Originally Posted by
Duragrouch
...
Chain indexing and even cogs: I read the link, interesting! I had thought the exact opposite, that odd-tooth sprockets are better to constantly even-out wear, but I ride a derailleur bike so it doesn't matter. But I had no idea that all the stretch happens on the outer-plate-links, not the inner, but I have to think about that, as even the inners use rollers and those could be displaced forward or back by wear of the inner swaged shoulders that they rest on.
Cold lube: I had been fascinated with Alfine 11 conversions of Bromptons (Scottish shop makes a new 135mm OLD triangle), until I read that the Alfine 11 does not use pawls, but rather roller friction, like an overrunning clutch. I have concerns about any transmission that uses smooth, lubricated, metal-to-metal friction, be it automotive CVTs, NuVinci/Enviolo hubs, or rollers. Some have reported slippage on Alfine 11 when climbing out of the saddle. But I mention this, because this might be worse (or better) with the lube at higher viscocity. Traditional auto automatic tranmissions have metal-to-metal clutches, and they mostly work, but can wear out. Modern dual clutch transmissions are different, having a clutch friction material against metal, with the better ones in an oil bath for a "wet clutch".
With the small amount of lube used by an IGH hub, and synthetic oils with little viscosity change over a very wide temperature range, being on the market for nearly 50 years now, one would think one would exist for IGH hubs, and not be cost prohibitive for annual changes. Synthetic 75W-90 gear lube is common. But having high shear stability to prevent lube breakdown, and also being "slippery", low friction, may be at odds with a lubed setup where you need to transmit shear force between parts, i.e., not slippery, while still lubricating other parts of the transmission.
EDIT: On another thread, someone said they had encountered an IGH lubed with automatic transmission fluid, and it swelled a plastic nut or seal enough to disengage from the threads and leaked. I also now remembered issues of compatibility with rubbery seals and such with synthetic oils. So I would qualify what I said above about that, and go by the hub manufacturer's recommendations.
I have been working on bikes for decades, only one season as a bike mechanic, the rest of that time on my own and on friends bikes. But that article was a real enlightenment, I would have assumed every link had the same elongation. But I read that before I bought my Rohloff hub 11 years ago, so I marked my cog so I could always put an outer plate link on that tooth when I first started using that hub. Rohloff hubs come stock from the factory with a 16 tooth cog, but other sizes are available. Thorn (the manufacturer of my heavy touring bike) fits 17T cogs to new bikes they sell, but I suspect that they do that so that every tooth wears the same. I have the 16T cog on my Thorn because I bought the frame and Rohloff hub separately. The photo is of my rear cog on my Rohloff several years ago before I flipped the cog. You can see really easily which tooth had the chain inner plate links because of the wear on the side of the tooth. That tooth also had a lot more wear than the adjacent teeth on the part of the teeth where the chain roller wear is. Disregard the yellow paint, I put that on it so it was easier to see which teeth I wanted to put the outer plate links on, the notch I cut in one tooth was difficult to see when the notch was filled with grime. But, if you change your chains promptly at 0.5 or 0.75 percent elongation, you likely have more even wear on all teeth. I run my chains on this bike well past 1.0 percent elongation, thus there is a bigger difference in how much wear there is on the different teeth because the outer plate links are much longer before I change chains.
I know nothing about Shimano IGH hubs, never owned one and never worked on one, other than my sisters hub when I told her to take it to a bike shop. (Brand new bike from the bike manufacturer, had a frayed shift cable, and that shift cable was hanging up in the outer housing preventing proper indexing.)
Rohloff charges a lot of money for their oil. I do not know what is unique about it, but substituting other oils has caused problems in hubs. Years ago I bought the 250ml sized oil bottles, have several more years of oil left. Regarding wear, viscosity, etc., there is not a lot of energy applied to the gears in a bike IGH. I am sure that e-bike motors put much more wear on an IGH, but so far Rohloff has been happy to join the e-bike crowd. On my vintage Triumph motorcycles, I used 80 W 90 non-synthetic gear lube which worked well. I had 140 W 90 synthetic in my Jeep pickup truck differential. I am not sure the rating on the lube I had in my Land Rover differentials, but I specified synthetic to the shop that changed the lube. I never changed the lube in my F250 four speed tranny or differential, but I am quite sure that was non-synthetic heavy viscosity gear lube based on how stiff it was in cold weather. My point is that on these motor vehicles, the gears had to handle a lot more horsepower or wattage than a bicyclist can put out and those motor vehicles had to last for a lot more miles. On a bicycle, I would never want a high viscosity gear lube, too much rolling resistance. I had 170k miles on my Jeep pickup when it went to the scrap yard, not many bicycles get that much wear but that is common now for motor vehicles.
I am clueless about what is special about Rohloff oil but it has a very low viscosity. On another forum years ago, someone ran a chemical test on it but they used the kind of test you use to diagnose motor problems based on what metals were in the oil after the fact, and some lab tech wrote on the lab report that it looked like some other fluid, I do not recall what. But that test really was not the right kind of test for trying to figure out what it was, you would need to identify what is common in other lubes that is missing in the Rohloff oil. I will just keep using the proper oil for the hub.