The Fool on the Hill: Rethinking the tricycle drive train

It's almost a year since I last wrote anything about the tricycle project, but I haven't fortgotten it. Today, I've been rereading my posts on the design, and especially this one. I've been looking again at this picture, of Seventy Seven's removable front subframe which carries its front wheel, steering, and transmission systems; and I've been thinking. This will be a short post documenting my thoughts.

Firstly, the removable subframe is a win: it makes it so much easier to work on, and to replace if necessary. The downside to that is that there must be a weight and strength cost to having it removable.

Secondly, Seventy Seven's removable subframe does something I do not need to do. It steers. Take the steering out of it, and it becomes much simpler.

Thirdly, it has three axles: the bottom bracket axle, the cross-shaft axle, and the wheel axle. Each of these is a potential location for either a motor or a gearbox. In practice, Seventy Seven places the 'gearbox' (in this case a conventional derailleur) on the cross shaft. Of course, there's some additional mechanical drag in having a cross shaft. But if it was worth having on a machine designed to break speeg records, not very much.

Punctures

But there's one thing that my vehicle has to do, that vehicles designed specifically for racing or speed record attempts do not have to do: my vehicle has to allow punctures to the front wheel to be easily fixed, by the rider unassisted, at the road side. A removable front subframe, while desirable in itself, does not help with that.

In my post Thoughts on Bent Cycles, I wrote

The best solution I've thought of so far is for the vehicle to have a front fork (which it doesn't actually need except for servicing, since dropouts could be built into the sides of the well) which pivots around the bottom bracket axis. If you lift the front of the vehicle, the fork will naturally pivot down, bringing the wheel with it. There needs to be some simple mechanism to lock it in the down position.

The wheel can then be removed and worked on like a conventional bike wheel, and when it's reinstalled, the fork can be unlocked and kicked backwards, and the front of the vehicle lowered onto it. It probably needs some sort of elastomer stop to rest on when in the working position. Other parts of the transmission can also be serviced with the fork in this position.

This still seems to me to be the best solution.

But going beyond that, there is merit in the three-axle solution: hear me out. If we have a primary chain on the right hand side of the transmission driving a a cross shaft to which a motor can be attached, driving a secondary chain on the left hand side, driving an epicyclic gearbox in the hub, then the Q factor of the whole assembly can be kept very low. As I've previously written, offsetting the epicyclic and dishing the wheel could make the usable Q factor even lower, since the bulge would be below the sweep of the rider's knees.

Another thought that is incompatible with that one is that a 'lefty' style hub, with only one left hand fork leg, could make getting the tyre off and on again even simpler, but it would mean you'd need to put the motor in the bottom bracket, and all commercially available bottom bracket motor systems I'm familiar with have huge Q factors. It's still worth thinking about, but is not my present proposed solution.

Handedness

Conventional bicycles all have the drive train on the right, and consequently all bicycle transmission components are designed to have the drive train on the right. Because bicycle drive trains include ratchets, you can't easily swap this round. In particular, running a derailleur on the left hand side of the wheel isn't going to work, except by making a custom mirror image mechanism, which would be absurdly expensive.

Similarly, driving a commercially available epicyclic from the left isn't going to work, and making a custom epicyclic would be even more expensive.

A derailleur on a single sided hub can be made to work — Mike Burrows highly regarded Windchettah design has this feature — but the support for the hub has to be on the right.

Well

A tantalising idea which I've been struggling with for a while is having a permanently bonded front wheel well forming part of the lower hull, and thereby considerably stiffening it, while having the subframe bolted only to the top of it. The problem is that the bottom bracket axle and the wheel axle are both longer than the desirable Q factor. If the subframe was designed to be removed through the bottom of the vehicle, the bottom bracket would have to be disassembled before the subframe could be removed, which... would kind of spoil the point. Conversely, if the subframe was designed to be removed upwards, the wheel axle would have to come up through the wheel well, which would mean that the wheel well would have to be wider than the full length of the axle... which removes the Q factor benefit.

I've been trying to think of ways of extracting the subframe forwards, but I think that would mean excessively weakening the front of the hull.

Or one could do without a wheel well altogether, but this would have a severe aerodynamic cost as well as letting road spray into the hull in wet weather, which doesn't appeal. A compromise which occurs to me is to have a 'skirt' made of a stiff tarpaulin fabric, possibly permanently fixed to the inside of the hull and fastening to the subframe with the sort of turnbutton fasteners that were used for the hoods of classic sports cars. That's at least worth trying.

Packaging, revisited

We can't really move the position of the bottom bracket, and consequently of the chainwheel, in the vehicle. We can't move the position of the front wheel much, either, unless we make it smaller, in which case the ride will be harsher. But we can move the crossshaft, and we can move the motor.

There's a lot of space in the nose of the hull which has to be there for aerodynamic reasons but isn't actually doing much. Unlike Woodstock, Beano, and Seventy Seven, the front wheel does not move relative to the hull, so drive chain tension can't bias the steering. So if I have a cross shaft — if — it can go in the nose, below and in front of the bottom bracket. This crossshaft could incorporate an epicyclic gearbox, or a motor. Unless it's far forward, it can't be very wide, however: whatever's on it has to fit between the rider's toes. In any case, the chain running back from it to the wheel hub has to fit between the inner sides of the pedal cranks.

A Bafang 250 watt bottom-bracket drive motor is 85 mm wide, which I think will fit between the rider's toes. This is designed to have a gear connection to a gearwheel fixed to the chainwheel, but that gearwheel is pretty large diameter (I think over 100mm, but I will measure one tomorrow), and that gear wheel could be bodged to drive any rotating thing, within reason.

Commercial epicyclic gearboxes start at about 147mm wide at the axle, which won't fit — but the axle can probably be kept outside the swept volume of the rider's feet. I am storngly of the opinion that a custom made gearbox would be massively out of the financial scope of this project!

So, let us suppose:

1. We have a short primary drive chain running forward to a cross-shaft in front of the swept volume of the riders feet, driving that epicyclic through its normal sprocket;
2. We take a Bafang motor, and adapt its chainwheel gear so that it will bolt onto the spoke flange on the left hand side of the epicyclic;
   1. obviously, this means making a custom mount for the motor, but that's a detail;
   2. we'd also have to make the motor run backwards, which might be slightly harder;
3. We bolt a chain sprocket to the right hand spoke flange of the epicyclic, and run a secondary drive chain from that sprocket, under the bottom bracket bearing, to a sprocket on the right hand side of the front wheel;

That's believable, and it's believable while keeping transmission gubbins out of the rider's eyeline.

The primary and secondary drive chains both run straight, so they can run extremely close to one another; there's no real risk of interference. So, provided the epicyclic is offset to the left, the Q factor can be kept extremely low.

Putting it all together

The subframe, with the transmission and wheel, are installed and removed as a single assembly; it is located and held by countersunk bolts inserted from under the hull, with their threads and nuts inside the hull.

The bottom bracket shell, on which I had earlier thought the fork should pivots, is therefore supported by the subframe; the fork would have been attached to the middle of the shell, between two bushes/bearings. The support for the cross shaft would then be mounted off the forks, meaning that the spatial relationship between the three axles was constant. This raised a problem for pivoting the fork down to fix punctures, because the primary drive chain would have to pass through the side of the wheel well, and I don't have a good solution to this yet, although a removable panel is possible. But in any case, the idea of having the cross shaft in front of the bottom bearing makes this solution less attractive.

So, firming that up: the subframe is made of welded or epoxied aluminium, probably hollow box section approximately 76mm x 25mm. It extends back from the nose (prow?) of the hull, above the front wheel, with a leg extending downwards from the rear end of this bolted to the floor just behind the front wheel and just in front of the seat. The 'bottom bracket' shell is hung under the upper box section element in two support bearings, at the outer ends of the shell. From the chainwheel on the bottom bracket, a short chain runs forward to a cross shaft as discussed above.

We now have two options:

1. Pivoting front wheel for tyre change

The key change with the forward cross shaft is that the forks for the front wheel, instead of pivoting on the bottom bracket bearing as I'd initially conceived it, could now pivot close to the axis of the cross shaft. This allows them to swing down to expose the front wheel for puncture repair; it also makes front wheel suspension possible. It means there needs to be a fairly long (because hinged fairly far forward) 'bomb door' through which the front wheel is lowered.

2. One-sided front hub

Bicycle hubs which have disc brakes have the disc on the opposite side of the hub from the chain, and for good reason: you need lubricant on your chain, you don't want it on your disc. But suppose we move the disc to the left hand end of the cross shaft — epicyclic hubs are available with disc mounts on their left hand ends — and don't have a ratchet in the wheel hub at all. Then we can have a single sided hub. We can even adapt a standard Lefty hub, by bolting a sprocket directly to its disc mounts. And if we have both a single sided hub and a detachable fabric skirt surrounding the front wheel (or even a rigid wheel well with a removable right hand side), then the tyre can be repaired or even changed without removing the wheel from the vehicle. It would be awkward, but not impossible. And removing the wheel from the vehicle would be easier too.

If we don't need to be able to pivot the wheel out of the vehicle, we don't need a fork pivoting on the counter shaft at all. It would be both stronger, and lighter, to mount the wheel spindle to a vertical member fixed to the subframe at the top and to the hull at the bottom. We also don't need a bomb door, and the hull would be stronger and more aerodynamic without one. A pivoting one would allow suspension which a rigid one wouldn't, but...

Are punctures even that big of a problem?

I'm inclined to think, however, that given the removable subframe, and given that with modern tubeless tyres punctures that actually stop you are now a very rare occurence, that I've been overthinking the need to fix tyre problems at the road side. I've not had a puncture that actually stopped me, on tubeless tyres, in seven years. So I think that if, as a man in my seventies, I have a problem with a front tyre, I just phone for someone to rescue me. I can afford it — provided it happens rarely — and it solves a lot of problems. But I like the one sided hub idea, 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.