See other failures under 000.html.


SKF spindle failed at square taper

From http://racknroll.blogspot.com/2013/09/moving-on.html as of 2014/08:

[9721694205_181d20a15e_z.jpg]  

SKF bottom bracket.

No usage history, injury report, or other information provided.

It is unclear from the picture where the failure initiated; usually there are "beach marks" for gradual failures, it appears in this case part of the spindle is missing: compare the small square taper against the large gray/grey area torn away from the spindle.

Compare also to this picture, and not the "flare" before the gray/grey shoulder next to the aluminum ring:

[Constructeur0700.jpg]  

Square-taper bottom brackets fail most commonly at the large end of the square taper, in the flat section, just at the edge of the crank. This is one of the highest-load points on the spindle (bending load; and no lower torsional load than anywhere else).

The failure may have been aggrevated by the design as the large end of the square swells rapidly to reach the inner face of the bearing seal, likely the same seal as shown in the ISIS bottom bracket here:

[isisskf.jpg]   [skfbb_exploded.jpg]  

Sudden shifts in diameter are called "stress concentrators" because the point of most rapid size change is subject to increased stress, simply because of the shape, and thus to more-frequent fatigue failures.

[dc9a16018aa3472fbdc0ec72f429db3e.jpg]   [Square-radius-fillet.jpg]  

A vaguely similar example was a titanium bottom bracket made of two "cups" welded together to form a hollow center, manufactured by (or for) Sidetrak, Inc., marketed under the name Erikson, and recalled in 1995. It was noted for being as light as other titanium bottom brackets of the day; much stiffer; and prone to break under relatively modest use -- and at the transition, which was not welded, rather than the rims of the "cups" where it was welded.

[X]  

This is not to say the square-taper SKF bottom bracket is particularly dangerous, but it is marketed/sold as a long-service bottom bracket, and likely the failure rate could be reduced using a generally-similar design but with more effort on fatigue issues.


See also FAIL-197.html.