Quantcast

Testing a method for making moldless-ish D tube skins

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

RPM314

Well-Known Member
Joined
Mar 27, 2015
Messages
735
Location
NY, USA
Hi everybody,
After being inspired by the work of Rol Klingberg on YouTube building his new ultralight glider, I decided to test out and idea that I had for making the D tube skin using the shear web and ribs as a male mold instead of hot wiring a dedicated male mold. This is of course a trade between build time and the quality of the resulting surface, and I don't claim to have worked out the best way to implement this process yet, so I welcome suggestions for improvements and input from those more experienced than I about how this process might scale up when you try to do an entire wing with it.

As you might imagine, the central problem is that once you pull the skin tight over the ribs, the leading edge will sag between them due to the tensile compliance of the foam core, and the leading edge radius will balloon out between the ribs due to the bending stiffness of the foam. The key idea I had is to sandwich the layup on both sides with a material that has high tensile stiffness and high bending compliance in order to control those undesirable tendencies. For this, I used some medium weight poster board from Staples, which is a bit thicker than cardstock.
20201021_092048.jpg
I started by making ribs out of some random foam and a shear web with composite flanges in the klingberg style. The rib spacing is 200mm on one side, 400mm on the other, and the sections correspond to the first 25% of a 750mm chord wing. (I botched the molding of this part pretty good, I know)
20201021_100834.jpg
Taping posterboard over the ribs, 1/8in divinycell is formed as close to finished shape as possible with a heat gun to minimize the foam's spring-back.
20201021_123324.jpg
Then one layer of vacuum bag goes over the ribs, then the posterboard, then breather, peel ply, layup, peel ply, breather, posterboard, outer vacuum bag layer. The vacuum bag layers are sealed together, the vacuum is pulled, and then the whole assembly is taped down tightly to the ribs and shear web. (full disclosure, I botched this vacuum bag pretty bad and didn't get good compression)
20201021_173519.jpg
Much of the D tube surface appears quite nice. That kink near the aft edge is due to my attempt to make an integral spar cap in the sandwich core, which is a story for another thread.
20201021_173603.jpg
Around the leading edge, there is a very slight bump on top of where the center rib was, but it's hard to spot. Both edges of the D tube are also turned up slightly, so on a full wing an inch or two of extra span might need to be built and then trimmed off. Across the longer rib spacing section, I laid a straight edge over the leading edge and measured the sag to be around 1mm.
 
Last edited:

TiPi

Well-Known Member
Log Member
Joined
Aug 25, 2019
Messages
337
Location
Peeramon (AUS)
Look at the wing and elevator construction of the SD-1 Minisport. The leading edge is formed by a foam strip covered in a couple of layers of glass, attached to a very light front spar. The front spar consist of a 1mm plywood web with 8x8mm timber caps facing forward and shaped to conform to the wing profile. All ribs are foam, the whole wing covered in 1mm ply. 12kg for a wing panel with 3m2 area and 6m wing span.
 

RPM314

Well-Known Member
Joined
Mar 27, 2015
Messages
735
Location
NY, USA
Is this the wing in question? Makes sense to me, making a rounded spar at the LE before bagging the skin would surely reduce the sag close to zero. I don't think plywood skins are viable for ultralight gliders though, I would probably stick with fabric.
 

TiPi

Well-Known Member
Log Member
Joined
Aug 25, 2019
Messages
337
Location
Peeramon (AUS)
Yes, this is the wing. I can upload some more detailed pics tonight.
For a sail plane (glider), you could just cover the front section to the main spar in ply and the rest in fabric (leading edge to trailing edge). Well used method on many planes and would give you a more precise profile in the critical first 25% of the wing.
 

RPM314

Well-Known Member
Joined
Mar 27, 2015
Messages
735
Location
NY, USA
For sure, that's historically the most common way to do it. But for the design I'm considering, the carbon fiber sandwich panel skin is limited by the minimum gauge of available materials to (nominally) about 4kg. A +/-45deg birch plywood skin for the same D tube would be limited by strength to 2mm and about 9kg of wood, which means you would have to settle for a sailplane of lesser wingspan to make up the weight.

Looking at the plywood from Aircraft Spruce, the materials seem to be about four times cheaper, which is nothing to sneeze at. I’ve never built a D tube with a plywood skin before, do you think that the bending stiffness of 2mm ply could be overcome to produce a leading edge of better accuracy than what I've made so far? I've heard of people soaking plywood and strapping it down to the wing for a pre-form.
 

TiPi

Well-Known Member
Log Member
Joined
Aug 25, 2019
Messages
337
Location
Peeramon (AUS)
I think 2mm is overkill depending on your speed. For low speed (<100kts) there are 2 ways:
  • add a very lightweight front spar with the correct leading edge shaped foam nose and cover the lot with your fabric
  • use 1mm ply as the leading edge nose part glued to the foam ribs
1604137067530.jpeg1604137128979.jpeg
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,257
Location
Saline Michigan
Use it as is for UL weight and airspeeds and it might be plenty strong.

You used what looks like a laminar flow foil shape, but this scheme may not achieve much laminar flow - too much surface roughness unless you fair and finish to profile with dry micro, profile sanding, and then seal coats and paint. Even then, the skin may balloon enough between ribs to spoil laminar flow. Run the calcs looking at airloads and then skin panel deflections. Might be lighter and much more laminar flow with a solid core. Run the numbers on weight including all of the glue to hold ribs and leading edge spar and then skin to the structure, then run estimates of skin deflections from your airloads to be sure you have a viable scheme.

I take issue with your calling this moldless - you made a simple mold to build the curved skin, even if there is no separate mold... Grin.

Billski
 

Riggerrob

Well-Known Member
Joined
Sep 9, 2014
Messages
1,591
Location
Canada
Arron I’m suggested a similar method except that he used a sheet of aluminum where you use poster board.
Remember to wrap the sheet of aluminum in peel ply and we it asid after your composite wing cures.
Squeezing a smooth sheet around your composite wing will save you 100 hours of sanding time.
 

RPM314

Well-Known Member
Joined
Mar 27, 2015
Messages
735
Location
NY, USA
I think 2mm is overkill depending on your speed. For low speed (<100kts) there are 2 ways:
  • add a very lightweight front spar with the correct leading edge shaped foam nose and cover the lot with your fabric
  • use 1mm ply as the leading edge nose part glued to the foam ribs
Use it as is for UL weight and airspeeds and it might be plenty strong.

You used what looks like a laminar flow foil shape, but this scheme may not achieve much laminar flow - too much surface roughness unless you fair and finish to profile with dry micro, profile sanding, and then seal coats and paint. Even then, the skin may balloon enough between ribs to spoil laminar flow. Run the calcs looking at airloads and then skin panel deflections. Might be lighter and much more laminar flow with a solid core. Run the numbers on weight including all of the glue to hold ribs and leading edge spar and then skin to the structure, then run estimates of skin deflections from your airloads to be sure you have a viable scheme.
Yeah, the back of my napkin says that 1mm would be enough for everything other than a snap roll. But I also just ran a check for stiffness, and while both carbon and birch check out for torsional frequency, 1mm birch skin has a critical speed for roll reversal around 90 knots compared to over 400 knots for the carbon (in other words, 6% reduction of roll authority at 100 knots). Given that flexibility can be a limiting factor on high aspect flying wings because it erodes the handling qualities (cough cough SB-13 cough cough), that doesn't sound good for the birch.
Splitting the difference with 0.5mm plywood and one layer of 100gsm carbon comes out to a pretty competitive weight around 4kg, and still has plenty of stiffness. If it can be built to a good finish and doesn't present a weakness during ground handling, then maybe?

I used the airfoil from the SWIFT sailplane, which is advertised as “laminar flow over the first 25% if it can get it”. What Klingberg is doing with his sailplane is to use a bit of sanding and automotive filler on the outer surface, and then covering with Oracal vinyl (65gsm) to hide the roughness left over from sanding. Probably still not laminar quality, but it's lighter and a lot less labor than micro fill, sand, and paint.

I'm still not sure how much emphasis to place on laminar flow for a foot launched glider. One pathway is fill and sand, but that adds weight which subtracts from the wingspan. The other route is female molds, which take a lot of labor and I have to be mindful of the tendonitis in my arms. What's at stake is around two points of L/D (out of 20-25) and a shift in optimal interthermal glide speed of about 3 knots (out of 55). This is why I'm thinking about ways to reduce labor so I can just get it built and enjoy flying. 🤷

A solid foam core for the first 25% of the wing made out of the lightest XPS available is about 6.5kg (just the foam), which I see as a non-starter. Maybe that could be cut in half by hollowing out a channel through the center, plus one layer of carbon it comes out around 5kg.

I take issue with your calling this moldless - you made a simple mold to build the curved skin, even if there is no separate mold... Grin.
Slander! It doesn't count if the mold is the flight article airplane. My lawyers will be in touch. 😉
 

Riggerrob

Well-Known Member
Joined
Sep 9, 2014
Messages
1,591
Location
Canada
Perhaps you could start by hot-wire cutting a D-shaped block of foam to combine the leading edge and main spar. Then hot-wire cut a second, smaller D, but leave it inside until after the (vacuum-bagged) skins have cured. Remove the inner D-shaped foam to reduce empty weight.
 

RPM314

Well-Known Member
Joined
Mar 27, 2015
Messages
735
Location
NY, USA
Possibly. The break even point in weight between that (outer skin only) and sandwich skin will be in the neighborhood of 10mm foam thickness, which might be a ground handling issue if the inside surface is unsupported.
 
Top