Note, this doesn't take into account rolling resistance or the like. A completely flat tire will, of course, cause efficiency to plummet.
Most of the upgrades we've suggested thus far have been expensive (ceramic), impossibly complicated (calculating the location of nodes), or impractical (lubing bearings with oil rather than packing them with grease, or using huge gears).
This entry, we'll examine a cheaper and necessary way to boost efficiency--optimizing chain efficiency through lube choice.
Let's start with a thought experiment. Suppose you were suddenly in a world where wind resistance remained constant, but every source of mechanical efficiency remained. You know you can put out 250w for an hour. Which of the following machines would be fastest?
|Chris Hoy's Fixed Gear Track Bike|
|Graeme Obree's "Beastie"|
The chart below shows what your times would be, assuming two things: constant power output of 250w and constant aerodynamic resistance equivalent to ring in the hoods on a regular road bike.
Tests show chains--even lubed, new chains--sap between 7-8w at 250w (that's between 2-3%). Of course, we have assumed a complete absurdity--wind resistance is much greater on a penny farthing than on a modern bike. Still, it's useful to show just how much energy we lose to our chain--almost at minute over 40k. Of course, just getting in the aerobars from the hoods saves us about 9 minutes, so clearly the penny farthing, however mechanically efficient, loses out to the modern TT bike.
(If you'd like to test out the data yourself, visit this calculator).
Those tests I referenced show a small amount of variance--sometimes chains sap 7w, other times 8w. If we can correlate anything to this 1w difference, we can actually find something useful out of this whole exercise.
The difference is correlated with temperature--cooler chains were more efficient than hot chains. As the chart below shows, at 150 degrees, chains tended to lose 8w, while at 0 degrees, the loss was just over 7w:
Why temperature effects mechanical efficiency is important, but there isn't much data on it, as far as I'm aware.
One theory explains cold chain efficiency by suggesting that more viscous lube is more efficient.
Another theory suggests that the colder steel is stronger, and stronger steel is subject to less flex, leading to less chordal action.
If you follow theory #1, you go out and buy gummier lube. If you follow theory #2, you might consider a more robust chain.:
Factors to consider when buying and using lube
The theory above suggests that more viscous lube is better, but tests suggest otherwise. We've already found that oil works better than grease on bearings because it is less gummy and more fluid.
Another tester found WD-40--a very thin substance--to be the most efficient lube:
A few other observations about this test: the worst and best performing lubes were aerosols. Wax performed worse than most oils.
Viscosity of lube is, therefore, not obviously correlated with efficiency or inefficiency. Nor is it correlated with efficiency over time.
Before you start lubing up with WD-40, recognize its failing as a lube--it is not a good lube to use for any extended period. Unlike other oils or lubes, WD-40 does not stick to parts, leaving them "lubricant starved," leading to higher friction and greater wear on components. 3-in-1, a thicker lube, was only slightly less efficient, but it coated moving components well in the test, and showed impressive durability.
You may also want to think about the durability of your parts--wax has been shown to keep debris out of parts, keeping them clean and running for longer than other lubes.
So what lube should you use? Well, it depends on the data, and I don't have access to the best data. That's at Friction Facts, but it costs money, and I'm a bit skeptical about the claim he makes that you can save 7w by using specific gear. Claimed savings over time is an even more dubious claim, since we've already found that fastest lubes often tend to be short-lived.
Factors to consider when buying a chain
As mentioned, chordal action can sap more energy than friction. Minimize chordal action in your chain by keeping it tight--making sure there is sufficient tension on it from the derailleur--and don't use old, stretched chains.
The important thing to note here is that weight is relatively unimportant when compared with strength. High-end chains usually weigh 20-50g less than lower-end chains, but the more important factor is strength. Unfortunately, manufacturers don't provide information on the strength of their chains. Chains weigh about 1/3 the weight of a typical bike frame, but most high end chains weigh in under 250g (e.g., SRAM's 1091R Hollowpin or Shimano's ). Far more important than weight, though, is their efficiency. Unfortunately, this data is either not there or costs money to procure.
If you don't have the money for a test, you can easily perform such a test yourself. Just borrow a Powertap and SRM, try various lubes and chains, and track the difference. The closer the Powertap number to the SRM, the better the stuff.
Here's my advice for chain care:
(1) Keep lubricated always;
(2) Keep clean always;
(3) For prologues (but nothing longer) and possibly short TTs try WD-40;
(4) Replace when stretched.
Nothing revolutionary, I know, but it'll not only keep your equipment, it'll make you faster. If you've got the cash, however, it might make sense to spend $50 on the reports and selecting more efficient equipment.
And if the loss of energy appals you too much, consider the penny farthing--you'll not lose any energy to a pesky chain, and you'll certainly gain the disgust of those who aren't anally retentive enough to understand the beauty of a nearly 100% mechanically efficient bit of machinery.