Having caught a cold and now recovering, I do not feel like going to the workshop to build anything. Better to stay warm and comfortable.
So, to pass the time in a creative manner I thought I 'd repost the build thread of my Avro 504 N/O. The detailed and rather long thread is of course lost togethere with "HPA 2.0"
The model is 1/20 scale, giving it a wingspan of 549mm and a wing area of 7,3 dm2. Target weight was in the 75 gram range (more on that later...)
It was meant as a replacement for competition use of my de Havilland 9 from 2010. The model was started in 2020, after sufficient documentation was collected in order to arrive to a well detailed model. It was completed just in time for the 2023 IIFI meeting in Nijmegen.
The Hellenic Naval Service received 6 aircraft of the type in 1925, in an attempt to reorganize after many years of military adventures, the first World War and the Greek-Turkish war of 1921-23 that followed.
From the build start one goal was to make the model convertible as well, in land and hydro form. This added some complexity but was indeed fun to achieve.
Also, the model would be flown free flight for the corresponding competition use but also radio controlled, when flying in smaller halls for trimming or fun. Too risky to fly it otherwise.
The method of switch to and from ff/rc has been described in a short article in May 2025 Aeromodeller magazine.
I started with the tail, which was rather easy.
Wanting to achieve thin scale-like outlines, outline accuracy and stability, I chose the full balsa core method, as is mostly used in larger models.
Thin 0,5mm balsa sheet, cut to size and ribs glued on either side to simulate the scale structure.
the outline itself is a wider curved piece of balsa, to simulate the tubing of the original.
The complete rudder is then sanded to airfoil ahape, using thin plastic as guides to limit the amount of material to be removed.
The same process was used for the horizontal tail, the stabiliser and the two elevators.
The sticks glued on one side remain oversize until the second side is done, to help with alignment
with the second side done, everything is trimmed to the core outline and the actual outline is added
and the stabiliser
tail finished and sanded
next came the fuselage.
I wanted a scale like structure so prefered to use basswood over balsa for the longerons, as this allows thinner cross sections to be used without sacrifising strength. There may be a small weight penalty involved but overall it is much better looking, especially in the open cockpit areas.
Some uprights in high stress areas were also made of basswood, the rear ones of balsa.
As usual, one side over the other to ensure complete repetition
there is some internal diagonal bracing that will be visible from the cockpit holes, so this was to be added.
Bt first, to avoid any deformation, a temporary balsa sheet piece was glued to each side for strength.
Thin thread, fuzz-free, was used fir the rigging. IT was CAed at one end and then, using weights to get equal tension, it was secured on the other ends.
the extending edges were cut and the balsa temporary parts removed with thinner, so the fuselage sides were ready to glue into a box.
Nothing unusual here, a slow process to ensure everything is true and square.
then came the first difficult task, to cut the formers for the stringers on the top and sides.
First attempts proved a disaster as the very thin and narrow balsa formers broke very easily, so I changed my approach and extended the parts beyond their outline, so that I could cut them to size after they were glued to the fuselage.
Before starting to glue the stringers a thick balsa block was glued to the bottom side of the fuselage to keep everything square. It would be removed only after the sides were done as well.
Black arrows indicate where glue was applied, so that the thinner would be applied for taking the block out.
The stringers were again made of basswood, sanded thin enough, aprox 0,8x0,5mm using a very simple sanding jig. It took some time though to get them all done.
all side formers added
About a year later, late 2021, work was started on the wings.
Quite a lot of ribs to cut. Being a rather process orientated modeller and an obsessed perfectionist, I cut each one separately, using two templates as guides. I find it much better than making a block of many and sanding it, becasue it is very easy to overdo it and damage the templates if one is not careful enough. Also the ribs might not be 100% exactly the same. Not that anyone would notice...
The fact that I prefer the spars to be inside the ribs and thus require accurate slots helps in that decision.
The rib front and rear ends were made oversize, compared to the corresponding leading and trailing edges. These would be pre-shaped and it is much easier to sand the ribs to their shape than the l.e. and t.e. to the ribs.
The less anomalies thre are on the structure the better the covering goes and shows.
a simple sanding jig for the trailing edge, using metal rods for the needed final cross section. 5cm long is enough, you sand and push the jig to a new position, sand again etc.
The aileron trailing edges were curved, so at first they were sanded to triangle shape, then stripped lengthwise, weted in water and glued to a form as laminations. Some final fine sanding was also needed afterwards.
shortly going back to the ribs, the slots were cut with self made tools, using carbon steel pieces glues to sticks. The carbon steel breaks nicely and straight, it is very sharp to cut balsa vertically.
seing here enlarging the slot for the aileron ribs, for the extra aileron leading edge.
Some fine sanding with small diamond needle files ensure a very clean and accurate outline.
The spars are soft balsa sandwiches, with a thin carbon piece in the middle. This provides great strength against deformation to the finished wing.
They were made slightly oversize and sanded after glueing to the exact cross section. Working with CA does not give a lot of time for proper alignment.
repetition, the high tech sanding jigs for the leading and trailing edges.
The leading edge was made of basswood. I think the extra strength came handy whenever the model hit an obstacle with the wing
Wow! Your attention to detail is amazing.
Thanks Philip...
wing construction. The oversize ends of the ribs are visible.
The ribs slot into the triling edge. The slots were marked and opened with a simple sanding tool of exactly the same width and fixed depth.
left wing panel, slowly coming together
the ribs are slotted into the trailing edge and to keep them in place extra sloted balsa pieces were made, to keep them vertical as well. Once in position they were glued to the leading and trailing edges.
The extra pieces allow temporary fixing of the ribs while the spars are being pushed through them. The ribs are thin, about 0,6mm so they might break if the spars do not go clear through their holes.
There are 4 wings panels, so it makes sense to auotomate the process as much as possible to avoid through-backs.
Tpo wing center section done.
Aluminium tubing glued on center section and wing panels, piano wire connects them and sets the initial dihedral angle.
...
second left wing panel done, moving on to the right ones.
I was a bit lazy at that point and did not make the laminated outlines one time. But I continued with the next panels anyway.
same old same old...
the aileron trailing edge had to be glued upwards towards the tip, because of the curved outline. A bit of trial and error until it looks about right.
First test assembly of all the main parts. It finally looks like an aeroplane.
And as one can see, the fuselage is still not finished, some things were being done in parallel.
wing struts were also cut and sanded to shape. 2mm thick basswood. The little lugs are for alignment to the templates, to be sanded away once the side outline has gotten its shape.
The outer wing strut pairs are different than the inner pair, in that they are not as wide, so there were 4 different templates to make for front/rear and inner/outer.
Once cut and properly sanded to shape they were sanded to airfoil section.
The wing ceter struts were also positioned to the fuselage for a first test. And to the wing center section.
additional aluminium tubing pieces glued to the wing center section for the center struts to attach too.
We are close to 1000 posts in the forum, I am not going to be the one to break that milestone, so I will pause for a while...
Jorgos, your Avro model is simply fantastic. I would love to see it in person sometime.
Best regards, Peter
Thank you Peter,
would be great to see you again too... And the rest of the Swiss friends.
And now we go to a somewhat unfittingpart, the rc system installation...
As I wrote earlier, the model doubles as a free flight and radio control, depending on the occasion and the flying site size.
This will be a 4 channel rc model, using a Parkzone brick with two linear servos for the tail and an extra servo for the ailerons.
The Rx is fixed on a plywood base that can slide in/out of the model, through the nose, for servicing purposes. The linear servo movement needs to be transformed to a rotational pull-pull system, through the shown bellcrank assembly
The bellcrank in pieces.
Temporary brass rod connectors from the servos to the bellcranks, to establish their correct length
The fuselage, finished some time ago, all stringers added and the temporary block removed, looks much better now.
The bellcrank in position.
The balsa sliding guides for the Rx tray inside the fuselage are also visible.
Change of subject, enough with rc...
The landing gear of the Avro N is somewhat complicated, as will be explained later.
To make matters worse, the wish to be able to interchange land and hydro system needed some thought. The whole system would have to be screwd in and out from the bottom, so a ply base was made, actually three of them. On for the landing gear, one for the float assembly and one as a temporary undetailed landing gear for the first tests.
The metal lugs are the attachment bases for the "V" landing gear struts.
The landing gear struts would be made of aluminium tubing with balsa fairings. The tubing ends would be formed to go over the lugs and be secured with 0,5mm pins.
...
The bottom connection of the "V" also took some tries to get it right. An inital thoght of a connecting plate did not come out that well, plus it did not solve the problem of attachnig the horizontal bar between the two "V"s.
An alternative was found, involving tubing of various diameters and bent wire connectors, which looked good and seemed strong enough and stable as well.
and the necessary delays for assembly pictures.
Wings are also finished here, basswood wing tips added at last.
The main landing gear base shown before was for the "V", which is the fixed part of the landing gear. The wheel axle is fixed on the forward struts, with rubber suspension. On the original they would hang lower because of their weight.
I wanted to replicate the effect, to make tha landing gear drop as well when flying. This meant again some more head scratching and engineering.
The front struts attach to a second base out of plywood, also removable for exhange purposes.
The next couple of posts are more about metal working than model building.
To make the final struts some jigs were improvised, for drilling and slotting them properly and repeatably.
The end holes had to be perpendicular to the tubing
once drilled, an orientating piece of 0,5mm music wire was added and the tubing was inserted into the sloting jig.
a very fine saw blade cut the slot. The jigs ensured all four parts came out equal.
next step was to sand the ends to a round profile, using pieces of aluminium as guides for the sandpaper
final step was to press the sides together, with a piece of aluminium sheet as a limiting spacer.
6 pieces were made, although only 4 would be needed, in case something went wrong. It is more difficult to go back for an extra piece, psychologically at least.
Wow! Excellent use of gigs! Love it!
I have to remind myself to use more gigs to solve these kind of problems :)
Hi Alex,
yes jigs can prove most helpful in many situations where repeatability is an issue. They do not have to be very elaborate, sometimes the simplest things work wonders.
Going back to the Avro, the cowl panels would be made of polystyrene sheet, to simulate the metal sheet of the original. Given the various curves, these would be vacuum formed, so a plug had to be made.
Starting with the side view and some cross sections, using plenty of balsa fillers and sanding, the basic shape was achieved
some more sanding and grain filling improved things even more
again some temp assembly photographs, quite a waste of time but necessary to keep the motivation and the build going forward.
Brass rod hinges have been added at this point to the horizontal tail surfaces.
the rudder was hinged differently, similar to the original
The cables to the control surfaces would normally be fairly easy to replicate. Since however they are to be functional and thus need to have as little friction as possible, it was necessary to route them properly.
The exits from the fuselage are clearly defined but they need to turn inside the fuselage, so an internal former was added, with short aluminium tubes bent to the required curves to guide the cables.
...
...
one of the weirdest parts I had to make was a universal joint for the front landing gear legs.
To it attach the front legs, the wheel axle and the horizontal tubes that go to the "V" struts.
Simple enough, only it had to allow free movement of each wheel up and down, like an independent suspension.
To design it was one thing, to make it was another. Especially in the size of a few milimiters.
Much to my surprise, carbon sheet worked great for this, together with CA.
The right part is the finished first one, the left one is the preparation for the second one. A lot of sanding to get there.
and the final result
The wheels were also made.
0,5mm plywood core with balsa halfs glued on either side.
I used me "lathe" for the balsa halfs, which is a Proxxon tool (like a Dremel) and an XY table to move the cutting tool around.
When the contour was finished the part would be cut and sanded to the exact height.
Here the jig used for rough cuting and later sanding
...
And finally, the model on its legs.
Main landing gear struts still temporary basswood ones
Back to the nose cowl...
After fine sanding a primer coat was sprayed and more sanding followed. I had no clear idea yet as to how I would do the panels, I was thinking of individual ones so made a base for the plug to be oriented in a variety of positions
time for some vacuum forming...
I even tried a large one-piece cowl and kind of liked it more.
some more parts were made and the best ones chosen for further work.
The main landing gear legs, made of aluminium tubing, carbon rod and polystyrene for the brackets.
The yellow sleeve is rolled tissue, to simulate the outer cylinder of the suspension assembly.
A jig (what else) to help align them with the other parts
The dummy engine was drawn in 3D CAD and sent to a company for printing and a dummy prop was carved from basswood, so it was possible to put all this together.
and again some assembly photos
I confess, I did spend some time looking at it, the thin scale like structure certainly looks the part.
The dummy prop shown earlier was not to be the proper one. The original one was laminated so I thought of doing the same. Not quite visible in the end, as it turned out.
the process has been described in the how-to topic about dummy props, so mostly photos here.
Yet another jig, for the rear face contour sanding.
...
And those main legs again, this time in their finishing stage.
I cheated a bit on their operation, I admit...
...
Incredibly beautiful, George.
What an eye for detail and craftsmanship!
The buildtread itself is also a pleasure to read and view, with beautiful, clear photos.
Cheers, Ramses
That landing gear is crazy beautiful and shows just how nutty we can get in this hobby! LOL Keep up the fabulous work.
Thanks Ramses, Glenn...
The time now is 2023 (some pictures taken on the 30th December 2022).
Dummy engine painting. The parts were given a coat of primer and some fine sanding.
Alclad lacquers were used, of varying shades
cylinders in gun metal
crankcase and cylinders
first cowl part finished and primed.
And somehow a defect appeared, the plastic deformed quite significantly. This wringle proved impossible to get rid off.
Time for a second one.
can still be used as a base for the motor.
Rocker arms and pushrods, 14 of them.
plug sparks and cables
Well, the A.S. Lynx looks almost finished.
Back to the propeller for the Lynx.
The original had brass sheeting on the blade leading edges. This would be simulated with plastic, vacuum formed to the prop form.
A simple base was made and for each blade some parts were made.
Careful trimming and sanding to shape.
It was important not to mix the parts up, hence the I and II on them and the prop hub.
Being vacuum formed, the plastic fited perfectly.
First varnish coat for the dummy propeller. 2 more followed.
A second cowl was made, with more care to avoid stresses that may appear later.
The bottom would also have the access hatch, necessary for the battery and the motor controller adjustment.
Temporay basla supports, glued with white glue, ta stabilise the rather flimsy part.
Internal plastic ribs to add rigidity.
a formed piece was glued on the inside, as a base for the cover, also shown in rough form
and the total nose put together for some more photographs
and the right side, this time without visible flaws.
The dummy propeller fits into the crankcase hole.
As I was writing at the time, when posting in the old HPA, I was begining to question how far was too far, in terms of detailing.
Making the strut end plates was a new one to me, especially since they would be functional stress taking parts.
Some CAD work to define the hole positions and a plywood jig for the starting holes of 0,5mm
mass production...
8 wing struts, 4 cabane struts, 12 in total, 2 ends, so 24 final parts needed.
drilling to final diameters, done by hand for each hole, in 0,1mm increments.
Some flattening and rough cutting
and sanding to shape, using tube guides for all the rounded edges. Jeweller's work...
back to some normal aeromodelling... Covering...
A simple base was made to hold the wings steady while covering the lower side.
Dope and thinners, using a thin piece of balsa to apply uniform pressure to the area being glued.
trimming the bottom side edges
top side follows.
Pre-shrunk tissue used everywhere, makes the task much easier.
and some finished wings
tissue shrinking, a simple base to allow full ventilation
By George,George!
I'm impressed beyond words by your meticulousness, patience and capacity to take pains!
I will have to lift my game, to be worthy of contributing to HPA!
I'm glad you're re-posting this immaculate build, George - for purely selfish reasons: I want to be reminded of how you made the rib tapes with stitching ;).
Stephen.
Thanks Jeremy, Stephen.
I hope the delay in continuing posting only helped build up some more anticipation :D
So, what comes next after covering? Why, stitching and rib tapes, of course...
For the stiching I built a simple balsa frame and glues a piec of tissue. Shrunk and doped, then thread was passed over it in parallel passes, the distance between passes defined by two rows of pins. In each row the pins are about 2,5mm apart from each other = the distance between individual stiches on the real plane divided by the scale.
With the thread ends secured, the thread passes were glued to the tissue, using white glue. Enough passes to make sure a strong bond has been achieved, this is most important to avoid dissappointments later on. White glue was chosen because it is not affected by dope and thinners that come later.
Mike ZK-AUD has suggested in the old HPA to attach the tissue to a cylinder and wrap the thread passes over the tissue, with the obvious advantage that the thread would be tight on th tissue and would be easier to glue. I agree, alothough I did not try it, so I am mentioning it here as an alternative.
Nest step is to slice the released tissue-thread part in thin strips. I use my small micrometer adjustable base. The width correspond to the rib thickness, more or less. The Avro had ribs of various thicknesses, so these were cut. Plenty of them given it is a biplane with many ribs...
lots of strips...
Next comes the first fun part.
Glueing individual strips onto the ribs. Glue one end first, let dry and then pull and glue all along the strip. When dry, cut off the excess ends.
Continue to the next rib, then to the other side, then to the other wings...
Great fun ;D
I use thinner for attaching, it softens the dope on the wing covering (did I mention int already had two coats of dope? no I didn't) and the dope on the strip.
And here is the first test of how strongly the thread was glued to the base tissue. There is nothing worse than tiny 0,7mm long thread pieces suddenly flying away from the glued strip.
stabiliser halfs and rudder being shrunk and then doped. Fixed on the board while drying
So, at some points all the parts have their rib tape strips glued on.
I chose yellow for the tissue on purpose, to contrast with the white covering while gkueing, for easier alignment. The strip should sit over the rib, of course...
Rib tapes are next.
Again, a simple piece of tissue, not doped this time is cut in 3mm strips, using the same micrometer adjustable cutting base.
Fun part #2.
Each rib tape is given a coat of dope, then placed over the rib stich strips and some thinner with a thin paintbrush.
The secret here is to press the rib tape onto the wing so that there are no "air pockets" or gaps between them. For best results I use a block of very soft balsa, to press the rib tape. The balsa is soft enough for each protruding rib stich to push through and allow the rest of the tape to be pressured.
I hope my description on that is adequate.
And that is that! An extra coat of thinned dope can be given, to seal everything together.
Simply amazing! My eyes would be bugged out of my sockets trying to do that kind of work! :o
Thanks Glenn,
to be clear, of course I use a large magnifying lamp while building. Impossible to do otherwise.
Back to the struts, the end plates have been bent and glued to each strut, doublechecking the type and orientation. In the end all 12 of them are different.
Thin paper was rolled at the strut ends to simulate the socket
Three light coats of wood exterior varnish brushed on, give a very nice finish
Tailskid, made of thin brass tubing.
Not the lightest choice for the rear section, but the only way to arrive at a delicate, yet strong enough, scale like structure.
Tail section showing the spring for the tail skid suspension.
And start of the fuselage covering, top curved deckings first
After the top decking was covered, it was time to add the cockpit coaming. Made of pink foam, roughly shaped first and then glued on with white glue.
The groove is for proper positioning on the top decking and to hide the balsa ends afterwards
to achieve a uniform overhang distance from the top decking, a balsa spacer was used to trim/sand the excess foam.
Next it was sanded round and pressed where the coaming stitching would be.
Finally it was sealed with thinned down white glue to a smooth finish for painting.
Open cockpit means the wood structure is visible through the cockpits, so this area was given some brown varnish too.
Next came the covering of the sides, the rear top fuselage and the bottom
So, that would mean the fuselage is finished. Normally. But there is some debate as to what normal is. Or how normal the builder was when he was building this...
Another surface texture detail that stuck to my head was the fuselage lacing. Having a clear 3D texture, I was not satisfied with simply simulating it by paint or leaving it out altogether.
So (the madness begins) first I measured from photogrqaphs the aproximate pitch of the lacing pins and drew the lacing paterns.
The drawings were printed on thin paper and the paterns were cut.
Each paper patern was secured with scotch tape on a balsa block / base.
Next, 0,3mm holes were drilled at the proper position and angle.
Then, short lengths of 0,3mm polystyrene rods were inserted into the holes and glued to the paper patern with dope. For the moment protruding much more than needed.
Fine thread, to simulate the actual lacing material was passed through and glued to the rods with 2-3 coats of thinned dope.
Skiped photographing the last stage, the rods were trimmed closer to the base and sanded almost to the height of the thread thickness.
When satisfied with the appearance, it wa stime to remove the paper patern from the base. This was done carefully, by cutting with a sharp razorblade all the rods under the patern, that were now glued to the balsa block.
Final preparation step, to cut the pattern outline, getting rid of the excess paper.
and, at last, to glue the resulting part onto the fuselage covering (doped by now).
Crazy? Yep...
Effective? Yep...
Forgot to add, after covering the fuselage sides, the holes for the control cable exits were cut and reenforced with tiny paper loops and the thread was passed through from the inside.
It would be impossible to go from the outside in, unless one had 1/20 scale workers doing the job, a joke that was discussed a bit back in 2023...
Wow, that is insane! In a good way :D
I will never be able to pull of something that detailed!
Excellent!
Thanks Alex,
do not forget of course that what you see is the result of many years of esperience and specialisation. So, never say never...
Following the various surface details it was time for the initial assembly of the struts.
Center struts first
fuselage bottom side, the landing gear bases are visible as are the holes where the screws go into to secure them.
A simple template for the wing struts, to bend to the correct angle the strut wire ends.
insert the struts to the corresponding tubes in the wing and bend them until they align with the position in the template. First bottom ends then top ends.
As can be seen, the wings have beem painted at this point
Fuselage also got its silver paint, white aluminium lacquer, actually.
Scotch tape mask to seal the cockpit holes during spraying.
After removal the cockpit coaming was hand painted in dark leather.
Center struts and wing center section also glued to the fuselage
horizontal tail also painted and in position.
After some unsuccessful attempts the wing roundels were made with decal paper sprayed blue and cut with a compass cutter. The white was sprayed onto the wings before application of the blue decals.
The decals conformed quite well to the texture of the covering and the resulting outline is very crisp, hats off to Mike Stuart for presenting this method on his website.
https://www.ffscale.co.uk/raiden.htm
The same with the fuselage black numbers, decal paper sprayed black, cut and applied to the fuselage sides. Alignment with a template where the numbers are cut out.
The white base for the fuselage roundel can be seen here, the blue is coming immediately after that.
Rudder also sprayed white in the center, for the same reason.
Fuselage roundels completed. Look even better than painted.
One migh think that the rudder might be too coplex to cover with decal paper, but it worked perfectly, even with the overlap at the edges to achieve a blue finish all around.
The 0,5mm wire is the hinge axle.
And the rudder in position as well.
Hallo George! Dein detailgetreuer und wirklich gelungener Bau des elektrischen Avro 504-Doppeldeckers zeugt von hervorragendem Realismus. :)
Lastwoodsman Richard
PS I tried to copy and paste the Greek translation ( mostly a bunch of question marks ), but it will only copy and paste the German?
Anyways, beautiful build!