Simple. - Use shapes where one could compress the fabric externally and evenly without bridges. Thus round/oval/hexagon: yes. Square/rectangular: only if required. I-beam/U-profile: no. - Hand-laminate the core, absorp quite a bit of excess resin by patting the laminate with cotton cloth. - Add PE foil with holes or peel ply and absorber. - Compress absorber. Maybe that shrink tape stuff is the ticket (never used it before), otherwise wrapping packing tape very tightly around the structure will do the job too. The tricky part is to wrap tightly and not disturb the laminate under the absorber. - Add heat. The epoxy will thin out tremendously and flow into the absorber a lot easier. Any composite professional is probably shaking his head right now, but these 'alternate methods' work fine in a 'once or twice a year we do a one-off composite part in the shed'-setting. BTW: IMHO a little resin-rich does not immediately mean very weak parts. Just a little weaker and a little heavier than optimal. Of course, one try to avoid that condition, but if there is an area which is not as good as the rest, well, don't bother. The test piece proved this; this piece was extremely resin-rich, the resin was made in two small batches (small batch = mixing ratio inaccuracies) and made out of mainly leftover pieces. It really takes effort to do worse than that. Still, the test piece performed well.
Resin rich parts will actually prove to be weak and unreliable.............which I think is something you are going to find out for yourself first hand.
I don't think I will have any resin-rich parts, but when I do first hand experience is the best way of learning I guess.
The first steps tonight. Just toying around basically. Bending the loops from 5/8" impact resistant PVC tubing. Was far harder than I guessed it would be; the pipe doesn't like to be bent in a 30mm (~1 1/4") radius. Applying heat was a delicate balance. Not enough and the PVC would not retain it's shape, too much and the pipe would wrinkle, crack open or lock the bending spring. They are copies of the originals, except that the inside dimension is slightly larger to accomodate for the composite fabric. Toying around with mounting the loops and positioning the loops. I intend to use more carbon and less aramid in these mounts because I can well use the good compressive and tensile strength of carbon and the risk of a heavy direct impact is not that high. A form like this, reinforced with steel at the subframe mounting points and tapering from loop pipe thickness down to 6mm (1/4") thickness at the subframe would provide a lot of strength and stiffness with only a few layers of carbon. I intend to use wide mounts like that on the top of the racks also. They give extra stiffness and strength. This form was made using solid 3mm (0.12") PVC sheet, heated a little with the heat gun to soften it. Because solid PVC sheet was what I had available. Not sure if I retain the solid PVC as-is; it is fairly heavy. Since I need to build up thickness using foam I might lighten the PVC by drilling holes in it before glueing the foam on, or replace it with a sheet of Depron alltogether.
Slight progress, about halfway done. I ended up tearing a piece of carbon/aramid cloth into separate fibers. Used the carbon fibers and some aramid in the length-wise direction of the pipes, and used the remaining aramid to tightly compress the lengthwise carbon/aramid fibers to the PVC pipe. Comparing the amount of cloth used and the laminate thickness vs fiber content in the datasheet of the cloth, I managed to reach close to 60% by volume of fiber around the pipes. Should do the trick, I hope.
It is interesting to note that high end bicycle frames all weigh roughly the same no matter what they are made out. Building composite structures is easy, but building composite structures that are actually lighter than metal is very difficult and requires lots of engineering and quality control work. Twin-shocker, fiberglass is very heavy material and CSM is about the worst of fiberglass on a strength/weight ration. Its purpose is to build bulk in applications like boat hulls. Despite common perception fiberglass does not often result in lightweight structures. When was the last time you saw anything that needed to be lightweight built out Chopped mat?
When I compare bicycle frame weights, I see carbon frames weighting in at about half the weight of steel, and aluminium/titanium are in between. Let's say 2kg for a steel frame, 1.5kg for an aluminium frame, 1kg for a carbon frame. Now, I am not a bicycle frame specialist, so maybe this is outdated and I am wrong. But there is another important factor to consider: these frames are made with state of the art materials and tooling, which is a huge difference from what is available to people wanting to build just one piece without the need to sell their kidneys. If a DIY guy in the shed was building a steel bicycle frame from standard sized tubes, no way he could get the weight down to 2kg and retain an acceptable stiffness. With carbon he could do a lot better. Not as good as the manufacturers with autoclaves etc., but very acceptable. The strength vs. weight ratio for even simple room temperature cured hand layups with only 35% fiber is much better than that of standard steel or even 4130. So regarding UTS or stiffness it is not that difficult to do better than steel. Same with high strength aluminium alloys, and with those alloys joining becomes a problem too. But yes, one must find out how to join the different parts of an assembly and how to arrange fibers to obtain the required strength in the direction needed. I am fairly certain that the strength and stiffness of the rack pictured above is OK. I did orient the fibers in the expected direction of force applied, made sure they were straight and not buckled, turned the aramid/carbon distribution towards carbon on the areas that are expected to be loaded in compression, and I did what I could to create a laminate with high fiber/low resin/little air content. But I am not so certain about the pipe/'mounting thing' joints, steel inserts and impact resistance. Time will tell. Twin-shocker, fiberglass is very heavy material and CSM is about the worst of fiberglass on a strength/weight ration. Its purpose is to build bulk in applications like boat hulls. Despite common perception fiberglass does not often result in lightweight structures. When was the last time you saw anything that needed to be lightweight built out Chopped mat?[/QUOTE]
I am a bit out of touch with bicycles as well, only know mountain bikes from a few years ago when pretty much everything weighed right around 3.5-4 pounds no matter what it was made of. Limiting factor was generally flex or beer canning with thin wall tube, not ultimate strength. I would disagree on the backyard builder aspect, most of the high end metal frames are basically built by guys in sheds, you buy the tubes from Reynolds and after that it is all about being a good welder/fabricator. Even old school lugged frames can be pretty light when built by someone who knows how and has access to good tube. Composite on the other hand took years for even the big boys like Trek to figure out, I saw quite a few mangled Carbon-Fiber wheels and frames in the late 90's and early 2000's. The early Y-frames where well known for punching the shock though the carbon frame, and woe unto anyone who dragged a carbon bike across the rocks clearing an obstacle, many would wrap in the chainstays in various materials to protect the carbon from this. The big difficulty with carbon is you have to arrange the fibers to the load path to see big gains, and to do that you need to figure out where the loads actually are. I bet the rack will be fine and pretty cool looking, I just doubt it will be much tougher/stronger for weight than simply using thicker wall or larger diameter pipe. The one place you may have trouble is that the pipe will bend in an impact while the carbon won't, so essentially the carbon takes all the load until it fails, then the steel starts working. There has been lots of work with composites in the boat world and results are generally a bit disappointing unless it is a mega-buck build with good engineering support. All this is not to say building things and trying new materials isn't worth while, I still have a pretty cool Carbon-Kevlar foam core skateboard I built when I was a kid. More an argument against expecting huge gains from exotic material.
You have yet to convince me that the "average guy" is going to be able to build a CF structure with a strength-to-weight ratio better than that of steel, in his garage. More importantly, you seem to be focusing on strength, but MANY applications are stiffness limited, not strength. The stiffness of woven CF is not all that great, necessarily.
I would imagine that bonding ready made carbon tubes into aluminum or steel lugs, would result in a pretty effective bicycle frame that was strong as well as light. However bearing in mind the fact that making the tooling required for compression or vacuum moulded frame parts from scratch, is probably well beyond anyone without many years of composite experience, I doubt very much its viable to build from scratch.
But when welding you use special tubes which are very thin-walled in the middle and gain wall thickness near the joints. These are not generic tubes. Lugged frames do need lugs that accurately match the tubing otherwise the brazed joint is not that strong. One could make these in the backyard, but it does involve lathes and mills. If you don't use the special tubing, building a 2kg steel frame would be pretty hard. Of course. These guys live on the edge and use as little material as they think they can get away with. When living on the absolute edge, it is easy to tip over. That is an undesirable property of thin high carbon laminate indeed. The reason why I wanted to implement more aramid fibre than I actually did. With some structures common sense helps. For example the loop of the luggage rack: luggage pulls the entire thing toward ground due to it's weight and wants to bend it inside. Vibration and rear suspenion action worsen this and creates forces, but these are mainly the opposite so the same fiber direction would do. Bending of the tubes themselves is not such an issue; after all the loop is 'braced' bij the luggage. The 'bending inside' of the loop and the torsion exerted at the mointing points is the biggest problem, so that is why I incorporated extra bundles of carbon fibre on the 'outside' (plane where the luggage sits onto the rack' and 'inside' (opposite side) of the loop. I also ran a lot of extra fiber to the spots where the mounting points meet the loop to cope with the local high stress caused by torsion. Forget about the cool looking; I don't think I can lay up another layer of carbon nice enough to serve as a visible finish. I will try, but if it fails I will just paint the entire thing black. Tougher for weight, well, the rack as pictured above weights 360 grams/12.7oz. This includes the fairly huge steel inserts around the mouning points. 360 grams with steel would mean very thin wall thicknesses, and the structure would crumble like a beer can when even looking at it. This is my biggest fear. A steel rack would bend when overstressed, essentially taking load off the structure. Carbon/aramid composite would not; it would bend a little and then break. I am not saying I can do that. I did some destructive testing on sample pieces of laminate I prepared, and it took way more effort to destruct them than a similar steel piece. Only compressive strength and impact resistance. It seems you cant have it all in one fiber. Of course not. Maximum strength/stiffness is only at the 0 and 90 degrees direction, and then there is the problem that the fibers are not entirely straight due to the weaving. But that is where unidirectional tapes and bundles of sereval thousand fibers comes in, as well as slightly prestressing the fibers to make sure that they are absolutely straight and won't buckle during resin application/compression of the laminate.
Because they have clever design teams and big computers to get the designs just right. This: ...is what is special about steel, and it is very very special. It makes lots of things possible and economical which would not otherwise be so.
You seem to be readily (and often) switching between "what is feasible for your average guy with composites" and "what is theoretically possible with composites." Of course composites can build a structure that is stronger/ stiffer, and lighter than classic metals; I have never denied that. That is why it has become the material of choice for next-generation airframes, jet engine parts, etc. However, don't loose sight of the fact that giant companies like Airbus and Boeing have spend hundreds of millions, if not billions, of dollars developing both the part designs and the manufacturing technology to be able to make some relatively simple parts (u-channels, skins, etc) out of unidirectional composite tape. I know this because I design and build those machines. However, NONE of this is transferable to the "backyard mechanic," which is what you were talking about, and what I commented on, before you changed the context.
I agree 100% with that.............pre-preg carbon parts using proper tooling and production methods are very very good, but simply are not feasible for someone to do at home. I get the feeling that a lot of people think that top end composite work is a bucket and brush affair, and will only find out this isnt the case, when parts fail...................
I am an engineer, so I tend to figure out what stuff is all about before doing something with it. But indeed, not much practical experience, and especially not in the failure department. But since I also intend to replace the subframe of the SV Adventure build with a glass/aramid/epoxy composite fuel tank, I am VERY eager to gain some experience with this stuff, especially in a stressed situation. This rack will gain me some insight at least. I do not intend to use composites there obtain the lightest structure possible. I want to use them because I can shape them in a way I cannot do with metal. Build a plug using foam by adding some and sanding it down again until I have the shape I want, yes, I can do that. Hammering some metal in shape, welding it up and make it look acceptable, no way. I still don't think so. But we shall see when it's finished and in service. I am not done yet. Progress will be slow also; I have another more urgent job (at least the misses thinks it is) going on; building roof carriers for the car. Could buy them, but they are on backorder for my model, cannot find them used for my model, blabla. That is something I do not want to fail, so I will build it out of steel. The unidirectional- fiber part of my previous post had nothing to do with what is theoretically possible, it is what I actually did. As is slightly prestressing them to keep them straight. But since you really seems to know what you are talking about: what makes a carefully manufactured homebuild part much weaker than a professional part, possibly even weaker than steel? This is not [size=-2](completely)[/size] stubborn-ness, but interest. The vendor of the composite materials I use (R&G Faserverbundwerkstoffe) also provides a simple calculation tool for their fibers/fabrics and epoxy systems. Here is a quick example with 4 layers of 160g/m2 woven carbon cloth, laid down in 0/45 degrees, using 35 volume-% of fiber, and their 'weakest' epoxy resin system requiring post-cure for maximum mechanical properties and higher Tg. Stiffness is about half that of aluminium.... :eek1 (BTW: CSM+polyester is even less stiff than pinewood along the grain) With a laminate stackup which matches what I did using higher fiber content, stiffness along the grain/length of pipe improves up to about 75 GPa, or only slightly better than aluminium, or 1/3rd to 1/2 that of steel. This does prove your point that woven cloth will not produce very stiff laminates. But the low density and freedom of shape allows large diameter structures (such as what I did around the mounting points), fair wall thicknesses, and still a low final part weight. Right? Or wrong?
If you want to make tanks, then forget about costly difficult to use materials, make tooling out of CSM, and try to come to terms with basic RTM techniques using CSM/poly. When you are confident with RTM using cheaper materials, try it using carbon/kevlar/epoxy.