> Here's a video that shows that when you finish up the bowline with the Yosemite finish, you are actually just untying the bowline
Rather, it's possible in mis-setting the knot, to destroy its integrity.
But there is a similar finish --take the tail around the eye leg in the
opposite direction-- that isn't vulnerable to this (mis)behavior.
> The function of most knots depends on their being properly set
> // ... properly set and dressed
Which is seldom explained in detail (or in any way much beyond
the vague advice to "dress and set the knot") in any source!
> Why people don't just use a simple bowline,
> which has been used by generations of climbers
> (and many more generations of sailors) escape me.
It should not escape you : the reason(s) why have been pretty well
discussed (and there are some posts in this thread giving further
testimony) --it can loosen and come untied, w/o some additional
security. Generations of sailors never used rockclimbing / caving
kernmantle ropes.
> To note: it's easy to check whether you have correctly tied the yosemite
> or edwards variants of the bowline by turning the knot over after tying it.
It's a shame that the bowline is typically shown from the perspective
(re which *face*) that it is --it SHOULD be oriented to that "turned over"
face, where more of the workings of the knot are easily seen! Consider
how the sheet bend is commonly presented --from THAT (turned over)
orientation ; the bowline should likewise be presented (and tied) that way.
I.p., the quick-tying method with the twist of the wrist works better
in this turned-over orientation.
>> As long as you don't put anything "in" the knot, like your harness,
>> or a carabiner tied to a stand, you actually just have a loop [rather, "the un-knot"] according
>> to the official science - but - as soon as you put something in there
>> that cannot be threaded through the knot, it's no longer a loop.
>
> I put it to you that this is trivially wrong. Tie an alpine butterfly and then load both ends
> without putting anything in the loop. I'd wager the rope will break before it will revert to its unknotted state.
You might want to stipulate that the material isn't HMPE rope,
then --not sure that that slippery stuff would hold to rupture vs. slipping,
but it might. (E.g., a double bowline can spill materiall around the doubled
rabbit hole and out, collapsing the eye (!!) in such material, in slow-pull loading.)
> Most bends can also be made into a loop knot.
--or all ?! It's a matter of taking a cookie-cutter view of the *knot*
(tangled area) and then specifying how the *ends* connect or not.
(Or, alternatively, of the angle of loading, neverminding whether e.g.
the legs of the "loop knot" actually connect to form an eye ; the behavior
of the knot depends only on their angle of incidence.)
> dangerously close to "manipulate a knot sufficiently and you can untie it".
> We already know that!
Except that one can't (with ends taken away from access) untie some
"knots", such as an overhand or figure eight or bowline. (Interestingly,
one can re-arrange the bowline to be "directional" in the opposite direction.)
> that the friction in minimal. In that (impractical) case,
> you could pull on the perfect Yosemite bowline in such a way
> that you get the simple figure-of-8 without rethreading.
As noted above, the situation of low friction is not entirely impractical :
it is seen in some modern cordage fibres already.
But you couldn't pull on the YoBowl in a normal way and get a fig.8;
you could pull on the two ends, and get an unknotted rope, though!
> On the rope I was testing with (9mm half rope) this slid at 2.3kN ...
It's this sliding, in dynamic loading, that concerns me re frictional heat.
> [it] continues to tighten as a [noose] which eventually failed at 9.4kN which seems quite an reasonable value for a somewhat older rope of this diameter. I tested a standard bowline with stopper to give a baseline and this failed at 7.9kN
(Why the stopper?)
These values are intriguing. Assuming the bowline's strength to be
about 65% of tensile, your one check would put the rope strength
at about 12.1 kn, and the noose at 77% .
> At no stage does there seem any potential for the mess to change into an 8 and this in fact seems impossible without loosening and manipulating the knot.
It depends on how the YoBowl is transformed by the setting
on its tail : it is possible to disorient it so that the tail is NOT within
the rabbit hole --it is pulled across/under the mainline--,
and THEN there is a topological (if not geometrical) "8" in the end,
which itself might take some further mangling to get into recognizable
form (vs. a sort of *two-circles* orientation).
> Ring pulled there seems no difference between the two ways (the correctly dressed Yosemite and the loop pulled up through variant) as they broke at 19.8kN and 20.1kN respectively.
Wow, THIS is a surprisingly high value --let's average it as 20kN.
Assume 12 kN for the rope (per above deduction)
and that the closed ring so loaded takes a severe bias in sides' tension
--i.e., that knot compression adds length (lessens tension) on its side
(which doesn't flow around the frictive end points of the tape sling
(and what's opposite this : a test device smooth metal pin?!))--,
and take the unknotted side to near 100% (or nearly 12kN);
that leaves 8kN for the knot,
putting the sheet-bend-like (rather : Lapp-knot(reverse sheet bend)-like)
knots at equal strength to the bowline. !!?
Hmmmm, that's a surprise.
> If [the] Yosemite bowline ... spare end falls out of the last tuck you revert naturally enough to a normal bowline.
Again, this depends upon how disoriented the YoBowl becomes
upon that tail-pull setting --it's possible to end up otherwise
(as in the _On Rope (1st ed)_ Knots chapter cover page image.
*kN*