"Mike Arnold's little AR-5 puts world-record performance in a tiny package."
By Peter, Lert Air Progress Magazine, July, 1995
Think of world speed records for propeller-driven airplanes and your mind turns naturally toward such brutes as the current incumbent, a Grumman F8F Bearcat modified to the extent that hardly anything is original except the data plate and the shadow. Moreover, the fact that what was already the best-performing piston fighter of World War Two required such extensive modifications shows that such speed records are hard to gain. It took about 35 years for an earlier Bearcat to wrest the title form its former holder, an equally modified Messerschmitt with such radical changes that it was an ME-109 in name only.
Given all the hype that surrounds a new absolute speed record, it's not surprising that some other, more-modest speed records are sometimes eclipsed. That's a pity, though, as in many cases these "class" records are, in their own way, every bit as impressive in terms of aircraft and pilots as their bigger and more-flamboyant counterparts. Moreover, both pilots and airplanes are "closer to home" and considerably more accessible. Thus, I felt honored when Mike Arnold, of Crockett, California, offered me the chance to fly his very own world-record airplane: the remarkable little AR-5.
I was excited, to. After all, few of us are privileged to fly anything that holds a world record of any type (although a number of my flying companions used to accuse me of owning the world's raunchiest 1953 Cessna 180). In Mike's case, the AR-5 has held the world speed record for class C1aO, propeller-driven airplanes weighing no more than 300 kilos (661 lbs), since 30 August, 1992, and it looks like it may continue to hold it for quite a few years to come.
Even more impressive than the record itself - 213.18 mph - is the fact that Mike, over more than ten years of single minded effort, managed to design and build an airplane that breaks an aerodynamic record even when it's sitting still in a hangar. It's the first manned aircraft with a total equivalent flat plate drag area (I'll explain that in a minute) of less than one square foot. This minuscule drag has allowed the AR-5 to reach its heady record speed powered by a totally stock, 65-hp Rotax 582 two-stroke -- the same engine that shoves much lighter Ultralights along at all of 45 mph. Even more remarkable, as aerodynamicist Bruce Carmichael likes to point out, the AR-5 goes that fast with fixed landing gear.
I'd heard about Mike Arnold and his remarkable project for a couple of years before I finally managed to meet up with him at one of Oshkosh's more select and obscure functions: the annual dinner meeting of the True Believers Society, a group that persists in the belief that there is indeed such a thing as laminar airflow over aircraft structures. These meetings generally combine (too much) food and (even more ) drink with at least one lecture or presentation so abstruse that each place setting should include a slide rule as well as the usual knives and forks. Often, there's some empirical aeroballistic research as bread sticks and dinner rolls are hurled at speakers who make outrageous claims.
Mike's unassuming presentation a couple of years ago was refreshing in a couple of ways. Not only was he able to show that one could achieve breakthrough design performance by almost fanatic attention to detail, rather than by any revolutionary new aerodynamic techniques, but he was also able to hold the attention of some 25 of America's finest aerodynamicists for close to an hour...without a single Viewgraph slide or equation!
Even so, more than a year was to elapse before I'd get a chance to be the first pilot other than Mike himself to fly the AR-5. First, the vicissitudes of travel schedules prevented it; then, when my schedule opened up, a broken throttle control forced Mike to make an off-airport landing that required minor repairs. Finally, Mike and I were able to synchronize our lives once again, and by careful planning, I was able to time my arrival in San Francisco Bay area to coincide exactly with last March's "Storm of the Century".
In the interim, however, I'd had a chance to learn quite a bit about both the design and construction of the AR-5 from Mike's fascinating set of videotapes (see sidebar). When I finally met Mike and the AR-5 at the Nut Tree Airport during the one-day respite between two monster storms, I was to find that the airplane more than lived up to all of its advance billing.
So what does it take to go faster than a Bonanza on 77% less horsepower ("and with fixed gear, too!)? Obviously, it takes a smaller, lighter airplane; but even more significantly, it takes one that's much cleaner aerodynamically. Sure, the AR-5 is smaller than a Bonanza -- but if it were to be 77 % smaller to match it's power, it would have a wing span of only 8 feet, and Mike would have to fly it in circles around his head at the local model-airplane field!
True, at 660 lbs (vs 3800 for the biggest Bonanzas) it's not that far from being 77% lighter -- but the real secret is that it has about 80 percent less drag than its production-airplane counterpart ("and with fixed landing gear, too!).
This is where the idea of equivalent flat plate area comes in. Since different airplanes not only are of different sizes, but also fly at different speeds, it's hard to come up with a single yardstick for comparing drag (which changes according not only to the size and cleanliness of a design, but also with the square of its airspeed). Instead, designers use equivalent flat plate area (usually abbreviated as "F"), which expresses total drag in terms of the size of a hypothetical flat plate -- say, a sheet of plywood.
The average lightplane has an F on the order of 10 square feet or less, which doesn't sound like much, but stop and think for a minute. Ten square feet is about the size of, say, a card table -- one of those folding ones we used to carry outside for a picnic. If you're carrying it on edge, it doesn't take much of a breeze to either rip it out of your hands, or knock you over -- 10 or 15 mph would be plenty. Now imagine how hard that table would shove you at 100 mph, and you begin to see why it takes so much power to go fast (and, incidentally, why installing twice as much power doesn't make you go anywhere near twice as fast).
Of course, a flat plate is literally as unstreamlined as you can get, and anything you can do to streamline it will reduce its drag. Otherwise, it would be much easier just to make airplanes cube shaped (are you reading this, Barnaby Wainfan?).
Streamlining even small items can make a very significant difference. For example, many biplanes profit by having streamlined stainless steel flying wires (actually blade like airfoil shapes) between their wings, rather than the simple round cables you might find on an ultra-light. The more you can reduce the drag, the faster you can go; but the faster you go, the more even a small amount of additional drag can slow you down. thus, it's pretty easy to get the first few mph out of a drag reduction program, but each additional knot may cost you more dearly in time, effort and sweat.
I don't even know if a world record was what Mike had in mind when he started to sketch the AR-5. According to him, what he wanted was "a little dogfighter', something he could throw around the sky for fun. At the same time, I'm sure he felt the challenge. As someone very familiar with the composite structures that go into homebuilts, he was also starting to immerse himself in the techniques and art of aerodynamic design as he laid out the parameters of what would be totally his own airplane, rather than a kit or design conceived by someone else.
Weight and cost were no doubt factors as well. To some extent, airplanes cost by the pound, just like any other commodity. More to the point, weight has to be lifted, and while we intuitively think of drag as only the parasitic kind -- the drag of nonlifting protrusions such as landing gear, struts and antennas -- the very act of producing lift itself also produces drag: induced drag. Thus, a smaller airplane is a lighter airplane, producing both less parasite drag and -- if properly designed -- less induced drag.
Moreover, a smaller and lighter airplane can be powered by a smaller engine. It obviously didn't escape Mike's attention that the 65 hp liquid-cooled Rotax 582 is not only significantly smaller and lighter than a Continental A-65, it's also aerodynamically much cleaner and costs -- even brand new -- less than half as much as a well-worn Continental.
There's a lot more to a design, however, than just making it small and light -- and each design decision involves tradeoffs and compromises. Take, for example, the question of wing size and shape. A long, tapered wing like that of a sailplane produces less induced drag, for a given amount of lift, than a short stubby one. One the other hand, it also requires a heavier and more-sophisticated structure for adequate strength and stiffness; so at some point, the benefits of a skinny wing diminish. A rectangular wing has more benign stall characteristics than a tapered one -- but at high speeds, it also has more drag.
Overall airplane layout is also important. For example, there are those who claim that a mid-wing location, with about as much fuselage above the wing junction as below it, is the most efficient. Others, however, hold that much of the drag of an airplane comes from the intersection between the wing and fuselage -- and a high - or low wing layout has only one intersection on each side, rather than two.
What about the landing gear? There's no question that retractable gear have less drag than fixed gear -- but is it possible to design a very clean fixed gear with a drag penalty that is small enough to offset the additional weight, complexity and cost of retractable gear?
These were the kind of questions that concerned Mike Arnold as he laid out the basics for the AR-5. The result is -- as can be expected -- a small airplane, with a wingspan of just over 21 feet. It's also light; at max gross weight, wing loading is only about 11.8 lbs per square foot, a bit less than half that of a Bonanza. The aft fuselage is fairly short; while this requires good-sized tail surfaces for stability and control, it's worth it in terms of decreased weight and wetted area. The nose is long, partly because of the light weight of the little Rotax and partly because long, skinny shapes tend to have less drag than short, fat ones (Gee Bees notwithstanding).
As some great artist once said, "God is in the details," and this is where the brilliance of the AR-5 really shines through. Take the landing gear, for instance. It's conventional, of course, with its two long legs, nicely faired and far enough outboard that they're not exposed to high-energy air from the prop (which, given its tiny size, isn't all that far outboard anyway). An additional benefit of this wide stance is that it makes the airplane surprisingly docile on the ground. Spring shock struts, with a couple of inches of travel, are built into the legs. The wheel-pants are what you'd expect of a high-speed aircraft: small and tight, with an extended afterbody and almost no clearance around the tire. What's different is the way they fair to the struts: Not symmetrically, but offset all the way to one side, so that there's only one drag-producing intersection rather than two. A long, curved, tubular spring fairs aft out of the tailcone, supporting at its end a tiny solid tailwheel.
Similar details about all over the airplane, including a flush-fitting fuel filler door and a combination fuel-tank vent and filler-drain scupper all of a half an inch tall -- everything carried out at a level of workmanship that rivals or surpasses the most expensive German sailplanes. Some of the most significant details, though, aren't all that visible; the overall shape of the fuselage and its relationship with the location of things like the wings and cooling-air outlets.
Mike studied most of the classic of aerodynamics to find out that it's not just the shape with which an airplane meets the oncoming air that's important; equal, if not more important, is the shape with which it leaves the air behind as it passes. Mike found that the AR-5 was just about equally fast with a big prop spinner, a little one, or none at all.
Sudden transitions in the cross-section of an airplane cause drag, whether they're increasing or decreasing -- a low-speed version of the area rule that makes Coke-bottle shapes more efficient for jet fighters. Mike was able to work these factors very much to his advantage. For example, the nose of the airplane is long enough that the airflow has a chance to align itself (aided by outflow from the radiator cooling outlets) before it gets to the wings. About the time it leaves the trailing edge, both the presence of the canopy and some very carefully shaped fillets keep the cross-sectional area near constant; then, as the aft fuselage taper, the job of maintaining areas is taken over by the tail surfaces.
The final important factor was not only the choice of an airfoil, but the precision with which it was executed. Hundreds of hours of painstaking work resulted in a wing on which irregularities are measured in thousandths of an inch and on which the airflow remains laminar -- flowing smoothly along in a sheet rather than tumbling in microscopic whorls and ripples -- back to an unprecedented 70 percent of the total chord.
Overall, the result of all these design decisions came out looking a bit like a cross between the longnose Ta-152 version of the Focke-Wulf 190 ahead of the wing, and an F4U Corsair from the canopy back, with elements of 1930s Bendix Trophy racer thrown in.
There's an old adage that says, "An airplane that looks good flies well," and I was to find that the flying qualities of the AR-5 certainly bear this out. At the same time, the airplane has a few idiosyncracies...which isn't unexpected in something that's really intended to be flown by only one person thoughout its career.
To get aboard, for example, you first put a towel down on the wing to avoid marring its pristine surface. Mike had to advise me exactly which (unmarked) postcard-size area of the wing had the additional stiffening to take my weight. Once in the comfortable fixed seat, I was presented with very basic instruments (airspeed, altimeter, tach, coolant temperature and EGT) and a handheld radio in a convenient little clip. There's a good-sized center stick; under the panel, it has an adjustable bobweight -- not to increase pitch forces or stability but just to prevent the weight of the curved stick itself from contributing to a G-induced pitchup. Somewhere up ahead under the float-gauge-equipped fuel tank are the rudder pedals and toe brakes; a two-notch flap handle is on the left sidewall. Pitch forces at any airspeed are light enough that elevator trim is neither needed, nor installed.
With the four-point harness fastened and the canopy locked, a flick of the ignition switch and a jab at the electric starter brought the Rotax to its usual slightly rough, blue-smoking idle at 3600 rpm to prevent gearbox chatter (the prop turns at half the engine), but that power was more than sufficient to trundle the AR-5 along at a brisk rate.
Direct tailwheel steering made it easy to fishtail for forward visibility; in fact, with that long , round cowl out ahead, the view is quite P-51-like on the ground. With the open tailcone right behind the pilot's ear and that tiny solid tailwheel rolling along Nut Tree's slightly rough asphalt, the effect was exactly like that of the old RCA Victor gramophone logo, with me playing the role of Nipper the dog.
Startup and taxi also brought home the AR-5's most significant idiosyncrasy, at least from the piloting standpoint: The throttle works backward! I'm not sure whether Mike did this to make things simpler or just to save weight (not that one can be profligate with the ounces -- the airplane weighs 488 lbs empty, and Mike weighs around 145, leaving only 27 lbs for fuel on world record attempts.)) But since the slides in the twin Bing carburetors on the Rotax are pulled out to open the throttle, a simple pull cable attached to a vernier throttle on the panel handles the chore instead of an elaborate linkage.
I found that as long as I used the vernier feature and remembered the old Spitfire pilot's expression of "opening the tap" for more power, things worked fine; but when I pushed in the center button and moved the throttle normally, I tended to get crossed up.
With no elevator trim and only a single "either it runs or it doesn't" switch for the Rotax's dual ignition systems, there isn't much to check before takeoff, so I line up on Nut Tree's runway 20 and opened the tap all the way. With only about 40 of the 65 hp available initially due to the aggressively pitched prop, acceleration isn't exactly neck-snapping at first, but the AR-5 moved out quite smartly nonetheless, and as advised by Mike, I raised the tail at about 45 mph. I expected a swing to the right (the prop turns "the wrong way"), but it was so minor, it was almost imperceptible. As soon as the tail comes up, the visibility is excellent, and the airplane leaves the ground very comfortably at about 60 mph.
Mike had warned me of another slightly odd effect: Since the prop is pitched for high-speed flight and turns pretty fast, it's at least partly stalled during the initial part of the takeoff. As speed increases and the prop finally "hooks on" to the air, the engine seems to sag momentarily; even so, the airplane continued to accelerate.
Mike had advised a climb speed of around 100 mph, at which the airplane climbed at better than 1000 fpm with excellent visibility. The air was quite bumpy, and the light wing loading made itself felt; on the other hand, the combination of good damping and very powerful controls made it easy to compensate for the gusts.
Pushing over in smoother air at 4500 feet, I let the AR-5 wind up toward cruise speed, and discovered more evidence of Mike's craftsmanship. Good laminar airfoils have a particular range of angles of attack over which they attain especially low drag (the so-called "bucket,"), and as the ship accelerated, I could feel it drop into the bucket very perceptively. This also let the engine wind up toward peak power; I throttled back...er, forward -- anyway, I "screwed in the tap" until things settled at 5500 rpm and an indicated cruise of about 160 mph, then I started feeling out the handling.
It was, in a word, delightful. Pitch forces are light, but not overly so for this class of aircraft; stick-free stability appeared just about neutral, perhaps just a hair divergent. At any rate, the phugoid period (if it even has one with me aboard) seems so long that I gave up trying to time it after about a minute. However, I weigh about 25 lbs more than Mike does, and after running the numbers a day or two after my flight, Mike noted to his embarrassed surprise that I was flying at a CG significantly farther aft than he ever had, despite some lead ballast on the engine mounts.
Roll rate is good (estimated at around 2.65 seconds for a full 360 degree roll), and roll acceleration -- the initial start of a roll out of level flight, which really gives the subjective impression of roll rate -- is even better. My only cavil there is that while both the ailerons and the elevator get somewhat heavier with increased airspeed, the ailerons "heavy up" slightly faster. Control harmony is well-nigh perfect at lower speeds, but pitch is a bit sensitive at higher ones. The rudder is very light and powerful throughout -- in fact, I probably wouldn't have minded if it were a bit heavier -- but there's so little adverse yaw from the ailerons that most evolutions, including aileron rolls and the dogfighting that Mike loves, can be executed "feet on the floor". I assayed several rolls, including one gingerly four-pointer (going negative halfway around would have stopped the engine, with its flat-type carburetors), and found that I could reach comfortably brisk roll rates without ever coming near the full travel of the stick.
I was also impressed, particularly in so light an aircraft, by how well it could maintain energy under increased G loads. "Windup" turns at full power could be held until the outside world started going a bit gray around the edges (for me, somewhere between 4 and 5 G) before the speed started falling. On the other hand, one can certainly tell when one finally leaves the laminar low-drag "bucket". Get below about 110 mph while pulling G, and the AR-5 feels like it's run into a wall of feathers. Of course, the "peakiness" of the two stroke's power curve contributes to this as well. As the airplane slows, the engine also drops out of its best-power range, and it takes what seems to be a long time to get things to wake up again.
Things are equally nice at the low end of the speed range. The airplane can be carefully worked into a stall (at 56 mph clean and 53 dirty) with almost no perceptible break, and it can be held level with ailerons and rudder while the sink rate builds up. Pulled up more briskly, particularly with full flaps, it tends to drop the left wing, although I was able to pick it up with rudder by about 45 degrees. Mike has spun it a half turn and reports no indication of developing problems...but I'll leave that to him.
I made a long, gradual descent back toward the Nut Tree. Mike had warned me that the Rotax tended to go lean at high airspeeds and intermediate throttle settings, requiring either more power (and higher speeds than I cared for in the rough air) or less airspeed. Arriving overhead for a high, 360 degree approach, I closed the throttle, waited what seemed like forever to get down to the 85 mph speed for the first notch of flap, and added the second on base leg. With full flap, the AR-5 comes down very nicely, with the nose down out of the field of view.
Some sink on short final called for a little power. Crossing the threshold at 80 mph, I did what I'd planned before the flight: I closed the throttle all the way, then immediately stuffed my left hand under my thigh to preclude my doing anything stupid with it.
As it turns out, it was no problem. Mike had advised me to make a wheel landing, since the three-pointer tends to scrape the aft ends of the wheelpants. Forward visibility was excellent all the way to touchdown, and the spring shocks in the gear legs soaked up the actual moment of truth to the point where I barely felt it. Despite gusts, the very powerful rudder and wide stance of the main gear made it a no-brainer to keep things straight until I could lower the tail, then pike it down with full aft stick for maximum tailwheel steering effectiveness.
So, that was my flight in a world-record airplane. I hope I get the chance for some more time in it. Mike certainly achieved his goal, whether primary or secondary, of a very small, high-performance airplane that's also loads of fun to fly. This is not one of those all-out machines, like an America's Cup yacht, that has to be kept in a velvet-lined case when not in use. In fact, with a cruise range of about 500 nm, it would make a neat little traveling machine; I only hope Mike can be persuaded to take it to Oshkosh this year. I'd be happy to ferry it there for him if he can't spare the time.
Peter Lert, Air Progress Magazine, July 1996