Thursday, April 16, 2009

Thomson Elite Setback Post Failure

Readers will note that I wrote about a Thomson seatpost bolting ear failure a few weeks back. 4 days ago, another Thomson Elite setback post was involved in an incident out on the field. The product owner reported that he had a close brush with an "aluminum colonoscopy".

Facts :
  • 31.6mm OD, 410mm total length with 5/8 inch or 16mm of setback.
  • Used on a Cannondale 29'er (offroad XC bike).
  • Made in USA with 7075-T6 Al alloy and anodized black to a glazing finish.
  • Post is machined and the head and post body are integral units.
  • User owned the product for 10 months.
  • Approx 2000 miles of use, mostly on pavement, some on fireroad and some singletrack.
  • Never been dropped, wrecked, knock over or even clamped in a repair stand.
  • Cannondale saddlebag was velcro strapped around the rails and post with nothing sharp in it contacting the post.
  • Owner weight = 190-200lbs. No indication of a weight limit in the instruction sheet.
  • Owner followed Thomson instruction sheet for installation and use "to the letter".
  • Saddle was installed far back on the rails beyond neutral position and the result was that the post was clamped on the forward 1/3 of the usable rail portion.
  • Owner was immediately replaced with a new post by Thomson over the Easter weekend.

Macroscopic features of the fatigue and final static rupture are easy to observe in the broken sample. The fatigue crack originated on the front side of the post body, propagated with loading cycles thus smoothing out the surface as it progressed. The static failure happened at the back end of the post, and these areas have a rougher appearance.

In this picture that shows the underside of the head, it appears that there was one 'dominant' fatigue crack on the tension side of the post that initiated this process. Observe the darker 'slit' near the arrow.

What's bothering me at this point are the following :

1) Thomson's marketing materials online claim that at around 250 in-lbs of torque on the post, the product will start yield phase and bend at the seat tube clamp. There is no evidence of any bending, either on the rails, or on the clamping site, from the original photos.

2) An elliptical bore leaves more material on the front and back ends of the post, in order to withstand bending stresses. Material is removed from the inner sides of the bore to reportedly, save 30-40 grams of weight. This structural design did not prevent this failure.

3) Does anodization reduce the fatigue life of this seatpost? In other words, did Thomson test the seatpost in post-anodized state? Anodizing significantly reduces the fatigue strength of Al alloys. As Al bends elastically, the anodized surface cracks and the crack grows into the body of the Al. Anodized aluminum only worsens the fatigue limitations of Al. You can't bend anodized Al significantly without cracking it. The cracks that develop on the coating are stress risers and potential sources for fatigue failure in the substrate metal.

4) Additionally, if one closely inspects the outer surface of the seatpost, there are grooves all along the length of the post. This may not be seen at first glance but can be felt with the hand. I'm not sure what functional purpose this serves, other than aesthetics. Is this done to reduce seat post slippage while clamped and loaded? Whether these features have any role to play in the breakage is something to be further studied.

If you use an Elite or a Masterpiece seatpost, it won't hurt to periodically inspect what's going on with the post. This type of failure is very dangerous largely because it is hidden and away from the sight of the user, happening well below and under the saddle. I hope Thomson takes this seriously and undertakes a root cause analysis before any one gets hurt. While replacing the product for free addresses people's concerns to some extent, it still doesn't take care of the problem.

* * *


Bluenoser said...


We had a strike at a steel mill hear years back which led to a shortage of steel pipe.

I was having problems with tearout while threading pipe and replaced numerous dies.

I used a scrap piece of Canadian pipe for a nipple and no tearout. Closer look showed the pipe from another country with lesser QC being used was full crap causing the tearout.

I'd be looking at the supply of material. Do they re-test if suppliers change?


Anonymous said...

I work for a company where 6061-T6 from Brazil is forbidden, that is the first thing I have to check in the certificate of compliance before accepting any raw materials.

Anonymous said...

...and I don't know the reason why we can't use those bars from Brazil