Todd Ferrante's Estarter to Canard Conversion

by Todd Ferrante  <back to homepage>
last edited 9/26/2005

Background: My first RC plane was a GWS Estarter.  As I crashed and repaired it, eventually it got too heavy to fly.  I think towards the end the fuse was more epoxy than foam.  I got the idea to discard the fuselage and convert the wing and elevator into a canard style plane.  Fate intervened when I was presented with the parts to make the Canmosa.  After one too many crashes, I decided to return to the original Estarter canard plane conversion, with the intent of making a stable platform for aerial photography and video.

Design goals:
I had a few specific goals in mind for this plane:
- Slower and more stable for aerial photos and video.
- Not as susceptable to damage from nose-in landings.
- More pitch authority.
- Improved canard pivot design.

Estarter Canard, mark 1 (9/11/05):
I whipped this plane together in a couple of evenings and did the first test flights the next morning and evening.  In the morning it was a bit too windy to fly, but I just couldn't wait to try it out. Hand launching into the wind took some gathering of courage, but the plane worked great with almost zero trimming.  I did about 5 laps around the field and brought it in for a gentle landing.

Estarter canard, mark 1.

In the evening the wind never did die out, but I got in a nice 10 minute flight. The wind was less gusty than in the morning, but was still strong enough that I wasn't comfortable enough to really work the plane. Doing some rolls were about as aggressive as I got. It definitely rolls better than the estarter, but with the dihedral wing, they are nothing like the rolls I got from the Formosa wing. The longer length of the plane (full length arrow shafts instead of half shafts) gave it much better pitch authority than the Canmosa. Both flights I had to land across the football field instead of along it because of the wind direction. With the breeze blowing, the plane didn't want to come down and it took a couple go-rounds to keep from coming clear across the field and landing on the asphalt track. Both times the landings were nice and level, with the plane sliding nicely along the short grass on the pop bottle cowlings.

Bottom side of the Estarter Canard.

The first design choice was to use full length carbon arrow shafts for the body, and to space them a little further apart. The longer shafts put the canard pivot point 22 1/2" in front of the main wing leading edge, compared to 9 1/2" on the Canmosa. The intention was to improve pitch authority and to reduce the plane's sensitivity to the CG location. From the two test flights I have flown, this seems to be the case. With the increase in length, the spacing between the shafts also needed to be increased, to maintain torsional stiffness. Another benefit of the wider spacing is that the CVS digital video camera fits between them.

Moving on to the nose of the plane. On the Canmosa I used the elevator flaps for the canard and the elevator bodies for the winglets. This left the canard undersized and the winglets oversized. For this plane, I used the elevator main bodies for the canard. These are also more streamlined than the mostly flat canard on the Canmosa. The canard pivot tube only runs ~1/3 of the way into the canard bodies. On the Canmosa, a canard-strike, on landing, twice produced broken canard tubes. This required stripping the tube from the canard and installing a new one. On this plane, the tube is buried in the canard foils, and replacing it would be next to impossible. I'm hoping that rather than the tube breaking, the canard itself will break off. Repairing the foam canard with epoxy would be much easier than replacing a snapped tube.

One advantage of having a nose wheel is that it provides some measure of protection to the nose of the plane on nose-in landings. Unfortunately, it tends to snag in grass, producing jarring landings. Due to the size of the rear mounted prop, I never landed (or took off) the Canmosa on pavement. I decided to eliminate the snaggy nosewheel and it's servo. To provide some measure of nose protection, I fashioned a nosecone from foam and the top of a plastic pop bottle. The fuselage shafts extend past the canard and stick into this nosecone. The idea is that the nosecone will absorb any hard landing damage instead of the plane. Kind of like a foam and plastic car bumper. I'm still contemplating the addition of a simple skid under the nose, if the camera requires ground clearance.

Estarter elevators for canards.  Foam and plastic nose cone.

The machined plastic canard pivot fitting on the Canmosa was indestructable, but weighty. On this plane it is fabricated from 3/32" thick birch aircraft plywood. Rather than a dead weight, it's almost part of the plan area of the canard. A more efficient design by far. If I decide to get really crazy about weight savings, I could cut lightening holes in the tray and cover them over with clear packing tape. I could also put a foam skin over the top for smoother airflow. We'll see after a few more test flights. The canard pivot linkage is different as well. On the canmosa, there was a link arm glued into the canard foam. I broke several of these on nose-in landings, when the canard would hit the ground. Hopefully, the nosecone will prevent these canard strikes. Rather than the link arm in the canard foam, I made an arm from plywood and epoxied it to the canard pivot tube. There were times on the Canmosa when I suspected that mushy pitch authority was partially caused by the flexy pitch linkage. This provides a much stiffer linkage and hopefully better pitch control. Time will tell if eliminating the mechanical fuse (the plastic link arm) was a good idea. Given the choice between replacing the plastic link arm and stripped servo gears, the link arm is the better choice.

Plywood canard pivot / servo tray.

After the nose, I moved back to the wing. Having used the elevators for the canard, I needed something to use for winglets. I took the vertical stabilizer from the Estarter and split it down the middle with an exacto knife and razor saw. These were epoxied to the upper wingtips, with the flat side inboard. They are a little weak, being so thin, and may get reinforced later.

The roll rate of the estarter was very slow due to the undersized ailerons. They are good for learning, but not comparable to the Formosa. I had heard of people extending the estarter ailerons to increase roll authority, and decided to do the same. The foam came from the lids of some large take-out boxes from one of my favorite joints, Smokin' Joes BBQ. The pieces were cut as large as possible along the diagonal of the lid. The resulting shape looks pretty cool, I think.

Winglets from split vertical stabilizer.  Aileron extensions.

I pulled the motor mount stick from the old estarter nose and mounted it to the rear of the wing. The wing is recessed so the stick lies parallel to the carbon arrow shafts. The stick is bolted through the wing attachment hole and epoxied into the wing pocket.

Motor mount

The top mounted wing on the Estarter has the aileron servo and linkage mounted on it's bottom surface. Normally protected by the Estarter fuselage, this linkage needed protection if I were to land the canard plane on it's belly. I took another plastic drink bottle, trimmed it to cover the linkage, and epoxied it right to the underside of the wing. When landing on mown grass, the plane slides nicely to a stop on the nosecone and rear cowlings.

Aileron servo cover from plastic water bottle.

This completed the airframe, but I needed somewhere to mount the battery and receiver. I would like to fill the entire space between the carbon arrow shafts, from nose to wing, with a piece of blue foam insulation. But, not yet having any, I turned back to the remains of the take-out boxes. I cut out the bottoms, scored the panels so they would bend, and wrapped them around the carbon shaft tubes. Packing tape holds the wrap tight. The receiver, battery and ESC are velcroed to the top.

I used my handy-dandy CG calculator spreadsheet to find the theoretical CG location. The velcro allows the battery to be easily shifted around so the plane balances at the right spot. I wrap a couple rubber bands around the battery, just for extra security. I balanced the plane right at the theoretical CG, and it flew perfectly right from the start. I figure the added area up front from the canard servo tray about balances the added area in the rear from the airleron extensions.

Flight Testing: I did the first major flight testing 9/12/05 at Riverview Farm Park, in Newport News. It was my first time flying there, and I have to say it is a perfect place to fly. At one end of the park, there is a tall (40 ft?) mound where you can launch from. It's cool to actually look down on your plane as it flys around below your feet. Helps with the crick in the neck, too.
There was a light, steady breeze blowing toward the face of the hill. The Canmosa, with it's oversized winglets, would have been continually turning into this breeze. The estarter canard's winglets are large enough to control yaw, but small enough to handle some wind without continually "weathervaning".
Pitch control is much improved. The Canmosa would pitch down fine, but was mushy bringing the nose up. This plane is smooth and responsive either way. It was always a little nervewracking with the Canmosa to pull up from a steep dive. You were never quite sure if it would pull up in time. With the estarter canard, this doesn't seem to be a problem. No problems yet pulling up, but I'm still leery of really testing the limits. Slow speed, nose high, stall testing showed the classic "nose drop" stall of a canard plane. I was a bit worried about this, as this canard has a lower aspect ratio and a thicker foil section than is normal. I was worried that it wouldn't stall before the main wing, but apparently it does.
This plane loops easier than the Canmosa, but it is still necessary to build up some speed and apply max power over the top of the loop. Coming over the back side, the plane either pulls up level or pushes back into inverted flight smoothly. A straight down dive shows no tendancy to push off either way. This is in sharp contrast to the Canmosa, which had a quirk of wanting to push into inverted flight out of a vertical dive.
The aileron extensions on the estarter wing drastically improve the roll rate of this plane over the estarter. (I'm sure the absence of an under wing fuselage helps as well.) As expected, the dihedral in the wing gives rolls that are more "barrel-ey" than crisp. But, they are much closer to a Formosa roll (imperceptable drop) than a stock estarter roll (30 foot drop). I'd say the drop was more like 3-5 feet.
Inverted flight was no problem. In fact, it took less canard input to fly level inverted than right side up. (This actually showed that my balance isn't quite perfect.) The dihedral wing does require continual tweaking of roll to keep the plane inverted, but not terribly much so. Felt kind of like flying the Formosa on a breezy day...
I don't recall specifically testing straight vertical climb rate. I'll have to check this next time.
Endurance was great! I didn't push it to the absolute limit, and was doing aerobatics much of the time, and still got over 25 minutes flight time on a 3S, 1800 mAh battery.
After the first battery, I rubber banded the digital video camera on the plane to take some video. I must not have had the camera secured well enough to the plane, as the camera was onboard when I launched, and missing when I landed. I never even saw it depart the plane, so I had no idea where on the ground to look for it. I had cleverly (I thought) located the camera right on the CG of the plane, so I could fly the plane both with and without the camera onboard. Of course this meant that when the camera came off, the balance of the plane didn't change, so I had no hint it was gone. Good thing it was only a cheap $25 CVS camera. I guess I better start putting my name and number on my camera as well as on the plane itself.

Estarter Canard, mark 2 (9/15/05):
After the loss of the camera, I obtained a piece of scrap blue foam. I epoxied the foam between the fuselage tubes along the length of the plane, and used velcro and packing tape to mount the electronics and run the wires.

Estarter canard, mark 2.

The blue foam is a little heavier than the white foam it replaces, but gives a much sturdier platform for mounting. It should also provide some amount of aerodynamic lift during flight. I'll experiment with new camera mounting methods when I obtain a replacement camera.

Electronics mounting.

I did some test flying and found a new quirk in the flight behavior that wasn't there on earlier test flights. Since the only change I made was filling in the fuselage with foam, I'm assuming this is the cause. What happens is that the flat fuse foam is pushing the nose around. like a control surface. When at positive angles of attack, this just provides more lift, but if you bump the plane into a very shallow dive (more of a decent, really) the nose kicks down into an ever increasing dive. To recover requires a hard pull out, or pushing through to inverted and rolling right side up.  The net result is that it's hard to fly smoothly level. The nose is either kicking up to near canard stall, or kicking down into the scarey dive.

Estarter Canard, mark 3 (9/17/05):

A combination of cutting out the front fuselage foam, and moving the CG slightly forward was successful in eliminating the pitch stability observed yesterday.

Estarter canard, mark 3, with digital camera

I made up a bracket for holding a servo that would press the shutter button on my Kodak digital camera. It required programming a custom mix into my Optic 6 so the servo, which was plugged into the rudder channel of a 4-channel GWS receiver, would be triggered by the landing gear switch of the radio. The regular digital camera weighs a lot more than the little CVS camera. It is scary throwing a $300 digital camera into the air strapped to a plane you have cobbled together from spare parts... But, with full throttle, the brushless motor with a 10 inch prop had no problem getting the plane in the air. I could easily fly around at half throttle.

Front view of shutter trip servo bracket

Rear view of shutter trip servo bracket

With this setup I flew three successful AP sorties.  The first was at dawn at my local field. On this flight I learned two things: 1) low light slows the shutter speed on the camera down so the pictures are blurry; 2) landing in dew wet grass gets your expensive camera all wet.  The next day I did another couple sorties. The first was in the morning, after the grass had dried out and the sun was brighter. The second was later in the afternoon at the Yorktown Battlefield.  Since this page is about the plane itself, the aerial photos will go on another webpage.  Until then, check out the links to the first and second threads I posted with pics, on RCgroups...

Update 9/21/05:
The Estarter Canard is officially retired.  I took it out for another AP flight and for some reason had trouble gaining altitude.  Since the plane was below the tops of the trees, I had to try to bank and turn within the trees surrounding the track.  It didn't quite make a tight enough turn... The left wing caught a tree trunk and just destroyed the main wing.  I contemplated repairing the wing several times before giving it up as a lost cause.  I really liked how this plane looked, too.  The digital camera survived the crash, but I'm not sending it up on a plane again.  It's just too heavy. 

After a crash into the tree line.

Update 9/25/05:
I took the nose from the Estarter Canard and transplanted it onto the Canmosa.  See the AP canard plane for further details...