Thread Tools
This thread is privately moderated by BAF23, who may elect to delete unwanted replies.
Sep 23, 2018, 08:03 AM
The sky is the limit
BAF23's Avatar
Thread OP
Build Log

Tipsy Nipper T66 Mk3 scale 1:3 Part1: fuselage


Chapter 1: Introduction to the Tipsy Nipper

Pic at cox

How can you squeeze a person of normal height and shoulders into that tiny cockpit?

Aircraft history

In 1952 designer Ernest Oscar Tips started designing his diminutive T66 in the Fairey aircraft factory in Gosselies Belgium. As the Hawker Hunter license production ran down he started building prototype OO-NIP that was testflown with open cockpit on 2 december 1957 by Belgium’s famous test pilot Bernard Neefs. In 1959 production started on the closed canopy Mk2 that were powered by a 45hp Volkswagen, Stamo or Hepu engines with exposed cylinder heads, 64 were built till 1962 (+78 kits delivered). Because Fairey started producing the F104 Starfighter, the Nipper production was then taken over by Cobelavia that produced 18 Rollason cowled-engine Mk3 Nippers in a hangar in Kortessem (just a few km from my town and model airfield) till 1966. From then on production of complete aircraft (and kits) was transferred to the U.K. Slingsby built them till in 1971 the Donington based firm went bankrupt (after producing 32 complete aircraft). Later kit produced variants even boosted a 85hp Jabiru engine. Nowadays many still are airworthy in the UK, Belgium and the Netherlands.

The design philosophy had been to produce a lightweight single seat aircraft that was cheap to produce and operate, and easy to fly. Although not the smallest nor prettiest, it was the smallest practical aircraft and even had aerobatic capabilities. Making it small kept the weight down (300kg empty) and some ingenious solutions were incorporated to get there. All had a tubular frame fuselage and rudder, with wooden wings and horizontal tailplane. In 1994 I had the opportunity to fly a couple of times in the 1961 Cobelavia produced airframe nr73, a Rollason Ardem powered Tipsy Nipper Mk3 registered OO-MLD (still flying around nowadays but in a Dutch color scheme and with tiptanks). It was natural to choose that aircraft in period colors as the subject for my 1:3 scale model. If you want to know more about flying the full-size aircraft, read the next paragraph, if you are only interested in the model, skip the following paragraph.

Flying the manned Tipsy Nipper:

The Tipsy Nipper had been designed in 1952 by an engineer working at the Belgian Fairey aircraft factory at Gosselies. It was first test flown and developed by the famous test pilot Bernard Neefs (of later F104 fame)who performed many spirited demonstrations with the Nipper, even at the Paris airshow of Le Bourget, prompting the sales. The production version only tipped 300kg empty on the scales, and was fully aerobatic with the mere 45hp Stamo boxer engine (modified Volkswagen). It performed a looping in only 200ft diameter, and its 6m wingspan with large ailerons gave it a good roll rate. Directional control was phenomenal because the whole vertical tail moved in one piece, making crazy flying, cross controlled passes, and flick rolls its best assets. It also helped control benign wingdrops during stalls and provided instant spin recovery. With so much area in front of the hinge, rudder application was light and precise.

The right-side-hinged total visibility canopy allowed you to prop the engine into life (no starter or electrics) standing between the wing and prop. That way your prevented the aircraft from moving forward (no parking brake) and used your left hand to prop, the right one on the throttle or the choke. After it came alive and ran smooth, you reduced the throttle to idle, walked around the left wing, lowered a narrow portion of the wing incorporating a step, and got in. Getting seated was no sinecure because the wing spar was a massive carry-through beam positioned over your thighs and knees. Once you managed to pass your lower members under it, you sat fairly comfortably in the 45cm narrow fuselage. Once seated, you moved your left arm back and down, and pulled up the wing-step that engaged in a sort of snap-lock. Because of the limited elbow room, straps had to be donned before closing the canopy. The restricted cockpit width had been augmented by leaving out the first wing rib, allowing you to stretch your elbows in the wing roots. The lower part of that extra room was made of clear Perspex, providing some vision below, through the very deep wing obstructing most of your normal down-vision.

Flight instruments consisted of an airspeed indicator, an altimeter and a compass. Engine instruments were an RPM indicator and oil pressure meter. Fuel quantity was a cork actuated needle protruding from the filler cap in front of the canopy, the tank was only 25 liters but gave it a range of around 300km at a 140km/h cruise speed. As you rolled forward, the Sandow-sprung main gear immediately widened, and the aircraft continued rather tail low, but visibility over the nose remained good because you sat high in the cockpit (neither seat nor pedals were adjustable). The small disk-brakes were very effective and were operated by a single bicycle type handle on the stick. Nose wheel steering was mechanically connected to the rudder pedals and resulted in a very short turn radius on the ground.

Pic Tipsy OO-MLD


The only check before takeoff was a test of the magnetos. Acceleration was swift, and it got airborne in about 300 meters, but climb was slow at about 500 ft/min (much depended upon the weight of the pilot, together with the fuel making up to 1/3rd of the takeoff weight)

With the engine just in front of your knees, it was noisy, but relatively smooth at high RPM without excessive vibrations. In the air it just felt like a toy, very subject to turbulence, but responding to the slightest control deflections. Controls were very harmonic and light, and it was a delight to perform aerobatics on a dime. One of my first flights was a long ferry to Coxyde airshow, but flying along the highway it was easy to become depressed when you saw than even trucks were passing you in a moderate headwind. Crossing Belgium took time in a Nipper at cruise speed.

Having no radio, I had made arrangements with the ATC people to show up from an unused approach direction at 500ft, at a precise time. Without crossing any runway axis, I then would descend to tower altitude within the airfield boundary, and circle it to get their attention and await for a green light for landing on the runway indicated on the T of the signal square (also in front of the tower). Everything went as planned, but I was surprised having to release the back pressure during the steep turn around the tower, I could easily have hit it because that Nipper steep-turned much tighter than expected (in about a standard runway width !)

After shutdown, a standard military AVGAS fuel truck came and the operator had to be very careful slowly transferring the 18 liters to fill the tank through the small hole (just the hose of the overhead bowser system contained already more). I then tightened the seat belt around the stick (no control locks) and lifted the tail up high to allow the now free main gear to spring back together, alleviating the stretch on the Sandows during our stay. Parking had to be done on a soft surface so you could drive tourniquets in the ground to tie the wings down, otherwise the slightest wind could cause serious damage. The picture on top of this page shows the pilot/owner Roger helping me secure the airplane that rests in its typical ground stance next to the farm in the middle of Coxyde military airfield. You can also see the sunlight shine through the Perspex under the wing elbow cavity at the wing root.

Choosing the model and powertrain

A model of that not so elegant but charismatic aircraft was on my wish list since a long time, but except for an RBC kit of less than a meter span, there was nothing but plans on the market. Finally during the 2016 glider meet at Bastogne I saw a Belgian kit producer (limited scale in his cellar) who intended to commercialize a 2m span traditional wooden kit. I saw him again at the first 2018 swap meet at Wavre where I could see one of his completed kits and the content of the box he had for sale. Unfortunately his molded nose at that time was for early Nippers with exposed cylinders. I talked him into producing a Mk3 nose but that would take weeks before it was ready. In the meantime he prepared a box that I picked up at his home so I could start the build and pick up the nose later on. None of the kits produced at that time had been assembled with an operating step or an opening canopy but as I wanted to make mine more scale such modifications were options I would dig into.

I still had an Axi4130-20 309kv in my stock and that seemed suitable for a 6kg electric model. With a box full of servo’s and 6S plus 8S ESCs, I only needed a prop to complete that model. I bought that specific aircraft to have an easy to transport, assemble and fly machine that I could operate from either tarmac or from prepared grass surfaces. The idea was to use it as a currency trainer capable of multiple not so perfect landings in a row, but also capable of performing Nipper type basic aerobatics. I thus aimed for about 200W/kg with a scale diameter prop that wouldn’t be screaming but propelled the model at a sedate scale speed around the pattern. It took a fiend and me some time to come up with a solution. I had 6 4S4000 25c batteries and 4 6S4400 45c batteries in stock. The choices were to either run the model on one 6S battery for some quick aerobatics, or two 6S batteries in parallel for long cruising or pattern work, a third option was to put two 4S batteries in series for all roles. After bench testing the various combinations with different props and measuring the outputs, we settled for the 8S setup (availability in my stock and dimensions in model) swinging an 17x8 prop. With an efficient XOAR prop this means 7500 rpm, 46Amp and 1338W during 223sec. A Biela scale-look prop turned at 6000 RPM and at 52 Amps produced 1513Watts for 197 seconds (with 30” reserve). At 70% power this would equate to 51km/h for about 5min flying time, sounds great for scale type performance on this model, even if that prop is a bit heavier and more noisy than the Xoar.

Chapter 2: Unpacking the kit and inventory

Everything was in a single large carton box (except the Mk3 cowling which at that time still had to be designed and produced upon my special request). You could easily put 2 kits or more in that single box and I do not know how it is stuffed for transport if it has to be posted (I picked mine up at the producers’ home in Belgium). Opening the box reveals more parts than I had ever seen in a kit, about everything is included like hinges, wheels, balsa sheets, dowels, steering arms, nuts, T-screws, strong bolts, you name it... , but because electric or petrol engines are so different, nothing is included for their mounting ahead of the firewall. The huge canopy was delivered with ample protective plastic both inside and outside of it.

All the wood was of excellent quality and the many CNC pre-cut large panels (most of plywood) had very clear cuts and parts were easy to remove. None of these were marked but every sheet item was clearly labeled on furnished paperwork. A checked parts list was also joined but the two reduced A3 sized plan sheets with very superficial assembly sequence depicted on diminutive insets are all you get to make an airplane out of all those bits and pieces. I must admit that if desired, full size plans could be provided but are not absolutely necessary for modelers with previous conventional wooden kit experience.

All but one weld on the optional scale sprung main gear were of good quality and the nose gear also looked good. The part with the bad weld was exchanged for free upon notification. The three 100mm tires are realistic and feel rather soft. The kit producer added me to the list of purchasers and through personalized e-mail you get access to previous builder’s experiences and pictures during their assembly, he is also quick in responding to any questions you have about the assembly. All of this practically takes care of the lack of plans and assembly instructions. To help other builders cope with the bare minimum of kit assembly instructions, I decided to make this build log even more detailed than usual and insert relevant pictures to illustrate the written text. After each paragraph title you’ll see that I mention day numbers, those are full working days without counting any (intermediate) drying time and are only a rough indication of the time that I needed to complete the model. Others could be faster or slower, depending on their level of desired perfection and details, and the glue they use.

Chapter 3: Preparation works (day1)

There are probably hundreds of ways to assemble a traditional wooden model, whatever is described next was my way and sequence of tackling the job, but might be completely different from the sequence other builders are choosing. Regular readers of my model blog know that I always start by making an inventory. My kit is MSN012, the twelfth produced and already incorporating some changes that are illustrated or/and described in the paperwork. All the wood sheets were present and I made a picture of everything before touching it further.

Pic 4542 what’s in the box?


I then took the paper sheets with the “numerotation” of the various parts and used a soft pencil to mark all parts with their specific letters and numbers when still in their carriers. The 5mm aviation multiplex plate didn’t come as depicted on the listing but in separate parts, but that was easy to see. The 3mm balsa sheets were also different: the one with the two WFTE’s came as depicted but the second and third sheets had been combined into a single long sheet. That wouldn’t have been a problem if they hadn’t mirrored/reversed the ER ribs versus the SR ribs. Be attentive and compare the orientation of the parts on paper with what you see on the sheets before you mark all the ribs. Nothing magic, but not straightforward either. A few extra parts for the canopy frame and stronger elevator sides are described on an additional paper and already incorporated on the sheets.

I then spent hours carefully cutting all the CNC machined individual parts out of the carrier sheets. I used a cutter to separate every part, then scraped away the remainder of the attach lips before lightly sanding all the sides to remove minor machinery blemishes. Larger parts had weight reducing panels that also had to be removed. Some people might just tear the parts from their carriers and assemble without sanding all around, and that could be done because the machining was well done. For me, the added work of treating every piece to perfection is just a joy and ensures me a perfect fit and finish later on. Assembling a wooden model in an apartment pushes me into condensing the “dust time” to a minimum, hence a full day for meticulous preparation of every removed part with vacuum cleaning after every sheet had been treated.

As parts of various assemblies can be found on various sheets, I looked at the plans and made separate piles of the groups of wood that were required for each sub-assembly. Each pile was then ordered in number sequence and held together by rubbers. All the wood was of prime quality and very lightweight, I just found minor blemishes on 3 multiplex parts (that were not load bearing critical). By the end of the day I had a huge amount of scrap wood in all thickness at the end of the room, and a neat number of sequential piles in front of that. I spread everything out for clarity, added the planking, stringers and metal hardware and spread the A3 sized plans in between. Following picture depicts that moment which meant the end of the preparation phase. Take care during assembly because most parts can easily be physically mounted upside down or forward to back, discovering that after gluing might completely ruin a sub-assembly, think twice before gluing any part.

Pic 4544 after preparation work


Important note: During the following chapters you will see and read my initial assembly method. During the flight test phase I discovered that some things better be altered earlier in the construction and those notes were at a later stage added in italic so you can avoid the problems I faced during the early flights. Be certain to build light aft of the center of gravity because the model needs serious additional weight at the front to get in balance The illustrations were sometimes produced later to show things better and do not always reflect the state of the model at that stage of the build log description.

Chapter 4:Rudder sub-assembly (RR) days 2 and 3

The reason I started with that was that having no full size plans, I needed the exact rudder dimensions to order the specific artwork from Calliegraphics. Without plan in real size and not all the rudder parts being pre-cut, I lifted my eyebrows till I looked at the inventory sheet under “empenages” and read that the missing parts RLE, RTE and RS were part of the bundle of stringers and had to be cut and shaped by hand. Luckily the plan depicts the exact dimensions of those parts, but to be sure about size and angles, I took aft fuselage plywood part F6 as a guide before cutting and gluing anything. That in turn required me to cut the aft fuselage unmarked small block of Samba to the correct angle. That block is the one in which the hinges have to come for the complete flying rudder to be attached to. This is a rather large surface and in case the model overturns during landing or is bumped during transport, large forces will be exercised on those hinges. Although the supplied hinges are of the nylon type with metal pin, any play in the hinges would cause excessive lateral movement on top of the rudder so I preferred to use even stronger larger (Protoplast) hinges. The downfall was that even larger hinge slots had to be made into both Samba hardwood beams, plus they were heavier and offset.

As it was easier to install the hinges before those beams were glued in place, those were the first actions that I performed for the assembly of the model. The hinges had to be installed perfectly straight in both wooden structures because the slightest deviation would be very visible on the high unsupported rudder surface. Having taken away so much material in the middle of the fuselage Samba beam led me to not only PU glue the hinges into the slit, but also to use (almost) sunken screws and bolts to keep the wood halves tight around the plastic. The main rudder beam was wider so its strength was better, the plastic hinges on this were just kept in place by PU-wood glue and small metal pins through the wood and existing holes in the hinges. Having never worked with Samba wood before, what stuck me most was how much dust it produced when being machined. That being done, I was glad I didn’t install the hinges after those beams had been glued in their sub-assemblies.

It was already very late in the afternoon when I finally was able to tackle what I wanted to start with: assembling the rudder itself. Using a square marked cutting mat on a very flat table allowed me to position the main vertical beam and glue the top ribs RR1a and RR1b perfectly square. Because all ribs are tapering towards their extremities it is important to support them wit scrap 2mm wood near their ends during the gluing. When that was dry I made the aft frame with the RTE of 298mm pushing onto RR1b and RR4, the latter being glued at the same time against the main beam. Because all those parts fit at various angles, some material had to be sanded to obtain the maximum contact area where glued together. Forward lower rib RR3a was then glued to the beam, using the loose positioned F6 fuselage dorsal rib as a guide for the correct distance and angle. Next the corner filler parts RR9 and RR7 were glued in place, again resting on some 2mm scrap wood to keep them centered. That sub-assembly was left to completely dry before the next actions were undertaken. Here you see a picture of the balsa tail puzzle parts before being further carved (for correct contact angles) and glued to the samba main beam.

Pic 4557 tail puzzle


When all angled parts were solidly in place, it was possible to adjust cross members RR3b and RR2b for correct length and fit, their angles only defined by the corner guides RR8 and RR6. RLE was cut as a 216mm balsa beam and glued into place, after which crossmember RR2a was made to fit in between according to the RR5 corner block angle and opposite to RR2b. After all this had dried I started to rough-sand the the side excesses of RTE and RLE down in the prolongation of all the RR ribs. On the real aircraft the leading and trailing edges of the rudder are small diameter round tubing but solidity of the wooden assembly dictated wider ribs on the model. RLE and RTE just cannot be sanded as thin as desired and I used common sense in rounding off the leading edge and flattening the trailing edge into a much thinner end. Upper and lower ribs were seriously rounded off, including the one over the fuselage that can cause the elevators to jam against RR3a with full up elevator and full rudder deflection (as found out during final testing before covering)after flight test substantially full up elevator to 30mm there was no jamming possible anymore. Outer corners of the rudder remain square as per original. Lightweight filler was then used on all the depressions, mainly the ones caused by the separation of the parts from their carrier sheets. That was the moment I could take the measurements for Callie to reproduce the markings of my chosen decoration. At a later stage an extra samba reinforcement was added as a support for the control horn but that could only be made after fuselage and tailplane were assembled and position of the horn had been selected.

Pic 4564 tail ready to cover


Chapter 5: The fuselage frame days 4,5 and 6

To assemble the fuselage you need a perfectly flat non adhering surface. The designer found a clever way to assemble about 95% the fuselage on this flat surface, the remainder are just small frame parts shaping the belly of the fuselage at a later stage. The main portion is mounted on strong plywood fake belly “chassis”. As it runs the full length of the fuselage it had been engineered in the two parts F7a and F7b, joined by a precisely cut V shape, and capped by the gear support plate F15. If these 3 parts are glued together and pressed at the same time, you will start with a strong perfectly straight base on which to assemble the rest of the fuselage. I used Titebond original wood glue for all the perfectly fitting wooden parts, and Bison PU wood glue for joints that could use filling from this expanding glue. Another major sub-assembly was produced at the same time, the two large side frames F0 were glued around horizontal frame F8 and vertical former F1, perfectly at straight angles if done at the same time. Both sub-assemblies were left to dry overnight to be assured of good strength before being mated together. If you already know which rudder and elevator servos you will use, it is easier to drill their mounting holes in plate F13 before gluing that in the fuselage.

Pic 4558 2 fus subass


After mating them I immediately continued with the aft formers F2, F3, F4, and F5, all capped by the long F6 spine. All of this can best be glued together to the basic “chassis” in one go in order to have all the correct angles. Thanks to the very precise factory CNC cuts everything sits nicely tight and amazingly no adjustment cutting is required before gluing that complex assembly. One note is that on your flat work station you have to lay the chassis in such way that the tail bumper part (of F6) overhangs and can dangle freely in the air. Just a few pincers are sufficient to hold F5 tightly against F6, and both F0 against F2 during the drying. During that time I used the parts list to select the foreseen 13 fuselage stringers and prepare them for application. At that stage the fuselage looked like this.

Pic 4559 fuselage with stringers next to it


If till that time the woodworm genes in you felt bored, they for sure will be triggered when cutting and applying some of the stringers. I suggest you start with the 10x5 samba stringers, of which you only have two and if you mess up one of the long FS ones you are in for some ingenious angled wood gluing to connect shorter pieces into a long one (guess how I figured that out). At the back of the long samba stringer you have to taper the top at an angle of about 10° to allow the shorter piece to actually sit on it. Watch the total height at that point because a balsa stringer still has to fit on top of both the samba pieces. That was the easy part, a lot more time was spent cutting, filing and sanding the back of the short samba pieces so they fit tightly and flat against F6, then widen to taper to nothing at the end of the yet to be installed samba rudder hinge end plate. Without any full size drawings this becomes a matter of trial and error. When you finally have the correct shapes, make an exact duplicate of the long and short FS stringers, and glue all 4 pieces in place. Be especially careful that those short stringers lay perfectly equal to the bottom of the horizontal stab cutout, any error in this might cause a deviation of the stab straight-angle versus the fuselage. When dry, you can slip the rudder hinge plate in between them en glue it solidly with pressure against F6 and on chassis plate. Congratulations, you have regained your woodworm qualification.

After that, cutting and sanding the six FSD crossmembers felt like child-play but remember that their role is to counter torsional and other forces so a tight fit into each corner is paramount. Next job was the cutting and gluing of the many FSR balsa strings forming the fuselage back. One length was used for the upper and lower stringer, another length for the middle one and the 3 partial stringers. Duplicate that system on the other side and you end up with a full 1 meter spare stringer. The fuselage was then rotated on its side and supported under the tail in order to glue the 1meter long balsa outside FSR stringer along the bottom fuselage, but don’t forget to first taper the aft end so its hugs the rudder hinge block. The balsa stringer is sufficiently soft (even dry) that it can make the bends in two directions between the aft fuselage and the cockpit sides if forced and kept tight against all contact surfaces during the drying process. A picture at this stage tells more than a thousand words.

Pic 4560 fus on side for outside stringer


After that had dried I turned the fuselage upside down on a wine carton and leveled it with scrap wood so the chassis was horizontal and the bottom formers could be glued in place. The lip of F11 had to be slightly sanded because excess glue from initial assembly of the chassis stood in the way, that’s how close kit-cutting tolerances are. F12b was glued flat just aft of the next former and F10b with 55mm spacing aft of F10.

Pic 4563 fus inverted with outside stringer and lower frames


This was also the moment to mark the drill holes for the scale main landing gear. No instructions are given for that but it is obvious they have to be centred on the still visible F15 reinforcement plate.
Work very precisely because later when you have to mount/dismantle gearlegs, the tolerance between pivot pins/screws and lower fuselage planking/covering is extremely small
. As there is no assembly leaflet it lets you figure out when you will mount the gear versus the two bottom stringers and how you will take care of the ample gear movements versus the covering of the bottom fuselage. If you want to keep your model light, you have to get some form of stringers in that area to later cover it with tissue. As I preferred to work more scale and increase solidity for manhandling the finished fuselage, at a later stage I used hard materials in that area and didn’t need any lower stringers forward of F11. At that stage I just glued the balsa stringers from the tail till F11 but later found out that manhandling the fuselage during assembly or transport easily caused those flimsy stringers to crack, so I replaced them by stringers of a stronger wood type. Not yet installing stringers forward of F11 at that stage makes it much more easy to block the model on trestles for further work on the fuselage. More about that gear area later in this build-log.

I then turned the fuselage on its other side to glue the last external stringer on the lower side. I then glued the wing attach box members in between the F0 cockpit sides, starting with FBW1, the one with the cut-out. Be sure to glue both with the wing attach holes at their lowest position. Failure to do so could ruin your day (and fuselage) at a later stage. Then insert the plain FBW2 part and glue into place. After excess glue has been wiped off on the inside, dry insert both WB dihedral main wing spar parts into the created orifice, this will push the FWB plates against the walls and ensure sufficient space to freely insert the wings on the field. Pinch the top of the FO sides tight against the FWB parts and wiggle the main spars free from time to time during the drying process. Some might wonder why I still hadn’t glued the FF Firewall at this stage, but uncertain desired nose-wheel height (affecting fuselage stance angle due to deep bending main-gear) and type of motor mount, led me to postpone that move to a later stage. If you think reading this fuselage assembly text took time, note that between the mating of both basic fuselage sub-assemblies and this stage, I needed a complete 17 hour workday to get the basic fuselage assembly to that stage !

Chapter 6: horizontal tailplane (SR), elevators (ER) and ailerons (WLEA) day 7,8

To assemble the horizontal tailplane you start by laying out its aft part SFTE on a totally flat surface, then push all SR ribs standing vertically in their slots, measure and cut the required lengths of leading edge SLE out of 10x10 balsa (with a little excess to cater for the sanding at an angle at their junction, and after final assembly at the tips. Check this dry assembly for alignment and keep SFTE firmly flat by use of a weight. Remove the leading edges and the ribs, apply glue to every aft rib part and push into place. After half of the tail done, drop the leading into place without glue and verify alignment and twist by looking straight down on it. Use a square ruler to position every rib perfectly perpendicular to SFTE. Repeat as necessary till every rib is perfectly seated, then gently push down on the leading edge to bond the ribs at the bottom and re-verify alignments. Perform the same ritual on the other half, ensuring both loose leading edges are perfectly aligned with each-other. Let dry for a while.

Pic 4566 ele vert


A piece of toothpick was then inserted and glued with PU between both leading edges to reinforce the 7° angle, then the complete leading edge was dropped-in and glued on previous vertical assembly. Minor alignment issues could still be straightened by using a book or so to push things straight at that late stage. Lightweight filler was then applied to fill any depressions in the trailing edge. Capping the aft and center section of the horizontal stab was straight forward but I recommend you make the center capping at least 1 centimeter wider on each side, top and bottom. That will allow you to cover the stabilizer (with fabric) after it had been glued to the fuselage instead of before. When everything had dried, sanding the leading edge into shape and removing the excess material from the straight extremities completed that symmetrical sub-assembly.

The elevators required more labor. After the ELE part is set horizontally, Ribs ER1 to ER8 can be positioned vertically for trial. Balsa ribs R1 and R8 can go to the scrap box because the same are also provided in stronger plywood. All the angled ribs are ill fitting on the ELE base-plate and to each-other so serious filing had to take place in order to drastically increase glue surfaces (5 filings per angled rib), also towards the future trailing edge. First glue outer rib ER8 at a square angle to ELE. The 10x2 samba trailing edge can also be prefabricated at that stage, but is then only used to check the alignments and rib lengths. With trial and error you’ll finally get everything to fit correctly. Then apply liberal glue to ER7 down to ER1 (also where they join each-other on an angled flat surface) and insert into the adapted ELE cutouts. Before the glue settles, eyeball along the future trailing edge that all rib ends are perfectly lined up. Use a little weight (a few pencils or so)in the open V’s to keep the ribs pushing against each-other. Let dry and repeat the work on the second elevator.

Pic4567 progress on ele


After thoroughly drying, I used the Dremel grinder to cut 2mm slits in the back of each angled ribs. This was a very secure job because there is hardly any material to work on and cutting too deep wasn’t an option either because there wouldn’t be any balsa left where the ribs touch each-other for strength. A mere 2mm is the max depth I dared to take out, but that still was a lot more support for the samba trailing edge to grab on.

I didn’t want to stay idle during all those drying processes and decided to alternate working on the tail with building the ailerons. Because the square ribs W11, 13, 15, 17 and 18 already have slits, I decided a different approach. The tight fit of those ribs allow them to be dry positioned on the WLEA leading edge base so the WTE trailing edge samba can be cut to size and trial fitted. Beware that unlike the symmetrical tailplane, the wings and ailerons have a flat intrados and a curved extrados. Care has to be taken to glue these ribs all the same way on the base so that the trailing edge will be in a straight line. With the glue still tacky on also the trailing edge, lay the ailerons with the intrados flat on the cutting mat and align WLEA and each rib along the square lines to obtain fitting ailerons. When dry, the task of fitting W12, 14 and 16 has to be undertaken. Again watch orientation and put the square angle on the intrados. Sand that short side into a corner so it sits nicely against the rib/LE corner. Measure where the pointy end joins the rib at the trailing edge and cut at an angle. Good luck taking off 2mm on the bottom so the top rests on the trailing edge and the back against it. This is plywood and it easily cracks where you don’t want it during the carving or cutting, and it is too strong to sand where it becomes very thin to mate the extrados end of the other ribs. Take plenty of time and cut very small bits off at a time. Start-off with W14, the shortest one to get a feel and if things go wrong, you can cut another W14 from the scrap wood. These cross-members can then be glued in place with the aileron flat on the mat.

Back to the horizontal stab: elevator leading edges are then sanded in a round shape whilst the aileron leading edges were sanded in the V-shape as suggested by the shape of the square rib fronts. As the trailing edge of the fixed stab SFTE doesn’t offer much grab area for the later to be installed elevator hinges, I used the 10mm scrap kit balsa to fabricate 4cm long blocks that were pushed and glued between the planking where the 3 hinges would come each side. The positions of the hinges has to be determined by placing stab and elevator against each-other, you’ll see that because the ribs do not align, there are only limited possibilities. The elevator leading edges provide sufficient grab and I decided to already glue the hinges on that side. The hinges are not the best quality because the metal pin often got bent during fabrication when cut off, so you better use pliers to bend back the distorted ones before gluing in place, failure to do so will cause additional friction during movement. Do not glue the elevators to the stab yet or it won’t slide through the aft fuselage anymore. The kit provides nylon control horns but no instructions nor parts to mount them on. I again used scrap kit ply to make triangular plates that I glued between the lower inner elevator corners so they were flush with the underside. When dry, holes were drilled for the horn screws and the backplate already glued to it, the horn itself will only be glued after the elevators have been covered. Beware that the furnished screws in the kit are very long for their job and fit really tight into the nylon backplates. It is advisable before any gluing to turn those screws fully in and out of their backplates to form a tread so the backplates won’t be torn from the wood after the control surfaces are covered and the inside becomes inaccessible.
Last edited by BAF23; Sep 23, 2018 at 08:16 AM.
Sign up now
to remove ads between posts
Sep 23, 2018, 09:18 AM
The sky is the limit
BAF23's Avatar
Thread OP

Tipsy Nipper T66 Mk3 scale 1:3 Part 2: Wing assembly plus scale details


Chapter 7:Wing assembly day9,10 and 11,12,13, 14

After a thorough plans and parts study plus some dry assembly trials I decided to deviate from the suggested sequence on the plan inserts. Some parts are labeled Left and Right so it is important making a correct pile of parts for each wing before starting. I noted that although the exterior cutouts for the 10x5 samba stringers were relatively accurate, the ones inside ribs W4 to W8 had to be seriously filed to let the stringers pass through. Making the fit too tight will create problems later on when you have to slide the WS stringers all the way through the wing ribs. In fact every indentation for the WS stringers better be individually checked for correct width and depth before starting the assembly or you’ll regret it at a later stage. Be careful when filing because ply separation lurks behind every corner. Do not remove the small trapezium protrusions at the bottom of the ribs at that time, they ensure proper wing warp during the assembly and are useful till after the balsa planking. On the plan it seems the wing warp is opposite to what it should be, this is only an optical illusion caused by the overall wing tapering. The model flies vice-less without wing-drop tendencies when built on the provided construction feet. Technically spoken, all WS are spars but for ease of identification I use the term wing-spar for the main box and the term stringers for the auxiliary spars.

The main wing-spar is built up from many parts but its main body is made of two plywood parts glued nose to nose, probably because of limited wood-length availability, which is neither good for strength nor for ease of assembly. Personally I would have engineered it as the fuselage chassis, with a V-cut to augment the contact length, but decided to go ahead with the kit as provided. none of the modellers having performed aerobatics with their Nippers have encountered structural failures so the strength of the spar and the wing field-assembly system is at least adequate
I first glued both W21 parts together on the side WS stringer, using many clamps to get everything perfectly straight. W21 being equal for left and right wing, it is up to you to decide which you use for which wing, but then look at the root where wing dihedral makes obvious what top and bottom are, and mark the front and back. First the stringers WS should be be glued at the upper front of W21 and left to dry on a perfectly level surface with a straight ruler on the side for check. When dry, cut the excess WS clear and sand the glue joint perfectly flat to facilitate later balsa cap bonding. For reasons that will later become obvious, I elected to first start the starboard wing assembly, even if the plan only shows the slightly different port wing. So far for an evening study and preparation work (not reflected in the day count).

Prepare all ribs and depending on the glue working time, either glue the ribs upside down on the main spar, then quickly turn everything around so all can dry flat on a level surface, or if using relatively rapidly grabbing glue, apply the glue to W4 to W9 and slide these ribs each in turn into position from under the right-side-up main spar. W10 can then be slid sideways on the assembly, and W2 and W3 glued from behind, allowing all to grab while level on their trapezium feet. You might use a weight on top of them to push all those feet firmly on the surface, but if using something long, remember that W2 and W3 are completely capped from front to back, so be sure to fill that distance and keep looking if all the trailing edges are well aligned. If not done on a cutting mat wit squares, be sure to use a square to align each rib to the main spar. Here is how the right (starboard) wing looked drying at that stage, with the port W21a and b plus WS being prepared alongside.

Pic 4573 wing pre-assy


After a couple of hours and still drying on the mat, you can glue aft wing-wide WS in position after sliding it in from outer to inner, leaving the inner part overhanging because you have no idea where to cut it yet. Be sure to push it in completely forward at the aileron positions because WFTE will have to slide all the way through every aft rib protrusion and lay flat at the back. Do not install WFTE at that stage but use it dry to ensure proper alignment and spacing of W6 to W10 during the dying process. At this stage you can also prefabricate the front W19 main wingspar by gluing W1a, W2 end W3 nose-caps perpendicular to it, noting that W1a is tilted because of the dihedral,do not mate this sub-assembly with the wing yet. Inner crossmember W20 can be dropped in its place and the protruding part can then be used as an anchor point for an elastic pulling the trailing edge WTE against the aft of W2 and W3 during the drying process. W1b is tilted for the dihedral, and angled because the fuselage tapers back at that point. Use left wing’s step plate WKSH as a template to help you glue this aft rib in place. Don’t search for a middle A1 rib, in the real airplane the inner rib is absent to make room for the pilot’s elbows in the inner wing. All the previously glued parts should be left to thoroughly dry before moving the wing assembly any further. At that stage the assembly should look something to this.

Pic4577 wing assy


Next step is to turn the wing upside-down and file away all excess glue or unnecessary protrusions (leave the little feet), both to facilitate the installation of the two lower WS stringers and the adhesion of them to the already produced materials. Extra glue can be allowed to flow along the bottom of the upper stringers. For the next trick, apply glue to all contact surfaces of WB (wing joining and dihedral plate), push it all the way through to W4 and press solidly against W21 for the initial contact time (right wing only). Wipe off any excess glue that might prevent the bottom stringer (now on top) to sink into place. Glue the latter into place and clamp solidly against WB and the main spar. Slide lower aft stringer through all slits till 5mm from the end, apply glue to future contact surfaces with all the ribs and slide all the way to the end. Immediate apply glue to all surfaces that WFTE trailing edge will come in contact with and push into the obvious slots. Again use clamps (hopefully you have sufficient) and let all your work dry overnight. After one day of work your right wing could look like this:

Pic 4578 wing invert


The port wing is slightly different in that it is equipped with an extending aft wing root that in the real aircraft can be manually lowered to be used as a step to board the aircraft. As this is such a unique and prominent feature, I wanted to make this functional on my model and this build log thus deviates from the kit as delivered.

Pic mld step down


The dowel that keeps the back of the wing in place on the kit could not be used anymore because it is positioned aft of W20, the pivot point for the ingenious step system. Because the trailing edge of that wing is interrupted, structural integrity is different and as seen on following picture of the full-size aircraft-kit contents, an additional angled crossmember was used to redistribute the trailing edge loads to W20 on the port wing. Also note the small vertical cross-plates between aft stringers and W6,7,8 and 9

Pic real kit exposed

The factory produced second production aircraft along the kit version contents

If you do not choose to incorporate a working step, build the port wing primarily as you did the right wing except that the all important WB carry through spar now has to be glued to W19 in order to slide next to the other WB in the fuselage box, and the fake step wing part is built as per plan and you can skip the following paragraphs of this build log. If on the other hand you decide to incorporate a functional step, bear with me and for the following modifications and assembly instructions.

Start as you did on the other wing by gluing spars W21a, W21b and the top stringer together. Then glue ribs W4 to W10 along that partial box. Take rib W3 and duplicate the aft part (behind W2) but extending into a sharp point about halfway the trailing edge. Label that new part S3. Take rib W2 and cut the complete part aft if W20 away. Label that part S2. Glue W3, W2 and W20 in their normal position. Cut and shape trailing edge WTE to fit between the aft of W4 and W7. Fabricate corner plate to grab around the trailing edge and W3, then glue all of that together. Use scrap plywood to fabricate an additional cross-member between the aft tip of W4 and where W20 meets W3, glue solidly into place (I used PU glue for that and the new corner at W3). Allow to dry completely.

Take both WKSH step plates and remove 5mm of their total width at their exterior side (they don’t lean on W3 anymore). As they were initially cut with the wood grain 90° to each-other, I elected to use the one with the grain in length as the lower plate because that is the one that has to cope with lengthwise over-pressure during maneuvering. Keep the dihedral in mind, making the lower plate 2mm wider at the thickest part of the step. Trial fit into position and eventually sand the angled aft part for correct alignment with the wing trailing edge. Produce a leading edge for the step, I used scrap 8mm balsa from the kit because it provides a good base for the hinges. Use W20 as a template for cutting the shape (angled at the inside for the dihedral and on top for the taper) but subtract the width of the S1 and S3 ribs. File away slits in the bottom to accommodate the hinges. Take kit provided W1b rib, cut forward lip away and glue that to the recess in W20. Sand top and bottom of W1b so it rests correctly (according to the dihedral angle) between both WKSH plates. Glue the 3 ribs and new leading edge together on the bottom of WKSH plate. Use scrap balsa to make walls around the actual footstep-opening. Use scrap foam to fill the inside to create some stiffness. When all is dry, sand upper surface flat and before gluing upper WKSH plate to it your assembly should look something like this. When using foam in the step you might consider using all balsa parts instead of the kit-plywood to save weight because the aerodynamic forces are not that great on the step

Pic 4587 the step


Remember you prefabricated the right-wing W19 nose assembly? According to the plan you should just glue it to the narrow end of both stringers of the main spar. That seemed insufficient to me and because the WB from the other wing only penetrates the first 5cm in the wing spar box, I decided to fill the void between WB and the future W19 by a perfectly fitting 5mm balsa piece. That was first glued and pressed over its total surface to W21. When dry, the balsa was sanded so nothing protruded from between both WS stringers, that completely flat surface was then totally glued to the W19/W1a/W2(a)/W3(a) sub-assembly using many clamps to solidly join everything against the WS stringers. To ensure the other WB could be inserted snugly in the remaining void, use scrap 10x5 samba bits to keep the spacing correct during drying.

Pic 4579 wing box filler and parts


On the left wing, WB junction piece is first glued to W19, perfectly centered to leave the desired space for the WS stringers to flank it, and just deep enough to touch W4 during final assembly. Again, use many clamps in the process. While drying, glue another custom cut 5mm balsa plate to fill the void between the future WB and W21, this time glue it completely against W21. Clamp it securely because there is no way to sand it in its recess. Leave to dry while gluing both bottom stringers in place. Before curing, remove the clamps and glue the complete W19/W1a/W2(a)/W3(a) assembly against the previously installed plates and stringers. Use as many clamps as you can to compress everything vertically and laterally, and leave to dry thoroughly. Before hitting the sack, I glued both leading edges WLE to the wings and was happy I did two basic wing structures in just two (long) days.

The reason I worked hard and fast up to that stage of completion was because so far most had to be built on a long flat surface to get everything straight. The hobby room in my apartment has no long perfectly flat table, so I was forced to glue a lot on a cutting mat on the table of the living room. Being able to restore the latter to its former glory and use after less than two weeks of abuse and sawdust was a blessing. Next day I trial assembled the model skeleton on the terrace and thoroughly cleaned the house. When looking at the picture of what I achieved in 11 days of work I remembered that in traditional woodwork modelling, 90% of the work gets finished in 10% of the total time, the remaining 10% need 90% of the total building time.

Pic 4583 skeleton on terrace


Step completion: On the wings, a furnished dowel is used to align the back of the wing to the fuselage, but I found one hole in the inner rib and one on fuselage plate F14 insufficient, even with an M8 screw and bolt. I therefore elected to make an additional hole in the second rib to spread the loads further into the starboard wing. The movable step on the port side makes blocking the wing in F14 impossible but during the first dry assembly I noted that W20 touched the fuselage just aft of the cockpit and at an angle. I figured that by gluing a slightly longer W20 forward of the existing one, it could penetrate the cockpit area where a few custom cut scrap plywood parts could keep the port wing in place, even under G-loads. So that takes care about the up and down lock, but in case of a crash, both wings want to continue forward due to inertia, and only the main wing spar remains to prevent that, forces it was not designed to cope for. I thus installed a system that kept the dowel from moving outboard on the right wing by use of a simple vertical pin and a washer, and on the left wing a horizontal pin that prevents the new W20 from sliding outward. No tools are needed on the field for that. To help spread the loads of the port wing and the step, I then glued scrap ply triangular reinforcements between the longer W20 and W2 and W3, on the underside so the hinges of the step could also grab on them along their total length (after I previously filed the slits for the step hinges).

Pic 4755 servos


Here the build-log continues for all. Glue the outer WSS servo frames to the wing, taking care to leave sufficient grab for screws to be turned into the WS and aft Samba stringers. The only purpose of the outside frames is to anchor the wing fabric around the removable WS servo panels. For practical purposes I deviated from the plan and choose to install the plates with the slits to the outside and at their most forward location, furthest away but aligned with the chosen control horn position. Take your aileron servos and fasten them to the inside of WS, I used hardwood blocks for that but there is very little excess room for them between the surrounding frame. Make the slits for the aileron hinges, unlike on other models where you can create some mechanical differential through hinge placement, you are here limited to the exact middle position due to the faceted leading edge of the ailerons. Use a bit of scrap balsa to augment the hinge grab on the wing trailing edge (just as you did on the fixed stabilizer). To facilitate wing covering, glue the hinges only to the ailerons at that stage, not yet in the wings. Horns were furnished without support plates, these again were fabricated using scrap kit ply and glued to the underside of the ailerons in the prolongation of the servo slits. When dry, holes for the screws were drilled and the horn backplates glued in place, the horns themselves were only installed after the ailerons were covered.

Chapter 8: wing planking day 15, 16

Some essential sanding is mandatory to flatten or deepening the surfaces before wing planking can be initiated. I first applied a full 10cm width planking top and bottom of both wings. The minimal intrados curve on the wings allowed the 2mm balsa planking to be applied dry, but in 3 times. I first glued the nose of the plain plank against the leading edge and the first centimeter of the ribs (using pins). When that glue was sufficiently solid I applied glue to the forward half of the ribs and pressed the planking into it with weights. After another drying time I applied glue to the remaining contact area between that plank and the ribs. This method was applied to both wings.

When all was dry I removed the pins and turned the wings right-side up. Because the extrados has a more pronounced curve and the bottom planking blocks the access to apply more glue to the ribs, I opted to apply pre-curved balsa in one go for the top. Selecting the softest balsa planks from the heap, I put them in the bath to let them soak for a couple of minutes in water. The dry balsa quickly became softer and using my thumbs to create a curved surface, more pronounced at the leading edge than further aft, I was able to gradually (take your time and don’t force anything) bend the balsa into shape. Once into shape I allowed each plank to dry for a couple of hours, keeping it straight laterally and curved longitudinally by pushing it against the floor/wall line, using weights to keep everything fixed while drying-out. After trial fitting, I applied glue to the leading edge and all rib contact lengths, pushed the pre-shaped plank in place and used a combination of rubbers, pins, finger pressure and weights to hold everything as should be during a thorough dry. Here you can see the works at the various stages I just described.

Pic kitchen floor 4588


Next I took care of the inboard planking, this is straightforward and can be done one full composite length at a time, using pins so the balsa capping is glued against each-other whilst following the extrados curve. I aligned the outside with rib W3 and made the inside overhang slightly longer because the final cut of that can only be done when the wings are slid into the fuselage (if you want a tight fit). If you look carefully at the plan you will see how the bottom planking hugs the straight fuselage line whilst the upper planking finishes in a bow to slightly augment the cockpit width. That is why I started with both extrados planking so that also could settle with the wings still on their little feet. Do not assume all panels are the same width, with the aft of the wings tapering closer along the fuselage to the end, you’ll need more planking there. To facilitate the transition on the trailing edge, sand the bottom aft of the contact area in a thin wedge before applying liberal glue and pins to fix the panels. In the same go you can cut the last D-box planking to fill the remaining gap over the main spar. Because of the wing decreasing thickness, that last long panel will slightly taper towards the end, be sure to dry assemble everything before gluing.

Pic4591 canopy plank cutout


Turning the wing around to do the bottom planking you either fill everything up to the fuselage line, or if you are more into scale you omit the wood between the fuselage and rib R2 from the main spar till the aft stringers, leaving just a few millimeters for the Perspex/acetate look-through-the-wing inserts to to be glued to. When gluing the aft inboard balsa, take care not to put any glue where the hinges of the step have to penetrate (if constructed). At this point the little feet that served to maintain wing warp to design specs can be eliminated. By this time about all your 2mm planking and most of your wood glue should be incorporated in the model, keep whatever is left for planking around the main gear.

Pic 4589 inner planking with step etc


Sand the wing as necessary to obtain a smooth transition between all planking and shape the leading edge. Use the hand palm and eyeball to ensure the entire intrados and extrados planking form just one continuous curved surface with no depressions nor steps. If you decide on a fixed canopy, next step is not as critical, but I wanted my canopy to be hinged on the starboard side as per real aircraft, creating a few additional problems.

As mentioned before, the inner planking on the top of the wing makes a bow to augment cockpit width, problem is that the canopy needs a narrow strip of material to rest on. The shape of the bow in the wing thus needs to be identical to the bulbous curve of the canopy at that point, no better way to find out but to use the canopy itself. First you’ll need to cut it for appropriate dimensions. If you look closely at the canopy in the light you’ll see a very faint outline of where the canopy meets the wing extrados. Highlight that with a non permanent marker and first make a rough cut outside of that line (I used the Dremel grinder) but keep the front vertical end intact for shape rigidity. Next you mount the wings in the fuselage and drop the canopy on it. Use scissors and sandpaper to trim the canopy bottom till it hugs the wing planking whilst its front end follows the bow of F1 frame, and the aft barely overlaps F3 (to ensure the shape). With the bottom of the shaped canopy allowed to drop freely over the wing extrados, use a soft pencil to mark the shape of the canopy bow on the planking. Then remove the canopy and draw a new bow a few millimeters inboard of the first line. Use a cardboard template or so to make sure the bows on both wings are identical (the canopy is flexible and will be made rigid to follow that bow by using a brass frame at a later stage). Once marked, cut and sand the extrados planking along that second line. Because things got complicated at that stage for the canopy fit and operation, I went to a nearby airfield where a similar real Nipper underwent maintenance. Although it was a Mk2 and only 2 serial numbers younger than the one I was making, it had substantial differences, much of them being aftermarket modifications by the various owners. I still took a zillion detail pictures for further finishing my basic model.

Pic 4635 in hangar with EFA


The bulbous shaped canopy is hinged in a particular way to allow the much curved shape to rest on the starboard wing when open. This was done in real by using offset hinge points some distance in the wing. On a model with removable wings this caused another problem. Transporting a single wing with a canopy dangling along it was no option. I thus decided to first having a copper canopy frame made by a friend, with attachments soldered to allow loose hinge pins to slide into, making the complete canopy assembly easily removable for transport. It rests unattached on the hinges and fuselage when open, but the same hinges and a locking pin on the port side (as on the real one) keeps everything tight when closed. It took hours to get it right because there is not a single piece of tube that is not curved in at least one plane. Contrary to the real Nipper where the canopy is not removable, we added a straight transverse tube at the back for added rigidity mainly during transport. After everything was much adjusted by trial and error, the brass frame was primed and painted, the paint then removed at the places it touched the clear canopy, and canopy glue was used to join everything in a much more rigid way, hugging the fuselage and wing contour lines. A contour line was then painted all around the finished canopy. An outside handle was then added . This was done by puncturing the port tube with a drill, but weakening it too much at the same points made me decide to omit the inside handle.


Pic 4756 canopy hinges

When dry fitting the ailerons I noted that no provisions had been made for the thickness of Oratex and paint so the aileron would slightly stick outside of the wing after final assembly. Looking at the outside of W10 rib I noted that it would be very difficult to make it completely smooth for painting where the stringers, leasing and trailing edge pass through it. I resolved both problems by gluing a piece of 0,6mm ply on its outside and then sanding it smooth with the rest of the wingtip. Ailerons were then completely aligned with the rest of the wingtip and the tip will be smooth to apply my model’s eye-catching type-name artwork onto it after painting.

Chapter 9: hardware installation day 17, 18

Before the model can be covered it is advisable to already at least dry-mount and test-fit all servos, produce the surface actuators and install the wiring where necessary. Regarding the ailerons we already prefabricated the servo holders during previous stages so the only job remaining is to fabricate the linkage (for which I used a 2mm rod with an adjustable kwik-link on the horn side and a custom made z-bend on the servo arm side. The electric wire on the servo was substantially shortened so there wouldn’t be extra wiring floating around in the wing. I then cut a length of wire, mounted a female JR plug on it and attached it somewhere near the opening of the servo plate. I then ran it along the aft wing stringers and attached it with a knot on each stringer crossmember, again so it couldn’t be visible or slam against the covering material during aerobatics. On the root side I made it extend 10cm so it could be connected to fuselage hidden JR plugs, but I used a male on one wing and a female on the other so there would be no chances to connect the ailerons in reverse.

If you built the version with functional step you’ll have to decide how it is kept in flight position. The real Nipper uses an antique ball with spring lock (like sometimes used in old furniture) in the wing rib. As this would be difficult to duplicate in scale I decided to use a canopy lock pin mechanism from inside the fuselage and just drilled a hole in the flap/step at the correct place where it locks it firmly into place for flight.

Assembling the wings to the fuselage for that, I found out that the furnished M6 hardware could not be used as such because the T-nuts are too long even if you could get their 7mm hubs through the 6mm holes of wing key support WFB1. The easiest solution was to use scrap balsa and after drilling out 7mm holes, glue that in front of WFB1, and have the T-nuts grab in the softer balsa instead of trying to get it in the hard plywood. There are no forces on those nylon screws so there is no structural problem doing it that way. Your wing should now be correct in shape and can be further finished (pore filler, primer or so) and fine sanded to prepare it for covering.

Same story for the fuselage, drill the holes for the servos and dry install all tail feathers (with their temporary mounted control horns in place). You can now use any straight bit to mark where holes have to be drilled in frames F4 and F5 for the Bowden rods to run as straight as possible from the 3 servos to their respective control surfaces. A pull-pull system for the rudder (as on the real Nipper) would also be possible but even more complicated because the cables run through under the fuselage floor and exit just ahead of the F5 frame, thus requiring another servo installation spot. With standard size servos, all 3 Bowden outer sleeves will rest on top of the F3 crossmember but for stiffness I added 3 ply extensions to guide the sleeves at their forward extremity. At the back I fabricated solid balsa plates to guide the sleeves out of the fuselage between the stringers. I then glued all (blue) sleeves to whatever they pass through. In retrospect I might have considered running both elevators (joined solidly together) from a single servo, that would be easy to do and save weight aft of the CG

Whilst that was drying I prefabricated the windows for under the wing roots. I used 1,5mm clear acetate that was cut to fit flush in between the balsa planking around it. I then used 0,5mm white polystyrene that I cut in narrow strips and glued on the inside of the balsa planking, with a few millimeters into the opening so the window could later be glued onto it. As the Nipper I am portraying had the throttle mounted outside the fuselage, I looked for a similar plastic bubble to cover it on the outside and cut that shape out of the window. The idea is that the bubble with throttle handle and pushrod to the engine will be permanently glued at a later stage to the fuselage whilst the windows remain permanently bonded to the wings, with a slight gap towards the fuselage and resting on the throttle fairing. This is the most practical way to keep the wings removable from the fuselage. Because the underside windows and large canopy will give a good view of the elbow room in both wing roots, I glued a 1mm balsa false rib against W2 left and right to close that space completely as per real Nipper. All corners were then finished with lightweight filler. Next photograph shows the result of all those scale modifications on the port side

Pic 4757 throttle box


It was easy to find a place for the receiver (not under the seat because the seat rests flat on the fuselage baseplate) and connect everything to make certain the throws could be achieved and all control surfaces moved freely and correctly. All prefabricated parts were then removed to facilitate further work. At this stage I also prefabricated the struts between fuselage and horizontal stab. These are not mentioned in the kit but as the stab is fixed on the model, I found it desirable to add them, as much for solidity as for scale appearance, but it adds undesirable weight to the tail. Spare balsa blocks between the stab planking and Samba blocks above the main floor at the attach points ensure a better grip for these struts. Samba bits were also shaped as to form a hugging support for the forward end of the stab to rest on because I didn’t trust all of the forces could be coped for by the thin planking glued solely to the F6 Spine. The upper aft part of it was so thin that it cracked trying to push the stab through it. I glued solid balsa rests left and right of that to shape the aft fuselage behind F5 and the rudder, this added extra rigidity and allowed me to further file away wood to allow a covered and painted stab to slide through correctly after the fuselage got covered.

I then spent the next 3 weeks applying cover and finish to the flying surfaces. I first applied liquid Oracover iron-on adhesive with a brush to all surfaces except the fuselage. Modern builders will probably do a fast cover (in a single day) of all those surfaces with White Oracover and then apply a few strips of colored Oracover where needed, but I personally don’t like this uniform immaculate too shiny finish so typical of large ARF models you mostly see at the fields nowadays. I much prefer to painstakingly cover my models with unpainted strong textured Oratex. I then apply two or three of coats of primer which I mostly sand away in between, then use a roller applicator to apply two coats of water based acryl paint. This method is much more time consuming and a bit of weight (although most evaporates during the drying) but produces a very strong surface that is still much flexible between the ribs and although the texture is mostly hidden, the surface is not mirror like and resembles much more how a recently painted fabric covered real aircraft looks like. To reproduce the colored streaks on the wings and stab, I was glad to have some pictures I took of the upper surfaces of the Nipper when the owner flew formation on me flying his Stampe to a gathering of old aircraft during the nineties.

Pic form 00004


It took me a complete day of trial and error to apply the masking tape in such way as to have all angles and proportions of the streaks resembling the ones of the air-to-air pictures. The translucency of the red necessitated 3 coats to obtain a uniform finish, for the blue a single coat was sufficient. After the painting I applied the remainder of the Callie artwork and glued the ailerons, movable step and perspex windows on the wings, finalizing their assembly.
Sep 23, 2018, 10:52 AM
The sky is the limit
BAF23's Avatar
Thread OP

Tipsy Nipper T66 Mk3 scale 1:3 Part 3: Gear, nose


Chapter 10: Motor and landing gear mount 5 days

Still awaiting the Mk3 nosecone after 3 months meant I also didn’t get to exchange the badly soldered main gear strut, but in the meantime Scale Dreams sent me the accurate plans of the soon to be completed cowling including the scale position of the thrust-line (without side- nor down-thrust) and distance between firewall and front of the cowling. That allowed me to fabricate the engine mount (crossing my fingers the delivered cowling would be conform to the plans) and complete the front of the fuselage. I studied how the nosewheel was supposed to be attached to the firewall but due to its pronounced forward rake and so far completed nose section, couldn’t figure out how to best install a steering mechanism and servo. I even considered mounting the nosewheel in front of the firewall instead of behind it, reason being that with a scale-size pilot with legs and feet where they should be, the steering servo would be somewhere at the feet and virtually inaccessible to service or replace after the firewall was glued to the fuselage, but that also meant I would have to cut out a good portion of the firewall to allow the steering arm to move back and forth. I also had no idea of the stance of the aircraft, the constructor warned me the rubbers of the main gear being too strong, the aircraft taxied too much tail high and had problems rotating for takeoff, he recommended lengthening the nose strut but I wasn’t fond of that. I still have the impression it sits higher than the real Nipper

The engine mount of the AXI 4130-20 that I had came with some spacers and 6mm holes to affix it. The spacers were useless because I needed an extra 7cm between the engine mount and the firewall. The displaced engine mounts I had in stock were either too short, not sufficiently rigid to cope for a 1500 Watt motor, or couldn’t accept the 6mm thread. I therefore resorted to using 6mm threaded wire cut to length, affixed to T-nuts in a reinforcement plate glued at the back of the firewall, and kept rigid by custom made plywood panels in-between them to form a strong extended motor-mount on which other things could be attached. The (still loose) firewall with all those things attached to was a blessing to work on it free from the model (especially for drilling).

Flight testing revealed that engine torque side-effects could easily be absorbed by some mix between throttle and rudder inputs, but that it is imperative to mount about 2° of physical down-thrust on the motor mount. Furthermore when balancing the model I found out it needed much weight in the front and at this stage you could already envisage mounting batteries etc forward of the firewall and integrated with the engine mount extension as I did on a later stage. As my modification only works on a wider Makk3 cowling, I won’t go deeper into this at this stage because I assume most other builders will choose for the narrow Mark1 cowling with exposed cylinders. Mark 3 builders better look forward to chapter11 to see how I at a later stage resolved this critical problem on my model

To better visualize how everything would end up, I decided to first mount the main gear to the fuselage but before that used a file to round off all the sharp corners of the aluminum gear beams sticking out of the fuselage. Later flight testing revealed the main gear suspension to be totally inadequate for ground ops from either hard or soft surfaces. I recommend at this stage drill four additional holes in the front beam as illustrated in following picture, more about a working gear setup later in the build log. Ensure that the front main gear beam will be easily removable at all times during the life of the model.


Pic4745 gear beam mod

Once mounted on the fuselage I noted that the legs didn’t move freely. It took me some time to discover that the pivot points were not hollow tubes but inside-threaded tubes, and that it was easy for the bolts to tighten themselves and jam the system. I removed the legs and filed some material away front and back of the pivot point, and upon re-installation left some play between the bolt and nut and the aluminum profile. Once installed for good (after test-flying), I used Locktite and new locking bolts to ensure proper free movement of the legs (will help them spreading under the airplane weight as per real Nipper). The gear hardware did not live up to expectations, the relatively scale looking tires turning anything but round and the few provided collars couldn’t be slid around the axles without first being enlarged inside. I added wheel collars on the axles inside of the wheels to reduce friction, and cut the excess axle length outside of the outside collar.

The nose strut in the kit consists of a strong 6mm wire with at the bottom a spring load absorbing compression strut that ends into a fork around the wheel. So far so good, but on the other end it gets its forward angle by passing through a nylon part that has to be bolted to the firewall by means of T-nuts and hex bolts. Those T-nuts are longer than the thickness of the firewall and for spreading the loads in case of the nose wheel encountering an obstacle or rough terrain, I added a piece of 5mm plywood in which the pins of the T-nuts could grab, and the larger diameter center section could pass. Installing this the way it was supposed to (aft of the firewall) meant the T-nuts and extra plate end-up in front of the firewall, the bolts against the nylon part aft of it. In case you ever have to remove the gear, a long 3mm hex screwdriver can be inserted from the cockpit through the pilot’s feet compartment. Don’t assume the nylon block with the hole for the strut to be symmetrical, mine wasn’t and had to be bolted slightly offset at an angle for the nose strut to come out vertical. It invariably comes out against the already much too small forward lip of F7a in which a hole had to be drilled. Because the initial stance of model looked acceptable with the nylon block all the way at the bottom, I decided to file away nylon at an angle so it could hug and be glued to the F7a plane, every bit of solidity helps in that weak narrow area (designed that way similar to the real aircraft, but that was a metal frame and not plywood).

Attaching the prefabricated firewall assembly to the model by simply gluing it (with PU-woodglue ) to the fuselage was only step one, polyester matting was then used between the firewall and the side fuselage panels to turn the mating into a sufficiently strong structure to withstand the combined forces from the powerful engine and the nosewheel impacts and distribute them to the rest of the central fuselage portion. I therefore also reinforced the lower assembly at its sides by cutting scrap plywood into triangles that were glued at an appropriate angle between the narrow lip and the firewall.

With the sides still sufficiently open I looked at a way to best install the nose wheel steering mechanism. In order for the raked strut to turn freely and spread the force on a larger surface (than a nylon corner), I used the grinder to modify old collars into a wedge shape conform to the angle between the strut and the top of the nylon, and another one between the strut and the bottom of the F7a baseplate. No collars were furnished to fix the 5mm strut to the airplane, and the bell-crank had only a 4mm mounting hole. Instead of weakening the flimsy looking bell-crank by further hollowing it out, I preferred to grind away 1mm diameter from the top of the strut to mate both for duty. The nose strut rake is so pronounced that actuating this bell-crank could only be done efficiently by installing pull-pull cables and the servo at the same angle. For accessibility I choose to install the steering servo inverted under the engine mount box and used scrap Sambal to fabricate the angled supports. After a bit of eyeballing I drilled holes for the cables to pass, then mounted springs on the servo arm to act as servo savers in case of inevitable shocks when operating from soft grounds. I also made a few additional holes for an 1,5mm Hex key to be inserted through the firewall to operate (tighten) the tiny screws that lock the bell crank and wheel collar to the strut. Needles to say, much trial and error takes were necessary to complete that structurally important sub-assembly of the firewall and its systems to the rest of the fuselage.


Pic 4678 motor and NWS mount

Maybe not really part of the landing gear, but a logical time to talk about the tailskid. The one in the kit only provides the backbone, but is too short to act as a tailskid because the rudder would hit first anyway, and is too narrow to conform to the original Nipper where it is (just as the rudder and tail of the fuselage) a simple aluminum tube, and thus just as wide as the rest of the frame. The tail of the wooden model fuselage being much wider because of the material used, leaves no other option than to widen the bottom aft part of F6 as well. On the real Nipper a metal extended skid is bolted under the fuselage support protrusion so I opted for a similar scale design and first cut a metal plate with holes for fastening and for tiedown as per real. I used a Dremel grinder to make slits in the F6 skid to accept 3mm threaded wire. When those had been made and tested for length, I produced two extra 2mm ply skids that I glued in sandwich around the original. The threaded wire could then be glued in the created channels, after which a retaining bolt was turned into position, forming the base for the metal skid to be bolted on with self retaining 3mm bolts. That is not only a scale addition as such, but an effective protection for the rudder during landing and transport.

Pic 4685 tailskid


With the gear being in place, I undertook the covering of the lower fuselage between the firewall and F11. As explained much earlier in this build log, I opted for hard material like on the real Nipper (that uses bolts on polyester plates to facilitate maintenance in that area). With manhandling the model for transport or on the field in mind, I used kit scrap 4mm plywood to plank in plain between F9 and F10b, and 3mm balsa between F10b and F11, and to fill the open triangles between the firewall and the fuselage sides. I first painted the forward U-beam and its adjacent inside area white because it will be visible through the slit. In order to respect the dimensions and not encapsulate the gear pivot pins and bolts, all the plates were glued in-between the frames instead of on top. These 14 pieces were first prefabricated to exact size and plate angles for good fit, and together weighed 70 gr, for me an acceptable penalty versus delicate fabric covering. I started by gluing all the side panels to the frames, but not to the aluminum transverse gear U-beams. After drying and sanding, I glued the lower flat panels in place, thereby completely covering the space around the main gear aft support, but leaving a slit open to allow the main gear front part of the legs to move freely in and out. The front aluminum crossmember support has built-in end-course nylon stops, so the plate does not need to be opened so far as to allow the rubber spring mechanism to enter the fuselage, it remains outside of it, whatever position it assumes.

Without any guidance nor parts, it’s up to the modeller to find a way to close the gap between the sides of the fuselage and the frontal lip of F7a. A study of pictures from real Nipper show a flowing contour that changes from the flat belly at F9 to the rounded firewall at the nosewheel strut. As you already read, I previously had started filling the almost straight panels of the gap where possible. I then used a large scrap 3mm excess plank and glued it to the front along the width of the F9 bottom, and around the nose strut flat on the underside of the firewall. I then closed much of the remaining gaps using many separate parts of the remains of the 10mm balsa sheets in the kit, cut to shape and against each other at various angles.

Pic 4679 fillup nose



Lightweight builders probably will use 1,5mm wet balsa to create this intricate shape, but the weight penalty for using thicker balsa buildups that are then sanded into shape is not that bad and greatly enhances strength in an area that probably will be used a lot during manhandling or supporting the model during work on trestles. I had to use daisy chain mounted long nylon binders and plywood beams to force the bottom balsa 3mm plate in the correct bow to follow the one of the firewall. I again used PU woodglue for the complete nose so it would fill the many gaps created by the various assembly angles. When dry, extensive course grit sanding allowed me to create a nice flowing solid belly section from the firewall till just aft of the main gear.

With all of this completed I assembled the model for a first weight and balance to find out where the batteries best be located to get a CG at the prescribed 140mm Flight testing later proved 130mm to be much more stable in pitch. I added 180gr of lead on the engine mount to simulate the cowling and upper fuselage planking and only 150mm with both batteries transverse all the way forward on top of each-other. Still without ESC nor wiring or receiver, total weight at that stage was 5,1kg.


Pic 4686 progress

Four months after ordering, the Mk3 cowling was finally ready but as it was a prototype, both the shape and finish left much to be desired and weighed 250gr. With at least 5 different engine types powering the real Nipper, there is a multitude of cowl shapes and no two ones are alike. Once back home I used the Dremel grinder to first open up the large intake mouth and exhausts, and by judicious cheating with the lines, shaping and much sanding, I was able to correct the shape into a visual resemblance of the Nipper OO-MLD with the enclosed Rollason Ardem engine. When test fitting the cowling on the model I was astounded that the cowling was 11 inches wide, it dwarfed the 17 inch prop and unless allowing much prop-flow to pass through the cowling, would seriously reduce its efficiency. The enormous gaping intake also allows the emptiness of the cavity behind and the ridiculously small electric motor to come into full view.

Pic4700 cowling trial


On the real aircraft the lower half of the intake is closed with a baffle positioned just behind the intake lip. My cowling didn’t have a lip so I fabricated one by gluing curved balsa along the inside of the of the mouth, then sand it into a rounded lip. This balsa was kept flat on the aft side to allow the baffle plate to be bolted on it. The latter was fabricated out of 2mm ply so this large flat surface would be sufficiently strong to withstand the combined forces of airstream and propwash. The plate was also glued to the lip for solidity once the fit had been ascertained. I first considered attaching fake engine cylinders to that baffle plate, but after painting the interior of the cowling plus firewall and engine mount flat black, the cavity didn’t catch the looks so much anymore and I preferred allowing the maximum of the propwash to flow freely through the cowling and exit after a minimum of obstacles through the large exhaust openings. The engine cowl had such bad finish that I spent a week and numerous coats of filler and primer before I obtained a flowing smooth finish.

Pic 4704 cowl lip


Numerous coats of paint (red lacquer doesn’t cover well) were then applied to get an acceptable result of this highly visible and so characteristic part of my Mark3 Nipper. In the end the cowling weighed 275 grams and is held in place by 4 screws into wooden blocks that I glued to the firewall. I used scrap 4mm ply to fabricate a frame to make an access panel for the electronic compartment.

Pic4708 elec comp plus intr panel


On a real Nipper this is the fuel tank and has a top with changing radius that was created by soaking a 1,5mm balsa sheet in water so it could be bent into shape around the frame. That hatch on the model could only be hinged at the front due to its geometry, and shuts closed by gravity and aerodynamic pressure. A fake filler cap and the funny wire that sticks out on top (that’s the cork powered fuel gauge) was then fabricated and by some judicious cuts and routing was later used as a locking mechanism for that swiveling top panel.

In the prolongation of that comes the instrument panel with its fairing. This has to be constructed in a way that the canopy rail can surround it. I first made a balsa baseplate to which I glued a front bow for shape, and a ply 2mm shaped panel in which I first had drilled the 18mm and 26mm holes for the instruments. I then cut slices from used plastic tablet tubes to create the instruments that are sticking out slightly. When dry I used a combination of balsa and cardboard to fabricate the particular shape of the fairing and glare panel. Everything was then painted satin black. Pictures of actual Nipper instruments were then printed in the correct size and glued to the back of the tubes, scissor-cut clear acetate glued slightly inside the tubes for glass simulation. Scrap aluminum was then cut into shape and bolted inside the starboard wing elbow recess and pictures of these aux instruments glued onto it. Switches, knobs and labels were added and provide a truly scale cockpit as per original OO-MLD during the period I had the privilege of flying that Tipsy Nipper.

Pic 4751 finished cockpit


Chapter 11: Final fuselage 2 weeks

Covering the fuselage was done by first applying Oracover glue with a brush over all the surfaces you want the tissue to adhere to, but not on the parts it isn’t supposed to touch. Although this sounds straightforward, it is complicated because of the spacer stringers and the intricate shapes of the fuselage. I started by covering the bottom “tub” from nose to tail, and along the stringers where the fuselage width gets serious. Excess Oratex was then cut off above that stringer and new glue applied on top of the Oratex where it was glued on that stringer. I then cut two large surfaces of tissue that I glued all along the fuselage from the bottom stringer to the top of the fuselage spine, one to the left and one to the right. With minimal judicious use of the heat iron, only to the corners, I was able to fasten the tissue without it touching anything but the side of the stringers or spine, and with a sharp knife I got rid of any excess tissue that could grab existing Oratex and create local double thickness with a undefined line that is impossible to get rid of at a later stage. Only after the minimal overlaps were attained, the iron was delicately moved over the tissue without pressing it down over the various fuselage parts. That ensured the Oratex would only adhere to the outside stringers and not to the planked fuselage sides. The Oratex thus shrinks over just the outside shape but this is delicate work. The arch on top of the fuselage has such a small radius that it was impossible to avoid flats between the widely spaced frames, but pictures of real Nippers show the same problem.

After the covering it became time to insert the stab in it’s final position, that required substantial filling-up to decently support the sides and was difficult because spoiling the finish of the already painted tailplane would be all too easy. In retrospect I think that a completely removable bolt-on horizontal tail assembly would be a better option, also for transport. Two coats of primer and two coats of white paint were then applied to the fuselage, allowed to thoroughly dry before blanking-off for the blue and red to be applied. The maingear was then bolted back into position and besides the rubber suspension, I used a safety wire to limit the movement in case of rubber failure, just as on the real Nipper. After 5 testflights I found a better solution for the main gear suspension. Instead of putting the rubbers around the kit suggested tubes, I attach them individually to bolts inserted through the front gear beam, and attach individual nylon end of travel limiters through the other pair of holes. The safety cord/wire is also inserted through the tubes in case a limiter snaps. Later flight testing proved that no gear suspension whatsoever is desired on the ground, unless in flight or at rest, the main wheels better have a spread of 50cm. For my airplane weight (almost 7kg) 2 rubbers of half-width are mounted in eachother for each leg. That was the only way to avoid wobbling during takeoff or crosswind taxiing, and facilitates rotation during takeoff while providing a correct ground stance with pilot aboard.

Here is a picture of the practical gear setup at rest (in the air, for transport and stockage)

Pic 4749 mod maingear rest


And a picture of the same setup under load (taxi, takeoff, landing)


Pic 4750 mod maingear spread

With the fuselage finished I was able to bolt the prefab outside throttle control box onto it, then run a fake throttle pushrod through the exhaust, and using scrap tubing I made some fake exhaust stacks that I glued onto stubs.

I then did a first complete assembly of the model which tipped 6200 gr on the scales, but even with the batteries completely forward in their compartment, the CG seemed too far back at around 170mm. Depositing lead in the middle of the cowling, I needed more than 400gr to bring the CG around the kit recommended 140mm mark. That was unacceptable so all hopes to quickly fly the model vanished, I studied the various options to move the batteries in front of the firewall. Although there was ample space in the cowling, I had no desire to at each battery change remove the prop and cowling, or slice part of the cowling off to make all too visible access doors

After removing the fake exhaust stacks, the cowling exhaust opening seemed large enough to slide a battery through it. I marked those positions and looked for a way to support the 4S4000 batteries left and right of the motor and its mount. Having those 400gr (each) batteries rest only on the cowling would not be a good idea in case of hard landing, so a support bracket was fabricated using the firewall and engine mount extension as solid anchor points. I thought running the battery wires in the fuselage directly in the former battery compartment, and blocking-off everything both visually and physically with the removable fake exhaust stacks, was the most practicable solution to resolver the CG problem whilst allowing battery changes with a minimum of fumbling.

In the end I had to enlarge the cooling air exhaust openings by grinding away some more material to allow the cowling to slip over the new battery holders, but only a purist would notice. The battery mount assembly was made by trial and error and consists of a 2mm plywood plate to which are glued both 3mm balsa battery coffins. Those are open on both sides, for cooling air at the front and for battery insertion at the back. To the underside of the basic plate I glued 4mm plywood transverse stiffeners and support triangles that spread the load onto the lower firewall, plus aft glue/bolt-on 10x5 Sambal wood blocks to anchor the complete assembly to the middle of the firewall.

Pic4722 coffin assy


For negative G’s/turbulence I ran a carbon rod over the top of the coffins through the engine mount box. That complete assembly weighs about 90gr and looks rather funny on itself, and once fastened to the nose of the model (but without the cowling) rather like a twin-engined Cri-Cri aircraft nose with ramjet engines. After being painted flat black it is hardly visible and the front exposure looks a bit like fake cylinders above the baffle plate of the cowling

Pic4724 coffin on aircraft


To prevent the batteries from sliding out during acceleration or climbs I designed a tool-free system whereby the orifice through which I pass the battery EC5 plugs closes and holds the fake exhausts against the aft side of the batteries, simple and efficient. Note how I already can connect the batteries at an early stage but only make the final 4S + 4S series connection with a single anti-blitz EC90 connector just prior to taxi, an important safety feature.

Pic4731 elec compt


The cg, throws and lack of down-thrust that were suggested on the first series of plans made the model almost impossible to fly, I needed all my skills to land it after the maiden. After 5 adjustment flights I noted following figures to obtain a stable model that is pleasant to fly: Together with some more lead in the nose (now 177gr total) I achieved a CG of 130mm behind wing leading edge. Throws were adjusted to 50mm up and 25mm down for the ailerons, 30mm up and down for the elevator with a neutral point just slightly up. 65mm left and right deflection for the rudder with +5% throttle to rudder mix. Zero throttle to elevator mix and an aileron expo to 35% (with 50% differential), the rudder expo 35%, and the elevator expo 25%. The model after all the mods and ballast weighs 6,9kg flight ready. This calculates to 197watt/kg which is a realistic figure for such scale model. The ensuing wing load amounts to 84gr/dm˛ and that provides more than adequate gliding capacities on this low aspect wing, it is not the brick I anticipated it to be. Use three half rubbers per gear leg for the damping and correct ground and air stance.[/I]

Chapter 12: Overall building assessment

A woodworms dream that fits together well with basic woodwork technique. The absence of full size plans and assembly guide worth that term is not insurmountable for an experienced modeller using sound common sense, but might cause serious problems to inexperienced model builders who do not think ahead or fail to analyse things before they start gluing. The balance between easy endless building and challenging moments requiring brain power is perfect. The model is more to scale than I initially expected and can be assembled relatively straightforward into a fun flying machine, or provides an excellent base for modelers who desire to add more scale details to the finish. Scale Dreams offers access to other builder’s assembly pictures and promptly answers in great detail any concerns you might have. Furthermore, any problematic parts are exchanged, but be aware this is not a factory but a hobby-business and although willing and capable, the guy behind the firm has another job for living and therefore you must exercise patience for any special or follow-up requests.

The quality and cut of the wood is excellent, but landing gear hardware quality is less than perfect. His first Mk3 cowling was of poor finish quality and the so called plans showed impractical CG and control throw information.

Almost everything is in the box, these are the items you’ll need extra to get a basic model flight-ready: engine mount, motor, fuel tank or batteries, 6 servos and wires, receiver, cover material and paint, lots of wood glue, 6mm nose-strut collars, plus imagination and dexterity if you want to make it more scale.

As I already incorporated all the essential feedback of my trial flights in this build log, you should now be ready for the maiden flight. If interested in the details of how I conducted and improved my model’s behavior step by step, you can read the detailed test flight reports in Chapter 13, but that is optional.
Last edited by BAF23; Oct 18, 2018 at 04:42 PM.
Sep 23, 2018, 10:54 AM
The sky is the limit
BAF23's Avatar
Thread OP

Tipsy Nipper T66 Mk3 scale 1:3 Part 4: flight testing


Chapter 13: Maiden and development flights

On the 12th of August 2018 I had the model with me during a BiGGS glider-meet in Battice. They have a 60x20m tarmac runway and after the morning briefing I performed the range check and taxi test and made a short maiden flight before the regular towed glider activities started. Ground handling was definitely not its best asset. For that flight I used 3 rubbers which reproduced that unique but different realistic ground stance during empty aircraft static and pilot carrying taxi or flight, but the spongy effect augmented by any nose wheel steering deflection (of the forward raked wheel) caused a duck-like wing waggle even without the aircraft moving forward. Due to the narrow gear and large flying surfaces, the aircraft has a tendency to quickly tip over on its nose and one wing tip during downwind taxiing, on uneven grass this becomes even worse when the nosewheel hits a bump or comes out of a depression (I cannot recall the full-size Nipper to display such bad ground behavior).

With such ground handling “qualities” I skipped any fast taxi to rotation attempts and elected to slowly open the throttle to get it off the ground as quick as I dared. I was happy it tracked straight as an arrow during the takeoff roll (my 5% throttle to rudder mix guess proved correct) and as I applied just a bit of elevator to rotate, it surprisingly rocket-ted in the air and I neutralized the controls. Luckily the power to weight ratio was more than sufficient to allow it to shortly climb at a ridiculously steep climb angle without loosing speed. A few meters above the ground I dared to apply a bit of down elevator to reduce the climb angle but it abruptly leveled off, it was obvious pitch sensitivity was way above my comfort level but after a while I got this wild horse sufficiently under control to perform a few patterns during which I hardly had to apply any trim and flew around at 1/3rd to half throttle. The short fuselage didn’t impair directional stability and the 50% aileron differential seemed on the spot to allow me to make turns without adverse yaw tendency.

With that over-sensitive elevator I had no desire to further explore the flight envelope except that I climbed at full throttle to a safe altitude where I performed some slow flying and a stall. I was surprised how good lateral stability and control-ability was during slow flight and further reduced power to the stall which was very docile. I them trimmed it out for cruise again and performed a CG dive check during which it even had a slight tendency to increase the dive angle, a sure sign of neutral to negative stability and the explanation of the pitch sensitiveness. I never expected that ugly duck to glide so well during the idle descend but these few handling movement gave me assurance that landing it could be done (even with the overly longitudinal stability problem).

I then flew a couple of low approaches out of loose 180° turn patterns and for the third one I felt confident for a landing on the hard runway. I flew a stable long approach with a little power and the nose a tad above the horizon. The idea was not to use much elevator (with a risk of over-correcting) but letting the aircraft gently settle itself using the ground effect and the large gear travel to absorb the rate of descend. The touchdown point was a little long on the available tarmac and as the aircraft settled softly without a bounce, I cut the throttle and rolled out. Just before the end of the hard runway I tried to turn it (at very moderate speed) but it immediately tipped over and a fellow pilot helped putting it back on its 3 wheels in my direction. Downwind taxi had to be done with care and as I intentionally left the runway for the grass I tipped it over 3 times, even using the correct aileron deflection for downwind taxiing.

Looking back at that maiden I still have mixed feelings, The scale gear looks very realistic but is unsuitable for operation on anything but flat surfaces. Downwind taxiing (in anything but very light wind conditions) should be avoided but the good news is that when tipping over, the long gear allows for sufficient clearance for the prop not hitting the ground. The model attracts attention on the ground and in the air but the factory CG of 140mm is too far aft for initial relaxed flying. Further development flights have to be performed to open up the flight envelope after a good balance has been found between CG position and elevator throw. If you’re not a scale nut as I am, choose for a rigid instead of a scale gear, especially if you operate from a not too well groomed grass surface.

Claude, another experienced early Nipper kit builder arrived later that day and told me he used longer gear rubbers to widen the gear width and he also did move the CG forward and reduced the elevator throw, I wish I had known that in advance. Soon after forwarding the part about the maiden flight to Scale Dreams he replied that in his February newsletter he had mentioned to reduce the elevator throw from 50 to 20mm either way, and to use 75% expo. No wonder my maiden flight felt “a bit touchy”, but I just relied on what was mentioned on the plans in the kit and omitted to dig in his old newsletters for the latest info. Reducing the throw to less than half was impossible without drilling extra holes nearby the center of the servo arms, but that finally did the job.

The second newsletter also mentioned a weak nosewheel axle and the first builder who stiffened up the main gear suspension with extra rubbers but that caused the main gear to remain in the “narrow high position” necessitating the lengthening of the nose strut, something I definitely don’t agree with considering the already problematic stability with the gear in the open position. I decided to go the opposite way, by removing one rubber (keeping just 2 on the model) and opening up the maximum allowable spread distance between them, I hoped the gear would be resting in a wider and deeper stable condition for ground maneuvering whilst only pinching up in the air or static on the ground. This would still provide for a little shock absorption upon touchdown but much less wallowing during ground ops because it remains out of the “elastic zone”.

To do list after maiden flight
Add lead in nose
Reduce elevator throw to +-20 with expo
Reduce NWS throw
Test suspension with 2 rubbers
Make transport/stock cradle

Furthermore I much reduced the movement of the nosewheel steering to minimize the tendency to tip over. After bolting70gr of lead under the front end battery plate, I reassembled the model for the second flight. The job was finished by making a compact convenient home/transport foam cradle for the complete model including the wings.

Pic0366 cradle


Second testflight

A day later the wind blew calmly in the axis of the Zwartberg tarmac runway and before it picked up I was ready to fly. The reduced nosewheel steering travel much eliminated its tendency to tip-over, on hard as well as soft surface, whilst still allowing short radius turns to me made at slow speed, but then it still seriously waggled its wings on the ground and had trouble leveling even in moderate wind. A powercheck showed 51 AMPs and1140Watt on the telemetry. After brake release I slowly opened up the throttle, the model tracked straight but after a while again leaped into the air without me applying elevator. I was surprised how much down I had to give to get it level again. I used all my down elevator trim to keep it level at 2/3rd power. Further reducing the power caused the nose to drop till I applied some up stick to stabilize the dive angle. It was easier to control the pitch angle with engine power than with elevator! The much reduced elevator throw combined with 35% expo meant I needed serious stick inputs to keep the nose pointed the direction I wanted.

After a few attempts at level circuits at different power levels I was assured I had sufficient control to proceed to the next step, so I climbed to altitude, stabilized in level flight and pushed it over into 30° of dive. As it didn’t give any sign of recovery I used up elevator to level off, the effects of increase in lead weight probably got eliminated by the full down trim elevator position in this short CG dive check.
It thus was obvious that the lack of engine downthrust (not mentioned on the plans nor newsletter) or throttle to elevator mix caused this longitudinal instability and that it hadn’t been solely the oversensitive elevator that had caused my troubles during the maiden flight. I climbed again for a stall check but with without idle power because I was afraid of running out of up elevator for the recovery. As I encountered no uncontrollable behavior flying near or in the stall, I came down to traffic pattern altitude for some turn reversals to check the adverse yaw again, it was non-existent.

I found it useless to further explore the flying envelope unless the pitch problem got fully resolved, so I decided to fly a couple of approaches to find an appropriate power setting for me to control the approach angle/speed to touchdown point with a minimum of power changes. As the second approach looked stable, I waited till the model was almost at the ground till I slightly raised the nose without touching the throttle. It settled very softly and I cut the power, waiting till almost standstill (10m roll-out) before backtracking. Despite the waggling of the wings I was able to taxi back downwind and was as bold as to leave the tarmac for the grass, where it again behaved itself without tipping over.

Even with a spare set of fully loaded batteries on hand, I decided to call it quits instead of attempting another flight with on the field guesswork about the throttle to elevator mix that would be incorrect after changing the motor downthrust angle later on. Driving back home I had my brain in overdrive to figure out how I could modify the main landing gear suspension because it are also the incorrect scale angles of the gear geometry that cause this undesirable waggle. On the real Nipper, the sandows between the gear are within the fuselage when at rest, on the model the rubbers are well below the fuselage to start with, causing too much up and down movement of a single gear leg instead of the rubbers keeping equal lateral pull on both rubber anchor points. Back home I made a short to-do list of shortcomings I wanted to correct before attempting a third flight.

To do list after second testflight
Increase lead in nose
Install backup power for receiver
Reduce elevator expo from 35% to 25%
Dial in a bit less NWS travel
Create mechanical engine down-thrust of about 2°
Watch ELE trim (full down now)
Dial in throttle to ELE mix
Design and modify main gear suspension system for gear stability
Modify angle in cradle by raising nosewheel but lower aft wing bed

Third testflight

A week later all conditions were met for a third flight (on tarmac). The reduction in nose wheel steering max deflection (I’m now down to 40% of the original setup) was beneficial to reduce the waggle tendency and still allowed me to turn around in a few meters. Taxiing in even moderate crosswind caused a bad wing low that was impossible to level using remote control, downwind taxi felt comfortable as long as the pace was kept at half walking speed.

Takeoff with slow motor spool-up was uneventful and I even had to rotate (at 3/4 throttle) to get it airborne. After establishing a shallow climb angle I slowly opened up to full throttle and was glad to see no pitch change with that. After two patterns I reached cruise altitude, reduced the throttle to almost halfway an trimmed the aircraft. I needed only about 1/5th of down trim travel for handsfree cruising. Applying full power didn’t change that, reducing power to idle it slowly but increasingly dropped the nose. Motor downthrust seemed to have pretty well corrected the previous flying problem, but the model still didn’t feel longitudinally stable.

I applied power and climbed to safe altitude where I again checked for level flight at constant speed before nosing over into a 40° dive. As I feared, the model kept on going and it even very slowly augmented the dive, a sure sign the CG was still too far back. This had already been confirmed during the second flight but changing too many parameters at the same time keeps you guessing what caused any change, so I flew the third flight with the same amount of lead in the nose, but with about 2° of engine downthrust and +5% of throttle to elevator mix. Installing a 2S950 battery power in the electronic compartment to duplicate the ESC BEC power had not added a substantial enough nose weight.

I climbed to altitude again to perform slow-flight and a couple of stalls. Even with the CG too far back, 20mm of up elevator was insufficient and the model didn’t want to stall even with the elevator fully up. That was a warning sign for me not to perform aerobatics yet and to just descend and perform a few patterns to feel the elevator response in final. Luckily I had reduced the elevator expo to 25% because it felt rather sloppy at slow speed. After 2 intentional low-approaches (during which I experienced some lateral instability in response to the wind), I called out for landing and flew a stable approach but was hardly able to flare due to lack of elevator travel. Even so, the touchdown was smooth and so far, that airplane never showed a tendency to bounce.

The lack of headwind component caused the aircraft to only very slowly deplete its speed (on the tarmac) and after a 60m ground-roll it overshot into the grass where it quickly came to a standstill. Being lazy, I opted to turn it slowly and return over the soft at a slow pace and using little steering inputs. I was surprised how much power I had to apply but apart from that, its behavior on grass was without flaws and the wings never dipped. The many fellow pilots on the field congratulated me as much for the scale flying as for the choice and realization of this very attractive model. I felt very happy about the immense progress in ground and flying capabilities and am convinced that with a bit more fine tuning this will be a pleasant to fly addition to my aircraft collection.

After shutdown I noted the values and looked carefully at the trim indicators versus the flight controls position before taking the batteries out and dismantling the model for further modifications back home. Before the next flight I will keep the downthrust, eliminate most of the mix, mechanically neutralize the elevator trim and have a bit of up elevator whilst augmenting the throw. For all those works, I wrote following to do list.

To do list after third testflight
Increase 50gr lead in nose
Increase elevator throw to 30mm
Guestimate elevator mechanical trim change according new CG and old tx setting
Glue and/or make hinges for pilot back panel
Investigate lateral (in)stability, Increase aileron expo (20% set)
Reduce throttle to elevator mix to 2%
Timer reset to 0 before takeoff
Dial in switch C warning when TX power on
Find system for better gear stability
Modify angle in cradle by raising nosewheel after lowering aft wing bed

Fourth testflight

End of August the designer’s club celebrated its 20th birthday with a great show. Four Tipsy Nippers were present, one had engine problems and I kept mine in the parking lot due to force 4 winds with rolling turbulence from nearby hills and trees. The two that flew partly or completely overturned upon landing on the well groomed grass, but survived with no significant damage. We made a nice family picture before dis-assembly and together enjoyed a nice strong local Gembloux beer before returning home. Even with the strong winds the event was a success, about half the models flew (I flew a second tow with the Foka to fill a hole in the program), and we were delighted by the planned flyby of a real Stampe and the four SF260 Marchetties of the Belgian Red Devils national aerobatic team.


4 nippers 079

A few days later the wind had calmed and blew in the main axis of my second club in Tongeren. After a field repair of a bad solder on the Amp sensor I further assembled the model and reflected how safe I had been to perform a thorough preflight inspection and installed that second power source. If that solder had disconnected in flight I would only have lost ESC and thus engine power, but kept power to the receiver and servos thanks to the backup battery through the diode.

As this was my first complete flight on a relatively smooth grass surface, I taxied cautiously, but even so there was no way I could keep the wings from pointing away by about 20° of bank from the slightest of breezes. I therefore elected to takeoff as much in the wind as I could but it still was frightening to accelerate to rotation speed with that uncontrollable bank angle. I therefore rotated earlier than planned just to get away from the ground, which worked fine, and immediately the wings leveled and after releasing the back pressure I was in a comfortable climb. The elevator sensitivity and response felt good but the ailerons still were a bit touchy throughout the flight.

After leveling off a bit higher than pattern altitude I trimmed the aircraft by applying just a few clicks of down elevator. Opening up from half to full throttle did not cause any pitch change so I was assured future go-arounds would be no problem. Turn reversals again happened without adverse yaw (50% differential and no rudder mix) so I soon gained altitude to check if I now had sufficient elevator travel to obtain a full stall. It barely did on idle power but with a bit of throttle the model could be flown (and controlled) at a ridiculous pitch attitude and walking speed. At full aft stick it stalled by dropping a wing, but very docile and immediate recovery ensured minimal altitude loss.

I then leveled off and trimmed it out at cruise power, then went into a 30° dive from which it showed no sign of natural recovery. To be sure I re-trimmed with power off in a glide and then pushed the nose over, no self recovery either. Despite the additional 50gr of lead in the nose, the CG still felt more back than I like (I prefer to fly with about 28%MAC whilst most modelers fly between 30 and 33%MAC) . This was also confirmed by minimal trim change at whatever speed. Just a bit too much of nose-drop when reducing to idle power was an indication that even the 2% of throttle to elevator mix was too much and could be eliminated.

I then felt confidence to further explore the flight envelope and performed a loop keeping positive G’s at the top because of loose stuff in the electronic compartment. Here it showed that plenty of power was available to make a loop diameter slightly larger than the real Nipper. I raised the nose too much for a left aileron roll and the model didn’t change pitch at all during the entire roll. For the right roll I then started with only about 5° of pitch and after rollout was totally level. The roll rate was very scale either way, confirming the 40mm up and 20mm down throw were good for the model. I tried a stall turn to the left and got surprised by the hefty rudder response, to be taken care off during later flights. After a couple of low passes in cruise power I got the warning of half battery power so it became time to practice a few approaches.

A farmer with his tractor working on the fields under the downwind and baseleg restricted my possibilities for a normal traffic pattern. I doubled the normal heights in baseleg and thus was forced to fly final in idle. I was surprised it didn’t fall like a brick but kept its speed very well, much better than a real Nipper. After two low approaches down to about a meter, I flew the next pattern to a full stop, therefore seriously reducing the speed in final. Reason for that was that I didn’t want to land on the nosewheel (others have reported bent or broken axles) plus I wanted to roll out as short as possible out of fear of turning over due to that impossible main gear geometry. When at a good meter above the grass, I halved the dive angle, the speed bled off rapidly without breaking the sink rate much. Even the 30mm of up-throw was insufficient to flare as I desired, but the model settled without harm and quickly came to a stop. Too smooth a touchdown would not allow the gear to spread completely thus allowing it to wobble during the roll-out. To taxi back I had to level the wings with my hands because the minimal wind already caused awkward bank angles on the ground.

That fourth flight was a success (in the air) and proved that even with that lousy scale gear, it can be operated from a grass field when pointing straight into the wind. I recon that for landing, a bit of power blowing over the elevator is desirable till the flare is completed. After that, the nose will stay high enough to perform a soft landing without bounce from the rubber dampers. Following works will be performed before the next flights.

To do list after fourth testflight
Add 50gr more lead in the nose (+55=177gr) + bat 2S2100(50gr extra)
Rx battery from 2S900 to 2S2100 (for endurance and balance) with plug adaptor.
Augment aileron expo from 25 to 30%, rudder expo from 25 to 35%
Eliminate THRO to ELE expo
Leave elevator expo at 25% till optimum CG is found
Check all soldering on power wires
Add x-member aft cradle
Affix loose components in electronic compartment

Fifth and sixth testflight

Early September I analyzed the options I had regarding that main gear problem. By modifying it to a completely fixed aluminum gear it not wouldn’t look good in the air and furthermore the model wouldn’t fit on my narrow cradle anymore, neither on top of my den, plus take more width in the camper, so that was out of the question. Having discussed with the designer of the kit we already came to the conclusion that having the suspension rubbers tied to each leg at the same time did nothing but aggravate the lateral instability on the ground. After examining the many possibilities of making the suspension from port and starboard gear independent from each other (even considering use of heavy-duty RC-offroad shock-absorbers), I decided to try keeping most of the existing provided hardware but make some simple modifications to the rubbers and their attach points.

I did not want to cut much of the belly open to drill into the alu gear beams but noticed that with a knife I could relatively easily free the front beam from the planking I glued it to where it exited the fuselage. I unbolted both gear legs and removed the pilot’s seat-pan to access the bolts and nuts that held the front beam in place. With the minimum slit width I had left under the fuselage, I found it easier to extract the beam sideways, which could be done by taking away just a few millimeters of balsa where at the place where the pins that hold the nylon cushions stick out. I then drilled 2 holes through the beam, vertical from where the rubbers were attached on the gear arms, cut the existing rubbers in half so they would fit in-between the beams, then used 3mm bolts and nuts to attach them to the beams. I affixed both rubber halves one into the other to augment their strength and slid the beam back into the fuselage.

A first unpowered test just pushing the completely assembled model (with batteries et all) along my terrace was not very encouraging, Even with independent suspension it still wallowed and the rubbers were too weak to keep the wheels at an acceptable width. Remembering that the safety wires between port and starboard had done nothing to improve stability because they too caused one gear leg movement to influence the other, I took the beam out again and drilled an additional two holes to which I could attach individual nylon endpoint straps. These ensure each gear individually is sprung by its own rubbers to absorb the landing weight, but restricts each leg to spread further than the normal ground stance with pilot aboard. As an additional safely, I added the safety wires that (like on the real Nipper) connects both gears together so in the event of failure of rubbers or nylon a leg couldn’t pivot too far up and out. This wouldn’t influence the other gear anymore because when it comes into action at least one nylon strap will have reached its endpoint.

I then added a third half-rubber around both previous ones and for the weight of my model this is perfect to allow the wheels to be together on the ground (static and transport) and in the air. Just as with the real Nipper, when you then taxi out (with the added weight of the pilot /fuel/batteries) the wheels spread apart during the first few feet and the tail lowers into a stance that better allows rotation to be achieved at the correct speed. On the model the gears spread till they reach the end of the nylon strap movement, firmly keeping them in place and completely eliminating any tendency to waggle during taxi, whatever crosswind picks up under a wing. The rubbers are sufficiently soft to allow the weight of the aircraft to quickly settle against the nylon endpoint again, whilst being sufficiently strong to absorb any normal landing force and return the gear legs to their narrow static ground or airborne position when necessary.

Rubbers and nylon limiters are looped around four 20x3mm bolts in the front gear beam, exactly vertical to the position they take when retracted (rubber) or extended (nylon), and around the optional scale gear provided rubber attach points on the legs. All rubbers are attached on one side, the nylons on the other, while the safety wire is routed through the attach tubes. The only drawback is that you need to dismantle the gear and slide the front beam out each time you’ll have to replace rubbers, but that is a small price to pay for having a workable scale gear with minimal modifications to the one provided (in option) by the kit producer. If you (just as on a real Nipper) lift the tail after every flight to alleviate the tension on the rubbers by allowing the gear to spring into the narrow position, and also stock the model that way, your rubbers won’t have to be replaced too often.

I also swapped the lighter 2S950 receiver battery for a 2S2100 placed against the firewall for added nose weight and daylong flying endurance. Together with some more lead in the nose (now 177gr total) I achieved a CG of 130mm behind wing leading edge. I eliminated the throttle to elevator mix and augmented the aileron expo to 35% (still with 50% differential), the rudder expo to 35%, but kept the elevator expo on 25%. With that configuration I went to my grassfield half September to conduct the fifth test flight in light crosswind conditions.

Starting from the narrow gear condition, the legs spread within 2 meters of ground-roll and I finally was able to taxi wings level over a prepared but not smooth surface, not having to take into account the wind nor nose wheel steering angle, what a luxury. After lining up and checking the wind (light x-wind from 90° starboard), I gradually opened up the throttle with neutral flight controls, but as the aircraft picked up speed, I saw the wind lift the starboard wing with the nose tending to dig in. Right aileron corrected that but by then I had eaten up most some distance so I applied full throttle and rotated positively. The Nipper climbed out steadily, much more stable with this new CG. Once at altitude I leveled her off and trimmed in pitch, roll and rudder to achieve straight and level flight at about half power. I noticed the engine/prop noise to be weird at about 75% throttle, but normal below and above that.

Time for the actual test card items. I started with the power range exploration, side on to check the pitch reactions when opening or closing the throttle, flying away from me for the yaw/torque effect these power changes caused, they were almost negligible. Next came the power-off stalls. I noted that even with full back stick the nose just dipped a few degrees causing it to self-recover and immediately enter a new cycle. Very funny to watch but also very comforting seeing that no wing drop occurred, what a docile aircraft! I then opened the throttle halfway, trimmed the model and pushed into a 45° dive from which this time it slowly raised the nose to prove my CG move to 130mm was bang on the numbers. I then performed smooth aerobatics: steep turns, slow horizontal eight, looping, cuban-eight, rolls, and stall turns, the latter showed that the rudder expo is getting better.

Time to come down for a few low passes for the crowd, and getting used to her behavior in the traffic pattern. After hearing that strange noise again during the go-arounds, I decided to full stop instead of making touch-and-go’s. The idle landing was uneventful, with plenty of elevator travel remaining during the flare. Taxi back was now piece of cake, confirming I finally had resolved those troubling gear problems. After shutdown I got an applaud from the crowd. Everywhere I show up with this unusual model, people are very attracted to it, and now that it flies similar to the real Nipper most of them saw during airshows during the sixties or seventies, their admiration grows even more. Here are the images from that flight, edited by the occasional cameraman to cut out all what happened at too high an altitude for his smartphone camera to capture decently.

Nipper (2 min 42 sec)


After resolving a few hardware problems with my batteries and transmitter, I performed the sixth and last testflight, which was more of a confirmation flight for the latest setup. Whilst charging the batteries I had been thinking about the takeoff problem and came up with the following reasoning. As the speed during the takeoff roll increased, the wing developed lift and the weight on the main gear diminished, causing the rubbers to become effective and slowly narrowing the gear, thereby allowing the crosswind to catch under the right wing and lift it. The weight on the nosewheel (batteries and lead) still kept it firmly on the ground, causing the model to dip dangerously forward to port. To cope for this phenomena I decided for the next flights to apply full up elevator during the acceleration phase of the takeoff roll, with the intent of slowly but gradually releasing the back-pressure when approaching takeoff speed. This system worked well and the Nipper remained wings level during the takeoff. I since also keep the sick full up after the landing and during taxi. While others reported broken nose wheel axles, I haven’t encountered that problem so far and continue operating on the original hardware. As suggested by another RC pilot, I might replace the rubbers by suitably sized and strength springs for durability.

That last test flight was mainly used to confirm previous assessments, but also to find out if the strange noise only came up at a certain RPM, or was also speed related. After some more exploring aerobatics I ended the flight with again a perfect landing. I then blocked the model between my legs and played the throttle again. The bystanders confirmed the noise did not come from the airframe, there was no reverberation going on and no vibrations felt as it transited or spent time in the noisy zone. After shutdown, motor nor axle showed any play at all. We could only conclude that the noise probably emanated from the interference of the propwash with the very closely positioned huge cowling which blocks 3/4 of the prop passing the horizontal, and only 1/3rd passing the vertical. With 40% power remaining in the batteries I still was able to develop 45 amps which resulted in 1360Watt. The model after all the mods and ballast weighs 6,9kg flight ready. This calculates to 197watt/kg which is a realistic figure for such scale model. The ensuing wing load amounts to 84gr/dm˛ and that provides more than adequate gliding capacities on this low aspect wing, it is not the brick that I anticipated it to be.

After that flight I reset-ted the fail-safe of the receiver to the new trim values and registered that the ailerons were perfectly level and had a throw of 50mm up and 25mm down, the elevator being very slightly up from streamline and with a throw of 30mm up and down, the rudder throw being 65mm left and right with still +5% throttle to rudder mix. Those values work for me and could provide a good starting point for test flying similar built model Nippers using a stable 130mm CG position.

I now consider the testflights for this model to be completed and am more than satisfied with the performance of this scale reproduction. It is well worth the kit price, and with a few changes and extra build time, results in an original docile yet aerobatic capable model that is capable to transport the kid’s pet animals and puppets during those lately very popular family events. With the choice of cowlings and the many attractive and highly visible color schemes, you’re not to encounter many similar looking models on a field at any time. I hope to post a better film of the Nipper’s flying qualities next season so stay tuned for updates.

To do list after sixth testflight
Adjust rudder neutral point to center ailerons (1mm)
Battery testing/repair and modify port batt cable fuselage entry
Adjust canopy opening lock
Aft port wing lock pin redesign?
Touch up white (glue marks) and red (bottom) paint
Eventually replace the rubbers by springs on the main gear

Further adjustments after flights 7 and 8

Back home I didn’t trust the engine noise and the motor was dismantled after removal. Two of the magnets were loose and had further damaged the interior. Furthermore everything was black and because I was sure not to have overheated the motor, this must have occurred with the previous owner who just sold it to me as as a perfect motor. With AXI having replaced that motor type by a V2 variant (of slightly different dimensions), I drove to one of the last model shops still selling the original 4130-20 motor to pick it up in Germany and installed it in my model.

The following day I undertook flight 7 on grass again the first weekend of October under fairly windy conditions. Taxiing to the cross runway was done carefully but at one point the crosswind lifted a wing a bit and the plane was close to tipping over and could not be put wings level again using only remote control. The flight was uneventful and proved the new CG was a lot better. More aerobatics were performed showing a better balance on the rudder but too touchy elevator. The landing itself was ok but turning away from the wind to taxi back, the wind got it again and this time the model even somersaulted. Luckily the strong rudder hinges withstood the impact well and after turning it upside down I pushed the model by hand to the parking to avoid further taxi mishaps.

After the elevator expo was raised to 50%, batteries had been recharged and the wind had decreased I got flight nr8 underway. Cautious taxi and takeoff directly in the wind caused no more problems but I still didn’t like the model climbing so steep after liftoff due to the necessary up elevator to keep the main gear spread during the takeoff roll. Flowing aerobatics were easier with the new expo and I performed my first intentional spin. The CG seemed good because even with full up elevator the pitch remained steady at a comfortable -30° during the five turns. Upon release of the full left rudder and the back pressure, the model instantly recovered without much altitude loss. At altitude I then performed some tight figure-8 turns, a real eye-catching maneuver in the Nipper. Pulling extremely tight caused a predictable snap by the high wing but at that high throttle setting no altitude was lost and the aircraft just kept on turning the other way, very comforting behavior to later perform that at low altitude within the boundaries of the field. I then came down to about 40 meters for a complete aerobatic sequence but my attempts at performing another typical Nipper maneuver, the side-slip , were not so successful. I was unable to find the correct rudder deflection to allow a good balance for a cross controlled level flight pass. That will have to be further investigated.

I flew a powered final and settled decently on the runway, this time taxiing all the way back to the parking without trouble. Batteries still had 40% remaining so their capacity is perfect for this model and my flying style. Looking at the model with a friend he asked if the gear couldn’t be spread wider because most pictures of manned Nippers show them deeper against the ground when taxiing. Such modification would make sense, offering more lateral stability and facilitating takeoff rotation by allowing it to fly itself of the ground. As setup at that moment, the gear was artificially kept from spreading further by the 2 nylon retainers and the safely cords. Making all of them a bit longer was no big deal but the strength of the rubbers would remain the same and the tendency to lift a wing in crosswind would augment. Reducing the number of rubbers from 3 to 2 could do the trick. In fact even one rubber or spring could serve the main purpose of bringing the gear together in the air and on the ground, but would be insufficient to dampen hard landings. The model being easy to land I am willing to diminish the absorption by trying it with two rubbers.

Update after flights nr9 and 10

Those flights proved the latest gear modification was a improvement and the model finally has the ground stability it deserves, even on uneven grounds, crosswind or turns at fast walking pace. Two half rubbers are more than sufficient because during takeoff roll the increased lift already caused the Nipper to become level again, so some aft stick still has to be applied. With the main wheels now spread at 50cm wide, the ground stance looks completely scale. The model lifts-off much quicker, drastically reducing the takeoff roll without sacrificing lateral stability. In flight and static on the ground (when tail has been lifted as per real Nipper) the gear springs back to it’s ultra-narrow track and during landing the widening of the track and remaining two half-rubbers ensure sufficient cushioning. Low Aerobatics are now a delight and low side-slip passes are becoming easier. Slow flying and minimum radius turns are very realistic but I am still surprised how well it glides (from pattern altitude) because I often come in too fast in idle instead of bringing it in with some power. This concludes the description of the long development flying period. Do not expect further entries unless something drastic happens.
Last edited by BAF23; Oct 18, 2018 at 04:28 PM.


Quick Reply
Message:
Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Tipsy Nipper Motor Burnout balch Parkflyers 5 Sep 18, 2001 05:18 AM
Tipsy Nipper steve f Parkflyers 0 Sep 06, 2001 07:02 AM
Tipsy Nipper, Prop selection for S400 2:1 gearing? Red Baron 47 Power Systems 5 Aug 01, 2001 07:38 PM
Ok guys! now I'm confused...Park Flyer Tipsy Nipper from Hobby-Lobby John E Electric Plane Talk 16 Jul 31, 2001 06:17 PM
HL's Tipsy Nipper cruzomatic Electric Plane Talk 18 Jul 20, 2001 07:11 PM