The Wing's the Thing

THE MARSKE PIONEER I - BY BILL DANIELS
Adapted form a 5/69 Soaring Magazine article.

Historically, flying wings have long sounded their siren call of low in weight, simplicity, and potentially, fabulous performance, but in the late 1960's, they had rarely if ever performed up to the expectations of their advocates. Perhaps a reason for this is that no one had ever really taken the time and effort to develop the necessary technology to convert the basic concept into a practical, safe and successful sailplane.

A prime difficulty in the past had been to devise a control system that coordinated well with the craft's natural stability characteristics. Jim Marske accepted the challenge and set about the long difficult task of developing a line of sailplanes that would first investigate and define the problems associated with the flying wing, and then solve them in a way that would prove the inherent superiority of the concept. The Pioneer is the second Marske tailless sailplane to fly, and incorporates the knowledge gained with the XM-l flying plank. Both the XM-l and the Pioneer I were low aspect ratio, low wing loading gliders, designed not for ultimate performance, but as proof-of-concept designs incorporating the maximum safety of the test pilot through very low take-off and landing velocities.

Additionally the Pioneer was to have been an outstanding performer under extremely weak soaring conditions, insuring that we could log significant time in these test machines.

Specifically, the Pioneer IA has a span of 46 feet (originally 40), an aspect ratio of 10.8, A 12.5 Ft. fuselage and a wing loading of 3.5 pounds Per sq. ft.

The controls were elevators hinged to the inboard trailing edge of the wing, conventional rudder/fin mounted on the centerline (originally, an all-moving rudder without fin), and spoilers for roll control.

The use of spoilers for roll control is rather rare and deserves special mention. They act by decreasing the lift and increasing the drag on the wing toward which the turn is to be made; thus, unlike ailerons, they have no adverse yaw. It was hoped that this would forever overcome the adverse yaw demon inherent in ailerons at high angles of attack.

Pioneer I Flight Tests

The initial flight tests were designed to investigate the ship's handling characteristics in the least hazardous way possible. We chose the El Mirage dry lake for this because we could make long auto tows and long straight glides without worrying about obstacles.

It was our good fortune in March of 1968 to have perfect weather for the flight tests. The air was cold, dense, and a turbulence-free wind was blowing straight down the lake at a steady 20 knots. This meant a ground speed of no more than 15 knots for the initial flights. Our good luck was to be prophetic. In eight flights that day we progressed from low, straight glides to 1000-foot altitudes and shallow S-turns without encountering anything but minor problems.

One problem was a slower roll rate than had hoped for, and the other was that the all moving rudder would oscillate at about 1Hz when free. The rudder could, however, be controlled easily so long as the pilot kept his feet on the pedals.

Pitch control, probably the most controversial aspect of flying wings, proved to be the most docile I have encountered on any machine. The most apparent thing is the lack of inertia, about the pitch axis, allowing the pitch changes to be made almost instantly without any tendency to overshoot the new attitude. A good analogy is the rack and pinion steering gear to be found on racecars. At first it feels sensitive but, when you get used to it, everything else feels sluggish. I came to like the Pioneer's pitch response very much.

The next day provided the first opportunity to try the Pioneer at Soaring. An auto tow to 1500 feet put me into a small thermal. I circled cautiously, working the lift as gently as I could. I wanted more altitude before trying the ship at steep bank angles. The climb took me to 11,500 feet ASL, (9500 above terrain), and provided the opportunity to systematically investigate the machine's handling with the safety of great height. (If I were to encounter severe difficulties, I was preferred to do so at an altitude where a safe bailout was possible.) Throughout the climb the Pioneer seemed to eagerly seek the core of the thermal almost without help. Even in this early test flight, the flying wing was endearing itself to me. It behaved in a way that inspired confidence.

One of the first things I wanted to know was how it stalled. Trying a gentle stall straight ahead, I found the stick reached the rear stop before a break occurred. With the stick held back, the ship flew straight ahead with no tendency to drop a wing. A pronounced low frequency buffet came from somewhere to the rear. Trying again with the nose well above the horizon resulted in a gentle break with surprisingly little loss of altitude, again no wing drop. Stalls from turns were no less docile. The ship maintained its bank angle and recovered flying sped immediately. It was becoming clear that while the machine could be momentarily stalled, it could not be held in a stalled condition.

This is apparently due to the fact that the elevator is part of the wing and is affected by the flow separation that starts at the trailing edge as the glider approaches a stall. As this flow becomes turbulent and separates, the up elevator authority is sharply limited. It is therefore not possible to increase the angle of attack to the stall. The effect is to limit the minimum airspeed to a safe value.

An objection is sometimes raised to this saying that if the up elevator authority is limited to prevent a stall, the landing speed will be too high. This would be true in a conventional glider.

However, the flying wing has another trick in its bag. During the flare before touchdown, the airflow under the wing produces a low-pressure area under the elevator. This is due to the venturi-like shape of the area confined by the under surface of the wing with a raised elevator and the ground. This low-pressure area occurs only when flying in ground effect and greatly assists the flare so the flying wing can be landed a very slow airspeed.

For the next hour and a half, I subjected the Pioneer to every test I could think of that might show up bad habits and as near as I could tell it didn't have any. I was able to determine that it was stable about all three axes, that it responded logically to all control inputs, that it would not spin however provoked, and that it retained its pitch stability on speed runs to100 mph.

The stick-free stability was remarkably similar to most other sailplanes. If the stick were released in straight and level flight at the trimmed airspeed, the ship would continue straight ahead for up to a minute, then gently enter a spiral. If the controls were released at a speed other than the trimmed speed, the ship would gently oscillate in pitch about the trimmed speed, with a 17-second period. This would damp out in three or four oscillations in smooth air. The oscillation was never noticed unless the stick was free. If the ship were placed in approximately a 30-degree yaw angle relative to the direction of flight and the controls released, the nose would swing back into line within three seconds, overshooting slightly.

A few points were not to liking, however, as might be expected from an experimental aircraft during its early flights. For one thing, while the roll spoilers' alone produced coordinated turns at 60 Knots, they tended to produce skidding turns above that speed. The effect was gentle, but noticeable, and required rudder input to keep the yaw string centered. It felt quite odd to need opposite rudder to make a coordinated roll into a turn at high speed.

In slow turns it showed the normal glider tendency to overbank and opposite spoiler was necessary to stop it. This was disturbing because it resulted in loss of lift when it was needed most. We were starting to re-think this roll spoiler thing. In spite of circling in thermals with one spoiler open, the ship seemed to climb extremely well.

On the plus side it was found (a surprise at the time) that the ship was taking less of a pounding during high speed runs than might be expected for a light wing loading machine flying in strong desert turbulence. The panel-mounted accelerometer showed a maximum of +2g, -1/2g for the 100mph runs, although pitch attitude changes were noted in turbulence. It seemed that the nose went up in down gusts, and up in down gusts, reducing the wing's angle of attack relative to the local flow quickly enough to prevent the machine from absorbing the load. This is no doubt due to the low moment of inertia about the pitch axes.

On several occasions, while flying in formation with other sailplanes, they broke off saying it was too turbulent to be going so fast. I felt things were fine.

On the slow end of the airspeed envelope, the Pioneer proved to be a surprising performer. I don't really know how slow the Pioneer 1 would fly because at the lowest speeds the airspeed needle rested on the peg and refused to indicate! Investigations of the airspeed indicating system failed to turn up a problem, so we assumed the speed was below the instrument's threshold. This assumption was born out in formation flights with other aircraft to the extent that they could not fly as slow as the Pioneer. This slow flight ability proved a tremendous asset in small, weak thermals, where it allowed extremely small diameter turns to be made.

From May to August of 1968, the Pioneer was modified to correct some of problems found in the original configuration. We also took this opportunity to lengthen the wings from 40 to 46 feet. This brought the wing area to 192-sq. ft.

Considerable effort was made to reduce the moment of inertia about all axes. This resulted in an overall weight reduction of 10 pounds to 430 empty even with the longer wings. A new airspeed was installed that would indicate down to 10 knots. As a result of these modifications, the ship is now called the Pioneer IA.

Pioneer IA Flight Tests and Modifications

First flight tests of the modified version began in August 1968. The handling qualities were generally the same as the 40-foot span Pioneer, with the exception that the rudder was stable and surprisingly effective considering the short moment arm. However, the roll rate was still disappointingly slow even though the spoilers had been moved forward and outboard.

After some thought we decided that the roll rate might be improved if air were allowed to escape (in a "puff") from within the wing as the spoiler was opened. We thought this would "trip" the flow quicker. We took care to use gaskets to seal the spoiler box completely when it was closed.

This small modification almost doubled the roll rate, cutting the time required to perform a coordinated roll from a 45-degree turn in one direction to a coordinated 45-degree turn the other way from 9 seconds to just under 5 seconds, measured at an airspeed of 50 knots.

The performance gain with the longer wings was quite noticeable and seemed to be everything we had hoped for. It should be pointed out that flying wings should be compared to conventional craft of similar span and aspect ratio, thus the Pioneer IA should not be expected to outperform the current generation of sophisticated fiberglass machines. Logically, the Pioneer IA could be expected to be slightly superior to a 1-26.

Flight comparisons seemed to suggest that we achieved something just under 30:1 L/D at 50 Kts combined with very good climb performance. With the new ASI, the minimum airspeed was found to be an indicated 28 knots. At this speed the sink rate was not significantly more than at the minimum sink speed of 40 knots. Soaring flight has been achieved on all except one attempt, with an average time per flight of just less than two hours. Most of these flights were made from auto tows.

The over banking tendency noted in the early tests seemed to be less noticeable in the Pioneer IA. This was counterintuitive so it was apparent that we didn't really understand the complete mechanism of over banking.

Glide path control on the Pioneer 1A was provided by dive brake/flaps hinged at the mid chord point on the underside of the wing. These acted to increase drag and to increase the coefficient of lift. At the speeds used for approach and landing, trim changes are negligible. At higher speeds there was a noticeable nose-down pitching moment.

We often heard from other pilots that the reflexed trailing edge of the wing looked like flaps in a negative setting. They speculated that this meant that the flying wing suffered from a low Coefficient of Lift. I took the outstanding climb performance of the Pioneer as refutation of this.

The only proper way to evaluate the success of a design is relative to the goals set for the design at the outset. In this respect I would judge the Pioneer to be a great success. It has proven to be easily soarable and has thus provided a great deal of information that could only be obtained from a prototype aircraft.

The Pioneer 1A flew on for several more years providing a set of data that will be of in-estimable value in the design of really "super" sailplanes.

Jim felt that we had accomplished a lot and wondered why no one else had taken this route. Why hadn't this been done before?

Prolog:

Looking back thirty years to those exciting days test flying the early Marske flying wings I find that I have lost none of the excitement and enthusiasm I felt then. However, I had hoped that the Pioneer 1 would be followed by a series of higher and higher performance racing flying wings. Other things in life just got in the way.

This is not to say that I am disappointed that Jim took the Monarch route. The Monarch is an exciting glider in its own right. The Genesis is a step along the path I had hoped for and the P3 and especially the P4 are exactly what I had hoped for so long.

Reading my May 1968 Soaring article again, I see that I might have understated the sheer joy of flying the Pioneer. It is hard to convey the sense of security engendered by its' stability and performance characteristics.

After moving to Colorado in 1969, I flew the P1A on several long cross country flights. The ability to use very weak lift and, if that failed, land almost anywhere was wonderful. On one flight, I reached 32,000' in the Pikes Peak wave. In another I flew almost 200 miles over mountain terrain that would have been very frightening in an expensive glass ship that required a large field for out landing.

In this time I allowed almost twenty pilots to fly the P1A and their comments have stuck with me. One particular comment haunted me for a long time. The succinct words were "As ailerons, roll spoilers make a great rudder." Spoilers do indeed make a great rudder.

Over the years I have thought a lot about the control system issues of the P1. It would be easy to say that the roll spoiler feature was a failure. In reality, we failed utterly to understand the relationship of control dynamics and circling flight.

Over the years, as a flight instructor, I have watched hundreds of students struggle to make sense of adverse yaw and coordinated flight. I have found that it helps to explain that the control responses produced by ailerons and rudder individually are not aligned with the longitudinal and vertical axes. At thermaling speeds, the ailerons rotate the glider about an axis that is inclined upward from the longitudinal about 45 degrees. The rudder rotates the glider about an axis that is inclined backwards from the vertical about 45 degrees. As long as these axes intersect at about 90 degrees, coordinated flight is possible, if not exactly easy.

The problem with the P1A is that the rudder and roll spoiler response axes didn't meet at anything close to 90 degrees. In fact, they were almost parallel. In effect, the P1A had two rudder systems and no ailerons!

In circling flight, there are other insights. Flight in a turn is asymmetric. In a turn, the inside wing travels a shorter distance at a lower speed than the outside wing. Without control input, the inside wing produces less lift and less drag than the outside wing so the glider will try to increase its bank and yaw away from the turn. A glider pilot will almost instinctively hold steady outside aileron to keep the bank at the desired angle and keep the yaw string centered.

This produces and interesting control dynamic. In effect, the pilot, by using the ailerons this way has made the glider asymmetric in a way that exactly meets the aerodynamic requirements of a turn. While holding outside stick, the inside aileron is lowered to increase the lift and increase the drag of that wing while the outside aileron is raised to decrease the lift and decrease the drag of that wing. The result is that the span wise distribution of lift and drag is restored to symmetry and the glider turns nicely with the rudder in an almost streamlined position. Gliders with great handling have ailerons that produce exactly the right roll and adverse yaw to produce this effect.

The rudder is still needed, of course, to handle turbulence induced yaw and changes in bank angle.

What does all this mean for flying wing control systems? Well, if you think of wingspans of 20 meters of so, the centerline conventional rudder will probably be inadequate unless it is very large or placed well aft. Neither seems practical or particularly inviting to me.
Considering the comment about spoilers making a great rudder, wouldn't it make sense to use that fact for yaw control?

Spoilers operated by pedal produce a rotation of the glider about an axis that is inclined backward from the vertical just like a rudder does. Combined with ailerons, coordinated flight feels exactly like a conventional glider with a tail.

I had the chance to experience this while flying a P2 that was modified by Jay Johnson in McCook Nebraska. This P2 had the conventional rudder replaced with spoilers operated by the pedals. A fixed fin replaced the rudder. Slow, tight turns required almost no spoiler and the combined effect of ailerons and spoilers produced a very fast roll rate of around 3 seconds in a 45 to 45 degree roll. I tried this at high and slow speeds and it felt absolutely natural.

I am convinced that this is the way to go for large span flying wings. The neat thing is that the moment arms of the ailerons and spoilers increase in proportion to the wing span so the system should scale beautifully. (Jim, how about a 30 meter P5?)