eirik@tekchips.UUCP (Eirik Fuller) (10/18/85)
In article <395@uw-june> wagner@uw-june (Dave Wagner) writes: >> >Can anyone comment on elliptical chainwheels? > >It seems rather obvious to me, that at any point in time the chain is wrapped >around exactly half of the teeth on the chainwheel, regardless of its >orientation. Think about it. > "Exactly" is a strange word to use in this context, considering the wide range of angles which the chain can make with the chainwheel, depending on the size of the selected cog and the length of the derailleur cage. This argument adequately accounts for first order effects, but the fact is that the top and the bottom of the chain both bounce up and down. Consider an extreme example: make a chainwheel with a huge aspect ratio (25 to 1; why anybody would use it is a mystery to me). With the long axis horizontal, the top of the chain will be approximately horizontal. With the long axis vertical, the top of the chain will make a steep angle to the horizontal, meaning that it will take more chain length to reach the top of the chainwheel. To be more precise, the chain will reach the chainwheel before its top. However, the length of chain needed to reach the chainwheel will still be more than if the long axis is horizontal. Furthermore, there will be more chain wrapped around the chainwheel. Think about it. I don't think your argument completely accounts for all effects. I can see two possibilities as reasons the derailleur doesn't wag: the second order effects cancel out, i.e. the reasons are purely geometric in nature; or, the derailleur cage has enough inertia, and the second order effects are negligible enough that the slack is taken up elsewhere, like the lower portion of the chain, whose tension is regulated only by the derailleur cage. It's been long enough that I can't remember what the bottom of the chain looked like in motion, except I know that the front of it definitely bounced, with the rear as a pivot. I don't remember how straight it was, or how much this varied.