Friday, April 02, 2010

19 Bicycle Chain Wear Test From Wippermann


Wippermann is a German bicycle chain maker. Their stainless steel chain Connex 10sx tops a list of chains in their on-going wear test to 1% elongation (click on above). If you'd look, the wear life variation is huge between the top and bottom of the list. I also believe theirs is the only stainless chain in the list so you can go "hmm?" to that.

The testing is being performed on a rig with a 52-tooth front ring and a 17-tooth rear cog, with an applied chain tension of 300 N. The test phases feature the following phases :

- An in-line 5 hour run with original lubrication.
- A offset test to both left and right side for 5 min at 50 Hz and then 10 hours at 100 Hz after independent applications of water, oil and sand to the chain.
- An inline 15 hour test at 100 Hz with above mentioned pollutants, which is then repeated every 15 hours after cleaning until chains show a 1% change (or 8mm increase from DIN specified length of 800.10 mm). This change happens due to wear of pins against the roller.

The following video put together by Cantitoe Road shows the test protocol. A text description of the phases can be found here.



What I'm hoping they'd add into the bevy of substances to soil the chain is de-icing salt used commonly in wintry regions. Some salt water spray would be appropriate. The chains could then be stressed to twice its recommended working load and then inspected for cracks or changes in cross-section.

Salt could be one of the prime culprits in the recent case of Shimano 6700 chain failures. We were all left guessing in that post and some of us even declared it a materials issue. Well, the news is that of the affected parties sent in his chain for hardness testing to a third party facility and the tests didn't point to a materials related defect. Puzzling eh?

But the people who work in marine environments know the almost invisible, catastrophic failures on loaded chains, whether they are galvanized or stainless. Microscopic cracks develop and grow within and sooner or later, you may have yourself a broken product.

I was also hoping that multiple sprocket combination could be utilized in this test, as opposed to just one fixed combination of 52/17T. I have written recently about chordal action in tiny sprocket sizes, those lesser than 17 T.

In that post, I calculated the variation in chain speeds of a 52/11 and a 52/12 combination and wondered about the long term effects this would have on sprocket wear. Dan Connelly asked me whether I had any data with me to back this issue. At that time, I replied that I had none.

It would be interesting to see whether the theory agrees with practice.


19 comments:

  1. Do you have any idea how these Wipperman 10s chains would compare to a Sram or Shimano 9s chain?Is there a huge difference in durability between the 9s and 10s chains that anyone has actually measured?

    ReplyDelete
  2. Nate,

    That would be interesting to see and I forgot to point out that. How much durability is lost with the narrower chain designs in 10 speed compared to 9 speed? Well, let me know if you hear anything. Jobst Brandt used to say that the narrow designs of today only give about 2000-3000 miles of life to 1% wear but his Regina chains (7 sp?) back in the day would live past 20,000.

    ReplyDelete
  3. Its a nice post
    http://infopediaonlinehere.blogspot.com/

    ReplyDelete
  4. Disclaimer: We are Wippermann's North American sales agents and an importer/distributor of Wippermann chains. We also generated the newsletter from which (we presume)Cozy Beehive drew information for this post.

    Some comments:

    1) The tests are on-going. You can see a detailed view of the test results to-date at: http://www.cantitoeroad.com/uploads/landingpage/connex/2010_chainwear_test_10speed.pdf?utm_source=Newsletter&utm_medium=Email&utm_campaign=April

    Note that the SRAM PC 1090R chain was in testing at the time the results graph was created (hence no data after the first 15 hours). The remaining 10-speed chains scheduled to be tested are shown in the results list. This results graph will be updated and wear test results posted as tests for each chain are completed.

    2)Yes, Only results for the Wippermann's 10sX stainless chain are shown; Wippermann nickel-plated and steel chains are also scheduled for testing. We don't expect these chains to wear as well as the 10sX, and it will be interesting to see how they compare to competitors. However, results to-date show that Wippermann's 10sX chain considerably out-wears the best chains offered by competitors.

    3) Why compare Wippermann 10s to Sram 9s or Shimano 9s? This seems exactly like an apples to oranges comparison.

    4) The suggestion to test chains with salt water spray to see if this is responsible for cracking is interesting but it is beyond the scope of this test. This test is a wear test.

    5) Wippermann is also conducting comparative shifting performance tests, tensile strength tests, and connector strength tests. All results will be released as they are are available.

    ReplyDelete
  5. Tomboniboi : Thanks for dropping in a comment. Yes I drew the graph from the Cantitoe Road website, which may have popped up from your newsletter. When do you expect all this testing to be complete?

    The salt water test is only a suggestion, and I thought since you're going the length you have been to test the chains, you might as well add one more item to the list.

    I don't feel comparing 10 sp chains to 9 sp chains are comparing apples to oranges. Sure, the chain maybe narrower but application wise, a 10 speed chain could be potentially used 9 speed systems. Or am I wrong? Most of us know that narrow chains don't last that long, so by testing 9 speed with 10 speed, perhaps a good case can be made whether to opt or go against using 10 sp. chains on 9 sp. setups. Feel free to correct me. I'm not sure if Nate was thinking along the same lines.

    ReplyDelete
  6. Hi Ron: The Wippermann wear tests are on-going. I don't think they will ever be "complete". They expect competitors will produce and release new chains, and they'll test each new chain as it comes on the market. We will post new results as they become available on an on-going basis.

    I think the salt water test you propose my well be of interest. If you want to propose some protocol, I'll forward it on to Wippermann for their consideration. Note, I am not an engineer; I'm a hail fellow well met hearty handshake firm clap on the back sales type, so I'm not in the position of being able to evaluate or comment on any test protocol you may propose, but I'll happily forward it on to Wippermann for consideration. My email address is tomboniboi@aol.com.

    Lastly, my apples and oranges comment referred to a comparison of Wippermann 10s to Sram or Shimano 9s. To make the comparison properly, I think you'd want to compare Wippermann 10s to Wippermann 9s; Shimano 10s to Shimano 9s, etc. I wouldn't be at all surprised if the results show that 9s out-last 10s in every instance where you are comparing otherwise identical chains from the same manufacturer. The interesting part would be to see by how much 9s outlasts 10s. Wippermann wear testing of 9s chains will follow the 10s testing.

    I'm sure this blog post will be closely followed at Wippermann. Please stay tuned.

    ReplyDelete
  7. 3cross8:59 PM

    Here is the Sheldon Brown cleaning method a day late.

    http://sheldonbrown.com/chainclean.html

    R.I.P. Sheldon

    ReplyDelete
  8. Obviously the 9s chains will last longer than 10s chains. Also, obviously a Wipperman 9s stainless chain will last longer than Shimano or Sram 9s chain. But like Ron said, I have a 9s system and sometimes I use a 10s chain because honestly, I think it shifts better.

    So, the question is... how much chain life am I losing by using a 10s chain over a 9s chain? I'd be surprised if I was the only person interested in this.

    ReplyDelete
  9. Really like the idea of objectively testing chain wear. Thanks for sharing.

    I have a lot of questions about the set-up and config of the test itself:

    1. why is a length of Wipperman chain (and 2 connex links) included in every test config along with the 63 links of the test chain?

    2. how did you select 50hz as a test frequency? 80-100hz would be the frequency in a typical real world cycling application?

    3. Is a test where you literally pour sand on a chain valid? That is not a typical real-world use and would seem to favor harder materials (like the st steel Wipperman, for example).

    4. How did you control for wear in the cog/ring? Presumably some of the chains that stretched in the first 15 hrs but were kept on the test jig caused wear in the cog/ring. Was the next chain simply hooked up to the test jig so that you had a fresh chain/work jig scenario? What was the order of testing? How was the order selected and did it have an effect on the results.

    ReplyDelete
  10. @Nate - I agree a narrower chain (FEELS) shifts better on my 7-speed freewheel cassette than a ?6-7-8? chain does.

    Of course the bike I use them on is a beater, and the chains are takeoffs from my more 'premium' bikes, so it is more a sake of convenience and compromise than anything else...

    P.S. Love Wipperman links for my bikes~

    ReplyDelete
  11. Salt spray testing would be of great interest to those in coastal districts or areas where salt is used to deice roadways. Stainless steel would likely fare quite poorly under these conditions due to stress corrosion cracking, particularly around the rivets.

    There are NACE/ASTM salt spray testing criteria that you may be able to draw on to develop a test procedure.

    ReplyDelete
  12. I am very fascinated at how the 10sx chain acted after 5hrs and how that is so different then all the other chains tested.

    While this model might in fact be more durable within the parameters of this test, it still should wear in a manner that is basically similar.

    - Ryan

    ReplyDelete
  13. I'll be away in South Carolina to attend a competition but please keep this discussion going and I'll be back in 3-4 days to tend to it!

    Cheers!

    ReplyDelete
  14. Jules2:09 AM

    I'm curious how the other wippermans fare since I just (3 weeks ago) removed a wipperman nickel chain because it seemed to be wearing way to quickly. I switched back to my workhorse the SRAM 1050 and have trouble-free since.

    ReplyDelete
  15. Anonymous12:11 AM

    I missed the part where they specify whether or not they use the same cog and chainring on all the chains or a different set with every chain? I am guessing they change the cog and chainring.......

    ReplyDelete
  16. tomboniboi@aol.com5:09 PM

    Responding to roomservicetaco's queries:

    1) A length of Connex 10sX is run with each subject chain as a control. Becasue Wipperman knows the wear rate and wear charactersitics of the 10sX, they use it as a reference to ensure nothing went awry during the test.(It is cleaned and measured together with each subject chain at the end of each testing period.) Since they don't have a person physically watching the test machine all the time (e.g., a 15-hour run may happen overnight)measuring the wear rate of the Wippermann 10sX provides an extra measure of security that test conditions are consistent.

    2) The first several minutes of the test are run at 50 hz so that the water, oil, and sand aren't thrown off by centrifugal force. After this initial application, the speed is immediately ramped up to 100 hz.

    3) Sand is used to accelerate wear. True you wouldn't likely see as much sand on a chain in a real world situation, but unless you do all your riding on an indoor track, you inevtiably pick up sand and grit that wears on your chain. Whether, and if so to what extent, this test favors harder materials is a curious question. Why possibly does it matter? The purpose of the test is to measure chain wear. It's not surprising that harder materials wear better than softer ones.

    4) A new cog, ring, and 63 link section of Wippermann 10sX chain are installed for each test to ensure identical conditions for each subject chain.

    I hope this answers your questions. More results will be posted shortly.

    ReplyDelete
  17. DIYDuffer5:21 AM

    Ron, Hi, A very interesting article.A s you know More sprockets on a cassette make for more chain wear. I just thought it was the less links in contact with sprocket but you have proved it is a nom- linear rate of wear -Dont buy CAPREO (9T prockets!!!) Of course Shimano were quick in seeing the larger potential revenue stream in aftersales when marketing 'compact cassettes' - sold to us as saving weight on the bike -in less durability sales for replacement replacements. [BIG S saw this right at the start of the Off road bike boom.
    Suntour were not as smart as the BIG S in this respect -history tells all]...and every years- we as a market swallow the same old Cr@p. (admittedly it is nice-but at what price??) I'll stick with my Six & 7 speed cassettes for every day bikes.... I recently replaced a road bike last years but had no choice but to buy Shimano 105/ Ten speed. I would have preferred as cassette with the smallest. sprocket of 13T or 12 T for durability and the jump between 2nd & top is not so big (I only have sparrow legs!!!). Is the Is the weight saving so different?????
    Rant over
    Now the SRAM/Camp & Big S on onto thei game they will be no turniing back.
    PS I had a 'Compact' chainset in 1983 on a Raleigh Tourer, Shame the front mech were not upto handling a 34T/50T front but thats another story.chiansette then!!!
    Cycling is supposed to be more enviromentally friendly but I believe I am just investing in landfill.

    ReplyDelete
  18. Wippermann has finished more testing of the 10 speed chains.

    The wear tests now include:
    * Connex 10s0, 10s8, 10sX, and 10s1;
    * Shimano CN 6601, CN 7801 & CN 7900;
    * Campagnolo Veloce & Record;
    * SRAM PC 1050, PC 1070, PC 1090, & PC 1090 R;
    * KMC 10L Silver, 10 Gold, 10SL Gold

    http://www.CantitoeRoad.com/connex

    ReplyDelete
  19. aaron popelka10:36 PM

    Ron-

    Great blog and post! I just found your blog recently and enjoy it greatly. If you wouldn't mind, I would like to pick your brain for a second.

    I am a bicycle mechanic, and a coworker and I have been having a debate recently on chain wear. Neither of us are engineers or have an engineering background, but I was intrigued after reading about this test. I was curious to know your thoughts on chain wear, and what causes it. Our debate is as follows-I postulate that the main cause of chain wear is the bushings rubbing on the pins and outer plates of the chain, and that this wear can be exacerbated by grit and dirt getting in the mechanism. I claim that riding in a gear that moves the chain faster, e.g. a small-small combination, is worse for the chain than riding in a big-big combination, because the main cause of wear is due to the chain and bushing flexing and rotating as they pass through the derailleur and gears. My colleague claims that while this is true, the main cause of chain stretch is from the lengthwise tension exerted on the chain from the bottom of the chainring to the bottom pulley of the derailleur, e.g. the chain's plates are literally stretching out and being pulled farther apart. Could this be true on hollow pin chains? Or do all chains stretch the same? And is such wear due to bushings rubbing against plates and therefore (slightly) decreasing in size and allowing more room for play?

    Alternatively, if a test were to be set up where a weak rider and a strong rider rode the same setup in the same conditions and same gearing for the same distance (but at different paces, because one is fast and the other slow), I postulate that the difference in chain stretch between the weak and strong rider would be negligible. My coworker believes that the strong rider would stretch his chain faster than the weak one. Would you be able to weigh in as to what you think the answer would be to either of these two questions?

    Thanks for your blog, and your time!

    -Aaron from Bicycle Revolutions in Philadelphia

    ReplyDelete

Thank you. I read every single comment.