Thread Tools
This thread is privately moderated by BAF23, who may elect to delete unwanted replies.
Aug 24, 2016, 02:01 PM
The sky is the limit
BAF23's Avatar
Thread OP
Build Log

Schneider Ka2b Rhönschwalbe scale 1:3

History and acquisition

In 1968 shortly after becoming solo on a towed Ka4 Rhönlerche, I took place in the frontseat of a Belgian Air Cadets Schleicher Ka2 Rhönschwalbe glider for my first winch launch. By that time their surviving 1955 acquired Ka2's weren't used much anymore and I was glad to have had that ride in one of the 75 Ka2b airframes that were produced between 1955 and 1957. This also explains why the Ka2b figured on my dreamlist of model gliders since years, especially when I found out that a traditional wooden kit was available from Schneider Models in scale 1:3. The size of that model, 5m33 span and a 2m75 fuselage length was as attractive as scary to me. The kit consists of only the plans and some precut wood and canopy mold, building it would require the living room table to be neutralized during about two winters because the wings and fuselage are too long for my hobby room. The existence of Ka2b OO-SZD on the Belgian register, after half a century still wearing extremely attractive SABENA (former Belgian World Airlines) colors, was also a strong incentive to “ever” possess and fly a Ka2b model in this scheme.

A fellow member of the BIGGS (large-glider association) built one in 3 years time and it flies like a charm, I was hooked but that model is rarely found secondhand and often the painting prevents an easy switch to the Belgian desired scheme. Early summer 2016 one was advertised on a popular German model forum but the price was too steep and it was near the Swiss border. The pictures showed it was very well built and except for the extreme nose, rudder and wingtips, was still in predominant white colors and reported to have made only a dozen uneventful flights.

After a few weeks the seller (he didn't build it) lowered his price and I contacted him. After a few mails discussing model details, present condition and an even lower price, he agreed to reserve it for me till I saw it with my own eyes. As he had only wing and tail covers but nu fuselage cradle, I borrowed dad's station wagon and opted for a one-day solo 1300km round-trip to get the glider home. Traffic enroute was awful due to many roadworks but I arrived at 2pm and got guided to his living room where the Ka2 rested fully assembled on the floor. It was a magnificent sight and after he showed me the details, shortcomings (mainly the warped wing trailing edges), operation of dual power module and the various servos, he took his receiver out and helped me fasten the large disassembled model for the long journey back. Transporting the model took the full available length and most of the aft cabin width of the car, thus preventing me from seeing anything through the interior or right rear-view mirrors, not exactly a safe situation driving during rush-hours on the crowded German highways.

The trip (non-stop legs) had taken me 16 hours total and I only disembarked the glider the next morning to avoid damage doing it in the dark.

Once back home I got confronted with the sheer size of this model, I even couldn't stock the wings vertically because the ceiling was too low, they ended up high on top of a large window dresser. I quickly fashioned a custom cradle so the model now permanently rests on a low dresser along the only free wall long enough to harbor the Ka2b. That is where all the work has to be done because I even cannot take it into my hobby room through the narrow hall.

Being impatient

Although I initially had no plans to start working on the model till winter, a few weeks later just for the sake of crushing down a fear, I tackled the problem of the nose end. Due to the lack of regular (strong enough) towships at his club, the seller had removed the towhook plus servo and replaced it with a ball bearing in the nose. The lead blocks also were removed and the two front frames modified to accept a Hacker electromotor powered by 2x6S5000 batteries. Before making the cart to allow the 15kg model to takeoff autonomously, he realized the large prop would complicate things and he put the model up for sale. Luckily he had kept both lead blocks, the long towhook and the actuating servo. He told me the front bearing would come off easily if I heated it up by sticking a solder iron into it, pulling it out and sliding the old hook back in. I took his word for it but during the next week I had nightmares about having to drill out the solid nosecone to get the job done, hence my desire to calm myself down by first removing the ball bearing.

To my utter surprise, within a minute of heating the inside of the ball bearing, I just could pull it out effortlessly from the nose, and the old 12mm diameter hook tube slid into the 6cm deep solid nosecone after a little persuasion. I couldn't find the place where the old servo nor the two lead blocks had been attached, and found out that I would have to live with the fact that the towhook tube was going to point slightly up instead of the preferred down-angle. At that stage I started thinking it might be a better idea to get the Ka2b ready for flight in its present condition so I could make the initial shakedown flights for CG and servo adjustments before winter, only tackling the cosmetic restoration during winter on an already proven and fully setup airframe. While searching the web for additional info I stumbled upon the page on which an Austrian modeler (the builder?) offered that very model for sale end of August 2015. As I bought it after the motorization modifications just 10 months later, my German seller definitely didn't fly much with it (but kept the cockpit puppet). Here is the link to the old illustrated add:

The assessment and some quick fixes

As with any restoration I started by making a detailed inventory of the model, its condition and vital mechanical and electronic devices. Because it is the smallest individual airframe part I started with the horizontal stabilizer. Small is relative on this model, it still measures 101cm wide with a 35cm root and weighs a hefty 703gr. That includes one servo in each horizontal stab side. Access to the latter is by removing 4 screws on a flush mounted plastic cap upon which the Hitec 5625MG servo become visible, mounted with metal plates in a solid cradle and using metal adjustable links towards strong nylon control horns. There is absolutely no play on the “holkehle” type elevator hinges nor on the actuating mechanism. The servos have double ball-bearings, metal gear, have a good reputation and produce 8kgcm torque, more than necessary for a single elevator. Both servos are wired through a single Multiplug plug with each of its sides glued into the horizontal junction with the fuselage, but with separate wiring for the port and starboard elevator surfaces and servos for redundancy. Three nylon M5x30 screws ensure this heavy complete horizontal stabilizer is quickly but solidly secured to the fuselage. Just as the rest of the model, the visible ribs are covered by white Oratex that has been very well applied and schrunck, but only given a thin white flat coat of paint. The plain horizontal stab part had a thin satiny type paint that had not been applied uniformly. There was a puncture in the port upper part (also in the transport bag) that had been patched-up with an ungainly looking Oracover piece of fabric. During the shake test I could hear something rattle in the starboard end of the fixed stab but was unable to pinpoint nor gain access to it.

The port wing weighed 2,914 kg. That includes both Hitec 5625MG servos. I had to rearrange the servo arm and the actuator for the spoiler so it could operate freely without end course buzz and with simultaneous movements on each side. For the spoilers not to get friction nor jam tendencies in their tight wing openings, I cut some wood away and sanded the flush plate on each of the four spoiler halves at an angle. Some of the extrados Oratex came loose from the middle of the wing and the trailing edge looked awful wit some serious “waves” at various places (a common occurrence on Schneider kits). As the seller told me it didn't affect the flying, I wasn't going to remedy that for the proving flights, but definitely will tackle that during the full restoration. The starboard wing was treated the same way and was also declared airworthy despite of the pronounced trailing edge problems along the inner third of the wings.

Although the aft fuselage didn't need any work, the front needed attention. After removing the seats I discovered two additional lead blocks at the bottom of the fuselage just aft of the forward stick, 348gr for the port and 402gr for the starboard piece. As they had been positioned 40cm back from the nosetip, I removed them with the idea of relocating them more forward for better weight efficiency. Also in the belly I found that the power cables to and from the Powerbox PB6310 sensor (rated two times 6Amps at 5,9v linear stabilized) had too many unsecured connections to my like. The cable towards the hook servo consisted of 3 different parts that had been joined, but I replaced those by a new ones in one piece with proper JR connectors at each end. The power cables from both LiFe 2S2800 batteries used small red connectors I didn't have, so I replaced those by Deans connectors for standardization with the rest of my 2S and 3S batteries. The balance connectors were the type for 4S batteries but with the 3 wires connected in a strange way, maybe they charged it that way? After repositioning them in a normal way I could use my battery tester to measure the capacities again. The batteries slide in a kind of polyester pocket sleeves that are suspended under the top and sides of the nose. Nothing but friction retained the batteries in place.

I didn't trust that and feared the batteries could slide back during violent tow maneuvers, the magnet-held instrument panel would never retain them and a serious CG-shift problem could then develop. I thus glued narrow Velcro straps with grip cords to keep both batteries firmly in their pockets, whatever attitude or jerks could happen in flight. All my electronics were mounted in the wing box on top of the fuselage, underneath the wing cover panel and allowing very easy access.

Cockpit and canopy

Just before she quit her business fabricating pilot figures, I placed an order with Alan Hawes for two 1:3 scale pilots (55cm long with 15cm shoulder width). I contacted Sandra to make sure I wouldn't end up with identical twins in the cockpits, but she suggested a female pilot in blue coveralls as the second pilot. That suited me well, as was the reduced price because the British pound had devaluated quite a bit after the Brexit was confirmed. Within 10 days the package arrived, for one fifth of the price of more renowned German suppliers, her puppets are lightweight, flexible and fill the cockpits well. On the other hand, facial painting and black rubber hands are not very well done and require attention from a demanding buyer. What I had not expected was that the woman-pilot head had been molded with a headset flattening her coiffure, not exactly suitable for a fifties-era scale glider. When installed in the seats I noted the latter were so scale that I would have to make parachutes in their back for the pilots to assume a correct posture, but their heads looked too small. Once installed in the cockpits, it also became clear that the two side mounted microphones had to be either repositioned or eliminated, I did the latter (no radios in most gliders until the seventies or eighties).

Unlike the real Ka2b, this canopy is made as a single unit that had to engage two pins at the rear, and is kept in place by a single pin at the front, the latter being operated by the scale canopy handles inside the cockpit (through the operating vent windows). The mechanism that actuated the pin rubbed and pivoted the magnet-held instrument panel so I corrected that. Modifying the one-piece brazed copper tube canopy frames to conform the real mechanism could be an option for later if deciding to split the front and rear canopies. The position and angle of the canopy bow between front and back seats is incorrect for the OO-SZD (too much forward) and seriously detracts to the scale aspect of the model. The position of the dual power switch under the back seat pilot's right foot along the front seat makes it a bit awkward to operate and necessitates the complete single-canopy to be removed/secured each time. Until I find pictures of the interior of the real OO-SZD, I will mostly leave the interior the way I acquired the model.

Hook and ballast

As already mentioned, the original towhook, twin heavy lead weights and a Hitec HS525BB in a loose mounting frame came in a bag with the model. No idea how/where these had ever been mounted, and that servo only produced 4kgcm of torque, grossly insufficient if you have to open the hook of a 15kg model in anger to separate from the towship during a mishap. After searching the web I decided on a Graupner DES707 15kgcm torque servo to fulfill that job. The foremost tip/bottom of the model was full of dirt/sand aft of the original? lead nose. If it ever did come in by hitting a dirt-pile it should have ended up aft of the long towhook tube. Did the model ever make a impact landing where the nose was damaged and therefore reconstructed as a powered glider? I couldn't see any damage at all in that nose so the whole situation still puzzles me, but looks solid. Awaiting the servo delivery I experimented with the 4 lead blocks and after some reshaping of the lead I found a way to stuff them around the hook in the most forward compartment. Note I also solidly welded/glued a metal ring around the hook so it cannot be pulled out of the nose anymore, and if the entire nose would separate, the actuating rod will automatically pull the hook open to ensure safe separation from the towplane. The hook itself is buried 9cm within the tube so sufficiently long nylon loops will be needed for flight operations.

Here is the list of weights that I installed at that time in the confines of the nose. fwd 850gr + on top fwd 265gr, top aft 125gr , bottom aft 230gr, bottom fwd 93gr, aft flat 830gr. That was 2400 gr of lead plus 2x250gr for the LiFe batteries all in the front 20cm of the 275cm long fuselage. The 2x250gr for the pilots and their seats are peanuts and contribute little to the total balance.

First balance

With all weights installed I took the model on the terrace to completely assemble it in flight configuration for a first balance test. The model is not only too big to assemble in the apartment, but also too large, heavy and tall for my Sig balancer. As during a large-glider meet in the North of Holland early-august I had been able to get the CG figures from another Schneider Ka2b owner (between 34 and 68mm from the leading edge along the canopy), I applied those marks with a grease pencil and used my fingers to lift the (about 15kg) model to check the CG. I was surprised that with all the already installed lead, it still balanced aft of those figures. Only after adding yet another 600 extra grams to the nose did I end up somewhere within the CG bracket. This was much more than anticipated and couldn't be quickly achieved. Measuring wing depths and interpolating with sweep angle and taper confirmed the CG figures I got were close to a 25 to 33 % of the mean aerodynamic chord and thus correct.

Looking at the awful wing trailing edge waves and realizing the repair of those and the new paint job would also add weight aft of the CG, I decided to call it quits and abandoned the idea of flying the model before before finishing the total restoration during the following months. The builder of the Dutch Ka2b told me he also rebuilt his horizontal stabilizer to make it much lighter to avoid adding too much weight to the nose. All this influenced me into stopping the balance attempts at this stage. Planned redesigns and restorations will alter present weight distribution so much that a second and third balance will be necessary before even attempting the proving flights to start refining the final CG that is suitable for my desired flight behaviors.

After taking pictures I disassembled the model and wrote a to do list about the things I would need to acquire and do to get this second-hand model to fulfill my wishes and standards. The time pressure to get it airborne was gone and I could resume flying my other models till the winter stop. Stay tuned for more restoration info when the model is completely refurbished and ready for flight again.
Last edited by BAF23; Aug 25, 2016 at 07:04 AM.
Sign up now
to remove ads between posts
Oct 29, 2016, 03:55 PM
The sky is the limit
BAF23's Avatar
Thread OP

Full restoration part 1

Tail mods

Most serious problem noticed during the first balance was the tail heaviness even with 2kg of lead in the nose. As I had no desire to add another 600 grams of dead weight, the only other solution was to put the tail on a diet. My initial idea of building another but lighter tailplane was discarded because I had no plans except for what I found on the Internet from a fellow Schneider Ka7 builder (identical tail as the Ka2b)

From that page I could click to the more detailed tailplane construction page (Höhenleitwerkbau) where I saw that without plans (material type and thickness) it wouldn't be simple to reproduce a similar but lighter tailplane of similar strength. Further digging in his build log I noticed he only used one elevator servo located under the CG, next to a rudder servo using long pull-pull cables to actuate the rudder. My model had three 60gr servos in the tail, 1m50 behind the CG. The three 2mm iron wires for the elevator and rudder hinges were even more aft and together weighed 37gr. All this was a bit of an overkill and every 100gr in the tail had to compensated by 200gr in the nose. Shaving off 100 gram in the tail thus meant a total weight reduction of 300 grams for the model. That is why I do not understand the tail had to be this massive, it was built as strong (and heavy) as a tank.

I wasn't sure how much weight I could gain by moving those tail servos under the wings and then requiring long Bowden cables or pushrods to the tail surfaces. Furthermore, I liked the way the horizontal tail could just be bolted on in the field without any electrical or mechanical connection to be made (the rigid MPX connectors just joined). Easiest way out of this dilemma was to physically reconnect both elevator halves and simply take out one of the elevator servos from the horizontal tail. Replacing the steel 2mm hinge wires by carbon ones also saved 25gr weight without dramatically reducing strength. The redundancy of having 2 separate elevators on two separate servos and channels is lost, but that is a risk I am willing to take with strong quality servos (just as the two other Schneider Ka2b builders/pilots I know also did).

I removed the ungainly patch that had covered a hole on the upper port side of the stab, discovering that a more shiny layer of paint was below the satin one. Apparently the model had been repainted at one stage or another, strange for a model claimed to have flown only a dozen times. I used PU glue to close the hole and when dry, the Dremel to get rid of the paint around the repair area so it could be made flush again by filler. To my surprise the Dremel kept on scraping deeper in thick solid white stuff before encountering the 2mm balsa I had seen around the hole before the repair. I decided to switch strategy and use sandpaper on a block to clear the whole area around it of excess paint, but first weighed the (now one-servo equipped) fixed horizontal tail at 449gr. Besides a lot of dust, the sanding produced little result, even using coarse 60 grit paper. I have no idea what was used (only over the upper half) but it was very thick and solid, kind of two component epoxy or so. After sanding less than a quarter of the top surface but not yet completely to the balsa planking, I measured only 10 grams lost, a rather disappointing result after all the elbow grease I sacrificed. By the time that horizontal stab was ready for priming and painting, the scale indicated only 408gr, another 40 gr had come off.

During those works I incidentally found out that seen within the interior of the starboard elevator servo, the horizontal leading edge former between the ribs was cracked horizontally. That could only have happened during a mishap and probably was the reason why that tailplane had been repaired so rigidly (and heavy) and why something is still rattling inside the hollow balsa structure. The individual elevators were kept as built but joined by way of a 5mm plywood block to restore the former leading edge integrity, and a 3mm plywood plate glued over the center section of the elevator's leading edge former carry-through beam. The complete horizontal tail before the finish tipped 570 grams, meaning it lost 133 grams since I started. More weight was then shaved sanding the vertical tail and from all around the aft fuselage. PU glue was then used to close the gaps between the aft tail vertical beam and the cover plates that hide the rudder hinge mechanics. Most of it had been poorly glued resulting in minor cracks along the overhanging thin plywood. The rudder then was completely stripped from its red Oratex material, filled and properly prepared for covering with new white Oratex to obtain a uniform base for the blue paint.

Interior fuselage works

Before touching the cockpit I took out all of the removable lead to gain access to all the empty front and bottom compartments of the nose. Using a long hollow tube I fed 200 grams of steel blaster balls for guns into the compartments and then held into place by allowing PU glue fed through a drinking straw to inundate the remaining space. Other small lead pieces totaling 70 grams were used to fill the smaller compartments in front and along the tow servo. The complete bottom of the nose is now one solid piece of material, greatly enhancing strength in case of high descend-rate impacts. There is still sufficient space to allow further lead to be added after the second balance attempt. I removed the long tube for the vario compensator and Dremeled off its copper attachment, the same for the side mounted old MHz antennas.

Still not being completely happy with the weight reduction I looked deeper (you can take that literally) into the aft fuselage and made up my mind to alter the rudder servo setup. I removed it from its tail position, as well as the heavy actuator arm, and pulled the servo wire out of the guide tube. That was another 86gr out of the tail. With only one elevator servo there was no need for two servo cables running the length of the fuselage so I cut it close from the MPX plug and that saved another 23gr. With only one servo cable to the back there was no more need for the 125cm long 12mm diameter plastic guide tube that had been installed on top of a long wooden stringer. Taking those out saved another 81gr. Those last changes cut another 200gr on top of the 133gr I already had eliminated from the horizontal tail.

Because the backseat pilot wears a parachute in his back, I figured I could move the aft seat forward a few centimeters, thus allowing space to mount the rudder servo on a bracket just behind the seat, but still having it accessible if necessary. The single arm of the Hitec servo was exchanged for a double arm, to which plastic covered steel pull-pull cables could be attached. These weighed only 10 gram, including the adjustable connectors. They were pulled through the aft fuselage and connected to the rudder that also had been modified with a CFK custom cut control horn sticking out on both sides. My initial thoughts about a simple cosmetic restoration had again evolved into drastic modification in both elevator and rudder controls, but were necessary to shave off weight from the aft fuselage that would get heavier when straightened out (more of that later). Following picture shows the extent of materials that have been either shaved away or discarded aft of the CG.

Primer ready 3

I wonder how that glider had flown before, the provided lead was much too light so I suppose the model was flown with an extrema aft CG by the builder. Probably the seller also found out that he would have difficulty balancing the model because his motorization blocked the forward fuselage for lead or his heavy batteries, all of which would end up in the forward cockpit, a foot from the nose and thus much less efficient to counterbalance the heavy tail assembly of the builder. I think all that is the reason he didn't fly the Ka2 and abandoned the motorization efforts, selling this nice model for a realistic price knowing that a lot of work still had to be done before getting it airborne.

Exterior fuselage preparations

When I first saw the model in real I was shocked by the “mattress” effect along the complete aft fuselage. After using some filler to flatten out the areas where the German registration letters had been removed from, I sanded the complete fuselage to remove at least the shiny part of the old paint. That wasn't as easy as expected because just as on the full size Ka2, the complete fuselage had been capped all-over by thin plywood around the formers. I suspect the builder later used dope or so because most of the plywood wasn't lying flat anymore but kind of sucked into the fuselage in between every former and stringer. Not only did it look awful but it also made filling and sanding extremely difficult and time consuming. Because the thin fuselage panels were very flexible a traditional hard filler was no option (it quickly would show cracks as it had on along the cockpit area) and I opted for flexible paint filler that was applied in seven subsequent layers that every time had to be sanded carefully without pushing too hard or the flex of the panels would result in convex instead of concave bulges between the internal formers and stringers. It took me ten days and much elbow grease (with a long straight sander) to get an acceptable result on the fuselage, using loads of sandpaper to get rid of the 1kg of applied water based primer, hopefully reducing much of that weight by evaporation and white dust that covered my terrace in between the high pressure Kärcher cleanings. In the end I got it sufficiently straight fore/aft and nicely rounded up/down as to start applying layers of white lacquer. If I wouldn't have done that, the shiny white would even have accentuated more the “waffle” effect. In-between the drying periods I already applied a few coats to the vertical stabilizer to check how well it covered and how subsequent layers changed the type of blue in the sun or shade.

Primer to paint fus 2

The fuselage under the wing and up to the nose had its own specific problems. There were multiple cracks visible in the paint, but luckily after sanding I discovered not many cracks in the plywood. I suspect the application of very rigid paint over the not so rigid fuselage plywood had caused those superficial cracks to develop during transport or manhandling. This proved that a form hugging cradle was essential for this model, during assembly, transport or storage. The previous picture demonstrates how far forward the cradle had to be placed (with the lead in the nose and absence of horizontal tail and rudder). It also confirmed selecting a a non-rigid type of primer and paint for the complete model was essential. Some of the real cracks were closed by PU wood-glue and later equalized with the rest of the rather faceted fuselage. The nose was another story. I didn't want to see flat faceted surfaces on the nose so after sanding away all the (thick) red and white paint I had to apply multiple coats of filler/primer on top of the existing two filler types, producing quite a patchwork pattern before I could continue. The Dremel and metal hand-file were then used to get a perfect contour around the (non-scale) tow-cable orifice.

Post sans 5

Fuselage finish

After a couple of coats of filler followed by sanding, I got closer to the rounded shape of the nose and filling of the deepest fuselage depressions. The filler was not that strong so for the final preparation I applied several coats of thick water based Acrylic primer which were each time almost completely sanded down with 100 and 180 grit paper until the smooth contours and surface were obtained. That took another full week and a fortune on sandpaper. The nose needed several layers but weight was not a problem there. They created a very smooth and hard base for the paint, but required hours (I think about 15 in total) of sanding with 180 grit paper on a sponge so it could follow the complex curves of the fuselage from nose to tail. All this sanding was performed on the terrace during a pleasant September month. It was hard resisting to go fly but I definitely wanted all this to be over with before colder temperatures came, the mess (white powder) I produced on my garden table, seats and terrace were the silent witnesses that winter is not always the best time to build/restore, and that when living in an apartment and practicing R/C flying, different priorities have to be envisaged or things could get postponed till spring (if wanting to keep the interior free of excessive dust).

During the drying periods I ensured to stay one step ahead by trying out the finish on the smaller horizontal tail and elevator. Small is relative with this 1:3 glider, the span of the tailplane is almost the same as the wingspan of my FMS Stearman, and the surface area is much bigger of the one of the biplanes both wings. The tailplane also got the filler/primer treatment and Acrylic lacquer paint was then applied with a brush. Mind you I used minimal treatment on the underside of the tailplane to save weight, but the topside had to look good. The paint layers also were sanded smooth with 320 grit each time, and all layers were applied with filth rollers. This gave a smooth gloss finish, also on the fabric elevator. All this meant I spent hours each day cleaning brushes, paint lids, roller applicators, paint immersion basins, fingers and other things whenever a new layer had been applied somewhere. All this had to be done in and around the kitchen sink between the preparations of my meals!

Final fuselage 17

The freshly covered rudder already got blue paint applied to check the color, but as other modelers before me found out, that SABENA blue is very hard to reproduce. The available pictures of any SABENA airliner all differ from each-other due to age or sun angle. SABENA used a few varieties of blue throughout the decades so I tend to rely on my memory to get the rich royal blue as used in the 60's and thus probably was applied on the OO-SZD when it first was painted during that era. From the pictures it looks close to RAL5005 signal blue, but on the model it looks a bit darker. Of course the blue was even more sensitive to application than the white, and three layers were necessary to obtain a uniform smooth glossy surface (as on the full size Ka2b) on the fabric.

As autumn weather arrived and the lacquer application to the fuselage had to be protected from dust and flies during the slow drying process (outside of the sun), the rest of the restoration was performed in my apartment. With the fuselage pointing down at a 45° angle and the tail up in the toilet room, I was able to barely maneuver the fuselage into my hobby room where it could be rolled upside down, but not turned around anymore. To obtain that smooth deep finish over the fuselage I applied 5 layers of white paint, sanding away most of it after each layer (with 220 grit paper), but thereby also filling all the remaining primer imperfections and building-up a uniform white layer over the various undergrounds. With paint applications only possible after 8 hours of drying (and subsequent sanding) it was another week of sanding with 320grit paper, lots of white dust and elbow grease, before I had a satisfactory glossy white fuselage that had to dry 24 hours before I could undertake the masking for the blue accent stripes. By that time I also trial-taped the Caliegraphics produced stickers so everything could come in place for a scale look.

Primer to paint fus4

With the thicker blue line tapering towards the tail and the curves the masking tape had to make on the nose and in front of the canopy, application had to take place with attention to detail unless paint would creep under folds in the tape (Tamiya narrow tape worked well for this). Using two thicker blue paint layers within 8 hours produced a uniform blue color and allowed me to peel off the masking tape an hour after the last layer, just before the paint would solidify over the tape, thus decreasing the chances of peeling off undesired overlapping paint. During those times another two layers of blue were applied on the vertical tail and newly white Oratex covered rudder, after which the Callie classy SABENA emblems were applied. All this worked out very well and hardly needed any corrections (with a scalpel). When I finally applied the last markings according to the pictures of the full size glider, plus affixed the horizontal tailplane assembly, even I was astonished by the attractiveness of the restoration result. Note that with the horizontal tail installed, the cradle has to be moved a substantial distance to the rear.

Primer to paint fus 9

canopy and interior

The clear canopy had a very small crack at the back and both complex PET assemblies fitted against each-other well but had not been glued together. Canopy glue quickly resolved those minor deficiencies, but when I used HG glue remover to get rid of the remains of the clear tape that had been used by the previous owners to solidify the cracked area, it rendered that area opaque because it melted the clear canopy. I quickly stopped and was able to polish that area clean/clear again with the Dremel and buffing products. It concerned only the last centimeter before the wing, but that had to be completely translucent because most people would look through that canopy section to see the incredible scale elements of the wing to fuselage assembly.

Next problem was more complex because it involved the fake separation between the forward and aft canopies which had been completely mismatched up by the builder. The way the model had been built did not allow for the aft canopy to be hinged up as per real Ka2/Ka7 and the transport restraints (fuselage height at that point) restricted me in modifying that basic design. The canopy still being an uninterrupted molded clear assembly was acceptable to me, but the rake and position of the copper inside frame was too far off and grossly spoiled the overall scale look of the model as seen in the opening pictures of this blog entry.

The clear canopy had been bolted and painted to a copper tube frame so removal would be tedious and prone to fatal damage. The sliding vent window positions couldn't be altered anymore. I therefore opted to insert a protection between the badly positioned copper canopy bow and the clear canopy, before using the Dremel grinder to separate the four bow welds from the horizontal frames. New holes were hand drilled in the base frame tubing to ensure firm connections (with PU glue over dowels) between the more aft position of the bow with increased rake, as far as the existing aft vent window allowed. After painting all interior framing white (instead of the previous gray) I used double white narrow tape to reproduce the fake separation of both canopies. Applying it was very difficult because the tape kept on bubbling over the top part because of the awkward curve. In the end the result was much more acceptable as when acquired, convince yourself by comparing the pictures at the start of this thread between the model and the full size glider, and what I was able to reproduce in following picture.

Final fuselage 15

The interior also needed work because the instrument panels of OO-SZD were much different from the ones of D9051. Thanks to detailed pictures I got from the owner of the real glider, I could see that over the years the shape of the instrument panels had been completely changed to accommodate higher knee positions (during stepping in?) at the front, and inclined a bit at the rear, both had undergone a complete change in instrument layout during the 2004 restoration. Whilst I could make a new front panel and rearrange the instruments I had, the full size glider still had that immense inclinometer with ball and fluid reservoirs, typical for 50's era gliders. As it was so prominent I had to reproduce it completely from scratch, using machined wooden parts, fuel tube bent in a bow and a steel-ball gun-bullet. The pale leather pouch/map holder of the real one was also added and the overall result looks stunning and adds a lot to the period/scale representation of the model.

All the rubbish behind the instrument panel will mostly be invisible when the pilot legs fill the open space in the picture). My new panel is not as long as OO-SZD's, but I was restricted by the need to hang it up solely on 2 magnets in order to get easy regular access to the LiFe batteries behind the panel. It also is the reason that I didn't reproduce the rudder pedals, they would interfere with the towhook mechanism and installed lead weights. Luckily during the 2004 restoration the prehistoric inclinometer had beenwas replaced by a more modern needle and ball for the backseater, so that was added to the 3 existing period gages on the model. The aft panel shape is not 100% scale but a modification of my old panel was much easier and would not detriment to the period looks. After a first installation I removed the panel again and modified it so it could be mounted at an angle as per real OO-SZD. As a detail I added the expiration date of 17/7/2011 as seen on one of the panel pictures.

Wingjob 29

Seats were taken out, repainted as per original and gray plastic seatbelt buckles painted metal. Removal of all the modern grainy gray paint in-between the frames and formers in the cockpit was not an option, but because after removal of the side microphones I had holes in the top vertical planking, I decided to fill the holes and sand those side panels flat and repaint them in lighter gray, and did the same to the vertical part of the seats against which the parachutes rested. With two pilots installed, the rest of the (deeper) fuselage would be mostly hidden so this shortcut was acceptable to me. The functional scale fuselage wing attachment was also painted white as per real, but the attachments on the wooden wing spar were left in natural wood, not as per full scale but it looks soooooo much natural/older/better.

wing job 20

Speaking of pilots, I had ordered 2 from Alan Haynes the month before they quit business. I didn't want twins and negotiated for a male instructor in the back and a female in the front (wearing a lightblue Sabena groundcrew type flying suit). Their 60cm length was OK for 1:3 scale, but their body and head proportions left much to be desired and had the same exaggerated slender body looks of Barby and Ken dolls. Furthermore the female wore a headset over strange-looking hair. During the 2016 glider meet in Bastogne I had my own head scanned in 3D but the guy had so much work I only got it in 2017 to be grafted on the Haynes body. Both bodies/clothes also had to be stuffed to get the proportions more correct and fill the cockpits as per real Ka2b. Parachutes also had to be made (from foam and old flying suit remains) otherwise the figures wouldn't sit correctly in their seats. All this might read like nitpicking but at this scale it is really important to get the interior right in those large cockpits, most model gliders have either an empty seat/cockpit, or a much too small pilot figure in it, much spoiling the looks when the model comes in to land or is seen static on the ground.
Oct 29, 2016, 04:59 PM
The sky is the limit
BAF23's Avatar
Thread OP

Full restoration part 2

Tackling the wing trailing edge problem

Although the previous owner assured me that the waving trailing edges of the wings didn't affect the flying qualities, I just couldn't look at that awful sight anymore and decided to correct it. After removal of the Oratex on the underside of the affected part of the wing I saw that although the end of the ribs were still perfectly aligned, the very thin wooden strips that were applied top and bottom capping the rib-ends had found a way of their own, resulting in very undulated trailing edges.

Wing job 7

With the ribs going all the way to the back and tapering balsa triangles solidly glued to everything, the obvious solution of a carbon strip all the way through was impossible because it would weaken the structural wing rigidity. A trip to the LHS (local is relative when you have to go to Germany to get a good choice) was all that was needed to purchase balsa preformed trailing edges of the appropriate dimensions (15mm wide and 4mm high). After protecting the still intact Oratex of the top (now on the bottom), I cut the new balsa trailing edge in lengths that could be pushed in-between the undulating planking. After trial fitting of each, I applied a liberal amount of PU glue to those trailing edges and pushed them into the trailing edge voids after prying them open with a screwdriver where necessary.

Wing job 4

Excess glue was immediately wiped away and the masking paper removed before the glue could get a grab on it. This was essential because even when painted, the white Oratex shows through sufficiently and any spillage would be visible on the top wing cover. After all parts were firmly in place, solid straight metal or carbon pieces were clamped on top and bottom along the treated trailing edges so the glue could dry with the formerly deformed strips solidly attached to the new straight inside trailing edge guides.

wing job 14

After 24 hours of drying I was glad to see that everything remained lined up even after the guides had been removed for a day. I thus didn't need to glue additional carbon strips to the back of the trailing edge balsa, but after covering and painting the wings, the clamps stayed on for weeks to help alleviate the long accumulated wood strain and to to ascertain that the reshaped trailing edge wood will keep and maintain its newly acquired straight lines for years to come. The second wing came out better than the first one because I didn't push the strips all the way in. On the first wing this had caused the less solid top cap strips to slightly bulge under the pressure.

wing job 16

After much preparation of the scuffed wing surfaces I used one coat of white lacquer on the underside to primarily fill the Oratex so dirt couldn't accumulate in the texture anymore, and two coats on the top to obtain the shiny finish of the full-size OO-SZD. All this was done in the limited confines of my hobby room, no sinecure for such large surfaces. By the time all the white had been applied on the nearly completed glider, the one liter water supported acrylic lacquer pot was almost empty.

wing job 19

Masking the wing for painting the blue stripes and panels was not simple because the pronounced bulge of the extrados made it difficult to make straight lines perpendicular to the airstream on that 7° forward-swept wing. Finally trigonometry, use of eyeball Mark one, patience and lots of realigning and pressing down on the fabric and depressions got the job done. I did not want to clean paint rollers 8 times so I used a blanket between both staggered wings (leaving the outside panels hanging in space) and squeezed the ailerons in between so everything could be covered in one go (with me applying the roller lying on my back below the wings for the bottom part). 8 hours later a second coat was applied and 2 hours later the orange tape was carefully removed before the blue completely dried out. That was especially necessary for the 5mm outside stripes flanked by wider tape, allowing that to dry completely most probably would have resulted in portions of the thin line to just lift from the shiny white wings (impossible to slightly rub the 5mm between the masking tape for the paint to adhere more).

wing job 21

There were more corrections to be made as anticipated, largely due to the Oratex/balsa joint ridges or forcing the masking paper in a bend to obtain the required straight line on the bulging surface. Painting it over was no option because it would show too much so a couple of hours of delicate scraping with the sharp point of an x-acto knife to remove the unwanted blue did the trick. The wings could then be fully assembled again and servo arm angles of both ailerons and spoilers changed for better throw symmetry. This again took a few hours because the limited deflection of the ailerons (caused by the minimal gaps between ailerons and wing) made it difficult to get some (differential) throws without causing the strong servo's to jam halfway their throw. Serious considerations and experimentation about mechanical versus electronic adjustments had to be made before an acceptable result was obtained, but the roll rate will be sloppy at the most, and rudder use necessary to augment this in anything but smooth weather.

I was anxious to check how much weight penalty the painting of the wings had caused. As a matter of test I weighed the flight ready port wing before doing anything to it (2914gr) and after the full restoration (2985gr), only 70gr increase for the additional wood and glue for the trailing edge repair and all those paint layers (over a 180dm² area) was a positive surprise. It proves that the acrylic paint carried by water weighs a lot in the pot but most of it is water that evaporates during the drying process (which can take up to a week before it all evaporates). That is also the reason I didn't hesitate putting 5 layers on the fuselage because most of the weight either was sanded away or evaporated during the drying process. That is unfortunately the only weight comparison I could make because I already added much lead to the nose before I started the fuselage restoration and extensive filling of voids.

Final assembly and balance

After working for 3 months (between 6 and 12 hours each day) I was impatient to see the model assembled for the first time in its new colors. It were the last days of October and the fog hardly lifted so I looked at a way to do it in my apartment. By moving a heavy relax seat, a few chairs and a barstool, I saw a single possibility to assemble this large glider with its tail in the kitchen, port wing in the dining room and starboard wing in the living room. The purpose of the assembly was threefold: one to check for fit and connections (both mechanically and electronically), two to gather the necessary hardware and tools to assemble at the field, three to work out the weight and balance in ready to fly configuration.

I first mounted the horizontal stab on the fuselage so the model wouldn't nose over in its cradle when I added both flight batteries to the nose. Both forward swept wings can then be joined together with hefty M8 screws and a common cotter pin. After ensuring the wires to the wings are out of the way, the complete wing assembly can then be dropped into place and fastened to the fuselage by means of four M5 screws and bolts, squeezing two solid spacers taking care of wing forward/aft compression forces between the rear attach points. A problem arose on the starboard side because all my efforts to straighten out all fuselage deformations also had eliminated the minimal gap between the fuselage top and wing bottom. Scraping off 2mm of the fuselage side for about 5cm length would have rendered the original shape, but also meant painting again. I therefore took another approach by cutting those 2 millimeters from the wings bottom along the thick inner rib. Everything then dropped into place including the four 5mm bolts and I first could admire the sleek lines of my new flagship.

Wing job 24

After the wing assembly I made the electrical MPX connections to the fuselage and powered up my Tx and the powerbox. Juggling with the electronic throw limits I was able to eliminate the most serious buzz from all servos. The elevator is so large that if not powered, gravity allows it to drop to full down. Needless to say its servo buzzes at neutral but if I only power up the glider just before takeoff, I count on the airstream to alleviate most of that gravity force when gliding. By intentionally crossing the wires of the pull-pull rudder system I had to inverse the servo movement and also made mechanical adjustments to the wires to have the rudder point straight with zero trim and have the correct wire tension to avoid any flutter. With the wings mounted I also was able to judge the differential aileron deflections and promptly eliminated the roll expo, anticipating a very sloppy response anyway. With all the electronics working correctly, I was able to attach the covering turtle deck, but found out that the nylon M4 screws were too short to grab deep enough in the main wing spar. New nylon screws were cut at correct lengths and the turtle deck now remains firm in place without risk of lifting from its own elasticity or from the airflow between the canopy and turtle deck. To complete the whole assembly I needed a metric hexagonal key, a 7mm spanner, a 1/2inch wrench and a flat screwdriver, all these were given a place in my field box.

My heartbeat went up attempting to figure a way to perform the balance of the model. How much additional nose-weight would be required and how could I get this heavy glider balance on two points that physically were on just planking between the leading edge and wing spar? Have you ever tried bending deep whilst straddling the fuselage and trying to use two fingertips to lift a 15kg+ model from the ground from under the wing? It was painful and I estimated the CG to be aft of the aft limit. I therefore added a loose 240gr ancient weight in the nose and checked again. This time it looked closer but how close? I finally resolved the problem by suspending the model between two bar-chairs, resting on rubber sanding blocks having a curved upper surface. That spread the weight over a sufficiently wide portion of the wing, allowed the model to dip forward or backwards, but because of the constant curve also allowed me to measure at the center point of the sanding blocks how far aft of the leading edge the CG ended up. The 240gr extra weight brought the CG to about 45mm back, meaning it would be close to the 27% MAC, the sweet spot I use for any maiden flight.

Wing job 28

With the model in its flight configuration it also was time to find out how much it finally weighed. I brought the bathroom scale along the model, first weighed myself, then also whilst carrying the model. The difference gave me a final result of 16,2kg. After calculating the wing area to be 180,34dm² and dividing both figures, the wing load ended up as 89,83gr/dm² which is only 10% less than the Blanik but 25% more than my Ka8. Those figures help me predicting the behavior of the glider in thermals and in the traffic pattern. Having flown F104 Starfighter for a living during the 70's I am not afraid of high wing loads, my B25 and Boeing 737 models fly well with 140gr/dm². This Ka2b will have a limited speed range and will not be a floater for landing, but a stable steady glider with more inertia and essential anticipation plus sufficient control deflections to get it off its straight and level “rails”.

With so much difficulty to get the model assembled in the house, I figured that with the removable wings and horizontal tailplane away, I was able to find the CG spot on just the sole fuselage. I made some experiments and found out that without wings but tailplane still assembled, the fuselage CG was exactly over the winch-hook. With both wings and tailplane removed (but batteries installed), it balanced at the point exactly under the bottom of the new canopy bow. With that information I will be able to perform any new weight distribution changes without having to assemble the complete model again (in the house), a very practical solution for later fine-tuning. Following picture illustrates this fuselage balance point.

Wingjob 30

That balance point is also important regarding the design and position of the custom cradle to block the fuselage in my camper during transport. I was astonished to find out how much weight ended up in the nose to get this glider into balance, batteries included it was close to 3kg , no wonder the total weight is a hefty 16kg, a figure beyond my physical carrying capabilities. I will either need a helper if it lands outside the field, or my towrope and Shambeck wing-wheel to move it on the field. Talking to other Schneider model builders I already knew most of his models came out 2kg heavier than advertised on his website. Apparently his data refers to an assembled kit but without electronics nor lead. Most glider pilots I know fly with an aft CG for performance, my choice of flying with a more forward CG for stability therefore results in extra balance weight, hence the 1kg increase over the weight it had when I acquired it. Regarding the modifications and tons of primer/paint to obtain a smooth fuselage I find that an acceptable tradeoff. For such a glider, scale looks is more important than performance.

During the assembled tests I also noted that the handholds at the wing extremities were not sufficiently high to prevent the aft tip to touch the ground at rest. On grass fields that is not a big problem, but sometimes a wing could scrape on a hard surface and that would cause difficult to repair damage to the thin aft tips. These handles had been very well made out of copper tubing I didn't fancy making new ones and just went for the lazy option, I turned them around their former aft screw and made a new hole further back. This deviation from scale was necessary because higher handles would make it more difficult to slide the wings in their protective sleeves. I then used PU glue and balsa strips to help permanently secure the odd-shaped old 240gr scale-weight in the most forward spot available, firmly against a former so it wouldn't come loose during hard landings or forward impacts. The last thing I want is having a 240gr iron weight wandering along the long cockpit in flight, a sure way to dramatically increase your blood pressure. That day I also spent more time on refining the adjustments on my transmitter and setting up the failsafe receiver with spoilers open, hook open and a bit of left rudder. Those failsafe settings will be refined after the proving flights but at least I have an (gu)estimated setting to fly the maiden.

During the last days of the daylight saving time of October I again gave up flying time to undertake the modifications to my camper to safely transport that model to large-glider meets and still have sufficient space to live for a week on location if the glider had to be sheltered inside every evening. The large transverse garage at the back of my camper was barely suitable for the Blanik, but this Ka2b fuselage and wings each have a length of about 2m75, that is wider than the legal limit for trucks and bus so it has to be stowed perpendicular to the driving direction. The mild temperatures (15°C), dry sunny weather with little wind were as ideal for model flying as for trial fittings of the Ka2 to ensure it is safe for long journeys. The illustrated report of the extensive modifications I had to make to safely transport this large model can be seen here

This completes the build log for the Ka2b, the model is ready for the post restoration maiden flight and a following entry about the proving and trimflights will probably only be posted next spring. Please come back then to follow this tread. December 2016 I showed my completed model during a clubevening. It took so much space that later that evening people had to sit around a table under the right wing for their aviation talks. This was the first time I could admire it fully assembled from a distance and had a big smile on my face that evening.

Last edited by BAF23; Apr 09, 2017 at 01:42 PM.
Jan 30, 2017, 09:26 AM
The sky is the limit
BAF23's Avatar
Thread OP

My miniature self in the cockpit

January 2017 I finally got the 3D reproduction of the head-scan that was taken half a year before. The size was correct but although filled entirely with foam bubbles, it weighed a hefty 110gram. The guy had foreseen a copper tube that could be inserted in a slightly larger tube in the head but unfortunately the fixed angle invariably dictated that the head would be looking up if the “spine” was parallel to anything but a pure vertical seatback. Furthermore, the lightweight puppet I purchased in the UK long before, had nice clothes and body proportions. When I decapitated her I found out that inside the light-blue coverall there was only loosely stuffed cotton wool and the lightweight rubber head just had the neck sewed to the lining. The coverall was very nice but completely sewn shut so access to the interior of the body could only be made by cutting the head off, leaving an orifice that could barely accept the short but wider neck of my figure head. Even with the long copper rod of the new head, there was no way it could stay in position once the glider would be flying or even transported by hand. The guy had done its best trying to apply facial and hair color but I was far from satisfied by the first look at it.

After experimenting with the new head on its copper tube in the cockpit and comparing it to an inflight picture of the real glider with somebody in the front seat, I got a better idea of the installation position, angle and elevation of the head versus the already mounted seat. Installing the puppet in the seat and just pushing the copper guide into the body produced an unacceptable result. After extensive brainwashing and measurements I opted to use two foam blocks, one that could be pushed into the body after having removed some of the cotton wool, and one that I could turn into a parachute and simply bolted to the seat. The trick was to insert the copper tube at such angle that the head ended up looking forward at the correct height, but that the tube protruded through the pilot's back so it could be inserted at an angle into the parachute pack. This allowed the heavy head to rest firmly on the seat when installed, the loose light body staying in the seat due to its shape and being attached to the head. After some trial and error installations I got it right, glued a plastic tube in the foam block before wrapping it with the tissue of an old USAF pilot bag so it resembled a backpack parachute.

While doing that I got more and more disappointed by the looks and colors of the delivered head and soon decided to use my available plastic model paint cans to attempt obtaining a more realistic image of me. The base colors being okay, I applied multiple thin layers of different colors over it with streaks of a hard brush to obtain depth and variety instead of a uniform application. To do that right required me to install a mirror on my work table to get my hair color, treats, eyebrows, skin wrickles etc as close as I could to my image. I used all kind of tricks with a knife, toothpick, 2-haired brush, dry dabbing with a towel, unequal application of Matt spray etc to get a result I could live with. I also painted the black rubber hands in flesh colors with red nail-polish but as they had been molded to the shape of thick gloved hands I was limited in what I could attain. After a week I quit but still wasn't satisfied about the shape of my mouth.

I couldn't wait any longer, inserted the head with the foam block in the body, and used toothpick ends to glue the remaining puppet's lining onto the foam block. After raising the collar on the coverall I was able to attach it with a couple of servo screws into the neck of the head. After lowering the collar again I poured PU glue into the space between the coverall and the short neck, and after it dried I was happy to have a relatively solid connection after the head transplant. I had somebody taking pictures of me with my scale replica, both dressed in light blue but with the flash or light angles we continuously obtained different skin and head colors both on me and on the puppet. Take it from me that in daylight it matches better than in these pictures.

I then put the seat back into the glider and found out how easy it was to slide the pilot into position with the copper tube protruding from it's back. I first had a few pictures taken without the canopy installed to show the unobstructed result

I still wasn't satisfied with the mouth and eyes so after the suggestions of a friend I got back to to the paint brushes and knocked the teeth out, removed the red lips, thinned them , raised the eyebrows, thinned the eyelashes and applied some darker skin accents under them for contrast. These modifications seriously enhanced the natural looks of the puppet.

I find it cute that a miniature-me fills the front cockpit of my flagship glider. As for the weight, here are the figures: pilot head as delivered: 110gr, complete pilot body with head: 245gr, complete pilot installed with parachute and seat: 375gr. Mini-Laurence now patiently awaits her maiden flight and sits against her parachute in the cockpit getting acquainted with the layout of the instruments. Because it still could take weeks she didn't strap in yet and keeps her ventilation window partly open.

A ski accident in Italy caused my right collar bone to be split in 4 pieces that had to be surgically realligned by means of a support plate and bolts. It is easy to understand that this seriously would curtail my ability to lift heavy loads for a few months, which in turn delayed the maiden flight of the Ka2b till after my recovery. About 3 weeks after the surgery I took the risk of forcing myself and with the help of my 91 year old dad we managed to perform a precise balance after I borrowed an adequate home-made ballancer from another modeller. I opted for that date around Eastern because it was warm and there was hardly any wind on my terrace.

After complete assembly I was surprised that even with the heavier pilot figure the model balanced at 60mm, seriously close to the aft limit of the recommended 34-68mm CG range and too far aft to my like for a maiden. Another 170 gr in the nose was necessary to bring the CG to the middle mark of 50mm. After disassembly and storage of the model I felt that I had overstressed both my right shoulder and my right knee, but I was glad that despite the pain this last job could be ticked off and I could give the custom balancer back to the rightful owner. The extra weight had to be inserted but not fixed yet because flight testing could reveal more or less weight to be ideal for the flight behavior. Stay tuned for the report of the maiden flight somewhere during late spring 2017
Last edited by BAF23; Apr 09, 2017 at 01:45 PM.
May 19, 2017, 10:48 AM
The sky is the limit
BAF23's Avatar
Thread OP

Maiden and first check flights

Wednesday 10 May 2017 all conditions were met for the maiden: my collarbone (and shoulder) felt about 85% OK, the Kruk towplane with King200 was flight-worthy, temperatures were mild and the wind light and moderate, and my 91 year old father was sufficiently fit to help me transporting the Ka2 parts into and out of the tight confines of my apartment and the camper. The president of the BiGGS international large glider association was present and helped me in the assembly on the field and performed a detailed preflight before the maiden. I liked the way he asked me to lift both wing extremities for checking that the top and bottom airbrakes would still operate under load in flight. This is a test I had never performed but I definitely noted it down for future use. After the satisfactory range check and a rough finger CG test we concluded the model was flight-worthy and talked things over with the waiting tow-pilot before we moved our big birds into takeoff position. My dad seemed more interested in the KRUK warmup adjustments than in my long meticulous but for him uninteresting steps toward the maiden.

Another club member took the opportunity to make a few pictures of me next to my new glider before...

I asked for a low-speed shallow climb to 300meters, which would give me just sufficient time for trimming out the glider, perform a stall, check the effect of spoilers on pitch, and making a few turn reversals to watch how much adverse yaw the ailerons produced before joining downwind for landing. Bert handed over his movie camera to Harry so he could assist me if necessary. Well that was the plan, but the reality was much different. Acceleration was normal and I could keep the wing from dipping during the first meters. The glider lifted clean off the ground without elevator use and settled at a seemingly comfortable height above the towship. Field layout called for a right hand turnout and I noted I had difficulty establishing half his roll angle during that. My ailerons felt sloppy and although able to maintain wings level, during turns my glider invariably started augmenting the bank angle without me doing any inputs, resulting in the glider cutting the corner which is no good if you want the rope to keep tight. Luckily the Kruk is so powerful that after the third turn we were at altitude and I opened the hook.

A quick succession of left aileron trim inputs kept the wings “about” level but that was inconsistent. I immediately noted the total lack of harmony between the flight controls. Turning to the left was very slow even using serious aileron inputs. Turns to the right were a little easier, but pitch was very sensitive and the glider had the tendency to pump. I tried trimming the rudder to the left to stabilize the bank problem, but altogether never really had things under control for a nice steady glide. All that uncoordinated flying seriously imparted the gliding performance and I was descending rapidly without ever being able to fly a proper straight trimmed-out glide stretch. A quick stall check showed a left wing drop but moderate pitch dipping. A check of the spoiler deployment showed me not to be afraid of using them because they produced little pitch changes. In the meantime the light wind had turned around so the towship dropped his cable in the opposite direction and flew downwind to also land that way. I on the other hand was standing on the edge of the runway from which we took off and I definitely didn’t want to land from a maiden on the far other side of the field.

When he overshot his landing and performed a go-around, I saw him turn close-downwind the moment I announced normal downwind, both in opposite directions. I became extremely stressed and was happy Bert waved him off to make room for me to land. Those were not the relaxed conditions you want for a first landing of your flagship. Turns to base, baseleg and final were stable and uneventful but without proper trimming nor proper stall check, I elected to fly the pattern fast with minimal flare. I already extended full spoilers in baseleg (the best stretch to observe pitch and speed changes). Line-up for final in the axis of the longest runway was good but steep (for previously mentioned reasons) and as usual for a maiden I aimed for a touchdown 1/3rd down the runway. I kept the nose down till just before touch, then applied a quick but positive elevator up input to break the rate of descent without realy holding the glider off the ground. This resulted in a normal (but relatively fast) touchdown and a stop in the middle of the runway after a mere 3-1/2 minutes after tow release. Just as a real Ka2b, with a pilot aboard it has much weight on the long and wide forward skid and that is an excellent brake for that heavy airframe. I ran to evacuate the Ka2 from the runway so the Kruk could finally land. My heartbeat and adrenaline flow were beating records and although people congratulated me for this “beautiful maiden”, I experienced different feelings about that flight.

Without touching my transmitter (trim) settings or shutting down power from the glider, I analyzed the position of the flight controls and trims to try to understand why it flies so funny with little or no effect from the aileron trim, neither to the left nor to the right. Rudder now was deflected slightly to the left and about a quarter of left differential aileron travel was showing. Elevator trim had been hardly touched and still showed streamlined. With that in mind I shut off the power from my Tx and the glider, and went into the clubhouse to drink a coffee with my dad and calm down, trying to “play again” the film of that maiden in my head. Unfortunately the cameraman had finger trouble and had no images to help me. Another guy made nice pictures during the maiden takeoff roll and that is what you see here.


Second flight

I realized all happened so quickly that I never had time to stabilize my glider during the entire flight, I was continuously running behind, correcting things without ever taking the time to see the results after half a minute. It was obvious that with so little consistent data but with a proven flyable glider, a second flight was essential to explore further into the possibilities to correct the flaws. Having been able to land it decently the way it was, I decided to take it in the air again without changing the controls nor trim positions. I asked the tow pilot to tow me faster, hoping this would help the roll rate and bring me higher above his possible slipstream interference. This time Bert operated the camera because he knew I could take care of myself after that eventful maiden. Another member lifted my wing for the takeoff (the grass was rather long that day), and off we went into the blue cloudless sky.

I had briefed that I would limit myself to almost flat turns, counting on the Kruk to pull me around the turns so I wouldn’t have to battle the ineffective ailerons. As from the first turn after takeoff I noted my ailerons still weren’t effective and to prevent the glider from banking by itself, I had to apply almost full left aileron to keep wings level behind the towship, not a very comfortable way of flying. Upon his rollout I was too late neutralizing my controls and started sideways excursions which I never got stabilized before he initiated the left turn. The Ka2 promptly developed a serious bank into the turn and I had to be quick to level the wings and keep the rope tight by moving outside again. By the time he rolled out I seriously got behind (the situation, never behind the towship) and the slow aileron response caused PIO’s (Pilot induced Oscillations) which became divergent to the point that I called that I was going to release. At that very moment, the 27kg break-cord in the rope at the towship side snapped, the cable falling into the adjacent field. All this is clearly visible in next video recording of that day.

Kruk en Ka2 Henis 10 mei 2017 Maiden flights Ka2b (6 min 49 sec)

As soon as I was clear, the glider flew much better and I got a grip on the ailerons as long as I flew calmly. When I finally had those trimmed out Bert suggested me to trim the elevator more down to eliminate the pumping. That worked very well and I started picking up weak but narrow thermals in a sky you wouldn’t expect that. That way I was able to fairly well maintain altitude whilst performing turn reversals with full aileron deflections, first without rudder, then with half and full rudder. The results were not encouraging. More thermals and after climbing 100 meters I had sufficient altitude to perform a CG dive test under fully trimmed condition (that was with nearly full left aileron trim and a quarter left rudder). I expected a very slow pullout and a left bank. To my surprise, with growing speed it never recovered from the 30° down pitch, and the higher the speed, the more right bank it developed, figure that ! I recovered from the 45° of bank and traded speed for altitude again. Next I came around for a proper stall check, again without touching the trims. With that right aileron a bit down and the left more up, I was ready for a right wing drop. Surprise again, it was the left wing that dropped sharply and caused the nose to lower at an alarming angle. Relax and think whilst working the thermals again, but I could not figure out where those very contradicting behaviors in roll originated from. That neutral longitudinal behavior would be easy to correct with yet more weight in the nose, but in the meantime it prevented the execution of spin testing.

I slowly got a grip on these very strange control deflections and stayed airborne well in the less than optimal conditions. When I entered downwind I again decided that as long as I didn’t understand its speed behavior, I would keep the speed up till close to the ground for a late flare. The sensitivity of the pitch made it difficult to maintain a steady glide angle, but this time I aimed for a shorter landing spot behind the threshold and again all worked well. After the landing I noted my timer indicated 25 minutes after release, quite a result for a new glider that cannot be trimmed to satisfaction. Parts of the flight and the landing are on the video, but I must have been too high for him to catch my CG dive test and the stalls, during which he was next to me and observed the strange behavior for himself with his eyes instead of through his camera. As at that stage flight safety was not a problem, I just noted the new elevator neutral position and dragged the Ka2 along the field for another confirmation flight.

Third flight

Because of the slight down elevator trim, I applied a bit of up elevator to get it off the ground and above the towship’s propwash. Once established in the climb I released the back-pressure and at speed she stabilized at a good height and angle. With what I had learned the hard way, I anticipated the movements during the turns and, with less stick deflections on the ailerons, was able to follow rather well in the turns. At the prebriefed release altitude I called and opened the hook upon which she immediately settled in a steady glide. I didn’t bother further exploring the flight envelope until the pitch and roll problems had been resolved, and just enjoyed the elegance of my Rhönschwalbe as I flew the minimal thermals at the end of that afternoon.

I felt like the queen of the sky but unfortunately I got dragged away from my sweet lonely position, because the crosswind forced the towship to use the diagonal axis of the runway and I stood in his way. That disturbance partly took my attention away but I still managed to remain airborne for 13 minutes after release. I adapted my traffic pattern for a landing across the field which incidentally allowed photographer Arsene to make some great realistic shots of this classy model.

As it was my last flight of the day and I would have had to drag it all over the field for dis-assembly, I deliberately made a left curved final so she would end up closer to where I wanted. The attitude and background were just superb in following shot

I continued my curved approach even including the flare and touchdown, but contrary to previous pictures you now can see it’s a model and not a real glider landing, cutting the weeds might eliminate that problem..

This landing again proved the docile behavior of this impressive 16,5kg model (if flown calmly and properly). By that time many club members were on the field and really admired my latest creation. After a quick cool celebration beer I started to disassemble that mighty machine because it takes time and patience to get it all in the correct sequence and angles into the camper. We then had another drink whilst filling in and signing the legal paperwork to send to the Belgian aviation authorities in order to get the authorization to fly models between 12 and 25kg (advanced model glider flying license mandatory). After arrival home, dad and I moved everything in my apartment so I could perform essential corrections after the maiden.

home adjustments

A couple of days later I started by the wings. By disconnecting the aileron links I was able to see the physical limits of the aileron deflections. Up didn’t go any further than what I had, but down I had a bit more room and after totally eliminating the electronically programmed aileron differential and adjusting the rod lengths and servo arm holes, I obtained almost symmetrical up and down aileron deflections which I hoped would boost the roll rates to acceptable levels. Of course the added induced and shape drag will cause some adverse yaw which I countered by dialing 10% rudder mix with the ailerons. I figure this will simply counter the drag and thus will not cause additional yaw movements during the critical tows. If causing trouble, I can always program the rudder coupling to only start after opening the hook. As the model had been test-flown without aileron expo, there was nothing more I could do. After seeing the nose rapidly pitching down when using full rudder, I reduced the rudder expo from 75 to 50% and adjusted its neutral position to slightly left of center.

All this was done in the apartment with the wings electrically connected to the fuselage but physically just parallel to the fuselage. I thus was unable to measure the angle of attack along various portions of the wings to check for wing warp, I’ll have to borrow an incidence meter and that will have to be done later when the model is again fully assembled in open space. Looking carefully at some of the pictures could point to a hint of washout on the starboard wing and none on the port, that could very well explain the illogical roll behavior between the dive test, the the stall and the aileron trim setting. Another suggestion of Bert was to weigh the finished wings individually and also measure each wing’s centre of gravity. That was an eye-opener, the port wing weighed 2943gr and balanced 92cm from the root, the starboard wing 3042gr and 90 cm from the root. I never expected that much difference and that certainly could explain the need for left aileron during the flights. Those distances correspond to halfway the lower spoilers so in order to better balance the wings I need to add about 100gr of lead somewhere in and around the root side of the port spoiler box. That is a lot of lead for there ! but might eliminate the aileron trim deflection and better stabilize the wings under tow. After flying the modified aileron setup I’ll probably try first with 50gr and observe if it makes a difference or not, before attempting to insert the full 100 gram lead into that port wing.

To get rid of the neutral longitudinal stability requires the CG to be moved further forward. There are two ways of doing that, lightening the tail or adding even more weight in the nose. As already mentioned much earlier in this build log, I never was happy with the very heavy weight of the horizontal stab system and already managed to shave some weight off, but not enough. As my intention is to produce an all new lighter horizontal stab next winter, the additional nose weight I install in the meantime has to be easily removable later. Space for “easy lead” is becoming sparse in the nose but I managed to insert and tighten a custom produced 100gr lead shape on the side and top of the heavier ones. That might not be much for such a heavy model but could well shift the stability from neutral to slightly positive, and that will probably eliminate the pumping in pitch and produce a moderate self leveling moment in CG dive tests.

As during the 3 test flights and extensive control throws and ground adjustments I had depleted my two 2S2800 LiFe batteries to 65%, Bert suggested that instead of carrying useless lead, I might as well exchange it for added battery capacity. I think this might work but I am limited by the strange shape of the suspended battery pockets and am afraid these could separate from underneath the top of the nose in case of a hard landing. I’ll definitely will also investigate that possibility after I find an equilibrium in that fuselage assembly. In the meantime I didn’t adjust the elevator neutral point, figuring the added nose weight will eliminate the need for down-trim. I also count on the better stability to reduce the sensitivity of the elevator, so I kept the 25% expo.

That was as far as I went at that stage. If modifying too much at once, you’ll never find out what exactly did the trick. I prefer advancing in small steps and observe the effectiveness of the corrections before undertaking further experiments. It will take me longer to get it completely right (for my desires) and in the meantime I’ll restrict my flights to calm weather and uncrowded fields, but so be it. Stay tuned for further updates after the summer.
Last edited by BAF23; Jun 28, 2017 at 04:12 PM.
May 22, 2017, 09:35 AM
The sky is the limit
BAF23's Avatar
Thread OP

Fouth flight

Two weekends after the maiden our club held its annual fly-in (all R/C categories) and for the Sunday show I assembled the Ka2b in between my B25 en Boeing737. Bert brought its incidence meter and together with civil engineer Dirk they leveled the Ka2b and started measuring the incidence angles along various points of both wings. The verdict was: from port tip to starboard tip: 0,5°/0°/fuselage/-2°/-2°, tailplanes both -2°. The good news was that the wings were not warped, the bad news was the 2° incidence difference between left and right wing. That confirmed my analysis of the flight behavior after observing the CG dive test and the stall. We talked it over with other experienced modelers on the field and came to the conclusion that the easiest way to correct that would be to reduce the port wings incidence by raising the aft attachment point on the fuselage by about 6mm, and filling-up the turtle deck a bit on that side. That attachment point consisted of 2 metal brackets bolted to fuselage frames, so relatively easy to remove and replace by two slightly longer ones.

That 2 degree incidence difference was enormous and I now faced the dilemma of flying the model during the show with the changes I had made after the maidens to find out if that helped, but also knowing for sure the basic aerodynamics were all wrong. The flightline was filled by Belgium’s finest R/C pilots and models, a relatively large crowd including my daughter and grandchilren plus various friend, with me participating the first time at such an event. Would it be wise to fly the Ka2b with those known shortcomings and risk holding the flight schedule up, plus my reputation? I hesitated for a couple of hours but whilst checking my slot for the B25 I saw that half an hour before a double tow slot had been booked by pilots I knew. I approached them and asked if they had a problem if between their tow and landing of both large gliders (one had a 9m span), the Wilga could tow me up for a short flight? Danny was a super towpilot and although he participated in the incidence discussion, he agreed to tow me if I wouldn’t hesitate a second to release if the oscillations started again. I thus setup the model with the flight controls at the new neutral trim settings and towed it to takeoff position after the others got airborne.

After they had released their gliders they had planned a short demo of the towships before landing, but Bruno landed his Zlin 252 immediately and looked at me. I was surprised, questioned Bert about this unexpected development with a tow-pilot that was not aware of my glider’s shortcoming, but then decided to go ahead and hook up because the show had to go on. I then rushed to the pilot box where I told Bruno I might disconnect early because of previous lateral stability problems. He opened the throttle and up we went. Neither the hundreds of eyes looking nor the the rapid change of plans gave me any stress and after a momentary wing dip early in the takeoff roll, it lifted smoothly and settled nicely in the climb. My guesswork on the adapted control setup seemed correct because I stayed the entire tow like on rails behind a smooth flying Zlin 252L, even during the turns.

After the release I made minimal trim adjustments and had a perfect glider which had no problems picking up the moderate thermals. Being high and dry with others still airborne I became bold and told Bert I would do a CG dive test. He was reluctant because of the other 2 gliders in the vicinity but a little later, using my peripheral vision I found a free area and pushed the nose down. Upon stabilization I neutralized the controls and was very happy to note that it now slowly raised its nose by itself, the mere 100 gr extra in the nose did the trick. Even better was the fact that it didn’t bank to the right anymore, proving the new equilibrium between more left rudder and less left aileron worked better. Bert suggested I also tried turn reversals and that was an excellent idea. The elimination of the aileron differential had drastically improved roll response, but the rudder mix of 10% was insufficient to compensate for the increased adverse yaw. In the meantime Danny and the 9m glider were already on the ground and I was eyeballing the remaining glider to get in sequence to land about the time his glider would be off the runway. I flew a textbook square pattern with a very steady final, followed by a full spoiler landing and the magnificent Rhönschwalbe stopping abeam the speaker’s position.

When the spectators applauded I made a little reverence but my inner smile was as wide as the runway, and my heartbeat was as slow as after a siesta. Bert later commented he watches my stress by listening to my breathing, and contrary to the stressful training flights I performed with other models the previous day, this time it was fully normal. We later talked the flight over with others that watched it carefully after knowing the incidence problem, and their idea was to continue flying it like that for the season and only make the port incidence modification during the winter period, I tend to agree and see what happens during the following couple of development flights.

Back home I noted my transmitter still showed about a fifth of left aileron and rudder trim. I increased the rudder mix to 15% and made a backup copy of all the settings. That demo flight (and detailed ground analysis by pros) seriously boosted my confidence in that model. Why the builder had overlooked the incidence during the otherwise very detailed build is a mystery, but it probably explains why he quickly sold the model, and the second owner also putting it up for sale shortly after modifying it with an electric motor. It gave me the opportunity to purchase a relatively cheap Schneider model which was not too difficult to turn into a beauty that flies well, with even better flight characteristics after the incidence modification. A serious structural problem on the tail of my Blanik the previous day revealed a previous repair by the original owner. Fiberglass cloth had been applied from the inside in the aft fuselage, but the surface probably had not been sufficiently cleaned and the original crack has surfaced again. Both shortcomings in less than 24hr made me seriously think if it still was wise to buy secondhand instead of building from scratch in the future.

Until I get better footage of the fly-in I only can give the URL of a teaser in which my Ka2b is clearly seen in good company.

Quick Reply
Thread Tools