Changing a Perfectly Good Airplane

Designing and building a modified RANS S-9, part 2.


In the first part of this article, we discussed changing the wing for improved inverted capability and increased roll rate. Michael also wanted an airplane with a great ramp presence, with the appearance of one of the current aerobatic monoplanes dominating the sport. He liked the aggressive look of the Extra, Laser, and Red Bull air racers. With the wing changes completed, he turned his attention to changing the fuselage and tail assembly to meet his needs.

Michael tries the S-9 on for size after adding sheet metal panels to the fuselage.

Redesigning the Fuselage

Fuselage shape is important in aerobatics. Beyond looking good, the fuselage must provide the pilot with excellent visibility and, when flown in knife-edge flight, the ability to hold altitude and maintain controllability. One of the seminal aircraft for aerobatics is the Pitts Special. When viewed from above, the Pitts fuselage is a symmetrical shape, an airfoil with a very blunt leading edge radius. The S-9 is similar but has flat, slab sides from the firewall to just aft of the cockpit, where it tapers to the tail. Michael wanted to change the overhead profile.

The overhead profile change began with the spinner and cowling. He used a large spinner and faired it into the cowl, approximating the front section of a symmetrical airfoil. He carried the expanding airfoil shape to the middle of the cockpit and then built a uniform, diminishing taper to the tail post. The air sees a fish-like shape as it moves along the sides of the fuselage. This shape, when held nose up by top rudder in knife-edge flight, will generate a significant measure of lift. In addition to the overhead shape, the fuselage cross section influences knife-edge performance.

The original S-9 fuselage cross section is nearly rectangular from just behind the spinner through the firewall and on to the mid-point of the cockpit. To help the air move across the fuselage cross section, Michael changed its shape from rectangular to a more rounded, nearly oval cross section. This eases airflow around the fuselage in knife-edge flight.

To achieve the desired vertical cross section (rounded bottom, curved sides, and rounded top), he carried the rounded shape of the cowling through to the front of the windshield. This rounded shape continues through the bubble canopy and transitions into the turtledeck. Michael copied the shape and structure of the Pitts turtledeck (aluminum skin over bulkheads) and adapted it to the aft fuselage structure. From the canopy’s aft end to the tail post, the fuselage sides and turtledeck are faired smoothly into the fin and rudder. To achieve these cross-sectional shapes, he added bulkheads and stringers to deepen the side and belly profile from the firewall to the tail post.

Developing the Cowl for Cooling and Aerodynamics

The overall shape of the cowl was determined by blending the lines into the airfoil shape of the fuselage, as discussed above. The cowl shape was also dictated by moving the radiator and exhaust system into the cowl and forward fuselage. The side profile of the top half of the cowl was determined by the location of the engine cylinder head. The side profile of the bottom half of the cowl was determined by the position of the radiator. The original S-9 cowl was designed around the Rotax 503 and 582. It has an annular ring opening just behind a smallish spinner. The exhaust pipe and muffler were simply hung on the right side of the fuselage ahead of the wing root. The radiator for the water-cooled Rotax 582 was mounted on the bottom of the fuselage below the pilot’s seat.

Wood strips were used to determine the profile for the cowl.

Some enterprising folks in Utah adapted the RANS S-7 cowl to accommodate a Rotax 912 installed in an S-9. This moved both the radiator and exhaust inside the cowl. In Michael’s case, he wanted to stay with the Rotax 582. Since the S-7 cowl could not easily be adapted to the reshaped fuselage, it was not a viable option.

Uwe Rothscheim, a German homebuilder, created a lovely red S-9 and moved the radiator and exhaust inside the cowl. Rothscheim and Michael corresponded and shared ideas about Michael’s airplane. In the end, Michael elected to use Rothscheim’s approach for the exhaust system move and took a different approach to relocating the radiator. Rothscheim mounted the radiator on the front skin of the cowl, and Michael wanted his completely inside the cowl.

Before applying foam, the entire airframe was protected with plastic.

Michael looked at LSA flying machines using Rotax engines. Nearly every airplane has the radiator buried in the cowl with a dedicated inlet for cooling air. The air is ducted into the radiator and out of the cowl. Michael decided on this arrangement for his S-9. He located the radiator horizontally below the engine and ducted air into and out of the evolving cowl shape. He received assistance from California Power Systems (CPS) in sizing the radiator. Using information regarding the recommended radiators from Rotax, Michael found a Honda Civic radiator that met the requirements. He added a cowl flap to the system so he could modulate the airflow through the radiator and manage engine temperatures. Next, he turned to John for help in developing the ducting to move air into and out of the cowl.

Applying polyurethane foam for the cowl plug.

With the foam applied, it’s time to start shaping the cowl plug.

John, like a symphony director, waved his arms and magic happened. Using cardboard and lath strips, he developed the inlet shape for the radiator. John reminded Michael that air does not like to turn. When it has to, it adds cooling drag in the form of resistance and turbulence. It turns out a lot of work was done with this technology in the late 1930s. John relentlessly shoveled the contents of old, yellowed technical reports into Michael’s brain. The inlet lip radius is a critical factor. The shape of the lip must be quite rounded to help the air depart the free stream coming around under the spinner and into the mouth of the inlet. With a little bit of TLAR (That Looks About Right) and John’s years of experience, the duct shapes appeared.

The completed cowl plug, ready for fiberglass.

The fiberglass cowl after it was removed from the plug.

With the radiator mounted parallel to the incoming airflow, a plenum, tall in front and narrowing toward the back, entices the air to flow downward through the radiator. The ducting on the underside of the radiator is narrow at the front and wider at the back. This causes the departing air, now warm with energy extracted from the coolant, to expand and pick up velocity as it moves into the expanding volume of the exhaust plenum. At the aft end of the plenum is a cowl flap. The flap can be moved to increase or decrease the size of the cooling air exit from the cowl. Behind the cowl flap is a tunnel built into the belly skin, beginning at the firewall. This tunnel helps the air transition back into the free stream with minimal increase in drag.

The radiator is sandwiched between the intake plenum (top) and the exhaust plenum (bottom).

A cowl flap modulates airflow through the radiator to manage engine temperatures.

The final shape of the cowling was now determined. The exhaust was mounted inside the first bay of the fuselage. CPS again helped Michael make the necessary adjustments to the tuned exhaust and fit it inside the airplane. Michael then installed the spinner and radiator air intake onto the engine.

The completed cowl after paint. The large intake cools the radiator.

Polyurethane foam was then sprayed over the plastic-wrapped engine and firewall. The foam was trimmed away and sanded to final shape, using templates to ensure symmetry about the vertical and horizontal centerlines. Once the foam work was completed, latex paint was applied to seal the surface. Mold release was applied to the plug. Fiberglass cloth and resin were installed over the plug to create the layups for the cowl. The cowl was split along a horizontal parting line and removed from the plug. The plug was removed from the engine, and the completed cowl was fitted with hardware for installation.

The cowl and right wheelpant installed on the fuselage.

Changing the Tail Group

The largest cosmetic change to the airplane occurred in the tail surfaces. Michael wanted to add aerodynamic horns to the rudder and elevators. He also wanted to sharpen the trailing-edge radius of the flight controls. Additionally, he wanted to move the trim tab into the envelope of the right elevator, instead of having it hinge off the right elevator trailing edge. No matching tab was installed on the left elevator in the original S-9. Finally, Michael wanted to change the outline of the fin, rudder, elevators, and stabilizers to more closely mimic the outline of the new wing.

The old trim tab hinged off the right elevator trailing edge.

John’s guidance for the tail group changes was simple: do not change the areas of the surfaces, and keep the hinges where they are. This precludes inducing control problems, like flutter or loss of feedback, into the control system. With this directive in mind, Michael changed the fin and rudder shapes while retaining the same area ratios. He also decided to add a piece of aluminum trailing edge material to the trailing edges of all the surfaces. The S-9 surfaces are all made up of 4130 steel tubing. The trailing edge tubes are 3/8-inch in diameter. Riveting on the V-shaped trailing edge piece sharpened the trailing edges to match the trailing edge of the new ailerons.

The tail surface modifications were made by cutting and rewelding the structure to achieve the desired shapes. Care was taken to ensure the area ratios were maintained. The original mounting structure, hinge points, and flying wire attach points were maintained as well.

The new trim tab is still on the right elevator, but the hinge has been moved and the tab itself has an entirely different shape.

Building the Airplane

The airplane was built over the course of three years, working in the spring, summer, and fall. Michael’s garage is unheated, so winter work in Kansas was not even a consideration. The fuselage was built first, followed by the wings and then the tail surfaces. John kept track of Michael’s work, making suggestions along the way. Once the structure was completed and all systems installed, it was covered with Stewart Systems fabric and fillers. Michael painted it with water-based polyurethane after his move to Oregon in the spring of 2013. Although there was a challenging learning curve, the final finish is excellent. The airplane received its airworthiness inspection in December 2014. After recovering from a Christmas knee injury, Michael flew the airplane in March 2015.

Michael changed the fin and rudder shapes, but retained the same area ratios.

Did the Changes Work?

Well, the proof is in the pudding. The airplane looks great and most people think it will cruise about 200 mph. Michael sure wishes it did! First flight revealed a nice little airplane with docile handling characteristics. There were some hits on the change goals and some misses.

First the Hits

The airplane absolutely meets the ramp presence requirement. It looks like it is moving sitting still, and most folks who see it ask what it is. Sea-level performance is great. The S-9 takes off and climbs 1000 fpm at 65 mph indicated on full power (6500 rpm for one minute) from the Rotax 582. At 2500 msl, it has a top speed of 118 mph at 6000 rpm (the 582’s maximum continuous power setting), as verified by the GPS triangle method. Cruise speed is right at 100 mph (5800 rpm, 2500 msl). In a glide, with the power fully off, the S-9 gives a brisk descent. It doesn’t come down like the Pitts, with the glide angle of a chromed safe, but it’s close. Since the 582 has a fairly rough dead idle, Michael always carries a bit of power so the engine is smoother and the descent is a bit flatter.

To lighten break-out forces, Michael added spades to the ailerons.

The controls are light in pitch and yaw; they are very predictable with good stick force throughout the speed range. Control movements are small, like the Pitts, and response is immediate. Stall speed is 40 mph indicated. It quits with a bit of aerodynamic shudder and falls off straight ahead. Touchdown speed, power off, is right at 65 mph. The airplane must be flown on due to its very flat sitting attitude and short main gear legs. Sitting on the ramp, the airplane is only 7 degrees nose up. A normal takeoff can be made without ever lifting the tailwheel! It lifts off at about 65, and to keep it from running away, the climb deck angle at 65 mph is about 18 degrees nose up (feels like straight up).

A full-stall landing results in a tailwheel-first touchdown, followed by the mains plopping onto the runway. The bungee gear is pretty stiff, so it hops a bit when this happens. Directional control is good with a strong rudder and Matco tailwheel and brakes.

In Oregon, high temperatures are few and far between. On an 85-degree day (hot for around here), all coolant temperatures are low (130 F) and EGT is about 1000 F. The cowl flap is generally closed except on climb-out. An electric fuel boost pump is used during takeoff and landing to ensure good fuel flow.

Aerobatics are fun and easy with wonderful visibility out the bubble canopy. The only blind spots are the shoulder-mounted wings. Inverted performance is great. The new airfoil holds inverted flight with minimum nose-up stick force. The stock S-9 requires a noticeably higher inverted nose-up stick force than Michael’s to hold level flight. Michael is still exploring heavier negative loads. He has flown it to -2.5 G with no adverse issues to report. Michael is no longer doing the “big boy” upside down stuff, so pushing is not a real issue. What is an issue is the ease of aerobatic maneuvering, and the new airfoil meets all its requirements. The airplane holds energy well and is slow to slow down when the power is chopped. Loops are small and rolls are a delight and quick, but require a strong right arm, as is discussed below.

Now the Misses

The folks at RANS advertise the stock S-9 can be built from its very complete kit in about 700 hours. Although meeting the original build hours was not one of the goals of the project, Michael’s builder log reflects nearly 1400 hours “at the bench.” Over and above the time needed to hammer the airplane together, Michael also spent hours making sketches and working with John to develop the changes made to the airplane. When making modifications, one change often drives other changes. This increases the time needed to design and implement the change. Lastly, sourcing new and different parts takes time away from the building process. In a complete kit, everything is there for the builder to assemble.

Weight management is a challenge. Michael gave the airplane the nickname of Fat Albert after it was weighed. It came in at 508 pounds, nearly 75 pounds over the design goal. The weight gain breakdown seems to be: wheelpants (Pitts units), the canopy and tracks, extra structure to carry the ailerons in the wing and extra weight of the ailerons themselves, aileron balance weights, the new cowl, added fuselage sheet metal panels and, finally, the water-based paint.

The roll rate did not meet expectations. The design goal of 240 degrees per second was not met. The actual rate is about 180 degrees per second. Roll forces are also heavier than expected. Michael built the airplane using the original aileron actuation system: push/pull cables. These cables, when forced to make a bend in the installation, are very draggy. Michael added spades (aerodynamic “power steering”) to lighten the break-out forces. Another local Oregon S-9 builder told Michael his ailerons lightened with use and wear-in. That retching sound in the background is John gagging over Michael’s use of the cables. At this point Michael agrees with John that he should have redesigned the aileron actuation mechanism. A better system would be push/pull tubes. Michael will correct this problem in the next iteration of the airplane.

When the airplane was initially flown in the cool spring of Oregon, internal cowl heating issues were not seen. However, as summer arrived, so did vapor lock in the engine-driven fuel pump. Two additional NACA scoops, courtesy of Van’s Aircraft, were installed on the cowl. One feeds fresh air directly to the fuel pump; the other directs fresh air across the exhaust muffler. Additional exit holes were added to the cowl to give this additional cooling air a way out. A rearward opening shroud was placed around the exhaust pipe where the pipe leaves the cowl. The aft face of this shroud is open and is a draw point, pulling hot air out of the cowling. These additions have eliminated the takeoff and climb vapor lock issues.

Do it again?

Michael knew what he was getting into, and the challenges of making modifications had already been experienced through his work in the aerospace industry. Michael substantially met the majority of his goals and had a great time doing it. To safely reach these goals, Michael needed the resources of a skilled engineer for the airframe changes and the friendly folks at California Power Systems for all the engine stuff. If Michael was offering advice, and you were taking it, he’d recommend getting technical support from the professionals who do this sort of thing for a living. The cost of having your work checked by someone who really knows what they’re doing—minimal. The joy of flying your safe and fun creation—priceless.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.