In Search Of The Holy Grail Bowline: A Bowline For All Reasons

[alpineer]

   You have a point here. However, why not have a secure nipping structure AND a secure collar structure, at the same time ? Regarding good things, most of the time "two" is better than "one". Bowline is a knot made by two good components : the nipping loop and the collar. If we decide to go forward, it seems reasonable to me to use both feet / to improve the nipping loop AND the collar. When you replace the tyres of your car, either of the front or the rear wheels, you replace both of them, don't you ? I never use one single anchor, or one single mooring line - and I admit there is a human psychology aspect in my thinking on this ! 

Simplicity, Simplicity, Simplicity. Complicating a structure further generally results in more bulk, more time required for tying, and more opportunities for mismanaging the tying process . At what point do you stop adding complexity to your structure? Do "we" decide to go "forward" further still, and take agent smith's cue and throw in a Yosemite finish? If you feel the need, you have those as options. 

[alpineer]

  Simplicity is imperative to me, too

   You had me fooled, X1!

[X1]

  At what point do you stop adding complexity to your structure?

  Good question ! 
  "I" would say, when "we" decide to move, "we" have to move both feet, just one step each. "I" think that the bowline is a two-components knot - so if "we" decide to go forward,"we" have to tie a more complex nipping structure AND a more complex collar structure. How much "more complex" ?  Nec plus quam minimum.
  Simplicity is a most complex thing do define ! Is an astronomical model that explains the orbits of planets with dozens of more simple cycles and epicycles, more simple than a model that uses only a few more complex ellipses ?
   More to the point, "I" tend to believe that the more complex nipping structure offers safety in relation to strength, and the more complex collar structure offers safety in relation to slippage. "I" have not seen any experimental proof that a double nipping loop, of whatever shape ( be it an Open Elbow, an untwisted or twisted Pretzel, a Clove hitch, a Girth hitch, a Constrictor... ) nips the penetrating eye legs harder and more efficiently than a single nipping loop. "I" think that the double nipping loop is needed for its greater strength and self-stabilizing properties, rather than for its greater gripping power. The main obstacle against slippage is the collar - so, if "we" wish something beyond the common bowline, "we" have to improve the nipping loop AND the collar.
   The two collars offered by the Janus bowlines, the two collars offered by the Double Collar bowlines, like the Mirrored bowline ( a Girth hitch(ed) Double Collar bowline), the braided collar offered by the Braided bowline, the Yosemite finish bowlines, all are means to the same end, a more secure collar structure. The problem is how to combine one of those solutions with one of the solutions that throw the other shoe, that of the safer nipping structure.
    To me, the common or "Eskimo" bowlines are perfectly safe 99 % of the time - so I think we are talking only for the rare rest of the cases, for people that are not going to change their routine, the retraced fig.8, for the next hundred years (at least), for people that do not read any post in this Forum, and for ourselves. So, let us not get too exited in defending our personal beliefs about what would be the best mousetrap. 

[alpineer]

As for "not critical ...", there should be something definite
said here about dressing --not an either/or/anything-goes
indication, as this seems to be.  (E.g., one could ask for one
form --for checking correct crossings--, and then specify that
this more open form be set into something more compact.
And, yes, it might be that the entanglement tolerates open
or compact dressings --which could be helpful if working with
such stiff rope that compactness was not easily achievable
(thinking of some ornery BW II old rope I have beside me!).)

I definitely prefer the open form, but expected tyers would "instinctively" scrunch the nip. But yes, I was not clear enough in stating that the nip should be left in the open form, although my photo #4 clearly suggests this dressed and set form. Or, I could have stated with more elaboration that it's definitely not critical . I will try to confirm with more testing.
Were you able to tame that old ornery B W II stuff?   

And among these tests was one (or more) that entailed breaking
the knot, such that your assertion of strength compared with the
(common / #1010) bowline has a basis?  --surprising my guess.
What material was used, and how many tests (for strength)?
(And did you set the knot so (overly) snug as SS369 shows?!)

The first strength comparison tests, admittedly crude, were done with cotton thread of expectedly non-uniform nature (what I had at hand that I could break by hand). Maybe a half-dozen tests were done, enough to satisfy my curiousity for the moment and cause me some mild disappointment re the results. The higher load tests were done with a friend, and used small dia. nylon accessory. The test appartus consisted of a screw-eye in a ceiling beam and a 3 ft. long section of 2x4 placed through the test loop for us to stand on and create the necessary forces to break the knot. Only two strength tests were completed because of the lack of control we had over the test apparatus. A minor injury was sustained by the friend. For safety reasons alone this is not a good means for testing knots.
  Those tests are preliminary at best. I would like to conduct more testing that would support the gathering of more relevant data.

Re: the strength test results, I too, was surprised by the findings.
Re: SuperSnug369 test specs, I believe I did in the cotton thread. The results were the same. I'll retest at some point.


 

[Dan_Lehman]

I'm having trouble with this "Elbow" [capitalized!] structure :
should it be familiar to us?  --else, any references to cite?
--dl*
====

Really?

http://en.wikibooks.org/wiki/Knots/Components
http://www.ropebook.com/information/knots/terminology
http://www.itstactical.com/skillcom/knots/knot-of-the-week-terminology/

Really.  (To that first site I put in some corrections --no telling
whether they'll survive ("bitter end" my arse  ).)

I see a helical structure, and don't succumb to dumbing down
to two-dimensional simplistic-ness.

You mention "no torsion" : I suggest that you do your loading
in braided rope and check this --I think you'll see obvious
torsion, with unloaded strands arc'd upwards in obvious
non-tension (there goes 50% !) and the opposite-handed
strands tight.  Once can see this in a fig.8 end-2-end/eyeknot.
Maybe this helps give #1010 its gain (it's lack of loss)?!
(Note that torsion in laid rope should affect the rope as
a whole --i.e., all of fibres gain or lose--, whereas braided
construction has fibres going in different directions!)

Now, the question of "Holy Grail"ness was put in particular
by seeking comparison of the OP knot with the mirrored bowline
--one that seems to have good slack-security and tested-in-HMPE
(lacking final tuck, alas) high-load (to rupture) security
(devoid of SuperSnugging!  ).  --as one example.


--dl*
====

[Dan_Lehman]

   It would be most desirable to get the security feature of a Bowline derived knot up front
rather than at the tail end of the tying process. There's a human psychology aspect to my thinking on this.
Considering reports of accidents implicating failed Bowlines, I suspect that most were a result
of not tying a tail security feature.
.:. Better for the security feature to be first --before tying the Bowline--  rather than last,

   You have a point here.  However, why not have a secure nipping structure
AND a secure collar structure, at the same time ? Regarding good things,
most of the time "two" is better than "one".  Bowline is a knot made by two
good components : the nipping loop and the collar.  If we decide to go forward,
it seems reasonable to me to use both feet / to improve the nipping loop AND the collar.

Heinz Prohaska --responding to the infamous (but often mis-told)
Lynn-Hill-falling-unknotted story (she didn't tie her intended bowine)
suggested using the anchor bend (fish.bend) as the base, which
would give a sort of Prohgrip /Blake's hitch grip to the tail inserted
through it --security at first closure of the tie-in eye!  His full knot makes
a "proper collar", IIRC.  Oh, the insertion of the tail to form/close the
eye is in the same direction that it goes in completing the base knot.
That sounds like about as "up front" as one can get --SOMEthing has
to be tied, after all.

For myself, I'm looking at two-stage knots, where the first stage
is some sort of hitch, hoped to be able to serve securely should
everything else somehow fall apart --though one must presume
that that situation implies that the base hitch too will be loose
(for how else ... ?) !  I've come to the conclusion that those
compound knots that don't look so good on this condition
nevertheless give slack-security so great that there is no need
for their being this back-up security element; but there is
one that seems secure when *open* --ironically, it is less
accommodating of forming so secure a compound structure!

These are ungainly and complex-looking, but their tying is
one of building upon simpler elements; the knowledge of
what one is doing should make this feasible.  The payoff
is having a tie-in that (1) securely endures to hold, (2)
might give better load absorption (in not so tightening
that is loses knot-compression), and (3) is easy to untie.


--dl*
====

[X1]

t...wo-stage knots, where the first stage is some sort of hitch, hoped to be able to serve securely should everything else somehow fall apart --though one must presume that that situation implies that the base hitch too will be loose

 A hitch tied on the returning eye leg around the rim of the nipping loop can serve this purpose - so, even if the nipping loop itself opens up, something would still remain closed !

These are ungainly and complex-looking, but their tying is one of building upon simpler elements; the knowledge of what one is doing should make this feasible. 

  If the "nipping structure" and/or the "collar structure", as a whole, happens to resemble a well-known knot, there would be a real or virtual aspect and sense of "simplicity", which would help the average knot tyer to remember how to tie, and to tie, the post-eye-tiable eyeknot utilizing it/them easily - a ready-made pattern, a built-in mnemonic, so to speak. That is why it pays to examine each of the two parts, the "nipping structure" and the "collar structure", as separate entities, although their entanglement into the "compound" finished eyeknot is usually very complex. That is why I have argued that the Constrictor bowline is, in fact, not less simple than the Tresse bowline - or, at least, it is not much more complex, to the point one should not consider it as a practical and convenient solution.

[alpineer]

Really?
http://en.wikibooks.org/wiki/Knots/Components
http://www.ropebook.com/information/knots/terminology
http://www.itstactical.com/skillcom/knots/knot-of-the-week-terminology/

Really.  (To that first site I put in some corrections --no telling
whether they'll survive ("bitter end" my arse  ).)

I see a helical structure, and don't succumb to dumbing down
to two-dimensional simplistic-ness.

You mention "no torsion" : I suggest that you do your loading
in braided rope and check this --I think you'll see obvious
torsion, with unloaded strands arc'd upwards in obvious
non-tension (there goes 50% !) and the opposite-handed
strands tight.  Once can see this in a fig.8 end-2-end/eyeknot.
Maybe this helps give #1010 its gain (it's lack of loss)?!
(Note that torsion in laid rope should affect the rope as
a whole --i.e., all of fibres gain or lose--, whereas braided
construction has fibres going in different directions!)

Now, the question of "Holy Grail"ness was put in particular
by seeking comparison of the OP knot with the mirrored bowline
--one that seems to have good slack-security and tested-in-HMPE
(lacking final tuck, alas) high-load (to rupture) security
(devoid of SuperSnugging!  ).  --as one example.


--dl*
====

   It's better for your health and well-being to succumb to "two-dimensional simplistic-ness" than to bitterness, Dan. I stated at the beginning of this thread that the proper elbow contains 720 deg. line torsion. I also see a helical structure. The tresse bowline has no helical structure (certainly not in the same sense). The line's construction is of no matter in the context of this discussion.
   The tresse bowline's elbow is not a "proper elbow". Nevertheless, it's as legitimate a structure as any other elemental form from which knots are built, and deserves to be acknowledged as such. It needs a name, and because of it's two-dimensional resemblance to the proper elbow (but without torsion) I call it open elbow. It doesn't contain any torsion*. Have you not coiled a power cord using the alternating handedness method? The coils exhibit local line torsion only. Throw the coils back out and all that torsion diappears, period. The "tressed elbow" elemental form doesn't contain even local torsion.

* Unless you want to be nit-picky about the small amount of torsion initially required for a three-dimensional linear structure to cross itself, it's perfectly reasonable to ignore this incurred torsion.

 

[James Petersen]

I'm having trouble with this "Elbow" [capitalized!] structure :
should it be familiar to us?  --else, any references to cite?
--dl*
====

Really?

http://en.wikibooks.org/wiki/Knots/Components
http://www.ropebook.com/information/knots/terminology
http://www.itstactical.com/skillcom/knots/knot-of-the-week-terminology/

Really.  (To that first site I put in some corrections --no telling
whether they'll survive ("bitter end" my arse  ).)
==

... "in ignorance of its history" didn't make it through the review -- guess it wasn't politically correct. 

[alpineer]

WARNING!  It should go without saying, but nevertheless it cannot be overstated,

--from none to all !!   
 

Wrong. "from none" would be "It goes". Such criticism! Over nothing. Or am I being too sensitive?

alpineer

[alpineer]

It's doubtful that a particular Bowline will ever be discovered that would achieve
"Holy Grail" status, but I would argue that the Tresse Bowline can come closest
to the ideal "Bowline for all Reasons", more than any other.

    I'm remiss in not asking : do "all Reasons" go beyond practical ?!!

--for this thread to be in (mere) Chit Chat and not Practical forum!
(--or Explorations..., New...)


 

  Yes, you would be remiss! No qualms here over it being moved to Practical.

[alpineer]

Of particular interest to me is how the knot would perform in HMPE Spectra/Dyneema.
Any takers?

For starters, there is the doubt that knotting such material
is at all worthwhile in terms of strength --all testing of it
so far has cast doubt as to knots being anywhere reasonably
strong as we would expect, from historical materials.
Of particular interest should be this video of Brion Toss's test
of not-quite-my-recommended "mirrored bowline" --he omitted
the reflected collar, alas--:
www.youtube.com/watch?v=Fy6Y2Xoo4Ak  [ in 5/32" HMPE 12-strand ]
.:. the eye knot holds, but breaks at what Toss guesses
is not-so-high load --the device wasn't calibrated(?!).

Yes Dan. But it's the Spectra Bowline 1 test video http://www.youtube.com/watch?v=ZFRQcExLA34 that I'm interested in for the very reasons you quote. I want to know; will the Tresse Bowline hold the eye's size and capture more of that elusive strength component?

alpineer

[Dan_Lehman]

Yes Dan. But it's the "Spectra Bowline 1" [= dbl.bwl + stopper] test video
http://www.youtube.com/watch?v=ZFRQcExLA34
that I'm interested in for the very reasons you quote.
I want to know; will the Tresse Bowline hold the eye's size and
capture more of that elusive strength component?

alpineer

Note that the double bowline shown in this video has its tail
stoppered (with what appears to be an overhand knot) ; IIRC,
that was not the case with the incomplete mirrored bowline (!).
Perhaps the "tresse bowline" --or the like knots I've shown above--
will tighten upon its SPart so as to prevent the material from
being drawn out under loading.  (Btw, I wonder how the knot
so slipping will behave in dynamic loading --I don't believe that
the slippage will become adequately fast ... !)  But one can't
hold much hope for strength, given the hard U-turn, I think.
(My hope for the "locktight" eyeknot's strength looks to there
being enough friction within the coil to draw off load over the
rather broad area of compression --perhaps something that,
if possible, varies with rate of loading : dynamic loading cheating
this hoped-for frictional assistance?)


When I saw this video, I was shocked --all bets were off,
my ideas of strength-building / material-behavior shattered :
"Huh, it can DO this??!".   
I could see that the gradual-bends tactic would amount to just
some wavy route to the U-turn w/o consequence (or maybe
there would be a price in torsion to pay!), just as running
a line over a good block puts in a U-turn but hardly diminishes
the force delivered beyond this point (i.e., no off-loading of
force in the block).

So, I didn't hold out much hope for what were already quite
inefficient-in-material-use knots of a tortuous SPart path
hoping to offload force over its long run --no, that would
all slide too well.  I did test one knot with such a *theme*,
but whereas I had in mind testing the fig.9 eyeknot (which
one test report suggested --knot name wasn't clear-- had done
well (for HMPE, which means about 50% !)), I opted to load
this structure in reverse, looking at that geometry being
one that gave a broader initial U-turn, and then ... <just hope?>.

In fact, this was my strongest knot in the test of 5 eyeknots;
but only by so much, and absolutely nothing to shout about
(= +/- 42% of "capwrap" tensile --a precise-looking value),
with others in the range from 33%-ish).

Interestingly, it appears that the break in the reversed fig.9
eyeknot came a little past the U-turn --or late in the turn--,
at a point where compression against the eye legs was great,
and maybe some movement added a kick of frictional heat?
(see attached photo of the broken & opp-end unbroken,
equally tied & pulley-set knot, thread-marked [nb, Agent_Smith :
Dan takes his own medicine!    --to some benefit, yes?!] )

In the attached photo, one can see the absence of the
ruptured SPart (lower knot) in the void of the *interior*
of twin strands flowing from the left through twin collars
(these are the U-turn of the eye legs --remember, it is
the reverse of the better-known "fig.9"),
and around and up over the knot to wrap the eye legs.
It appears that the break comes after a half-wrap?!

My thread-marked points are color coded, and as follows:

W) white threads (bit yellowed in image, a white-balance issue)
mark the point where the SPart & resp. eye-legs exit the knot;
these will show how much material yield there is on knot
compression, and any slippage (I think that most reports
call this "slippage" which to me seems wrong-headed --there
must be some feed out of material when the knot compresses,
but I don't see that as "slippage", which has connotations
of insecurity --but perhaps its all a matter of balance, hmmm?!).

P) Pink-red threads --one per knot-- mark the suspected
point of rupture (my surmise (what have we learned if we
can't surmise accurately --and the guess & test game should
be an iterative one of improved surmise!).

G) Golden threads mark a 2nd-guess rupture point,
based on some surmise of the yield of material to compression.

What I see in this photo is that the SPart entered the nub,
moved rightwards -> upwards -> back downwards -> and
then in coming up around towards the viewer broke at
about the low point of this wrap, as viewed.  The puff of
white ruptured fibre ends will fold down to about this
point, and the examination of the broken SPart vis-a-vis
its markers with the intact upper knot similarly tied &
loaded also place rupture here.

The P threads started out nearly where the G threads are
at rupture --I'd say that they were just a little before
this point (i.e., slightly closer to exit, roughly right atop
the 2nd-crossed (SPart flowing into nub) twin eye leg).
And the break was even farther along the SPart's path
than this point, by about a half wrap (180deg).

Note that eye legs pulled out just a little, evenly.  And
perhaps the only real measure of "slippage" is the white
thread at the tail --was it pulled into the nub?  Here, it
is not; perhaps a mm or two at most.


--dl*
====

[alpineer]

It is interesting to note that the very similarly curved SPart
of the knot (my "#56" shown in my next msg.) broke only
1 strand but had the same strength; one might have thought
that if the forces could be so evenly apportioned to break two,
it would almost necessarily be at a greater force!  (Perhaps
there are other factors at work to explain these results.)]

I would surmise that a second strand was on the verge of failing just before the first strand failed. Beyond that, I would think there's not much to say which would be of any statistical significance from a single sample.

[X1]

   I have tested the four variations of the Tresse bowline : the "left" and the "right" hand ones, and the "under" and "over" ones ( "under" / "over", regarding the way the first diagonal element of the "8" shape nipping structure crosses the second - the variation shown at the first post is the "under" ). I can confirm that this nipping structure is very stable, and grips the penetrating segments very efficiently, EVEN WITH A LOOSE COLLAR. The mere presence of the two legs of the collar, even if they are not tightened, is sufficient to keep the returning eye leg in place. I have seen that the "under" variation is better, indeed, because the standing part remains closer to the axis of loading of the returning eye leg, and does not twist it as much as it does in the "over" variation. I have also seen that the tail of the "left hand" Tresse bowline is squeezed more efficiently by the standing part s first curve, AND by the returning eye leg s first curve ( the variation shown in the first post is the "right hand", which I believe is a little inferior regarding that matter ).
   So, I believe that all that is needed in order to have a "safe" bowline, is a more safe collar. The only way I have been able to improve the collar, and retain the basic element of this bowline, its double nipping structure, is the braided collar shown in (1), copied from the ingenious "Braided Bowline" presented at (2).

1. http://igkt.net/sm/index.php?topic=4321.msg27020#msg27020
2. http://igkt.net/sm/index.php?topic=4283.msg26651#msg26651