OK so this is a touch off topic
That particular shaft has alot of axial fibres in it. Fibres which run mostly along the length rather than at +/- 45 degrees as you'd expect in a torque shaft.
Reason being is twofold:
A) Its really long. Needs to not whirl (I.e. resonate) at high speed and hence needs a high flexural stiffness.
B) That particular shaft is a so called "compliant shaft". You see how its before the system torque limiter? The system torque limiter does pretty much what it says on the tin, but this shaft doesn't benefit from its protection. The reason for that I used to know, but I've forgotten. Something to do with failure modes and "catchers".
Anyway...that first shaft can withstand full motor torque and will just stall the drive unit if needed. However it also has to have a low torsional stiffness so you get the right torsional vibrational behavior in the system. The net result is a shafr which is bloody strong in torque (>4000Nm) but which can also wind up by more than 180 degrees before failing. It also has a high flexural stiffness.
All of that is achieved by having an usual laminate design which has loads of high strength material at all sorts of non typical angles.
Incidentally, before this kind of carbon fibre tech was developed (I
think we were actually the first to qualify it) compliant shafts were a more complicated affair of two concentric metal shafts. One narrow solid shaft made from titanium or managing steel up the middle to give you the required low stiffness/high strength balance, then one hollow tube like a drainpipe riveted over the top to give you the vibration.
This is a textbook example case of how using carbon 'properly' can do so much more than make nice looking bodywork.
