, replaced by a thin overcast above. That layer dissolved near Minneapolis, just before we crossed the area where the charts showed we would be expected to cross the benign cold front.

Over the Rockies, one’s thoughts turn to tactics to deal with the unlikely possibility of diverting to a high elevation alternate in response to a system problem. The topographic information provided by the Garmin G1000 MFD helps one picture the situation below, with a precise digital readout of maximum and minimum elevation within the selected range, along with a black bracket to reinforce the same data graphically on the legend, and a white horizontal line against the blue color representing the sky to show your altitude relative to the terrain below.

Performance figures, noted at two points enroute:

March 17, 2009 1720Z

MP 31.5 inches
RPM 2480
FF 16.0 gph
TIT <1615 dF
EGT <1575 dF
CHT <340 dF
OIL TEMP 171 dF
OIL PRESSURE 45 psi
TAS 210 knots
IAS 140 knots
GS 280 knots
FL250
OAT -36 dC (IAS -1 dC)

March 17, 2009 1900Z

MP 31.5
RPM 2480
FF 16.8
TIT <1625 dF
EGT <1585 dF
CHT <360 dF
OIL TEMP 176 dF
OIL PRESSURE 43 psi
TAS 218 knots
IAS 148 knots
GS 290 knots
FL250
OAT -36 dC (IAS -1 dC)

Use of combustion in transportation began by burning fuels like wood and coal in the open to boil water to make steam to drive pistons to crank a shaft. Later combustible petroleum-derived fuels were burned directly over the piston heads. Pistons have been mostly replaced now in favor of turbines, with continuous combustion around the circumference of fine-bladed fans. Pistons continue to have advantages for some applications such as light aviation and road vehicles however. Though they are heavier per horsepower than turbines, with more moving parts, they are less expensive to manufacture and more fuel-efficient.

Old methods.

In 1983, when I learned to fly in an ancient C172, the only engine-related instruments were an oil pressure gauge and an oil temperature gauge. We were taught not to lean from full rich until we reached cruise, and never below 3000 ft. The method was to lean until RPM decreased, then enrichen a little. Imprecise. But given that this was a normally-aspirated engine, it was sufficient to keep us out of trouble. Besides, more precision needed to wait on improved instrumentation along with the improvements that fuel injection brought.

The turbocharged Cessna P210, N267LM, that Sape Mullender and I ferried from Schwäbisch Hall Germany to Waterloo Iowa in 2000 had an Economy Mixture Indicator (EGT) and a set of operating instructions P210 pg 1 and pg 2 designed to keep internal cylinder pressures at a safe level by keeping fuel flow sufficiently rich of peak.

With tuned injectors it became possible to ensure that all cylinders were receiving close to the same fuel-air charge. With improved instrumentation, it became practical to monitor key temperatures in each cylinder. Indirectly, that gave the pilot indications of internal cylinder pressures and enabled us to operate piston engine more efficiently by giving us the lean of peak option.

New methods.

I had been reading John Deakin’s Pelican’s Perch, but when Scott Marti first took hold of the mixture control in a Columbia 400 and pulled it back from 38 gph to 18 gph to demonstrate the big pull, I must admit I experienced an involuntary shudder. Old habits, especially in old dogs, take time to change. Now we all have flown thousands of hours using these new methods, and we know they work well. But they must be thoroughly understood if we want to use them most efficiently, and they are not reducible to a simple set of instructions.

Dr. Wayne Isom, who I frequently fly with, makes a comparison with new interns learning diagnosis. They come out of medical school with a set of useful numbers tatooed on their brains, like pO2 96% to indicate satisfactory respiration. He watched one miss a collapsed lung by focusing on that number and missing a rapid respiration rate.

More than once I have had pilots tell me they run their engines at 31.5 in MP, 2450 rpm, with FF at 18 gph. Like that 96, those are useful numbers, but not the whole story.

Future methods.

Low initial cost and low operating costs will keep piston engines in small aircraft for some time. We are likely to see changes in fuel. A company called Swift Enterprises for example, currently has a formulation called 702 in testing at the FAA William J. Hughes Technical Center at Atlantic City Airport in New Jersey. Preliminary tests showed the mixture, made from biomass, meets or exceeds the standards for 100LL avgas without any petroleum-derived components, and without lead. We are also likely to see variable timing on ignition systems with the advent of automated engine management systems such as the one called PRISM that GAMI is working on. The PRISM times the spark by directly referencing the internal cylinder pressure. We may see better designed fuel injection systems as well. Diesel engines like Thielert’s may become more widely used. They offer piston efficiency burning Jet A or road diesel fuel. Their Centurion 2.0, for example, is a turbocharged 4-stroke water-cooled 135 hp engine with a compression ratio of 18:1. Over 100 aircraft are equipped with it.

Animated model.

To focus discussion, a Flash animation has been developed, copyrighted by Principia Inc., 2009. Access is free to individuals for private use. Organizations or individuals who wish to offer the animation to the public will require a commercial license. To obtain a login, email Principia Inc.

Go to Principia client area to login and open the animation. It will persist in your browser’s cache, even without an internet connection, until you manually clear the cache or turn off your machine.

A Flash plug-in is available from Adobe if not already installed in your browser.

In the cylinder view, you will see a representation of the two flame fronts spreading out from the two spark plugs. Of course, that is an idealization. To see an actual video taken inside an operating cylinder, go here on YouTube.

Five red buttons on top control (from left) page, either panel or cylinder; pause crankshaft rotation; step one frame; zoom; select individual frames.

Three gauges and four sliders are active controls: MP, RPM, FF, SPARK, IAS, OAT, and OCTANE.

Fuel flow (FF) can be set independently by clicking on the FF pointer and moving it to the desired level.

Seven graphs display results of changes in variables, from top left, turbine inlet temperature (TIT), exhaust gas temperature (EGT), cylinder head temperature (CHT), internal cylinder pressure (ICP), brake horsepower (HP), air to fuel ratio (A:F), and brake specific fuel consumption (inverted), a measure of fuel efficiency (1/BSFC); colors are used to depict temperature changes, ranging from yellow (cooler) to red (hotter); colors are used to indicate air/fuel ratio ranging from light blue (leaner) to dark blue (richer); redlines on TIT and CHT are indicated.

Failures are selected with toggle keys along the bottom: INJECT/VAPOR for effect of reduced fuel flow from either plugged injector or vapor lock; PLUG for one failed spark plug; CHT-IND for a failed CHT probe; DETON for detonation; PREIGN for preignition; VALVE for an exhaust valve stuck partially open; and CRACK for a crack in the exhaust manifold

Variations.

MP (throttle) increase results in higher potential internal cylinder pressures as indicated indirectly by TIT, EGT, and CHT data.

RPM (engine and prop revolutions) increase tends to increase fuel flow due to increase in engine-driven fuel pump speed.

FF (fuel flow) adjust to show impact of air:fuel mixture changes.

Spark timing can be advanced or retarded to illustrate the impact of earlier or later ignition.

OAT (outside air temperature) increase tends to cause internal cylinder pressures (ICP) to rise by decreasing air-cooling efficiency; also tend to limit maximum LOP fuel flow by limiting available oxygen mass in the mixture.

IAS (indicated air speed) increase tends to decrease ICP by improving air-cooling efficiency.

Octane increase decreases the rate of flame front progress.

Examples.

Hot and high cruise: MP 32 IN RPM 2500 IAS 150 OAT -5 C (ISA +30 at FL250).

Already only about half as dense at this altitude, cooling air is also warmer than standard. Result is less efficient cooling, resulting in higher pressures and temperatures in the combustion chamber, reflected in higher TIT, EGT, and CHT indications. Induction air, being warmer, is less dense than standard, even after being compressed by turbochargers. It therefore contains fewer oxygen molecules, and is able to oxidize fewer fuel molecules. In light of those two factors, LOP fuel flow will need to be lower than standard to keep cylinder pressure (as indirectly indicated by TIT, EGT and CHT) at an acceptable level.

Cessna’s PIM for the 400, RC0500005HIM, Revision level H, pg 5-32, suggests a fuel flow of 15 gph might work under these extremely hot conditions, to keep TIT at a value at least 50 dF below peak. It might. But rather than setting that fuel flow, and accepting whatever TIT results, the prudent pilot will reduce fuel flow until TIT cools to 1625 dF or below, and monitor CHTs to ensure they do not rise much above 380 dF. More prudent still to set fuel flow to keep CHTs below 380, even though it might mean sacrificing airspeed. Suppose that 15 gph turns out to work. The 7.5:1 compression ratio in for the TSIO-550-C produces 13.7 BHP per gph LOP, yielding 205 of 310 maximum, or about 66%.

That same page in the same PIM predicts 231 KTAS under these same conditions, which yields 147 KIAS. Interestingly, that airspeed prediction is marginally better than what is predicted at standard conditions, despite the predicted lower fuel flow and lower BHP. What gives? Less dense air at warmer than standard temperatures has one nice positive outcome: higher airspeed due to lower parasitic drag.

Autogas: hypothetical but educational. Substituting regular US auto fuel for 100LL avgas.

FAA-approved STCs are available for most lower compression aircraft engine installations, but no turbocharged engines and not for Part 135.

Why just lower compression engines? The lower octane rating of auto fuels signifies among other characteristics, their lower auto-ignition temperatures. That is the temperature (pressure) at which a gaseous fuel mixed with air will supply the energy to support combustion. In a diesel engine, we want a relatively low auto-ignition temperature. That is because a spark plug is not used to trigger ignition. Instead, liquid fuel is injected directly into the cylinder once the piston is starting down after compressing air in the combustion chamber and thereby raised the temperature above diesel fuel’s auto-ignition temperature by compression. The injected diesel fuel immediately vaporizes and combusts, creating additional heat and pressure to accelerate the piston down.

However, in engines designed to burn avgas or autogas, a mixture of fuel vapor and air is pulled into the combustion chamber on an intake stroke, then compressed on the next stroke, during which process, it must resist the urge to spontaneously combust before ignited by the spark plug at the end of the compression process. Otherwise, the cylinder could be damaged. This is known as pre-ignition. For 100LL avgas, that auto-ignition temperature is about 440 dC. For auto fuels, it is much lower, about 260 dC. Hence the limitation to lower-compression aircraft engines.

About 1900Z on Saturday December 6, 2008, the pilot of Cessna 206 N6053B about 10 miles north of Traverse City requested a clearance for the ILS 28 instrument approach to land at Cherry Capital Airport (KTVC). The approach plate is provided in the links section at the end of this article.

His communications with Minneapolis Center make it clear he had encountered structural ice. The pilot was not able to land after flying the approach, lost communication contact with ATC, then lost control. The airplane was destroyed and the pilot killed when he struck a mobile home east of the airport.

He had departed South St. Paul MN Municipal Airport (KSGS) about 1615Z on an IFR clearance, cruising at 11,000 feet, headed home to Glens Falls, NY (KGFL), a distance of about 870 nautical miles. He had left a voicemail for Dick Bovey, his instructor in Glens Falls, in which he said he had received a weather briefing that led him to believe he would be able to fly above a storm then over the Great Lakes. We assume he meant he talked with a Lockheed Martin briefer, but we do not know at what time. Those briefings are recorded, but not available to the general public. There was an AIRMET ZULU for moderate icing below 16,000 feet over Wisconsin and Michigan along his route, issued at 1445Z by Chicago, and expected to continue beyond 2100Z thru 0300Z. Since he departed about 1615Z, he may have received his briefing before 1445Z when an earlier AIRMET ZULU issued by Chicago at 0845Z, valid at 1500Z and expected to continue through 2100Z, limited the icing forecast to altitudes below 12,000 feet.

A detailed study of weather conditions at the time of the accident by meteorologist and CFI Scott Dennstaedt will be available soon at AvWxWorkshops.com. In it, Scott concludes that the flight encountered lake effect icing conditions over Lake Michigan. For more information, contact Mr. Dennstaedt at Scott@chesavtraining.com, or (803) 802-2591.

The newly-minted instrument pilot, Keith Harris, alone in the plane, reported structural ice to Minneapolis Center and said he was turning around and heading back west. That might have been a better plan, with Green Bay, 100 miles to the west, reporting VFR conditions, while Traverse City was low IMC with one mile visibility in snow at the time he reported an increasing ice accumulation to Minneapolis Center. He rejected that plan, and accepted vectors to the approach, possibly because he had already accumulated ice along his route to the west. The radar track shows him climbing to 14,500 shortly before he requested the approach into Traverse City, probably in an unsuccessful effort to climb out of ice.

N6053B was not approved for flight in known ice (FIKI), and had only the usual limited anti-ice provisions in such non-FIKI aircraft. There was pitot heat, which helps prevent the pitot tube from becoming clogged with ice. A clogged pitot tube causes erroneous airspeed indication in the cockpit which can make it difficult for the pilot to maintain control. There was an alternate static source in the cockpit that would continue to provide accurate static air pressure in the event the two outside pressure sources became clogged by ice. If ice interfered, the reliability of the altimeter, vertical speed indicator, and airspeed indicator would be compromised. Both of those remedies would require that the pilot remember to use them. Given his inexperience and the effect of stress on his thinking, he may not have. One item of automatic ice protection on his airplane is a spring-loaded door in the air intake plenum that is designed to open if the filter covering the normal induction air intake becomes blocked by ice.

He reports that his windscreen is iced over. He did have a defrost control that would redirect cabin heat toward the bottom inside of the windscreen. But it is not designed to clear structural ice from the outside of the windscreen. The amphibious floats on the airplane made it less able to deal with an accidental icing encounter like this by provided additional surfaces on which ice could collect.

His communication with Minneapolis Center as he is vectored to the approach made it apparent he had a glideslope, but no localizer indication, even though both signals was being broadcast normally. There is a link below to these communications, edited to remove silences and communications with other aircraft. Ice may have already been building up to a level where it was beginning to interfer with the “whisker antenna” on the vertical stabilizer that is responsible for receiving those signals.

Shortly after he missed the approach, the air traffic controllers lost communication with him. It is possible that ice caused both of the COMM/GPS antennae to break off. Without help from controllers, and without a GPS signal, his best tactic would have been to use the compass and autopilot to fly away from the icy clouds being generated over the lake. But it is difficult, even for an experienced pilot, to continue to think clearly in dire situations, and he did not take that course. Equipped with floats, he may have been considering a water landing, though that would have been difficult in the strong winds and low visibility reported at the time, even if he had been able to locate the lake. At the time of the accident, there was a wind at the airport of 13 knots gusting to 26, with 0.5 miles visibility and vertical visibility of 500 feet in snow.

The radar track shows him flying around in circles about 15 miles east of the airport at low airspeeds for some time before he impacted trees and then the mobile home about 2015Z.

When he purchased this airplane in 2006, I provided Keith an introduction to the 206 and the Garmin G1000 avionics suite. In February 2007, I gave Keith and his instructor and designated examiner Dike Bovey additional familiarization training with the G1000 and the King KAP140 autopilot. Keith earned his instrument rating in February 2008.

Thanks to Doug Robertson and Airport-data.com for the use of the photograph of 6053B moored on Lake George this past summer. Thanks to Live-ATC.com for the audio, and to FlightAware.com for the radar data.

I offer Keith’s wife and family my heartfelt sympathy for their loss.

Links:

N6053B moored on Lake George, summer 2008

AvWxWorkshops.com

N6053B flight track

ILS28 at KTVC

ATC/Harris communications part 1

ATC/Harris communications part 2

I made a sort of living for a brief time in the 60s rebuilding air-cooled normally-aspirated engines that powered the thousands of VW beatles and vans on the road then.

So when engine monitors like the JPI began to appear in the cockpit, I had a basic understanding of what I was looking at, and I was intrigued.At first, I used these monitors just as more detailed EGT gauges. I would look to see that the engine really was running a little rich of peak after I leaned by the rough method all of us learned if we learned to fly a piston engine any time in the past 50 years — pull the mixture knob out until the RPM start to drop off, then push it in a little.

Then when even more detailed engine data began to appear in the Cirruses I was training in, I promised myself I would find out more about what goes on inside an engine. If we wanted to set up for “Best Economy” cruise, the Avidyne MFD engine page was prompting me and my trainees to go where previous rote engine leaning training had warned us never to linger for long: lean of peak. In older airplanes with normally-aspirated engines and EGT gauges, the POHs uniformly instruct pilots to run at peak EGT to achieve best economy.The turbocharged engines I first encountered were the Lycomings in the single-engine Cessnas. To achieve best economy, the Cessna POHs recommend running those engines at peak TIT. Then when I took Columbia 400 factory instructor training in Bend Oregon in January 2005, I was introduced to what was then a new and controversial idea, though it was then and still is standard procedure on the Columbias. It was dubbed “The Big Pull” and it originated in the Advanced Pilot Seminar engine management seminars. The idea is to pull the mixture knob out quickly and smoothly past peak TIT and into the lean of peak zone. Originally, the POH restricted LOP operations to power settings of 65% or less, but by the time I returned to Bend the next year for certification in the G1000 Columbias, TCM was allowing LOP in cruise up to 85% power.I had been avidly reading John Deakin’s engine management articles in his Pelican’s Perch series on Avweb.com. So when I got a chance to take an early look at a proposed online version of the popular APS course, I jumped at the chance to learn some more of the underlying science. I took another long look at the online course after it had been in public release for almost a year. By now (Fall, 2008) the online course has largely replaced the live course offered over a weekend in Ada Oklahoma. The live course will be presented this month October 24 thru 26.

This article will be a review of the concepts presented in that course, specifically as they relate to the operation of a turbocharged engine. After that review, I will offer a few suggestions on improving the format of the online course.I now require completion of this online course for all my clients transitioning into the Columbia 400, unless they test out with a score above 80% on the attached turbocharged engine management quiz.

The course begins with a review of the operation of an internal combustion engine and helped to re-familiarize me with some of the terms engineers use to discuss them. We are reminded that the spark is fired before the piston reaches top dead center, typically about 20 degrees before, and that this is denoted as minus 20 Theta. Combustion in the cylinder reaches a peak sometime after the piston has started back down, and the timing of that event is critical to the efficient and safe use of these engines. A mixture of about 40 degrees F. rich of peak TIT results in the fastest burn and the highest peak pressures. Temperature of the gases inside the cylinders at peak are about 3000 degrees F. EGT is much lower because most of that heat is dissipated through the cylinder walls and fins by the time the exhaust valve opens. EGT (and TIT) is not a good measure of peak combustion temperatures, but it is a fair measure of the timing of the combustion event in the cylinder. Best power, defined as maximum horsepower for a given MP and RPM, is obtained at about 75 degrees F. rich of peak TIT. Best economy, depending on power setting, is reached between 15-40 degrees lean of peak for relatively low power settings, and 40-90 degrees lean of peak for higher power settings.Early in the course, the concept of effective timing is introduced. ThetaPP is defined as the degrees of rotation beyond top dead center at which peak cylinder pressure occurs. It can vary with the composition of the fuel introduced into the cylinder. For example, autogas has a shorter latency period than 100LL and therefore promotes quicker progress of the flame front. Latency period is the time between the firing of the spark plug and the point at which combustion is well-started. When the rising temperature and pressure in the cylinder cause pockets of the fuel-air mixture to explode before the flame front can reach them, this is defined as detonation. To use autogas without creating dangerous levels of detonation, the spark event needs to occur later to make up for the faster burn time. The presence of somewhat less than about 2 grams of tetraethyl lead in each gallon of 100LL is there only to slow and extend the latency period. Contrary to popular belief, it has nothing to do with lubricating valves. By changing the blend, oil companies may be able to produce a fuel with almost as long a latency period as 100LL, reflected in the octane rating. The urgency of that search is increasing as it is becoming more difficult for companies to justify the cost and health danger of using leaded fuel.

Prop pitch directly limits engine RPM at high power settings, and higher RPM is conducive to later ThetaPP because when the pistons are moving faster, they stay further ahead of the advancing flame fronts. Another way of looking at it is that higher RPM leaves more room for the combustion event and therefore leads to lower peak pressures in the cylinders. The ideal point for the peak pressure has been found to be about +16 Theta. With a turbocharged engine, the pilot has the option of reducing RPM while maintaining quite high manifold pressure settings. Since there is excess air in the mixture when operating LOP, it follows that the mixture control also controls horsepower in this zone. So it is easy on the engine to reduce RPM when desired to extend range, provided mixture is kept lean enough to ensure that TITs remain in a safe range (see above) when expressed as degrees Fahrenheit LOP.  Additionally, the CHT should not be allowed above about 380°F, and the TIT should be kept at or below the published limit.The maximum certified continuous power setting in the Columbia 400 is 85% of 310, or 264. When operating LOP in an engine with a compression ratio of 7.5:1, you get 13.7 hp/gal. That means your fuel flow limit LOP is about 19 gal/hr to remain within that 85% MCP.

For simplicity, I introduce transitioning pilots to two cruise settings in the 400, Go Fast, and Go Far. Obviously, if you are really in a hurry, then you will be willing to live with high fuel flows and Go Really Fast. Then you will stay 150 degrees or more ROP. But going LOP only costs a few knots of airspeed, and it is much easier on the engine, so both of my introductory cruise settings are LOP. TIT is the guide, with a watchful eye on CHT. TIT redlines at 1700 on the 400, so 1650 is as hot as you want to run continuously, and I teach 1600 to 1625 to allow a little buffer on that. Want to go fast? Most of us do. To get to my Go Fast setup, provided the engine is set up reasonably accurately, leaning to a TIT of about 1625 will give you a fuel flow of about 18.5 to 19.0 gal/hr, with CHTs running below 380. I leave RPM about 2500 to 2550, just below the redline of 2600, and I back the MP off a little too, to about 32 inches, which puts it below the 33.5 inch maximum cruise setting. If any temperatures trend up, I fine-tune the mixture by leaning a bit, which will immediately cool the TITs and eventually the CHTs as well. At this setting, I see airspeeds on the 400 ranging from 180 TAS at low density altitudes all the way up to 220 TAS up in the flight levels.

Got a nice tailwind and you are thinking you can make that 1200 nm trip without a fuel stop if you save a little fuel? Then Go Far. As long as you are LOP, there are plenty of variations on this, and depending on just how strong that tailwind is and how far you would like to extend your range, you might decide to back off even more than this. I like to keep the RPM fairly high, so I recommend about 2400 rpm with about 30 in MP, and mixture leaned to about 15 gph, which will only cost you about 7-10% of your TAS.

LOP principles impact other modes of flight, starting with taxi and runup. If mixture is leaned aggressively, the gross mag and ignition check is more likely to reveal a weak spark plug, because with the leaner mixture, a stronger spark is needed. For this reason, APS recommends doing a mag check in the air at cruise LOP to get early warnings of weak ignition. On one mag, all EGTs should rise together and stay up. The flame front does not spread as fast or as far when ignited by just one plug. So when the exhaust valve opens at the bottom of the power stroke, the gases released are hotter, since the combustion event has ended more recently than it would have had it been set off by two sparks.

In the initial climb in the Columbia 400, I stay with the factory recommendation to climb all the way to cruise at full power, with mixture set full rich. If you are already LOP and fuel flow becomes restricted in the climb, temperatures and pressures drop, posing no threat to the cylinder. If the altitude change is minor, and the engine is already set up for LOP cruise, I do the climb LOP, carefully monitoring TIT and CHTs.

In descent from LOP cruise in a turbo like the Columbia 400 it makes sense to stay LOP and resist the urge to push the mixture in to full rich. Except for the increase in OAT and air pressure, the effect of which is minor at lower power settings typical in the descent, the engine is not aware of the decreasing altitude, so there is no need to add fuel to compensate. Worry about shock cooling appears to have no basis in fact. APS points to the long engine life typical in tow planes, which repeatedly climb at full power, then descend immediately at idle, day in and day out, with no evidence of damage from thermal changes.  Or Bob Hoover, who used to go from full power to full feather and back to full power repeatedly during his act.  His engines usually made TBO with no cylinder work.On final in the 400, I follow the APS recommendation to leave the prop alone, but I move the mixture ROP, leaned a little for the upcoming taxi. I have found that if I leave the mixture LOP, the engine is prone to quit on the rollout, which can be inconvenient.

In the event of a missed approach or go around, there is plenty of time to check that mixture and prop are full forward and the backup pump is on before adding power for the climbout.  On the other hand, I do not attempt to change a trainee’s habits if he/she pushes prop and mixture full forward on final, since there is no harm done to the engine at that point.

After landing, I follow APS and ignore the 5 minute cooldown waiting period that Columbia calls for. The thinking is that as much cooling as will ever happen has already happened in the low power approach and glide to the runway. A quick look at the engine page before shutdown will confirm that the 1-2-3 shutdown criteria will have been met in the taxi to the ramp. That 1-2-3 is an acronym for TIT below 1000 degrees, oil temperature below 200, and CHT below 300.  In fact, the data shows that temperatures will start increasing again while turning off the runway.

Finally some comments on the look and feel of the online APS seminar:

Overall, the online version is a very nice piece of work, and I commend APS for being pioneers in the use of the web for aviation training. The layout has a scrollable text box on the left side of the screen for each slide, most of which are static, but a fair number of which are animations, including some video footage taken in flight. At the end of each section, there is a brief quiz. There is a glossary you can refer to at any time to refresh your understanding of the technical terms. When I went through the course, I spent a total of about 10-15 hours on it, spread over maybe a dozen sessions. It is easy to log back in and pick up where you left off last time. The cost of access to the course for 60 days is $395.

What could be improved in future versions of the course? Some of the slides are a little difficult to see in their full detail, and adding a zoom capability would solve that. The animations could also benefit from zoom capability as well as the usual stop/start rewind/fast forward capability found on most video players. After you successfully pass a section quiz, you cannot go back to review information in that section. Access to the entire course via a table of contents would be an improvement. There is footage showing how readings change on a test engine’s instrumentation as throttle and mixture are changed, and as autogas is introduced. This would be more effective online if it were an interactive simulation, rather than a video recording of a test engine. And finally, there are video recordings showing changes on instruments during flight. Online, this information would be more clearly presented via an interactive simulation built from engine data dumped after the flight from the Avidyne MFD in a Cirrus, for example.But all of this will take time and money of course.

When and if the Prism electronic ignition is approved, and GAMI starts making serious money, maybe we will see these improvements in the course, along with others I haven’t thought of.

Links:

Pelican’s Perch

Advanced Pilot Seminars

Engine Management Quiz 

We are investigating which of the three single-engine turboprops would be best suited for the 600 nm trip my dad often made for Lily Tulip Paper Cup Co., from White Plains NY (KHPN) to Augusta GA (KAGS), with three coworkers, and two pilots, for a total payload of 1200 lbs, at a realistic average weight of 200 lbs for American men.

Goldilocks would break baby bear, the Meridian, with that test load, even at the higher 5134 lb maximum takeoff weight permitted on serial numbers higher than 157. With a typical empty weight of 3500 lbs, our load brings zero fuel weight so close to the maximum takeoff weight, that all the Meridian could do would be to fly around the departure airport for a short while. So baby bear is out of the competition.

The Pilatus 12, papa bear, has more than enough capacity. With a maximum takeoff weight of 10,450 lbs, and a typical empty weight of about 6200 lbs, the test load puts the Pilatus at 7400 lbs zero fuel weight. To be conservative, we will allot 250 lbs of jet A for taxi and climb to FL250, another 720 lbs for 2 hours of high speed cruise at about 260 KTAS, 170 lbs for descent and landing, and a generous 360 lbs for reserve, totaling 1500 lbs of fuel, for a take off weight of 8900 lbs.

Double the distance to 1200 nm and add a 60 knot headwind, and you are at the Pilatus’s limit, close to maximum takeoff weight because of the extra fuel required, and too slow to be practical for a that trip.

How about momma bear, the Socata TBM? If she is capable of making the test trip comfortably, the company could save on purchase price and operating expenses. With a maximum takeoff weight of 7400 lbs and a typical empty weight of 4800 lbs, our test load brings us to a zero fuel weight of 6000 lbs. The TBM850 needs about the same 200 lb allowance for taxi and climb to FL250, but once there, the ISA performance charts appear to show a significantly higher true airspeed of 310 vs the Pilatus’s 260 knots, but with a significantly higher fuel burn of 450 lbs per hour, vs the Pilatus’s 360 lbs per hour. Awarding the TBM about 30 minutes less time at cruise, my calculations show only slightly less fuel required for the test trip, 1400 vs 1500 lbs for the Pilatus. That fuel load will put the TBM inside the maximum weight allowed for take off, but just barely. So only if this crew and these passengers will not routinely need to fly further than the 600 nm test trip would this airplane work out.

That is unlikely, making the Pilatus the only truly practical choice, which is why there are more than 800 of them in service worldwide.

Let’s suppose you are involved in a business that involved frequent trips to locations less than 1000 nm apart, typically by yourself, but sometimes with up to three colleagues. Let’s also say those destinations are often closer to smaller airports that do not have scheduled service.

You might benefit by setting up an air service for your business. If you only use your airplane(s) to transport people directly involved in your business, and do not offer the service to the general public, your operation would fall under the less stringent FAR Part 91 requirements. Even so, you would be well advised to act as if it were regulated by the more stringent rules in Part 135 or 121.

To maintain a consistently high level of safety, Part 121 and 135 operations are carefully designed and managed. Professionals specify and acquire suitable aircraft, set up thorough maintenance procedures, design appropriate operational standards, and carefully train and supervise pilots..

Preliminary 2007 aviation accident data released last week by the NTSB continue to show Part 135 and 121 operations are substantially safer than Part 91..

Scheduled Part 121 operations logged a total of 18,700,000 flight hours, with 24 accidents, no fatalities. For non-scheduled Part 121, the totals were 605,000 hours with 2 accidents, one fatality..

Part 135 commuter operations logged 302,000 hours, 3 accidents, no fatalities; for on-demand Part 135, the totals were 3,668,000 hours, 62 accidents, 43 fatalities..

Part 91 operations logged 23,835,000 hours, 1631 accidents, with 486 fatalities..

When I was a kid living in Valhalla New York, my dad often had business trips to a Lily Tulip paper cup manufacturing plant in Augusta Georgia. That meant getting to La Guardia Airport (KLGA), flying to Atlanta Georgia, and then driving a rented car from there to Augusta. Westchester Airport (KHPN) is just a few miles from where we lived, and Augusta (KAGS) is a serviceable airport too..

Making money in a scheduled Part 121 air service between airport pairs like Westchester and Augusta, about 600 nm apart, would mean setting pricing as high as that small market could bear to pay for shorter travel times due to direct routes and avoidance of the major airports where delays are increasingly common. And that would mean choosing appropriate equipment for that route. Airplanes would need relatively low capacity and efficient operating characteristics. But they would also need to be to be reasonably fast and comfortable. A Gulfstream or Falcon, with fifteen available seats, might be a good choice provided the airplanes could be kept busy. Keeping them busy would probably mean expanding service to other similar nearby markets to allow equipment to be moved easily to accommodate fluctuating passenger loads..

On-demand charter service for a small group of business travelers on a route like this could use smaller equipment like a Hawker or Citation with about six seats, gaining efficiency without significant sacrifices of speed or comfort..

But companies with enough internal demand for flights to and from less popular destinations frequently find they can justify the expense of running their own aviation operations. If Lily Tulip had stayed in business, maybe they would have hired professionals to set up a good tight Part 91 aviation department following FAA guidelines from Part 135 or 121, and then guys like my dad would have had quick safe trips to places like Augusta Georgia..

Next: a look at three single-engine turboprops, the Pilatus 12, Socata TBM 850, and Piper Meridian. We will see if any of them might be suitable for hypothetical 600 nm Lily-Tulip flights from KHPN to KAGS with up to four passengers..

Contact us to find out if an internal aviation operation would make sense for your business. Email info@principiainc.com or call 917-841-2362..

Links:
2007 NTSB Aviation Accident Table

I recommended that a client sell his airplane last week, but he has not taken that advice.

A green private pilot with an instrument rating that is barely wet, I had agreed to work with him on a transition into a single turboprop. He had already selected not only the model he wanted, but the exact bird, a 2002 Piper Meridian, PA46-500TP.

The first time he ignored my advice was when he refused professional help in the purchasing process. Had he accepted that help, he might have realized that a turboprop is vastly more airplane than required by his flying needs. Failing that, he might have been steered to a more capable model. But failing that, he would very likely have been steered away from the particular plane he bought, which was overpriced by at least ten percent.

He and I had worked together already for a long time, first on his instrument rating, then on his transition into a faster single-engine piston airplane. So I knew his penny-wise, pound-foolish weaknesses. The die was cast when a friend recently bought a turboprop, and he needed to make amends. However the friend has the flying skills and financial resources that allow him to easily handle a Socata TBM, the model in the middle of the turboprop market, between the Meridian and the Pilatus 12.

When it became clear that he was going to buy the Meridian, I told him it was an impractical decision, but agreed to train him. We took the ground and simulator training course at SimCom together. In the simulator, he proved that he needed more work before he would be competent in an essential skill for this and any other airplane — hand-flying approaches in the clouds. But I was resolved to fly with him as long as it took to reach competency. We flew the airplane home together and he began making immediate plans for long flights with passengers. I began to realize that his goal was to log the minimum hours that would permit him to solo. I became alarmed when I realized that we were rapidly approaching those mimimum hours. When I work with a candidate for an instrument rating or a transition, I can decide when a pilot can safely advance to his goal. But when the goal is defined only in terms of logged hours, I need to rely exclusively on the pilot’s agreement with my judgment of his abilities. 

On one of our flights together with passengers, an occasion arose that allowed me safely to give him a clear view of how much he had to learn. We were being vectored to an ILS approach when the glideslope indicator on his electronic HSI failed. The ceiling was high enough, and the final approach course was free enough of nearby obstacles, so I asked him to hand-fly the approach using the backup glideslope. It was a wild ride, accomplished only with much coaching. But in conversations afterward, he focused on the possible causes of the instrument failure rather than on the need for more partial panel training.

Had he been willing to concede he needed put off the goal of piloting the airplane solo for the foreseeable future, and resolve to fly it only with me or other competent co-pilot, then I would have been willing to continue training him. But with no sign of that concession, and with the time rapidly approaching when I would have no control over the operation of his airplane, I recommended he sell it.

Tim Slifkin hit a bird at 6000 feet, over the Hudson River. It was night, and he was on his way back to Morristown NJ from Nantucket MA in a Mooney cruising at about 140 knots, The windscreen cracked on the right side but held together. The first-time passenger sitting on the right was startled, but reassured by the pilot’s calm demeanor. Fifteen minutes later, they landed at Morristown. An examination of the airframe revealed signs the bird had been sliced by the prop, glanced off the engine cowling, and then off the windscreen. Feather samples were sent to Rutgers University, where the bird was identified as a Canada Goose (Branta Canadensis). Adults have a typical wing span of about five feet, and weigh about fifteen pounds.

While you currently see single-engine turboprops like the Pilatus PC-12 in passenger charters, you do not see single-engine jets.  One major reason may be FOD, foreign object damage. Air entering the Pratt and Whitney PT6 on the Pilatus must pass through a screen and make a 180 degree turn to enter from the rear. This reverse-flow design protects the delicate turbine blades from birds, ice, and other objects that can be sucked into a conventional jet or turbofan engine, where reverse flow is not an option.

As part of the certification process for aviation jet engines, four pound chickens, or reasonable facsimiles, are fired into the engines from air cannons at 180 knots to simulate bird strikes. Contrary to popular belief, an engine does not need to keep on producing power after such abuse. It simply needs to survive long enough for an orderly shutdown. But that is not good enough if the airframe does not come with a backup engine.

There are two companies working on single-engine personal jets, Cirrus and Diamond, both using the Williams FJ33 engine. There is also the PiperJet, planned around the Williams FJ44-3. But I do not think Piper will survive to build that airplane. (More on that subject in a later post.)

On the Cirrus version the engine is mounted on empennage, with the intake just behind the cabin and the thrust nozzle between two stabilizers in a V-tail. The Diamond uses two intakes mounted to the left and right of the cabin, just above the wing. The engine is integrated in the empennage, with the single thrust nozzle mounted on the bottom.

In 2000, Sam Williams, founder of Williams International, manufacturers of the FJ33, along with well-known aircraft designer Burt Rutan, were granted U.S. patent no. 6089504 on an airframe design that attempts to address single-engine jet FOD. The basic idea is to protect the intake by placing it close to the fuselage and behind a forebody (the cabin) where large objects like birds are less likely to be able to reach. Reached by email, Mike Van Staagen, Vice President for Advanced Development at Cirrus stated categorically that this patent was not used in the design of the Cirrus Jet. However, to the untrained eye, the intake on the Cirrus Jet does appear to be protected to some extent by its position behind the cabin, particularly at the high angle of attack required at takeoff.

The question remains how Cirrus and Diamond plan to address the overall issue of FOD, including bird strikes, and damage or malfunctions caused by ingestion of other foreign matter including hail, rain and ice; also other causes of compressor stalls and flameouts in other turbine engine installations, including crosswinds and the sideslips that pilots are prone to use to deal with them.

Pilots familiar with the Boeing 727 know that engine #2, which has an intake with an S-turn, is more prone to compressor stalls during takeoffs in crosswinds than either #1 or #3, with straight intakes. High crosswind components across the #2 intake disrupt the airflow into the intake, where it is already disturbed by the turn. This tends to lead to variations in the pressure field created by the compressor blades. Those variations can ripple through the engine and in extreme cases cause compressor surge and flameout.

Turbine designs include features designed to help minimize compressor stalls. All utilize bleed valves to relieve excess pressure and some have a system for changing the angle of the stator vanes, which are fixed to the engine body and are not part of the turbine, to help increase air pressure in case of disrupted airflow through the intake.

All turbine operating instructions call for igniters, normally switched to OFF or AUTO for normal operation, to be turned ON for takeoff and for any other situation where engine failure would be particular troublesome or something might possibly interfere with combustion, as when penetrating heavy rain or turbulence. Provided one has enough altitude, restarting a piston-driven propeller airplane is a routine operation. Restarting a turbine in which combustion has ceased is not routine or quick and therefore requires significantly more altitude than what is needed to restart a piston engine. A restart is not an option when an engine fails on takeoff in any airplane. Turbines and turbofans are at higher risk of engine failure on takeoff due their vulnerability to ingested foreign objects and crosswinds. Up until now, except for the injector seats employed in single-engine military jets, the only acceptable solution to the risk has been to install a minimum of two turbine engines, either one of which can supply sufficient thrust for a successful takeoff.

We need to know more about what Cirrus and Diamond are planning to do to deal with this issue.

Links:

Video of bird strike

Sam’s and Burt’s patent

Beginning May 25 and ending May 31, 1999, I transported a lovely old green and white Cessna 180 K, N180BB, with a big Continental O-520 engine from Hayward Air Terminal near Oakland California to Dublin, Ireland. Enjoyed it immensely, and managed to avoid ground looping. There are said to be two kinds of taildragger pilots: those who have ground looped, and those who will. I am pleased to still be in the second category.

The last leg, twelve hours over the desolate and unforgiving north Atlantic, was the most memorable. I got a wake up call at 0400Z, 0130 local time for a 0230 briefing by the FSS at St. John’s, Newfoundland. For my crossing, they had arranged a nearly perfect day, with strong tailwinds and practically no significant clouds. The only flaw in the weather was a disturbance passing north of the field that was kicking up rather frisky breezes. I took off in the dark on Runway 20, with the wind at 260 degrees and 20 knots gusting to 30. Runway 29 would have been some help, but it is closed this summer for resurfacing. On the other hand, if I had to deal with a 20 knot crosswind in a taildragger, I would rather have it on the right. I got off ok, even at 25% over gross, and climbed into the arms of a lovely southwest breeze that carried me and my little craft safely all the way to Ireland.

I had ground speeds around 150 knots in the climb out of St. Johns, with the cold front bearing down on the airport and compressing the distance between the isobars. The crew at the FSS had given me the very thorough briefing typical of the Canadian weather service. I took their advice and flew southeast a little while down at 7000 feet until clear of the broken cloud layers circulating around the low, then turned east and climbed to 10,000 feet on course in smooth air. It wasn’t long before the stratocumulus layer ahead was lit up pink by the sunrise. That is a beautiful and heartwarming sight. I suppose one feels that if the sun still rises, then things are likely to go well. This is a line of thought that is not likely to occur to the average person, rising from their familiar bed to begin their familiar routine. It is more likely to occur to a pilot setting out alone over the ocean, acutely aware that he is entrusting his life to a mechanical device and to the whims of nature.

As the day brightened I saw that the stratocumulus layer below was nearly 100 percent overcast, with only occasional breaks through which glimpses of a forbidding gray ocean were briefly visible. Overhead a thin wispy cirrus layer gradually gave way to clear skies. For the first half of the trip, the ground speed rarely dropped below 160 knots, and the undercast gradually broke up to a scattered layer by 35 degrees west longitude.

My clearance was 48N50W, 50N40W, 51N30W, 52N20W, 52N15W, DOLIP, and then direct to Shannon (EINN), which is 52N08W. I had filed Dublin (EIDW) as an alternate, planning to go on if fuel was not a problem once I got near Shannon. I plugged EINN into the GPS, hit direct, and used that great circle for guidance, not about to zig and zag over the Atlantic to make precise waypoints prescribed by some clerk at Shanwick. As it turned out the waypoints were very close to the great circle.

A transatlantic flight quickly leaves both radar and VHF radio range. This is why HF radios, with their theoretically longer range, are a requirement, along with position reports. The HF radio which tested out ok on the leg from Bangor to St. Johns, turned out to have a very limited range. I was assigned the primary frequency of 5618 with backup 2899. For the first hundred miles of the trip, I could hear fairly clear conversations on 5618. Not understandable, of course, since the temporary installation straps it to the top of the ferry tank behind me with no connection to the audio panel. By the time I needed it for a position report at 50N40W, the frequency had fallen silent, and there was no response from Gander Radio.

The solution was to request relays from airline crews overhead on the air-to-air frequency, 131.8. The conversations often led to other topics like what kind of a Cessna was I, how much fuel did I have on board, what was the ride like way down there, how much did they pay me to do this, etc. But the initial conversation always went something like this:

Me: Any aircraft on 131.8, Cessna N180BB, relay request.
Them: N180BB, Air Canada 869, go ahead with your request.
Me: N180BB POSITION 49N46W 0730Z FL100 ESTIMATING 50N40W @ 0830Z NEXT 51N30W
Them: (readback)
Me: Readback correct.
Them: Let me pass this on to Gander. Call you back in a minute.
Me: Appreciate it.
Them: OK, Gander copies your report.

My position reports were passed on by Canada 3000 311, an Airbus 330; Elite 379, a B757; Air Canada 869, Speedbird 503, a B767; KLM 687, United 971, Speedbird 209, and United 905.

Strapped down to the rails where the rear seats normally are mounted was a temporary fuel tank made of aluminum sheet metal, about 3 feet square, full of 124 US gallons of 100LL. Added to the 84 useable gallons in the wing tanks, with the 12 gallons per hour consumption rate checked on the transcontinental shakedown flight, this bird had the capacity to fly for 14 hours with the required transatlantic 3-hour reserve. Thirteen hours at 130 knots would reach the closest Irish airport, Shannon. As luck would have it, that steady westerly wind shaved two hours off the 14 hours it would have taken to make Dublin, the final destination for this delivery flight, so there was still 5 hours of fuel left at landing.

Managing the fuel flow from ferry tanks is often problematic, and this one was no exception. It was installed by a reputable company, Telford Aviation at Bangor Airport in Maine. However, during the test flight from Bangor to St. John’s Newfoundland, when fuel was supposed to be flowing exclusively from the ferry tank to the engine, the digital fuel flow gauge began to show high and erratic rates of flow. A check of the engine analyzer showed no evidence that the cylinders were being flooded, so I concluded that the fuel flow gauge had lost its senses. A little further on, the gauge on the left wing tank began to show a pronounced decline. This was odd, since the main fuel selector valve, which controls flow from both wing tanks, was in the “off” position. There was also a faintly visible vapor plume streaming back from the vent under the left wing tank that I happened to notice while checking for signs of ice during a spell in the clouds.

When the engine suddenly began to starve, long before the 60 gallons in the ferry tank should have been depleted, the digital fuel flow gauge was exonerated. Switching the main fuel selector to “BOTH” restored the fuel flow. In a conversation later that evening with Telford, a plausible explanation emerged. The main fuel selector valve leaked in the “OFF” position, permitting the fuel from the ferry tank to flow both ways–toward the carburetor, and up into the wing tanks. As pressure built up in the wing tanks, the left tank began to siphon through the vent. The solution Telford proposed was to draw the wing tanks down substantially before starting the ferry pump, and then monitoring the wing tank gauges to be sure not to overfill them.

During the crossing, I flew for 3 hours on the wing tanks, then switched to the ferry tank with the pump off, and the main fuel selector valve off. Flow stayed at the normal 12 gallons per hour, and both wing tanks continued to show a decline, particularly the left, confirming the leak theory. When I was close enough to Ireland to risk more tampering, I opened the main selector and switched on the ferry pump. Gradually, the wing tank gauges showed a rise, which was encouraging, since I had more confidence in the gravity-feed from the wing tanks than in the pumps required to get the last one-third of the fuel out of the ferry tank.

I imagine pilots who do this all the time become quite expert at sorting out ferry tank problems. One little gadget I had along with me on the recommendation of the old hand who installed the tank was a manual pump to be used to build up pressure over the fuel in the ferry tank. This would help push fuel to the engine in case both ferry pumps failed. The thought of electric failure halting fuel flow was not comforting, but at least I had a tool to deal with it, however awkwardly.

With my three tastless Newfoundland-made sandwiches and my 1.5 liter water bottle, I was prone to Lindberg fantasies, especially since the flight was only a few days after the May 27 anniversary of his daring flight. A message from my one and only GPS about half way across brought Charles to mind when it said “no GPS position.” I kept on my heading, thinking in the worst case, I would eventually sight some part of Ireland, meanwhile marveling at Lindbergh’s chutzpa. After all, he navigated all the way from Long Island, New York, to Paris, with only a wet compass. About 20 minutes later, the GPS found its bearing again, and led me unerringly to Shannon, and then on to Dublin.

I wonder how Charles dealt with the awkward problem of urination on his 33 hour flight? I had along a collapsible plastic canister with some sort of chemical powder in it that was supposed to turn liquids into gel. Didn’t work. Furthermore, there was no reliable way to seal the thing up when it was filled. I managed to dig a plastic bag out that I had stored my wet bathing suit in, and wrapped the canister in that, tying it up tight. That did the trick. But getting your immersion suit and your jeans down and then back up while jammed between the yoke and the ferry tank behind you is no easy task. There were several uncommanded descents when my knees accidentally hit the yoke. Notice I made no mention of underpants. I left them out of my wardrobe to make this operation at least a little less awkward. I wonder how they manage in space?

There are pilots out there who deliver airplanes year round, often two or three times a month, to far off destinations in all kinds of weather. Among them are Hardy Kalitski, Don Ratliff, Ken Dawson, Denny Craig, Margaret Waltz, and the late John Carlson. The trip I describe here would be routine to any of them. If any of them read this, I hope they find I have captured some of the reality of their business.

Links:

Video of sunrise over Atlantic

Route map

An account by Sape Mullender of another transatlantic ferry flight