1/24 scale DH Mosquito for rubber power.

[Prosper]

Today I did lots of 'spot welding' (dots of thin CA on each of the many joints in the wing); made the aileron shroud seen in pic 2, and a leading edge tip member - this is a bit of a burden at 0.2g, but has served very well in the test model, which has taken some thumping wingtip landngs without either tip breaking.

I used a long sanding block to give final shape to the ribs at the leading edge before it's sheeted, see pic 1 - actually the long sanding block is the dark thing at the bottom, propping the wing up.

Stephen.

[Prosper]

Pic 1. shows the leading-edge sheeting bound to a mould. The sheeting is 0.6mm. I used light 0.8 or 1.0mm sheeting for the test model, to save time - but now I hope to save weight by taking the time to thin various parts. Also visible in the picture are the moulds or forms for the wingtips, along with the 0.6mm sheet blanks ready to be wetted and bound to the moulds.

Pic 2 shows the leading edge sheeting released from their binding, and 3) shows one of the wingtip sheets bound to its mould before drying. The curved ouline of the mould makes binding very difficult and it ends up as a mess of wrinkles and sticky tape.

Last picture is of the leading-edge sheet in place on the wing. Some judicious trimming will be needed until it fits snugly enough to be glued down for good.

Stephen.

[sandfac]

Stephen,

Really enjoying following this build and flights of the testbed. The Mosquito is a favorite airplane and it's nice to see it done with your precision and craftsmanship.

[Prosper]

Many thanks Oliver.

Yesterday I made the sub-spar that runs where the radiator face is - and forgot to take photos. I also did the coarse part of sanding the trailing edge to shape. Then today I've spent ages tying myself in knots over the radiator intake. I took the best measurements I could from suitable Mosquito photos but I'm not confident that it'll look right when finished. That's the difference between a quick test model and a properly-finished job - it takes twice as long bcause of piddly details!

Next, because the radiator intake is narrower that the radiator face itself, I cut a piece of card the same size as the radiator face, to check that I could insert it thru the intake then erect it and press it against the sub-spar where the radiator face is. That's so's I know I'll be able to paint or draw (or something) a representation of the radiator and smuggle it into place after the tissue covering's done. Fine! It went in and sat vertical with a bit of prodding, snug against the spar. Duuuuh . . . how to get it out again? Hmmmm, hadn't anticipated that. The answer is that you have to destroy it in the process - see last pic.

Stephen.

[TheLurker]

Nice work Stephen

...so's I know I'll be able to paint or draw (or something) a representation of the radiator...

.
May I draw your attention to my favourite trick for a radiator grille/mesh? Surprisingly convincing even when seen close up.  A very thin film of (spray on) varnish is needed to stop the pencil lead smudging.

[Prosper]

That's ingenious, Genius! Kudos, as they say. It won't work for this project but I'll remember your wheeze for future use. The radiator here has only vertical elements and no grille as such except for the inmost little part. I'll probably try to render it pretty dark, as it looks in most pictures. Talking of which, looking out a picture to illustrate the radiator, I found this (pic 1), which I'd only glanced at during research . . . and now I'm really angry:

  After gawping at every image and every video from every angle I could find, I'd satisfied myself that the wing section of the portion of the wing housing the radiator is completely smooth - a continuous curve. Yet this drawing shows a distinct bulge on the lower surface. My belief that there was no bulge led me to make an intake slot which looks too narrow, to my eye

"but I'm not confident that it'll look right when finished."

. Now, this is only a drawing - a sketch, so to speak - but it looks pretty authoritative. The bulge would explain how to make room for a wider slot. Too late now, unless I take a step back and rebuild this area of the wing.

Stephen.

[Prosper]

Apart from the radiator business, I sheeted the leading edge. This was with 0.6mm soft wood, and I expect it will be thinned slightly more by the final sanding fair, prior to covering.

Pic 2) shows this sheeting from behind. I normally choose a mainspar position further back, say 1/3 of wing chord, and sheet ahead of this. That means the sheeting is glued to the mainspar. In this case the mainspar position is dictated by the way the engine nacelle mounts to the wing. It's further forward. So the L.E. sheeting extends behind the spar, and is glued to very thin spanwise strips (I think they're just too flimsy to be called 'sub-spars). I think this will be strong enough, and the spanwise strips should support the sheeting well enough that it doesn't sag between ribs due to tissue shrinkage. It worked OK on the test model.

When I first started contemplating the design I presumed that I'd have to cut away the mainspar where it crosses the nacelle, in order to give room for the rubber motor. When it came down to detailed design I was pleased to find that this wasn't necessary. Left of pic 2) shows where I've partially filled the gap between the mainspar strips, and added cross-pieces behind the spar. These will have holes in them to accept pins on the detachable nacelle.

Pic 3) is the inside of a wingtip skin. This is nominally 0.6mm too, but it's been somewhat crushed round the edge by the tissue that bound it to the mould or form. I hope it'll be strong enough. I don't remember edges this thin on the test model. Mind you, I won't be flying this model over entirely unsuitable ground as I have the test model. That has suffered both tips hitting hard and rutted ground multiple times with no breakage (final pic).

Stephen.

[Prosper]

I've been unable to work on this since my last post, other than in bits of time here and there - which I've used to get the left wing to the same stage as the right - nearly done. There's now a switch of focus. I don't want to install the wingtips until all the multiple spanwise strips are in place. I don't want to install those until the pendulum aileron pushrods and bellcranks are in place. I don't want to install those until the engine nacelles are made and fitted. So next I'll be tackling the nacelles.

The bulge would explain how to make room for a wider slot. Too late now, unless I take a step back and rebuild this area of the wing.

Regarding the radiator intake slot, I've decided not to modify it, though I'm not too happy with what I've got. After seeking through my picture collection again for useful evidence, I couldn't find any photo suggestive of a noticeable bulge under the wing . . .  but could find to the contrary, as the 'screengrab' attached seems - to my eyes - to indicate. With the cowlings off, the line of the underwing is cearly seen.

The ghostly hand swimming into the picture isn't in fact an apparition but belongs to one K. Weeks. The Mosquito is one of the 'new-builds' - a brand-new Mossie made from scratch. This does raise the question of what concessions and compromises the builders might make, and what gaps might exist in their knowledge. Maybe they have every de Havilland drawing and only use the original specs. Either way, the area concerned isn't one that 'comes out well' in period photos - at least at the perspective angles I'm looking for. So I'm going with the 'new-build' evidence.

Stephen.

[Prosper]

Nacelle wrassling.

The nacelles are less curvy than drawings like to show. All in all when making the balsa plug for the test model, I had to spend a lot of time working out the shape, leaning heavily on photos with panel lines evident and on u/c doors. It came to light that the aft end is 'hollowed out' - see pic 2, and 3 shows this quite well. Drawing u/c doors on the plug, then sanding the plug and re-drawing, until the ink lines looked like the real thing in photos, was my method. Same for panel lines (Mosquito nacelles were of  metal, not wood).

To save effort with the test model I made a plug that was symmetrical and it had to serve for both left and right nacelles. This made working out the blending of nacelle and wing difficult. The wing taper in thickness, the extended inboard wing leading edge, and the dihedral, mean that the L and R nacelles are somewhat different in order to fit their respective wings. This time I'm hacking away part of the plug and will attach an extra piece specific to the L or R body.

[Prosper]

Now see pic 4 which shows the balsa plug in the foreground and one of the test model's nacelles behind it. Look at the curves of the plug, then see how the model nacelle has failed to keep the curvature it had when initially moulded. It can't be down to tissue shrinkage because the fore part is flattened even though it has no tissue covering. I wonder if the exceptionally humid weather for weeks on end is a factor, and the model repeatedly landing on soaking ground. I suppose I need to provide more supporting structure - AKA more weight and more time.

Stephen.

[OZPAF]

it's a pity that you can't get away with just one nacelle,Stephen   That's an interesting way of judging curvature  - using the shape of surface lines - undercarriage doors and panel lines.

I'm a lousy 3d carver/shaper- I usually resort to 3d dwgs and the use of the old "bread and Butter" approach. Your system is far more accurate as it doesn't depend on the accuracy of the drawings.

I think you should leave your work partly uncovered and displayed in a transparent box.

John

[Prosper]

it's a pity that you can't get away with just one nacelle,Stephen

In the event, I saw that this wasn't necessary.

Your system is far more accurate as it doesn't depend on the accuracy of the drawings.

Yes, I think so, but it does need some basic 'solid' information, say nacelle length and width for example. All the rest can be discovered by study of many photos and getting ratios. A really good side-photo (good in the sense of not being distorted by the camera lens) can yield the basic shape and then the study of photos at different angles can show up 3D aspects like cross-sections. 
 

The picture shows the nacelle plug (which I made for the test model) with a razor saw in it - I made several cuts on the right side of the plug. Thin card can be pushed into the cuts and a pen can be run round the card where it meets the balsa. This yields a true cross-section as seen in the picture. These card cross-sections not only help in attaining an acceptably symmetrical plug, but also make the templates needed to shape formers for the nacelle internal structure.

Now I've gone further and sawn the nacelle plug right in half. Am I mad? Well probably. But that's not important right now. In fact as I hope to show in coming posts, this will make joining the moulded balsa sheet parts of the nacelle quicker and more accurate. It happens that the original saw-cut wasn't at all perpendicular to the plug's centreline, so when I completed the cut I had to add some balsa and sand it square - this can be seen as a thin balsa wedge in one or two pics. On top of that goes another 'endplate' of thick wood. This is just to stop all the handling and tissue-binding etc. from rounding off the edges of the plug.

[Prosper]

Also see the 3mm 3-ply balsa slabs which were shaped to be the nose formers. Hard balsa, glued with aliphatic.

Now, binding balsa skins to the form. Each nacelle has three main panels - front, rear, and top cowling. At least that's the idea. It worked for the test model. However I'm already on my second attempt of a front skin. It's got to conform to some pretty heavy-duty double-curvature. The test model skins managed this with just a couple of wrinkles which had to be cut out. I think it all depends on the individual piece of balsa sheet used, and practice. That's the second attempt in the pictures, and it's drying in my hotbox as I write this - along with a rear panel. See that there are some wrinkles right at the nose (they're just visible as corrugations). These may well be acceptable - the wood shrinks and flattens out somewhat as it dries. We'll see.

Stephen.

[OZPAF]

This is a fascinating approach Stephen - courageous as well. Cutting into a carefully shaped form!

John

[Prosper]

Thanks John! I've done the cutting-halfway-thru-the-form a good few times before. It doesn't seem to be a problem.

Where was I? Oh yes. I found that when the bindings were unwrapped, the wrinkles at the front of the nacelle panel were still present, so I resorted to brute force. I rewetted the local area and used a hot piece of metal to iron the wrinkles. Hot enough to hiss at it brushes the wet balsa. The now barely-discernible wrinkles can be sanded out - see before and after pictures.

I moulded four of the six nacelle panels yesterday and the remaining two today. It's a slow business because of the depth of windings and sticky tape required - this slows evaporation so the 'mummies' need to spend a couple of hours in the hotbox. In most cases 45 min seeems enough.

I'm learning. Initially, and also with the test model nacelles, I bound the centre of the wet balsa panel to the plug first, then on to the front and back ends. I changed to binding first the ends, then the middle. Hey presto, no wrinkles or puckered lips at either end - but a couple of folds at the middle. To deal with these I turned to what I should have done all along: cutting relief slits, as I do with balsa/aliphatic double-curved panels. These need filling, but the overall result is better - no touch-and-go brute force.