I have personally seen 3 tost ring failures at my gliding club this season,
(the
last one only about 3 hours ago!) these were all the large rings on the
strops,
not the ring pair that goes to the release, unfortunately I have not been
able
to recover any of the rings to see the failure mode, however I was
examining
the rings on the other strops and found a couple that were obviously of
welded construction, and one was clearly oval rather than round, I don't
know
the age of them but they get very heavy use.
Ben.

At 03:16 07 June 2017, Frank Whiteley wrote:
On Tuesday, June 6, 2017 at 2:51:26 PM UTC-6, John Carlyle wrote:
Frank,
=20
When there are "conical male/female shapes at the failure point" it
gener=
ally indicates tensile overload failure (caused by exceeding the
materials
=
tensile strength). A giveaway would be necking on either side of these
shap=
es, which might be what you're calling failure "at the weld taper".=20
=20
Fatigue looks very different - the fracture surface shows a smaller
smoot=
h portion surrounding the origin, with the remainder of the fracture
surfac=
e appearing granular (perhaps what you're calling calling
"crystallization"=
.. Fatigue failure is caused by thousands, perhaps tens of thousands, of
loa=
d variation in the presence of a small defect in the part (usually on the
s=
urface).=20
=20
These are two completely different failure modes, and have two
completely=
different causes. It would be interesting to see pictures of the actual
fa=
iled ring and the fracture surface.=20
=20
-John =20
=20
On Tuesday, June 6, 2017 at 4:26:34 PM UTC-4, Frank Whiteley wrote:
We had a TOST large ring failure while winch launching at a CAP
Fligh=
t Academy a few years ago. It failed at the weld taper and the ring also
b=
roke opposite the weldment with the appearance of fatigue crystallization
a=
t the break. So, it's possible several launches were done with the
failed
=
weld before the ring failed at the second point. It was reported back to
t=
he vendor and also to TOST IIRC. Locally we discussed if perhaps some
coun=
terfeit supplies were getting into the channel. The academy leader
passed
=
away some time after the academy so I don't know the result of the report
b=
ut perhaps the vendor will recall the event. I'll check.
=20
Frank Whiteley
=20
I'll amend my post to say that it was not evident that the failed
ring
=
was welded, but that separation had conical male/female shapes at the
failu=
re point. Looking for images in the archives, but nothing so far.
=20
Frank Whiteley

John,

I think you've described it accurately. Still checking around for images.

=
They may not exist. This happened around 2009 and the rings certainly
had
=
not seen that many cycles. Not aware that anyone attempted to load test
th=
em, but who knows?

Thanks for your comments.

Frank Whiteley