A few more thousand waxed miles under the belt and a few more observations...
Wax Colour
The wax I've been using began as a typical opaque colourless / white variety, subsequently turned grey by the moly powder added. Over time it seems to have taken on an increasingly brown hue; and what little discussion I can find on the net suggests this is because it's been over-heated / "burned" and should be replaced.
I'm somewhat dubious of this explanation as I've never really cooked it (it's maybe seen 90 degrees C tops occasionally), and have had wax smoking hot in the past / at the point where it will vapourise and burn like a little chip pan fire, and don't recall any discolouration of what was left.
The colour seemed to get even more yellow / brown after I recently started using a new chain, which I think provides a clue. This had been half-arsedly rinsed in paraffin to remove the sticky factory grease, however handling afterwards suggested that there was still a fair amount left on the chain before it went in the pan.
The obvous conclusion is currently that the yellowing is the result of grease on the chain dissolving in the wax. The puck is currently about 1" thick and the yellowness is mostly confined to the top 1/4" or so; I'm not sure if this is indicitive of the lighter hydrocarbons in the grease rising to the top as it cools, or that the colour is otherwise homogenous but disguised by the heavier moly powder that gravitates towards the bottom of the puck.
Equally it's possible that the wax could be oxidising / breaking down over time / repeated heat cycles - but I can't find any information on how likely this is.
Is this a problem? Not that I can tell.. the wax still seems to be doing its job so I'm in no hurry to replace it and will keep topping it up as the level in the pan gradually drops.
Metallic Contamination
As covered in
my post from June 2023 I'd previously encountered some significant metallic contamination settling out at the bottom of the wax puck.
At the time it was speculated that this was likely from a couple of chains that had been previously run for a good distance with conventional lube before being waxed, rather than from items that had been waxed from new.
When this post was made the puck would have seen about 7-8k miles worth of waxing, while in the 2.5yrs elapsed since the contamination was removed I've waxed another 10-11k miles worth of chains.
Checking the underside of the puck recently has revealed next to no metallic contamination, with that present apparently restricted to one or two obvious flakes and nothing like the uniform glittery layer that made up maybe the bottom 1/8" of the puck last time.
So, it appears that this contamination was indeed the result of waxing previously-oiled / significantly worn chains; the takeaway being to only wax new chains if possible and if you must wax a partially-used, previously oiled chain to ensure it's really thoroughly cleaned beforehand (if this is even possible).
This potentially raises questions about what happens to wear-induced metallic contamination in chains. Presumably with an oiled chain the steel fragments worn off the chain remain captive between mating wear faces of the chain; held in suspension in the oil. Conversely since such fragments don't seem to appear in the pan post-wax they must be going somewhere – I assume being pushed out of the chain with the wax as it's displaced during use.
This also makes me wonder whether the mechanism of wear differs significantly between oiled and waxed chains. We all know the former allows external contamination (dust, grit etc) to be drawn into the chain by capillary action and act to abrade the corresponding surfaces in the chain.
Since wax evidently hugely reduces the ingress of such external contamination, I wonder if the metallic particles that result from wear are perhaps a lot smaller...?
Chain Longevity
Finally the distance covered on various waxed chains has allowed me to make some anecdotal observations about chain life.
Brompton:
KMC E8 and original SRAM PC10 (8 speed), both written off at 0.5% wear at around 2-2.5k miles on each chain. Both low-mid range chains, subject to a pretty hard life as used in all weathers while the drive train is pretty close to the ground on this bike and running around small (13, 16T) sprockets at the rear (50T chainring) which are known to increase friction / wear. The chains are also a bit shorter than normal at 100 links compared to the 112-114 links typically found on a full size bike with conventional derailleur drivetrain.
A fairly significant amount of variation in wear found between different links, with an approximate correlation with where each link is loaded relative to the pedal stroke.
Fuji:
KMC X9 and X9 EPT (9sp); both low-mid range chains and of comparable quality to those on the Brompton above. First partially worn when I got the bike so difficult to assess mileage to failure. Second got to about 4k miles from new although it was excessively worn in some areas so should have been written off before this point – perhaps reaching 0.5% wear at 3-3.5k miles. Used in all weathers, typically running on a 36T chain ring and at the smaller end of the cassette – the 15T and 17T sprockets apparently seeing the highest median use.
Very large amounts of variation in wear between links, with some still showing at inside the 0.5% wear limit with others clearly exceeding this by a good amount.
Given the similar chain quality and operating conditions to those on the Brompton, I suspect this comparison offers a reasonable insight into the differences in lifespan between chains used on the Brompton and more conventional drive trains.
Genesis:
KMC X11 (11sp), a mid-range chain that’s touted as having greater wear resistance than the X8/X9. Potentially had an easier life than those above as mostly fair-weather use on larger chainrings / sprockets (50/34 and 11-34 cassette). Currently sat at around 4.5k miles with very little suggestion of wear, and it wouldn’t surprise me if this got to 6-7k before it needed replacing.
Obviously difficult to make comparisons to those above given the number of different variables involved.
It's interesting to note that so far none of the chains that have reached / exceeded 0.5% wear have done so at anywhere near the mileage suggested by measurement early in their lives. I suspect this is due to measurement error (as discussed earlier in this thread) and likely exponential wear behaviour - i.e. the more worn the chain, the faster it wears - so wear rate accelerates as the chain heads towards the end of its life.
I’ve started running X11s on the Fuji’s now so it’ll be interesting to see how these wear compared to the same chain in the Genesis application / seemingly lower-quality chains previously fitted to the Fuji.
How's everyone else getting on?