A CANARROWING EXPERIENCE
by Dan Fritz
As many of you know, I have been experimenting with a 2 meter canard for the last two seasons. Since we are trying to rekindle participation in the S.O.A.R. Achievement Award Program, I am submitting it for consideration in the Annual Design Award category.
The airplane is called a “Canarrow,” for obvious reasons. Unfortunately, there is no surviving hardware for display. I folded the original proof-of-concept prototype on a winch in 2005; the more refined second prototype was shot down by a fellow club member while high and downwind (we never found the remains despite hours of searching). Accordingly, several photos and a full building plan have been provided in the “Photos” section of the web site. Despite these problems, the design has been quite successful. The airplane is stable and easy to fly. It launches and flies conventionally. I used it to win the 2 meter class in this year’s McIntyre Series. I took it to one out of state contest where it finished sixth overall in RES, despite being a 2 meter. I encouraged everyone to fly the second prototype. Richard did inverted thermal turns with it. Karl was flying for all of the flight photos. GordySoar the famous internet blogger took a couple of flights with it in Iowa. Some of the juniors and even Peggy Crosby got some stick time with it. No one had any trouble. The overall performance of the second prototype at least equaled that of a conventional built up 2 meter.
The design has its origins in a balsa stick glider I made in the early 1980s. Canards were in fashion then. I swept the forward wing to avoid landing damage. The swept canard looks like an arrow, hence the name. I was reminded of it again when I decided to build something for the inaugural McIntyre Series, which got me thinking about the classic 2 meter dilemma of wing area versus aspect ratio. I figured the best way to have open class figures for each was to start with an open class wing. The idea was to chop off the outer panels leaving a 2 meter center section and then attach the severed tips to the fuselage in a lifting configuration. The Canarrow is technically a true canard because pitch stability and control are provided by a forward lifting surface, but it is more correctly a tandem wing airplane. The CG is entirely between the two lifting surfaces. If one were to reassemble the forward wings to the tips of the main wing, the result would (excluding tip blocks) span 112 inches, offer 752 square inches of area and have an aspect ratio around 17. I realize things are not so simple, and drag is higher (there are twice as many tip vortices for one thing). Nevertheless, there is some validity to this concept. The design attempts to minimize its weaknesses where possible. The tip vortices from the forward wing flow over the top of the main wing at the dihedral breaks where some of their energy is recaptured. Sweeping the forward wing proved remarkably successful in protecting it; the forward wing was the only part to survive the crash of the first prototype, and I used it without modification on the second. The forward wing lies entirely behind two lines drawn from the nose to the tips of the main wings; assuming the craft lands on a flat surface, something else must break before the canard will.
I tried to keep the design easy to build. The fuselage is a semi-monocoque in that most of its strength comes from a thin (1/16 inch) skin, and she shape is maintained by fairly robust and closely spaced formers. However, the corner joints are heavily reinforced with triangle stock balsa that does contribute to the overall strength. Despite the high bending moments generated on landing because the mass of the main wing is so far from the nose, there were no fuselage failures on the second prototype even with contest landings. The longest single part is the fuselage side panel, which measures 48 inches and can be cut from a single sheet of balsa. The airfoils are all flat bottomed. Wing construction is simple open bay on the lower surface and a combination of sheeting and cap strips on top. I used a built up trailing edge, but solid stock balsa would work just as well. As the photos show, the original vertical stabilizer was too small. I miscalculated the CG point (too far forward) and, consequently, overestimated the directional stability provided by the main wing’s dihedral. That was a simple fix, but by the time I got it balanced right there was a ton of weight in the tail and the flying weight was 67 oz. The second prototype used a shallower fuselage to minimize side area ahead of the CG, which allowed for a slightly smaller vertical stabilizer and reduced weight. I rearranged the internal components with the battery and rudder servo in the tail. It balanced with almost no added weight. I must have built tanks in a former life, because it still came out at 45 oz. Most of the weight was in the main wing, which had lots of carbon; it could and did survive fairly aggressive winch launches and even zooms. A more weight conscious builder could easily get things down to the mid 30s. The only other change between the two prototypes was a slight reduction in dihedral on the second airplane, and that seemed to fly better.
In the air, the Canarrow does fly like a bigger airplane than a 2 meter. It ranges well and is quite buoyant. The biggest difference between the Canarrow and a conventional design is that pitch does not vary as much with changes in speed. Accordingly, it is important to pay attention to pitch trim since small changes will not result in the nose pitching up or down, but will simply change the flying speed. The deck angle of the fuselage stays remarkably flat, even while climbing in lift. Both prototypes had a single spoiler on the main wing. The first one was twelve inches long and too small. I lengthened the spoiler to eighteen inches on the second prototype and had excellent results. Unlike a conventional design, deploying the spoiler on a Canarrow causes the nose to pitch up sharply. The elevator on a canard moves in the opposite direction than on a conventional design (the elevator moves up for a nose down command and vice versa). Accordingly, the pitch compensation for spoiler deployment is to also spoil the forward wing with a “down” elevator command. A simple linear program mix was able to keep the fuselage level with any amount of spoiler deployment. The combination proved effective not only for coming down but also spot landings.
I have included a building plan with this article. It is a fairly small .pdf file, and everything is on one sheet. Kinko’s can print full sized plans for about fifteen dollars. Anyone who builds a Canarrow and balances it according to the plans can expect good performance from the first hand toss. The second prototype flew at exactly the same balance point as the first, thus confirming the repeatability of the result. I would be happy to provide building assistance to anyone interested.