Redesigning the Gleipnir.

[DerekSmith]

Gleipnir : Pass the cord twice around the bundle bringing the ends back to the middle.  Make a twist in the middle strand and pass the ends through it from opposite sides(or pass slips through it).  Haul the ends to tension the bundle, and as the tension reaches the middle nipping loop, the Gleipnir works its magic and grips the ends without any slippage.  Pull harder to tighten the bundle and the Gleipnir grips the newly tensioned ends without any slippage.

I think that the Gleipnir is a treasure of a knot and I use it regularly, so I am almost reticent to admit that I found myself considering its weaknesses and ways of improving it.  But I did, so here it is for your consideration and critique.

Uber Gleipner or UG : Middle the cord and pass the bight end around the bundle to meet the two ends.  Twist and fold the bight end to form a round turn.  Pass the ends (or slips) through the round turn from opposite sides.  Haul the ends (or the slips) to the desired tension. Done.

It works exactly the same as the Gleipnir and the improvements are arguably trivial - you only go once round the bundle with a doubled cord instead of twice with a single cord, but the real improvement is that the grip is much greater.

So why is the grip improved ?  When you look into the functioning of the Gleipnir, you see that the load applied to the ends has to travel around the bundle in order to arrive at the nipping loop.  If there is a lot of friction (or snagging) then the capstan effect can seriously reduce tension by the time it reaches the nipping loop.  It is that tension that is critical to keep the loop nipped closed, yet it is compromised by having to travel all around the bundle first.

In the proposed UG variant, the tension applied to the ends applies tension directly to the round turn nipping loop, nothing is lost in having to travel around the bundle, so the grip is at its maximum.  (If you want still more grip, make two round turns in the byte end).

Of course, that extra grip comes at a cost.  The knot exhibits the self embracing and locking functionality of the Constrictor, so if overtightened, it can become seriously hard to get behind one of the ends in order to pull it back through the nipping loop(s) to let off the tension and therefore release the grip.

What do you think?

[SS369]

Hi Derek.

Do you feel that this UG is type-of-cord dependent and sensitive to the diameter of the object?

Having just tried it with 6mm accessory rope on a 3.5 inch cardboard tube I don't get the lock.

BTW, itthis combination doesn't work with the OG (original Gleipnir) either.

One observation, though, is that the OG is easier to tighten I think, because of pulling the tails in opposite directions.

SS

[DerekSmith]

Hi Derek.


One observation, though, is that the OG is easier to tighten I think, because of pulling the tails in opposite directions.

SS

Hi Scott,

Aren't you pulling the UG tails in opposite directions?  You should feed them through the nipping loop from opposite sides (exactly the same as the OG), then pulling them one against the other...  I have tied this UG around solid objects and even then with no spring from the object, I still get the UG to lock tight.

Derek

[DerekSmith]

Hi Derek.

Do you feel that this UG is type-of-cord dependent and sensitive to the diameter of the object?

Having just tried it with 6mm accessory rope on a 3.5 inch cardboard tube I don't get the lock.

SS

Cord specific ? Yes, without question.  The knot by design does not shed any load into any external objects, so in the extreme if the cord has poor or no self friction, then it will be impossible for it to hold.  In contrast, hairy hemp would likely be impossible to release.

Object / cord diameter ratio specific ?  From principles, this ratio should be irrelevant, because the knot does not know where the cord goes when it leaves the knot.  The worst case might be considered for the knot to be free hanging (i.e. not bearing on any surface for added compression).  I have tested this in a few cordages and it held well in 'free suspension'.

Derek

[xarax]

   I, for one, can not understand the description of this knot, for one or more of the following reasons :

   1. I do not believe that knots can, or should be, described verbally. ( I mean "only", with words, and with words only, of course ).
   
   2. The Gleipnir itself has different forms. The last time I was trying to persuade, with words, dan Lehman that the Gleipnir could be tied in another way, it took me ages to achieve this ! However, at the end there were his pictures which clarified the matter - and it was proved that, indeed, there are more than one forms of the Gleipnir :
   http://igkt.net/sm/index.php?topic=1449.msg10564#msg10564
   http://igkt.net/sm/index.php?topic=1449.msg10580#msg10580

   3. I do not believe that there is any variation of the Gleipnir which has been not tied already - and shown, by pictures, in this Fotum. I believe that the Gleipnir, as a binder or as a hitch, has been explored - although its different implementations are not tested, of course ( However, which knot has been tested ?  )

[SS369]

Hi Derek.


One observation, though, is that the OG is easier to tighten I think, because of pulling the tails in opposite directions.

SS

Hi Scott,

Aren't you pulling the UG tails in opposite directions?  You should feed them through the nipping loop from opposite sides (exactly the same as the OG), then pulling them one against the other...  I have tied this UG around solid objects and even then with no spring from the object, I still get the UG to lock tight.

Derek

My bad. I did not pass the tails through from opposite sides. It works as you've indicated. Sorry to waste your time Derek. 

SS

[DerekSmith]

No problems Scott,

It is my fault, as Xarax says, I should have posted images.  Unfortunately, at the moment I do not have that facility.  Still, you got there in the end.

What do you think of it?  Do you think it has already been discussed as Xarax suggests?

Derek

[SS369]

Here's a photo of it.
This is what you've proposed, correct Derek?

SS

[xarax]

   If this is the " Uber Gleipnir"  (sic) / " UG Gleipnir ", then I should have to try hard to figure out how on earth is it different from the binders shown at :

   http://igkt.net/sm/index.php?topic=2981.msg17782#msg17782
   http://igkt.net/sm/index.php?topic=4763.msg31498#msg31498

   When two knots are different, as the "common" Carrick bend and the Carrick X bend, we say they are the same. When they are the same, we say that they are different ... I wonder, is this a "failure of the English language", or what ? 

  ( And, please, do not wake up Dan Lehman to show him this binder... , and tell him that it is Uber Alles !    )
   

   

[DerekSmith]

Good memory you have there Xarax.

Indeed, Dan has preceeded my post by some three years.

The proposed UG is definitely the Lehman-Binder - well done Dan.



In deference to the original creator of the Gleipnir though, I still prefer the name UG.

Derek

[xarax]

   It is not a matter of good memory - which, unfortunately, I lack !  It had happened to me to have tied this knot, as a hitch, when I tied the "simple-hitch-a-la Gleipnir" - but I had not presented it, because it is [ as a hitch / qua hitch, on the surface of a round pole ] inferior to "the hitch finally presented". I had explained why this is so many times, so I am not going to do it again here.
   Now, as a mid-air binder, it is good - but I had found that it is better when/if it is tied with one more wrap/turn, so the "nipping tube" is longer (1), and can engulf better the very useful twist of the one end around the other seen in the double "simple-hitch-a-la-Gleipnir". There are posts in the above mentioned threads that explain this.
   Luca s Prusik binder(s)(2) and SS369 s simpler Cow binder(3) are very tight mid-air binders - which also happen to be very tight as hitches.
   I, too, had recently tied yet another mid-air binder, based on a twin crossing knot, which also seems to me very tight. (4)
   The issue of which mid-air binder is more secure and easier to be adjusted, under tension, is still open (5). However, if nobody spears some minutes to tie the binders we already have, or even some seconds to just look at their pictures, he will not be able to learn anything, or to proceed, I am afraid.
    ( I am talking here about the many two-nub binders which do not utilize a mechanical advantage, like the binders shown in (6), for example )
 
   1. http://igkt.net/sm/index.php?topic=2981.msg17791#msg17791
   2. http://igkt.net/sm/index.php?topic=4821.msg31427#msg31427
       http://igkt.net/sm/index.php?topic=4821.msg31447#msg31447
   3. http://igkt.net/sm/index.php?topic=4821.msg31428#msg31428
       http://igkt.net/sm/index.php?topic=4821.msg31752#msg31752
   4. http://igkt.net/sm/index.php?topic=4819.msg31391#msg31391
   5. http://igkt.net/sm/index.php?topic=4818.0
   6. http://igkt.net/sm/index.php?topic=1870.msg21229#msg21229

[xarax]

  Of course, we can tie any "nipping structure" on the middle of the line of such a one-nub / two-lines binder, and then let the free ends go through it, from opposite sides. I was always interested in nipping structures that are TIB, that is, topologically equivalent to the unknot, so anything tied in this remote, from the ends, point of the line, the middle, can be tied/formed and untied/released at an instant, and without leaving any trace - any still knotted structure, like a "relic" overhand knot, for example, which should be tied in a first, and untied in a second stage.
   The situation is not very different from what we have in a bowline-like knot, tied as its 'nipping structure, on which the "collar structure" is attached. As noticed by Mark Gommers (1), the Clove hitch, or the Constrictor, can be clinched hard around the penetrating lines, and almost jam (*). That is why I was reluctant to use them as nipping structures in the case of using such knots as nubs of hitches, tied around round poles. However, as nipping structures of one-nub-tied-in-the-middle-of-the-line binders, the disadvantage may well be considered, and may actually become, an advantage.
    I have also tried such nipping structures in the case of adjustable nooses - which can serve as hitches, too. The Pretzel-like nipping structure used in the very tight and almost perfectly balanced adjustable loop shown at :
   http://igkt.net/sm/index.php?topic=4464.msg28357#msg28357
does not work very differently from a Clove hitch - but it does not jam as easily.
   I have NOT tested ANY of those binders in a tug-of-war test ( and by "test", I mean a systematic, controlled, repeatable, and statistically important series of experiments, NOT a one-off attempt, to "validate" a belief in one more knotting myth ! ). However, I think that the Clove hitch ( which I used in another way in the Bull Clove hitch (2)) may be proved to be the easiest to tie ( although veeery difficult to untie ! ) and the most tight, for its size, simple solution.
   WHO is going to test all those binders ?    The good thing is that all those knots would eventually slip in lighter loadings than the loadings we are normally confronted to in the case of destructive strength tests - AND that the same line can be use many times, provided the nub would not tied on the same part of it... So, I hope that we will learn about some tests, before the end of this century...

   (*) 
   A plausible "explanation" of the jamming of the Clove hitch, is attempted at :
   http://igkt.net/sm/index.php?topic=4347
     
1.  http://igkt.net/sm/index.php?topic=4507.msg28775#msg28775 
2.  http://igkt.net/sm/index.php?topic=4748
   
 

[xarax]

   While I was taking some quick and dirty pictures of the Clove binder I was talking about in my previous post ( so I would not do what I blame : to only talk about, and not to show the knots at the same time...), I remembered something which I should had mentioned, although it is not anything really novel : In this binder, as well as in many others of the same type, I have seen that it may be better to pull the ends when we have not yet crossed the tails = twisted them around each other - so we will be able to pre-tighten or tighten the binder, without been hindered by the brakes !  Those "twists" of the ends inside the nipping "tube" are great brakes, and they are very effective as a "lock" of the free ends, but they may absorb prematurely some of the tension we try to insert into the wraps and the nub...So, it may be better if we keep the ends parallel to each other inside the nipping tube while we are pulling them, and only later make them twist around each other - by applying the usual move we apply to the end of a trucker s hitch, when we lock it in between the opposing bights.
   So, I decided to show the pictures in a new post, and describe this technique - just in case there is somebody out there who will read about it...

[Dan_Lehman]

Gleipnir : ...  Haul the ends to tension the bundle, and
 as the tension reaches the middle nipping loop,
the Gleipnir works its magic and grips the ends without any slippage.

One quibble with this, noting the emphasized part:
there is sometimes a *delay* or difficulty of getting
sufficient tension around some object (overcoming
frictional resistance) to achieve adequate nipping.
(And, should the bound entity --esp. pertinent where
this isn't a solid object-- change in its bound measure,
the binding if slackened can fail (a reason to tie off
the structure, in such cases, or use some other binding).)

Uber Gleipner or UG : Middle the cord and pass the bight end around the bundle to meet the two ends.  Twist and fold the bight end to form a round turn.  Pass the ends (or slips) through the round turn from opposite sides.  Haul the ends (or the slips) to the desired tension. Done.

Quite well described,
and everybody should get it right,
from "only" these words, without further
guidance!  Now, there are different crossings
of the tails, but these are fine points awaiting
elucidation as to what difference they might make.
(Further : frankly, I find your particular image much
harder to comprehend(!); I imagine a larger object,
such as a sleeping-bag/bed-roll (carpet roll would
be something needing stout resistance), and a longer
lead-in to the turNip, and good view of the tails
oriented roughly parallel to binding axis, hauled out.)
 

... the improvements are arguably trivial ...

In the proposed UG variant, the tension applied to the ends applies tension directly to the round turn nipping loop, nothing is lost in having to travel around the bundle, so the grip is at its maximum.  (If you want still more grip, make two round turns in the byte end).
...
What do you think?

Precisely the redress my original misgivings sought.
And I've tried other variants, with the same design
goal : to have the tensioning have immediate effect
on the turNip ("nipping loop"), avoiding the mitigation
of force that might come from friction around what's bound.
(And Xarax presented a version in which the original's
turNip was inverted, taking the delivery up away
from the object then down-around, gaining I think
some slight but appreciable force and locking friction.)

But this revision still suffers a problem with uses
where one might be e.g. disposing of the material
with the bound contents : that of efficiency in the
consumed length (as one would need sufficient
length to enable the set-up and binding, but the
have at least one long tail lost (the other end can
be cut close and the spool of twine keeps it for the
next task)).

--dl*
====