Wednesday, September 30, 2009

A golden nugget from years ago

There's been a length debate over at about Powercranks and their effectiveness. What else is new, right? One of the big marketing claims of PCs is that you'll gain a couple of mph in speed which represents "up to 40%" increase in power. The debate has always been how much is merely training effect versus something PCs may actually do. That is, do PCs, as claimed, increase efficiency and then power? Or do they just provide the motivation to train harder and you'd get to the same place with regular cranks? Having used them, I fall into the camp that they don't work magic. If you've peaked on regular cranks, you've peaked. PCs can certainly help you hit your limit just like regular cranks, but they won't push beyond that.

So what did I find in the archives? An old threshold effort from my first month of training with power and in the first 8 months of my bike riding. A little background first:

  • I started riding around September 2001
  • I crashed in mid Feb 2002 and suffered a broken hip
  • Within 3 weeks of the crash I was on an recumbent exercise bike getting my flexibility and moderate strength back
  • Within 5-6 weeks of the crash I bought a Computrainer so that I could get back on my real bike and regain full fitness
  • Within 4-6 weeks of the crash I bought Joe Friel's cycling book and learned the basic principles of proper training methods
In late April 2002 I did a "40k time trial" on the Computrainer. Looking at the data, I went pretty hard. My heart rate during the effort averaged 157 bpm. That's in the range of my typical HR for an hour effort (155-165 depending on fatigue and/or heat); I've got a slow beating heart. My HR also got up to that level within 2 minutes. These days, after 8+ years of cycling it can take a good 8-10 minutes to reach a steady value. My HR also drops like a rock - 60-80 beats in a minute once the workload is removed.

Now my Computrainer has always read a bit low. The 40k effort was 174 W average on the Computrainer for 77 minutes and 176 W for the first 60 minutes. I have calibrated my Computrainer against a Powertap and the "correct" value for 60 minutes would be 188 W. Fast forward 8 years and my hour power is in the 275-285 W range. So my hour power has increased some 50% through training effect alone. If those Powercranks worked the way they were advertised, I should get another 40% gain for a whopping grand total of 110%!!!!!

Can someone say snake oil?

Sunday, September 13, 2009

Stages of friction

There's some renewed talked on the slowtwitch forums about chain lubes and a claim that no lube is just as good as lube. I thought it would be good to provide a brief explanation of friction and wear. Perhaps the easiest way is to show what's going on with the qualitative picture below which is for repeated motion in a single wear track:

When you have fresh surfaces, either with or without lube, stage I exists. Even bare metal-on-metal contact exhibits relatively low friction early on. The duration of stage I is a function of many things, such as the materials, surface coatings, temperature, pressure, atmospheric conditions (air, pure nitrogen, argon, vacuum, etc), and liquid/solid lubricants. As wear particles develop, you enter into stage II. Stage III is a continuation of the wear particle buildup. Eventually, the number of wear particles leaving the wear track is balanced by the particles being generated and you reach a steady-state (stage IV).

The time is takes to go from stage I to stage IV is a function of many things, just like the duration of stage I. It may be anywhere from a single lap to many laps depending on the contact stress, materials, etc.

Some materials may exhibit a decrease in wear as signified by stages V and VI, though without active removal of wear particles, stage IV becomes the dominant player. From my previous entry you can see that for a sample metal-on-metal contact, stage I was short lived and stage II/III was also a short event. The metal pairing reached a steady state level quickly. It should be noted that had I taken a simple brush and wiped the surfaces down to remove the majority of the wear particles, the friction would have temporarily dropped down to stage I levels. Thus, active cleaning is a great way to keep friction low for bare metal-on-metal sliding contact.

In general, liquid lubes will keep the friction down to stage I levels. The lube acts as a transport agent for wear particles. It should be noted that you can still have wear of the materials, and indeed I have tested some VERY expensive grease (as in a few thousand dollars per pound) which continued to wear the surfaces significantly yet kept friction nice and low. However, any lubricant will have a finite life. Starvation occurs eventually and friction jumps up to metal-on-metal steady state levels as shown in Figure 4 of the link above (the June entry).

What does this mean for chains? When you hear that chain squeaking you can be pretty sure you are in stage IV and it's time to clean and relube. The surfaces are lube starved. Can you get a chain to have low friction without lube? Sure. But the chain has to be designed properly. That may mean proper material pairings, an active cleaning process (e.g., a debris wiper), or proper surface treatments (however, you could argue that something like gold plating is actually a form of lubrication).

So remember, it's not about the initial friction but how long the initial friction lasts. If it lasts a short time, then the question becomes how good is the life of my lube if one is used.