Yep...and also the Berner "blade" style pulley cage :-)

And, in the case of the eTap shown above, how about design the battery cases to create a Kamm-tail foil shape with the derailleur body?

---

http://bikeblather.blogspot.com/

Silly question....Has anyone taken the rear derailleur off at a wind tunnel to see how much it is costing from a drag perspective? If not having it there at all is only .1 w better than there won't be much advantage to optimizing it.....if it is 10 w better than you should get to work :)

Good thought ....

---

Cheers, Neal

+1 mph Faster

Agreed, Tom. I still have one of those old Shimano 600 rear derailleurs and think they're beautiful. The styling of more current derailleurs definitely doesn't look very aerodynamic.

I do not see the design changing much however for a cable actuated version I do believe that the more direct feeding of the cable like MTB versions would be better like some others have suggested. Why do we need such a huge loop. I am running Ilink Mini to make the loop smaller but it still does not need to be there.

Also the RD has way too much material as I have seen people reduce the weight of the RD to under 100g and still have it functional BUT that is just resolving the weight issue and by removing a fair bit of material it does not solve the overall issue of size.

I think they could be optimized BUT I do not see the overall design principle changing any time soon.

Can they ever get a gearbox small enough to fit into the BB area and maybe optimize the frame to still be aero?

This ^^^. You could probably operate the expanding gears with servo in the bottom bracket and rear hub. It would have virtually stepless and uninterrupted gear ratios so you could simply set a target RPM and have the system constantly trim the gear reduction to within 1-2 RPM based on rear wheel rotational speed. Huge advantage on a mountain bike. No hand controls needed for shifting.

Now, to get the chainrings and cog to be mechanically efficient despite the gaps.

You'd still need a very small tensioner because of the lag in shifting. But it could just be mounted to the chain stay and also act as a guide to prevent dropped chains and make wheel changes easier. The system would also allow changes in wheelbase with horizontal dropouts and much easier wheel removal.

---

TrainingBible Coaching
http://www.trainingbible.com

On the other hand if you have a chain tensioner you don't need expanding chainrings for both the crank and rear hub. :)

You don't need a chain tensioner for that setup. As the front expands the rear contracts and vice versa. They would be synchronized so that the total number of teeth engaged by the chain between the two would remain constant.

I know. You missed my (weak) at attempt at humor. I was just pointing out that you're replacing a simple, reliable design consisting of a spring and two jockey wheels with a complex design requiring multiple motors and precise computer-controlled synchronization of complex mechanical stuff.

Yeah the cost/benefit probably isn't there for the majority of the market. It *might* be there for high end road racing, time trialing, triathlon, and maybe even mountain bike racing (XC in particular).

The synchronization isn't particularly hard to accomplish. None of the parts are actually very expensive but the assembly of those parts would invariably be expensive.

I don't know that there would actually be an efficiency hit imposed because you're "missing" 2, 3, or even 4 teeth. If there was an efficiency hit it would be very very tiny and more than offset by gains in efficiency from perfect chain line and chain tension.

The act of pedaling would stir up any air getting there quite considerably I would think. I don't know that you would get anything meaningful out of it and I am sure it would be entirely relative to foot size and crank length and shoe style, etc.