Making a Wing Spar
 

Here is a quote from the Wikipedia describing a wing spar:
"In a fixed-wing aircraft, the spar is often the main structural member of the wing, running spanwise at right angles (or thereabouts depending on wing sweep) to the fuselage. The spar carries flight loads and the weight of the wings while on the ground. Other structural and forming members such as ribs may be attached to the spar or spars, with stressed skin construction also sharing the loads where it is used. There may be more than one spar in a wing or none at all. However, where a single spar carries the majority of the forces on it, it is known as the main spar."
https://en.wikipedia.org/wiki/Spar_(aeronautics)

I will be making a main wing spar that is to be incorporated in to a foam core glider wing with a span at about 2,4m or approx 95inches, the wing is to be made in two halves with a single center wing joiner construction for easy transportation. I am going to use some methods and materials that might not be applicable to the common modeler, but what I hope to do is to show some basic principals in composite processing/construction and hopefully it could be interesting as well.
I will try to explain what I am doing and why I'm doing it as I progress but if there should be anything that seems unclear - no honest question is silly, quite on the contrary. Also if you just want to comment that you would do things differently and perhaps want to explain your experiences or theories that would certainly be welcome too.

First to make is the sparcaps, I am using a mold to make them and it's a quite simple mold. It can be made from MDF plates (typically wall and ceiling plates) or similar, then covered with tape to make a mold-able surface and waxed.
Here is two examples; one quick, cheap and simple and one more thoroughly made, they produce the exact same result.

Attachment 2212718Attachment 2212719

I have made some carriers for the lengths of carbon tow, it will be become quite a few lengths when all is ready and some order in the chaos is necessary. I have limited space so I need to have these at another table and be able to carry them on to my workbench when I actually need them. The carriers is thin slices of foam that is covered with protective plastic to ensure a smooth and clean surface.

Before I lay them down on to the carrier I try to unwind and spread the tow a little so that it is easier to lay it in to the mold nice and flat and so that the fibers ends up being as straight as possible. There will be 5 of these lengths that runs all the way from root to tip in each sparcap, on top of that there will be one additional tow each 10 centimeters or about every 4 inches so that the sparcap gets gradually thicker towards the root. These are the bottom sparcaps, the top sparcaps is about 7,5 centimeters or 3 inches between the additional tows so the top sparcaps becomes slightly thicker towards the root than the bottom ones.

Attachment 2212771Attachment 2212772Attachment 2212773

To get the tow compressed down in to the mold I have made some slices of high density foam, they act as sort of a piston and make sure that the surface of the tow becomes flat and consistent. I adjust them so that they have a very slight press-fit with a layer of peelply (rip-off fabric) around it, if they are too loose the tow will run up along the sides of the mold and if they are too tight they might scrape away the wax and make the epoxy stick to the surface. Last I punch vent holes in them to allow the excess epoxy to escape from the tow while it is being compressed - this is essential, I use a rather large needle and I sharpen it with a diamond disk to prevent it from tearing the foam.

Attachment 2212905Attachment 2212906Attachment 2212907Attachment 2217201

I have plenty of epoxy in the mold before I add the tows, then I apply a new coat of epoxy on top of every tow length. With the tows spread like I have it is easily saturated but if it is not practical to spread it (this may differ from different manufacturers) then the epoxy might need to be worked in to the fibers with the brush to properly saturate it, you can also make yourself a tow saturater device where the tow passes through a cup of epoxy and between some guides making it automatically fully and properly saturated directly from the roll.

The mold is 7,5mm tall and it is almost filled at the root-end when I am done, as a reference 1/4" is 6,35mm so it's just a little more than that. On top of the foam pieces I have a breather/absorber to make sure the epoxy gets out and on top of that I have a floor board piece to make sure the foam pieces gets pushed down straight and flat, the plate on top (floorboard piece) also expands the vacuum area and increases the compression of the tow. The vacuum is applied gently at first just so that the bag closes, then it's possible to align things a bit. When everything is nicely aligned and set I apply max vacuum.

Attachment 2213148Attachment 2213149Attachment 2213150Attachment 2213151Attachment 2213152Attachment 2213153Attachment 2213157

The epoxy I am using for this have an extremely long curing time so it will stay in the mold for about a week or so.

It have been in the vacuum bag for 48hours now and the last 24hours it have been between 35 to 40 degrees Celsius in my workshop so I have taken it out of the bag and I will just let it be for some more time before I pull out the sparcaps. The vacuum is only strictly needed until the epoxy have become plastic for perhaps 6 to 12 hours depending on the epoxy, the rest of the time I usually have the vacuum applied for the workpiece to be kept in place in the mold or perhaps on a glass plate so that it can not twist or move while the epoxy fully hardens.

Even though I have done this quite a few times there is still things that can go wrong, the peelply around the foam pieces that compresses the tow was a little too widely cut as you can see on the pictures in the last post the edges rises above the foam pieces and bridging could occur. Bridging in this context is that the excessive peelply edges might get trapped between the bag and the mold when the vacuum is applied in a way so that it cannot move along with the foam pieces down in to the mold, the result could be something like a tilted or curved top surface of the sparcaps. It is always exiting to see how it turns out when it's time to pull the part, it just never gets boring.

Attachment 2213304Attachment 2213305

About the shear web I have already made what I intended to use for this spar but it is essential and shows better what composite processing is all about than my rather special and improvised method of making sparcaps, so I thought I'd make a material plate that would be suitable for a shearweb. The shearweb needs to encounter two forces; compression forces and shear forces - to prevent the spar from bending. Vertical grain balsa is hard to beat for it's light weight and compression strength compared to just about anything, and it is relatively easy to get hold of and not too expensive so that is what I'll be using. I take one sheet (100x10cm) and I cut it in 5 parts to make a material plate of 50x20cm, I only cut it and see that the edges fit reasonably together - no gluing, no sanding, no nothing else.

The direction of the wood grain should be vertical between the sparcaps to make the most of the balsas exceptional qualities, the spar will eventually be wrapped in fiberglass at 45/45* so this takes care of the shear forces and at the same time prevents the sparcaps from lifting from the shearweb at high loads. I am applying a light weight layer of fiberglass on each side of the balsa that will hold it together when cutting it in to narrow pieces, from a spar structure point of view it would be beneficial to the spar construction if the fiberglass is at 45/45*, however from a practical point of view a layer of 0/90* would make sure the narrow lengths of vertical grain balsa remains nice and straight after it is cut and it might make it easier to handle.

I am using one layer of 55g fiberglass at 45/45* and one layer of 49g at 0/90* on each side of the balsa, I hope some of the readers would perhaps be so kind to help me translate these fabric weights approximately in to imperial measurements - that becomes just a little too complicated for my simple mind to grasp at the moment.

When I have saturated the first side of the balsa and fiberglass with epoxy I have a perforated self releasing plastic film on top of it, then I turn it around and lay it on to the absorber/glass plate. The absorber/breather could very well have been a couple of layers of paper towels, I am using some suitable polishing material/cloth that have the correct width to my material plate and it makes a nice and even layer. What this does is to compress the fiberglass fabric on to the balsa and at the same time squeezes and pulls out any air bubbles and excessive epoxy, the result is a very light weight and very stiff, straight and strong material plate. It is difficult to feel the difference in weight with your hands between the original balsa sheet and the laminated one, but the strength and stiffness is something else indeed.

Attachment 2213467Attachment 2213468Attachment 2213469Attachment 2213471Attachment 2213472Attachment 2213475Attachment 2213476Attachment 2213478Attachment 2213479Attachment 2213480Attachment 2213481Attachment 2213483Attachment 2213484

...

The shearweb material is now ready to be cut in narrow lengths to make the spar core, with a thin layer of fiberglass like this it is easy to cut with a break-blade knife or similar.

Attachment 2214070Attachment 2214071Attachment 2214072


The sparcaps came out alright as well, they have now a thickness at about 2,45mm at the root or approximately 1/10" and about 0,7mm at the tip or 1/36".

Attachment 2214076Attachment 2214077Attachment 2214078Attachment 2214079Attachment 2214080Attachment 2214081Attachment 2214082Attachment 2214083Attachment 2214084

Now I need to make the spar joiner.
I will make it solid in several vertically aligned pieces and mount them together, this way I don't risk that a delamination of the parts reduce it's actual strength. I am using the same method as with the sparcaps but with a slightly different procedure, here I have saturated the carbon tow before I lay it in to the mold. Also I don't use the peelply around the foam piece on top of the carbon, the top surface (actually the side of the joiner piece) needs to be milled absolutely flat afterwards anyway so the foam is simply permanently glued to the carbon surface.

Attachment 2214549Attachment 2214550Attachment 2214551Attachment 2214552Attachment 2214553Attachment 2214554Attachment 2214555Attachment 2214556Attachment 2214557Attachment 2214558

I have milled the top (actually the side) of the spar joiner pieces so that they are absolutely flat and consistent in thickness, they have now a thickness at about 1/8". A note of warning about handling carbon fiber parts like this; the edges are often razor sharp and the fibers can bury them self deep in to your hands so no quick and careless movements along the edges, using some kind of gloves might be a good idea.

Attachment 2215142Attachment 2215143Attachment 2215144Attachment 2215145Attachment 2215146

Now I am basically ready with all the materials and parts for my wing spar and I will take a little break.
Next will be to sand the surface of the sparcaps to remove the mold edges, cut and make ready the shearweb pieces, mount together and make ready the spar joiner, and make the joiner box/pockets.

I thought it would be good to do some explanation and fill in some blind spots now that I have had a little break and have done some other things, leaning back and getting a little distance to it makes it easier to view it all with fresh eyes.

About the first thing I did when making the sparcaps I said that I was using epoxy with extremely slow curing time and this is because it actually takes some time do this, I spent about two hours from starting to mix epoxy till it was in the bag. This does not mean that you cannot do this with a regular laminating epoxy and yet get a high quality result, it will just end up a little thicker and not as extremely compressed. There is a couple of things you can do to slow down the curing of the epoxy a bit; open the door (or window) to get the temperature in the workshop down a degree or two if possible, have the epoxy bottles in cold water perhaps (it will make the epoxy thicker but it will also slow down the curing process), when mixing the epoxy mix it in reasonable small batches so that the epoxy is as fresh as possible at all times. Of course it pays off to do everything you can do of preparations before you start to mix the epoxy - make ready everything you can think of that is to be used or that will go in to the bag before you start.

About using fabric at a diagonal angle (-45/+45*) it is important to get the direction of the fibers the same at the top and bottom of the workpiece if it is to become and remain straight and flat, this might be a sandwich structure material plate like I have shown or perhaps a foam core of some sort with fiberglass at both sides. However if it's something like a cowling, a fuselage or a boat hull that have a more self stabilizing shape this effect might not be that prominent and there would perhaps seem to be little or no real difference if you consider this or not. Anyway, the reason is that most fiberglass fabric is not perfectly balanced, it means that there is perhaps 25 fibers (just to say a random figure) in each tow at the length of the roll and perhaps 24 fibers in each tow at the width of the roll. The epoxy contracts by a small amount when it cures and this will cause the fabric to contract slightly differently in the two directions, so you want the direction that contracts the most (or least) to be parallel at the top and bottom (or both sides) of your workpiece.

About making the spar joiner this would typically be made in a complete piston type of mold with clamps or bolts to compress it rather than using vacuum like I have done, vacuum is most beneficial with larger surface areas and a spar joiner like this does not represent a large surface area so you would get a greater pressure on it using clamps. If I was to make my joiner mold again I would make it like the sparcap mold with two molds in one and use a board on top of the foam pieces to increase the area and pressure like I did with the sparcaps, the pressure on the sparcaps have been probably about twice the amount of pressure compared to the joiner pieces because of the increased area that the floor board represents. Anyway, I like to do it this way because it is a lot simpler and less messy than using a complete clamped mold, also those joiner pieces that goes together to make one solid joiner is likely to be just a little stronger than it actually needs to be even with the moderate compression so I think this will be good enough.

To make a simple example on the pressure involved when using vacuum; at -0,8 bar or -800 mBar of vacuum (23,6 inHg or "inches of mercury") the total pressure on one square meter would equal 8,0 metric tons (using the weight of the earths atmosphere). So if you look at my shearweb plate that is 1/10 of a square meter, that makes a total pressure of 800kg or 1760 pounds lbs on that balsa sheet. If you have a small car that would do it too.

About the tool that I'm using - the vacuum pump; the easiest kind of vacuum tool to use is something like a laboratory device, this is typically a relatively small size and quiet running unit with a built in cooling fan and that is designed to be running continuously. Unfortunately these are often surprisingly expensive, there are other alternatives like a vacuum system with a tank and a vacuum switch that works similar to a typical garage compressor but you might have to make it yourself based on plans and instructions and it would mean some effort and time needs to be put into it before you can actually use it. I could make a thread about making an affordable and easy to make digitally controlled system using one or two 12volt mini pumps and a programmable Arduino board at a later time if there should be any interest for it. However, if you have the budget for it I would recommend a device that is designed to run continuously and that is capable of pulling something like 23 inHg or more, this means that you can just plug it in to the wall outlet and start processing immediately.

Finally I would like to say to those of you that gets an unpleasant feeling by thinking all of this is necessary to make anything in fiberglass for your RC hobby; it is not necessary with a vacuum setup, it just makes it easier to achieve a really good result in some cases and it opens up some new opportunities as well.

I am finally back at my project and I have made ready the spar joiner, now I can make the joiner box/pockets that goes outside it and that will become a part of the wing spar root.

Attachment 2217206Attachment 2217207Attachment 2217208Attachment 2217209Attachment 2217210Attachment 2217211

The spar joiner pockets are often made by wrapping fiberglass or carbon fiber around the joiner, I am however doing it differently because if you consider the joiner a long and narrow piece and that the epoxy contracts when it cures you are begging for trouble. I will be making this in several steps rather than doing it in one go so that I don't risk that the pockets get stuck on to the joiner and becomes perhaps impossible to remove.

First I have a layer of very thin and flexible food packing foil on to my work surface, on top of that I have my layers of fiber reinforcements. The first layer that becomes the inside of the pocket is a light weight layer of Kevlar/Aramid (61g/m2) at 45/45*, this is a safety feature and not strictly necessary. It becomes sort of a protective lining between the joiner and the outside carbon fiber in the pocket and it will help preventing the stiff carbon fibers from breaking at the relatively sharp corners both when it is being processed and in case of shock like a hard landing. Next I have a layer of 160g carbon at 45/45*, then I have two layers of 140g unidirectional carbon at 90* (vertical fiber orientation), and at last a second layer of 160g carbon at 45/45*. The food packing foil stays on to the inside of the laminate (towards the joiner) so that the epoxy never really gets in direct contact with the joiner at all.

Attachment 2217220Attachment 2217221Attachment 2217222Attachment 2217224Attachment 2217225Attachment 2217226

The thin food packing foil could potentially trap air inside between them and prevent the laminate to be properly compressed around the joiner, it would be beneficial to have a piece of thin paper or similar just outside the ends of the joiner that goes past the edges of the food packing foil so that the air have a secure way to escape at the ends of the joiner. I thought of this after it was in the bag of course but it seems the result was still pretty good anyway, I was prepared to do it over again a second time but I will be using this one.

Attachment 2217229Attachment 2217230Attachment 2217231

I have joined the pocket parts. I have used epoxy glue with some fine milled carbon fibers mixed in to it, the spar joiner needs to be properly waxed before doing this. The trick is to let the epoxy cure to soft plastic and then pull it off, if you wait too long letting the epoxy get fully hardened it might become more difficult. I have used something called 10 minute epoxy but it's actually quite slow curing, I let it cure for about 1,5 hours and at this time it have become plastic but it is still slightly soft and flexible. Excess epoxy outside the pocket can be scraped off, I have made a hole at the end of the pocket and I use a small hammer and a small bolt and carefully knock it loose and pull it off, it might help to try and twist and bend it a little first and you might hear some slight cracking sounds when doing that. When both pockets are made I put them on to the joiner again and let the epoxy fully cure, even though the fit is quite tight indeed I have actually put it in the food saver vacuum bag to make sure that it is forced in to shape when it fully hardens.

A food saver vacuum machine like this is actually splendid at composite processing and quite affordable, they typically pulls about -0,8bar (23,6inHg) at full effect and perhaps -0,6bar (17,7inHg) with the anti crush setting. I have opened up a couple and no matter how it looks on the outside or what the cost is it's the same 550 motor pump inside, it's a really simple pump unit but it works. The vacuum bag roll plastic is a little stiff so it's probably best for work pieces without too much compound shapes and such, also a closed tube bag means you need to be able to move the workpiece with the peelply and absorber layers in to the relatively narrow bag. I like to use it for some things (like this) because it is extremely simple and quick to use, also the bag is permanently sealed so there is no connections attached and you can move the workpiece to a shelf and out of the way.

When the pockets are fully hardened I will grind away the joint edges and reinforce it a little, finally I will adjust the length of the spar joiner and the pockets and I might fine tune the fit so that the pockets meets properly at the center.

Attachment 2217662Attachment 2217663Attachment 2217665Attachment 2217667Attachment 2217669Attachment 2217670Attachment 2217671Attachment 2217672Attachment 2217673Attachment 2217677Attachment 2217678

I have finished the spar joiner pockets...

Attachment 2217896Attachment 2217897Attachment 2217898Attachment 2217899Attachment 2217900Attachment 2217901Attachment 2217902Attachment 2217903Attachment 2217904Attachment 2217905Attachment 2217906Attachment 2217907Attachment 2217908Attachment 2217909

I have made ready the shearweb. I have sanded the surface of the parts lightly with a bar sander to make them reasonable flat and smooth without taking away too much material and then washed with acetone, I have just used 5 minute epoxy to glue the pieces together. I try to apply a very thin coat on both sides, when using the thick epoxy glue it very quickly gains excessive weight. When mounting the spar core to the sparcaps I have used my slower curing 10 minute epoxy glue mixed with some fine milled fibers, this bond is more important and it needs a little filler.

Attachment 2218469Attachment 2218470Attachment 2218471Attachment 2218472Attachment 2218473Attachment 2218474Attachment 2218475Attachment 2218476Attachment 2218477

I have built together my wing spar, now all that's left is to wrap it.

Attachment 2218774Attachment 2218775Attachment 2218776Attachment 2218778Attachment 2218779Attachment 2218782

So this is my wing spar basically finished.

I have not wrapped it yet and this is because I thought I wanted to install a small joiner pocket at the tip as well so that I can experiment with different wing tips like vertical winglets or perhaps a slightly larger tip panel for those calm days at the slope etc. However I will set this aside now because I have other things to tend to and I don't know when I will be able to get back to it, so I thought I'd make this sort of a finish line for now.

I will make the wrapping of the spar in the same way as I made the joiner pockets using fabric at 45/45* where one layer goes all the way from root to tip, one layer that covers the root to about 1/2 of the spar length and a third layer that covers about 1/4 of the wing spar from the root. This way the shear web becomes slightly stronger as it progresses towards the root.

Other than that I have made this wing spar as strong as I could because the glider I want to make will be intended as a slope racer / dynamic soaring plane, it will likely have something like 1kg lead ballast in the fuselage on a good day and it might see some pretty violent turbulence and hard maneuvers at high speed. If it was made for something more like a thermal or electric glider it could be made much more lightweight using the same basic structure. If my focus was to make it lightweight rather than very strong I would make the sparcaps with the same thickness at the tip but I would have greater distance between the shorter tows to make the tapered thickness slightly less pronounced, I would use only one layer of 50g fiberglass at 0/90* to bind the vertical grain balsa shearweb together, and I would use micro balloons instead of solid milled fibers as a filler when mounting the parts together. The only thing that really needs to be carbon fiber is the spar caps and the spar joiner that is made from tow.

Attachment 2218992Attachment 2218993Attachment 2218994

Hi ibuild,
​​​​​Thanks for this very informative thread,
i am making replacement wings and tailplane for a 4m Fox glider.
The original wing and tailplane were badly damaged when we received this model. The wings look like they have suffered from flutter. Not surprised after seeing the original construction.
The original mainspars have a bluefoam verticle web with 50g glass either side. There was no spar caps and the skins look like 50g glass 1mm balsa 50g glass with a carbon triangle at the wing root. The wing socket does have some plywood around it

I have 3d printed plugs for the wings and tailplane and made the molds. For the layup of the wing I was thinking of going with 25g glass, 80g glass at 45 degrees, 1.2 herex and 50g glass at 90 degrees. No herex at spar positions. Peel ply for the live hinges.
Would you use the same spar construction descibed above for a hollow composite wing?
Can spread tow carbon tape be used for spar caps? If not what what weight of carbon tow do you recomend?
Regards

Arthur