More lube data (actually part 1...)

I recently completed a set of tests to see if there was a difference in lubricants. The lubricants tested included lubricants directly marketed to the cycling market (e.g., Pedros or Pro Link) as well as general household products (WD-40, 3-in-1). A leap of faith is taken in terms of this testing in that it is assumed that the materials sliding against each other don’t matter to the degree as the lubricant. A 440C ball was slid against an Inconel sample on a pin-on-disk tribometer. Dry sliding tests were performed to establish a baseline (dry friction above 0.60). A wear track was developed, and this wear track was used for all tests rather than starting a new track. This was done to keep the effect of roughness in the testing. After a lubricated test, the Inconel sample was cleaned and stripped of the lubricant. A dry run-in test was performed to verify the cleaned sample was free of lubricant and that the dry sliding friction was at least the initial dry value (i.e., over 0.60). Contact force was set at 5 N. Tests were run wet, immediately after application of the lubricant.

The results, quite frankly, surprised me. One of the products was a wax lube. I certainly expected it to degrade quickly, and my expectations were certainly met. But there were a few other surprises as seen in the graph.

A few results stand out. Quite obvious is the increase in friction with KryTech Wax. But surprising is the poor performance of Pedros lube. It performs on level with the wax, and this is a rather popular lube. Obviously not a good one for chains. Another surprise was that as the test went on, Purple Extreme, my current lube of choice, degraded as well. Yet 3 lubes performed very well, either continuing to decrease with time or holding steady. Pro-Link, another popular chain lube, performed best of the bicycle oriented lubes. However, the biggest shock is the performance of household products. 3-in-1 Oil, a household oil, performed well, but not well enough to a product known more as a solvent than a lubricant. That’s right, good old WD-40 performed the best of all. I was completely shocked by this result. Prior to breakdown of the lubricants, some friction values can be predicted, shown in the Table below.


The table confirms what the plot shows; WD-40 and 3-in-1 are the leaders in low friction as well as consistency (as indicated by the low standard deviation). The expected range in the table represents 3-sigma values. Admittedly the low values, particularly for Pedros and Purple Extreme, are suspect. In reality you likely would not achieve friction that low.

So what oil is the best for performance? Well, that depends. Practical experience with WD-40 is that it tends to wear off quickly, but this may be more a function of the aerosol product rather than using the same formula in liquid bottle form. Pro-Link performed well overall, but the cost compared to the household products may not be justified. Nonetheless, just looking at the raw data, you could certainly say the order of preference for lube would be WD-40, 3-in-1, Pro-Link, Purple Extreme, Pedros, and KryTech Wax.

One shortcoming of the test was running the tests in the wet state. In the future I plan to apply the lubricant, leave it for a time, wipe off excess lube, and then perform the test. This is more directly applicable to the typical application for cycling.