A wing printed in LW-PLA

[Simon Chaddock]

You might think that is not particularly unusual. You can buy the stl files from Plane Print do make virtually a complete Sea Gull in LW-PLA but it does have a carbon tube spar.

What I was looking ar was to make a wing stiff, strong and still be light enough to be just printed.

It is well known in conventional 3D printing that strength and stiffness of an item is several times more along the line of the bead layers than across them. This applies just the same for LW-PLA.

So how print a wing with the layers going span wise. Obviously printers have a limited size bed to the wing would have to be made is sections and fixed together.

The most practical method I have found is to print the wing it two pieces flat on what would be the spar line.

 

Just a short 50 mm test length that creates a Clark Y wing  section with a chord of 6.75" (172 mm).

There are no ribs but an automatic 'gyroid' infill that provides close support to the wing skin.

The final result is impressively rigid and has a smooth ripple free surface accurate to 0.05 mm.

   

The 50 mm long wing weighs 7.7 g but a 250 long section would take a good 6 hours to print both parts as LW-PLA  has to be printed quite slowly. 

The next problem was to consider how to do an aileron.

It was quite simple to modify the wing tail part to finish is a recess.

The intention would be to use the accuracy of 3D printing so the aileron could be a close fit.

The aileron itself had a flat under surface (thank goodness I chose a Clark Y!) so it can printed flat on the bed.

Rather than short test piece the wing with aileron would have to be printed out at the full 240 mm length to prove it actually worked.

The 240 mm wing tail and aileron showing the fit that is possible.

The problem is how on earth to hinge it.

To be continued.

  

 

 

 

 

[Andy Stephenson]

Check out how Eclipson do it with pins and loops.

[Simon Chaddock]

Andy

As far as I can tell from the Eclipson video it appears at least for their free plane that it is printed conventionally with components "rising out of the bed" meaning that carbon? reinforcement is required to achieve the required stiffness. They do look nice though. 🙂

 

My intention is to create a relatively simple 1m span plank wing that is just printed and the pieces glued together.

The design process is pretty simple using a CAD wing section "sketch" for each of the front and rear sections. 

  

 

The 'span' of each section is set in CAD but is likely limited by the printer bed dimensions, 240 mm in my case.

 

That is the limit of the basic CAD design work. Once converted into an G code file the parameters of the print are all set in CURA.

It is printed on what would be the spar face so the filaments run along the span rather than across it which is more conventional.

https://www.youtube.com/watch?v=X3bH_Irsx5M

The difference in strength and stiffness is remarkable so hopefully a 1000 mm span "just printed" wing is doable.

Quite a bit more to do. Mostly many hours of printing. 😉    

[Simon Chaddock]

The wing would be made up of 4 240 mm panels with a short dihedral centre section.

For simplicity the aileron would be the length of the outer panel. I hoped the accuracy of printing would allow the aileron to be a good fit into the wing.

A CAD view of the aileron radius recess.

 

The aileron matches it. It is actually 238 mm long to allow for hinges at each end.

 

Printed out showing the sort of fit possible.

With simple acetate 'pin' hinges at each end a short video showing it is really free moving.

https://www.youtube.com/watch?v=bFyXPgph32M

I later substituted printed hinges that made it easier to position the pin at exactly the centre of the recess radius.

The aileron is still free moving to over 45 degrees each way. The actual gap is no more than 0.3 mm. More a technical achievement than anything else as I am sure the close gap will make absolutely no difference to the performance of the wing!

The 240 mm wing panel with the aileron weighs 33 g.

Still musing over whether to actually print all the bits to complete the full 1000 mm wing. It will take about 20 hours printing and use up quite a bit of my LW-PLA. 😟  

[Simon Chaddock]

After many hours printing I have the parts for the complete wing.

The tricky bit was the short dihedral centre section. This is the front part.

It gives a dihedral of 6 degrees.

It is slightly oversize so the inner wing panels plug into it so the glue joint is in shear.

It took a bit of trial and error to get a good tight push fit.

 

The complete printed wing.

Each panel is 240 mm, the 50 mm centre means it is actually 1010 mm span!

Next is to consider the aileron servos. As the wing relies entirely on the skin for strength and stiffness the servos may have to simply glued on the underside.😲

It will be several days before the glue achieves maximum strength before I can test to see how rigid, or not, it really is.   

Edited March 12, 2023 by Simon Chaddock

[Simon Chaddock]

The final challenge was to install the aileron servos.

To avoid any cutting of the skin they are simply glued on the underside with small streamline fairings which also double to help retain the servo.

The completed servo installation.

Like the servo the wires are glued down too. A bit crude but it works.

Note the thin (0.5 mm) PLA strip added to the underside to provide a bit more bending stiffness. 😉

The wing fixed by bands on the Wing Dragon fuselage for the test flight, whenever that might be!

The load testing so far has suggested that the wing will fail in compression with the skin buckling well before it tears in tension, probably by a factor of 2 or even more. 😲

Not really very surprising given all the loads are taken by a thin 0.8 mm skin. Adding closer skin support or doubling the skin thickness would help but it rapidly adds to the total weight.

Keeping to a simple print and without an inserted spar I might have a potential solution but the initial wing needs to be tested first.

We shall see! 

[Simon Chaddock]

Only a short video but it does fly.

https://www.youtube.com/watch?v=_6xlZA_1luc&t=1s

It was a short video because an RH aileron horn broke free. I saw the aileron flutter so landed asap quite under control on one aileron.

Repaired I flew later at 5 pm, in the rain, and looped it no problem.

I make a 'proper' video when the weather obliges.   

 

[Simon Chaddock]

As I indicated right at the beginning printing the wing with the filaments going 'span wise' greatly  increases the ability of the skin to resist tension but being a thin skin even though well supported causes it to buckle at a much lower compressive load.

The solution is to provide a square section compression 'spar' top and bottom, still in LW={LA and printed at the same time as the rest of the wing along with a positive warren brace shear web between them.

The result is the leading edge wing section becomes a "D" box structure.

The skin ahead of the spar is supported by a gyroid infill as on the original wing.

For a modest 2.5g increase in weight this section can survive a bending load well beyond that of the original printed wing.

Using this design suggest the same 1010 mm wing would weigh 156g complete with servos rather than 148g but it would be significantly stiffer.

 

At this point I am still aware that using a simple 3D design process is limited to creating a 'plank' wing that even in LW-PLA is still heavy for its strength.

As an example the wing I built from XPS sheet foam for the same fuselage is stronger and stiffer yet only weighs 112g complete.😲

Printing a complete plane is an attractive proposition but as a "builder" I think I will stick to printing just the parts where the positive advantages of printing outweigh any penalties of using it.

        

[Nigel R]

its impressive nontheless.

 

what sort of parts do you envisage continuing to use raw 3d printed parts for?

 

 

[Simon Chaddock]

Nigel R

By far the best use I have found so far, in the sense it makes the best use of its capabilities, is for EDF ducts in LW-PLA and scale nose cones in ordinary PLA.

The duct for the Lippisch P13a 50 mm EDF is a typical example.

Not easy to make in any material but in LW=PLA  the complete duct less the EDF totals just 33g.Once glued together it is plenty strong and stiff enough to be safely handled and with care the fuselage was actually built around it.

 

Something like a Sea Vixen has a huge radar nose cone.

Printed as a hollow single wall structure in black PLA.  

It weighs 5g. It will not survive much of a ground impact but better to damage a nose cone than any part of the fuselage structure. It only takes the touch of a button to print another.😉

It is of course possible to print lots of sorts of smaller bits but it is done more for convenience than any practical advantage.

 

Of course if you really must have a complete scale 9 cylinder radial inside the cowling then 3D printing is an valid solution.

Not a trivial undertaking as it consists of nearly 60 individual printed parts glued together. 

 

 

 

 

 

[Nigel R]

many thanks simon 👍

[Simon Chaddock]

A short update. Earlier this month the LW printed wing had a serious crash, a vertical 'nose plant' from 40' as a result of a an elevator servo failure.

As the wing is retained on the fuselage by rubber bands the damage appears fairly superficial however closer inspection shows there is much internal 'shock' damage and distortion.

https://www.youtube.com/watch?v=9yNSQvahgew

The spar remains stiff but the large area of internal damage significantly reduces the wing's torsional resistance.

 

In simple terms it could be said that the wing design was just too close to critical as it ended up being damaged "all over" rather than failing at a weak point which would likely be easier to repair!

Oh well never mind.