New threads and interesting conversations directly in your inbox. Sign up now and get a daily summary of the latest forum activities!
Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by stanislavz, Dec 3, 2019.
Tons of fabric planes look like cast plastic. Just takes time.
Theres an option if you want to get fancy...
Router/cnc corners/longerons out of foam, wrap to what you think you'll need. Recessed flats for panels (think like 1ply/3mm foam/1ply) sides/top/bottom. Will need some formers and likely braces - same idea plan it so you can cut it with automation & vacuum bag. Then assemble - flox and tape all together - bottom panels probably dzus or similar - mechanical fasteners so you can access for build and maintenance. Think RC diecrush construction but with glass/carbon over foam (ties into that black plywood discussion)
Manageable piece sizes, automated, design it so you can vacuum bag most on the bench without molds - maybe include jigging in the formers that can be cut off at final assembly.
My wild thought. Ymmv
Please complete and flight test your current design before you decide you're finished, Bill. : )
These discussions of corrugated composite panels... Why would doing this to whole panels be a good way to build a skin panel? In choosing "best" or even best compromise, I must emphasize that WEIGHT IS THE ENEMY. Pick things that make for lower total airplane weight every time you can... It will take off quicker, climb better, fly higher, maneuver more easily, carry more fuel/people/stuff - just about everything an airplane does, it does better if it weighs less and still does the jobs you set out for it.
In sheet metal you might decide that you need more stiffness in one direction and so putting any of a number of stiffeners on that run that way makes sense. In composites, I gotta say "do the analysis". Figure out how much weight you think you will save, think about the ways you will add some weight - whether you will actually save any weight in the end - and how much fuss that will add to your process. In both structural and non-structural panels, a simple cored panel is pretty darned light. For instance, a non-structural panel with 2 BID on each side, vacuum bagged on 1/8" foam panel is about 0.40 lb/ft^2 where there is foam and about 0.35 if there was no foam. Make a bunch of foam trapazoid sections and stick them on, fuss the cloth over every one of them so it fits perfect, and lose half of the foam weight, and if executed perfectly, will weigh about 0.375 lb/ft^2. Trouble is that you are now using more cloth go up and down the chamfered edges, you can never get it to lay in there perfectly which results in more resin bridging all of those gaps, and you have a bunch more opportunities to make scrap panels. By the time you are done making a bunch of these, you will be lucky to save any weight at all in exchange for a LOT more fuss in the process.
Other issues come up with it too:
Every time I make a panel that will attach to another someplace and some way, yeah, I chamfer the edge of the core and go glass to glass for an inch along the edges. This is lighter than putting a high density core around the edges, but other stuff happens. The Class A Surface usually puckers at the edge of the core - true the panel has to be filled and faired anyway, but that line of slight pucker takes extra fuss to fair properly, and now, instead of one line near the perimeter of the panel, you have a bunch of these lines everywhere. Yeah, you could spray a gel coat into the mold first to prevent this, but gelcoat weighs three times as much as dry micro.
Then the panel becomes very soft in bending where you have removed the core. The thing is almost a wet noodle compared to the cored regions. In my shop, we usually apply a ply or two of glass tape in the areas where we have omitted the core to make it less floppy there. The single layer of tape weighs more than the deleted foam.
So, why do I omit foam at all? Well, if the panel is a continuous rib or frame that is tabbed to another panel, I do not omit foam. I do omit foam at free panel edges, like an access panel or the edge of a skin where it will bond to another skin or an underlying structural member. It does make the panel heavier at that location, so I need a good reason to do it. If it is a wing skin crossing over a spar, the spar cap weight can be so much lower if the spar is 3/4 inch deeper that the un-cored section allows a much larger weight reduction. And panel edges that are not glass-to-glass will delaminate - edges must be closed out, so omitting core there is the lightest way to do it and have it remain durable.
No, If I were trying to cover the outside of an airplane with non-structural panels, I would go continuous core except at edges and where I had structures I had to be clear of. And then I would still have a heart-to-heart discussion with the design lead on how if we were doing all that outside weight anyway, why not just make it structural and skip the separate internal structures? I changed the subject again, didn't I? Grin.
Stan wants to build in composites, and no other airplane is quite right... It is a new airplane, but he wants to do it.
I will give some serious objections to a wooden Tailwind - the steel tube fuselage is light, very sturdy in an off-field landing, repairable, and fun to build. A wooden fuselage might be pretty, but the aftermath of setting down that fast landing bird in rough country is more likely to end in funerals with plywood than with steel tubes.
Then comes the argument about composites vs wood. I can selectively reinforce composites and make a very sturdy cockpit. Tough to achieve at all in wood... The Falco is sure pretty, but they do poorly when you do not put it on a runway.
I do believe that the Sportsman scheme for the fuselage is pretty darned good. Combine it with a slick composite wing and tail and some first class aero work at the intersections and cooling, and it could be a VERY capable successor to the Tailwind.
The Sportsman evolved from the GlaStar. The first design studies for the GlaStar included a composite wing. That was deemed to be too heavy, so the folding wing, removable HS/elevator, and the rudder became metal. Note the multiple constraints and trade-offs involved; kit plane (no solid cores to be cut), folding wings and removable HS/elevator for highway tow-ability, quick and easy conversion of LG from tricycle to conventional to floats, safety cage around cockpit, and reasonably STOL (Not in Super Cub category, but a Sportsman stalls, flaps down, at 42 knots, and will break ground in 400+/- feet.)
Don't get me started. If any of you have Hollmann's books, please do not use his advice as written, get some good training on the technical areas and then recognize how far off the mark the man was in much of his writings. As to his recognizing that a steel tube cockpit with monocoque elsewhere was good, North American beat him to that revelation with the AT-6 fuselage originally designed in 1935, and they most likely did not invent it either...
Flat beam, solid round, and round tube landing gear were all covered in Wittman patents and originally flown in Wittman airplanes. Cessna did not invent any of it, and Steve's patents are long run out, so we can all use these elegant landing gear.
Oh, I intend to.
I still like the scheme anyway.
Wood for wings - maybe. For fuselage i do agree and its a - no no. I have started to build ragwing stork - but, sold it after tail stage, due to understanding how unsafe it is.. And realistic, what wood is fast to build, but not cheap. All threads up here on searching for modern plywood shows it.
Tailwind or not - it is not about make it from plastic or no. Spend some time today - xflr5 with some fuselages as airfoil shape
And results are funny, but logical. Bended up tail do creates negative lift at 0 aoa. To compensate this - main wings must bu at higher aoa.. etc..
So yes, its have to have an common and correct aoa for fuselage and wind. Tailwind or Ol ironsides - its not a big case. It is - more about winning of design over tech of some kind.. But - i have driven an Alfa romeo in past time. So i have chosen nice before tech before.
But not like ch-701 or other options, where tail is bended up. And do create some draggy areas..
And if would like to be fancy, i could do just this :
This sportsman ? https://glasair-owners.com/glastar-...portsman-cage-replacement-in-cameroon-africa/
The very lightest fiberglass covering is 4oz per square yard Razorback fiberglass fabric. I did a repaint on a Maule covered with Razorback fiberglass. The glass was fine but the paint was badly cracked. Razorback is out of business. Harder to use with limited shrinkage.
Thanks for the informative, useful post (though I kinda wish it had found a home in the in the "black wood" thread ).
My observations and comments
A flat, cored, panel is very light. And if we have a truly flat surface that we'd like to cover in composites, it can be great answer. So, why consider an uncored laminate skin with built-in unidirectional stiffeners (whether they are undulating corrugations, intermittent hat sections, or intermitent straight up-and-down ribs)?
1) It could flex to make relatively tight simple curves (aka 2-D curves) without modification. In this way it is like thin plywood.
2) It could be made in large sizes and shipped in the rolled-up state. This can't be done with cored composite panels. The reduced shipping costs could make centralized production of stock sheets feasible.
3) A stock product. If every sheet is basically the same and cut as needed when the aircraft is assembled, a centralized manufacturer or a DIY homebuilder can make all the panels rapidly. This is handy where workspace is an issue and setting up for a large vacuum bag operation is inconvenient (because all the flat sheets can be made at one time and stored for later--rolled up if required).
4) No issues with panel close-outs, joggles, tapering. Without a core material, there's no threat of delamination, water intrusion into the panel, etc. Where the sheets need to overlap, the builder knocks off an inch or so of the ribs/corrugations with a straight bit in a router (riding on a smooth sheet placed over the not-to-be-cut ribs). Hit the newly rib-free back margin with a sander and it is ready to bond to an underlying sheet, a former, rib, etc.
Weight: "It depends." Obviously, like a cored panel, face lamination could be made thicker or thinner depending on the weight/toughness required tradeoff. What is acceptable in a glider or SD-1 won't be acceptable in a STOL acft that might be kept outside in the hail. On the "toughness"/"hangar rash" issue, we'd need to see if these back-side ribs provide local support to the face that is more akin to the support provided by PVC foam or softer XPS foam. As we've discussed, the stiffer/higher compression strength PVC foam apparently allows an approx 50% thinner outer skin to be acceptable (6 oz CF vs 12 oz CF). If the (closely spaced?) small ribs provide local support akin to Divinylcell, then a thinner outer skin might be acceptable, which would help offset the weigh of the up-down CF/glass in the ribs. Also, in all the areas where the back "stiffener" laminate is attached to the front laminate, that front sheet is effectively increased in thickness and (presumably) more damage resistant.
I think it may be most useful to think of an uncored but stiff-in-one-direction composite panel (with built-in small ribs/corrugations) as analogous to thin plywood for design purposes. There are a lot of great aircraft designs (and parts) that can be covered with 2-D curved panels. In the right situation, composite panels could be a good alternative to metal (and the "thousand rivets" with their two thousand matching holes and a thousand protruding rivet heads) and/or plywood (which is a great material, but does have significant challenges in availability, shipping, dimensional stability, and rot). Thin plywood is used in many fine airplanes, I hope it isn't heresy to suggest that it is possible modern composites in a sheet form might be able to replace this older sheet composite and provide some advantages.
I am asking this to myself from time to time. And where is not an direct answer..
All goes to two extremums - Cored load bearing panels. Or trusses and fabric covering. And as Billski mentioned me few times - if it is tough enough to stay alive in hangar - it will needs minimum reinforcement to become load bearing.
In a logical homebuilder medium is an composite version of ch-701 fuselage. It is floppy and not nice in real, as i have seen. In carbon it be built in a better "feel", but it is totally illogical. And yes, you can add more stiffeners. Ie - each 10 inch. But if done separately - to much mess. Done in one step - will need some of head scratching on joining it..
But done from cored plates with molded recesses (just paper tapes glued to table before molding) - will make joining panel an easier one.. Just some ud on corners in any step..
Overall it’s hard to beat an aluminum semi monocoque airplane for building for general shape, strength, and simplicity to build. It’s not real elegant to build. There is elegance in building with wood or tube steel. Marginally harder because you have to up your skills slightly, but shape has to suffer unless you are a master or you fake the outer. Composites have the shape won. Not simple to build even though they seem simple to build. Shaping the mold has some elegance, laying stuff up is just a contained mess.
The beauty of composite and rag and tube is its pretty quiet work. Aluminum is like factory work. Lots of noise. Tooling can be pretty crude for rag and tube, and the beauty is you spend less time tooling up and most time building. Works great at home. Composite is all about the tooling. Everything has to be perfect before the first drop of glue. No going back if a mistake. Complete do over. Aluminum is in the middle. Forms need to be made for some shapes and some is draping sheet over the forms.
Everyone has a natural point of entry. If that’s not what you want, things just got a bit harder. Getting something completed is hard enough without trying to reinvent the wheel. Building something as designed is a big project. Building something against its natural intent just makes it multiples harder. Now you are trying to force something that was not meant to be. It will end up as an unneeded compromise. I am a Tailwind nut. Love the wood wing. I love the simplicity of the fuselage. The only thing I wish is it was 10% bigger. I’m not in to that much work, so when I get around to it, it will be plans with Clement mods. Just about perfect as it could be.
This engineer believes that the Black Wood thread is chasing a nice concept ... but it will take more creativity to succeed at than this engineer has.
As to rolling it or bending it as you say, let's look at the limits on the concept:
ex = y/rho, where ex is elastic strain, y is distance vertically from neutral axis to outer fiber, and rho is radius of bend to neutral axis;
We can re-arrange this to be rho = y/ex;
Epoxy composites have strain at failure of about 1%, so the min radius we can roll a single cured ply of 6 oz graphite cloth (0.007" thick) is about 3.375", 2 plies is about 6.75", and three-ply about 10.125;
When you try to roll it to that radius and slip, you break it, right now, right here. To have much margin, you must roll it into a much larger radius. For instance if you use a 12" radius on the single ply, you are at about 28% of strength (and 24" for two-ply, 36" for three-ply);
With a FOS of 2.0 in the material you are allowing yourself 50% of strength, which means your load carrying capacity is all done within about 22% of the strength of the material.
Let's remember some other things about graphite:
The fibers have a negative coefficient of thermal expansion in the direction of the fibers, but a positive across the fibers - really odd behaviour, but it is real. If your graphite parts are operating at a different temperature than the temperature at which they were cured, they have internal stresses added in which further takes away from your usable strength;
Graphite does not tolerate non-uniform loading very well - any errors in fit that are forced to fit is big stresses and cause failures.
The upshot is that graphite-epoxy is great stuff, but to use its characteristics, it has to be built to shape, assembled without building in undesirable stress gradients, and you must take into account residual stresses from deformation at assembly and from thermal-dimensional interactions. You are proposing a lot of built-in stresses before it even sees operation. Good luck with processing, handling, and getting a usable build at a decent weight.
Oh, and all of the important stuff in my above post ignores any stiffeners bond along the skin. For the portion of the roll where the stiffener is on the cured plies, the radius it can be bent is close to infinite. This does not just mean that it will bend a certain radii through the region with no stiffener, then straighten out through the stiffener, then go back to curving, but it also means that the glue line holding the stiffener to the cured plies is stressed pretty strongly too.
All is not lost. If you are really clever you can cure your plies and attach your stiffeners to the plies curved in the right direction and at about the right radii to reflect the application. Then everything will be far less stressed when rolled for shipping and when applied to underlying structure. If this sounds like molding it to shape before shipping, yeah, it might be. I do not know, I am not creative enough to have solved this one yet. Oh yes, I have patents and have served on patent committees, so I do have some idea about the real nature of creativity. So far, this one eludes me. Maybe someone else will figure it out and demonstrate it to us.
In the meanwhile, built to shape looks like the way to build composite assemblies. Even Mike Patey thinks so (see video posted above).
So for "business plan" - you need three side formers - one for two sides, and two for top and bottom. And a nice flat sheet with good surface, but flexible to take shape of formers. Ant it is not as nice and easy, as it looked from the beginning.
Just a question, related to Tailwind and this "post-cured composite bend" - Do corner tubes are slightly bended outward ? Its looks like this on all pictures too. But wont it flex then ?
The braces are there so when you weld, it does not suck in. Once released, the longerons will be straight. it will be scalloped if you don't.
How much to bow out depends on a lot of things. The heat and size of the weld, diameter of the tubing, the angle of the tube to the longeron tube, etc.
Building the largest coal fired boilers in a electric power plant with a firebox of 220' x 220' x 318' tall, a single section wall panels of tubes are about 15' wide and 80' long. Have to fit the panels together with a 1/16" gap on the sides and weld the tubes together on the ends of each panel. Panels set a warp in them for not being stored or shipped correctly on railroad cars. Start at the top and start making stitch welds, but the panels can get several inches wide down the wall from the warps. We have put over 200 tons of pressure trying to bring the gap to spec's with no success , but use the shrinkage of the stitch welds to do the job. Close spacing, heavy and hot stitch welds for max shrinkage and small, wider apart and colder stitch welds for little or no shrinkage. After the wall panels are stitched welded to the 1/16" gap, then they are seal welded.
Sometimes that shrinkage can be your friend.
My boss’s dad did those types of jobs. Lots of refrigeration plants too. Moving tubes with the heat is why you and he were the professionals and I’m just a hobbyist. Interesting stuff if you have the lifetime to dive that deep. I can appreciate the work. I have most of his personal welding truck stuff including the truck. 66 Dodge with a 66 Lincoln welder. Moved the welder to a trailer.
Separate names with a comma.