Thread Tools
This thread is privately moderated by BAF23, who may elect to delete unwanted replies.
Jan 11, 2019, 05:22 PM
The sky is the limit
BAF23's Avatar
Thread OP
Build Log

Build log of a JPV Foka4 scale 1:3


Pic of the real Belgian Foka4 anno 1963 somewhere in Belgium, still wearing its world championship race number



Part 1: Prefabrication by French hobby builder Jean-Pierre Voisin

After a stupid priority choice following an unfortunate combination of events, my 4m span Fliegerland Foka 4 plummeted vertically in the ground minus its port wing during the September 2018 BiGGS meet at St-Truiden (illustrated story here) https://www.rcgroups.com/forums/show...02&postcount=3. As Fliegerland was out of Foka kits for a minimum of 6 months I shopped around and had to go for a 5 meter span version which I ordered from JPV, a Normandy retired builder (he has no website). A few months before that, I saw a fellow BiGGS pilot maiden one at Pottes and the quality of flight and the model seemed very good. The weekend after the crash I got in touch with the builder and within days I order a pretty much prefabricated kit consisting of a molded fuselage, styro wings with Samba (Abachi type) sheeting over 100gr glass-fibre ready for finish, ailerons and 400mm spoilers installed. Wings and tail rigging made, wheel installed, clear canopy without frame included. Removable rudder with built-in servo installed. That is a lot of work this amateur but very capable builder delivers (halfway via a trucker) for about €1500, after 10 weeks of intense labor. Basic fuselage and foam wing cores can be ordered for less than 500 euro if desired (or for repairs).

Regarding the choices I made: I wasn’t going full scale on this project but wanted a solid model that was easy to transport and assemble, with the correct outlines but practicability paramount. Single mechanical large spoilers only on top of the wing were simpler than smaller ones on top and bottom. The full sliding canopy looks cool but weakens the nose and restricts space for ballast and access to batteries, so I ordered the opaque front to be glued solid to the fuselage, leaving the clear canopy to be fastened non-scale wise on the field but providing full field-access to everything in the nose. Once closed for flying, the fuselage shape looks completely scale. As this is also an aerobatic model of fair size, I decided to install high voltage servo’s for more power directly driven by double 2S LiFe batteries. Following servo’s were purchased: Rudder: Hyperion DS20UMD (12kgcm), Ailerons and spoilers: PichlerMaster DS3010HV (8kgcm), Elevator and Towhook: Hitec D645MW (12kgcm).

During the build, JPV sent me weekly updates with pictures of the progress on my model. It was nice to see how well equipped he was to produce the molds and assemble my Foka on a rigging that ensures perfect squareness of all surfaces. Here is just a selection of the many pictures I received, illustrating well how better he is equipped compared to myself to undertake that project and ending with a qualitative very strong but still relatively light scale-looking glider (he mentions about 11kg).

Pic2657 Making the furrows for the aluminum wing key


Pic 0443cr Servo placed inside the rudder because of extreme rake of the latter


Pic 3206 Half-fuselages still in the molds, being joined


Pic0341 Fuselage in the rig, foam wing cores cutting


Pic 0319cr Tail rigging


Pic 0615
Wing rigging


Pic 0224 Foam wing assembly


Pic 5507
Wings vacuum sucked, reinforced with UD290 strips and glassed with 2 diagonal 100gr tissues, heavy weights as precaution for pressure loss during curing.


Pic 1046 Wings covered with Samba wood, weighing exactly 1651gr each (without servos)


These images clearly illustrate that the value for money is excellent for this artisanal product and that no matter how good you are, it would take you longer to get the same result because he has all the rigging at hand. Mind you, while he was working on my Foka, I was working on my “secret” glider-project and before his delivery, had to carry my 93 year old father’s luggage during a two week transatlantic cruise. (Barcelona-Teneriffe-Bermuda-Miami) on the stylish cozy Azamara Pursuit ship, there are worse things in life to keep busy. Enroute between Santa Cruz and the Bermuda with force 9 gale wind, JP sent me pictures of the model after the final rigging, I was delighted. That model looks really impressive and the elegant tail never suggests this is a scale model of a 1960’s designed glider.

Pict 121205


The day after my return home, JP announced a French befriended trucker (and model pilot) would bring my Foka to a town not too far from the Belgian border the following Sunday. It was very well packed and the fuselage looked awesome. All this got delivered a mere 10 weeks after initiating the order and build, remarkable! My 500km drive was well worth the assurance the outsize Foka got home without transport damage.

Pic 714cr


Next morning it felt like Santa had passed overnight and I spent half a day unpacking and inspecting my new toy. The build quality was good and seemed very beefy. The wheel seemed strong enough for a wheelbarrow, just as the 2mm thick 30mm diameter AU4g 2017T4 wing key. I got two clear canopies and the finish with the Samba wood on wings, stab and rudder was truly superb. Gaps for ailerons, spoilers and elevator were minimal, to the point I’m afraid these physical throw restrictions might hamper the maneuvering of this aerobatic capable glider. I used a spare cradle of my Ka2b to temporary stock the raw fuselage in between my other large gliders, then fabricated bubble covers for the wooden surfaced parts before putting those aside till later.

Pic 718


Read the following parts 2 to 5 in this tread for further model completion.
Sign up now
to remove ads between posts
Jan 11, 2019, 06:31 PM
The sky is the limit
BAF23's Avatar
Thread OP
Part 2: The fuselage

After the love at first sight was substituted by sober analysis, I noticed that the fuselage halves mating, outer shell finish and fillets for wings and tail were of good but not high quality and would need further work before finish could be applied. That was not a problem, but comparing the shape of the fuselage to known plans and pictures of real Foka’s revealed that the cockpit tub (and fuselage top) seemed much too wide and not to scale. Most people wouldn’t notice, but cheating will have to be done with the pilot figure to conceal that anomaly. Real Foka’s have narrow fuselages with a pointy top and wider belly (a bit like a pear as can be seen in the picture below) in which a pilot has to be squeezed before the narrow canopy can be slid closed, but the JPV Foka is rather obese in the shoulder area, more apple-shaped.

Pic Harry with Foke


This also affects the canopy and I am glad I already had made my mind that I would make it as a semi-scale model. In order to work comfortably on the fuselage I decided to start building the custom cradle first, instead of after model completion. I used a combination of wood plus hard and soft foam, taking into account that the 66cm long wing key had to fit between the supports, and the model had to sit correctly on my rack after ballast was added in the nose. The nose skid also appears to be rather high, and that was even before the aluminum skid was added (because the fuselage is rather thin at that sensitive portion). On the other hand, I used that unexpected depth to locate the towhook mechanism. A hole was cut about 10cm back from the nose so the orifice would be almost invisible (real Foka’s only have a single hook on the port underside further back, that serves for either winch or towed starts). A large-size aluminum tube-hook assembly was then reversed front to back and grinded so nothing would stick out of the hole with the tube pointing slightly down. To spread the forces on the hook I filled the forward space of the gutter with a length of square hardwood which I blocked with expanding PU wood-glue whilst the aluminum tube was fixed with strong UHU-Endfest. The 12kgcm Hitec D645MW HV servo was affixed on a wooden pedestal just behind the hook, with a wooden dowel in-between to ensure that even if either component comes loose, the hook will open by itself.

The other identical Hitec 645 servo was for the elevator and had to be installed in a (JPV prefabricated) custom cutout under the rudder. That place is deep, narrow and under the 45° sweptback fin, not the easiest place to work in. A 2,5mm pushrod runs straight aft. The designer made a small hole for it to exit the fuselage, but that meant the control horn would be highly visible much more aft than the fuselage at that spot. I opened up the hole to create a vertical slot through which the control horn and Kwik-link passed (with full down elevator), but still be fairly close to the aft fuselage in cruise. I used a sandwich type control-horn screwed all the way through the elevator because I was unsure of how solid the foam core was between the planking. Testing for arm lengths at both sides (for correct throw) was tedious because each time the servo arm had to be separated from the servo to allow the linkage to be adjusted in length, it had to be extracted through the gap, all in a canyon just slightly wider than the servo. With the servo in place there was no way to run the power wires but to make the extension wire all the way through the long fuselage, so I fabricated the wires for rudder and elevator at that time. What takes 20 minutes on most models took 1-1/2 days on this Foka, and everything (except the metal hook) was taken out again to allow proper painting of the inside, also under the rudder because that cavity will still be visible when the rudder is deflected on the ground. Most modelers wouldn’t care about that, but I do it for the peace in my mind because I get upset looking at aesthetic shortcomings of my models on the field. Here is a picture of the servo’s for elevator and rudder taken at a later stage, looking down from the aft vertical stabilizer.

Pic 0735


I used 3mm ply to fabricate the frame for the canopy. I then drilled two 3mm holes through the frame and the fuselage at the left and right forward corner. Bits of piano wire were glued at right angles into that frame part so the frame could catch and lock in the forward position. Another hole was drilled in the middle on top of the back frame to accommodate the sliding pin of the opening latch from the fuselage. That got prepositioned but not yet glued to allow easier sanding and painting of the fuselage. The clear canopy-acetate was then cut to the correct outline, a tedious job because none of the frame lines are straight in any direction. Furthermore, JPV hadn’t mounted the aft of the nose portion at a square angle and that caused various complications during the fitting of the canopy and interior. The accurate cutout for the vent window was very delicate and installing the sliding vent window is a work of patience. I first drilled 0,8mm holes in the upper and lower guides, then threaded these by slowly turning in the M1 screws and taking them out again. The window was then positioned in the opening made the acetate and the lower guide pressed into position after which the position of the 3 mounting holes was marked on the canopy. These holes were then drilled out with a 1mm finger drill, then using a 2mm finger drill a depression was created to accommodate the conical screw heads so everything would be flush when assembled. An eyebrow pincer was then used to guide the tiny 1mm screws into the hole while the tiny screwdriver pushed and turned to engage it in the previously prepared holes. With the lower guide in place and mounted flush, the sliding vent window was put back into position and the same sequence was then applied for the upper guide. Only do this on a perfectly clean working table and floor using a white towel to work on. I twice had one of those microscopic screws slip away and would never have found them again without the previously mentioned precautions. That is one of the most tedious assemblies on any scale glider, after that the rest seems peanuts (I had to do it twice because at a later stage I wrecked the canopy by accidentally stepping on it).

Mating the clear canopy to the home-made frame also required some ingenuity because of all the critical angles and shapes of that frame. Lots of dry assemblies were made before all the necessary parts were prepared for the gluing process that had to be done on the fuselage itself in one go. Plastic bags were inserted between the frame and the fuselage so any dripping glue could not cause both to stick together. The frame was then held into position by clamps to prevent pressure on the canopy to move the frame during the 24h drying process, and to have the length of the frame hugging the slightly concave shape of the fuselage. Clear canopy Formula 560 glue was then applied all around the frame and the canopy delicately put into place. Many rubbers were then slid over the fuselage nose to ensure the canopy was pressed everywhere against the frame. On the lower side I therefore added a rubber hose to better push it in place. The picture below illustrates how all this was done.

Pic 0719


After drying for about 8 hours, the glue was sufficiently solid for the rubbers and clamps to be removed using a wooden rod to push through the open vent window. The frame was then separated from the fuselage before the canopy glue completely solidified. Additional glue was then forced into any remaining openings plus in all the inside corners for added strength. Clamps were again used to force the clear canopy to snugly cling the frame where this had not been successful the first time around. Due to the rake of the aft panel and the pilot headrest being fixed to the fuselage (I fixed mine on the seat for better accessibility of the wing wiring), I decided to cutout a major part of that panel, thereby losing its lateral compression qualities. This necessitated 4 additional pins between the frame and fuselage rim to ensure canopy rigidity under all circumstances. After drying for an additional 24 hours, the lower canopy surfaces were sanded smooth and the inside canopy was taped to allow gray paint to be applied. The fake canopy closing handles with red knobs were later inserted on the inside to complete the canopy assembly.

At that time I also inserted horizontal wooden dowels between the fuselage surfaces that will be pressed against the wings due to inertia in case of wings hitting a pole in the fields (backward compression) or the nose against a wall (forward compression). Experience proved a solid transverse dowel at the front and back prevents difficult fuselage repairs in those critical structural areas. On my previous Foka a few hard landings caused the inertia of the wing-weight to compress the fuselage downward, causing cracks around the adjacent bow just behind the canopy. The wheel being more aft on this model, I couldn’t link it to the wing key and just glued some solid vertical beams between the wing key holder and the bottom of the fuselage close to the wheel. The extra weight at the CG is a small penalty to pay for the added fuselage rigidity.

Pic 0725


A 1:3 scale pilot from Warbirdpilots.com was purchased through the French Topmodel retailer for an acceptable 175 euro, together with safety belts and other scale interior details. The pilot is 56cm tall but as with many scale puppets, is too svelte with a rather narrow head. Luckily the glider-type hat widens the latter and the former can be corrected by stuffing his shoulders and thighs under his nice clothes. The leather flight jacket is very well done but his face and hands have to be color embellished to eliminate the plastic look. The figure weighs 425gr and can be articulated to adopt the required stance for any cockpit. Noteworthy is the ease with which the hands can be rotated to grab a stick or handle. Maybe not commonly dressed for the glider community, but how many (ex-)military pilots don’t hang around in civilian clubs still wearing their military leather jacket when flying? (and without the jacket our pilot still sports a nice complete long sleeved blue shirt).

Time to start on the interior. The pilot fills the space below the canopy so well that little is seen from the seat or lower side panels, I therefore elected not to spend much time reproducing a scale seat-pan that wouldn’t be seen anyway. Despite having only a few interior pictures and a sketchy scale drawing I was able to first construct the fixed side panels, then a removable floor with seat and headrest. A period instrument panel was ordered from scale-cockpit.at because all I had in stock for this scale was Czech instrumentation.The side panels were a challenge because very little info concerning those is available and there is not a single (known) full-size Foka4 remaining in Belgium.

Pic 0728


Because on this Foka the wheel is mounted well aft of the CG (as per real Foka), any takeoff and landing will cause much friction on the nose skid and the polyester bottom would rapidly be scraped away when operating on hard surfaces. The model needing weight in the nose anyway, I decided to install a complete 2mm thick aluminum skid between the hook till the wheel (a real Foka uses a wooden skid). After shaping it to fit the underside of the fuselage shell, I drilled holes to screw it in place. I also drilled a hole at the correct off-center location where the real Foka has its single multipurpose hook. Winch launchers are not my preferred mode of operation, but I made it sufficiently strong to withstand such forces, but the main purpose of that non-scale hook is to rapidly connect my ground tow cord to evacuate the landing area without having to fumble with the transmitter to close the electric hook.

Pic 737


With all the hardware in place I then completely assembled the model for the first time in order to roughly balance it for the suggested 114mm point. It needed a hefty 2+kg for that so before further finishing the interior I stuffed lead in the nose. I started by filling the lower side cavities around the hook and under the pilot’s feet with 410gr of air-gun steel balls. These were held in place by allowing ample quantities of PU glue to flow around them. I took that opportunity to anchor long inverted M4 screws in this so as to be able to better block the additional lead plaques that were further shaped according to the nose shape, to complete the fixed and removable ballast.

I then custom cut six different 3mm lead plates so they would hug the fuselage nose contour (at their height) still allowing access to the hook mechanism and servo. The idea was to place both 2S2100 LiFe batteries (106x50x11mm ea) side by side above that access panel. As the batteries weigh 110gr each and the balls 410gr, the contoured basic fixed lead had to a minimum of 1450gr. Each individual plaque was then numbered and weighed before all were dropped in sequence over the embedded long screws and fixed with bolts. P1=204, P2=238, P3=234, P4=270, P5=315, P6=330 for a total of 1598gr. The bolts are long enough to accept more plaques that can be added or removed to fine-tune the balance after the test flights, 2,5kg prepared ballast should be sufficient to start with for the final weight and balance. All of the weight is kept in the lower half of the nose for stability, leaving plenty of space to add a working airspeed sensor in the nose-end as per real Foka.

Pic 0729


Final shaping of the tips of the rudder and horizontal stab was then performed with sandpaper, and the rudder hinge-rod shortened and shaped to be hidden completely within the rudder top. All the tail feathers were sanded smooth to be readied for priming. After masking the inside of the canopy, all of the nose-interior till the wing-key was primed. Whilst drying I started preparing the outside of the fuselage, horizontal stab and rudder. After sanding the blemishes and shiny gel coat away (for better adherence of the primer), I applied filler to the remaining imperfections but overall, the basic delivered shell had been very well produced. After sanding the interior and seat I applied a second coat of primer, continuing with the interior of the aft fuselage under the tail and the recess in the vertical stab behind the rudder. This may be an overkill but I like my surfaces to be smooth and uniform in color even through slits or peek holes. Although the primer is relatively thick, I apply it in two coats, the first with a brush, then sanded smooth almost to the basic subsurface, then a second coat with a roller. At that stage I also applied ample primer on the horizontal stab and rudder, both had been laminated with Samba wood that absorbs the primer like a sponge. Use of any other kind of filler could reduce the adherence of the primer. I count on the primer to help saturate the wood so it becomes stronger and doesn’t dent during future operations, handling and transport.

As the (white) primer (and paint) are water based, they add little weight but fill any imperfections and are easy to sand. Most builders nowadays use shrinking self-sticking materials, but with the orange final paint (it doesn’t cover that well) I prefer to have a uniform primer everywhere under the lacquer, that is easier to repair if necessary. I was tipped too late to use gray primer under orange or red. Traditional build and finish take longer, but have definite advantages over expensive modern materials, the proud of craftsmanship and non-toy like appearance being some of them. After a few days of priming and sanding, inside and outside were sufficiently smooth (but still roughened with 400 paper) to be painted by roller. The interior got two coats of traditional gray color, on the exterior I started with the masking for the large white panels. Masking was not essential but knowing it would take a few coats to get uniform colors, I preferred to load them up so after masking again I could apply an equal amount of paint layers of white and orange side by side and still have a smooth finish. I let those dry for a few days before applying masking tape on them to apply the orange paint.

Pic 0731


Many builders complaint they cannot remove the masking tape without collecting parts of the previous layer in the process, but most don’t have the patience to let the coats thoroughly dry. If the paint canister mentions dry in 1h and overcoatable in 5h, that doesn’t mean the paint is completely dry after 5 hours, I take at least 24h before I even consider using masking tape on it, and at least 48h before I delicately then remove it. Most paints (even modern 2K ones) take two weeks to completely harden out. Alternate jobs and let applied paint thoroughly dry before handling it to anything but soft touch. Between the many coats and masking between colors on fuselage and canopy, I used wet 1000 grit sandpaper to eliminate blemishes and augment adherence of the next coat. One month after taking delivery of the model it already looked like this just before Christmas (I make my own presents ;-)

Pic 0733


I got the Caliegraphics artwork just before New Year and applied them to the fuselage greatly enhancing the overall appearance. Electric wiring was then routed through the long fuselage and to MPX plugs just inside the wing roots but further work on and in the fuselage had to be suspended until the scale instrument panel got delivered from Slovakia. I then lost another week installing and painting a new canopy after I inadvertently fell on the first finished assembly after I lost my equilibrium stepping on a loose screwdriver on the busy floor of my small work room. As from that time I was much more careful and stocked the individual Foka parts on racks when I wasn’t actually working on them.

Pic 742


Part3: The wings

As can be seen on the pictures made by JPV, the wings were very much finished (with installed spoilers) but without servo’s. With no instructions or markings other than two tiny holes on the bottom indicating where the wiring could be found underneath the wood planking, I was very apprehensive cutting and prodding in there. I delicately started in the hole closest to the starboard spoiler. Millimeter by millimeter I carefully opened up the hole to find out where the electric wiring came from, and where the spoiler actuator Kwik-link could be found in the foam. Cutting was hard because just under the 1mm Samba wood were the very stiff and solid span-wise polyester reinforcement strips. Under that was nothing but soft foam and electric loose wires for both spoilers and ailerons, none of those could be tampered with because those wires were placed before the vacuum sandwich of the wing construction.

To minimize weakening of the wings I selected small (35,5mmx30mm) but powerful (7kgfcm) flat (10mm) high voltage metal gear servos for which a minimal opening was made in the wood/foam. The same was done for the aileron servo’s in the thin part of the wing, about halfway the ailerons, and linkage to the horn was made with a tested approximate neutral and guestimate mechanical differential throws, but the arms were not yet shortened, that will only be done after the development flights. Stringers were then glued in the corners for servo attaching, and vertical bits glued to support the 1,2mm ply cover plates. Space was at a premium especially for the aileron servo’s in that thin portion of the wing. The choice was either cutting out a larger hole to accommodate the JR connectors next to the servo, or making the hole deeper (till the wood on the top of the wing) to allow the plugs to rest in place between the servo and the cover plate, I choose for the former on the spoilers and for the latter solution for the ailerons.

Pic 741


Every Foka has rather prominent end-plates at the wingtips, these are multi-purpose preventing the over-pressure from the intrados from rolling over to the extrados causing loss of lift and drag (these are like reverse winglets), allowing a falling wing to only contact the ground on the end-plate bottom without risk of damaging an aileron, offering a tie-down hole for windy conditions on the ground, and offering a short stub at the back for a wing-walker to grab hold during ground moving (optionally a transport wheel can be inserted through 2 additional holes). On the real glider those plates are screwed onto the tips for easy replacement. On my model no provision for that had been made and the delivered 5mm ply end-plates were supposed to be glued to the wing ends that had no ribs installed.

I remembered the trouble I had inserting or removing the wings from the Fliegerland Foka from their cover bags because of the pointy aft part extending so much behind the wing plan-form. For this model I preferred to adopt removable end-plates, also for ease of replacement after landing too often on tarmac, plus to keep the transport bags smaller and easier to fit in. I therefore resorted to PU glue flattened M3 Tee-bolts to the foam in between the Samba layers at the tip, and using 5mm copper tubing inserted and glued 10cm into the foam to help support the end-plates. The ailerons had to be sanded a bit because the assembly was so tight it rubbed against the plates. The 3mm diameter copper handhold tubes were then inserted through long drilled holes in the back of the plates. Using light glue, they will be easy to replace because in normal conditions they will hit the ground first after a wing drop. The overall result is a practical and useful scale addition that is as prominent on the ground as in the air.

Pic 738


After the hardware was installed I removed everything, including the close-cut spoiler caps, and applied primer on the wing surface. The concave intrados Samba sucked the primer inside the wood, solidifying it but filling the deep nerves as much as the ridges, necessitating serious sanding and a second thick coat of primer to flatten the surface. Sanding took away much of the primer but quickly clogged the paper whatever the grit, because the primer hadn’t completely dried out in 12 hours. After that, 3 coats of orange were applied to obtain a nice equal color over all wood and equal to the one on the polyester fuselage. In the end I used following grits between all layers: 120, 180, 280 and 320. With the limited space in my room and the single trestle, I could only do one wing side at a time so all the previous had to be repeated 4 times, with adequate drying time between layers, and lots of sanding dust that each time had to be cleaned in the hobby room before the next coat application (January is too cold for sanding outside or in unheated garages). I finished by applying a fourth layer to the top of the wings after another thorough sanding sanding to eliminate all profile imperfections. For the complete model I used 800 centiliter of primer and a full liter of lacquer. The finished wings with all hardware installed weigh 1972 gr each.

Part 4: Final assembly

In between the drying I painted my pilot and various bits and pieces such as the control horns. The pilot was then stuffed and firmly attached to his seat with a scale looking parachute harness but not with period seat belts because I couldn't find the correct beige colored straps nor click system buckle (yet). As the turn buckle isn’t functional and I want my pilot to be removable for use in another glider or temporary replacement by children’s puppet/pets figures, I permanently fastened some straps from the turn buckle assembly to open ends, and others from the seat frame to the buckles to tighten the straps. Both are then joined like in real gliders and tightened. In normal conditions the, this pilot or any other puppet will remain strapped to the seat assemble and removed/inserted as a whole when the model is assembled/disassembled on the field or work has to be done on the electronics or receiver underneath.
Sunglasses from another crashed pilot were then hot glued to the repainted head, and the loose glider hat also attached with a tad of hot glue to his head.

Pic 743


Fake knobs for canopy and spoiler operation were then fashioned from a pearl-type bracelet, painted and glued in position. An old stick was modified and bent so it could be glued to the seat pan. A fake plate was then fashioned for the rather typical rudder pedal adjustment system on the right side of the fuselage. A period speed and other limitations panel was then glued against the same side panel as per real Foka.

Six lead panels were inserted and blocked/attached in the nose The batteries were then positioned in between and MPX plugs soldered and attached for connection. As These are two identical LiFe batteries and the chances one drains the other are equal to the chances of a simple diode or sophisticated power box to fail, plus everything being high voltage capable, I made the choice to use two simple cheap electrical switches for on/off. My reasoning for this is that even if one cheap switch fails, power will continue to pass through the other battery/switch, and when charging I will note one battery has used little power and it’s switch has to be checked. I consider this simple setup as fully dual power with less chances of total electrical failure because I eliminated the common single point of failure hardware. These switches were hidden behind the ....

A FrSky G-RX8 receiver was then mounted under the seat and connected to the power and servo’s. I then mounted the horizontal stab and wings on the fuselage and performed a weight and balance of the finished model (except the instrument panel). The model balanced on the recommended 114mm CG with 6 lead plaques and one battery, resulting in a flight weight of 11,8kg an thus a very acceptable wing load of about 89gr/dm². With the addition of the weight of the second battery, power switches instrument panel, pitot tube and battery/airspeed sensor modules, an equal amount of weight can then be shaved off that 6th lead plate without having to balance again. Even with a cockpit and pilot, 2,2kg of additional weight was necessary for ballast in the nose. Here is a picture of the weight and balance during the first final assembly of the 99% completed model, the shape of the cradle clearly shows the exaggerated fuselage width.

Pic 744


Looking back at the model and its completion I must say that for a mere 2000 euro all in I ended up with an impressive and attractive brand-new model that due to the extensive prefabrication by JPV, was ready for flight in about six (intense) work weeks (about 250 work hours). Although I am very satisfied about the quality, fit and finish of this JPV kit, I am not about the “wide -body” shape of the fuselage. The peculiar long narrow canopy shape is impossible to achieve on the much too wide fuselage and seriously spoils the scale look of the finished model except when viewed in profile. Following that I do not regret my decision to choose for the nose full instead of the long sliding more scale version. The nature of the Samba covered wings requires good filling out and sanding before either modern cover materials or paint is applied. No cockpit interior at all is suggested nor included leaving it up to the builder’s desires and imagination to produce one or not, but as it is very visible under the huge canopy transparency, I think is is well worth considering.

Part 5: Maiden and development flights

Delayed till favorable spring weather, please come back later for that report.
Apr 08, 2019, 04:33 PM
The sky is the limit
BAF23's Avatar
Thread OP

Flight testing


Part 5: Maiden and development flights

After I finally got the correct scale instrument panel from https://scale-cockpits.at I carved a suitable pedestal from a block of foam and after installation was able to fine balance the model to the builder’s specs.

Pic4876


On april 6 2019 the weather was fine and after a range check plus buddy model-check by another experienced pilot, I was ready for the maiden at the TMV grass runway. I first performed a couple of flights with my easy to fly Flair Ka8b to check my new FrSky transmitter operation and programming. When satisfied about the electronics and my rusty post-winter fingers functioning, I assembled the Foka but intentionally didn’t mount the wing end-plates yet in case of landing or rolling out outside of the prepared surface.

pic5152cr8


The moderate wind was blowing fully cross which necessitated an angled takeoff run and diagonal landing. After gathering my courage, I connected the towline and joined Eric, the experienced Rans S19 towpilot. Because he quickly veered into the wind during his ground-roll I experienced difficulties keeping the wings level during the first meters and was lucky they didn’t catch the soil. Once airborne, everything quickly stabilized and I was able to follow his first turn without difficulties. Halfway the second turn, the towship experienced a momentarily fail-safe receiver condition, audible through the engine retarding momentarily, and visible by the towline being released at the Rans’ side. He already had lost another line in similar conditions a few tows earlier. Although it was my maiden flight and we only reached half the briefed altitude, I kept the towline and dragged it till overhead the field before releasing it.

By that time, the remaining altitude was too low to execute a CG-dive test and I just checked out the controls and spoilers, then performed a clean stall before I circled down to enter the pattern. My initial feeling was that I had a very stable glider that was sloppy on the controls and reluctant to stall. I joined downwind a little higher than required, knowing I had relatively large spoilers to help me down after crossing the tree line. My final was stable as a rock, and the touchdown smooth. With the main wheel well behind the CG, the large skid immediately dragged the Foka to a normal stop, brakes being unnecessary with such design. As the bystanders applauded for the successful maiden, I resumed breathing and was very happy that the unexpected tow-incident didn’t affect the rest of that critical maiden flight. You can see a movie about that maiden in following link.

Foka4 Laurence Maiden 01 (3 min 0 sec)


I invited our group for a celebration drink in the bar where we talked about the fail-safe problem of the towship. We agreed to reprogram it so his hook wouldn’t automatically open, we were only two glider pilots that could quickly disconnect if we saw trouble. As my maiden had been too short to analyse control responses well, I mounted the wing end-plates and prepared for a second try without altering anything to the transmitter nor glider.

Pic5163cr8


In the meantime the big ASW28 got towed without a glitch and I collected the cable to hook it up after the landing of the towship. That flight was not filmed but during the climb we experienced again fail-safe behavior of the Rans, twice very shortly after each-other. This was audible but as we were sufficiently close to see the cable not developing too much slack, I kept the connection and we agreed to continue the climb. That flight I released around 400m and I grabbed some good thermals that allowed me to climb even higher, demonstrating the excellent thermal capabilities of that Foka thanks to its HQ2.5 profile tapering from 14,5% at the root to 10% at the tip.

As I was more interested in finding out its behavior, I positioned for a first CG-dive test with 30° angle, from which it quickly auto-recovered. I re-trimmed the glider for a little higher speed to counter the turbulence and re-positioned for a second test, with 45° dive angle. This time the recovery was much slower but still positive, proving the designer’s suggested 114mm CG was right where it should be (for my type of flying). Next I undertook some more stalls but it took a long time to decelerate and dip the nose, even with the stick almost full aft, no danger of stalling in final with this baby. I then executed some full aileron turn-reversals, observing little adverse yaw but a roll-rate that was about the same as my big lazy Ka2b. That was unacceptable for an aerobatic glider and confirmed my fears of lack of mechanical throw range during the model build. The best way to find out the effectiveness of the elevator is to perform a looping. The glider passed well but required much more stick deflection than my other aerobatic birds, especially during pullout. I then performed another looping immediately followed by a split-S and that gave me sufficient food for thoughts concerning control responses. After staying up for about 10 minutes I abandoned the thermals and entered the traffic pattern for another textbook landing.

As the tow pilot had changed the antenna choice in his transmitter and performed an uneventful test flight without towing anybody, he asked if somebody wanted to hang behind him to duplicate the previous troubling conditions. I obliged but asked him to wait a bit because I followed Bert’s suggestion of reducing the electronic aileron differential, which I lowered from 50% to 20%. The tow to 450m was uneventful and that flight got filmed again. I needed the altitude because I really wanted to further explore the Foka’s flight envelope and characteristics. A series of loops, split-S and a dished-out barrel roll confirmed me the roll rate had augmented, but still was insufficient. After a normal stall I also performed one with full spoilers, then setup for a spin. All these low-speed maneuvers took long to attain, were then very benign, and promptly recovered with no vices and a minimum of altitude loss. Even with full up elevator and full left rudder deflection, the model recovered from the spin by itself because the nose quickly fell through. Last thing I wanted to check was rudder effectiveness. I therefore setup for stall-turns but during the first one I let the speed drop down too much before applying full left rudder and despite a huge rudder throw, I never saw any yaw, just the model falling backwards and recovering by itself. During the second attempt, I immediately applied full left rudder when I reached 90° pitch, but although this time I observed some yaw, it again almost simultaneously fell over backwards with a crooked recovery till the rudder got neutralized. I figured the enormous sweep-back angle of the tail caused nearly as much pitch as yaw effect during deflection, and I abandoned the idea of performing stall-turns with the Foka.

By that time I only had little altitude left before I had to enter the pattern in downwind. The crosswind was getting stronger and pushed me closer during downwind, requiring me to skip the straight base-leg and make one continuous final turn in which I already popped the spoilers out to loose altitude. Everything looked fine but getting in closer I noted I was going to end up short and retracted my spoilers again, hoping this would be sufficient to touch on the field. Unfortunately although the speed seemed more than sufficient, I saw the nose drop a bit and pulled back slightly on the stick. I didn’t dare to make large corrections so close to the ground but should have. The sloppy elevator control (as noted during the aerobatics and stalls) had no effect until pulled seriously back, so the model plunged and bounced into the high grass just a meter short of the edge of the field. I then released the aft-pressure and allowed my Foka to settle gently as the speed decreased, only pulling the spoilers when on the ground. That was an anti-climax after 3 nice flights and meant it was time for reflection back home. Watch the movie of that third flight here:

Foka4 Laurence Test flight 03 (5 min 52 sec)


That same evening I analyzed the days’ flights while looking at the images and my settings on the transmitter. I remember having shown doubts about the constructor’s suggested limited elevator throw, but as I was afraid my twice larger throw might result in a very nervous elevator during the maiden, so I had dialed in 50% expo. That obviously was too much, causing very sloppy pitch control unless applying full stick deflections. I decided to keep the elevator throw but for a first step reduced the expo from 50 to 20%. As the reduction of ailerons differential from 50% to 20% had improved the roll rate without being too twitchy nor causing excessive adverse yaw, I took away all the differential but dialed in 20% instead of 5% mix into the rudder. All those changes will hopefully result in faster stick responses and a flight behavior representative for an scale-aerobatic capable glider. Further info will follow after the completion of my test program.
Last edited by BAF23; Apr 08, 2019 at 04:40 PM.
Apr 19, 2019, 04:43 PM
The sky is the limit
BAF23's Avatar
Thread OP

further flight testing


A week later was the first BiGGS gathering of the year at the Pampa BAF shooting range. The weather was terrible, we even had snow showers on Saturday afternoon and an ice-cold relatively strong wind all weekend long. Only a few die-hards ventured in the air and as I wanted to experience the changes I had made to my Foka, I joined the short line and hooked on behind Eric’s Rans19.

pic pampa9


The strong wind and thermals made the air very bumpy and I had to work to keep her steady in the tow. After release it quickly became obvious that the changes I had made to the transmitters were a step in the good direction, the Foka reacted better to the controls. I made a super soft landing but then retreated to the clubhouse to warm my hands. After a quick lunch I got up for the fifth flight, this time going for some aerobatics. The elevator felt better but the roll-rate had not improved significantly. After another good landing I immediately hooked up for the sixth flight. No aerobatics anymore but good thermals. So good that I had to fly two full minutes with full spoilers to stay below the altitude of the nearby airfield parachutist aircraft passing by in a climb. After ten minutes I was getting cold and felt the wind picking up (some saw dust-devils in our final approach. That was enough testing for the day and I even didn’t need spoilers during my pattern and final. After another good landing (even with canopy unlocked and a broken wing-endplate) I took the model to the camper and called it the day.

Once home I reflected on what I had seen and felt, and concluded that only major corrections to the mechanical freedom of the flight controls would help me attain the maneuverability I expect from this aerobatic glider. With cloth aileron and elevator hinges under the Abachi top of the top surfaces, I had to very delicately cut away half a millimeter of the wood forward and aft of the hinge, just a tad too deep and the hinge would be damaged. The result was well worth the effort, the aileron now moved freely 45° up instead of the mere 20° before. Towards the bottom it was even worse but the problem was different. The deep recess had to be cut at a wider angle along both the aft of the wing and the leading edge of the aileron, again keeping in mind the thin hinge cloth at the bottom. I opened up the sides till I got about 20° of down instead of the 10° before. I then applied the same procedure to the other wing and also applied the same technique on the much thinner elevator. Of course I needed to repaint all the hinge area’s with lacquer paint no overlaps are visible to the naked eye.

I also noted that after even smooth landings or ground transport, the canopy became unlocked. I therefore removed the pin mechanism by one with a longer pin. That also made me reflect that if ever losing the canopy in flight, I could also loose the pilot on his seat, because the latter is only held in place by gravity. I after a canopy loss the model would hit a turbulence, chances are that pilot and seat would also fly loosely through the air, and the model with insufficient nose weight would be uncontrollable. I therefore also installed a pin mechanism to lock the pilot seat in position in the cockpit. As the battery cables were too visible in the cockpit front, I adapted (extended) the battery box to better enclose everything and painted that interior again, totally hiding the cables from view except the small balance cable inserted in the voltage sensor for the telemetry.

I remember that during a tight ground turn, one of the wing fence copper hand-holds had digged into the soil. I saw no apparent damage but after landing I saw the aft half of the plate dangling. Those end-plates looked very scale were were impractical on the field. I made two new plates by sandwiching a 1mm GFK plate in-between 2mm plywood plates, firmly glued with PU after etching all surfaces. All holes were then drilled and the plates painted white, but no copper tubes were added at the back. I will use those new strong plates for daily flying, but also have the scale plates on the transport plate for occasional display purposes, both in the air or on the ground.

I then assembled the model and reprogrammed my transmitter with aileron differential, expo on ailerons and elevator (both having now twice the throw as before), and eliminated most of the aileron-rudder interconnect. I won’t mention any values at this point because I’ll need a few more test flights before I find the optimal settings. At the end of the setup I discovered that the rudder servo had a tendency to oscillate badly around it’s neutral point. I replaced the servo in the rudder with an identical one but the problem persisted. The phenomenon happened as well using the transmitter as when using a servo tester, but only with the servo in place in the rudder. Replacing it with another brand servo was not that obvious because it had to fit in the very tight box in the rudder, and the number of teeth and position of the rotation axle had to match perfectly the arm that is solidly glued into the fixed vertical tail. To fit their position in the box, each servo you install has to be first deprived of its side-mounting arms, rendering the servo useless for other applications. Furthermore, the servo’s need to be of the High Voltage variety because of the nature of my chosen electrical setup. The third servo (of another brand) that I installed behaved better but sometimes also juddered a bit. I concluded the problem was caused by resonance of the rudder, with the digital servo fighting to keep the neutral position. After much searching I found an old but brand new German Yuki brand CYS-S0130 analog servo having the correct number of teeth in my servo stock. That cheap (14,9€) metal gear servo had been advertised as having 12kgcm of torque at 6,2v and 13kgcm at 7,4v, meaning it easily withstood the high voltage of my 2S LiFe batteries and was sufficiently powerful for rudder operation. Testing after trial installation proved everything worked normally again. Digital servo’s sometimes are no improvement at all and for most of my uses, analog servo’s perform just as well and with less power consumption. I better buy some more before all are replaced by digital ones on the market.

Pic Odoorn199

Checking the flight controls before takeoff at Odoorn

A week later I tested the rudder in flight, but the CG still was too neutral to my like. Another week later, I attended the long BiGGS weekend at Odoorn in the North of the Netherlands. The Foka performed very well and the CG dive check showed some improvement after adding 14gr in the nose. It still was erratic in pitch from tow separation till established in a steady glide, I might try a little down mix between hook closed and elevator with a 4 second slow resume after hook opening (to eliminate the pitch-up caused by higher towing speed). Aerobatics passed well, including the stall turn, proving the cheap analog rudder-servo worked much better. My flights were consistent between 10 and 20 minutes under very weak thermal activity. After about one hour of flying (in 5 flights without switching the battery off), I charged the batteries and was happy to see that the power consumption had been a mere fifth of the capacity. All landings that weekend had been very good good but I discovered a new problem. Due to the wheel having been mounted through a (tight) cutout in the bottom of the fuselage, but without any wheel box around, the fuselage filled with local grass and dirt being swung forward through the wheel grooves. Enclosing the wheel and closing every remaining orifice necessitated 7 custom produced plywood parts that had to be glued in-between the existing gear reinforcement wood and metal and the polyester skid, a time consuming affair. During assembly the last flying day of the meet, I noted one of the nylon screws holding the horizontal stab in place was crooked, this was later replaced by a metal M4 screw. As both modifications were behind the CG, some additional weight was added in the nose.



By the time of the annual september meet at Bastogne, I had ironed out most of the the model's flight settings but a week later still further reinforced the landing gear frame in the fuselage. If not touching down too smoothly, the weight of wing tended to squach the fuselage down a bit, causing the cockpit side panels and canopy frame to deform. This problem was corrected by installing solid wooden dowels vertically between the gear-frame and the wing key sleeve. I flew it again a few times at the end of the season meet in Tongeren and finalized the electronic and mechanical flight control settings to validate its soaring and aerobatic performance for the next season. That concludes the build log and the development flights of that beautiful scale glider.



LATEST SETTINGS:

Ideal CG location for me: 112mm
Elevator throw: +27mm, -25mm, 40% expo
Rudder throw: 95mmL, 90mmR, 10% expo, 15% mix from ailerons
Aileron throws: +35mm, -15mm, 40% expo, 50% differential
Last edited by BAF23; Sep 29, 2019 at 05:00 PM.


Quick Reply
Message:
Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Build Log Scratch built 65" De Havilland Mosquito Scale 1:10 - build log JornWildt Scale Kit/Scratch Built 1227 Aug 12, 2019 09:58 AM
Gallery 1/64 scale RC converted HotWheels cars build log vtolman Micro Cars 0 Apr 16, 2016 11:13 AM
Build Log Build Log: 1/10 scale DC-3 astroboy Scale Kit/Scratch Built 308 Jan 22, 2012 07:31 PM