Kraig Willett writes:
>>> Cool stem failure video (165 kb *.wmv file) here:
http://www.biketechreview.com/stems/images/stemage.WMV
>>> and the accompanying blurb here:
http://www.biketechreview.com/stems/extremeload.htm
>> It's not obvious from reading the explanation what the load and load cycling was during
>> this test.
> It's a blurb that occupied a free weekend some time ago - it's not a journal article.
I don't understand what you mean by that. If you did tested the stem then you might give a few
sentences on the parameters involved.
>> I take it th buzzing noise is the frequency of load cycles
> Nope, it's an air compressor.
I thought the sound track had something to do with the test. That was a bit misleading.
>> but I think it might be helpful if we knew:
>> 1. What the static load used was.
>> 2. What the cyclic varying load was.
> to ~1000lbs
>> 3. What the frequency was.
> Pressure on for around 4 seconds and off for four seconds (you can probably prove me wrong by
> looking closely at the video, but that is approximately correct from memory).
Oh! That was the blurriness of the picture. Usually such tests are done with a steady load with a
superimposed vibratory load so that a reasonable statistical number of load repetitions are
achieved. 1000lbs in the direction perceptible from the video is not a realistic load for a
handlebar stem considering that stems are loaded by torsion more than any other load.
>> 4. What that means in millions of stress cycles.
> Definitively, beyond a reasonable doubt with a 99% confidence level? - probably not a whole lot
> unless one has lots of time and money. I have neither. Some might argue that a protocol that does
> not go into the millions of cycles acceptable for comparative and developmental purposes as long
> as the failure modes are representative.
There were no million cycles at the rate you mention, or the test would still be underway. The
failure mode using 1000lbs may not have anything to do with failure in use, or it may do so but as I
said, this is not the principal loading on a stem.
> FWIW, the failure mode in this stem was the same as a CT stem that I tested for several hundred
> thousand cycles.
How was that test constructed?
>> Without that, I'm no wiser than before seeing the test or its result.
> But you were pretty wise to begin with, and I didn't write it for you specifically. It was aimed
> at a much broader audience who might not even know that stems can _possibly_ fail - or where to
> look when inspecting one.
In condescending to a less informed audience, you left out essential elements of a credible failure
test. If you write it up, it should contain the information essential to the test and its
credibility.
> C'mon Jobst, it wasn't even a _little_ cool seeing that stem snap like a twig?
If you doubles the load, do you think it would do otherwise? How about 1/4 the load and over a
larger number of cycles? I'm sure I can cause a failure in any commercial stem with 1000lbs
test load.
>> From the failure pictures I see that the weld bead broke rather than tearing out parts of the
>> base metal but that still tells me little.
>> Please explain.
> Any other questions?
I take it you came into this with foregone conclusions and don't want to go back and reconsider any
of it. This is not giving the stem a fair hearing. I am not convinced of your assessment on the
basis of this test. It may or may not be a good stem but without comparing it to some other well
accepted stem, the test is not revealing anything useful.
Jobst Brandt
[email protected]