Stevenson 750 Report
Above is a photo of Doug Stevenson's Corvair powered CH 750. Doug built and flew this aircraft in Southern California, out of
the French Valley airport. The engine was completed and test run at Corvair College #18, after which I
delivered it to Doug and the two of us did the initial steps of the installation on his firewall. Doug went on to complete
the aircraft and the installation in the following months. The aircraft completed three flights. The first one was terminated
due to an electrical short underneath the panel. Doug turned 180° back and landed at the airport. The problem was traced to
part of the voltage regulation. This was only discovered in flight because on the ground the engine had the ability to overwhelm the brakes
at any RPM above 2,000. When the airplane was on climb out it could turn more than 2,800 RPM, and the charging system provided a voltage
that had not been seen on the ground. In our Flight Operations Manual we insist that everyone tie their aircraft
down and run it
at full takeoff power for 2 minutes prior to ever flying. After correcting the electrical issue, Doug did just this and verified that
the engine ran very well at full power.
The second flight was about an hour and was completely uneventful. All of the
operation of the engine and the installation worked as intended. He specifically commented on how smooth the operation of the
engine was and that the aircraft had outstanding power. It had no cooling issues and this was demonstrated by sustained climbs
of thousands of feet at wide open throttle.
On the third flight of the aircraft, all systems appeared to be normal and demonstrated regular operation. Before the flight, Doug
fueled the airplane and checked the quantity with a stick he had made for the purpose. He had been unable to get the electronic
instrumentation calibrated to read the actual level in the tanks. All three flights of the aircraft were done without any way of
knowing how much fuel remained in the aircraft once it was in flight. After flying for approximately 45 minutes, Doug decided to
return to the airport area. He descended from 5,000 feet at a reduced power setting for several minutes. He recalled that this
RPM was approximately 2,400. However, the descent rate meant that the aircraft was producing very little power at this RPM, had a
high manifold pressure drop in the carburetor, and should have had the carburetor heat applied. Doug reported that he did not
use the carburetor heat. After several minutes of descent, the engine stumbled and quit. He attempted to restart several times,
on both ignitions.
Each time the engine kicked over slightly, but would not continue to run.
He found himself outside of gliding distance to the airport, and elected to land on a road.
Avoiding power lines, he landed the aircraft and it went over its nose and ended up on its back. He had executed the glide
at the best glide speed. However, he had difficulty flaring the aircraft on touchdown. The plane was within the published weight
and balance envelope at the time of landing.
The difficulty was caused by a characteristic of power off approaches flown in Zenith
701, 750 and 801 aircraft. The flare must be started with reserve energy generated by gliding well above the best glide speed
before flaring. Pilots who elect to seek transition training in Zenith aircraft are made aware of this characteristic. It is not a
difficult concept for pilots who normally make power off approaches in general aviation aircraft. However, many pilots are unfortunately
trained to only make power on approaches; in an engine out situation, in any general aviation aircraft, these pilots are at a severe
handicap in understanding the handling of the aircraft on landing. An in-person study of Doug's airframe and listening to his story
of the last minutes of flight demonstrate that the majority of damage on the airplane was caused by its inability to pitch up and
check the rate of descent at the last moment.
I spoke with Doug after the accident and we made a plan to get together and conduct a more in depth investigation. By this
time, the NTSB already issued a preliminary report that simply indicated that the engine failed to make power. In spite of the
disclaimer that is printed on every one of these reports that states in part “this report may contain errors,” this led a number of
Internet speculators to suggest that Doug had experienced a mechanical engine failure. From speaking with him in person, I knew that
this was highly unlikely; besides, I personally test ran the engine for about an hour before delivering it to him. We have more
than 40 Zenith aircraft flying an almost identical engine installation, so I suspected that the explanation was outside the basic
engine and installation. I got on a commercial airliner and flew out from Florida to California,
had a new propeller shipped in, and went to Doug's with a camera, a notebook
and an open mind. After a little prep work in the hangar, we filmed this video of the engine running perfectly:
When watching the video, you can see that the engine runs, idles, starts and stops normally. I took time to film the Dynon engine
instrumentation so builders could see the engine's normal operating parameters present. About two thirds of the way through the video
you can see Doug
switch back and forth between both the A and B ignition. It runs the same on each one. The only alteration to the engine's systems was
the installation of the new propeller and connecting the fuel line to the 1-gallon clear plastic container sitting atop the cabin.
The blue color of the fuel identifies it as 100 low lead. This demonstration effectively eliminated any discussion that there was a
mechanical problem inside the engine, or that any of the systems of the engine caused it to cease running.
Doug had a very strong suspicion that he had run out of fuel. He is very experienced with engines
and is a pilot of considerable experience. His description of how the engine almost restarted was consistent with an engine that
was running out of fuel. He contends that he took off with enough fuel on board for the duration of the flight, but his inoperative
fuel gauges would not have told him if he was losing fuel through a gascolator leak or some other source.
Doug stated that he was pretty sure he had 13 gallons of fuel in the aircraft before the last flight. The normal
fuel consumption of a Corvair engine in cruise flight is about 5 gallons per hour to produce 75 hp.
This is consistent with most other 100 hp aircraft powerplants. However, Doug's engine is a high-performance version of the Corvair
engine, with substantially increased displacement. Additionally,
the 750 is a high drag aircraft that can climb at a very impressive angle given a substantial amount of power. Doug stated that
much of his flight was climbing under these conditions. Under such conditions, a large displacement Corvair like Doug's could
conceivably consume 10 gallons an hour. This could even be increased by an overly rich setting in the carburetor. It would be
difficult to establish the fuel consumption for this specific aircraft from its single previous flight of duration.
If the consumption was substantially higher than Doug was suspecting, the inoperative fuel gauges would not have warned him.
The fact that the engine stumbled after several minutes of low power operation leads me to wonder about carburetor ice.
The conditions for it were possible at the time. With characteristic frankness, Doug simply stated that he did not utilize
carburetor heat because he felt that the RPM of the engine was still high enough that it was not required. In our
Conversion Manual and the Flight Operations Manual we explain that the potential
for carburetor ice is reflected in the drop in manifold pressure, not any specific RPM. The NTSB data
includes weather information at the time, and it is fairly easy to extrapolate what the temperature at Doug's altitude was.
Again, few people have studied carburetor ice as closely as I have, and in this specific case, I maintain that
it was potentially an issue.
No aviation professional ever states that an investigation determines the ”cause” of an accident.
Instead the goal is to eliminate things that did not cause it from consideration, and examine the remaining possibilities to determine
a “probable cause.” This isn't semantics; the difference in wording illustrates a completely different philosophical perspective
as to how an investigation is conducted.
It is ironic but true that preliminary reports are issued to prevent public speculation, but human nature being what it is,
the data presented is often distorted by people to support their own pet theory or conclusion. We have all seen major accidents
where the TV reporter shows up, often while firemen are still on hand, and asked the first aviation official they find
“what caused this?”
This is why any real aviator gets his information from actual first-hand investigations and reports
rather than TV news and Internet speculators.
Above is the right side front view of the aircraft. Notice how severely distorted the first bay of the fuselage is. Both the
nose and main gear and their mounting points were destroyed in the accident. To support the aircraft for the test run, we blocked
it up on pallets and held it down with
the orange ratcheting tiedown straps seen in the photos. No work whatsoever was done to the engine or its installation before
the test run, with the sole exceptions of bolting on a new propeller
and putting the 1-gallon gas can atop the cabin in place of the wing mounted tanks.
The left side view shows the same damage to the fuselage. Despite this damage,
the only injury that Doug suffered in the accident was a strained shoulder from where
his shoulder harness restricted his forward movement.
Doug's aircraft was equipped with a 68 inch, two-blade, ground adjustable Warp Drive prop.
The blades of this propeller are solid carbon fiber. Doug stated that the engine was not turning and the blades were horizontal at
touchdown. Nonetheless, each of the blades were broken off. Even though his Corvair engine is equipped with a very stout fifth bearing,
and I could detect no run out in the crank after the test run, we're going to still pull the engine down and inspect
Doug's gascolator is the stock one supplied by Zenith with the airframe kit. I checked it,
and found it free of debris of any type. The rubber hoses and hose clamps are supplied
with the kit. Obviously they work. This said, I am personally in favor of utilizing braided steel
hose with AN fittings throughout the aircraft. It is more expensive and slightly heavier. Our
Installation Manual outlines the sources of braided hose and how these hoses are built and installed. Additionally,
I have always been in favor of using the highest quality gascolator
available, which is made by Andair. Like Doug, builders are free to use whatever they like.
Our example installations, and instructions in the Installation Manual, are designed to illustrate
the best way we know of to install the engine. I am not arguing that more economical systems can’t be made to work;
obviously they can. The focal point of my work is aimed at showing people the best way to do things, not just simply what will
work, and never is aimed at
demonstrating what you could get away with.
Doug's aircraft ended up on its back by rolling over the nose. This is the spinner and backing plate from his engine. Our engine
installations utilize a number of proven off-the-shelf components from other manufacturers. These are all specified in our
Zenith Installation Manual. The spinner and backing plate are available from Van’s Aircraft and are common
to the majority of RV series aircraft.
This is the Nosebowl from Doug's aircraft. This is a component that we make
and sell. It comes with its surface finish complete ready to paint. The damage shown here is from the accident. The Nosebowl
air inlets come from us un-trimmed. The slower the airplane,
the larger the holes need to be for air. This Nosebowl is utilized both on Zenith 601s and 750s. Typically, they are trimmed to
4.25” diameter for a 601 or 650, and 4.5” diameter for a 750. As shown here, the holes are only 3.75” or so. Additionally, we
put air inlet rings behind them to blend the flow of air entering the cowling. Despite seriously undersizing the holes, and having
an engine that produces 120 hp, Doug's engine ran cool. This illustrates two things: First, the Corvair's excellent cooling
potential, and second, that our designs and installations are never engineered to just barely do the job.
At all times, I remember that our primary customer is neither a test pilot nor an aeronautical engineer. They are homebuilders,
and it behooves us to provide them with things that have a wide margin of safety, and great reliability. This goes all the way
down to the size of the inlet hole on a Fiberglas Nosebowl.
This is a picture of the aft part of the doorframe on Doug's aircraft. The stain is from 100 low lead fuel leaking on the door.
Initially Doug thought that this might have indicated that the drain valves in the tanks located above the doors may have leaked
in-flight. After careful consideration, and listening to his complete experience, I suspect a different cause. The airframe ended up
resting on its back. The Zenith doors are hinged at the top.
Each of the doors ended up resting on the underside of the wing because the gas strut mechanisms that held them in place were broken
during the accident. This would bring this component in contact right where the stain is with the underside of the wing drain.
Doug stated that when the aircraft was rolled over, the doors momentarily stuck in the up position. I suspect that at this time a
small amount of residual fuel in the tanks drained through onto the doors.
This is a look at the stock Zenith fuel selector and hose arrangement installed in the aircraft by the pilot’s left ankle. Again,
I would greatly prefer to see braided hose and AN fittings on Corvair powered Zeniths. For a number of good reasons, the 750’s fuel
system continuously drains both tanks at the same time. This valve simply works as an on/off, not a tank selector. Anyone with
professional training in accident investigation will tell you that there is a long list of aircraft that have been downed by human
error operating this system of tank selection. This interconnected system gave Doug a hard time calibrating the quantities in each
individual tank and making his electronic instrumentation read this correctly. This led to him flying with inoperative fuel gauges.
Before the flight, he did test the depth of fuel in the tank with a calibrated stick that he made. But this would not allow him to
notice any fuel leakage in flight, nor unexpectedly high fuel consumption from operating a very powerful engine on a high drag
airframe. Any accident is a chain of events, and you only need to undo one link to prevent it. Having operative fuel gauges would
be the easiest link to undo in this chain of events.
Here's a look at the rudder pedal area in the aircraft. The floor is dented upwards more than 10 inches. This is predominantly
caused by the nose gear being folded underneath the aircraft. Despite the impact destroying a lot of the airframe, the cabin area
remains intact. No aircraft can be designed to protect its occupants from all impacts, but some are far better than others at this
task. In my two decades in aviation, I've had the unfortunate education of having seen an awful lot of crashed aircraft. My degree
in aeronautics from Embry Riddle is from the same program that generated many of the best-known accident investigators in the
country. I'm in a pretty good position to say that this aircraft has well above average occupant protection in an accident.
The specific detail I would like to point out here is the location of the Odyssey battery. In Zeniths, we've traditionally
recommended putting the battery behind the pilot seat. If you're going to put it on the firewall, like this, I would recommend
taking very serious steps to make sure that the positive battery terminal does not contact the metal structure of the aircraft when
it becomes distorted from an impact. Doug's installation ended up with only an eighth inch
between the positive battery terminal and the distorted firewall. The only insulation he had on the terminal was a few wraps
of electrical tape. Had this contacted, it would have discharged the entire energy of the battery, even if the master switch was
Had any fuel been present, the likelihood of a fire would have been incredibly high.
Here's a pilot side view of Doug’s instrument panel; the primary instrument is a Dynon.
The white space on the right is a holder for an iPad which provided the moving map display. This is another good view of how distorted
the floor is. Distortion is not bad, it is energy absorption visually displayed. What is unacceptable is a structure breaking up.
Several things can be seen in this photo. First that we ran the engine without making any improvements or attachments to it other
than hooking the fuel line up to a gas can and putting on a new propeller. All the engine controls, including the throttle on the
extreme left side, were left alone. I literally did not touch anything else on the engine, make any adjustment of any kind, charge
the battery, or touch a wrench or screwdriver on anything.
I simply turned the ignition switch on, primed the engine by moving the carburetor throttle twice, pressed the starter button and
the engine ran after 2 seconds of cranking. This effectively demonstrates that there was nothing wrong with the engine internally,
nor any of its connections to the aircraft. We did two engine runs about 5 minutes long each at various power settings. The engine
not only made full power but also idled, and restarted normally. It ran equally on both ignitions. All of this can be seen by
studying the video at the http://www.youtube.com/watch?v=Q2zYNP-IBeE YouTube link.
A very important consideration looking at this photo: This is visual evidence of why you never want to use a hard aluminum line on the
floorboard of an aircraft. The rubber hose that came with the kit was utilized to run from the fuel valve to the firewall pass-through.
In several places it was pinched very tight between the distorted sheet-metal. Had it been a standard piece of 5052 aluminum tubing,
I am sure it would have ruptured. If there had been fuel in the tanks in an accident where the aircraft remained upright, this leak
would have poured fuel into the damaged cockpit. It is far more desirable to have a flexible line, preferably a braided steel one, in
this location where a distorted floor from an accident would rupture a rigid line.
This is a pilot side rearview of the engine installation. The vast majority of the details of Doug's aircraft were done in
accordance with our detailed 125-page Installation Manual,
written specifically to cover the installation of this engine on Zenith aircraft. In the foreground of the photo is the
large aircraft oil cooler that we utilize on 750 and 701 installations. This complements the Corvair's excellent cylinder
head cooling. The baffling shown is the kit that that mates with our cowling and the Corvair engine. It is a traditional system,
just like the ones seen on certified aircraft. It utilizes the entire area above the engine as a plenum for the cooling system.
This exact design is now flying on more than 40 Zeniths.
Here's a view of the underside of the panel of Doug's aircraft. The two blue cylindrical items are the redundant Bosch ignition
coils. Although the wiring looks a little disorganized, it is made of the correct material. Most homebuilts have panel wiring that
is organized about this level. I could make a serious case for aiming higher than this in terms of organization and presentation.
This said, the wiring was not a source of trouble on the day of Doug's accident.
Above is the nose gear strut bent backwards underneath the aircraft. The Exhaust Pipes are on the
outboard sides of the photograph. They were undamaged, and only a single tube in the Motor Mount was
bent despite having the entire weight of the aircraft on it
during the slide to a halt and the roll over onto its back. At the very top of the photo is the air inlet on the underside of the
carburetor. In our cowling kits, the air filter is integrated into the bottom of the cowling, which had been removed before we
chocked up the aircraft and put it on pallets.
Here is an underside view of the engine and carburetor. The gold item is our Billet Oil Pan fitted to
the underside of the Corvair engine. The black item is a Marvel Schebler MA3-SPA carburetor. This is a certified carburetor that
originally was utilized on Cessna 150s. Doug's carburetor was overhauled and yellow tagged by a certified fuel system repair
station before he installed it. After the test run, I pulled the drain plug in the front of the carburetor and carefully
checked for any type of debris or obstruction. There was none visible. The blue fitting on the side of the carburetor is
an AN fitting on the end of a braided hose. A number of our builders utilize the braided hoses we suggest ahead of the firewall,
but don't follow through with braided hoses inside the aircraft. As I said before, I would greatly prefer if Corvair powered
Zeniths all had 100% braided hoses throughout.
Braided hoses made for these fittings do not care about ethanol in fuel. This cannot be said for all forms of rubber hose.
In California, ethanol is allowed to be put in fuel at levels as high as 15%. The Corvair engine itself does not care about this.
I recommend operating your entire test runs on 100 low lead in order to remove one more variable from all of your original flight test work.
The accident investigators who looked at Doug's aircraft initially focused on the aspect that he had previously run the aircraft
on auto fuel and had a small quantity of it in the tanks at the start of his last flight.
They did not have available to them the option of running the engine like we later did to eliminate many of the things that
might have at first glance been thought to be an issue.
The orange scat tube on the left side of the photo leads directly to the carburetor heat muff. The other end of this hose is
connected to the air cleaner assembly in the lower cowling. I have long counseled Corvair operators, and light aircraft operators
in general, that they should make a habit of utilizing carburetor heat at any substantial power reduction.
I have written this in countless places, to the point where many people new to aviation falsely conclude that the Corvair
has a special need for carburetor heat. The carburetor seen here and its type of installation is identical to a small
Continental engine. I do not have any understanding of why pilots are ever reluctant to utilize carburetor heat as an anti-ice
measure when operating light aircraft. The majority of pilots rationalize away its use and get away with this. This shows you
how rarely you actually need it, but when you do, people who operate with this mentality will find themselves in a bind. It is
far better to utilize it as it was intended, and thus it will already be on when you need it. On the day of Doug's accident he
descended for 3 minutes at a reduced power setting without ever employing the carburetor heat. After this 3 minutes came the
first sign of the engine stumbling.
Rank amateurs when reading accident reports like to jump to conclusions or try and discover “the cause.” This is neither how nor
why accident investigations are conducted.
At the very best, an investigation will issue a “probable cause,” which is an entirely different mentality than stating that you
know something from an absolute perspective. Much of accident investigation is about eliminating possible causes by verifying
the operation of systems after the fact. The primary example here is the running of the engine with a new propeller and a gallon
of fuel, which effectively eliminates any discussion that would blame the engine internally or its installation subsystems.
After careful consideration, my feeling is that the aircraft simply exhausted the fuel supply. There is a serious possibility
that it may have experienced carburetor ice, as conducive conditions existed and carburetor heat was not utilized. When reading
any of these notes, come back to this paragraph and understand the difference between stating a cause and conducting a real
Above is the fuel gas cap on one of the wing mounted tanks. These gas caps are internally vented. The light blue stain on
the gas cap is from fuel leaking out when the tank was upside down because the airplane was laying on its back. The initial NTSB
report notes that fuel was seen leaking from this spot on the inverted aircraft. However, this does not necessarily indicate
having a significant quantity of fuel in the tanks. Doug personally spoke to the firemen at the scene. This is the person who
stated that fuel was leaking from
the tanks. But the comment was taken out of context. The firemen who said it pointed out that fuel was intermittently dripping
from the caps. There was no hazmat cleanup done,
and the complete quantity that leaked might have been tiny.
Above, the passenger side rear spar attach fitting and fuel line connection. This is the standard rubber hose that comes
with the kit. Although I did not see any direct evidence of leakage here, the edges of the hole were not smooth, and builders
need to aim a little higher on any spot where any rubber hose lays on aluminum.
Doug and I posed for this photo at the conclusion of the testing. If anyone reading this took away only one thing from
this discussion it should be this: Doug is a painfully honest man, which is not a common commodity these days. Very little of
the things that we know could have been learned if Doug decided that he did not wish to participate. He openly shared all the
information and invited me into the hangar, and gave his frank appraisal of all the events. He did so out of a genuine belief
that no one should ignore the lessons that are available here. I have been working with aviators for more than 20 years, and
this is a nice ideal that most people would swear allegiance to, but if it came to their own accident, they might be very
reluctant to admit anything that would allow criticism by a Monday morning quarterback. This wasn't even a consideration for Doug.
"Real freedom is the sustained act of being an individual." WW - 2009
Now At The Hangar