Concept for an Ultralight Aircraft.
Shortly after college, in 1981, I began work on an Ultralight design. The FAA rules at the time limited an ultralight to 155 lbs empty weight and be foot launchable. Those were the only rules. The foot launchable rule seemed like it was not enforced.
I took an after hours course in advanced composites at Boeing and wrote a program to estimate properties of composite layups. I realized that these new materials may allow a good aerodynamic design while meeting this weight restriction.
I then wrote a program to find the stresses on a wing section using a lumped mass approach for the compression and tension due to bending moment spanwise along the wing. Then, with scaling data for the GAW series wing section, the program made the wing as thin as possible for the given amount of graphite reinforcement as it progressed out from the root. A conventional configuration resulted in a wing which was too heavy. The solution was to make it a tandem configuration. The program was enhanced to optimize the front and rear span, root chords and fore/aft position so as to provide the stability margin over the CG range and minimize weight. The planforms were constrained to provide the optimum lift distribution and the 3/4 chord point constrained to a straight line so that the flaps and spoilers had straight leading edges. The program generated the configuration seen on this page.
During this time, several people became interested in this project. I formed a partnership with three good people who supplied funding and effort toward building the prototype. We decided to enter our project in an EAA ultralight design competition. That is when we generated this preliminary design document and sent it in to register for the competition. Speedolight.pdf
Shortly thereafter, the FAA changed the rules for the Ultralight category. The "foot launchable" requirement was dropped, but the maximum power off stall speed was reduced to 24 kts. The maximum speed was also limited to 55 kts. The allowable empty weight was increased to 254 lbs and fuel capacity limited to 5 gallons.
The program was then enhanced to constrain the stall speed to 24 kts, with all the other features remaining the same. The configuration thus generated looked identical, but was almost twice the size. The minimum weight was too high with the original wing construction technique as the solid foam core became too heavy with the larger wing area. Also, at this size the kevlar skin construction became very expensive. The wing section was then redesigned to use skin which was a sandwich of 10 mil fiberglass on each side of a 1/8 in foam core. Two spar sections of the same skin were added at about the 0.2 and 0.4 chord (I don't remember exactly) positions. The foam was deleted in the skin between the two sections and layers of graphite tape were added to these areas. The fuselage was also the same fiberglass sandwich with graphite reinforcement. The generated design then just met the 254 lb requirement.
There were skeptics as to whether the tandem wing configuration would be stable in pitch. I wrote another program which did a 6 degree of freedom (pitch angle, pitch rate, horizontal and verticle positions and velocities) simulation of the aircraft with any given initial conditions. It used a 6 point predictor/corrector algorithm (insert some famous mathematicians name here) for high accuracy. It accounted for the downwash of the front wing as the rear wing encountered it based on the movement through the air. All components of the aircraft were accounted for as far as inertias, lift and drag. This program took 2 hours per second of simulation on my VIC-20 computer. The program in basic took up all the memory with room for only one print statement. I directed it to the cassette drive. Plots of this data showed, even with the craziest of initial conditions, that the plane was very stable. A byproduct of this simulation was a good estimation of steady state performance. Holding the stick full aft at full power would result in about an 1100 ft/sec climb with the pitch angle slightly oscillating as the front wing was stalling and recovering.
The simulation showed that the top speed was still about 100 kts. I looked into some new electronic pressure sensors. In order to avoid a very high cost sensor, I designed a venturi type pitot tube to feed the correct pressure range to a reasonably priced sensor. From there we would build an engine governor to hold 55 kts. One of my partners, Robert Morris, was to design the electronics. I went to a seminar on the new category rules with the FAA present and a representative said that it would probably be acceptable to add a speed governor. This feature would act as a safety feature, making the aircraft easier to land as the pilot would not have to touch the throttle until it was time to touch down.
We rented part of a boat shop and started work on the molds. We had the fuselage plug nearly finished and made a first attempt at the wing molds. The wing molds were not working out because I was using 2in thick sheet foam cut out to the wing profile and gluing them with alignment dowels. This was not accurate enough, so I redesigned the wings to linearly vary through just two ribs between the root and the tip ribs. This kept the planform to a good approximation of the original continuously varying planform. This way we could build a wooden skeleton and fill it with wet pour foam, then use a long sanding bar across the wooden ribs of the form.
It was at this point that we decided to hire the owner of the boat shop to build our wing molds. He had constructed the original wing molds for the Glasair homebuilt. He wanted about $26,000 to do the molds. We had people ready to invest in the project, but the Ultralight market was going soft. The initial frenzy was over. The best selling Ultralight was selling for about $5,000 and they were only selling 10 per year. We estimated our cost of construction to be about $8,000. The selling price would need to be about $16,000. If we only sold 10 per year, there would not be enough revenue to support the manufacturing effort. The performance exceded the best motor glider on the market, but to get it certified with the FAA as a motor glider to FAR part 23, it would optimistically take an estimated 5 engineers one year of work. Probably more than $1,000,000 worth of effort. The best selling motor glider was selling for about $60,000 but they were only selling 3 per year. Again not worth the investment.
Thus the project ended. I really appreciated the contributions of my former partners. James Greenwood was an excellent draftsman and engineer. Dwight Rouso did much technical review of the math involved. Robert Morris kept track of the business/accounting side. All provided much input into the technical details and labor involved in mold building. If I had had enough money to complete it myself, I would have liked to build one for my own use. The programs that I wrote and the final design are now long gone.