On Monday, May 29, 2017 at 8:30:35 PM UTC-7, Jeff Liebermann wrote:
On Mon, 29 May 2017 08:40:13 -0700 (PDT), wrote:
The inside layers are the sources of danger. High end bikes use many very thin
layers and the lay-up is completed as rapidly as possible. A bubble can be
caught or a section of cloth may not be properly wetted. This makes an area
where movement between the layers can begin and this can break down the
surrounding resins.

Very true. However, this "movement" is what allows ultrasonic and
vibration testing to find flaws in adhesive connections much easier
than finding breaks in welded connections. If there is no mechanical
connection between parts, ultrasonics will not transfer between them.
Even if the two parts are physically touching and comprise an
interference fit, ultrasonics will not couple all the energy through
the connection. Only when there is some adhesive, that will transfer
the energy across the boundary zone, will there be coupling. This can
easily be seen on a pass through or reflection ultrasonic test. It
will also show up on a vibration test, where the resonance pattern
will be very difficult if there are voids or discontinuities in the
bonding.

Presumably, such testing is done at the factory, but might be lacking
or sample tested to cut costs. I don't know. I've never visited
China to see for myself. My guess(tm) is that the inspection is so
simple, quick, cheap, and easy, that it would foolish to eliminate it.
At worst, simply shaking or banging on the frame, and looking at the
resonance waveform on an FFT vibration analyzer, to compare it with a
known good frame, would be sufficient to detect a problem.

Incidentally, in my dealings with a local plastic furniture factory, I
ran into similar testing. The testing wasn't to detect glue voids or
other imperfections. It was to detect when someone forgot to glue a
joint. That happened a few years previously causing a chair to
collapse under someone, allegedly causing grievous injuries sufficient
to inspire expensive personal injury litigation. After that was over,
I'm fairly sure that the factory considered QA and inspection rather
seriously.

Jeff - on the Apple assembly line they were actually whipping workers to work faster. Do you really think that that attitude is somehow less prevalent with bicycle manufacturers? I would bet that one out of ten bikes are quality tested with ultrasound or vibrational testing which I would have little faith in since the lay-ups are so thin that if the problem isn't obvious you would be hard pressed to detect it with vibration. Remember that the inside and outside coats are probably fine. A small bubble couldn't be detected by either method.