Ballerina PET

[Clive Hall]

The fuselage is now sufficiently advanced to be ready for wing fitting: next job, build the wings. The rib sets are cut and waiting.

The wing planform is sketched on paper, mainly to give the rib spacing when fitting to the spars.

[Clive Hall]

Visitors over the holiday blocked a lot of building time, but now the wing panels are making progress. Apart from being enlarged this wing is close to Peter Miller’s original design. The main difference is in the fitting of two servos close to the ailerons rather than a single central servo with pushrods and bell cranks. The paper tubes are wide enough to contain the servo extension leads complete with plugs. They may not provide much spar stiffness, but they do help to make the ribs firmer in position.

The next post will describe an unorthodox method for fitting the wing leading edge balsa sheet skins that form the D-box structure.

[Peter Miller]

Look very nice.

Personally I never lift a wing from the board until all the leading edge sheet is no. That makes sure I don't geta wrp innit butthen I can get warps in anything if I am not careful.

[Martyn K]

looks great - sorry I missed the start

Subscribed

Martyn

[Clive Hall]

The two wing panels are temporarily clamped together pending a decision. Which should I do first, glue them together or fit the leading edge D-box sheeting to the separate panels?

If the sheeting is done first it will not then be possible to clamp root ribs together or the wing joiner to the spar while the glue sets. If the join is done first the sheeting will be more difficult to fit, but the strength should be better and the fitting of the wing dowel and wing bolts will be much simpler.

[john stones 1 - Moderator]

Sheet top or bottom then join, I sheeted top whilst pinned to board, but either way as long as panels are kept flat wing should turn out warp free.

John

[Peter Miller]

If you follow my sequence yo don't have to clamp the ribs together. Build one woing completley incluing R-1. Before adding top sheet to second wing and R-1. With the wings joined add the second R-1 ans then top sheet to the second wing.

[WolstonFlyer]

Looking really good Clive, it is a very neat build!

[Clive Hall]

I decided to join the wings before sheeting them. The reason is that my method for attaching the sheet includes a means of controlling any tendency to warp, so I took advantage of the opportunity to clamp the joins. This has not arisen before, because for a long time my wings have been either one-piece without dihedral or separate panels attached to a fuselage by ali tube. I was content to remove these wing panels from the board before sheeting the leading edge; it was never going to be a problem because simple jigs prevent warps.

To ensure accurate incidence angles and freedom from twists the wing frame is held down on cradles, one pair for the top, and another pair for the underside, each pair being a root cradle and a tip cradle. For long wings any number of intermediate cradles can be used. These match the rib profiles and support the frame in position while the skins are being pressed into place.

The wing sheeting technique I use is unorthodox, but it has proved fully reliable and accurate over the past ten years on a number of models. It involves the use of contact adhesive which removes all need for pins, clamps or weights to hold the skin in place while the glue dries. In this technique the glue is allowed to dry before the skin is applied, but it does mean that the skin must be placed with precision as once it touches the frame it sticks and will not come off, so there are no second chances. It relies on the jigs.

The procedure is first to position each skin accurately on its wing panel by pinning one edge to the spar. Next drill holes to fit a cocktail stick through the skin and the spar, one hole at each end, and press the sticks lightly into place. (Inset – fuller view)

Lift the wing from the cradles complete with the skin, turn it over, and then while pressing the skin against the ribs with a ballpoint pen, mark all the places where the frame touches the skin. Now repeat the process for the other side of the panel. Do both panels, then there will be four marked skins ready for gluing.

[Chris Barlow]

I'll admit if I tried it that way I'd end up with propellers not wings!

Do you pin the wing down before laying the skin on?

[Clive Hall]

Chris, it's not as scary as it may look. As soon as I have fitted some skins I will post the rest of the method in detail. The wing is not pinned down, but the cradles are. I pin them to the bench and then place a wing panel on them with some weight to keep it from moving, and then apply the glued skin to the glued frame. Be assured, I have used this method many times and it is easy and accurate. The pre-gluing of the parts is the one bit that takes time; the rest is a piece of cake.

[Clive Hall]

Just a few of the models with wings built by this method. All flew well.

[Chris Barlow]

You've got me interested! Never considered contact adhesive for wing skins before.

[Clive Hall]

I think the reason why contact adhesive is not considered for wing sheeting is that the sheet, from the first moment it touches the frame, has to be fitted with absolute accuracy or it will never fit. Also the gluing routine is not obvious at first. The cocktail sticks ensure the required accuracy, the cradles prevent warps, and the rest is easy.

[Peter Miller]

I would also say that contact adhesive needs more area for strength as it is pretty weak over small areas.

[Clive Hall]

I suppose the joint strength depends on which contact adhesive is used, and on what materials.>>

I first thought of using this method on a model with span 85 in, estimated weight 17lb, power Zenoah 38cc petrol, so I did a strength test on a sample of a skin joint. The results removed all apprehensions.>>

The actual figures are not available now so I have done a new test and estimated the force needed to tear the upper leading edge skins off the 62 inch Ballerina. The test piece has a gluing area of 1.46 square centimetres. I discontinued the tension test at 7 Kg because it was becoming dangerous – and there was no point in pushing it further. The glue area of the leading edge contact of a single wing rib is 2.8 times the area of the tension test..>>

I used a spreadsheet to calculate the total glue contact area on the Ballerina leading edge top surface. By comparing this with the test piece it seems that it will require more than 900 Kg to pull the skins off. I have no worries about contact adhesives for wing skins.>>

[Peter Miller]

I don't know if the total area is the critical factor. A 3/32" wide rib might start the separation process.

I have done tests where I glued 1.4" sq. Sppruce to plywood using Aliphatic, Epoxy, Superglue and just for the hell of it, Thixofix. Contact area being 1/2" X 1/4".

In all cases except contatc adhesive the wood broke. With contack adhesive there were a coule of wood fibres but that was all, the rest was the glue failing

[Chris Barlow]

I've used contact adhesive a lot laminating Formica to MDF and it never failed unless it got wet.

My only concern if using it for wing skins would be whether it would be strong enough in sheer strength to prevent the wing fron flexing and twisting. Having never used it for this I really wouldn't know....yet.

[Clive Hall]

When I did the original test on skin adhesion with contact glue the test piece was a replica of a part of a wing, a piece of rib with a section of skin attached. It was written up in the other magazine, of which I have a copy somewhere which I may yet be able to find.

I think the total glued area is relevant as the start of the separation process is likely to depend on a peeling effect, which may not be able to begin as all the skin is held on with no opportunity for an individual part to lift to begin peeling.

The original load test was done with a fisherman’s scale with only point contact under the skin next to the rib, yet the strength was impressive, with no peeling tendency. I think the bottom line here is the number of models that have had long flying lives, have been subjected to some violent flying, and have never shown the slightest sign of skin separation, with the performance of the test pieces pointing to the same conclusion.

[Clive Hall]

With regard to torsion resistance in a wing, the contact glued specimens have all had excellent torsion resistance, wings as stiff in torsion as any I have ever experienced. In this connection we are talking balsa/balsa or balsa/ply and these combinations seem to work particularly well with EvoStick contact glue.

[Clive Hall]

The pull-test piece was two 12 mm square faces of medium balsa joined with contact glue, the grain arranged to simulate a sheet-to-rib join. On a straight pull it held more than 16 lbs direct tension, several times the weight of the complete model, yet its area was less than 100^th of the contact area of the glued skins.

For thesake pf completeness I tried a shear test, the test area again being 12 mm x 12 mm. Result; it failed at 25 lbs, (11 Kg), with the failure being 80% balsa breaking rather than the glue separating.

I think the airframe might collapse some time before this kind of loading is reached. How much more do we need?

[Peter Miller]

Well that seems to prove it. If it works for you far be it from me to argue but I think I will stickto my methods.

[Chris Barlow]

I suppose a lot of the strength will depend on application and technique.

For example if the adhesive is too thick it will flex luke rubber. Too thin and it'll soak into the balsa leaving insufficient glue to bond to the joining glued balsa.
Will be interesting to see anyway.

[Rosco]

Geepers, to far advanced for me....all that technical stuff!

Nice work though. I can appreciate the science behind it all.

Back to your build.....any ideas on the scheme. Everyone us adding their own personal touch which is the fantastic side of this build.

At first I was a bit disappointed with the Ballerina being chosen but now I think it's been a good one!

Anyway, nice work on the build Clive.

Looking forwards to some 'closer to completion' pics.

Rosco

[Clive Hall]

Rosco, I want some dark colours to make the model easier to see, so I have been looking at the Travelair Mystery Ship for inspiration. This is black where Peter’s model is red, and red where his is white –more or less There is some white pin-striping between the colours which should be an interesting challenge.

More pics soon when the wing skins finally go on, then I attend to the tail unit.