Tricycle construction notes and resources

This isn't a normal blog post. Normal blog posts are typically rarely updated. I intend to update this fairly regularly as ideas occur to me.

Composite techniques resources

The Easy Composites YouTube channel is an excellent source of information on how to build composite structures, which I have leant on heavily and will probably continue to lean on heavily. Easy Composites are, obviously, a commercial company selling materials for composite structures in the UK; I will probably buy my materials from them because frankly their technical resources are so useful.

Forged carbon fibre and the four bar linkage knuckles

The knuckles at the pivots of the four bar linkage are complicated awkward shapes and need to be strong. Manufacturing methods I have considered include machined aluminium (can't do myself, likely expensive); and carbon fibre laid up over a 3D printed armature. Recently, however, I've watched a series of videos on 'forged carbon' — pressure moulded solid parts made from tightly packed unwoven strands of carbon in an epoxy matrix:

1. Make Forged Carbon Fibre Parts Using Compression Moulding
2. Forged Carbon Revisited: Mould Design & Strength Optimisation
3. How Strong is Forged Carbon Fibre?

This would be heavier but tidier and much stronger than parts wrapped over an armature. Stronger than needed, in all probability; but relatively easy to make.

The upper parts of the knuckles are completely unlike the lower parts. Furthermore, to allow for triangulation to prevent unfair torsion on the upper knuckle, I'm proposing to make essentially wishbone side bars, meaning there are three entirely different knuckles each side: rear upper, front upper, lower. While the right hand parts are mirrors of the left hand parts, and therefore moulds/armatures can be prepared from the same CAD files, the actual moulds or armatures are different and must be made separately. So this is quite a big deal.

Finally, each knuckle is two separate pieces, with a hinge between them. There are therefore twelve distinct parts, in two sets of six which mirror one another. Each part needs its own complete mould, which can be (and is most easily made) 3D printed. But this is a lot of 3D printing, and means I may have to add the cost of my own 3D printer to the project.

Forged carbon is stronger than aluminium for our purposes, as well as considerably lighter. 'Optimised' forged carbon — i.e. made with some continuous tows along the lines of stress — stronger still. And it will have a better finish than carbon laid up over an armature, without being greatly heavier. The Campagnolo cranks on my race bike are forged carbon fibre, and frankly, if it's good enough for Campagnolo, it's good enough for me.

A complexity here is that the parts have to incorporate through holes for the pivot axis, and, ideally, bearing seats for the pivot bearings. How we put these into a forged part I'm not certain. They could be drilled through the part after moulding has been completed, but that would require quite a complex fixture to hold the part in the precisely correct orientation during drilling. Possibly the bearing seats, at least, could be cast into the part by using a four-part mould — which would also make getting the part out of the mould easier.

Ideally the lower knuckle has to incorporate mountings for the stub axle and for the disk brake caliper, so it's a fairly large item; but I think it's within the size that I can 3D print the mould parts for.

Forged carbon fibre stub axles?

I have one genuine Cannondale lefty axle on order, and I also have three on existing Lefty forks which I can take precise measurements off. I could buy two more, but they wouldn't be cheap, or have them turned up from billet steel or aluminium, which wouldn't be cheap either.

Of I could make them out of forged carbon fibre; or I could make them as forged carbon fibre sleeves to fit over a carbon fibre tube. These would be both stronger and lighter than turned metal parts.

They need precise bearing seats, so the forged carbon would probably have to be lathe turned after forging, but that doesn't seem impossible. It would still be cheaper than turned aluminium.

Flax, balsa and bioresin

Composite structures can be made using natural fibre reinforcement, chiefly flax, instead of glass, carbon or kevlar fibres. Flax is preferred out of the natural fibres available because of its long, consistent fibres. However, it's not nearly as strong as glass, and much weaker than carbon. Although the fibre itself is lighter, because it is weaker, a usable material will need to use more layers of flax and will therefore end up heavier. It also will not produce as rigid a material. However, as a natural fibre, it will biodegrade.

Similarly to using flax instead of carbon as a biological reinforcement fibre, balsa wood can be used instead of Nomex as a sandwich core. Balsa is about three times as heavy as Nomex but is considerably cheaper, and actually the weight of either of these materials is relatively small in the context of the whole structure. Balsa is (obviously) biodegradable, whereas Nomex is definitely not. So balsa is definitely worth considering as a core material. However, fitting Nomex to a complex curved form is likely to be substantially easier than fitting balsa.

Finally, 'bioresin' — epoxy resin made from mainly plant-based feedstocks — is available and is claimed to have very similar properties and performance to petrochemical-based epoxies, but is about 20% more expensive. Whether bioresin results in less toxic waste in the environment, however, I don't actually know.

From what I've read I don't think an all flax layup would give me a strong enough, light enough structure. However, an inner laminate of flax inside an otherwise carbon layup would have substantial benefits of sound insulation — carbon shells are notoriously noisy.

Laying up on a male mould

I had been advised — and had accepted — that the only way to get an acceptably fair finished hull would be to lay up in a female mould, which would mean making a male plug, laying a female mould up over that, taking the mould off the plug, and then laying the hull up inside the mould. However, this video shows a streamlined hull laid up over a male plug, and the resulting surface seems to me entirely acceptable. Not having to make a female mould would halve the amount of laying up work that would have to be done, and halve the amount of materials used.

The plug here was made of XPS foam, whereas I'm still proposing to make my plug of wood; However I don't think that, for these purposes, this is a significant difference.

Highly adjustable ergonomic rig

At 4:58 in this video we see a highly adjustable ergonomic rig — equivalent to what I've been calling an 'ergonomic prototype' — built by Barney Townsend's team at South Bank University. I hadn't planned to build my ergonomic prototype with even nearly this degree of adjustability, but it looks as if it might be worth it!

Getting in/getting out

A conventional way of getting in and out of such vehicles is to stand astride the hull, squat, put one's hands on the gunwales, and lower one's posterior into the seat, then finally lift one's legs in. Someone — typically someone else — then lifts the upper shell and positions it on the lower shell. There are several reasons why this (probably) won't work for me on the vehicle as I have so far been conceptualising it, which are:

1. If, as I've been proposing, the windscreen/visor is already attached to one side of the hull before one gets in, one can neither stand astride the hull nor put a hand on the gunwale of the side to which the screen is attached;
2. If, as I've been proposing, the hull is essentially a monocoque and entry is only through the windscreen aperture, then the gunwales will be considerably higher than on vehicles like Beano and Woodstock;
3. I'm elderly and don't have great upper body strength or flexibility.

If I need to beg strangers for help with getting in and out, this isn't a practical vehicle.

So let's rethink. This thing is a tricycle, with a tilting hull. If one is sitting in the vehicle, opens the side-opening screen (which we'll assume is attached on the right), turns the steering to the left thus tilting the hull to the left, one ought to be able to swing both legs over the left hand gunwale. Getting up from that position might still be a struggle, and would certainly require effective parking brakes! Being able to brace on both gunwales would significantly aid in this, which means that the side opening screen, although light weight and simple, doesn't work.

A fully removable windscreen or top shell also doesn't work, because it would be large and light weight, and would be bound to blow away just when you had got settled in the vehicle and wanted to reach out to install it.

Which means there needs to be a rigid window frame for the screen, making it into a door, which is bound to add a little additional weight; and it needs to pivot either at the front, where the pivot would be in the most busy and aerodynamically sensitive part of the vehicle, or at the rear, behind the seat; I'm obviously thinking the latter. It would still be large and lightweight and problematic on windy days, but at least you couldn't lose it.

So I think getting in by standing astride (after opening the windscreeen door) and lowering one's self in probably works, but may mean that the gunwales, at least at the front of the aperture, need to be lower than I was thinking; but I think tilting the hull and swinging both legs out the same side is likely to be the best way of getting out.

Tags: Cycling, Tricycle