Understanding Chain "Stretch"

Status
Not open for further replies.

turboimark

New member
Messages
5
Reaction score
0
Location
Birmingham, AL
Understanding chain elongation and how to predict failure

Hey guys,

So I have been riding around the woods for a while now and it's time to change my drive chain and I figured I would give a brief walk through on understanding when/how your chain wears and how to predict failure.

I see a lot of guys seem to be in one of two camps. Either they a) run the chain until it snaps or they b) change it every -x- amount of hours regardless of condition.

In short, there IS a way to predict chain failure and it can help save you a ton of downtime and also a lot of money on unnecessary chain replacement.

And the best part of it? It's FREE. That's right, all you need is a tape measure and your eyeballs!

CHAIN PITCH

"Pitch" is the measurement of the chain from center of pin to center of pin of one link. This is the standard way every chain is measured. There are 4 main standards for chain measurement: ANSI, BS, Motorcycle, and "Engineered". For the sake of this, we will focus on "ANSI" standard chain.

ANSI standard chain is the most commonly used chain in the world and extremely common in go karts and pocket bike applications. You normally will see these chain sizes written as #35, #40, #41, #50, etc. which is the ANSI standard for chain.



All ANSI chains are based as a fraction of the number 80. So a #35 chain is 35/80 = .4375 pitch

#40 chain is 40/80 = 1/2" pitch chain.
#50 chain is 50/80 = 5/8" pitch chain
#80 chain is 80/80 = 1" pitch chain

And this goes all the way up to #240 chain which is 240/80 = 3" pitch chain.
So, how does this help me?


HOW TO PREDICT FAILURE


Well, we know now what the STARTING LENGTH of the chain is new out of the box.

So with that logic, we can measure the used chain and use the percent difference to determine when it will fail.

EXAMPLE

So for example if we have a brand new section of #40 chain and we measure 10 links of it, then:

10 links x 1/2" per link = 5" total length (NEW IN BOX)

So if we ride this kart for a little while, then re-measure, we may find the 10 links to now be 5.10" total length.

NEW IN BOX LENGTH = 5.00"
USED CHAIN LENGTH = 5.10"

If we take the percent difference (old-new / new) *100 = ((5.1-5)/5) *100 = 2% elongation.



So in this case, you should start looking at replacing the chain.

As a rule of thumb, chain should be replaced between 2.5-3.0% elongation.

The reason behind it being that chain wear is linear from 0-3% but after 3% the hardness generally has been worn off of the chain and the chain will snap unpredictably. I've seen guys run them further than that, even up to the 7% range but it's very obvious there is stretch in the chain at 3% and that will be evident through slapping, rattling, etc. Also, being out of axle adjustment (takeup) is a big indicator as well.

tl:dr

Measure your chain regularly. If it has grown 2-3% then it's time to replace it or else you will find yourself stuck out in the woods.

Resources:

http://chain-guide.com/toc.html
 

Attachments

  • ANSI table.JPG
    ANSI table.JPG
    112.7 KB · Views: 8
  • chain elongation.JPG
    chain elongation.JPG
    100.6 KB · Views: 5
  • IMG_3553.jpg
    IMG_3553.jpg
    205.6 KB · Views: 6

Hellion

Disc brakes are for cowards
Messages
5,544
Reaction score
858
Location
Location: Location:
Absolutely fabulous. A great info blurb, nice and concise. Just learned a better way to understand what I already knew (or thought I did). :thumbsup:
 

JTSpeedDemon

2019 Build Off Winner!
Messages
4,026
Reaction score
148
Location
Texas
Yes, very interesting and superbly well done!:thumbsup:
I'm very thankful you took the time to write up this article.
Who else thinks this should be stickied immediately?!
 

Kartorbust

Well-known member
Messages
3,931
Reaction score
214
Location
Utica, NE
I agree this should be stickied and posted in the FAQ section of the forum.

I wonder, would the 2%-3% work on motorcycle chain, or do those have their own set standards?

Does it also matter if it is o-ring or not?
 

turboimark

New member
Messages
5
Reaction score
0
Location
Birmingham, AL
I agree this should be stickied and posted in the FAQ section of the forum.

I wonder, would the 2%-3% work on motorcycle chain, or do those have their own set standards?

Does it also matter if it is o-ring or not?

No, the 2-3% elongation is regardless of standard. The reasoning behind that is because chain is case-hardened to a certain depth, regardless of pitch. Generally the first few .001" range of upper material is hardened and when it wears through you're running bare untreated metal on metal.

The cool thing too is that you can wear the chain and use all of your take-up (axle adjustment) and if it's not at the 3% mark yet, then just remove 2 links and bring the takeup back in tight if you have the room. It's common practice in industrial applications.

---
IMO O-ring chain is very outdated and you only see if in street bikes nowadays. In the past, tolerances were not as tight so they would use o-rings to help delay contamination. Regardless, the o-rings eventually wear out.

As far as stretch, the o-ring has no bearing on percentage. They are wedged in the sidebars so then failing does not add any more or less to allowable percentage.
 

itsid

Moderator
Staff member
Messages
11,564
Reaction score
118
Location
Ruhrpott [Germany]
BUT..
roller wear and chain stretch are two very different and mostly unrelated topics.

chain stretch is not due to rollers wearing down,
it wouldn't allow the chain to snap it wouldn't even stretch the chain at all
O-ring chain takes up for some of the wear to still protect the pin with worn rollers
and has nothing to do with manufacturing tolerances.

what stretches and thus weakens the chain is pin wear or fatigue for that matter
and at times plate stretch (on low quality chain)

So personally I'd say measure both roller diameter (averaging over the entire section of links) and actual length of segment
and decide based upon both independently..
and if you find any individual roller having excessive lateral play
you'd better order new chain now to be prepared next week.

a low load high abuse chain might not stretch past the 1% mark but can still have fully worn rollers which cause the sprocket to wear eventually start jumping etc.
and taking the slack out only makes it worse (the wear not the jumping part).
so such chain also needs to be replaced before it wears on the sprocket teeth,
no matter if stretched or not.
Since a worn sprocket prematurely wears the next chain rollers..

So the only thing to go by is roller wear at times..
Since once a pin is allowed to collect rust or dust
(O-rings offers some protection remember?)
it grinds down the pin hidden from view
and you cannot see or predict it's failure really

excessive play on the roller can be a sign of such hidden wear
since the pin and roller should wear evenly when ground against each other,
half of the play comes fromt he worn inside of the roller the other half is the outside of the pin
(which must be prevented)

'sid
 

turboimark

New member
Messages
5
Reaction score
0
Location
Birmingham, AL
BUT..
roller wear and chain stretch are two very different and mostly unrelated topics.

chain stretch is not due to rollers wearing down,
it wouldn't allow the chain to snap it wouldn't even stretch the chain at all
O-ring chain takes up for some of the wear to still protect the pin with worn rollers
and has nothing to do with manufacturing tolerances.

what stretches and thus weakens the chain is pin wear or fatigue for that matter
and at times plate stretch (on low quality chain)

So personally I'd say measure both roller diameter (averaging over the entire section of links) and actual length of segment
and decide based upon both independently..
and if you find any individual roller having excessive lateral play
you'd better order new chain now to be prepared next week.

a low load high abuse chain might not stretch past the 1% mark but can still have fully worn rollers which cause the sprocket to wear eventually start jumping etc.
and taking the slack out only makes it worse (the wear not the jumping part).
so such chain also needs to be replaced before it wears on the sprocket teeth,
no matter if stretched or not.
Since a worn sprocket prematurely wears the next chain rollers..

So the only thing to go by is roller wear at times..
Since once a pin is allowed to collect rust or dust
(O-rings offers some protection remember?)
it grinds down the pin hidden from view
and you cannot see or predict it's failure really

excessive play on the roller can be a sign of such hidden wear
since the pin and roller should wear evenly when ground against each other,
half of the play comes fromt he worn inside of the roller the other half is the outside of the pin
(which must be prevented)

'sid

Right, elongation and fatigue are two separate things.


In my experience in the field people mostly conflate "stretch" and elongation so just to be colloquial I call them the same thing in this post since 99% of people will never know the difference.

If the chain is fatiguing (actually stretching sidebars without wearing rollers) then there is definitely overloading happening and another problem entirely with the design (i.e. you need a bigger chain or a dual strand setup).

IDK about measuring the rollers separately. The only time I've ever seen worn rollers but low elongation was in cases of long flat conveyor chains where the roller for some reason was dragging along a long flat surface and locked up instead of rolling. Something like an plywood kiln chain or a car body assembly conveyor. It's extremely odd to see rollers wear but no elongation on drive chain unless the customer just bought absolute oddball Chinesium garbage.


Anyways, I didn't want to get too off into the weeds. I'm writing a Chain stretch 101 forum post for amateur go kart builders, not teaching a graduate level course in material science.

Peace
 
Last edited:

itsid

Moderator
Staff member
Messages
11,564
Reaction score
118
Location
Ruhrpott [Germany]
:thumbsup:
true tue..
just wanted to leave it as a "note"

since there is well known case for worn rollers in the yard karting world:
Some karts have very odd sprockets, some 'uhm mediocrely talented' POs decided to weld the sprocket to the axle for example or someone has a Carter Clover hub (with an odd bhc for the sprocket) so he cannot easily find a new one.

Such sprocket can potentially cause prematurely worn rollers...
if used past it's life and it's not the chains fault at all ;)

that's why..

'sid
 

batman3n1

New member
Messages
9
Reaction score
0
Location
Santa Maria, CA.
Question

Understanding chain elongation and how to predict failure

Hey guys,

So I have been riding around the woods for a while now and it's time to change my drive chain and I figured I would give a brief walk through on understanding when/how your chain wears and how to predict failure.

I see a lot of guys seem to be in one of two camps. Either they a) run the chain until it snaps or they b) change it every -x- amount of hours regardless of condition.

In short, there IS a way to predict chain failure and it can help save you a ton of downtime and also a lot of money on unnecessary chain replacement.

And the best part of it? It's FREE. That's right, all you need is a tape measure and your eyeballs!

CHAIN PITCH

"Pitch" is the measurement of the chain from center of pin to center of pin of one link. This is the standard way every chain is measured. There are 4 main standards for chain measurement: ANSI, BS, Motorcycle, and "Engineered". For the sake of this, we will focus on "ANSI" standard chain.

ANSI standard chain is the most commonly used chain in the world and extremely common in go karts and pocket bike applications. You normally will see these chain sizes written as #35, #40, #41, #50, etc. which is the ANSI standard for chain.



All ANSI chains are based as a fraction of the number 80. So a #35 chain is 35/80 = .4375 pitch

#40 chain is 40/80 = 1/2" pitch chain.
#50 chain is 50/80 = 5/8" pitch chain
#80 chain is 80/80 = 1" pitch chain

And this goes all the way up to #240 chain which is 240/80 = 3" pitch chain.
So, how does this help me?


HOW TO PREDICT FAILURE


Well, we know now what the STARTING LENGTH of the chain is new out of the box.

So with that logic, we can measure the used chain and use the percent difference to determine when it will fail.

EXAMPLE

So for example if we have a brand new section of #40 chain and we measure 10 links of it, then:

10 links x 1/2" per link = 5" total length (NEW IN BOX)

So if we ride this kart for a little while, then re-measure, we may find the 10 links to now be 5.10" total length.

NEW IN BOX LENGTH = 5.00"
USED CHAIN LENGTH = 5.10"

If we take the percent difference (old-new / new) *100 = ((5.1-5)/5) *100 = 2% elongation.



So in this case, you should start looking at replacing the chain.

As a rule of thumb, chain should be replaced between 2.5-3.0% elongation.

The reason behind it being that chain wear is linear from 0-3% but after 3% the hardness generally has been worn off of the chain and the chain will snap unpredictably. I've seen guys run them further than that, even up to the 7% range but it's very obvious there is stretch in the chain at 3% and that will be evident through slapping, rattling, etc. Also, being out of axle adjustment (takeup) is a big indicator as well.

tl:dr

Measure your chain regularly. If it has grown 2-3% then it's time to replace it or else you will find yourself stuck out in the woods.

(Disclaimer) I'm new here. my CT200u-EX uses a 420 chain which wasn't on the table. How do I compute stretch.
 

itsid

Moderator
Staff member
Messages
11,564
Reaction score
118
Location
Ruhrpott [Germany]
by quoting the longest post you essentially already stretched the thread a fair amount w/o much value ;)

just saying ;)

420 chain is half inch pitched
the rest is in the post you just parotted for no reason :D

'sid
 

Snaker

Active member
Messages
341
Reaction score
177
Great writeup Turboimark. I've worked with a lot of chain through the years, both on my stuff and industrial. But haven't gotten into the technical end, more of a change it when it needs it. Some thing I want to add is how the stress (load) is transferred, using a dirt bike as a easy to "see" example.

Looking at a dirt bike from the left side (engine drive sprocket to the left, rear wheel driven sprocket to the right), the stress is only at a couple links near entry on the drive and a couple links near exit on the driven. A very stretched chain makes it glaringly obvious.

I could put the bike in gear, roll the bike back till the top of the chain was tight between sprockets and yet easily pull the chain outward away from the driven sprocket teeth just a few links in from the exit.
So the last link to exit has the full load which it transfers to the next link in line to exit.

A new chain will have this transition happening but the transition will happen when the tooth is just starting to roll out of the last link, so the stress is deeper into the bottom of the tooth. As the chain stretches the transition happens later and the chain stress moves further out towards the tip of the tooth, which accelerates the wear .

Some of these characteristics also happen with drive belts.




On your 2-3% general wear I have a question. About the only chain I deal with these days is on my bicycle. I have a go - no go tool

https://www.parktool.com/product/chain-wear-indicator-CC-3-2



That tool specs out no go choices of .5% or .75% depending on how far one wants to push it.

Question is how do those small tolerances jive with your 2-3% general specs?
 
Status
Not open for further replies.
Top