The Story Behind the Machine

Bob Waldmiller

Let the Games Begin!
Introducing the "One-Design" and Its Influence on Excalibur
The Wing Is the Thing!
Weight Reduction Time...Again!
90% Done and Only 90% To Go!
Inverted Fuel System, Inverted Oil System, and Flight Controls...

It all started when I received my first aerobatic ride in a Pitts S2B. It was the mid 1980's and I was introduced to aerobatics by Glenn Maben. One day we took a trip to Olean NY where we each received (paid for) a ride (instruction) with Bill Thomas--an exemplary aerobatic instructor. Glenn had already been flying a Decathlon for many years but this was my first adventure into the world of aerobatics. At $120 per hour I wasn't going to wait long to get 100% immersed into this sport and I told Bill to forget the easy stuff like wingovers and stalls and get into the real stuff like rolls, loops, snap rolls, hammerheads, and inverted spins. I know he was watching to see if I was about to turn green but I never did; I was having too much fun. I was hooked on aerobatics; I couldn't get enough of it! Unfortunately, I didn't have an aerobatic aircraft of my own nor could I afford to rent one.

I had been building a KR-1 for several years and it was within a year of flying. Unfortunately, after that aerobatic demo ride in the Pitts, every time I looked at the KR-1 all I saw was an airplane that I knew I'd be dissatisfied with. When I couldn't stand it any longer, I started looking for alternatives. That's when I saw the Corby Starlet. It was in the Wicks catalog and it was a little strange looking but the more I studied it, the more it grew on me. It used a VW engine and, conveniently, I already had a VW engine with an engine mount. The wheels and brakes could come off the KR-1 too; they'll work just fine. I figured all my major expenses were transferrable so there wasn't much of a loss if I destroyed the KR-1. I took the chain saw to it and that was the end of the KR-1. A Corby Starlet was in my future.

I sent away for the info package in 1987 and perused it for about a year. I had just joined the US Air Force and I hadn't exactly settled into my first assignment until early 1988 so I just kept thinking and studying the info pack. The only thing that disturbed me was the 4.5g limit on the airframe. I committed myself to the Corby Starlet anyway and bought the plans. Studying them, I found some of the answers to my questions regarding the stress limits on the Starlet. First off, the wooden wing spar didn't provide much safety margin when one considers the FAR Part 23 standard for aerobatic category airplanes. They need to be designed to a limit load of +6 and -3 g's with a factor of safety of 1.5 giving an ultimate load of +9 and -4.5 g's. I clearly needed to evaluate the design with respect to my own personal requirements and I also knew it would be a big job. Little did I know just how massive the undertaking would be!

Let the Games Begin!

(Click on the thumbnail pictures for larger versions)

The first step, of course, was to define my design requirements. They were as follows:

  • 10 g design limit load at 750 lbs.
  • More power...I planned to upgrade my 55 HP (1680cc) VW to the HAPI 82 HP monster after the first year of flying.
  • Minimize the empty weight of the aircraft.
  • Use more capable airfoils on all flying surfaces.

Before I started building anything, I had to get my design requirements into a set of working drawings. I wanted to retain as much Starlet as I could so I attempted to evolve the design instead of starting with a clean sheet of paper. Right off, I knew the bigger engine would add more weight forward of the firewall so I immediately added an extra 5" behind the pilot's seat. Moving the tail further aft would counteract the engine weight and give me more tail volume at the same time. I wasn't sure if I actually needed more tail volume because the original design had reasonable tail volumes to start with. In addition when I ran the calculations for tail damping factors, I found that Excalibur could be expected to have good spin characteristics. Also, based on Don Wilkinson's comments on the aerobatic capabilities of the Starlet as published in the Oct/Nov 1984 issue of Corby Starlet Newsletter, I was confident that Excalibur would also have good spin characteristics. (Flight testing will tell the real truth, however).

Since aerobatic loads impose severe torsional moments on the wing as well as bending moments, I felt it necessary to change the manner in which these torsional loads are reacted out of the wing and into the fuselage. I designed the wing to use an auxiliary spar just ahead of the ailerons and located a rear spar box to pick up the load behind the pilot's seat. This contrasts significantly with the Starlet's method of reacting the torsional loads via the D-tube leading edge and dumping it into the main landing gear beam. Along with the second spar box, I put in the 2" stretch in the cockpit area and made other adjustments to ensure the CG would end up where I expected it to be. At this point I drew up a new fuselage. The wing would come later. My biggest concern was getting the hardest design work done first before I lost interest in it and a fuselage is definitely more difficult to design than a wing...believe me!

It took lots of brain scratching to find ways to keep the airplane light. One area was in the landing gear. Since I changed the manner in which the wing torsional loads are reacted into the fuselage, the only purpose for the original landing gear carry-thru beam was to bolt on the two-piece steel gear legs. If I didn't have a two-piece gear, I didn't need the landing gear beam. I decided to use a one-piece aluminum gear. The lower longerons were heavily gusseted forward of the main spar box for just that purpose and a "hinging" arrangement was devised that would prevent the gear from twisting the lower longerons into splinters.

One might think that the aluminum landing gear would be a lot lighter than the steel gear. This is a common misconception. Although aluminum is roughly 1/3 the density of steel, you end up using a lot more of it to get the required stiffness (or flexibility in the case of landing gear) while keeping the stress levels down to manageable levels. In order to have a higher working stress limit, I specified 7075-T6 in the manufacture of my gear. 2024-T3 would be heavier since the working stress levels must be lower and 6061 is downright unacceptable. One comment I have received from many people regarding my choice of 7075 is its notch sensitivity and brittleness. This is a valid point. However, if the gear is kept free from nicks, this is not a problem. Besides, steel leaf springs on your car or truck suffer from the exact same problem--which isn't really a problem is it? Overall, I'd say I saved roughly 5 pounds in the landing gear department on Excalibur.

Note: While you're looking at the pictures, you should take note of a few things I did mostly for appearance. First, there are very few sharp corners in the fuselage. The lower corners of the fuselage are all rounded, for example, as are the inside corners of bulkheads and most cutouts. I wanted to give my airplane a level of craftsmanship that could be considered "show quality". Secondly, there is evidence of attention to detail everywhere on the airplane. Aluminum parts are generally polished and/or alodined, and steel parts are or will be painted white. Although I did these things more for the "show quality" aspect of it, I also wanted to force myself to make every attempt to keep the weight out of the airplane. Being a little "nit-picky" allowed me to do just that.

Having effectively eliminated about 5 pounds from the gear legs and mounting arrangement, I looked for other areas to save more weight. The next obvious would be in the wheels and brakes. I already had the Azusa 5" wheels from my KR-1 so I elected to use them. The Lamb tires fit perfectly! The mechanical brakes were retained after I saw the Pulsar arrangement. They use external band linings around the outside of the drum and are cable operated from heel pedals under the rudder pedals. At first I was skeptical about them but after trying out the brakes in the Pulsar, I deemed them adequate. In addition, I've heard few, if any, complaints about the arrangement from Pulsar builders who use them. Both the Pulsar and Excalibur have about the same gross weight and landing speed so without further consideration, I put them in my airplane. Right now I'll justify my decision based on the fact that it's too easy to add weight to the airplane by installing unnecessary items but it's extremely difficult to remove weight. So it's best to start out with the lightest solution available and work up from there. We'll see just how happy I am with them when I start taxiing the airplane around under its own power.

What's next...the tailwheel? Sure thing! A couple of Pitts drivers I know say the Haigh tailwheel is the only way to go. They say it makes driving a Pitts on the ground so easy that you'll almost lose your taildragger competency. Plus it puts less demand on the brakes until you need to stop or turn around in your parking space. I was sold! I ordered a look-alike replica of the Haigh unit from Aircraft Spruce but the workmanship was unacceptable so I sent it back. The replica was supposed to be about 1/3 the price of the original and used aluminum parts to save weight. I ended up designing my own and incorporating a nifty little feature-a cockpit releasable tail tie down ring. I knew I wasn't going to put a starter on the engine so this feature made sense for hand propping the airplane.

Before I started building anything, I did a stress analysis on the fuselage and determined that it was necessary to add a couple of extra diagonal braces in the fuselage trusswork. Mainly in the cockpit area and behind the seat. Due to the fact that I planned on using a one-piece wing, the lower longeron gets cut at the main spar. Some of the load must be carried through the spar by the attach bolts. So be it! The stress analysis showed the structure to be very robust and capable of reacting whatever forces the wing and tail could impose on it. I ended up with a minimum factor of safety of 1.2 in the lower longeron forward of the main spar box with a +10 load and the airplane CG 3" forward of its max forward CG position. This was the worst case load and should never be the case if the airplane is operated within its weight and balance limits.

Note: The factor of safety was computed without taking into consideration the benefit of the 3/32" thick plywood skins in the forward fuselage area! Thus the structure is theoretically safe before it is covered with plywood. WOW! To make sure the wood joints are as capable as the spruce structure, I installed spruce gussets in every joint and covered them with plywood. There are no joints in the airplane that rely only on the plywood skin for gusset strength.

With the fuselage design finalized, I refocused my attention to the tail. The single biggest change was the airfoil change to the NACA 0009 section on both the horizontal and vertical tails. I redesigned the spars and added nose radii to the elevator and the rudder. I wanted to close the gap for performance and aesthetics. There were no other major changes. However, the new horizontal stabilizer mounting changed to a durable machined fitting instead of the weldment used previously. I had to raise the horizontal tail inch for clearance reasons. Overall, Corby Starlet builders will easily see the similarity between Excalibur and the Starlet in the tail area.

By 1994, I had completed most the items described above. The wing had been designed but not built at that time. I was still looking for different aerobatic capabilities. I wanted better inverted performance than the NACA 43012A section could deliver and many years ago, when I was still going to cover the wing with fabric, I settled on the NACA 1412 airfoil section as used on the Decathlon. I figured if it worked well on the fabric covered Decathlon wing then it would work well on the fabric covered Excalibur wing.

Also about that time, I disregarded the notion of using the VW engine. More power was the key to aerobatic performance on those vertical up lines and the little VW wasn't going to give me much in the way of competitive performance. My alternative was the small Continental engines. I started looking at C85's and then thought, for the same weight I can have a C90! In order to keep the weight a low as possible it had to be a C90-8 engine without a starter or generator.

I needed more advice on those small Continental engines so I talked to Jon Sharp, pilot of Formula 1 air racing "Team Nemesis"; a friend of mine for several years. I was having trouble locating a C90-8 and Jon suggested going all the way to the mighty O-200. Whoa! That's quite a weight penalty from the C90-8! Jon assured me that the O-200 had areas where weight can be removed. I was certain, however, I couldn't make up the 30+ pounds of weight over the VW installation I had originally intended for Excalibur. The airplane was going to be nose heavy and I was getting into the mind-set of adding ballast in the tail to compensate for the O-200.

Introducing the "One-Design" and Its Influence on Excalibur

There had been discussions in the early 90's about starting a new class of aerobatics in which all the participants would fly airplanes conforming to a single specification-"One Design". The objective was to level the playing field such that the true competition was between the pilots; not the planes. Dan Rihn, initiator of the idea wanted the "One Design" to be a low cost, inexpensive to operate airplane with capabilities that would allow a novice aerobatic pilot to begin competing in the Sportsman category and still be capable enough to continue through the Intermediate category and on to the Advanced level of competition.

One of Dan's early sketches was that of the Corby Starlet or an airplane with a similar appearance to the Corby Starlet! I believe the Starlet was Dan's starting point as a minimal baseline but one that met his design criteria. Initially he specified the VW as the ideal low-cost motor and then upgraded it to the O-200. The structure was initially wood but in the end, the One-Design ended being a steel tube fuselage with a wood wing powered by a Lycoming IO-320. The final product was drastically different than the Starlet structurally but dimensionally still very similar.

At Oshkosh 1993, I saw the prototype One-Design before it made its first flight. Immediately noticeable was the fat, really round nosed airfoil on the wing. Jeez, that thing looked tough! When I finally saw it fly less than a year later, I was very impressed with its aerobatic performance. Roll rates in excess of 360 per second! Climb rates better than 2500 ft per minute. And best yet, it looked good to the judges on the ground. WOW! I was absolutely convinced more now than ever before that the O-200 was the engine Excalibur needed. I was already committed to finishing my Excalibur but I was quite envious of the One-Design's performance. The One-Design became the benchmark by which I constantly compare and evaluate Excalibur.

The Wing Is the Thing!

In May 1994, Russ Erb published an article called "One-Design Airfoil Analysis" in the EAA Chapter 1000 Newsletter in response to my question of "what makes it tick?" I didn't have the actual airfoil coordinates but, by observation, I reasoned that the airfoil was simply an ellipse with tangent lines drawn to the pre-determined trailing edge location. What Russ found was what I suspected all along: The lift, lift slope, drag, and pitching moment coefficients were not drastically different from a similar looking NACA 0016 airfoil in the normal range of flight angles of attack. However, the stall characteristics were perfect for aerobatics when autorotative maneuvers such as snap rolls and spins are considered. I liked that and since everyone else was using similar "ice-cream-cone" shaped airfoils, so would I.

So here's my current situation. I have a big heavy Continental O-200 on the front end of an airplane designed for a VW engine and even with the tail stretch I mentioned earlier, I'm certain my CG is going to end up too far forward. What to do about the nose heavy condition? Well, the main wing spar location was fixed by the fuselage design and had to remain where it already was. Likewise the seat and firewall locations were already far beyond the point where changing them would be a reasonable thing to do. I had already built the fuselage and I wasn't about to rebuild it. I opted for a more creative solution of moving the wing forward but not changing the spar location. My solution was to simply move the wing ribs forward on the spar. This would effectively move the mean aerodynamic chord of the wing forward to a position that would be favorable for my CG problem. Of course, that meant I had to re-design the wing...again...and the airfoil would look more conservative that the radical "ice-cream-cone" due to the max thickness point moving aft. It was now 14.5% thick and had a blunt leading edge and a squared off trailing edge. The planform started to evolve into something similar to the One-Design with a little leading edge sweep and the dihedral was removed for aerobatic performance.

I had read and re-read an article I saw on the Stephens Akro wing flutter problems and I had also opted at that point to fully skin the wing with plywood to provide greater torsional stiffness in my wing. There is no longer any fabric covering on any primary structure; it's all plywood. Of course, the plywood skin is also heavier than the fabric covering but at least the weight is almost entirely aft of desired CG position for the airplane which was favorable. Still, I was adding weight to the airplane which I really hated to do.

Weight Reduction Time...Again!

The heaviest thing about a plywood covered wing is the skin. Unfortunately, even though a wing skin is lightly stressed, the skin must be stiff enough to resist buckling. I needed the plywood skin to carry wing torsional loads so going back to a fabric covered wing was completely out of the question. How about high tech composites? Most people don't realize just how heavy a high tech material such as carbon fiber/epoxy really is; it's roughly three times heavier than birch plywood. So why don't I just make the carbon wing skins 1/3 the thickness of the plywood skins? Well, when you do that, the carbon skins are only 25% as stiff as plywood for buckling resistance. The bottom line is a solid carbon fiber skin, designed to the same buckling criteria as a solid wood skin, will be heavier. So next time someone tells you that they wouldn't consider flying a wooden airplane, you can tell them that wood is, pound for pound, more efficient than carbon fiber when used for wing skins...and it's cheaper!

The wing spar is a totally different animal, however. A wood wing spar is quite heavy when compared to a carbon fiber equivalent. I did the math and there was no way to justify a wood spar in Excalibur. Carbon fiber has material properties (strength and modulus of elasticity) that are an order of magnitude higher than Douglas Fir and even though the density of carbon fiber is 3 times higher, you only need 1/10th the amount of material. Unlike wing skins, a spar doesn't generally have limits placed on it for buckling problems. Its shape takes care of that. The result was I could build a carbon fiber spar that would carry my 750 pound airplane to 10 g's and it would weigh only 19 pounds! That's probably half the weight of the Douglas Fir equivalent! All in all, the lighter but stronger spar made up for some of the weight of the plywood wing skins.

90% Done and Only 90% To Go!

I figured I did a pretty good job of keeping the weight out of Excalibur. In its current configuration as shown in the pictures above, it weighs 380 pounds. That includes the engine which weighs nearly 200 pounds alone! I still need to add the wing skins, ailerons, canopy, cowlings, fairings, and paint but I figure there is less than 90 pounds of weight to go before the airplane is complete. That would give me an empty weight of 470 pounds which is outstanding considering the One-Design weighs in at more than 800 pounds empty. Add pilot, parachute and some fuel and I should have a competition weight of 750 pounds. The One-Design is at least 250 pounds heavier. Yeah, it has more power too but all things considered, Excalibur will certainly hold its own against the competition!

Inverted Fuel System, Inverted Oil System, and Flight Controls...

I have yet to fully design these systems so a detailed description is not possible at this time. In summary, however, I can quickly describe the inverted fuel system as quite simple. It is a flop tube in the main fuel tank feeding an Ellison throttle body injector. The inverted oil system is a lot more complicated and I have only conceptualized it as being a dry-sump system with a remote oil tank. The flight control system is half complete already and includes a fully push-pull tube system in the elevator circuit and most likely a push-pull tube system in the aileron circuit. The rudder is cable operated. There is virtually no similarity to the Corby Starlet in the flight control system.

So now you know. Excalibur definitely has the Corby Starlet in its family tree. The similarities are numerous but at the detail level it's a very different airplane. There are no interchangeable parts with the Starlet design with one exception: The rudder and elevator hinges are 100% pure-bred Corby Starlet! The DNA is an exact match in that particular area!


EAA Chapter 1000 Home Page
E-Mail: Web Site Director Russ Erb at erbman@pobox.com

URL: http://www.eaa1000.av.org/pix/waldopix/excalibur_story_web.htm
Contents of The Leading Edge and these web pages are the viewpoints of the authors. No claim is made and no liability is assumed, expressed or implied as to the technical accuracy or safety of the material presented. The viewpoints expressed are not necessarily those of Chapter 1000 or the Experimental Aircraft Association.
Revised -- 24 May 1999