Author Topic: Omnidirectional-loaded loop with omnidirectional tails?  (Read 1824 times)

agent_smith

  • Sr. Member
  • *****
  • Posts: 1177
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #30 on: May 09, 2020, 09:55:19 AM »
per jimmyh:

I recommend that you start a new topic thread.

The content has drifted off topic and has at times become a personal narrative.

For example, you have attempted to use social proof to back your narrative and thus attempted to create ill will.

I'm thinking that a moderator will step in if this continues.

Quote
The behavior of a knot in ring loading is the exact same as it is when you cut the eye legs before pulling on them. I can show you a picture of a butterfly loop in one of these two configurations, and unless I zoom out enough to include the clipped ends or intact eye, you won't be able to see any difference because there isn't one. Either it works both loadings or it works for neither, as they are the same thing.

I am trying to understand how this relates to the simple question that was originally posted. You are interpreting omni to also encompass circumferential  loading of an eye - which (in my view) goes beyond the intent of the OP. It would be a stretch to assume that the OP also intended to seek a knot that could sustain loading along multiple axes - including circumferential loading of an eye... I say this because you are getting into technical territory that most ordinary users would not be concerned with.

I therefore think that your proposition re circumferential loading (hoop stress) of the eye of a Butterfly is traversing territory that belongs in a new topic post.
I also note that you appear to narrow your definition of what you call 'ring loading' to be in a singular defined axis, which would favor your apparent insistence that it mimics an offset loading profile.
By your interpretation, there appears to only one axis in which the eye can be expanded? An alternative view might be that it acts like hoop stress - and could exert an outward force in all directions simultaneously. But, if you wish to only apply an expansive force along a defined single axis to favor the notion of offset loading - that is your prerogative. But, again - you should start a new topic thread to explore your propositions.

Quote
"Omni" means "all". Either it's good for all loadings or it it's a poor suggestion for OP. Take your pick.
Hmmm,  and?

Quote
After getting around to reading some more of the current threads, it appears that this *is* normal for you now (though I thought it wasn't in the past?)
?
Another peculiar comment - this isn't normal and its personal.
I think you will receive a warning from a moderator if you persist in this type of personal narrative.

Quote
Since it doesn't seem like a fluke, I think I'm no longer going to put any effort into trying to make a conversation work with you. If it works, then great. If not, I'll just stop responding
Another peculiar comment - and one that is getting personal (again).

I think you have issues which cant be resolved in this topic thread.
This isn't a conversation - it is inappropriate behavior that is disguised in a topic thread.

I have repeatedly advised that you commence a new topic thread - one where you can expand your ideas.
If you do this - you should refrain from including thinly veiled misbehavior with your narrative.

SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 1879
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #31 on: May 09, 2020, 01:40:16 PM »
Lets return to the original topic.
Personal opinions of each other are forbidden and will not be continued. The thread is becoming diluted and not germane to the question asked.
Of course you may continue using personal messages or emails...

SS

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3903
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #32 on: May 10, 2020, 08:25:56 PM »
Quote
Would that this were so,
but you need like rope & forces and I don't
accept the transferability of % behavior of
one rope to another (we see already the cases
contradicting this).

I'm glad you don't. That makes it interesting :)

It obviously isn't going to be a *perfect* match, but how far off have you seen it get when comparing (for example) nylon vs nylon? I would be pretty shocked if a simple bowline could be made to jam in any nylon rope, or if a double overhand could be made to resist jamming in any nylon.
As I noted above, not re material per se but the
construction --i.p., kernmantle where flattening can
occur (if it's not really packed, like some caving ropes ARE)--
can influence jamming.

I don't know about "jamming" so much, but I have found
nylon (presuming) BWL(s) somewhat jammed --i.e., there
was evident tension within the knot, in the SPart (diminished
diameter, a consequence of stretching) that was not outside
of the knot, and that greater outside diameter'd rope thus
sort of was *stoppered* from entering & loosening the knot.
(The particular seashore-found case I recall was not hard
to manually loosen.)  Otherwise, though arguably beside
a pure "jamming" situation, the BWL can partially capsize
into a pile-hitch noose --SPart's nipping turn opened to be
an obvious helix vs. loop-- and some tight knots there.

And with an operational test for jamming --i.e., manual
loosening-- one likely sees jamming where per % total
load it wouldn't be counted (where my hands hurt trying
to loosen, but if I can use some tool, it comes out).

As for the strangle knot,
I'm with some ambivalence : on the one hand,
it's a knot intended to be secure-when-slack
(which something jammed sure is), but it also
has more material in it than a mere overhand,
and so ... more to work with at loosening!?
The tightness of the outer wrap(s) comes
via passage through friction and so gets
some mitigation; the simple overhand has none
and gets full(er) force (though, with less contact
patch area).
(I recall once on some jammed tangle being happy
that I could insert the tip of a hook into some strand
and then apply a pulley'd force on that hook to open
it up enough ... !)

Quote
On the other hand, if a knot breaks at 50% tensile strength in one rope and 65% in another of the same material I wouldn't be too surprised.
But in the case of different sources,
one has to question the test & reporting methods!
Is one tester basing % on rating,
another on tested value?  (Or both on rating,
but the ratings are whacko, one size rope vs. another!)

Quote
Quote
.:.  It's an interesting conjecture --provably true/false?
Well, heck, now your tying up my evening thinking!
(Right after I invent the perpetual-motion device.)

It sounds like you understood what I was getting at. I suspect it's provably true, though turning my intuitive and visual (and not foolproof) "proof" into something formal and rigorous has never been my strong point. I'll have to think if I can get closer to that.
I can add one to this : if the Tibby eye knot is TIB,
will the eye knot formed by fusing Tibby's SPart & Tail
and cutting its eye (to become new SPart/tail) also be TIB?
I know of cases that are so (Angler's/Perfection loop),
but wonder if all are.


--dl*
====

jimmyh

  • Jr. Member
  • **
  • Posts: 24
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #33 on: May 11, 2020, 05:02:19 AM »
Quote
As I noted above, not re material per se but the
construction --i.p., kernmantle where flattening can
occur (if it's not really packed, like some caving ropes ARE)--
can influence jamming.

Hm, interesting. How flattening influences jamming isn't obvious to me, but it makes sense that it could. The difference between "static" and "dynamic" nylon is construction too, and elongation would definitely be expected to influence propensity to jam.

Maybe flattened material interacts in complex ways depending on geometry, but I expect that a change in elongation would mostly preserve the relative ranking of knots and therefore would allow a lot of what is learned in the convenient test material to carry over to the end use material.

Quote
Otherwise, though arguably beside
a pure "jamming" situation, the BWL can partially capsize
into a pile-hitch noose --SPart's nipping turn opened to be
an obvious helix vs. loop-- and some tight knots there.

Yes, I probably should have been clearer that when you're measuring other properties things could change more. For example, slack security depends quite strongly on the stiffness so construction isn't a small term. When testing things at high tension (e.g. jamming, strength) the stiffness matters less because the force produced by the tension can be much higher and start to drown out any effects of the stiffness. I would expect capsizing to mostly follow the latter, but the tightness of the initial dressing depends on the stiffness and that can impact capsizing.

Quote
And with an operational test for jamming --i.e., manual
loosening-- one likely sees jamming where per % total
load it wouldn't be counted (where my hands hurt trying
to loosen, but if I can use some tool, it comes out).

Yeah, this is something I feel was missing from earlier conversation. It's a pretty soft curve there, and while plenty of climbers take big whippers and still manage to untie their figure 8s there's still room for improvement on that front by using some sort of secured bowline. When I had that butterfly loop jam on me after using it to pull a vehicle out of a ditch, I'm not entirely sure I wasn't able to recover the rope with pliers, but I know for sure that it was way harder than acceptable when there are other solutions like a TIB bowline which would have been easy to undo.

Quote
The tightness of the outer wrap(s) comes
via passage through friction and so gets
some mitigation; the simple overhand has none
and gets full(er) force (though, with less contact
patch area).

I have to admit, I'm not used to thinking about optimizing knots in this direction! My money would be on the strangle knot but you're right it's not trivial and I've never checked.

Quote
But in the case of different sources,
one has to question the test & reporting methods!
Is one tester basing % on rating,
another on tested value?  (Or both on rating,
but the ratings are whacko, one size rope vs. another!)

Yep. Science is hard, and details matter. At least knots are simple enough that with some effort we can isolate most of these variables.


Quote
Quote
Quote
.:.  It's an interesting conjecture --provably true/false?
Well, heck, now your tying up my evening thinking!
(Right after I invent the perpetual-motion device.)
It sounds like you understood what I was getting at. I suspect it's provably true, though turning my intuitive and visual (and not foolproof) "proof" into something formal and rigorous has never been my strong point. I'll have to think if I can get closer to that.
I can add one to this : if the Tibby eye knot is TIB,
will the eye knot formed by fusing Tibby's SPart & Tail
and cutting its eye (to become new SPart/tail) also be TIB?
I know of cases that are so (Angler's/Perfection loop),
but wonder if all are.
Yes, it works for them all. I thought some more about how to come up with a proof, and that's actually where I started.

If you fuse the two SPart's together you end up with an "unknot", and the only difference between a TIB method of tying the original knot vs the fused and split version is which end you're allowed to loop around. However, to get to the other side one can always loop around in the other direction. For example, doing the "fuse the SParts and cut the eye" operation to a TIB  bowline makes for a trivially TIB knot. You just place the collar right over the old SParts/new eye. It's the regular version that's a bit of a trick since you're not allowed to loop around that end, and so we just go the other way.

Continuing from there, all TIB knots become equivalent to the "unknot". With the tails joined, "through loading" just becomes "ring loading" of the new eye, and my conjecture simplifies to "for all unknots, there exists at least one eye through which ring loading is offset loading". This is the same as "...there exist at least one eye for which there no passages of rope through the eye" (because if there were, then when tensioning that passage would fall on the opposite side as the rest of the knot, and therefore the knot would not be entirely offset) or "through which there are no over/under crossings" (because over/over and under/under crossings are just over/under/around and not *through* the eye, so ring loading pushes them all to the same side of the line of tension).

However, two knot diagrams belonging to the same knot can be related through only 3 operations called "Reidemeister moves". Type 1 "Twisting", type 2 "moving a strand over another strand", or type 3 "moving a strand over crossing". The only of these that changes the crossings is type 2, and that can only add or subtract over/over and under/under and can never produce or correct an under/over as that would require moving one strand *through* the other.

Since all TIB knots are equivalent to unknots and all unknots can be tied/untied using only Reidmeister moves, if a knot is TIB, there must exist some point from where a Reidmeister move can be applied towards untying it. By virtue of allowing Reidmeister moves towards untying, there can't be any passage through this point that it can be snugged up against, and any such point must form an eye. If there exists no such point, then it cannot be untied through Reidmeister moves, isn't an unknot, and therefore isn't TIB. The contrapositive of this is my initial conjecture: if the knot *is* TIB, then there *does* exist some eye which will produce offset loading when ring loaded, and the only way to have a TIB knot which isn't offset loaded is by not using all of the loading profiles on all of the eyes.

To reiterate the relevance here, this thread is about "omni (all) directional loaded loops with omnidirectional tails". It turns out that we cannot have this in a TIB knot without having offset loading on at least one of the eyes (counting "through loading" as ring loading the "eye" of the unknot). This means, when searching for potential candidates we have a few options:

1) Relax our "omni" requirement a little bit and create an extra eye "tail" that we don't use (at least not for ring loading). This way we can still have an eye and two standing parts that can be loaded in all directions, even if not all possible loadings of our multiple eye knot.

2) Relax our "omni" requirement a lot, and give up either ring loading or through loading.

3) Accept that it won't be TIB.

4) Accept that it will be used in offset loading, and make sure it behaves itself in this way.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3903
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #34 on: May 11, 2020, 06:46:43 PM »
the proper nomenclature?
One can only go so far in knots nomenclature
before trying to figure out "proper" becomes
a major headache!  --so many (mis-/ab-)uses,
conflicts, overloadings, ... !!


Quote
As a general rule, overhands with something passing through them are more more prone to jamming when both the incoming and outgoing lines are on the same side of the thing passing through the center. As far as I can tell, there are two reasons for this. One is that the line has to bend around a larger angle before getting to the part where it binds and the capstan equation says that more angle means less tension, and less tension means less jamming. The other is that there is a bit of a "collar" formed, and there's always that space under there that doesn't get cinched down with tension. This allows loosening in the same way one would loosen a bowline.
What I refer to as the "pretzel" vs. "timber hitch" forms,
as you've described them.  Although the latter can leave
that space you note, it can be the case --esp. w/stretchy
material-- that the collar around a part gets very tight
and holds tight on the force-diminished diameter,
jamming the relaxation past that point, unstretched
material beyond this point *blocking* against the
tight collar --small space created or not!

Sometimes the pretzel OH form has enough strands
surrounded that one can loosen the structure by
wiggling the enclosed parts back'n'forth.


Quote
Too bad there's no TIB Ashley's bend [#1452] equivalent of the butterfly loop.
For the latter, one can make it via what some
have called "the twirly-flop method" --an old,
traditional twisting of the rope then tucking
the bight through prior twists folded together--
if one, just pre-eye_bight-tucking, turns one
of the halves --only the right one can be done,
so it will be obvious (turn other & you'll have nothing)--
over, creating a Fig.8 half, and then having the
SParts' draw being the same (clockwise/or anti-).

For Ashley's, there is a twin loops fat version,
and then something akin to these knots with one
overhand & one fig.8 half and nice looks.


--dl*
====


jimmyh

  • Jr. Member
  • **
  • Posts: 24
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #35 on: May 17, 2020, 06:53:32 PM »
Quote
What I refer to as the "pretzel" vs. "timber hitch" forms,
as you've described them.

That sounds like a good way to refer to them.

Quote
Although the latter can leave
that space you note, it can be the case --esp. w/stretchy
material-- that the collar around a part gets very tight
and holds tight on the force-diminished diameter,
jamming the relaxation past that point, unstretched
material beyond this point *blocking* against the
tight collar --small space created or not!

Sometimes the pretzel OH form has enough strands
surrounded that one can loosen the structure by
wiggling the enclosed parts back'n'forth.

Yeah, it's not the only factor going into jamming, and the amount of "stuff" in the overhand is a big part too. For example, the simple lehman locked bowline jams when loaded from what is normally seen as the "tail", even though it's a timber hitch overhand. It's a shame, since otherwise it might make for a good TIB omnidirectional loop knot.

Quote
For the latter, one can make it via what some
have called "the twirly-flop method" --an old,
traditional twisting of the rope then tucking
the bight through prior twists folded together--
if one, just pre-eye_bight-tucking, turns one
of the halves --only the right one can be done,
so it will be obvious (turn other & you'll have nothing)--
over, creating a Fig.8 half, and then having the
SParts' draw being the same (clockwise/or anti-).

For Ashley's, there is a twin loops fat version,
and then something akin to these knots with one
overhand & one fig.8 half and nice looks.

I"m struggling to follow here. Are you saying that you can tie a knot that is one half overhand and one half figure eight with each half being the "timber hitch version"?.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3903
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #36 on: May 17, 2020, 10:48:52 PM »

Quote
For the latter, one can make it via what some
have called "the twirly-flop method" --an old,
traditional twisting of the rope then tucking
the bight through prior twists folded together--
if one, just pre-eye_bight-tucking, turns one
of the halves --only the right one can be done,
so it will be obvious (turn other & you'll have nothing)--
over, creating a Fig.8 half, and then having the
SParts' draw being the same (clockwise/or anti-).

For Ashley's, there is a twin loops fat version,
and then something akin to these knots with one
overhand & one fig.8 half and nice looks.

I"m struggling to follow here.
Are you saying that you can tie a knot that is one half overhand
and one half figure eight with each half being the "timber hitch version"?.
Yes!  (just had to get out some rope and see for myself)
Again, in the traditional method, just prior to inserting
the eye bight through the twist-formed opposed bights,
give the one bight that can take it a half-turn which
puts its side into a to-be Fig.8, and tuck.

(E.g., in Animated Knots' tying sequence,
at step-image #11 --which, yes, already has the bight
insertion, but just imagine that yet to come (as this
image shows the knot otherwise in the desired state)--,
one would turn the LEFT side near edge away-over
the away side, making it a Fig.8 (upon bight insertion).)

And, yes, both sides then have "timber-hitch" vs.
"pretzel" geometries for easier loosening.  (And
also the same-rotation draw of nipping turns.)


--dl*
====

Keystoner

  • On Walkabout
  • Full Member
  • ***
  • Posts: 87
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #37 on: May 22, 2020, 01:08:24 AM »
What a ridiculous thread to try and follow. Just a ridiculous conversation here.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3903
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #38 on: May 22, 2020, 10:44:05 PM »
Further to :
I"m struggling to follow here.
Are you saying that you can tie a knot that is one half overhand
and one half figure eight with each half being the "timber hitch version"?.
Agent_Smith provided a handy image for reference,
above.

EDIT NOTE: Some ideas added for the original poster.
The twisted Butterfly is something I have had in the pipeline for over a decade...
ref. Reply#15
Ah, in A_S's image of the "Twisted...", just back out the
eye bight, and un-twist the upper SPart's fig.8 back into
an overhand; this will result in the SParts rotating/twisting
the tails in the same direction (as w/#1452, 1408, zeppelin).
That's what I mean.


--dl*
====
« Last Edit: May 22, 2020, 10:45:04 PM by Dan_Lehman »

bushrag

  • Jr. Member
  • **
  • Posts: 13
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #39 on: June 29, 2020, 07:13:44 PM »
Im the OP and attempting to get back on topic, does anybody have any experience with rethreaded 8 or a doubled double fisherman? If you can access the tail, I think it might be a viable solution. I experimented and it seems to stand quite well. To tie I simply formed a double fisherman where the loop would begin, then form a loop and enter the double fisherman from the standing side, tracing the knot and pulling all the slack through. See photo.


SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 1879
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #40 on: June 29, 2020, 08:36:27 PM »
Hi Bushrag.

The main challenge with your latest offering and so many other possibilities, is the untie-ability after loading.

SS

agent_smith

  • Sr. Member
  • *****
  • Posts: 1177
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #41 on: June 30, 2020, 12:34:04 AM »
Hello bushrag,

I went back to your original (opening) post to try to gain some insight or perspective on the exact application that you intend for such an eye knot.
Quote
Is there a knot to make a loop that can be loaded in all directions with tails that can be loaded in all directions?

To specifically answer your question in relation to your presentation at reply #39, the answer is "Yes" - your presentation will be capable of sustaining a load in all directions.
However, as SS has pointed out, your presentation will be vulnerable to jamming.
However, if the load you intend to apply is not significant, then you will be able to avoid jamming.

Technically, your presentation isn't a corresponding eye knot of #1415 Double Fishermans bend (or "doubled double fishermans").
It's actually derived from #516 (Double overhand knot).


I would point out that your presentation is not TIB (Tiable In the Bight).
Tying your presentation requires access to an end.

...

Some further comments:

[ ] It is not made clear what your intended application is
[ ] You used the term 'omni' in the title of this thread topic. It is not 100% clear as to what your understanding or notional concept of this might be - in terms of precise loading profiles.
For example, we could interpret 'omni' to also mean circumferential loading (ie hoop stress) of the 'eye'. Think of this as a force that loads the eye in all directions - trying to expand the eye like a balloon.
Perhaps you have conceptualized this in a different way?
Maybe you meant a cross-loading of the eye, axially aligned with the SParts (standing parts)?
Or, a cross-loading of the eye, aligned perpendicular to the SParts?

To be honest, I am not sure how to interpret "omni" in your particular case.

On the face of it, it would appear that #1053 Butterfly is the obvious and simple solution - and its 'TIB'.
Maybe you have a specific reason for avoiding the Butterfly as a solution?
I would comment that circumferential loading of the eye of a Butterfly will cause distortion of the core.
Also, cross-loading of the eye (axially aligned with the SParts) also causes distortion of the core.
I don't know of any specific load testing in this way to assess what happens as load is progressively increased (ie does it result in jamming or perhaps capsizing of the knot core?).

Another option is #1059 Constrictor loop.
It will jam...but, your presentation will also jam so maybe jamming isn't a selection criteria?
(you haven't specified if resistance to jamming is a design criteria).

bushrag

  • Jr. Member
  • **
  • Posts: 13
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #42 on: June 30, 2020, 06:59:23 AM »
Elsewhere on the internet someone introduced the use of amsteel blue tied into a diamond knot soft shackle in order to add a belay loop to the climbing harness.

I then decided to look for other ring-loadable eyes/loops or knotted structures that would be semipermanently tied through the harness tie-ins as a DIY belay loop and eliminate the weak point of a soft shackle employed in this way (fast removability is not necessary) . I know that full rated 22kn is practically not achievable with cord but at least it could use the rationale for belay stance or rappel only not fall protection.

Why have loadable tails? I just kinda thought why not eliminate an extra piece of gear and use the tails as lanyards, rappel extensions, cowstails, etc.

agent_smith

  • Sr. Member
  • *****
  • Posts: 1177
Re: Omnidirectional-loaded loop with omnidirectional tails?
« Reply #43 on: June 30, 2020, 09:31:32 PM »
Ahhh... we now have a context and application - thanks!

For a minute there i thought you are a rope access operator (ie industrial rope access) and seeking a better way to integrate a dual leg lanyard into your full body industrial harness.
But - you are an outdoor recreational climber (by the sound of it).
To be honest, if you a lead climber (trad routes) - adding extra 'stuff' to the front of your climbing harness might cause other issues... eg its already a 'full-house' if you have say 2 EN892 'half-ropes' tied into your harness (double rope climbing) plus a full 'rack' of gear.

Indeed, it seems that maybe you are into canyoning or caving?
Adding a 'cows' tail to your harness for canyoning would make sense - particularly at belay ledges during a multi-pitch descent.

However, if you don't actually lead climb - but only 'second' a lead climber up multi-pitch routes - then adding a 'cows-tail' lanyard to your harness wouldn't cause entanglement and over-crowding issues.
Your presentation based on #516 is a little bulky in my view.
Given that you aren't concerned about jamming - I would use #296/#1412 Ring bend (which can also be tied in a corresponding form that is semi 'offset').
The ring bend is more compact and is stable and secure when properly dressed and cinched.

I would also recommend Sterling high strength 8.0mm diameter cord (the same cord as used to configure an 'AZTEK').
Sterling high strength cord is fit for purpose and hard wearing - perfectly suited for human fall protection applications.
Sterling also sell high tenacity cords which are also excellent for your application: link https://sterlingrope.com/store/climb/cordage/high-tenacity-cords (either 6.0mm or 6.8mm).