Thursday, November 8, 2012

For the Analy Retentive Cyclist: Part 1, Reducing Friction

Suppose you put an SRM (or other crank-based power measuring device) and a Powertap on the same bike and went for a ride.  Supposing they are both calibrated correctly, they should give you the same number, right?


The Powermeter will give you a lower number than the SRM.  Precisely how much lower?  Well, that's the point of this series.

This post will examine how to minimize the energy that should go into rotating your wheels, but instead goes into heating your chain, your hubs, your bottom bracket, and all your sliding parts--that is, how to reduce energy lost to friction.

But first, an overview...

Mechanical Efficiency

A bicycle is the most efficient form of transportation because it transfers almost all the energy from your feet directly into forward motion.  At it's best, it is 99% mechanically efficient.

Unfortunately, the bikes you and I race with not bikes at their best.  They are much less efficient; usually, they are only 85-89% mechanically efficient.

The reason for this?  Gears.

Unless we take up track racing gears are necessary, but the energy loss we accept when we ride bikes with gears--and this is the real point of this--can be limited.

While fixed gear bikes are almost invariably 99% efficient, there's a wide range of efficiency for geared bikes; estimates differ as to how much you lose with a drive train--somewhere between 3% and 15%.  Geared is worse than fixed, but bad geared is far worse than good geared.

That's good news, since it means that, for those of us with the time, money and a values system that is as divorced from reality as is possible to optimize our drive trains, we can go a lot faster.  At 200 Watts, for instance, a good drive train can mean 24 Watts.  That's over 1 mph increase in average speed, a couple minutes off a 40k time trial, the chart below suggests.
That's a whole lot more than aero helmets get you over 40k (i.e., 30-60 seconds savings).

In other words, in a time trial, ensuring your drive train is optimized for the maximum mechanical efficiency is probably as important as a time trial helmet.

So how do you maximize efficiency?  Well, first, we need to look at how you lose energy in the drive train.

How energy is lost

Well, that's not quite isn't lost--it just isn't going where we want it to go; that is, into forward motion.  Instead, it's heating sliding objects on the bike up, as friction or bending and oscillating parts, what I'll call flexion.  The two enemies of efficiency, unless you are a fire starter or pole vaulter, are friction and flex.

Just how much energy is lost through the thermal effect of friction?  Well, here's an example of a guy putting burn marks in a sheet of paper from simply banging two metal ball bearings together:
The energy from a simple collision causes enough heat to burn paper.  Imagine the energy created from a continuous sliding together of the kind you have in hubs, in cable pulles, and in the bike chain itself, which is a series of dozens of sliding parts.

Heat is one way to understand the energy lost; sound is another way.  A loud drive train is necessarily an inefficient drive train, since energy has made noise rather than accelerated the wheel.  Every noise you hear on your bike (and in your car, or on the metro, for that matter) is the sound of energy being wasted.  The vibration of guitar, and the acoustic energy they transmit is probably the best way to picture what happens to a frame, chain, or other bike part that is struck or stretched or stressed.

Waves travelling through strings at difference frequencies--that's one way of showing how energy travels through materials by flexing them.  Strings allow sustained energy (and, therefore, sound), while other objects (drum heads, for instance) absorb energy relatively quickly, producing a staccato effect.

These are two illustrations of friction and flex-- if energy is going into flexing or heating, that is energy that would otherwise go into the rotational force of the rear wheel.

Bikes will never be 100% mechanically efficient, so our task will be to minimize it, given the materials and budget we have.

You'll find energy losses in almost all your components, as this chart shows:
This is an oversimplification--pedals, of course, flex a bit when you step on them, addition to losing energy as they turn on the spindles with each pedal stroke.  This indicates what testers have shown to be the things that contribute to measurable loss of mechanical efficiency.

Now we've identified where there is energy lost, let's look at ways to minimize that loss.

Fighting friction

Pedals, hubs, your bottom bracket, and the little wheels in your rear derailleur--these components are all spinning parts that lose energy to friction (the chain does as well, but more on that in a bit).  They all use ball bearings to minimize friction loss, so they are already extremely efficient, but there are some things you can do--things that cost money---to reduce the amount of heat your components produce as they slide against each other.

Ceramic Bearings

In the last decade, some cycling manufacturers have begun using ceramic bearings, a material harder than steel and smoother--resulting in less energy going into bearing wobble and thermal gain, and more energy simply directed to where it's supposed to go, they say--into rotation.

For bottom brackets, the claim is that steel bearings have a tested drag of 4%, but ceramic bearings have a tested drag of 0.5%--a savings of 5w at 100w output.

And that's just the bottom bracket; a bike fully equipped with ceramic bearings (add to the bottom bracket chain pulleys, hubs, and pedals) would save even more energy.

So they claim, but even if it's isn't 5 watts, iinstalling ceramic bearings will almost certainly save you a small amount (prompting a nice April Fool's article about the UCI imposing friction limitations on bikes), so you're getting dropped, don't think ceramic bearings will help you win races.  If money is no object or you are just an obsessed asshole (no disrespect intended), it will yield a few seconds over 40k, but probably not as much time as a TT helmet and a good coach.

Here's a bit of evidence from Cyclingnews (Source), a bit of showmanship,

Bearing lube

Most mechanics recommend "packing' your bearings with grease.  This can be auto grease or synthetic grease--I found many recommendations.  This is probably best for the preservation of your bearings; even the so-called "sealed" bearings introduced in the last decade aren't impervious to water and dirt, and grease does a good job of maintaining the seal so less stuff gets into the bearing.

However, the most efficient lubricant is not grease--it's oil.  That's because the thick coating of grease tends to slightly impair the free flow of the ball bearings.  As even skateboarders know, the thinner your oil, the faster your balls will spin.

Oil isn't exactly practical as a bearing lube, although it works better with ceramic than it does with steel (ceramic bearings can theoretically "pulverize" debris that gets in their way, and are less likely to be scratched or deformed than steel bearings).


The reason an SRM will always produce a higher reading than a Powertap is because energy is lost between the crank (where the SRM measures) on its way to the rear hub (where the Powertap measures).

Some of this energy loss is caused by friction.  To reduce friction, use ceramic bearings.  Use oil if you're completely and utterly serious and kind of a dick about it.

Next up, improving the efficiency of your drive train--your chain, your gear choice, and selecting the most efficient cadence.

1 comment:

Tim Rugg said...

Good stuff Kevin. Thanks.