Bob Coolbaugh’s 1911 Curtiss Pusher Summary Flight Test Report
By Bob Coolbaugh
October 13, 2010 — After 15 flights and some major flight control modifications, we now feel we have a plane safe to fly – not predictable and controllable, but at least safe to fly. Andrew King and I have swapped flying and he may have a different perspective than I about this bird’s unique handling properties. Still, we both agree that this is a totally different animal from anything either of us has flown.
In broad terms, I knew when I first pushed the throttles up that I could be in for a painful experience at worst (the Hammondsport Curtiss Pusher crashed on its maiden flight 6 months ago) or, at best, Mr. Toad’s Wild Ride. Some clichés about test flying ran in a loop through my mind the night before. Some still lurk in my dreams. The thing that will kill you is that you do not know what you don’t know. Flying on the razor’s edge. Flight at the edge of the envelope. There are no old, bold Naval Aviators. Grampa Pettibone’s admonition – Jumpin’Gehosaphat! What was that old man thinking!! Terminal velocity can be reached at a speed considerably slower than a Saturn 5 Rocket. Also, the unequivocal statement by several “experts” that the Curtiss will never fly. One even went so far as to warn us that we’ll kill ourselves in it.
Well, I’m here to tell you that all those things have elements of truth, even though it is difficult to envision in a boxkite that only travels at 60 mph. Hard to conceptualize flight at the razors edge if you’re supersonic at the edge of space.
Weight and balance only proved that the CG and MAC were totally out of line with modern FAA concepts of 22-34 % Mac. We added nose weight to get to 40% empty and 34-38% with pilot. The highly under-cambered wing and the double pitch inputs of canard and elevators, the variations in outrigger boom lengths and a near total lack of historical data left us “best guessing” our operational CG. Thank goodness it was manageable, though still requiring down elevator pitch attitude for the lighter pilot (me).
Andrew has some forward elevator in level flight, but he effectively shifts CG toward center MAC with weight ( the size does matter concept). Note that I do not mention that we are holding push pressure or talk about the pounds of push force required. There is no force! In fact, there is no feedback. You use arm strength and coordination to constantly figure out what the new neutral point is. In the slightest turbulence, even this can become a chore. The nose will hunt with the slightest provocation. You dampen the oscillations by constantly finessing the input until it is corralled. Then you do something totally unthinkable like attempt a turn, and the nose takes off on another fun excursion. Needless to say, we have to work on this, while still staying true to the look and feel of an original Curtiss Pusher. Sure wish Glenn was still around to talk to, that’s for sure!
Take-offs are very quick at about 300-400 feet of roll. At 970 pounds empty and with the 125 HP continental and Sterba 74 x 40 prop, we have a beautiful combination of power prop and weight. The long wing with the added 30” cells as in the Ely-USS Pennsylvania airplane, has plenty of lift and climb performance is better than expected. Without an airspeed indicator, you have to feel when the plane is ready to fly and then determine an elevator setting which will ease the nose off. The main wheels are well behind the leading edge, meaning you have to raise a lot of weight a long moment arm to get a climb started.
Once the nose comes off the ground, though, you’d better be ready to reset the pitch, as it climbs in a pretty flat attitude in ground effect. We had wondered at all the drag caused by the 130 flying, landing, brace and strut wires hanging all over the plane. We have proven that idle power in flight requires prompt attention to a very pronounced nose down descent attitude. The plane will stop flying if you close the throttle and hold the nose up. Images of an F-4J underpowered and coming down at the ramp come to mind.
That and a safe falling from the sky. The unsettling thing about the take-off is the pitch hunting, which we are learning to anticipate and dampen with experience. It is still going to happen, we are just getting better at reining it in. Then there is the transition from ground effect, where the plane is pretty solid to real air, where it has to fly at the capriciousness of the winds. We are flying in generally calm conditions, but there are still wind gusts and thermals to contend with once airborne. This causes upsets. For pitch problems, see the paragraph above.
Aileron control began as what I would call unmanageable simply because I did not have enough arm strength to muscle the control wheel against the air forces over the very large ailerons. There is no balance to the ailerons. In fact, they are hung 1.5 inches AFT of the hinge line in order to clear the rear bay’s landing and flying wires. At 8 feet long and 2 feet wide, there is a lot of surface stuck like glue in trail once you get 50-60 mph flowing over them. At 45-50, I found I could at least get back to the runway.
After the first 12 flights, six of which were mine, I was ready to put the thing up on a pylon at the entrance to the airport. Andrew King had a sleepless night of pondering and came up with a plan that proved elegant in its simplicity and which has totally changed the aileron forces. If you ever lose your mind and try to build one of these things, I’ll go over Andrew’s design with you. We were still able to keep the look and feel of the original Curtiss control set-up, so the airshow audiences will still be amazed and awestruck.
Back to the root cause of all this aileron whining. Curtiss used a shoulder yoke hinged off the seat frame to control the ailerons. The shoulder yoke was on about a 20 inch lever arm off the seat frame. The aileron control cables were attached to the back of the yoke, behind your back. Lean left and the ailerons turned left. Lean right to turn right. The pilot had all his upper body strength and the 20 inch lever arm to muscle the ailerons around. Innocently enough, I decided to run conventional aileron controls via the control wheel and rudder controls via rudder pedals. This modification to modern convention was because we planned on cross country flying from the beginning and we knew of a few accidents caused by modern pilots flying with ancient control inputs.
When things go wrong, we wanted our years of ingrained reflexes to react predictably to counter upsets without having to first rationalize aircraft handling differences and then applying the fix. In this case the aircraft designer (me) created a problem that the pilot (me) had not anticipated which caused some palpitations on the first flights. I think Andrew felt the same. Routing the aileron cables over a series of pulleys, up the inside of the control stick, over more pulleys and around the rim of the 15” control wheel gave us the look of the original Curtiss.
What it didn’t give us was the leverage to move the ailerons in flight. With a real grunt I could move the wheel about 10 degrees. If the wing got tossed over in turbulence, correcting to level flight became “wait-and-see, ease the nose over, opposite rudder if it works, hang on a hope everything comes out OK”. That’s the Mr. Toad’s Wild ride experience. It was especially difficult in a right wing down recovery. Remember that the pitch demon was hard at work all this time, too, adding to the fun and excitement. In effect, the control wheel was the last pulley in the loop and it had a 7.5” radius of effective leverage cranking around to overcome wind resistance over two unbalanced 16 square foot slabs and 55 mph.
Drawing from his experience banks, Andrew designed a set-up with a new control wheel and a 5” pulley, giving us an effective 3 to 1 ratio improvement over the 15” wheel pulley. The down side was that it now takes 3 times the wheel turn to get the same, earlier throw, but the roll muscle required is about one third less. I will gladly sacrifice the ability to smartly roll into a 60 degree bank for the simple pleasure of knowing I can make a 15 degree bank turn with some semblance of control. We now have a night to day difference in aileron control.
Rudder control turned out to be a little stiff but totally predictable. As you can imagine, adverse yaw is a real factor. The large ailerons overhanging the ends of a very long wing, the lack of a fuselage for “wetted area” and a small rudder all contribute to adverse yaw tendencies. Rudder first, then add in ailerons is the secret to controlling yaw and making reasonable turns. The plane will fly crooked all day long if you let it. We use a yaw string of yarn attached to the front canard post as a guide. That part is still not intuitive. If the string is trailing off to the right, you have to step on the left rudder to align, unlike stepping on the ball in an inclinometer.
When you’re more concerned with where the nose is having its fun, the aileron thing and adverse yaw, stepping on the nose of the yaw string is not a natural reflex yet. Rudder cable control originally was via cables wrapped around the control wheel, through fairleads to a large rudder horn attached ahead of the rudder hinge. In order to run conventional rudder pedals to this rudder horn we had to cross the cable via pulleys to correct the direction. This led to a nightmare of crossing wires in an already crowded bay aft of the wing. A call to Karl Ericsson at Owl’s Head Transportation Museum solved the problem. Their Headless Curtiss replica has rudder pedals and the cable routing is to a rudder horn attached behind the hinge line as in “modern” Piper Cubs. No more crossed wires and multi-pulley jumbles. A very small rudder horn was all that was needed. Having conventional pedals allowed another concession to safety, Cleveland brakes.
The landing approach is flown power on, in a long descending straight away. Particular attention has to placed on dampening the pitch excursions or you are definitely not going to make a safe landing. Fly a long final and sort all the variables before you get to 5 feet. I have found a descent rate at 45-50 mph and reduced power controllable. I target a spot short of the touchdown point and add power to fly along the ground, easing forward without introducing pitch oscillations, until I can roll on nearly flat. Then, power off and a very short roll out on grass. Roll out on pavement is not much longer. As long as you are not messing with ailerons and pitch, the rudder is very effective at tracking straight in the landing “flare”.
Speaking of flare - this is not a recommended procedure. Consider that the mains are very far back and you are heavy and ahead of the wing. When the mains touch, there is not enough elevator/canard to hold the nose up while you ease it to the runway. When the mains hit the nose will plunk down smartly. This will bend stuff because there is no suspension other than the rubber in the tires. If the nose comes falling over and you yank the stick to stop it, you are in danger of getting airborne at a very slow speed or, at best, dragging the rudder on the ground. I have gotten underpowered in the flare and you had better believe the speed pays of too rapidly to cushion with a late burst of power.
Enough of this. If you want out of this rehashing, please let me know. I will not be offended. This is just such a different flying experience I wanted to share it with pilot friends I am pretty confident will not try this at home! Best to you – blue skies and calm air, please.