16

My bike is equipped with Shimano Ultegra CN-HG701-11 chain. After only 600 km I measured it with ParkTool CC-2 and it gives 0.75% wear. ParkTool says

For 11 and 12-speed chain, replace at or just before the 0.5% reading.

enter image description here

Then I measured the outer distance between 10 links L2 enter image description here

and the inner distance L1

enter image description here

The actual length between 10 links is

L = (L2 + L1) / 2 = (134.44 mm + 120.0 mm) / 2 = 127.22 mm

Wear = (127.22 - 127.0) / 127.0 * 100% = 0.173%

Many sources including Sheldon Brown suggest to replace the chain just before chain pitch grows over 0.5%. Should I trust the CC-2 or there are more accurate tools for quick check? Maybe my CC-2 is damaged or I use it improperly?

9

A lot of chain checkers read the chain being worn to x% even on a new chain (esp. on higher speed chains). The variation in new chains and reading inaccuracies can make tools essentially useless. Following these tools will lead to you replacing your chain sooner than necessary.

A better quality chain tool can help, or using calipers/a ruler as you are using. See this article from Zinn, for example -- essentially, you need to get an idea of the errors your tool makes, and you can be conservative with the replacement requirements (analogous to oil change intervals for cars; this can be a bit wasteful). Or, get a more accurate measurement (analogous to oil life meters in newer cars). It looks like he uses the Progold or Rohloff tools, which are inaccurate, but by getting an idea of the errors it makes, you can make a rough guess if you need to look closer for measurement, or just say screw it and replace the chain (even if its early)

The most accurate way seems to be hang the chain off the bike and use a calibrated ruler/tape measure over several links (in the hanging case, they suggest 100+ links, the longer measurement reduces measurement error). If you don't want to do that, a good ruler/caliper on the bike and measuring lengths.

This link explains how various common tools are inaccurate (including yours). Tools which are accurate are Shimano TL-CN40 and TL-CN41 and Pedro’s Chain Checker Plus. To quote Sheldon Brown on the reason why:

The major cause of chain "stretch" is wearing away of the metal where the link pin rotates inside of the bushing (or the "bushing" part of the inside plate) as the chain links flex and straighten as the chain goes onto and off of the sprockets. ... There are also special tools made to measure chain wear; these are a bit more convenient, though by no means necessary, and most -- except for the Shimano TL-CN40 and TL-CN41 -- are inaccurate because they allow roller play to confound the measurement of link-pin wear.

18
+500

Your high CC-2 reading likely relates to how much pressure are applying. The CC-2 device is not particularly solid and its readings are sensitive to how much pressure you apply to the pivoting gauge.

From the manual:

Instructions

  1. Set pivoting gauge so “0” is visible in viewing window.
  2. Lower CC-2 so fixed pin rests between chain’s inner plates.
  3. Lightly push pivoting gauge just until it stops. Forcing gauge will damage the CC-2’s pins, resulting in inaccurate measurements.

I also have moved to measuring chain stretch via Vernier calipers, however my approach differs from yours as I try to mimic the cc-2 type inner measurement but with more precision (see below for methods). According to my calibration chart your L1 measurement suggests a chain stretch of 0.4%, which seems like a lot for 600 km, but not impossible. I would try repeating your L1 measurement a couple more times to be sure. Either way, you will likely need to change your chain soon.

Using Vernier Calipers To Measure Chain Stretch

Because you are also attempting to use Vernier calipers to estimate chain wear I thought I would share a method I derived for using Vernier calipers. I decided to follow the same method as the Park tools (inner measurement - your L1 approach - for 10 links) as I wanted to include wear on the bushings (as this is what the cog interacts with). As many have pointed out there is no perfect way to measure chain elongation.

I have, over the years, struggled with this question, consulted with tool brands on the designs of chain-wear indicators, and have measured innumerable chains with many different methods. I have yet to come up with a definitive way to do it. I have ended up choosing one chain checker that I found to be the easiest to use, and I use it frequently and try to figure out its idiosyncrasies to ensure first and foremost that I replace my chain soon enough that it doesn’t fry my cogs, and secondly that I don’t replace chains too soon and throw good money down the drain by disposing of perfectly good chains.

-- Lennard Zinn, Technical FAQ: Chain wear measurement

So my approach has been to take an approach and improve the precision so that I could use it to do things such as estimate chain wear rates and predict service dates.

inner measurement Example of an inner link measurement (10 links apart)

I settled on your inner measurement approach (pictured above) and calibrate my Vernier calipers against a CC-2 device, so I could translate a linear measurement into an approximate % wear. I did this by creating a calibration regression chart converting the Vernier L1 measurement to a stretch percentage (see below). I figured that while the CC-2 can be difficult to use consistently, it should at least be generally accurate. Note that my calibration measurements involved only a very light push on the CC-2 gauge.

Calibration

I repeated three measurements of the CC-2 gauge at the 0%, 0.25%, 0.5%, 0.75% and 1% gauge readings. I then took the average and did a simple linear regression against the gauge readings (below).

Vernier/CC-2 Calibration Chart

The regression estimates were then used to create the following reference table:

chain stretch reference table

I believe this is a reasonably decent calibration chart as I hit 0.5% or 0.75% on the chart, the other Park Tool CC-3.2 pass/fail gauge nicely slides in at the 0.5% and 0.75% sides respectively. The conversion table is of course not perfect as some chains came with an initial stretch measurement less than 0% (i.e., the CC-2 gauge would not fit in place). As such, 0% does not represent true zero. That said, we are really interested in the 0.5% or 0.75% wear, which the chart seems to approximate well (for how Park Tools measures chain wear).

As you can see your measurement of 120mm suggests 0.4% stretch. Typically I will measure the chain with Vernier calipers in three random places and take an average, then compare against the chart. I also record these measurements and regress them against the current distance or duration to get a better estimate of the current wear rate, as well as projecting service intervals.

Predicting Service Life

I typically take periodic measurements over the course of the chain life in order to predict my Service date. Below is an example for a KMC X11-SL chain. As you can see it will be time to change the chain at about 3,250 km. Given how much wear you have observed in 600 km you may wish to reconsider your maintenance regime or chain brand.

enter image description here

Wear Rates by Product Tier

You can also use this approach to compare the performance of different product tiers within a chain brand or even between brands. Below is an example of 10 speed Shimano chains: Tiagra vs Ultegra vs XT. Note I use duration as chains are on different bikes that travel at very different average rates of speed so I didn't think standardizing against distance would give an apples-to-apples comparison. All three give the same wear rate estimate (slope of 0.0003) which suggests this a sensible comparison approach.

The XT chain was run on a gravel bike and was in dusty conditions for most of the time. The ultegra chain was run in clean road conditions and the Tiagra on dirt (similar to the XT). More recent XT chains (not displayed) seem to have the same initial reading as the Ultegra chain so I may have lucked out with the XT chain in the plot.

It is scary how bad the Tiagra chain is, out of the box it's stretch measurement is nearly 0.5% and I would be surprised if I get 40 hours of riding out of it (less than a month commuting) if I wait to 0.75% before I change.

All three product tiers have the same wear rate (slope), but it would seem that Shimano's strategy is to have the lower priced tiers have a worse starting point so that their life is shorter. If the lower tiers (e.g., Tiagra) had a tighter tolerance chain life would have been the same as the higher product tier, despite being constructed with "lower quality" materials as suggested by the marketing department.

Wear rate comparison

  • 1
    The lower models having more stretch initially is a classic strategy for manufacturing (e.g. intel makes CPU's and bins the lower performing ones under lower model numbers). I don't think 10 chains is a big enough sample though. Nevertheless +1 for the numbers. – Batman Jul 12 '17 at 2:40
  • It looks like the Tiagra is the best chain with the least stretch overall. – Kevin Thibedeau Jul 12 '17 at 15:39
  • 1
    @KevinThibedeau - I think you are misreading the graph. Tiagra is the chain that started with an initial stretch measurement of 5.7256 which is a starting point very close to the 0.5% stretch milestone (5.728). The XT chain is the had an initial stretch measurement of 5.7037 which is essentially zero. The XT started with the lowest stretch measurement. I updated the graph so the legend ordering followed the chain positions in the graph. I also changed the Tiagra symbol type to make this clearer. – Rider_X Jul 12 '17 at 16:26
  • 2
    +1 for the wow factor, you have done some serious thinking on this. i just ride my chains until it is painfully obvious that it needs to go or i get a dip in performance lol – Nate W Jul 12 '17 at 19:46
  • 1
    I've awarded the bounty to this answer because of the sheer level of commitment it takes over a long time to repeatedly measure and record all the details. I lack the patience for such work, and lack the tools to accurately measure a chain (I have tried!) – Criggie Jul 17 '17 at 0:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.