There are a lot of myths circulating in various forums about the importance of chainline. Chainline refers to how straight the chain runs between front and rear sprockets. In single speed set-ups and where an internal hub gear is used, it’s often possible to get a near perfect chainline. Sheldon Brown explains chainline in three articles, the lead one of which is here. With derailleur gears, the general guidance is to set up a straight chainline based on the position of the middle sprocket on both back and front, where there are 3 on the front, or in the middle on the back and between the 2 front chainrings in the case of a double, or aligned with a single front chainring. The problems with running the chain at an angle include lower efficiency, greater chain wear and noise. But just how big an issue is it?
This paper, written in 1999 by members of the American Society of Mechanical Engineers (ASME) throws up some interesting conclusions. The reason that a ‘poor’ chainline is cited as reducing pedalling efficiency is that there is increased friction causing energy to be wasted as heat. The paper doesn’t argue with this point but the conclusion is that these losses are negligible compared with non-thermal losses. The paper describes how an experiment was set up to measure efficiency relating to 3 characteristics of a chain drive: the ratio of sprockets and their relative alignment (chainline), variations in input power and rotation rate (cadence) and lubrication, or lack of it. The detailed results make very interesting reading. The authors concluded that chainline offset and lubrication have a negligible affect on efficiency, at least under laboratory conditions and no significant efficiency differences could be detected between a lubricated and dry chain. (I wonder if a dry chain might actually last longer because grit and grime would not adhere to it and cause greater wear?) The major factors affecting efficiency are spocket size, larger sprockets providing better power transfer, and chain tension. Most importantly, there is a reciprocal linear relationship between chain tension and efficiency, the higher the tension the higher the efficiency. The latter makes a huge difference. With a chain tension of 305N the experiment revealed a drive efficiency of 98.6%. When the tension was reduced to 76.2N, efficiency fell by nearly 18% to 80.9%. Don’t worry about getting your chainline accurate to within mm, don’t worry about lubrication, but do keep the chain well tensioned. I don’t know how the figures above relate exactly to chain slack but 305N is pretty taught!
Last year I bought a restored A.S. Gillott frame from a dealer in South London. It has one or two very minor marks from where it was packed for delivery, but it’s pretty near perfect. A 1948 model, it may well have been built by Ron Cooper, who worked for Gillott, and I know that Ron was involved in its restoration, replacing one of the drop-outs.
What I didn’t realise until the frame arrived was that the dropouts were designed for an Oscar Egg “Osgear” set up. It’s not possible to fit a conventional derailleur freewheel or cassette. I decided to make it a single-speed bike and found a nice pair of Charge track wheels, the rear one of which had a flip-flop hub, so I was able to build up the bike to be ridden either as a fixie or with a single-speed freewheel.
I decided that this was going to be my prettiest bike, so matched the brake and gear cables to the red of the decals and even bought a red and black saddle to complete the work of art. Looks lovely….rides like a shed on wheels!
I think the problem is partly the razor sharp saddle (back to a Brooks B17 any day now) but also the geometry of the bike. I haven’t measured it but the top tube seems particularly long. I feel like I’m stretching over the bike, even to ride on the tops of the Cinelli drop bars. And my feet feel as though they’re too far forward, wherever I position the saddle. It’s time to try a new set up – one that will make the bike as good to ride as it is to look at.
Having appreciated the joys of Sturmey Archer hub gears last year, when I road my Thanet Silverlight with a 1949 4-speed hub from Barcelona to Bristol without a hitch, and inspired by the sight of the restored Bob Jackson bike mentioned a couple of days ago, I’ve decided to rebuild the Gillott frame. Even though you have to stop pedalling for a second to change ratios with internal hub gears, each change is precise, silent and perfect – not like the rather crude process of dragging a chain across a bunch of sprockets to end up with yet another imperfect chainline. So the frame will be fitted with a Sturmey Archer S-RF5(W) 5-speed hub in the rear wheel (with a downtube shifter as a nod towards vintage styling), a B17 saddle and either a Mary bar or Mungo bar from On-One. The bars are on offer at the moment and are so cheap I decided to buy both and try them out. The Sturmey Archer was chosen because Shimano hubs and others aren’t suitable for the relatively narrow dropout spacing of vintage bikes (125mm OLD). This combination should give me a lot more flexibility in setting up the bike for greater comfort, and make it practical for commuting. I’ve bought a couple of 27 x 1 1/4 wheels, which is the size the frame was originally designed to accomodate, and I’ll rebuild the back one with the new hub – that was a cheaper option that buying a pair of rims, a front hub and all new spokes.
The bike in 5-speed form will appear here soon….if all goes well.