Learning from CAD

As I reported in my last essay on the tricycle project, I now have a CAD tool which I'm learning to use. So, learning to use it, I'm beginning to be able to visualise the tricycle, if not yet concretely, at least as an assembly I can look at and study. And doing that brings up several things.

That removable subframe

Mating surface with the hull

I wrote earlier about building the transmission for the tricycle on a removable subframe. Now that I have a model of that subframe, several things about it become apparent.

First, fitting it precisely to the fairly tight compound curves of the inside of the hull

1. will not be at all easy; and
2. cannot really be done until the hull has been moulded.

But it would be really useful to be able to build and evaluate (and if necessary modify) the final subframe on the functional prototype before committing to the very considerable expense and effort of moulding the hull.

So: if a small area of flat floor can be provided at the front of the hull, on either side of the front wheel aperture, to which the subframe can be bolted down, then I can make the subframe to be bolted down to a flat floor, and those same bolt holes will allow it to be bolted to the functional prototype (whose 'hull' I envisage to be a 50mm x 150mm wooden beam).

Can I make that area of flat floor in the hull?

Yes. There is a variant of expanded polystyrene known as 'XPS', which can be cut with hot wire, glued with epoxy, laminated over with carbon or glass fibre, and easily sanded into shape. I learned about this material while investigating how to make the aerofoil legs for the linkage. So, after the hull has been moulded, I can cut a piece out of a 50mm thick sheet of this foam and sand it to fit into the nose of the hull as a floor. I can then trace the wheel aperture onto it, cut the aperture out, and laminate it into the hull with carbon fibre.

I had thought of neat ways of embedding bolts into this foam block, but engineering friends have persuaded me that this won't be strong enough, and that through bolts from the bottom of the hull are needed. If there are to be through bolts, they should be countersunk for aerodynamic reasons, and it will probably help to mould the cones for the countersinking into the hull, because that will precisely locate the holes.

Structure

I had assumed that I would need to weld the structure of the subframe from aluminium tubing, or rather, since I lack either the skill or the equipment to weld aluminium, I would need to pay someone else to do so.

But laminating carbon fibre over a shaped armature is something I can do myself, and XPS is an ideal material for making that armature. Metal hardpoints will need to be bonded into the structure to hold at least the wheel hub and the bottom bracket, and probably to mount the subframe to those bolts through the hull. Carbon fibre can cause rapid galvanic corrosion in many metals, but fortunately I read that stainless steel and titanium are resistant to this.

So while an aluminium tube subframe remains an option, I am now leaning towards a laminated carbon fibre structure instead.

Man was born free, but is everywhere in chains

What isn't shown on these diagrams is the chains, and boy, are there a lot of them. One chain, obviously, runs forward from the chainring to the left-most sprocket on the epicyclic; this is the 'primary chain'.

'Left most sprocket'? Isn't the point of epicyclics that they only have one sprocket? Well, bicycle epicyclics are designed to fit in the hub of the wheel; and, actually, this one could be put in the hub of the wheel, now we've arranged everything else as we've arranged it. The motor could drive a second chainring on the cranks, although that would mean a slightly wider Q dimension. The main reason I want the epicyclic separate from the wheel hub is that I'm still wanting to use a one sided hub, and epicyclics are not designed for that. But if the epicyclic is away from the wheel hub, I need to transfer drive from the epicyclic to the wheel hub, and for this I need an additional sprocket mounted to the body of the epicyclic, possibly bolted to the spoke flange... but we'll come back to that.

Because we have a one sided driven wheel, and because I don't want the brake disk close to a drive chain because of the risk of lubricant contamination, the brake disk will be mounted on the end of the epicyclic — and, fortunately, modern epicyclics are available with mounts for brake disks. This means, of course, that there can't be a ratchet at either end of the chain — the 'secondary chain' — which links this second sprocket on the epicyclic to the sprocket on the driven wheel. So the epicyclic will always rotate with the wheel.

However, that's not us done with chains. We need a motor drive in the system somewhere, and that motor needs not to put parasitic drag on the system when it's not being used. So there must be a ratchet or a clutch between the motor and what it drives.

I've identified a motor made by a Chinese company called 'L Faster' which has an integral reduction gear and a ratcheted sprocket drive. It's available in either brushed or brushless variants, in (legal) 250 watt power, and either 24 volt or 36 volt versions. It's fairly inexpensive. I don't know how good it is but it would fit where I need it to fit. This isn't a final decision, but it is the motor that I've modelled in my design. It would sit above the epicyclic, and would drive the epicyclic via a third chain.

Adding sprockets to the epicyclic

Which means a third sprocket on the epicyclic. Mounting sprockets to the body of the epicyclic — which isn't designed for this — will be tricky. If the sprocket is mounted to the outside of the spoke flange, the primary and secondary chains would be running worryingly close to one another. If it's mounted to the inside of the spoke flange, then at least part of one of the spoke flanges needs to be cut away to get it on, meaning that if this design proves unsuccessful, the epicyclic cannot subsequently be repurposed in a more conventional bicycle.

In any case, we don't need to mount just one extra sprocket, we need to mount two.

The alternative that I see is to turn up, in aluminium, a sleeve which is a close fit on the body of the epicyclic, with mountings for the two sprockets on it. The sleeve, obviously, can't be slipped over the end of the epicyclic because of the spoke flanges. But it can be cut in half, the two halves assembled to the epicyclic, and the sprockets then slipped on to hold the two halves in place.

The sleeve obviously must not rotate around the body of the epicyclic; but it can be fixed by pins or screws through the spoke holes in the spoke flange. The inner diameter of each of the two sprockets needs to be larger than the spoke flange diameter, so they're quite large sprockets. But this solution seems to me to work.

Conclusion, for now

So: I seem to have a reasonably compact, workable, buildable design for the transmission/drive subsystem. It's not the most compact possible. If I put the epicyclic into the hub of the wheel, I could drive it directly off the chainset, and drive a second chainring on the chainset from the motor, which could be located under the bottom bracket. That would work, and would be both more compact and simpler. The downside is that it would definitely be impossible to replace a tyre without taking the subframe out of the hull. But it's going to be pretty awkward to do that anyway.

Share on social media

I'm no longer offering links to kleptocrat owned social media. If you're still using it, please think about whether you should be.