Dragonfly MK-II 3 airfoil conversion

Discussion in 'Member Project Logs' started by trifoils, May 21, 2016.

Help Support HomeBuiltAirplanes Forum by donating:

  1. May 21, 2016 #1

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    The quickie style aircraft caught my eye when I was a little kid. I saw one at a local fly-in and couldn't believe something so small could actually carry a man into the air and back again. As years passed the idea of building or buying one came and went several times. There were many selling points - stall resistance, simple structure, speed, elegance (landing gear on the wing tips - is that possible?)

    By the time I was in school finishing my degree I had read enough pilot reports of extremely high landing speeds, poor ground handling, broken canards from PIO, and the bug / rain trim problem. Somehow I made it to the yahoo Q-bird forum (this was over 10 years ago) and these problems were well known. There were many ideas for how to update the design but no real breakthroughs, except one user by the name of leon who had experimented with adding a horizontal stabilizer in X-Plane, the flight simulator. I decided to get the sim and experiment with the same idea.

    It worked very well and changed the personality of the plane, especially in slow flight. Stock tandem wing planes had such a slick glide path and low approach angle it was not easy to set them down where intended, but adding the stabilizer opened up many possibilities for flaps and other control system changes, including using the same airfoil on both wings. I played with my own version of a tandem wing + tail for a couple years and came up with a refined hot rod with real STOL capability.

    Years passed and I never built the prototype I designed in X-Plane, but recently decided to investigate how a stock Dragonfly MK-II would react to having all pitch control via conventional stabilizer. It worked very well in the sim and small RC models, so I have decided to try it on a real aircraft. The aerodynamics behind the idea (original custom design and the retrofitted dragonfly MK-II) can be seen in the X-Plane forum:

    T-tail Quickie Derivative - Custom Designs - X-Plane.Org Forum

    dragonfly MKII - Custom Designs - X-Plane.Org Forum

    The aircraft itself was a craigslist find, and will need a full inspection before proceeding with the modifications. It was finished but never signed off and the engine has a few hours of run time. I will be changing the panel and landing gear, in addition to adding the tail.
     
    Last edited: May 21, 2016
    Joe Fisher, Peterson and delta like this.
  2. May 21, 2016 #2

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    The existing airframe was built from the TASK quick build kit, with premolded fuselage parts and an aluminum header tank. The wing layups were assisted by a professional kit manufacturer (remember the KIS series?), but the peripheral layups, bonds, and rigging was done by a true amateur. Many of these secondary jobs will need to be repaired or redone completely, and the plane sat out in the sun for almost four years in Camarillo after staying in a hangar for twenty more. The engine (a hapi / mosler 82 hp "magnum") was made in 1990. The builder created some beautiful hoerner wing tips for the rear wing, which produces hardly any lift during the critical phase of flight (landing), so they essentially add drag with no payoff. The front wing looks and feels very solid, with a great contour on the upper surface.

    I've been a composites manufacturing engineer since 2008, dealing mostly with damaged parts, process control and operator training with some bonding, fastener joining and cad design thrown in for variety. The few red flags this plane threw up at me when first viewing it have been minor issues that can be easily repaired, though the overall condition is very sad. I partly took this on out of pity for such a potentially beautiful aircraft.
     

    Attached Files:

    Joe Fisher likes this.
  3. May 22, 2016 #3

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Since I'm sure it may come up (and already has in a private message I received), the basic concept is to build the Amsoil Racer with a conventional (instead of "T") tail, using a dragonfly airframe. For initial flight tests, the canard will be geared to the elevator with perhaps slightly less throw.

    Here is an RC model in the smallest practical size (things fly a lot worse the smaller they get), that proved the airworthiness of the idea. We were not allowed to take photos or video in the hangar where I flew it but this pic should show the layout. It flew on rudder, elevator and throttle only, and had a level of stability and control unlike anything I've ever flown. "It's on rails" was one of the comments I heard from an observer. I could bank and yank it out of a corner of the building then skim the floor, avoiding equipment and parts like an A10 creeping through a canyon.

    This is the final iteration of what began as a stock tiger moth that had the wings moved, rear wing size trimmed, tail size trimmed, and ultimately flew circles around its former self with absolutely no tendency to stall in high alpha. You'll have to excuse all the tape repairs. We had several midair collisions per day in the hangar and this thing got shredded a few times, but was much more durable and recoverable after an impact, following the changes. Though it's not a one for one dragonfly + tail exact scale model, it proved the concept through many adjustment cycles and the only thing that ruined its flying qualities was too much up thrust after smashing into a pole, causing it to stall in climbout and drop a wing (easily recovered from).
     

    Attached Files:

    Joe Fisher likes this.
  4. May 23, 2016 #4

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Here is the start of the teardown and prep work for the modifications. The HAPI engine has no corrosion on the inside but I can't figure out if it has a fifth bearing or just a plain oil seal on the front of that sleeve behind the prop shaft.

    Moving an airplane is a big job, but this airframe is so small it was possible with a 12' flat bed trailer.

    I began scraping off paint today to allow a new layup where the stabilizer will go, and found the builder did a great job of filling and priming, but he used the heavy stuff made for car finishes.

    I spent several minutes sitting in the cockpit, changing cushions around to get the best position, and really love the comfort and security of the interior. It could use a little more visibility down the sides but I doubt I'll care once it's airworthy. The TASK fuselage has a great contour that necks down toward the tail. My biggest impression of this aircraft so far is that it looks, feels and is built just like a sailplane. It could be a quick little thing if that motor really produces 82hp.
     

    Attached Files:

    Joe Fisher and MikePousson like this.
  5. May 25, 2016 #5

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    The canopy was out in the sun for several years and has little fissures visible when you get close (it's actually clear colored; there was a tan sheet behind it in the photo). They look like cracks but are not as thin, and I wrote off the canopy as unusable when I first saw the plane, but now I'm not so sure. They are only on the inside and can be sanded out, but it takes hours with 150 grit to clear them off. Sanding is easy work that does not degrade when drinking beer, so I've been going at it for an hour an evening. If it doesn't work the only loss is my time and labor but if it is successful I will avoid having to buy a new $600-800 canopy with all the work and risk involved in removing the old one from the frame, trimming and bonding in the replacement.

    Sanding while buzzed puts me in a meditative state of mind when I contemplate the rest of the project, and I think I will change the hinge setup for a tip-up with the instrument panel attached to the canopy frame. It will make it much easier to get my legs in place when entering the plane.

    Also I think I will cut off those beautiful Hoerner wing tips from the back wing and reuse them on the front, then create the traditional mildly upturned tips for the rear. It might lower the landing speed by a couple Mph, and it just looks good in my opinion:

    Dragonfly - Nest of Dragons

    I'm really looking forward to the finished product. Dragonfly owners seem to be very happy with their planes.
     

    Attached Files:

    Joe Fisher likes this.
  6. May 30, 2016 #6

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Here is a panel mock-up to check if I like it. There will be a 7" Avmap in the center with a flowing dashboard over the passenger's legs.

    I've been peeling away the layers of filler on the tail and the builder was serious about getting the contour he wanted, because even now I'm barely down to the dry micro layer. No real surprises or disappointments so far.
     

    Attached Files:

    Joe Fisher likes this.
  7. May 31, 2016 #7

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Here is the mixing station ready to go. Little dedicated work areas like this can save a ton of time in the long run. The original Rutan composites training course has plans for a cutting station for the cloth, which I would make if I was doing an entire wing; I'll just do it on the pool table for the stabilizer.

    The original prop looked weathered but after some sanding, most of the wood is solid and intact. It cleaned up to a nice shape. I will wrap the outer half of each blade with a thin layer of fiberglass at +/- 45* for torsional rigidity and to keep the ~1/8" thick tips from splitting if I hit a bird.

    It is a 54 x 46, which should be good for 150 mph at 3400 rpm.
     

    Attached Files:

    Joe Fisher and BJC like this.
  8. Jun 4, 2016 #8

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Wings temporarily reassembled to get an idea for the overall alignment and where the tail should go. Here is the dorsal fin taking shape.
     

    Attached Files:

    steveair2, Joe Fisher, Himat and 2 others like this.
  9. Jun 5, 2016 #9

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Got the prop tips wrapped this evening. It was a fun job that took a couple hours. The process starts with precutting the glass then painting some resin on the surface, laying the dry cloth down, shaping it around any curves, painting on more resin, squeeging out the excess and finally adding peel ply to the exposed edge to keep it laying smoothly during cure (this is a huge time saver). The finished layup should be completely transparent with the woven texture visible at a low angle. It's OK to have a random brush bristle left on, according to the original RAF documents.

    This was a practice run to get me back up to speed with wet layups (most of my career has been prepreg type stuff).

    Here are the pics:
     

    Attached Files:

    steveair2 and Joe Fisher like this.
  10. Jun 6, 2016 #10

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    The trimmed blade, ready for microballoon filler:
     

    Attached Files:

    steveair2 and Joe Fisher like this.
  11. Jun 7, 2016 #11

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    One final refinement possible when adding a horizontal stabilizer to a tandem wing is a reduction in span of the rear wing, since the front wing is no longer doing the work of pitch trim and control it can fly at a lower proportion of the total lift, compared to before. According to the dragonfly plans it creates about 2/3 of the lift, and that 22' span rear wing is always working half as hard.

    All this can change when you have a 16' moment arm on your pitch control surface, instead of the ~3' it originally had, using the canard only.

    Part of what makes this configuration fly so sweetly is it seems to automatically pitch up just enough during speed changes to keep from diving excessively, yet still mush through a stall rather than breaking sharply.

    Here is a top view of the dragonfly MK-II-T with a 4' reduction in rear wing span. It flies excellent in the sim and matches what the R/C model predicted, plus it makes up for the weight and drag added from the stabilizer.

    I won't be trimming my rear wing until after many flight tests, and very gradually at that. Here is the final size, and the original STOL IO-320 powered hot rod that inspired it all (which I will still build after this one is in the air):
     

    Attached Files:

    steveair2 and Joe Fisher like this.
  12. Jun 9, 2016 #12

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Here is the hot wire saw ready to go, made from the power supply and inconel wire available from Aircraft Spruce.
     

    Attached Files:

    BoKu and Joe Fisher like this.
  13. Jun 13, 2016 #13

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Some progress cutting the cores:
     

    Attached Files:

    Joe Fisher and jmt1991 like this.
  14. Jun 15, 2016 #14

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Here is the spar core joined and ready to go, with the ply kit laid out.

    It will be a full box beam (unlike the "c" channel common in the rutan designs, and the rest of this dragonfly) because I might as well close out all sides since I have the spar by itself. It will also give it a little more torsional rigidity (resistance to twisting) which is a good idea with my elevator so far away and so large, compared to the rest of the control surfaces on this (and rutan) design(s).

    I also did it this way to avoid having to cut through the front half of the vertical stabilizer, and rather just cut a small rectangular pass-through just forward of the vertical stab's spar. The final shape will probably be similar to what's on a P-51 Mustang (might as well copy something fast, since it's just an educated guess from a drag standpoint).

    I'm trying to psych myself up to lay it up tonight, so I came to the forum to look at everyone else's projects before I start mixing the epoxy.

    The pics:
     

    Attached Files:

    Joe Fisher likes this.
  15. Jun 15, 2016 #15

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Layup complete (took about five hours).

    The process started with covering the core with a 50-50 mix of epoxy and microballoons, then making the first wrap of +/- 45 cloth around three sides, then the stack of unidirectional glass & carbon, and finally another wrap of +/- 45 cloth around three sides to close it out. Finally, it was wrapped all around with peel ply.

    This was my first layup where it was not possible to leave the part in one position the entire time, and it was a delicate job because of constantly having to mess with plies, due to handling. If I could do it again I would try to make a holding stand that attached to the core from the ends.

    Tomorrow I'll unwrap it and see how it looks. It feels pretty light.
     

    Attached Files:

    Wayne, Joe Fisher and narfi like this.
  16. Jun 19, 2016 #16

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Dorsal fin layup complete. The process was the same as before: the gaps, corner transitions and holes in the core were filled with a thick mix of epoxy, microballoons and cabosil (which keeps it from dripping down). Then the rest of the core was covered with a thin mix of the same stuff. Two plies of "Rutan style" satin weave cloth were added at 0-90 and +/- 45 and finally the edges and top were covered in peel ply to smooth the transitions and help hold the plies in place to keep air from getting in. You can see I even had to tape it in place at the top. Finally a squeegie was run over it for several minutes until no more air came out (it makes a popping sound).

    Mike Arnold has a great video about this basic process here:

    https://www.youtube.com/watch?v=U2PX1p7nbw8&t=6814s
     

    Attached Files:

  17. Jun 22, 2016 #17

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Heads are disassembled and laid out for measurement. There are a few known issues with the HAPI "Magnum" engines beyond the obvious cooling problems.

    The valve guides need ~.004 in. clearance to the stems, and this was not always the case with some engines due to a worn out reamer used in the final years of HAPI, according to the Mosler bulletin sent out shortly after they acquired the company.

    Also the springs are in the 150 lb range which is way too much for an engine with a mild cam profile running at 4000 rpm. This may be partly what is behind the claims by some engine owners who experienced cracked heads or stretched valves. Something in the 90 lb range should work fine.

    And lastly, the two piece pushrod tubes with internal springs have a habit of chafing the pushrods.

    My engine luckily has the required .004 in. stem clearance already but probably the heavier springs and definitely the two piece tubes. It does not have the magnets in the flywheel which have a reputation of flying off.

    There was a known crank failure mode in the initial Magnum engines because it used the same internal parts as the 60 hp version, but it was all redesigned after about 1987 and Hapi claimed the new version was 230% stronger. I checked the hub interface and it seems to have a 1/4" diameter pin pressed into a hole in the hub to handle the torque loads, instead of a woodruff key which can cause a crack to form after several years. I tried pulling the hub off with a huge three jaw puller but it didn't budge. I'm going to accept it as is, since my engine was made in 1990 and the hub looks quite a bit different from what the 1835cc engine had (a bit thicker all around).

    No additional corrosion has been found inside the cylinders, which still have the hone marks. I'll get some valve springs and pushrod tubes, reassemble it and do a very careful job of baffling the engine. I've got a quad CHT in the panel to monitor it during test runs. Rex Taylor claimed the cooling issues these engines were known for was due to baffling leaks, and based on what I've seen in GA style baffling, I'm not surprised. There are many leak paths with standard aluminum sheet and silicone seal type baffles. I might just lay up a composite (post cured) baffle / plenum to avoid this problem.

    The pics:
     

    Attached Files:

    steveair2, Joe Fisher and cluttonfred like this.
  18. Jun 23, 2016 #18

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    I will try to answer a couple questions that came up in a private message.

    First: why do beams need shear webs, and why do we wrap a composite spar with +/- 45 fibers after all the unidirectional fibers are down. When I first saw all the plain weave bidirectional fabric wrapped around some professional grades spars recently, I wondered the same thing. It is because the parallel fibers need something to tie them together when under a load, or they will slide apart. The resin is not strong enough by itself to keep everything together.

    I like to visualize a phone book being bent. The papers slide across each other until you squeeze hard enough to keep the pages stuck together. Same is true with fibers in a composite beam, spruce spars in a built up beam, or a metal I beam. The web, be it +/- 45 degree bidirectional cloth with a core under everything, vertical grain wood planks, or a vertical sheet of steel, is there to keep the upper and lower spar caps from sliding across each other when the beam is bent.

    Another question I will try to answer in another post is why does a tandem wing or canard system (or any stable airplane) work, what makes it stable, and how does adding a stabilizer to a tandem wing make it better. The short answer is it has to always face the incoming airflow under any situation, and it needs to raise the nose if airspeed goes up and vice versa for the opposite.

    Feathers on a dart are the most obvious solution to this problem, and canards / tandems do it by loading the front wing at nearly twice the pressure as the rear. The front wing is always the most heavily loaded, and it goes down the further back the additional wings are. This is why conventional planes are more efficient than canards - because the most lightly loaded surface(s) are also the smallest. Three airfoil planes give up some efficiency here but make it up in maximum possible lift coefficient, which allows the designer to reduce the total surface area for a given stall speed. John Roncz has an excellent article on this and I will post a link if I can find it.

    I chose a three airfoil system for this reason, and because it greatly improves the slow speed maneuverability of a tandem wing arrangement. I will prove all this with the Dragonfly MK-II-T prototype, which will flash an unmistakable profile as it streaks across the sky.
     

    Attached Files:

    steveair2, Joe Fisher and cluttonfred like this.
  19. Jun 24, 2016 #19

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Here is some progress on mounting the stabilizer. After leveling the plane and checking the incidence of each wing (they were both already aligned at zero to each other), the stab incidence was set at roughly negative 1-2 degrees and the outline drawn in place. I contemplated setting it at zero like everything else, but it will be working the hardest in slow flight, especially if the plane is loaded toward the front of the CG range, and I want all the power it can give me. This will cause the elevator to be deflected down a few degrees in cruise flight but it's a drag penalty I'll accept, in exchange for more elevator authority in flare for landing. If you've seen the stabilizer on any Boeing, it rotates for pitch trim because you get much more lift by changing the angle of the whole surface, compared to just deflecting the elevator.

    The spar pass through was cut with a dremel and the spar was temporarily aligned and held in place with bondo on the top, then normal glass tape joints over fillets were added to the front and back. Finally the leading edge cores were bonded in place with thick micro (one part mixed epoxy, 3-4 parts glass bubbles with a dash of cabosil).

    I'm trying to wrap my brain around what kind of fairing it will need to fill the space over the top of the tail cone. Times like this I just glue in some foam and start carving, and my subconscious mind comes up with the answer, after a while.
     

    Attached Files:

    steveair2 likes this.
  20. Jun 26, 2016 #20

    trifoils

    trifoils

    trifoils

    Well-Known Member

    Joined:
    May 21, 2016
    Messages:
    111
    Likes Received:
    100
    Location:
    Tehachapi & Palmdale, CA
    Hotwiring the elevator hinge cove:
     

    Attached Files:

    Mavigogun, steveair2 and Joe Fisher like this.

Share This Page



arrow_white