Last winter I decided to build a foam core fibreglass bike that would be used as this seasons race bike. This bike is an adaptation of several European bikes I had seen pictures of, the Ross low riders, and Brahm Moen's Carbon M5. Using 3/4" yellow construction foam I constructed a 3 layer sandwich with a layer of epoxy and 6 oz regular fibreglass cloth between each of the layers. The foam was cut over size laterally and epoxied together with the cloth between, a steel headtube and front boom tube were placed into the front of the sandwich to catch the pedals. In the rear there is a piece of aluminium to catch the rear suspension mount point.
This whole package was placed between two weighted sheets of plywood to cure. Once it had set up, about 18 hours for epoxy resins, the whole thing was trimmed with a band saw to shape. Because at this point the only glass and epoxy was between the layers I was able to use a sander and file to round up the parts of the foam and to finish off the basic shape of the frame.
I then laid up over the top of the outer layer of foam three layers of 6 oz fibreglass cloth and epoxy. I have found over the years that the best way to work with fibreglass is to paint or roll the epoxy directly onto the surface you want to add cloth to and the place the dry cloth on over the wet epoxy. By hand you then smooth the cloth down and get it into the proper place and get the edges smoothed down well. If there are some areas that did not have enough epoxy to wet the dry cloth out completely you take a small brush and add extra just to the areas that need it. I have found that you can do 2 to 3 layers this way very easily and you get a very good bond with minimal epoxy. Once this was finished I allowed it to cure and then assembled it and tried it out. The people who saw it at the first meeting it came to will attest that the fibreglass alone was not stiff enough to be functional in this type of a design. At that point the bike weighed in at about 21 lbs complete, this was about the target I was looking for. To stiffen up the whole bike I doubled the front boom size to about 2" and added a single layer of 20 oz Carbon Fibre to the entire frame. This was laid up dry just like the fibreglass cloth had been, but was harder to wet out as the cloth was so heavy. Locally IASCO, in St. Louis Park, carries carbon fibre cloth in 5oz and 20oz weights, I was hesitant to spend the time to add two or three more layers to the frame so opted to add a single layer of the heavier cloth. It didn't matter much as I was already over my target weight and now I was more concerned with decreasing flexibility. This works just like it was supposed to. It still has a small amount of flex laterally but vertically is more then stiff enough. It did add about $150 to the cost just for 2 yards of cloth.
The next frame built to this design will be done with three layers of the 5oz cloth and will skip the exterior fibreglass cloth. It just adds weight and expense. This should be almost as strong as what I now have and save 6 to 7 lbs bringing me back to the target weight I was looking for. The carbon fibre cloth was a lot stiffer then fibreglass cloth because of it high weight but was still quite manageable and layed up quite easily over the frame. It also took extra wetting out after application to make sure it was completely saturated with epoxy resin. Because carbon fibre can be cut with a good pair of normal scissors, one doesn't have to buy special ones like you do for Kevlar cloth. For the strength gained this was the best solution. By the time you figure your time into the design it would have been much cheaper to just use the carbon fibre right from the start. The design is set up to try to minimize frontal area by getting the rider as low as possible and to put the rider in as aerodynamic a position as possible. The front wheel is a 17" Molton rim and tire that resides between the riders knees, while the rear is a nominal 24" behind the seat, that is used on the Terry bikes. Both of these have good high quality tires and tubes that are easily available from local sources. I went on to incorporate a centrally located swingarm and rear suspension system under the seat that also allows for a stepup compound gear.
I felt that with my weight of around 200 lbs, a body on the bike, and the small high pressure tires, I would need some give if I was to keep wheels from trashing all the time. I also wanted an easy way to get large gears without resorting to large chainrings. The compound gear set does this very effectively. The pivot point for the suspension and the compound gear set are a shortened bottom bracket shell set into the foam matrix so that a freewheel with the pawls removed could be threaded over the adjustable side bottom bracket cup that sticks out with a standard length bottom bracket axle. This was done after the original sandwich was laid up just by drilling a hole in the frame and epoxying it in place.
The rear swing arm is an aluminium H shaped structure that Mark Stonich was kind enough to TIG weld up for me. I constructed it out of 1"square .068 wall tube and it has held up well so far but I feel that in my next try I will use larger section tube in the vertical dimension to give it less flexibility and greater strength. The rear brake is mounted to the swingarm and moves with the rear wheel. I have also found that the rear tends to flex a bit when hitting a pothole while leaned over at speed. In the near future I will add a second suspension mount parallel to the first on the outsides of the swingarm, instead of the single midline mount now on the bike. This should help to reduce some of the tailwag that is a little disconcerting now.
Because of time limitations this summer I haven't been able to get racing with it yet and only have the body about half done at present. The low seating position and high stem tend to limit frontal visibility as far as avoiding potholes goes, but it is very fast and responsive so you can generally avoid them once they are seen. The 8" off the ground seating position means that your center of gravity allows you to out corner almost anything on the road, but does take a little getting used to how to balance it at first. Once that is figured out the bike is just a joy to ride.
In looking at the final product of some 150 hours of labour, I find that I will definitely do this again. It looks a little rough but is very strong and functional. It again, as I have gained my reputation in the past, took minimal measuring and jigging to come up with a serviceable bike. Total cost for the materials could be held to around $300 by doing it the right way the next time. All in all it was a lot of fun to do and rides like a dream.
Don't let carbon fibres reputation as "high tech" scare you off from trying this very versatile and easy to work with material in the future.
Dave Krafft
Institute for Plasma Research
University of Maryland, College Park
The epoxy curing is temperature sensitive, but there are so many formulations that it's hard to generalize.
There are a lot of room temperature curing epoxies. If your workspace is above 20 Degrees Celsius or so, there will be no problem. The cure time will increase the lower the temperature.
I've never had any trouble with getting a cure (The coldest I've worked with has been around 10 Celsius) - In fact all of the temperature related problems I've had have been in the other direction - if you mix too large a batch the heat from the reaction in the center of the vessel can't get out fast enough, and the reaction speeds up, basically running away
- I've had nominal 1 hour cure epoxies cure in 5 minutes because I mixed too much at once (burned my fingers, too).
Apart from boat repair and tailbox building, I spent a lot of time studying epoxies for encapsulating magnets. I also started building a bike by epoxying pieces together, but I haven't finished it (and may never), so I can't report on the suitability of epoxy in that application. Since I'm pushing epoxy as the way to go, I should share some of what I found.
Clean surfaces are very important: epoxy won't stick to greasy things. This can be very helpful - I used Crisco (hydrogenated vegetable oil) as a mold realease quite successfully - saved a lot of money, too. I've also used wax paper and candle wax to keep epoxy from sticking. Use lots of crisco (I'm confident that margerine, axle grease, etc would work fine too) - if the epoxy flows past the crisco it can displace it and clean the surface well enough to stick. If you are casting in a mold, make sure that the mold comes apart in a way that lets air in - the suction due to air pressure is large, and can make it very hard to get the item out of the mold. Molds can be sealed with caulking (like the stuff used on bathtubs), or roofing sealant - be sure to use neoprene based sealant (I tested Super Flash Neoprene brand sealant, and found it worked well - I assume other neoprene sealants would work, too).
Do not use an Asphalt (tar) based sealant. I found that the asphalt sealant was displaced by the epoxy, and there was some evidence that it disolved in the epoxy (!).
An idea I haven't tried out (yet) is to line molds or parts with thin plastic kitchen wrap (reynolds wrap/saranwrap) and just let the epoxy stick to it.
There are a lot of different fillers that can be used to thicken epoxies. I've used microballoons (to add bulk without adding weight) and thixotropic silica (Cab-O-Sil).
The microballoons are good for filling in areas where you just want bulk - you add them to the epoxy, mix well, and apply. The cured epoxy sands easily, and takes paint well (though it may soak up more than plain epoxy). The thixotropic silica thickens the epoxy so you can apply it in areas that are vertical (or upside down). There is still some tendency to sag, but much less than otherwise.
The epoxy I was using in this application (boat repair) had the consistency of warm honey when unfilled, and I could get it to about the consistency of spackle (wall plaster), or even a little thicker. Mind you, when it's thick it takes a long time to mix more filler, so it's a race against the clock to add all the filler before it cures. A lot of the time you'll want to use microballoons and silica - add the microballoons first, since they don't hamper stirring half as much as the silica does.
A lot of this stuff is covered in _The_Epoxy_Book_ available from System Three Resins.
They are a good company, very tinkerer-friendly. Reading their book saved me huge hassles when I was working on the boat. Also take a look at "The Hall of Shame" on their website - it certainly made me feel better about my misadventures.
I'll mention Fibreglast Developments again - They have much the same stuff as System Three, and are also tinkerer-friendly.
For anyone looking at production work, Fibreglast also offers seminars and courses.
It's been posted on the hpv list before, but for completeness I should mention R.Q. Riley's page on foam-sandwich construction:
This is where I got the inspiration for my tailbox.
I strongly recommend _The_Epoxy_Book_ to anyone attempting to work with epoxies.
If you are willing to look outside the tinkerer/hobbyist market, there are some terms that you'll run into that confused me when I first started looking (excuse my units):
Peel Strength: (pounds per inch)
The Epoxy is used to stick a strip of flexible material to a substrate. The strip is pulled perpendicular to the substrate - the force in pounds required to peel the strip off, devided by the width of the strip, is the peel strength. Peel strength is therefore a measure of how easy it is too peel two epoxied pieces apart.
Yield Strength: (pounds per square inch) The epoxy is cast into a rod and the ends are pulled apart. The force required to break the rod, devided by the cross sectional area, is the yield strength of the epoxy. Epoxies can have remarkably large yield strengths, but be aware that yield strength (and all epoxy properties) is sensitive to mixing the resin and hardener in exactly the right ratio.
Hardness: (Shore D, A)
Unfortunately I've misplaced my detailed notes on hardness, but here's what I can recall off the top of my head. Most epoxy hardnesses will fall on the Shore scales, either Shore A (softer materials) or Shore D (harder). The scales actually have some overlap.
Measurements are made by pressing a round nosed pin into the material being tested - the relationship between the depth of indentation and the force on the pin is translated into a number between 1 and 100.
Higher numbers are harder: 58 Shore D is about the hardness of a bowling ball. Above 90 Shore D starts to get pretty brittle. Below about 40 shore D is pretty rubbery, and begins to overlap the high end of the Shore A scale. Shore A is most useful for rubbers and foams - any epoxy whose hardness is measured on the Shore A scale is going to be lousy in structural applications. The usual epoxy from System Three or Fibreglast will have a hardness between 60 and 80 Shore D, OK for almost all velomobile applications.
Viscosity: (centipoise)
Viscosity of the uncured epoxy is only an issue if you are casting or vacuum bagging. Around 2000 centipoise is like warm honey, lower numbers are less viscous (flow more freely). Above 10,000 centipoise you're getting into something
more like treacle - avoid if possible.
The information above is mostly from memory, so double check before risking life or limb based on it.
If anything more occurs to me I 'll post it. Boatbuilding and aircraft building use many of the same materials and have many of the same constraints as velomobiles, so it's worth looking at books and websites on those subjects, too.
Try http://www.boat-links.com/boatlink.html as a starting point.
From: "Mike Nelson" <miken@greeleynet.com>
Subject: Low-tech composite construction
Date: Sat, 19 Aug 2000
Fiberglass over foam is a fun way to work, but a couple of warnings need to be brought up that I haven't seen mentioned here yet. First if you use Kevlar, the finished laminate can't be sanded or you'll end up with a balled up mess of kevlar. One work around is to put all your kevlar layers on and then a layer of 'glass on top that can be sanded smooth. Secondly kevlar is unbelievably strong in tension but nearly as bad as glass in compression. Very few structures can be built to only take tension, thus you end up with a structure only as strong as glass for roughly ten times the cost. Glass over foam does indeed work well just stay away from polyester resin, It eats styrofoam. Use epoxy instead. Also the toxicity problems from the epoxy to you appear to be accumulative, they will develop over time, be careful. Gloves or barrier cream and good ventilation are a must. Carbon fiber is at least three times stronger than fiberglass in both tension and compression and many times stiffer as well at a similar weight. That's why you see so much about it here. In short, it takes a really well designed and manufactured structure to beat the simplicity, cost and availability of the typical steel bike. Having said that I'm currently working on a carbon fiber swing arm I am hoping to trim 1 1/2 to 2 pounds off my current steel design. We'll see.
From:
"Mark Marsh" Mark@bold.demon.co.uk
Subject: RE: [trikes] Carbon Trike Design Post
Date: Fri, 19 Oct 2001
Hi Eric, Your design looks quite similar to my own carbon / kevlar trike (but prettier) so I might be able to help you avoid a few problems. Suspension, to make the rear end of the trike laterally (tortion about the long axis of the trike) stiff using your design, you would have to use quite thick chainstays which would make for a harsh ride. I made a seperate swingarm which worked extremely well (when locked out the ride is very bumpy but adding an elastomer spring makes for a plush ride with very little pedal induced bob). If you do this there are two things to consider.
1. The power side of the chain run should intersect the pivot axis, this minimises the effects of pedalling on the suspension. The pivot does not necessarily have to be a bearing, I am conidering using a composite 'hinge' constructed from a thin but wide section of kevlar that will flex vertically but be stiff tortionally for the next incarnation.
2. It may be necessary to go to a 20 inch rear wheel to maintain a good weight distribution (I have a 26 inch rear wheel and it is V easy to spin out in the wet).
I dispensed with the seperate seat and simply strengthened an existing design with two deep channels 3" either side of the centreline (which allows the chain to pass very close to the bottom of the seat, maximising ground clearance).
I would suggest that you consider using kevlar for the main body of the trike, it is as strong as carbon but has much better failure properties (it tends to start tearing itself out of the epoxy matrix before it breaks, giving a gentle failure). It's main disadvantage is that it is less stiff than carbon and harder to work. The stiffness can be compensated for in the design and although it is harder to work with, it doesn't form the horible sharp splinters that carbon is prone to whe you use a wet layup. It's also half the price!