EAA - Experimental Aircraft Association  

Infinite Menus, Copyright 2006, OpenCube Inc. All Rights Reserved.


Tools:   Bookmark and Share Font Size: default Font Size: medium Font Size: large

EAA Experimenter

[ Home | Subscribe | Issues | Articles | Q&A | How To | Forum Review ]
[ Hints for Homebuilders | Glossary | Polls | Around the Web | Submit an Article]

David Roe’s Hummel Bird

AKA "Diva"

By Patrick Panzera, EAA 555743, ppanzera@eaa.org

David Roe's Hummel Bird
David Roe’s Hummel Bird parked on the ramp at COPPERSTATE 2003.

David Roe has a degree in the arts and was an illustrator for years. After getting “burnt out” on that, he became a jazz and blues trombonist and did a little singing as well. David comes from a flying background. He has memories of helping his father buck rivets. His Dad would bring home old airplanes that were in need of some TLC and David would be enlisted to assist. David always wanted to fly but never had the desire to do anything commercial with his skill. Every time he tried to make a profession out of doing something he enjoyed, it became work, and as he put it, “turned me into a miserable human being.” Flying keeps the smile on his face. David has logged more than 1,500 hours with his single engine land rating.

David finished building his ½ VW-powered Hummel Bird in 1995. It was a straightforward, plans-built, stock Hummel Bird with conventional gear. The plane was flown for about 900 hours, at which time the “stock” ½ VW engine developed a crack in the engine case.

David Roe's 1995 Hummel Bird
David enjoying the fruits of his labor in the (then) newly completed 1995 version of his Hummel Bird. Note the stark contrast in cowl designs between the photo above and the photo at the top of this page. After 900 hours of flying with the ½ VW engine, David and Doug replaced it with the one-of-a-kind Aeromorph. Photo courtesy of David Roe.

David wasn’t enthralled enough to put the effort into making a fix, so he started looking at engine options. He turned to friend and neighbor Doug Reed, who was the builder of the original ½ “VeeDub,” which was a mixture of Doug’s imagination mixed with ideas borrowed from Morry Hummel’s, Bill Spring’s, and Gary McGill’s plans. These three sets of plans were enough to give him a feel for the design, and Doug exceeded many of the design parameters.

Aeromorph 75
The Aeromorph 75 started out life as a pile of aftermarket and otherwise stock VW parts, a conglomeration of motorcycle parts, as well as custom, one-of-a-kind, hand-made components.

Doug and David stayed in close contact after the two-cylinder engine was built. Several days after David told Doug about the crack, he got a call from Doug, who pitched the idea of the new “Aeromorph” engine, a design which had been kicking around in Doug’s head ever since he built his first HAPI VW engine. He was always thinking there was a better way to go than with cutting a VW engine in half. It was probably the looks of the ½ VW engine sticking out in the breeze that inspired Doug to build the compact Aeromorph 75. Starting with standard VW engine, through heroic efforts, Doug narrowed the engine substantially. Using short piston skits, custom-connecting rods, and a radically de-stroked crank, Doug was able to achieve a 21-inch wide, four-cylinder engine for David’s Hummel Bird. (We’ll feature this engine in a future issue of Experimenter. ~Pat)

What’s in a Name?
David prefers not to refer to his plane as strictly a Hummel Bird. When he first built the plane it was good for 120 mph. In its current state of design it’s a 170-mph plane, and he therefore affectionately refers to it as a Hummel Bird Diva. “She’s definitely a girl and she has an attitude,” he said.

The Cooling System
“The cooling was wonderful. I never had any cooling problems with the heads sticking out in the breeze,” David added. “It was wonderful and an easier installation, but Doug sold me on the new design, and truthfully he’s probably wired more for the research and development and had the necessary mindset, more so than I. It was more work than either of us thought. He said I could have it done in six months, but it took over four years.”

Round cowl
The round cowl gives the plane an entirely different look.

The engine installation was initially designed to be fan-cooled. Doug was thinking of a zero-drag or very close to a zero-drag cooling system, which, to make a long story short, didn’t work out. They were back to opening up the air inlet and using traditional baffling and exhaust augmentation for the cooling system. There’s an augmenter channel running underneath the airplane from which, as David described it, “the exhaust is being dumped underneath my butt and it can get pretty warm at times.

“I ended up buying some Fiberfrax® ceramic fiber insulation and some insulating foam, but I tell you, I was squirming in the seat while on the ground, that’s for sure.”

The Airframe
The airframe had to be modified quite a bit in order to take full advantage of the new, narrower engine. Amazingly, the engine was narrowed to 21 inches wide - valve cover to valve cover - which corresponds to the seat-back bulkhead, the widest portion of the airframe. The firewall was then grown concentrically to match, which had to be done to an already finished plane. That’s why the blue stripe on the side of David’s plane seems to just start abruptly, as everything forward of the stripe is new since the plane was originally completed and painted.

“It was a trick within itself getting everything to work,” David said. “We have a false skin, or I should say, double skins from the firewall to the line midway to the canopy.” David believes that this doubling up of the fuselage contributes to the strength of the firewall, and it certainly helps create a cavity for concealing a lot of plumbing and electrical.

Round cowl
Shown here is the beautiful wing root fairing David designed and built for his “Diva.”

The point of all this work was drag reduction, and David did a lot of fiberglass work to meet this goal. The largest parts were the nose bowl for the new cowling and the wing root fairings. When David first built the airplane, it was a pure learning experience since he started out knowing very little. “I still know very little except the dynamics involved in wing root fairings,” he continued. “I initially just put metal there with no thought of what was the correct method. I later realized that to do it right, I really needed to make them from fiberglass in order to get all those compound curves in it.” The root fairings seem to be successful, although he doesn’t have much in the way of flight data to compare the before and after. The airframe has only been flown in its previous and current configurations, and none of the modifications were installed and tested incrementally in order to find out the individual result of each.

Hummel Bird

When asked if he has ever tried any glide tests, David responded, “I did do some glide tests with the other engine because I had several engine failures that put me onto the road, so I’m done with glide tests.”

wing root fairing
Shown here is another view of the wing root fairing, as well as a view of the canopy in the open position.

At the time David built his plane, the canopy was the only prefabricated part available from Hummel. The airplane has a very clean tail fairing which also houses the tail position marker light and strobe. David made the assembly by hand, which is part of the aerodynamic cleanup program.

Safety Feature
Another attention-getting feature of David’s plane is the Schweitzer-style tow hook. For a moment it catches the viewer off guard, since most will think that David is towing gliders or banners. Then David reminds that his plane has no electric starter, and that this hook allows him to secure the tail for hand-propping. Once started, David can get in the cockpit and release the tail using the tow hook release.

Hummel Bird

In the early stages of operating the first engine, learning how to start it before finding the magic recipe where it just pops, he wasn’t quite there yet and had a little “experience.” The plane is propped by standing in front of the wing but in back of the propeller. One day, David was trying to hot-start the engine. He thought that he had flooded the engine, so in order to clear it he went to the cockpit and applied full throttle. He then went around in front of the engine to pull it through a few blades. “I could have sworn that I turned the master off,” David said. “I could have sworn I had everything off...so standing in front of the engine, with my left arm I made one flip of the prop, it instantly fired up and went to full throttle rpm, just inches from my abdomen, but no further, as the tail was tied down. So in my opinion, the tow hook paid for itself. I mean, it saved my life.”
The Diva version of this plane now has an electrical system, unlike the former version. The plane now has 20 amps available, and David figured he’d take advantage of it. “I really like the late afternoon and evening flying, so I installed navigation lights and strobes. I tried to light it up like a Christmas tree since the thing is real hard to see in the air. There have been several times where I was flying formation, the lead plane, and the guy off my wing looks down to check a map, looks up, keys the mic, and asks, where did I go?”

Another serious mod David performed on his plane was to increase the wing fuel tank volume by removing a rib, which increased his fuel capacity from seven gallons to 17. “I have 5 ½ gallons in each wing plus six-gallon capacity in the header tank, which gives a total of 17. But to do that, we had to take out a rib on the innermost portion of the wing panel. We then replaced the skin in that area with 0.040, a thicker skin. Oddly enough, the torsional rigidity of the wing panel is now greater than it was before, but it didn’t contribute much to the fuel tank cells. While I had the wing opened to fabricate the fuel tanks, I lined the underside of my top skins [which are only 0.016] with ¾-inch Styrofoam board. I used a spray glue that didn’t attack the foam. It weighs nothing, and it made an immense improvement to the torsional rigidity of this wing.”

Hummel bird

The Cockpit
While the plane was down for the engine change, David decided that it was in need of new engine instrumentation, and with the space limitations, there was only one way to go: a combined engine monitoring system. David chose to install the Grand Rapids engine information system, which has worked very well for him. He’s running exhaust gas temperature and cylinder head temperature on all four cylinders. The new instrument panel is fiberglass.

“What I did in order to get room between the header tank and the panel was to move the panel aft about three inches. This also allowed for the installation of a transponder and radio. I installed the Microair M760 VHF transceiver and matching T2000 transponder. I now have a transponder, which I never had before, and it’s kind of a kick. I’ve taken my little plane into Albuquerque International, which has Class C airspace. What a thrill! Now I can also use flight following and other services which weren’t available to me before.”

In addition to these cockpit mods and fabricating his own throttle quadrant, David has installed interior lighting to illuminate those items that aren’t self-illuminating. The old airspeed indicator needed to be upgraded, because with the new, more powerful engine and the aerodynamic work, the old airspeed indicator was getting “pegged” on a regular basis. David was quite happy to make this change.

Hummel bird

The Wheel Pants
Another of the many modifications that came about on the airframe with the new engine installation was lengthening the landing gear. In so doing, the wheels are naturally moved forward a bit, which actually helps the ground handling. The gear was lengthened to allow for a larger-diameter propeller while retaining adequate ground clearance. A longer propeller was selected to absorb some of the additional power that the new engine produces. Another reason for lengthening the gear is because David was having difficulty with three-point landings. “I could never use full aft stick without slamming the tail into the ground,” David said. There are no flaps on this plane. According to David, you just side slip and cross your fingers.

With the new, longer gear, some fiberglass was needed to dress everything up. Although David was recycling the pants from the previous version of this plane, the gear fairings needed to be addressed. This allowed him to incorporate “landing lights” in the leading edge of the fairings. Although David doesn’t do much night flying and doesn’t need them much for that, as stated previously, he likes to be seen. Observers have reported that they work great at dusk and make the little Diva look like a big airplane.

Hummel bird wheel pants

The wheel pants were given to David and probably came off a Long-EZ. They had to be radically modified to fit his application as there were massive holes on the tops. The skirts and the blisters are David’s work. “I really warmed up to using clay a lot wherever I can for fabricating intricate forms. They were great wheel pants to start with because they were just wide enough to accommodate the scissors with this landing gear.” The landing gear was built with weldments of a somewhat crude telescoping tubular design with an internal die spring and scissors, which made work easy when lengthening it. All David had to do was use a longer extension-limiting bolt.
Weight and Balance and Mass Balance
David and Doug initially thought they were going to have to install the engine as close to the firewall as possible. As it turned out, they knew this airplane was going to be operated above the listed redline (Vne). David remembers asking Morry Hummel what the red line on the airframe was, and he was told 140 mph without balanced surfaces. David has balanced his surfaces 100 percent, but Hummel really doesn’t have a Vne for this condition. Fortunately, all the extra weight added to the control surfaces on the tail was adequate to offset the additional weight of the engine, and they were able to set the prop within one-half inch of the specified station. David said, “That was just dumb luck. It’s not because we’re brainy or anything.”

Hummel bird

The new fiberglass fairing David built for the top of his rudder allowed him to put the balance weight far ahead of the hinge. The counterweights for the elevator are inside the fuselage tail cone, and the aileron weights are visible from under the wing. Since the aileron weights are aft of the center of gravity, this weight also helped get the plane to balance out.

Hummel bird
Aileron mass balancing is done by way of a tube jetting forward of the hinge line.

NACA inlet
David’s plane is the picture of “attention to detail.” Little things like the quality of the installation of the NACA inlet bear this out.

Hummel bird

Test Patch
There’s a little black triangle painted on the upper surface of the left wing, about mid-span, near the leading edge. When asked about the significance of this, David replied, “I have trouble coming up with a really great smart-ass answer for that.”

The short story is that it involves trying to do something by himself that really required two people. “I was trying to replace a rivet in that area, and the bucking bar slipped and produced a big old wart.” David is using his mistake as a learning experience, and the area has been filled with Poly-Fiber’s SuperFill®, a lightweight epoxy filler. “I wanted to try it, to see how it holds up on aluminum.” To make the test somewhat “worst case,” David painted the test area black, a color that’s usually quite hard on epoxies.

So next time you see David at a fly-in, be sure to ask him about the significance of the little black triangle. Maybe by then he will have come up with a snappy answer.

Some of you may have winced when you read that David routinely flies at 175 mph indicated, at altitude. Some simple number-crunching puts his true airspeed near 200 mph at 8500 feet on a standard day. And please don’t take any of this to be an endorsement for operating beyond design limits; David ended up paying his dues.

Can flutter be expected? The control surfaces have been fully balanced, but the structure of this plane is nearing Part 103 type construction. So it would seem logical that something had to give, and it did, during the first opportunity that David had to fly his new Hummel Bird modifications, wide open throttle (WOT), at an altitude less than 7,500 feet msl.

Based from Santa Fe, New Mexico, where the field elevation is 6,344 feet and altitude being his friend, preferring to have a couple thousand feet between him and the terra firma, David routinely flew at 8,500 feet and above.

“I had never encountered any sign of flutter before at the higher altitudes that I am accustomed to flying in, but down in the thicker air at 2,500 feet, things were different. I was returning to the Casa Grande Airport with my friend Jeff Scott (in his O-200 powered KR-2S) as a flight of two. We had just finished a wing-to-wing speed run of the two planes at WOT, and were coming back in for a low pass or two, ‘just ‘cause.’ It was on my downwind leg for the airport that things got a little weird.”

Having preconceived notions of what flutter might be like, David always thought of flutter as being a violent “flapping,” with the yank trying to be ripped from his hand. But his experience was something more like an unusual engine vibration being transmitted through the stick, than a violent yank.

Looking out at the left aileron, David assumed that his vision was blurred, as he couldn’t bring the fuzzy edges of the aileron into focus. Realizing what was really happening, he throttled back the engine and slowed the plane down to an airspeed that brought everything back into a balance condition. On the ground, he and Jeff looked a little closer at the ailerons and found some oil-canning and some less-than-stellar movement with the control balance arm, but there was little if any slop in the control linkages themselves.

“My flutter was a result of trying to make a 100 mph airplane go 175 mph, without properly ‘rigidifying’ the control surfaces.” During the airframe modifications for the new engine, David certainly had the forethought to balance the controls to 100 percent as previously mentioned. As an aside, he also tested the ailerons for torsional rigidity by having a friend hold one end while he applied torque to the other end. It seemed fine at the time, but obviously it wasn’t.

The original Hummel Bird ailerons have no internal ribs or stiffeners of any sort. They are just a sheet of 2024T3 wrapped into a triangular wedge and are six feet long. The Vne for the Hummel Bird was published as 140 mph “without balanced controls.” But balance is only part of the equation.
“I flew the airplane home taking note to stay below the 160 mph indicated airspeed where the gremlin wanted to take over. It’s interesting to note that on my departure out of Scottsdale [Arizona] at about 3,000 feet msl, I got distracted with traffic and inadvertently reached the magic number, and felt the buzzing starting to come on. Down the road, on my descent into Santa Fe at about 10,000 feet msl, I managed to blast past the 160 mark again with no sign of flutter.”

Once home, David made arrangements to fly the plane over to Los Alamos where he could have his EAA chapter buddies take a look at the situation. The solution was unanimous; the ailerons needed to be rigidified. David was given extra instructions to measure the amount of torsional stiffness before the corrective surgery. “By holding one end stationary and applying a certain amount of force on the opposite end, I took a measurement of that deflection. If after the fix, the measurement I had taken before the fix could be reduced to one-fourth of that, I should be way out of the frequency range that was causing me problems.”

The prescribed fix for the ailerons was pretty straightforward; use two-part expandable urethane foam to fill the triangle. With the aileron removed from the airframe, David removed the outer end cap of the aileron and poured it in. “If you’ve never used this stuff before, it’s really kind of neat. The expansion rate accelerates with temperature, and unlike the spray foam in the can, it cures pretty quickly. Multiple pours are a breeze. Just wait 10 or 15 minutes at 70 degrees, and you can continue your pour. Be careful when handling, though, as it is still rather soft and can be crushed if proper care isn’t used. By an hour or so, it gets pretty firm. The end result was a solid, firm, very rigid, and light aileron. The control, of course, must be rebalanced.”

It’s also worth noting here that care should be taken to restrict the tendency for the skins to bulge due to the expansion of the foam. A crude jig made from long, straight edges mounted on both sides of the control surface with wide masking or packing tape wrapped around the entire package worked well for David.

The results of his efforts are limited, but they seem to do the job. David didn’t obtain the one-fourth measurement of the initial torsional benchmark that was suggested; he only got about one-third. “I see no sign of twisting at all in the aileron and have total faith in its integrity, so I suspect that my methods were at fault.” On the initial test flight David managed to bring the indicated airspeed up to about 185 mph at 8,500 feet msl with no sign of flutter. That was no real surprise, though, because there wasn’t a problem at altitude. The following week David took the plane back over to Arizona for a family visit. At an altitude of about 3,000 feet msl, David accelerated to 180 mph indicated, with no sign of the gremlin. “This is 20 mph over the 160 mph that was the beginning of my problem area, at that altitude.”

Things seemed fine for quite some time, but the demon raised its ugly head again while flying in Ohio. “The idiot (yours truly) that built the counterweight arms for the static balance of the ailerons didn’t get them symmetrical.” The left arm (the one in question) was hanging down about an inch lower than the right aileron arm, which was aimed more directly into the relative wind during straight and level flight. With the left arm hanging down in the relative wind, it would begin to vibrate with increased speed at lower altitudes, exciting the aileron into a flutter condition. Again, at higher altitudes (7-8,000 feet), no problems - not to say that he wasn’t on the verge of having one.
Now that David adjusted and raised the left arm even with the right arm, there have been no further indications of flutter. “It now points into the wind with better penetration of the air and less resistance and susceptibility to begin vibrating. At least, that’s how I see it,” David told us.

I’m sure that this flutter information will get a few people stirred up, and the best place to “vent” might be the discussion group at Oshkosh365. See ya there!




Hummel Bird


Modified Clark Y

Horsepower (estimated)



Wing Span (feet)



Wing Area (square feet)



Fuselage Length

13’ 4”

13’ 4”

Empty Weight

450 lbs

270 lbs

Gross Weight

700 lbs

530 lbs

Vne (mph)



Cruise Speed (mph)



Stall Speed (mph)



Takeoff Distance



T/O over a 50’ obstacle



Landing Roll



Fuel Capacity

17 gal

6 gal

“G” rating

about 4



Copyright © 2014 EAA Advertise With EAA :: About EAA :: History :: Job Openings :: Annual Report :: Contact Us :: Disclaimer/Privacy :: Site Map