[Thanks for the well-wishes, Andre; better today!]
Kyle and Berto's approach is not unreasonable as a first approach for a racing team...It would be interesting to make a similar controlled comparison between sets of derailleurs with 3000m/5000km on them in the hands of the average commuters...The results will be very different.
Agreed, Andre, agreed! I have seen commuters' grunged-up derailleur drivetrains that were so packed with mud, grit, and months (years?) of neglect, the derailleur pulleys would snatch-seize and release a half-dozen times with every turn of the crank. Maintenance is just not high on the list of priorities for many people. There is no question in my mind that under neglectful conditions, the shielded, internally-geared hub would very soon eclipse the derailleur drivetrain in efficiency and parts service life. All but the drive cog and chain are shielded, after all, and there is no deflection away from an ideal chainline.
Reading the results of the research brought another thought to mind, though I've never seen it investigated. The authors make a point of saying the drivetrains were less efficient in higher gears, and this held true for derailleur drivetrains too. It was a bit of a revelation/shock for me to realize high gear on my Deore cassette is exactly the same size as the 11t derailleur pulleys. In the past, those pulleys were 10t, and I've always been fascinated at how tightly the chain has been forced 'round the small radii of the tension and jockey pulleys. They aren't under a great deal of tension, but that tight wrapping has got to be source for a huge amount of wrap-induced flex-wear, as the chain about folds double on itself one way and then the other with every passage through the pair of pulleys. I know the greatest cause and source of oil-fling is this very same tight wrap. If it is tight enough to actually eject oil from the rollers, then it also has to be a major source of wear. As a derailleur user, I find myself wishing for the much larger, red plastic pulleys Shimano fit to their budget line of steel derailleurs on department-store bikes. Methinks they're onto something.
I also recall seeing Roger Durham's comparative test data on derailleur pulley bearing friction in the late-1970s, when he came out with his line of Bullseye shielded cartridge-bearing alloy pulleys. They aways made a perceptible difference for me and lasted far longer than the old sleeved bushing design which is largely still with us. Only Simpex/Huret got it right with cones and ball-bearing pulleys, but they couldn't match Bullseye for shielding and low maintenance. If pulley bearings affect friction, what is the effect of pulley size on chain wear? It has to be huge. Of course, Rohloffs without tensioners have none of these concerns.
We aren't racing teams, much as some would like to pretend...
Darn! There go my traffic-light-sprints-for-the-finish....
All the best,
Dan. (...being won over to the Rohloff's virtues a post at a time by Andre...)
