Ring loads on Bowlines

In the paper I linked to in this thread, the simple bowline [1010] and the left-hand (cowboy) bowline [1034.5] were load-tested. Ashely seems to have disparaged the cowboy bowline, and at least some others that I know of followed, without quoting any evidence.

I know of no evidence other than my own observation of the two loops behaving differently as to the ease of untying after being loaded or resistance (or rather lack thereof) to unraveling when shaken with little load - perhaps they do and the #1010 has some advantage in these departments, although I haven’t observed it.

The evidence in the article seems to show that the two loops are identical under standard load but highly significantly different when subjected to ring load (I imagine with the regular standing end unloaded completely) - both come undone, but the #1034.5 requires some 5 times higher load to accomplish that (and I think importantly the #1010 came undone at the loads equivalent to some 200kgf - not that difficult to accomplish in practice).

Here is an illustrative video by RopeLabs that show the slipping and a similar trend (less pronounced in this one test than the difference reported in the paper):

https://www.youtube.com/watch?v=Mf0wZRxNcQ0

This raises an interesting question: there are 4 ways of locking the bowline in the ‘woven’ fashion (Scott’s lock) - Mark Gommers illustrated that somewhere in the past and it makes sense topologically. Two of them start with #1010 and the other two with #1034.5 (the difference in locks is that one version wraps the WEND? (I like the name) around the loop part prior to weaving it locked and the other one doesn’t). Is there any evidence that those locked constructs inherit the properties of their base knots when ring loaded?

Referencing this research paper: https://www.researchgate.net/publication/344689340_Revision_of_Commonly_Used_Loop_Knots_Efficiencies
And replying to mcjtom…

Its hard to keep the word count to a minimum when critiquing a published technical paper.
Therefore - all the usual warnings of ‘don’t read - too much technical content ahead’ applies!
Although this is a technical forum dedicated to knot geeks and all things knotting - so in my estimation, the long word count
ought to be acceptable given the nature of the viewers/visitors demographic.

The problem with virtually all of these type of papers authored by academia - is that the primary underlying premise is that MBS yield (strength) is the only way to measure and understand knots. It is almost a pathological condition (particularly within the broader context of using knots in life critical applications - eg climbing/canyoning/caving/rope rescue, etc).

The authors of the paper defined ‘knot efficiency’ as:
Knot efficiency is defined as a proportion
of static breaking strength of a rope in which
the knot is tied (marked as x) and static breaking
strength of the same rope without the knot
(marked as y).
Knot efficiency (n = x/y) is usually expressed
in a percentage.

This definition is cliched and just plain wrong in my view.
And yet, it dominates the thought processes of most laypeople (ie it is the default way of thinking about knots).

Knot ‘efficiency’ (in my view) is measured by the following metrics:
amount of rope consumed to form the knot (less is better)
footprint (overall size/volume) of the knot core
stability and security
resistance to jamming
utility and loading profiles (the number and effectiveness of potential applications)
TIB (whether the knot is also Tiable In the Bight)

Interesting note: At para 3.3.7 the authors actually touch on these metrics - but fail to synthesize their significance in terms of ‘efficiency’.

The authors also define knot terminology incorrectly.
For example, their definition of the term ‘tied on the bight’ is muddled.

Their use of the term ‘wend’ is also peculiar. Tail is a better term in the context of what ‘emerges’ from the completed knot core - and has zero load.
The term ‘wend’ was invented to describe the end of the rope while it is being manipulated to tie and form the knot (which is a temporary state).
Their use of the term ‘stand’ could have been improved to SPart (standing part) - a bit of digging around could have uncovered this term.

The authors also use the term ‘ring loading’ - which (in my view) is not properly defined.
Ring loading appears to be constrained to an eye knot (ie loop knot).
And the authors appear to assume a transverse loading direction.
Which actually means the eye is loaded perpendicular with respect to the axial alignment of the SPart (ie at a 90 degree angle with respect to the SPart).
In my view, some assumptions were made.
The term ring loading could be construed as circumferential loading (ie hoop stress).
Note that an ‘eye’ of a knot can also be loaded axially with respect to the SPart (ie load is in axial alignment with the SPart) - which could be considered to be ‘normal’ loading.
I therefore hold the view that ‘ring loading’ is somewhat nebulous.
A better term (for the authors) is transverse loading.
Note that rope rescue technicians can and do load eye knots circumferentially (the Mobius Butterfly knot is efficient for this purpose).
(Meaning that the eye of the knot is pulled in all directions, outwards, which expands the eye).

And furthermore, the issue with transverse loading of the simple #1010 Bowline is that it is in fact cyclic loading which triggers rapid loss of security.
#1034 1/2 is more resistant to cyclic loading in the transverse direction. This is in contrast to loading that is ‘static’ (unchanging, and not cycling).

NOTE: There will be some members of the IGKT forum who will vehemently disagree on my proclivity for defining loading directions more accurately.
Also, knot terminology will always be contentious… for example my preference for ‘eye’ knot instead of ‘loop’ knot can cause convulsions and anger in some.
There are those who will ferociously cling to Ashley’s definitions - in no small way being influenced by tradition.

The ‘General conclusions’ (at #4) in the paper is formulated in a way that is unremarkable and almost meaningless.
For instance; … “With a probability
bordering on certainty, we may conclude that
the efficiency of loop knots is not constant as it
is implicitly presented across the vast majority of
published works, but it is a decreasing function of
static breaking strength of the rope.”

This is an unremarkable comment.

A little better is the second point:…
Furthermore, electron microscopy and high-speed thermal
imaging revealed that knotted rope is subjected to
temperatures of extraordinary extent as the knot breaks
(Sect. 3.3.10).

Although the proclivity of rope to heat up and melt is nothing new.

At 3.4 (Future research) - I would encourage the authors to closely and thoroughly examine the following:

  1. Why some knot structures are totally jam resistant (and contrasting this with knots that are prone to jamming)
  2. Dressing/geometry of a particular knot - and how this influences its jam resistance (eg different forms of #1047 F8)
  3. The influence and effect of varying the number of rope diameters inside the nipping loop of ‘Bowlines’.
  4. In eye knots that are loaded in axial alignment with respect to the SPart (ie ‘normal’ loading profile) - the effect of parallel eye legs versus eye legs set at different divergent angles.

In reference to Scott’s locked Bowline:
Yes - I consider there to be 4 different variations of Scott’s locked Bowline.
I don’t use the term ‘woven’ to describe the geometry - because this can be confused with a ‘woven Bowline’ (a different knot).
Only one of Scott’s locked Bowlines is ‘TIB’.
Yes - the Scott’s locking maneuver definitely improves resistance to cyclic transverse loading (in all 4 variations).
NOTE: When declaring there are 4 different variations of Scott’s locked Bowline, this is true within a defined chiral domain (eg Bowlines with ‘Z’ chirality).

EDIT NOTE: There are risks involved in giving feedback of a technical nature - because people are involved and there is an ever present threat
of hurting someone’s feelings. In my estimation, the key is to separate factual opinion evidence from personally directed attacks.
All I do is present the facts as I see it - but within the realm of knot science, many concepts are either evolving or challenging old traditions.
When you challenge old ideas or beliefs, it is inevitable that there will be individuals who feel angered or outraged (and so the natural human response is to retaliate).

This may make sense intuitively, but has it been ever meaningfully demonstrated/quantified, to your knowledge, especially if starting from #1010 vs #1034.5 makes a difference? It has for their ‘unlocked’ versions.

…and the passage prior to it.

A bit off topic, but I think that they show that while all knots decrease the nominal MBS of the rope and to different degrees (that’s well known), increasing the rope MBS on which they are tied further increases the proportion of the rope’s MBS that a knot is eating away. That’s novel, or at least I haven’t seen it reported before.

I’ve seen some notes about higher %s for smaller dia.,
but we need cited cases. IIRC, the Dave Richards’s testing
of 12mm low-elong. & 10mm dynamic & 7mm “accessory” (also
rather low-elong.) had pretty similar % strengths for BWL,
grapevine, fig.8 ek, butterfly, and fish & sheet bends
–though these latter sometimes slipped.

OTOH, it occurred to me long ago that just bumping e.g.
the strength for tension of a line need not have any implication
for the fibre’s resistance to compression. (There were some
remarks that hi-mod rope fibres were “weak” in compression,
but frankly if one looks at their breaks vs. nylon/polyester
I think you’ll see them as strong --it’s just that their tensile
strength is so huge that in %-of-that comparison, yeah,
they get low numbers.

(I’m currently wondering what difference it might make
were a version of HMPE made in which fibres were much
larger than the current super-fine ones!?)

–dl*

For reference, a discussion on informal testing on Sheet, LH Sheet, and Lapp bend as primitives involved in ring (transverse) loading of simple bowlines (but I’m still not sure how that translates to their locked versions other than what I can casually observe).

https://igkt.net/sm/index.php?topic=3117.0

The 2 different Bowline types ( innie or outie tail)nip differently is the way i have always seen it.
The ring loaded/cross-axis/transverse loading profile is worst angle of pulls axis on the rope structure of any standard Bowline, especially with what is usually the SPart now just a loose Bitter End rather than usually most rigidly tensioned rope part to nip against now gone. Single sided termination Hitches and dual sided continuance HHs are very directional in how they lock down. Look for critical changes in about any 90degree change in any mechanics like cross-loading eye rather than inline loading, or going from Bend(inline powered locking) to Bind(lock from forces across rope, not along rope length). Eskimo/Anti-Bowline (i L-earned as Jacked Bowline) turns it’s lock 90 degrees to accommodate the 90 degree change of force empowering lock in ring loading.
.
The tail out /leftie/cowboy style in regular usage puts primary/pre-arc forces to sandwich a vice grip on just the eye side of the collar, this switches to nip of the Bitter End by primary forces in the ring loaded /cross-axis/transverse loading profile.
.
A normal/tail in/ what would be tails, Bitter Ends same side in Sheet Bend; has no primary force type nip when loaded across eye only. And the end with bulk of rope distort around to sit as capstan for Bitter End that reels around it to freedom. Leftie/cowboy Bowline form distorts so , but primary forces nip Bitter End from reeling free so easily.
.
The secondary force on one side for nip in normal ‘innie’ Bowline ring loaded is as like secondary/lesser nips to both sides that fails in Square Knot used as Bend, but here applied to 1side that isn’t Bitter End. Turning force for lock 90 degrees for Round Binding makes Square usable for binding giving primary force locks that Bend usage doesn’t. Sheet Bend turns 1 Bitter End of Square 90 degrees(instead of the force turned 90) to give primary force lock to convert to Sheet then can be used as Bend, where Square can’t.
.
They say a journey can be greater than the destination. Perhaps knot strength/efficiency is over rated in nominal usages with plenty of headroom and no impacting. But, knot strength/efficiency is a readout of what is going on inside a knot that we could not otherwise know. It is how we test and rate structures in other materials and also shows rope is just another structural material following and teaching the same rules in the end. We have L-earned much in this chase, just as racing has improved car efficiency and safety, even tho (most) drive differently on road in most popular car usage, safely inside the now defined power band, not at extremes. Testing such limits is kinda how we feel our way along, and number to compare outcomes.
.
Untying fails I would look at as not so much percentage of tensile fails, like do for breakage fail. Seems friction change would more likely alter untie fail than tensile fail.