Geoff.D wrote:[XAP]Bob wrote:Trying to work out *how* it slips - since there would appear to be a pretty strong tension in both directions.

Then realised that the rolhoff is of course a torque convertor, so one chain set will see FAR higher tension, and a lower chain speed, than the other. But the tension is all that matters in terms of sliding along the rails.

If I understand you, Bob, you're saying that there'll be a difference in tension between the input chain and output...and this resultant causes the slippage?

Yes.

The output chain is at a fixed gear with the rear wheels, the input chain is at a fixed gear to your feet (ignoring chainrings for the moment)

I will assume that the input and output sprockets are the same size (I know this is a bit off, but bear with me)

I will also assume that "direct drive" (or close to it) is in the middle of the gear range (haven't looked it up)

In general the input chain speed and tension are approximately constant (simplification of the pedal stroke, but you get the idea).

The speed of the chain doesn't matter in terms of the tension force that is exerted on the hub (through the sprocket) and the return side of both chains is assumed to have zero tension.

In a low gear (slow wheel speed) the output chain is transmitting the same power (assuming no loss in the IGH) through a much slower moving chain - this is therefore at high tension, and there is a net force (OutputTension - Input Tension) pulling the hub rearwards.

In the direct drive (medium wheel speed) case the input and output chains are moving at the same speed and (assuming no loss in the IGH) the same tension, so there is no net force and all is well with the system.

In a high gear (high wheel speed) the output chain is transmitting the same power (assuming no loss in the IGH) through a much faster moving chain - this is therefore at low tension and the net force (same calculation as above) is negative - i.e. the force is acting to pull the hub forwards.