RE: [CR]Filed cranks: the result [was: Stress cracks on Super Recordcranks?]

Subject: RE: [CR]Filed cranks: the result [was: Stress cracks on Super Recordcranks?]
Date: Tue, 17 Oct 2006 13:13:11 -0700
In-Reply-To: <>
Content-class: urn:content-classes:message
Thread-Topic: [CR]Filed cranks: the result [was: Stress cracks on Super Recordcranks?]
Thread-Index: AcbyCopU634v1/DYS+qak4bKsRXBWQAGvSOA
From: "Mark Bulgier" <>
To: "ryan hildebrand" <>, <>

Ryan Hildebrand wrote:
> With all your encouragement and instructions I took a small
> chainsaw sharpening file to my cranks. [snip]
> Aesthetics aside, what do you
> think? Would you ride these cranks?

Not everyone should ride on known-cracked parts, but some people are aware enough, willing to inspect often, and/or crazy and place a low value on their own lives. I am, I suppose, all of the above, and have done it many times. I'm guessing someone who can take a crank off and file it as you have done (looks good!) is likely someone who can be trusted to clean and inspect often.

The cracks will grow relatively slowly at first - though the farther a crack progresses, the less uncracked metal remains, so the stresses increase and crack growth rate accelerates.

I would say that if there is ANY cracking visible at any point in the future, toss it - you gave it a shot with the filing, and if you didn't file away all the cracked metal, then it's toast, don't try it again.
> Has anyone heard of an LBS offering a crack detection service
> (e.g., Magnaflux)? Anyone know of liquid penetrant home
> kits/sources for?

Bill Davidson, custom framebuilder out of Elliott Bay Bicycles in Seattle, used to have a dye-penetrant kit and knows how to use one. Don't know if it's a service they offer and it might be too expensive.

I just did a brief search for web sources for dye-penetrant and it looks like a small kit costs about $50. This will be a many-lifetimes supply of course, but I don't think they make it in smaller cans. Here's one source I found just now, have never dealt with them though: rd= Penet&d=1&d1=&d2= or

I have done it and it's not hard, just follow the instructions completely. The results can be inconclusive if your technique is at all sloppy or rushed, but done right it's usually quite clear and you'll see the true extent of the crack. The Wikipedia entry on it gives an overview:

Mark Bulgier
Seattle WA USA