Extra Project
by Mike Hurley
Project explanetion
It has long been thought that the Extra 300 series of aerobatic aircraft are among the truest, most neutral flying rides available. In models, I think that fact is even more evident, with the Extra winning more national and international IMAC and TOC titles than all other aircraft combined. I’ve owned three small glow versions and I must concur. For all-out precision I’ve found nothing better.
A new era of aerobatics has emerged that pushes the physics envelope and raises the bar of flight performance beyond anything dreamed of ten years ago. "3D" flying with a scale aerobatic plane was introduced to the world by Quique Somenzini at the 1994 Tournament of Champions and the rest of us have been playing catch-up ever since. 3D performance requires a bit of modification to the size and amount of movement on all flight control surfaces. Inherently, the standard scale Extra does not excel in 3D flight and most of the offerings on the market are not modified to excel in this style of "unrecognized" aerobatics.
After spending literally months researching the market for an Extra that could deliver both precision flight for competition and advanced control surfaces for 3D, coupled with my personal desire for the largest possible aircraft that would still fit in my SUV.....I came up with nothing.
The only way I could have my dream plane was to design it and build it myself. So, with the help of friend and long time IMAC flier/designer, Norm Cassella, I decided to do just that! Norm’s been designing aircraft since dirt was created. He flew in the TOC in the early 70’s, has won the IMAC nationals in unlimited class several times, and is a past president of IMAC. I couldn’t have a better project partner. Ideally, my Extra would be close to 100" span and still be light enough to be powered by the most reliable and inexpensive gas engine available to us, The Zenoah G62. The plane must have a final weight of under 20 pounds. Norm has extensive experience with his own Laser with these exact proportions so moving his planform to the Extra should be a piece of cake.
I did the entire draft of the plane’s outline myself. I spent about 2 months researching the full scale 300L, IMAC’s rules and guidelines, and other models currently on the market. I did discover that several of the planes we now fly regularly in IMAC do not fall within the rules as written. Quite a surprise!
Our Extra will have a span of 97" and has a target weight of 19.5 pounds when used with an engine of 5.5 pounds. I have calculated the wing area to be about 1610, with a root of 21.5" and a tip cord of 12". The tail surfaces are airfoiled and retain a scale outline. The ailerons will optionally be full span and the hinge lines on the empanage have been drastically changed. I decided on starting with a cowl, canopy, gear, and pants from Ohio RC’s 30% (87") Extra 300. The parts fit in this project like they were custom made for it. The design was meant to utilize basic engineering principles with a combination of carefully chosen materials to create the most strength and stiffness at the lightest possible weight. We decided to make two prototypes using completely different engineering and construction techniques to test which one we liked better. The final outline on both planes will be identical. Norm’s first plane will have a G62, mine will have a BME 5.8 or equivalent. I will be building a third plane after our initial tests and any deemed improvements will be incorporated. I may try another engine at that time, but the real push is to get unlimited performance with the G62.
Many of you probably already know that 3D flight requires the ability of the aircraft to vector the engine’s thrust in such a manner as to enable precision flight without relying on the wings to keep the aircraft aloft. In other words the plane is flying in a completely stalled manner at all times while still maintaining control. To do this requires a thrust to weight ratio of about 1.5: 1 minimum, better would be 1.75 to 2: 1. The control surfaces must have the mass, throw, and strength to efficiently and forcefully change the direction of that thrust, and to hold the weight of the craft. My Extra is designed to optimize these characteristics without the need to change the CG from optimal precision flight settings. Because of this and the fact that the scale outline has not been altered, it should also retain all of the scale Extra’s stability when flown with regular control throws. Slight but legal changes have also been made to reduce pitch coupling. Roll coupling has not been a factor with the stock Extra design.
Both of our fuselages are done now except for the foam parts. All of the templates for the wings, empanage, and turtledecks are done. Both planes feel very rigid and the weights are right on target at 3 pounds with cowls and gear. We will start cutting foam in the next week or so. Look for updates as the project progresses.
002: The Box/stick construction with hatch and turtledeck. Rear of fuse uses boxed sticks and front lightply is also made rigid by boxing the sides. The main construction starts with the inside wall and is boxed to the outer dimension. This method is both light and very stiff.
003: The two construction methods side by side. The one on the right is a more conventional construction with the outside of the fuse being built over the plan. The motorbox extends to the wing tube and ties in with the landing gear plate. Hard landings will have force spread over a wide area and have to overcome compression to cause damage.
006: you can see the conventional lightply/stick framing on the right (turtle deck not done) as compared to the boxed structure on the left. Both are very light.
011: Norm Cassela checks the plan against the built fuse.
The Extra 300L Project has been coming along well.
I’m very happy with the
final proportions and the adherence to scale. The cowl fit was a little
tight with the BME100. Even with the small plug option I still had to open
up a relief for the tiny spark plugs. I choose well tested NACA
symmetrical airfoils that have a Mean Aerodynamic Center (MAC) of about
28%. I decided on a double dimension wing with a thinner tip and an
increasingly sharper leading edge toward the tip to improve snap
characteristics. I also changed the leading edge taper by moving the tip
forward 2 degrees for increased slow speed stability. I went with a 13% root
(thickness is 13% of the chord). At the tip it is thinner at 12%. Ailerons
size was increased slightly to 3 ¾” and brought to within 4 inches of the
fuse for better stalled flight (3D) control.
Although I stayed exactly to scale with the basic outline (other than some
relief of the elevators for the rudder) you can see that the empinage hinge
lines have been moved quite a bit. Both the rudder and elevators were double
beveled for increased throw… Yes, that is the scale size rudder! Norm is
starting to call me 3D-man, This tail should provide excellent authority to
perform even the most demanding freestyle maneuvers. At first I made an
empenage that fit the dimensions of the wing but after they were completed
it was obvious that all of the tail was too thick. I settled on the tail you
see here with a rudder of NACA0009 at the bottom and NACA0010 on top. The
horizontal stabs are 11% at the root and 10% at the tip. It’s hard to tell
in the photo but the entire empenage is removable. The wing was raised by a
half of an inch and the horizontal stab was lowered by ¾” for improved
neutral flight in knife edge, still well within IMAC 10% deviation rules.
The incidences are set to 0-0-0 with 2 degrees right for the engine. The
thrust line is in the scale location.
The projected weight outcome of 20 pounds is still on target, Assembled the
way you see it in the first photo, minus wheels, pants, radio gear, and
covering, it is under 16 pounds. The span is 97” with a final wing area of
1680 sq.in. Length is 84”. Wing loading at 20 lbs will be 27.4 oz/sq. ft.
even at 22lbs it would still have a very light loading of 30.17 and with the
BME100, power will not be a problem. The hatch and turtle deck are both
hollowed foam with a 1” thickness. When sheeted with the 6 lb. density 1/16”
balsa those structures become very rigid. The hatch has a 1/8” balsa base
and is very light. I still have a few details to finish before I start
covering but I’ll be back after it’s covered or after the first flights to
let you know how it turns out.
I now have about 20 flights on my prototype Extra 300L and Norm has about 6 flights on his. I can say that the planes are done and at this point, I don't see an end to the evolutionary process.
Before I get to my initial flight evaluation let me fill you in on final specifics.
I like to stick with one type of radio system and I use
JR. The trans is a 10SXII talking to a 950s receiver.
I made all of my extensions from custom electronics unassembled gold connectors using 22
gauge silicone covered 64 strand copper wire and each extension was twisted. JR HD charge
switches control the two 1700 mah AU 4.8 packs plugged directly into the receiver.
For servos I was lucky enough to get my hands on some of the first DS8411 digital servos.
These are JR's new flagship servos producing a whopping 155 oz. of torque and a fast 0.18
second 60 degree transit measured at 4.8 volts.
All of the control rods are less than 4 inches so I didn't
see a need to make up carbon fiber examples. Hangar Nine aluminum arms and Rocket City
hardware round out the lineup ensuring a no slop control system.
I had a hard time with the control mechanism for the rudder because the surface is so
thick at the control point causing interference if I tried to line it up with the hinge
line for the best geometry.
I punted and took the easy way out so I'm not getting anywhere near the throw that I had
planned for. I haven't actually measured in degrees.
Up front the ignition box is attached to a nifty hatch that is held in place with a 440 cap screw through each side of the motorbox. The plug wires are routed through guides that are also fastened with the same two screws. Take the plug caps off, remove the two cap screws and the whole unit comes off as a single piece. A 1400mah AU 4.8 pack powers the ignition system and is mounted just inside the motor box.
I made a rod for the choke arm from carbon tube that exits the cowl through a rubber
grommet. It's very light so vibration is not a problem and starting is so convenient.
The spinner is a polished Tru Turn.
Right now the BME 100 is swinging a Mejlik 26x10 but I'm sure I'll try more props as time
goes by.
My initial flights were done with Amsoil mixed at 100:1 but I recently mixed a batch of
petroleum based oil so the engine could break in more readily.
My first tach readings give me a peak RPM of around 6400. I'm not extremely happy with the
amount of vibration I'm getting so far but my light airframe could be contributing to the
problem.
Getting the engine in proper tune has not been a walk in the park either.
Balancing the plane was a crap-shoot since it was a new
design and I have no idea how that value is determined in the real world.
I went for the modeling standard and placed the balance at 1 inch in front of the leading
edge of the wing tube.
The plane felt solid on the first flight but I really didn't look for anything more than a
safe landing.
Lines were solid and locked in, up lines are straighter than any plane I've ever flown.
The down lines are just as good with no tendency to pull out.
On the first few flights I could not get the plane trimmed exactly right, the elevators
are so sensitive.
It wanted to flare quite a bit on the landing approach so I assumed it was tail heavy, but
more on that later.
We were trying to evaluate the snaps both inverted and positive. We tried it on uplines,
downlines, 45's and horizontal. After a couple of times the plane started to over rotate a
half turn with a slight hesitation before the over-rotation.
It was really bad in negative snaps but seemed to work ok on level positive snaps. I
wasn't sure about fuel consumption so I decided to land.
On the approach I noticed the linkage to the right aileron was hanging loose... The metric
machine screw that holds the servo arm on had fallen out. We must have been flying it that
way for ten minutes doing snaps and all sorts of crazy maneuvers. I can't believe it
didn't flutter and destroy the aileron or maybe even the whole wing!
No damage at all.
Knife edge flight had a disappointing amount of pitch to
the belly trying to dispel my notion of a lowered stab being that flaw's solution, again,
I will later find out the real problem.
My real curiosity was in how it would do in 3D flight, after all, that's the purpose for
this design.
My flying skills in this area are growing but still quite limited. Luckily a contest was
coming up within driving distance and there would be many pilots attending capable of
putting the Extra through it's paces.
What is it that makes Carden's Cap so stable in 3D flight? Get the plane into a
stall, add full elevator, blip in a little power, and the plane is rock solid.
Apparently there is more to 3D than a bunch of throw and big control surfaces, but you
gotta start somewhere.
My first attempts at anything close to 3D was to see how it handled in an elevator.
After all, that maneuver is the basis for all of 3D.
It was obvious from the start that something was not going according to plan. Once the
plane entered into the stall any attempt to keep the tail below the nose caused the plane
to rock wildly a la the falling leaf. It looks kinda cool but this is definitely not what
I'm after.
Following a really long and hot day at the CAPI IMAC
contest I twisted Branwell's arm enough for him to give it an evaluation flight.
Immediately the plane was performing a torque roll ( wow what a great airplane! nothing to
do with the pilot. ) But beyond that Branwell's finding confirmed my earlier decision that
something was amiss.
The plane was just out of control in any type of stall. Branwell felt that the plane was
extremely nose heavy and that the control surfaces were not doing their job.
The following week was experimentation time. Knowing that the plane wanted to flare quite
severely on the landing approach caused some hesitation at moving the CG aft.
I found that I could not get the plane properly trimmed in pitch so I changed the trim
resolution cutting its total movement to 30% of its original travel and making each click
of trim that much finer.
Voila! The plane flew completely locked-in and the flare was eliminated.
The new JR servos have precision centering like never
before making minor adjustment very sensitive. After moving the CG back just one inch to
the leading edge of the wing tube many improvements were immediately noticed.
Pitch on knife edge almost disappeared renewing my faith in at least that part of my
design. The mix is down to a scant 3% both directions. Inverted elevators lock right in
and waterfalls require almost no rudder correction.
I was unable to perform a flawless waterfall with my Cap but now the Extra made me look
like I had been doing it for years.
Upright the elevator still has a little rocking motion but nothing like it was
before moving the CG. I think I can control this amount of movement but I could try
moving the CG back a little more or reflexing the ailerons for even greater stability.
Sealing the hinge gaps was just about miraculous, very tight knife edge loops are effortless, as a matter of fact the plane will almost pinwheel.
Even though the two planes look nearly identical ( Norm's wing stripes are straight, mine curve aft at the root) they are really very different in both construction technique and in wing design.
Norm is testing the plane with an unmodified G62.
His plane had a final weight of 20 lbs. 12 oz. And mine was slightly heavier at 21lbs 2
oz. -- About a pound more than we would have liked but still quite acceptable giving a
wing loading of a little over 28 oz per sq. ft.
Can you say floater?
Norm is utilizing a high aspect ratio wing
featuring an airfoil with the MAC about 5% further aft than mine and a sharper leading
edge.
The ailerons on my plane are constant cord where Norm's taper.
We found that his plane seems to be a little more stable in the elevator stall but it
wants to roll left just slightly. I had to hold in right aileron but the plane is very
stable. My thinking is that the P factor is putting more pressure on the raised left
portion of the stab causing the tendency to roll.
I think the airfoil on his plane is the reason for its stability and is 1superior for 3D but I still like the planform on mine. Maybe we'll try a hybrid wing to see how it performs.
Norm's plane also tends to over rotate after
negative snaps that have very little airspeed --but no more than my Aeroworks Edge540.
This could be attributed to a couple of things.
The JR 4721s could be flexing slightly upon centering, or the narrower wingtip may just
not have enough area. Also Norm likes to let go of the sticks after his snaps allowing a
little center wobble, but that doesn't seem to be a problem when he flies other planes.
Hey I'm searching here!
After just a few flights Norm was able to tune into the snaps and the issue was just about
moot.
Check out those great Pilots! 1.5 oz after painting.
The scale dash panels were scanned from the Internet and are actually from a 300L.
Like mine, Norm's plane locked into pattern maneuvers very well with neutral flying tendency in every attitude. Norm flew the IMAC unlimited schedule which will challenge any plane this year with its multiple snaps in the most difficult places. The power requirement is substantial. We were both very pleased with how his plane performed with the stock G62. It will pull through all of the maneuvers with just a slight slow down at the most difficult points.
I am having blast with my prototype Extra. It will push me to practice 3D with more regularity and push my comfort zones to learn this new avenue of flying. We'll continue to tweak the planes and find their limits. And what we learn from these two planes should improve the next rendition. Designing and building your own plane is like no other project. The incentives are greater as are the rewards when you watch your child take to the sky. I highly recommend a project like this to anyone.
