By Bruce H. Carmichael, Kitplanes Magazine, Oct., 1993
In August, 1992, Mike Arnold set a new 3 km FAI world speed record of 213.18 mph in Class C1aO for airplanes weighing less than 661 pounds at takeoff. He set the record in his brand new design, the AR-5, an all composite homebuilt sport plane powered with a 65 hp Rotax. Seeing the photo and caption in the October issue of the EAA's Sport Aviation magazine certainly aroused my curiosity.
While I was trying to figure out how to get in touch, the phone rang, and it was Arnold, who was directed to me by our mutual friend, Stan Hall, the sailplane designer, builder and pilot. Arnold offered to let me study the AR-5 and asked if I could do a representative waviness survey on his wing. I jumped at the chance. The day after Thanksgiving found me in his compact shop in Crockett, CA, where he had the AR-5 assembled.
FIRST IMPRESSIONS
The gleaming white plane immediately said "smooth, light and tight" with a near absence of protrusions. It was up on saw horses, allowing me to crawl under.
The bottom of the airplane is every bit as smooth and protruburance-free as the upper part. Arnold will jolly well get drummed out of the airplane manufacturers's club for that as they have traditionally mounted all the drag-producing garbage underneath where they think no one will see it.
I saw no ripples in reflected light, indicating a very low degree of surface waviness. There are no unnecessary paint stripes to trip the laminar boundary layer. All intersections are beautifully filleted, and joint lines are flush and barely discernible. It looks like it was carved from a solid block of ivory. This is the work of a master craftsman and on a par with the finest high-performance sailplane technology.
THE WING
As my friends in a group named TWITT (The Wing Is The Thing) say, the wing is the thing. Arnold's 21 foot span, aspect ratio 8:1 wing employs an NACA 65-418 section at the root and an NACA 65-215 at the tip. The plain flap of 25% chord covers 50% of the wing span, and where the wing fuselage fillet would have cut it off short at the root, Arnold cleverly extends it with a lower-surface, split-flap addition.
The 22% chord, top-surface hinged ailerons cover 40% of the span. No curtain or foam rub seals are employed, but the hinge line gap is very small. The leading edge of the ailerons, elevator and rudder are slightly extended outside the contour line as recommended by Hoerner to reduce drag. Elevator and rudder are mass-balanced at the tip by a narrow horn balance extending to the leading edge. They are all external balances and, of course, protrude when controls are deflected
The streamwise running gaps are not rub-sealed. The control surfaces are fastened on with raised head machine screws for easy removal. The cusps have left in the ailerons, but Arnold reports excellent later control forces. The controls are well harmonized on all three axes.
The horizontal tail uses a low-drag, 12% thick section at the root and 10% at the tip. The vertical fin uses a 10% low-drag section. The tail surfaces are very generous in area as Arnold designed this as a personal light airplane, not a racer.
THE FUSELAGE
The 14.5 foot long, 23 inch wide, 36 inch deep fuselage has very pleasing lines. The slightly less than 1 foot diameter flat open cowl face is faired elliptically back to the point of maximum fuselage width. It is interesting to note that flat faced Navy torpedoes, faired in this manner, show no increase in drag over a streamlined nose of equal wetted area. This fuselage with the internal flow system sealed off should show no increase in external drag due to nose form.
The forebody is smoothly faired into the conical afterbody. A beautiful free blown canopy is generously faired at its leading edge into the basic fuselage, thus softening the adverse pressure gradient approaching the canopy and preventing excessive thickening of the boundary layer. This type of leading-edge filleting is also found on wing and tail surface intersections.
The full-width canopy hatch is foam-sealed along the transverse but not along the longitudinal seams. The canopy seams were not internally tape-sealed during the record attempt. The hatch is secured by two piano hinges on the right, and a single latch on the left. The pilot can release the latch in flight, and the canopy then raises only half an inch off the left rail.
THE WING FUSELAGE JUNCTURE
To give away as few drag points as possible on this important juncture, Arnold did not allow the fuselage to contract until very near the wing trailing-edge. He had read an article by an old research pilot, flight instructor and friend, Ray Parker, on the importance of this. There is a stagnation-relieving fillet at the leading-edge, a corner radius fillet to prevent the large boundary layer-thickening found in a square corner and an expanding fillet on the aft portion to soften the adverse pressure gradient. The forward fillet is kept small to prevent excessive supervelocity near maximum wing thickness, with its attendant flow deceleration problem aft of maximum thickness. The bottom of the wing is flush with the bottom of the fuselage, thus virtually eliminating two intersections. This juncture should help reduce the interference problem common to low-wing aircraft, particularly at the higher lift coefficients, and, no doubt, it helps also at high speed and low lift coefficients.
LANDING GEAR
This is the cleanest fixed landing gear that I have seen. It was placed outside the propeller slipstream so that it does not see the higher dynamic pressure of the prop blast, nor is it subjected to the non-zero angle of attack from the propeller blast or swirl. The low-fineness-ratio fairings of the gear legs meet the wheelpants flush with the outer side, thus eliminating two more intersections. An expanding fillet is employed at the rear of the leg/wing intersection.
The wheels protrude from the pants a mere 2 inches; the steerable tailwheel is tiny and unfaired. It is located at the end of a circular-cross-section rod. The airstream sees an elliptical cross section of this round rod due to its angle with the flow.
CABIN AIR INTAKE
The protuberance under the fuselage admits cabin air and looks like a miniature P-51 cooling duct. The intake is spaced out from the fuselage skin so that, should carbon monoxide become entrained in the fuselage boundary layer, it will not be drawn into the cabin. The only protuberance under the wing is a single dogleg pitot tube made of ¼ inch round aluminum tube, unfaired.
POWERPLANT
The liquid cooled Rotax 582, according to the manufacturer's specs, delivers 66 hp at 6600 rpm. It has a very large exhaust system that Arnold has managed to house within his fuselage to the right of the engine . A portion of the open cowl face feeds air to cool it, exhausting through a 6-square inch exit duct raised about ¼ inch out of contour. Cooling air flows over the rounded intake lip and dumps into the internal cowl plenum. The engine exhaust exits through an unfaired, circular-cross-section steel tube that protrudes through the fuselage wall and turns downstream.
The expansion chamber had to be made elliptical to fit, and Arnold thinks it may have detuned the system and cost him a bit of power. Arnold tried a large-diameter spinner with the inlet air entering through the circumferential entry, but he found no change in top speed.
SURFACE QUALITY
Having failed to see any large surface waves in the wing with reflected light, I next rocked a machinist's 6-inch scale on edge over the contour in the chordwise direction. I could not feel any jarring due to flat spots or large waves.
Next, we put a piece of masking tape marked off in quarter-inch intervals on the surface in a chordwise direction located 5 feet out from the fuselage wall. We read the dial indicator operating between fixed legs spaced 2 inches apart. A plot of the readings on the vertical scale, amplified 500 times compared to the chordwise location horizontal scale, revealed the following:
We could now find the spar where the largest peak-to-peak reading of 0.0036 inch over a length of 3.6 inches gave a ratio of one part in a thousand. A second wave nearby had a peak-to-peak reading of 0.002 inch over a 1.5 inch wavelength, or 1.33 parts in a thousand. Farther forward on the chord, a wave of 0.001 inch, peak-to-peak reading over a 4 inch wavelength, yielded a ratio of one quarter part in a thousand.
What do these numbers mean? It means that Mike Arnold has smoothed his low-drag wing to a greater degree than any production composite sailplane and equal to or better than the best achieved by dedicated sailplane pilots after a winter's work filling and resanding their wings.
SO, HOW DID IT GO SO FAST?
How do we explain a man-carrying, low wing, tractor-propeller-driven airplane with a fixed landing gear and generous tail surfaces achieving this record-breaking performance on such low horsepower?
The FAI rules permit losing some altitude on the way into the course and over the measured run. Arnold had to hold his entry speed to 220 mph to stay within his flutter clearance. Although he could have entered at 492 feet, he chose to enter at 300 feet to be safely within the requirement. By letting down to 30 feet at the end of 4 km, he would have a descent angle of 1-degree, 18 seconds. His weight component along the flight path would give him an extra13.6 pounds of thrust to help stave off the decay of entry speed toward normal maximum level flight speed. By observing his readings during the runs and comparing with the ground data, Arnold was able to get a first-order correction for his airspeed system. He feels his maximum level flight speed was 207 mph.
Using this speed, 65 hp and 82% propeller efficiency, the AR-5 drag area comes out to 0.88 square feet. People have been trying for a long time to get a man-carrying airplane under 1 square foot of drag area from such calculations. The latest Formula 1 racers also appear to be in the vicinity of 1 square foot.
Dividing the drag area by the wing area of 55.125 square feet yields a traditional drag coefficient of 0.16. Drag area numbers have the disadvantage of combining size and speed, while the drag coefficient based on wing area moves around as the ratio of wing area to total area changes from design to design, but if we divide the drag area by the total wetted area of 236.4 square feet, we arrive at a wetted-area drag coefficient of 0.0037!
What does that mean? The lowest figure I had seen previously for a propeller-driven airplane was 0.0040 for the P-51 fighter, which had a larger Reynolds number, which means lower skin friction drag. In addition, the P-51 has a flush retractable landing gear and zero cooling drag. Mike Arnold's AR-5 is the most efficient man-carrying, propeller-driven airplane I know about.
CONCLUSIONS
The above comparison may puzzle traditional aero engineers. But the conclusion is that composite construction executed by a person of Mike Arnold's training and skill not only eliminates roughness drag, but the reduction of surface waviness permits achievement of the design extent of laminar flow. By using the 65-series airfoils, Arnold has almost 50% of the wing wetted area operating laminar with its greatly reduced friction drag. His very clean fixed landing gear located outside the propeller wake, his very low wing/fuselage intersection drag, and the absence of protuberances and associated form-drag all help. His efficient Cato propeller with twist distribution adjusted for fuselage interference as suggested by propeller guru Eugene Larrabee also helps. It appears that his cooling drag is also very low. The low total wetted area of this plane completes the picture. I couldn't resist doing a component by component drag buildup calculation. The sum of the external drag areas was 0.86 square feet, leaving 0.074 square feet or 8% of the total for cooling drag. With a less conservative fuselage drag estimation, the margin for cooling drag became 12%.
THE PERSONAL CONTRIBUTION
It is now obvious that as interesting as the advanced craft are, the people behind them and their stories are even more interesting. Arnold has been fascinated by high-performance aircraft all of his life. He carefully memorized the best features of the best racing planes and fighters and spent countless hours drawing his dream midget fighter-personal aircraft that would be fast, maneuverable and safe.
Arnold's degree is in movie making rather than engineering, but he has found and read the great books that present the real physical picture in understandable form. In particular, he studied "Fluid Dynamic Drag", by Hoerner to the extent that he took it on vacations with him. He then apprenticed himself to Fred Jiran of Mojave, California, to learn to work with composites. Jiran repairs and fine-tunes racing sailplanes, so Arnold got to see the finest drag-reducing tricks of the high-performance sailplane manufacturers.
He learned clever ways to build one-off prototype planes using the Burt Rutan method. Arnold's goal is to keep things as simple as possible while giving great attention to drag reduction. Engineering judgment and superb detail design are found in the AR-5. Arnold is a prime example of a man with a dream and goal who has made all the long, hard acquisitions of knowledge and hands-on experience necessary to reach that goal. There are many lessons to learn from his history as well as from the AR-5.
{Author Bruce Carmichael is an aerodynamicist who is well known for his work in the area of drag reduction and high-performance sailplanes. His comments are recorded in Mike Arnold's,"Why it goes so fast" and "Moldless, low drag wheelpants" videotapes, but he is not commercially involved with the AR-5. --DM, Ed.}
(Kitplanes Magazine, October, 1993)