Polikarpov U-2/Po-2 build article.

by alibopo | November 7, 2014 | (11) Posted in Projects


Polikarpov U-2 / Po-2

NOTE 16/01/2015 - There's a plan link at the foot of the page. :)

UPDATE 28/03/2015 - I've been collecting and updating all my articles, planes and plans on my own website on this page; RCplanes It's still a work in progress, but there's a few extras and beta-stage projects that don't appear here on the Flite Test website yet.

Having decided to build another biplane, I chose this aircraft because of its distinctive design and rich history.

The aircraft was designed by Nikolai Polikarpov and the prototype first flew in January of 1928. Serial production in the Soviet Union started in 1929 and ended in 1953. The original U-2 classification was later altered to Po-2 to acknowledge the designer.

License-built versions, CSS-13’s, were still produced in Poland until 1959. The upper estimate for build numbers is around 40,000, which apparently places it at the top of, or very close to, the all-time record holder for production figures of a single design.

Pre-WW2 it was used as a trainer, a crop-duster, and many other civilian and military uses before serving right through WW2 as an air ambulance, an aerial reconnaissance platform, an artillery spotter, a target tow and a light bomber, to name of few of its roles. Due to its STOL capabilities it was used for supplying Soviet partisans behind the front line and, famously, it was used as a night bomber flown by female crews, cutting their engines to glide near-silently over their targets to harass German positions. Though vulnerable to more modern fighters the Po-2 was highly manoeuvrable and flew at speeds slower than the stall speeds of the opposing fighters, making it a difficult target. Having said that, its slow-flying made it very vulnerable to ground fire. A single bullet in the right place was enough to down the plane, but equally it could be riddled with bullets and still keep flying.

As late as the Korean War Po-2s were flown in combat by the Korean People's Air Force. UN forces called the Po-2's night attacks ‘Bedcheck Charlie’. There were some notoriously successful bombing raids where UN planes were destroyed on the ground by Po-2 attacks. The wood-and-fabric-construction of the Po-2 gave only a minimal radar echo, making it hard for opposing night-fighter pilots to find.

The Albanian Air Force operated Po-2’s until 1985. This little potted history is just a glimpse of the longevity, reliability and functionality of this very successful aircraft.

From the modelling perspective there are practical reasons for choosing this aircraft. The fuselage is square-sided, making it easy to copy in foam board using the familiar box fuselage and turtle-deck formers with paper skin. The cabane system used on my AVRO 539B will transfer to this build with only a few modifications. Regarding detail, the only significant drawback is the exposed radial engine on the front, but that should be reasonably easy to fake-up using black nylon washers on wooden spines.

In terms of flight characteristics, the plane should be very stable. The staggered wings give a very large effective chord, a bit like the Baby Blender. It also has dihedral on both wings and a long fuselage, which all help with stability.

So, where to start.

First I found a good 3-view on the internet…

…which I imported into Adobe InDesign, a Desktop Publishing package. This software uses an identical process to nerdnic’s Photoshop method. The two pieces of software share the same basic drawing interface, which allows layering of the drawings as they are created.

I made my ‘workspace’ the same dimensions as my A1 foam board (594mm x 841mm) and scaled the 3-view up and down to see how the larger components would fit on the foam board.
I hoped the plane could be made from two sheets of A1 foam board and after trying a few wing sizes in the available space I came up with a 1034mm wingspan version (just under 41 inches). I think this is the maximum size I’d want in my car, and I doubt I’d want the plane any bigger. For comparison my AVRO 539B has a wingspan of 840mm (the long edge of the A1 sheet), and my Velie Monocoupe/Morphocoupe a wingspan of just under a metre (two half wings from the shorter edge of the A1 sheet).

After copying all the profiles and sorting all the assembly details and geometry; fuselage, wings, cabane frame, interplane struts and undercarriage, I printed up my components on A3 paper and stitched them all together with clear tape. Note the top wing, shown in the foreground here, is made from asymmetric parts; one side includes the centre flat that will sit on the cabane platform, the other side is just the outer wing.

I transferred the paper detail onto the foam board…
…and then began to cut out all the components. This should be familiar to anyone who has built the core Flite Test models. At this stage you might choose to remove the curve from the rear fuselage sides – I’ll leave an optional line on the drawing. Removing the curve makes fitting the rear section of turtle deck a lot simpler, but does narrow the tail making it slightly weaker.

I needed a way to integrate the cabanes into the foam board fuselage. The method I chose uses two transverse ‘rail’ mounts which are hooked onto skewers embedded in the fuselage sides.

I marked an initial position for the skewer on the insides of the fuselage. In this first position I quickly realised the skewer wasn’t far enough down the fuselage side, for reasons I’ll demonstrate shortly.

This was the final position I chose – the dimensions will be correct in the plans, but as this is the test build there’s always changes happen as the build progresses. I cut a length of skewer, with a reasonable overhang at the back and a little projection at the front to allow it to ‘catch’ on the corner fold where the nose starts to taper. This ‘length’ helps to spread the forces coming from the wing , through the cabane frame and into the fuselage. On the top deck you can see some distinct nicks in the foam at the front, intended to accommodate the zip-ties. These were just a ‘fiddly’ nuisance and this is now just a slot on the drawings.

(Apologies for the fuzzy picture.) With the skewer in position, I drew along its sides and then made angled cuts following these outside lines. I can’t give exact sizes, as all these skewers are slightly different.
This created the first ‘v’ shaped hollow…

…which I then scored with the skewer until it fitted flush in the slot.

Skewer glued in place and ready for the next stage.

The cabane mount is going to catch on the skewer but I don’t want it to stick out of the side of the fuselage, so the first right-angle bend is only about 4-5mm long. This 2.3mm wire produces a tremendously strong and almost rigid cabane frame. I think you could get away with lighter wire, 2mm, but the frame would have a little more ‘spring’ to it. Nothing you’d notice in flight, but the extra flex might cause the thread and CA joins to release over time. No problem if you go for a copper wire and solder join.

The next bend needs to place the wire flush with the top of the fuselage deck. This is a ‘B’ type fold, so I took the wire to the edge of the folding slot. I’ve given a size in the drawing, but you really need to check against your fuselage for a good fit.

For the next bend the mount needs to span the inside of the deck. There’s a little ‘wiggle-room’, but make it too wide and you’ll need to dent the fuselage sides to get it to fit, or make it too loose and you leave less space for your power pod.

After the last bend (to match the first leg) it’s time for a test fit. All looking good, though in this photo I see I need to bend the nearer leg in a little to make sure it’s square. I wouldn’t want it to spring open the sides of the fuselage.

Here’s the wire in place for a final test fit. I used a sharp skewer and a ‘wiggle’ to poke holes in the foam where the ends of the wire hook over the skewer. Take care not to go all the way through the fuselage. It’ll be much neater if the ends of this wire remain hidden. After you’ve made-up a second ‘identical’ cabane mount, you can glue-up the box fuselage. You could glue it up with the mounts in place, but that would be very fiddly. Don’t worry, there’s a ‘neat’ way to do it after the fuselage is formed, so go ahead and glue it up without them.

To get the wire in place you need to stop the ends sticking out and catching the fuselage sides. This is done by pulling the mount’s legs together against the natural spring in the wire. I’ve used zip-ties for this. Now you can see why I moved the skewer further down the fuselage, it allows the legs to be longer. If these legs were too short it would be impossible to get them to bend and still spring back into shape.

Position the mount carefully and snip the zip-ties to allow the legs to spring into position.

Mounts installed. As with all the builds using a removable power pod, now is a good time to fit your CUSTOM pod to get the exact position of the tabs and slots. I only later decided on a custom pod, so read ahead, see what this looks and make it up now! The tabs should locate in the normal positions shown, but if your pod is slightly wider or narrower it’s not so easy to squeeze it into place - better to make it fit now. This first turtle deck former closes the slot for the front cabane mount, so I installed it to keep the mount in place.

Then I installed the nearer of the turtle deck formers and cut lengths of tongue depressor to fit between the formers. Two popsicle sticks side-by-side would do the same jobs. Note the centre former has cut-outs to fit over the tongue depressor. I’ve looped a couple of zip-ties around the cabane mounts to show how everything fits. You need enough space to get a zip-tie through and around the mount so don’t put those wooden elements too close to the edge of the fuselage.

Moving-on, because the fuselage sides form a curve, I ran the curving inner deck all the way to the tail. As I mentioned earlier you can eliminate the curve and make the sides straight, with a single shallow fold just where that last big turtle deck former is, but I like the curve, and it does make the tail a little stronger. There are alternate tab and slot cut-outs in the drawing of the horizontal stabiliser. The solid ones are for the curving fuselage, so you need to choose the holes you need. And if you don’t use those alignment holes (I didn’t), you need to trim the tabs from the ends of the fuselage sides.

At this point, don’t glue that very last little turtle-deck former until you are ready to fit the tail. As this was a test build, I also left extra material around the lower wing cut-outs until I had an actual wing to test in the space. This is probably good practice as different people’s wings do seem to turn out a little different.

Here you can see a ‘classic’ power pod, but later I decided to use a custom pod to help with a battery fitting problem I encountered.

Horizontal and Vertical stabilisers are quite simple, following familiar techniques for hinges and joining. Because the horizontal stabiliser is a big span, I added a skewer to reinforce it. I bedded this in just like the previous skewers in the fuselage sides. There’s the familiar popsicle stick, spanning the two halves of the elevator. (I keep forgetting, but it’s usually better to cut the hinge bevel on the ‘plane’ side, which gives more material on the elevator for fitting control horns etc.) I later modified the component drawing to include a bigger locating slot for the tab on the vertical stabiliser, and added slots for tabs on the curving fuselage sides.

With such a large balanced rudder there’s not much length of hinge, so I decided to reinforce it with tape.

All hinges cut and acting freely, so it’s time for a test assembly of the tail. Looking good, even though this version doesn’t have all those fancy alignment slots and tabs. The front of the Vertical stabiliser locates in the end of the fuselage deck, but it still took a bit of care to get the tail assembly centred when I glued it all up. Of course the first step is to join the vertical and horizontal stabilisers, making sure they are square.

Tail assembly ‘carefully’ glued in place, and I added a couple of little deck formers right at the back.

Actually a later picture, but now is a good time to add the tailskid mount panel into the tail. I suggest you leave the skid off until later to make it easier to handle the fuselage.

The end of this little tailskid panel locates on and centres the lower ‘keel’ element of the rudder. 

Using the ‘lower wing top plate’ to help establish and maintain my fuselage width, I added a cross member just behind the wing. Once it’s set there’s a second plate fits inside the fuselage, right on top of the first plate. The wing elastic skewer sits on top of this second plate making it extra strong.

In front of the wing I’m adding a hatch-come-undercarriage mount. This is a little box folded up to fit between the fuselage sides.
I’m gluing tongue depressors across the base of the box for the wire undercarriage to rest against.

The tongue depressors are cut to length so that they project enough to sit on the fuselage sides.

A test fit of my 1500mah batteries shows I have barely enough clearance, and these are unusually flat batteries. The only way to make more space for larger batteries is to modify the power pod.

This is my solution. As well as being shallower, the hollow faces downward.

There are the locating tabs on the top, ready to fit in the slots in the box fuselage deck.

The pod is held-in by short lengths of skewer with zip-ties that wrap around the cabane wires. OK, it’s not a quick release system, but it is replaceable. The ties can be snipped and re-fed.

Here’s how. First bend back the narrow feed tail of the zip-tie…

…then push the folded end through, so that the tail snaps open against the turtle-deck former…

…and pull back until the tail re-appears on the other side of the wire. Grab the tail with long-nose pliers and pull it through. Easy! (OK, it’s a bit fiddly, but it’s do-able.)

Ready to re-fit the pod. You can also see the locating pins for an ‘original’ pod. These are not required now, but it would just make a mess pulling them out.

Post-maiden service complete - battery strap fitted, and more down and side thrust added to the motor. I’m ready to tighten the zip-ties.

I also added a bit of weight to the nose to help get the CG right. I’m using one of the heavier foam boards and the downside is I often need even more weight to counteract the tail moment. A build in Dollar Tree or similar foam board will be a lot lighter than my plane, and may not need this extra weight. Were I building this again, I’d hold off fitting the nose and cockpit turtle decks until the plane was fully fitted out and almost ready to fly. Then I could fit any balance weights without removing the pod.

Back to the actual airplane build… now is as good a time as any to make-up the undercarriage. I used a 2mm wire for this.

All the bends are shown on the plans. ‘bend A’…

‘bend B’…

…and ‘bend C’. You can make the frame from a single long wire, or break it up into smaller sections. I chose a single long wire.

Here’s a test fit with the metal frame held together with zip-ties. Next stage is to join the overlapping portion with thread and CA glue.

This undercarriage system uses rubber bands or prop rings to hold the axle in place and provide a little suspension. Not essential but I use a length of control rod snake to make a sleeve to cover most of the axle. The wheel can’t go over the sleeve, so this helps to keep the wheel from rubbing against the suspension rubber. A washer would work as well. Note that the last 15mm of that bottom corner of the frame has been bent to make it near-enough vertical, again to stop the ‘pointy’ end of the frame digging into the side of the wheel. A lesson I learned from ‘chewing-up’ the insides of my AVRO wheels.

Complete frame, axle and wheel assembly.

Once the skewers are fitted, elastics hold the undercarriage in place. This system worked well on my AVRO 539B. It may not be the neatest solution, but it’s durable and forgiving if the plane has a rough landing. Previously when I’ve used a ‘hard’ fixing, I’ve ended up damaging the fuselage. A word of caution, this is not a battery tray. The skewer mount at the front is not strong enough to carry the weight of a heavy battery; the battery needs to be strapped to the pod.

This picture shows how removing the front skewer allows the undercarriage mount to swing clear to fit the battery.

Let’s get that cabane assembly made and mounted.

There’s bending guides for the cabane frames in the plans. The two front frames are made up as shown in this picture, with no overlap. The third frame is identical in shape, but the wires at the top have a good overlap so they can be joined.

There’s a ‘neat’ space-saving way to join the two front frames by placing the wires together with the gaps on either side. It’s shown here, and on the assembly sketch in the plans.

Bend up your three cabane frame wires and the two top wing support rails, and then fit the frames to the fuselage.

Here’s the dry assembly. As well as holding the frame to the fuselage, zip ties are great for holding it all together. Once the frames are all sitting symmetrically side-to-side, the 8 zip-ties connecting the frame to the fuselage can be tightened reasonable firmly. See how the bottom of the frames rest against the tongue depressors. This helps spread any downward pressure into the fuselage, and stops the foam deck being crushed or deformed. The upward movement is resisted by the hidden cabane mounts, which connect into the fuselage sides.

Once you’re completely happy with the shape and symmetry of the frame you can start to join it all together. Get those bottom zip ties nice and tight and then snip off the ends to make the next stage easier. Thread and CA is a laborious method, but it does get the job done. A little patience is all that’s required. The thread does not need to be ‘super tight’ - once you add the CA it fills up all the gaps and bonds the threads together making them extra strong. Thin CA is best for this as it soaks into the threads, and if you use a kicker, wait a few seconds to allow the CA to be absorbed by the thread before setting the glue. Join the front pair of struts first. You can see how that ‘neat’ staggered join works, minimising the bulk of the join. As you add the thread joins, you can snip-off any zip-ties the join replaces. After these first four groupings of thread and CA I later added another four, two in the middle and then one more on each of the short legs.

Check the back frame against the front pair for a reasonable fit before binding it up. Here you can see the simple overlap of the top wires held in place by a single zip-tie.

This is not rocket science, and not precision engineering. As long as it’s reasonably neat, it will look fine on the finished model.

I could set-up the angle of the top mounting rail using a full-size copy of the cabane geometry drawing and positioning it flush with the top of the box fuselage, but I think seeing the wings in place will help, so let’s move away from the fuselage and get those wings built.

The top wing is made from ‘handed’ parts. One side includes the ‘flat’ wing centre. There are two dihedral ‘turns’ on the wing, which sit either side of the centre. I aligned and taped the two wing sides together before making my leading edge bevel cuts and the shallow cuts and scores along the wing camber fold lines.

My folded spars never come out as reliably as the laser-cut kits do, so I don’t cut the wing locating holes for the tabs until I have the spar assembled. Equally, I tend to cut one side of the spar profile, then fold the spar halves over and use my first cuts as a guide to cutting the second side. That way there’s no problem of a mismatch if it doesn’t fold just right.

Not shown in the build sequence, but I actually have a 2.3mm piano wire reinforcing element spanning the centre 300mm of spar. This is shaped to the correct dihedral and sandwiched between the spar sides. I cut a slot in the centre of the spar on one side, just the same way as I embedded the skewer for the cabane mounts, and then glue the wire into the slot before folding and gluing the spar halves together. By all accounts, the wing doesn’t need it, but my planes tend to be heavier than the dollar tree ones, so I like to be sure. The wire also helps maintain the dihedral angle and stops the spar from deforming during the build process. To stop the ‘cold’ metal setting the hot glue too quickly, I sit the wire in my kitchen oven for a few minutes at its lowest setting (40 degrees C). Not so hot I can’t pick it up, but warm enough that the glue stays soft while the wire is bedding-in to its slot.

The joining tape is still in place, keeping the wing ‘halves’ in alignment, so fold the entire wing, pressing it into shape and establishing a constant profile along its full span.

Cut through the tape holding the top surface of the wing panels together, just as far as the leading edge. And then on the other side of the centre panel cut through the foam on the top of the wing, again, only as far as the leading edge. This frees the centre section, ready to be folded into shape.

Score and introduce a fold at the edge of the centre panel, ready to make the dihedral.

Holes cut for the strut mounts – just big enough to slide the arm of a ‘9gram servo’ through.

A careful dry fit of the spar gives me the final positions of the tab cut-outs on the bottom of the wing.

Same again for the other side.

After cutting out the tab holes for the spar, it can sit loosely located in both sides. This puts the centre section of the spar in the correct position, and it can now be glued in place. After the spar is set, slightly flatten the trailing edge of the bottom panel to increase the glue contact area before folding the centre section over and gluing it down. I only glued the leading edge fold, the top of the spar, and the trailing edge join. I didn’t bother with gluing all the curve folds, and I’ve hung the back edge off the table as I pressed it into place to allow for any slight under-camber.

With the centre section set, it’s time to make-up the right wing panel. Before working on the right wing section I lifted the left edge of the centre section with an off-cut of foam to allow the dihedral to form. Note the left wing section is still hanging on its taped join. Be aware this is vulnerable, you don’t want it tearing off and messing-up all your careful alignment work. Once you’re happy that the positioning of your off-cut is creating the correct dihedral angle (spar sitting flush along the top of the wing), lift the wing and run glue under the spar. Now set the wing down. Make sure the off-cut is back in the same place to form the dihedral and press down on the spar as it sets.

Now you want to fit the wing top panel. As it comes over you need to trim away material to get a flush fit. Do this bit-by-bit, gently laying the wing panel over the centre section and marking material to be removed.

Keep marking and trimming for a good fit. This is not a constant-width cut. It is widest at the fattest part of the wing and tapers away to nothing at the leading and trailing edges.

Before closing the wing you need to install the interplane strut mounts. Angle the front mount backwards and the back mount goes straight up and down.

Here they are, glued down from the inside. Do a test-closure of the wing and see how much the back mount interferes with the top panel. You will most-likely need to dent or even trim the top panel to get a good flush fit. Remember to put that ‘dihedral’ packing-piece back under the centre section before closing the wing over. Also you need to hang the projecting interplane mounts off the edge of your work surface to allow the bottom panel to sit flush with the table. It’s a bit fiddly, but not impossible.

With the right side done – it’s time to move to the left. This should still be aligned and ready to go with the tape still in place. As you prepare to glue the spar down, remember to run glue into the open join between the wing and the centre section. Using the same off-cut spacer under the centre section as you did before, hold everything firm to help the spar set at the correct dihedral angle. Once the spar is set, follow the same mark, trim and fit procedure to get the top panel to fold over. Fit your interplane strut mounts and then glue down the top panel. Carefully push glue into the two top joins and ‘smear’ finish them; finally add tape over the joins to finish the wing assembly.

I cut and shaped short sections of tongue depressor to reinforce the back edge of the wing for the elastics.

I ‘cheated’ with the bottom wing and didn’t bother with a centre flat section. The wing assembly is much the same as before. Join the two symmetrical wings halves with tape before cutting the leading edge bevels and the shallow cuts and scoring of the folds for the curve. Now dry form the whole wing to get an even shape across its span. Now cut the tape along the top of the wing so you can make the wing up one side at a time. Start left or right, it doesn’t matter. Here I started on the left. I left the ailerons intact – they’re already marked on and I’ll cut them out later. First I made sure the dihedral turn on my spar was sitting dead centre and then cut the tab locating holes on both sides of the wing. Then I glued the spar in place on one side. Fit your interplane strut mounts in the top of the wing, slightly angling both towards the centre of the wing. Install the servo at the end of the spar and position the servo wire alongside the spar, adding an extension or Y-harness as needed. Now you can glue down the first top panel. There’s no problem with the strut mounts hanging out from under the wing, but still let the trailing edge of the wing hang over the edge of your table to let any under-camber form as the glue sets. With the first wing half assembled, notch a little hole for the servo wire extension/Y-harness and lead it clear.

Use a spacer to raise the completed wing half and establish the dihedral with the loose end of the spar sitting snugly in place on the bottom wing plate. Once you have a set-up that is stable and easily found again, run glue into the centre join and under the spar. Carefully re-position your dihedral spacer, and press down on the spar to let the dihedral set correctly.

Install your strut mounts and servo and then repeat the bit-by-bit procedure to get the second wing top fitting flush against the first.

Once you have a good fit, feed/join your servo wires and glue down the second wing half. 

Now fit the lower wing top plate, making sure it is square and central when it sets.

I glued-on a tongue depressor to reinforce where the elastics go over the back of the wing.

Now I have the two wings I can get an idea of how it’s coming together. I’ve still to adjust the position of the rear cabane frame to set the angle of incidence of the top wing, and I need to make a few little adjustments to the cut-out in the fuselage for the bottom wing, but it’s looking good so far.

I didn’t get a picture at the time, but here’s my full-size copy of the cabane geometry drawing positioned flush with the top of the box fuselage. You can rotate and position the rear strut to get the correct angle of incidence before binding the mounting rails in place. I had a little rethink about my initial CG and the angle of incidence, so ignore the figures in this photo, they’ll be right on the amended drawing.

All bound and fixed, but even at this stage there’s plenty of ‘wiggle room’ regarding positioning of the wing. It can be moved side to side, forward and back and even rotated slightly if your cabane assembly is a little off.

Once you are happy with the wing alignment, add some centring pieces to the top wing to help position it on the cabane assembly. I’ve shown the wing out of its mount, but you would add these corners while the wing was in place. Now every time it goes back on, it will be pretty close to where it should be. These will deform with hard use, but they can always be replaced or repaired.

The interplane struts are fiddly and take a little patience. After aligning the top and bottom wings against the fuselage, I made-up the centre strut on each side and worked from there, adding the rear pairs and then the front pairs. These are really only there to help maintain the wings’ alignment in flight, so a little flex here and there won’t matter. Don’t make them tight, as this will just change the dihedral on both wings, making the bottom dihedral sharper, and the top dihedral flatter.

The struts attach using modified Z-bends and right angle bends with swing-in keepers. It would get a bit crowded around the mounts it you used swing-in keepers at both ends. On the right you can see two modified Z-bends ‘happily’ sharing the mount.

There’s a good sequence to attaching the interplane strut wires, which only has a swing-in keeper sharing the mount with a modified z-bend the one time. This is shown on the right here. Make sure the modified z-bend is well above the mounting point of the swing-in keeper.

Based on the CG I calculated, and ‘loose’ component testing, I found I needed the weight kept forward, so I fitted my rudder and elevator servos in the rear cockpit space with snakes running to the rear. The tubes from cotton-buds are a good fit for this size of wire and you could use them at the tail and straws for the rest of the run. This space is not accessible once the turtle-deck is on, so you need to think it through and have the wires and/or snakes in place before you close the deck. The important thing is to eliminate any chance of the wire bending when the servo is ‘pushing’.

I ‘filed’ a hole into the foam with the side of a drill bit to bring the snakes out just under the horizontal stabiliser. A small, round rat’s tail file would also be good for this job. Because of the need to pass through glue, just poking a skewer through will be quite hard. Glue the end of the snake in place to keep that little corner of the tail-to-fuselage join strong.
Here’s the rudder connected-up.

…and the elevator. Control horn glued and screwed in place! Belts and braces…

Not much more to do. Here’s the aileron cut free, the hinge bevelled and reinforced with hot glue, and the control horn connected up to the servo. Again I forgot to cut the bevel on the other side to make more space for the control horn.

The turtle-decks are poster paper, taped and glued using the familiar Flite Test system. Start at the tail and work forward. I was a bit quick and missed the photographs of this stage, but here’s the template in position, showing how the back section centres around the vertical stabiliser.

Because this ‘deck cover’ passes around a curve, I snipped little nicks in the edge of the paper to get a better fit. Normally I glue these overlaps down, which might have made the joins neater, but I wanted to use the ‘Flite Test’ clear tape system that folks are familiar with. Then I saw Josh gluing down the turtle-decks on the FT minis using hot glue, so I should have just stuck to my original method. Josh added a neat twist though, instead of starting at the side, he started at the top of the former which seemed to make it a bit simpler. Have a look at he mini's build video and see what you think will work best!

Next there’s a short section behind the second cockpit, this sits on a square-sided section of the fuselage, so is simple to fit. If you’re going to mount a machine gun on there, stick another little bridging support between the curved formers to stop the top of the paper buckling.

A later picture, so the cockpit section is already in place, but the next step is to fit the nose. Press the nose piece into place to find where the large former finishes and mark this line on the paper. Snip out little triangles up to that line as shown (but all the way round). This allows the paper to fold over the former and lie flat. You might want to put this off until the very end of the build in case you want to add nose weight on top of the pod to get the CG sorted.

With the nose section in place, start fitting the cockpit and cabane section. This needs cut-outs to go around the cabane wires.

Just keep removing material until the turtle deck sits in place, in full contact with the formers when pulled tight. Glue and tape into place.

Here are all the turtle decks fitted. I’ve included a template for the windscreens if you want to add those. A little ‘padded’ cockpit edge made from split and stripped round electrical wire would be another simple feature to add a bit more character.

The wheel spokes were filled-in using my ‘olde-style wheels for olde-style planes’ method, and I added the teardrop fairings on the rear undercarriage wires. These wrapped around the springs/shock absorbers on the real plane. A ‘wing type’ nose bevel in the centre of the 40 mm face of a 30 x 40 mm blank will make a rounded leading edge, and a deep bevel on each outer edge will wrap around the wire to make the elongated teardrop shape. I stuck these on with foam safe CA, and to finish I wrapped and pinched clear tape around the trailing edge to sharpen it up.

The only other addition to the plane is the dummy motor. Each engine cylinder is made-up from a stack of two different sizes of black nylon washers glued to a short piece of dowel. The dowels are first glued into a thick cardboard ring.

I printed the Russian stars onto self-adhesive parcel labels and cut them out and stuck them on. By a lucky coincedence, this 'everything white' finish is the winter colours for this and many other Soviet planes during WW2. I’ve seen a version of this plane with skis fitted to the existing undercarriage frame, and that might be a fun mod for winter flying. There's also a float plane version, and if someone had the patience to waterproof everything, this would make a great seaplane. On mine, everything has been sprayed with mat varnish, which will keep off most of the dew-damp and fog I encounter.

The dummy motor still needs a final coat of paint, and I’ve still to work out how to attach it (magnets?) but it definitely adds to the look of the plane.

Built ‘carefully’ in lighter foam board, I can see this coming out at around 800gramswith a 2000mah LiPo. Even at just over 1Kg my plane still manages to fly very sedately, so at lower weights it will perform even better, and fly for longer. There are no slow-motion effects in the maiden flight video, and the winds were about 5-6mph, so that should give you a good idea of the plane’s potential.

For the maiden I started with a few trim ‘clicks’ of right rudder, and a few ‘clicks’ of right roll on the ailerons. My elevator was flat. I also set-up a reasonably gentle aileron-rudder mix. I started in low rates, everything on 50%. Initial full-range motion on the ailerons and elevator was 40mm (20 up, 20 down). The rudder has a full range of 50mm. First launch was with too much power and the plane climbed aggressively and had looped over my head before I cut the power and got it safely down. I put in a whole load of down elevator, moved my battery forward a bit and reduced power to try again – that’s the first flight you see in the maiden. As I got the plane under control it was clear the rudder wasn’t doing much, the ailerons were OK but a bit ‘anaemic’, and the elevator was too aggressive. But it was flying so I kept it in the air to trim it out and get the feel for it. It still wanted to climb, so I kept feeding-in down elevator trim until I ran out of range. The rest of the flight was Ok, but I had to manage the elevator constantly with a little down elevator to maintain level flight. Under power, the plane wanted to climb even more, so I reckoned I needed a bit more down thrust on the motor. The other possibility is too severe an angle of incidence, but I’ll fiddle with CG and down-thrust first before altering the wing angles. I landed when the battery alarm sounded and examined the transmitter. I seemed to have removed all the right trim on the ailerons, but the rudder still had a few clicks of right on it. I went into the sub-trim menu and adjusted my elevator into an initial down position, which allowed me to get my trim range centred. I upped the range of throw of my rudder and ailerons in the dual rates to 65% for the ailerons and 80% for the rudder. I dropped the low rate on the elevator to 40% (I also set the elevator high rate to 75%). I moved the battery forward a little more and set the plane off again. That’s the next launch in the video (1:17) - the plane just floats out. There was a little elevator trimming, but once that was done the plane was pretty steady. The rudder still wouldn’t do much until I used a touch of aileron to bring the wing over (especially on right turns, suggesting the need for a little more right thrust on the motor) but it did not detract from the experience, which quickly became very enjoyable. I tried some loops, which were easily accomplished and uncomplicated with no bad tendancies as it came out of the loop (sadly the spec-cam decided to freeze before then). But I didn’t feel confident with rolls, next time.

The CG on this flight was 84mm back from the leading edge of the top wing.

Wingspan 1034mm / 41 inches
1270KV 200W outrunner with a 9x3.8 SF prop (also fine with a 9x4.7 SF prop)
AUW with a 1500mah LiPo - 1070grams / 37ozs.
Time in the air from launch to battery alarm for my 1500mah LiPo was 12.5 minutes of mixed (but mostly sedate) flying. Which I felt was pretty good. This improved by almost a minute with the after-maiden modifications.

Any omissions or glaring errors in my history, plans or build sequence please let me know. Any questions, please ask. Please only attempt this build if you are confident with building other more complex foam board models like the spitfire or cruiser.

Cheers, alibopo.

Here's the updated PDF plans - including sketch detail on some of the build elements; Polikarpov Po-2/U-2 

UPDATE (27/11/2014) - these plans were originally drawn for a 6mm thick foam board. I've edited the plan and adjusted for a 5mm foam board, which is more common. I'm told DTFB is closer to 4mm. Anyway, it's not rocket science,and foam board is quite forgiving, but check your foam board against the marked slots etc. before making the fitting cutouts. Gross outlines are the same but you'll need to make allowances as you cut slots for folding-up the box fuselage and holes for any connecting parts, especially the horizontal to vertical stabiliser connection.

Regarding tiled printing Adobe Reader XI (free) will print a 'POSTER' version, which is the same as tiling.

And here's a link to the maiden -


marc60 on November 7, 2014
Chapeau, very nice build. I like the details like the motor cylinders
Log In to reply
alibopo on November 9, 2014
Hi marc60, thanks for the comment. Regarding the motor cylinders, it's funny how these ideas seem to pop out of nowhere. I was looking at little nylon tubes on Ebay to make stand-offs for the motor (before I went for a custom pod), and spotted those washers - two different sizes, like the fins of a cylinder head, it seemed very obvious when it hit me. Though not in any way scale, these small details do make a difference. Cheers.
Log In to reply
Bayboos on November 7, 2014
That's a very impressive build log, I really appreciate it. I have just one question. It seems from the pictures that you are not using DT fom board. Rather than that, it looks like the foam you are using is much thicker and the paper is of higher quality. Is that correct? And if so, can you tell us what exactly it is? Do you know how much stronger and heavier it is compared to DTFB? Any answer will be highly appreciated.

Keep up the good work, and thanks a lot for all the details.
Log In to reply
alibopo on November 7, 2014
Hi Bayboos, absolutely right, this is significantly heavier foam board, but not really any thicker - it's around the 6mm mark. But yes, the paper is thicker, which lends a bit more strength to the builds as well as weight. For the actual figures, my foam board weighs-in at 0.39 g/sq inch compared to the Dollar Tree stuff at 0.19 g/sq inch - I don't think it's got an actual trade name, it's just sold as 'foam board' and it's what I can get in packs of 10 in the UK through an Ebay supplier. Buying foam board in the UK can be quite expensive, the only way to reduce the price is to buy in bulk packs. The actual weight and comparison of weights has been discussed before in this excellent article by apnewton from Australia (also look at the threads that follow);
...and a response I did in my article when I had an AHA! moment and realised I was building with much heavier foam board than was available in the US is here;
My builds do seem to have longevity on their side, and I think they are better suited to the windy weather we get in Scotland, but I do get envious at times by the low build weights some people quote. :) Cheers, alibopo.

Log In to reply
Bayboos on November 7, 2014
Just like my Storch (speedbuild kit) that weights 650g (complete, without battery) at 1.5m wingspan :) Please remember that DTFB is approx. 4mm thick, with the difference being clearly visible on the pictures.

I understand you perfectly, living in Poland does not make things easier for me either. I actually tried several types of covered foam, but I found all of them way too heavy. RC planes below 1kg don't usually need extreme strength, while they can benefit from being light.

I myself ended up using Depron in most cases - probably the only type of foam that is cheaper here than in US - for it's high quality, sufficient strength and weight benefit. I find it having just one disadvantage: compared to any type of covered foam, it's way more flexible. I counteract that by using balsa spars and/or covering it with cheap (and thus light) paper myself. Especially the last technique is very interesting. Since I usually need to transfer the plans (printed on paper) to the foam, I can just glue the paper on and leave it there. When using 3mm Depron, I can usually get enough strenght at the total weight close to (or even under) the weight specified in FT plans.

Now I just wish I could have more time to build more planes...
Log In to reply
alibopo on November 8, 2014
I actually did not realise DTFB was only 4mm. When I've printed up and used Flite Test plans they seem OK for the build, but I was probably cutting with an eye for my foam. Interesting! Most of these plans have quite a thick line on them, so I suppose just cutting to inner, centre or outer part of the line will give a reasonable variation. I've seen a very good demonstration on YouTube of 'uncovered' foam getting covered in 'masking' paper used by vehicle re-sprayers. Water based PVA (white glue/wood glue) is used to soak the paper. I suppose we just have to work with what we have, but I don't know if I'd have the patience for that. I think building in foam board is a great system, even with the heavier types. When I built my Velie Monocoupe I based it on a set of plans for a traditional balsa and 'gas/glow/etc' plane. It came out at roughly the same weight as the balsa version would have so I never thought anything was 'wrong'. You've spotted an interesting issue, I think I'll have to add a cautionary note regarding the thickness of my foam to the plans and article.
Log In to reply
alibopo on November 27, 2014
Hi, I took you comments on-board, and I've adjusted and updated the drawings for a 5mm thick foam board. Hopefully that's a 'middle ground' for most folks. Cheers.(27/11/2014)
Log In to reply
Yogenh on November 7, 2014
I would like to have the plans for this one
Log In to reply
alibopo on November 8, 2014
Hi Yogenh, plans are available from the link near the foot of the article. Cheers.
Log In to reply
alibopo on November 27, 2014
Hi Yogenh, just to let you know I've updated the plans, adjusting everything for a 5mm thick foam board.
Log In to reply
Yogenh on November 27, 2014
Thanks a lot will have to give them a try
Log In to reply
Berkut47 on November 7, 2014
is it possible that you could make tiled plans for all of us that don't have the skill to do it ourselves
Log In to reply
alibopo on November 8, 2014
Hi Berkut47, if you use Adobe Reader XI (free download) you can choose the option 'POSTER' and this will print in tiles. Cheers.
Log In to reply
Berkut47 on November 8, 2014
thanks comrade
Log In to reply
808aerosquadron on November 9, 2014
alibopo, great build article and yet another interesting and beautiful airplane.

Keep up the good work and "Keep'em flying."
Log In to reply
alibopo on November 9, 2014
Thanks for the encouraging comment. This is a great community. Cheers.
Log In to reply
hellcat on February 8, 2015
very nice build and plenty of support images to assist and make life easier
Log In to reply
Nose Dive on September 14, 2015
Hi Alibopo, I applaud your unusual and interesting choice of aircraft. The second I saw the first pictures I knew I had seen it before. "Die Nähmaschine" (the sewing machine) was also one of the nicknames the German Wehrmacht bestowed on this plane. During night raids, all the Germans would hear was this unnerving rattling sound. This build, again, looks fantastic. I already downloaded the plans of one of your earlier builds, the SE5, and started building it. Now I can't wait to start on this one. Great job!
Log In to reply
alibopo on September 18, 2015
Hi 'nose dive', thanks for the comment. The PO-2 is so elegant in the air, it's a joy to fly. The SE5 is actually the later build, and I've been having a lot of fun flying it. With the long nose, it's much easier to balance than the PO-2, which needed a bit of nose weight added. These tend to be lighter wind planes, because they get bounced about a lot by gusts, but I recently added flight stabilisation to the SE5 and with the flight smoothing that gives I've been taking it out in slightly stronger winds. Both are great planes for super-slow stall-turns and loops. Rolls are a little more challenging as they take a bit of time to complete the manoeuvre. (If a second set of ailerons were added, I'm sure that would speed things up.) If you watch the video of the SE5, you'll see I messed-up the roll a couple of times by feeding-in too much reverse elevator while inverted, but I've got the hang of it now. These planes will never be 'snappy' fliers, but that's their real charm for me. They both fly inverted, but not efficiently. I'm thinking of modifying the wing design in future builds, maybe reduce the depth of the wing and flatten it out a bit. 'Flying Penguin' is developing some nice profiles. The real 'star' for me is the little AVRO 539b. That was my first biplane, and it really can be thrown about the sky. I've maybe over-engineered the cabane assembly (or not -- all my biplanes are still flying!), and I might re-work that to make it easier to build. I'm thinking I can probably get away with just gluing a gift card beneath the box-fuselage deck and running my cable ties through that and around a skewer, instead of all the metalwork. I'm sure it would be plenty strong enough. I'd love to see your build flying once you finish.
Log In to reply
Nose Dive on October 6, 2015
Hi there,

I noticed that I enjoy building these planes so much that I
barely go out to actually fly them. My first build was the FT Storch.
It turned out pretty nice. This plane however, turned out to be way 'over my head' since I had zero experience in flying these things. So, before I'll start building your bi-wings I went back to the scratch table and built the FT Tiny Trainer. I finished it but I still have to go out and fly it.

Actually I was almost done tracing and rescaling the drawings (in A.Illustrator) for the SE5 when I found your posts . With pictures and everything!!! No need to finish that one then. ;-)

For some time I have been thinking about buildling the De Havilland Mosquito.
I am addicted to that plane. I have build several plastic model kits of this plane and
would love to build a RC fomie of it. I believe that the FT Cruiser is based on the De Havilland "DH88 Comet racer" and does have some similarities with the Mosquito

I as well have been paying attention to flying Penguin.
I find his approach of curved fuselages and surfaces super cool and very inspiring
and I very much would like to use them in a Mosquito.

Before I go crazy browsing for usable drawings of 'balsa-build-Mosquito's' I have a question:

When you build a plane, how do you determine the size of the plane versus the size motor, prop and electronics you want to use. Is there a 'formula' or guideline to figure that out?

greetz, NoseDive
Log In to reply
alibopo on October 7, 2015
Hi - no special magic. Because of the type of plane I'm building I look for a slower KV motor around the 1200 - 1100 - 1000 - 900 range. That usually throws a larger prop; 9 - 10 inches. Faster spinning motors use teeny-weeny props, that would look silly on a biplane.Then I look at the weight of the plane. Building in my heavyweight UK foam I can get a monoplane out around the 600 gram mark or a biplane around the 1000 gram (including battery) - I build heavy planes ... anyway, most motors have a thrust value in grams, relating to the prop and a 2S or 3S LiPo. It's just basic power to weight. Some of my planes come close to 1:1 (thrust v flying weight), but most of the time I'm flying around half throttle. Real flying, none of that 3D (helicopter) stuff.
Basically my 600gram planes will fly on an 8AMP Blue Wonder motor (400 grams thrust). They fly very nicely but the performance isn't spectacular. Bump it up to 12 amp and it's a bit more sporty. I'm flying the heavier 1000gram planes on closer to 20 - 25 amp motors, but I'm underpropping and under-loading them. Running flat-out is inefficient and eats batteries. I like time in the air just cruising around. Hope that helps. Alibopo
Log In to reply

You need to log-in to comment on articles.

Polikarpov U-2/Po-2 build article.