Great exploits and photo-article Aidan, opens up a whole new world -
fascinating stuff. Doesn't look very complex!
Hehe! No doesn't, does it!
The simplicity of its construction belies the complexity of hydro and thermodynamics going on there... which is why it's such a fascinating device.
It took quite a few years to find the trick to getting these things to act as an auto-lock clutch i
n a controlled and predictable way. Without the slots and holes and other clever tricks, you just get a steadily decreasing torque Vs speed curve as fluid viscosity, even silicone fluid, decreases in the normal way with temperature - quite the opposite of what you really need.
These things work at transient internal pressures up to 100 bar and the plates deform in 'sets' and pairs. The fluid 'fill' is quite critical, in viscosity and volume terms. Too much and it'll burst, too little and it won't lock early enough - and then might overheat.
The working analysis has been advanced a lot in the last decade and is now pretty well understood, though is still a mixed mathematical 'model' based both on theory and empirical correlation, giving a way of scaling and designing couplings for a variety of sizes, loads and speeds.
More advanced viscous couplings now have servo-control of locking behaviour - quite a long way from the original Ferguson/Rolt patents.