Winter bend

[xarax]

This can be seen as Ashley's #1452 with the tails drawn out of the collars;

  Or ( in the supposedly "weaker" reversed form ) Ashley s #1451 s symmetric cousin with the tails drawn out of the collars.
 
  And I say "cousin"(*), because its mechanism is exactly the same with the ABoK#1451 s, only its tails are leaving the knot towards opposite directions ( as they should  : the Standing ends are converging / approaching / coming / entering from opposite directions, and an end-to-end knot is but a knot, not a permanent union/merge of lines ! - so the "natural", and prettier, IMHO, thing, is Tails diverging / distancing / leaving / exiting towards opposite, too, directions. In this sense, the re-tucked through the collars ABOK#1452 is also a streamlined, pretty ABoK#1452... ).
  What I wish to hint is that ( in its supposedly "stronger" normal form ) this knot is Ashley s #1452 bowled-over-by-retucking, not bowled-over-and-re-tucked, as the knot presented in (1).
   Many knot tyers say "retucked", and mean "retucked through the central opening", and they believe that this is the more secure and stronger retucking of all the possible ones. (  That is wrong, and, among other things, it may even be a cause to miss important knots....). Stronger it may well be, because it increases the volume of the central core, around which the Standing parts will make their first turns. However, regarding security against slippage, the convenient retucking-through-the-central-opening can be weaker !
   The tails can be driven through any of the openings the 'parent" knot presents, "collars" being two of them. In fact, the central opening may offer some "protection" to the choking of the Tails, because of the amount of material swirling around, which can absorb, by its inner friction, a great portion of the forces that are meant to play the role of the last line of defence against slippage. ( I am speaking about the behaviour of ordinary, and rather stiff, material here, of course, not about the easy flattened slippery Dyneema...). The same reasoning had lead me to question the security of some forms of "Link" bowlines, where we have exactly the same situation, with two O-shaped interweaved nipping loops :

maybe prefer taking the tail out through the *center* of the turNip, which shows as an inviting hole-target;

  Nooo ! This is an inviting trap ! Beware of its seductive siren song ! I have sensed, seen, felt and measured, that the tail can not be nipped efficiently there - because the other segments that would encircle it would also play the role of a weaved cocoon, of a protective knitted shield. They will absorb a great portion of, and they will obstruct the constricting power of the nipping loop [ read : the core of the bowl formed by the embracing Standing parts ], to reach and bite it hard. ( I am always talking about kermantle, stiff climbing ropes, of course.) I always deliberately drive the tail away from this black hole - even if this would have been the unstrained, "natural" path a less curved working end would have followed, or if it, by the very bulk of the tail placed there, would had allowed a less curved standing part.
  So, to secure the tail in the more efficient way, we must find a place ... where it will be bitten hard by the standing part s... first curve(s) - AND where it will not pass through those two dangerous spots : This central black hole ( when there will be other segments that will encircle and "protect" it ), and the "soft spot" near the nipping loop s crossing point ( where it will be nipped by the two converging legs of the nipping loop only, and not by its rim ).

1. http://igkt.net/sm/index.php?topic=4777.0

(*) Curiously, it is also the reverse of "Carrick s cousin", named by H. Asher ( The Alternative Knot Book, p.63-64 ) - so, is a cousin of both .  Miles calls it "Spherical bend" ( M. A14), and says that it "forms the most ball-like" of all the symmetric bends he presents in his book - but I really wonder why ! Even in its capsized form, in which he shows it in p.104, it does nt look soOo spherical to me...

[xarax]

I for one believe that the interested tyer would try the knot out both ways and glean from it what they can.

   The interested tyer would see that, when he will load the knot the one way ( the way suggested in Reply#7 ), the two collars at the "top" side would move towards the middle of the knot and would approach each other, so they would be transformed into two obliquely placed - in relation to the axis of the knot - rings encircling the pairs of adjacent ends ( resembling the rims of the overhand knots in a properly tied fisherman s knot ), and they would finally "kiss" and become parallel and adjacent to each other. When he will load the knot the other way, it will remain more flat, in a state resembling, more or less, the dressed but unloaded knot shown in the pictures. As I had also mentioned at the same reply, I have no idea whatsoever about which way of loading the bend would result in a less slippery and less weak knot...   

[enhaut]

I would appreciate a step by step how to do; "photo montage" of this knot.

[SS369]

Having tied this again just now in the form of a sling, with the tails in the lower position (as presented in the original photos) I find that when using some 6mm accessory rope that is firm, round, and rough, that after I have snugged it reasonably tight the the tails continue to feed into the nub under tension. (Arm and foot applied.)

My assumption is that this doesn't bode well for the slippery materials.

When pulling on the tail ends (same sling), the the movement does not seem to be as great though I suspect the load profile in the nub may be too sharp and cause a local high pressure area that has been noted elsewhere to be the bane of that material.
It has been shown that Dyneema performs less than favorably when a sling has a knot in it and then subjected to a dropped load, as shown here > http://dmmclimbing.com/knowledge/knotting-dyneema-vid/.
And then there is creep to consider for constant loads, which could exasperate matters, especially the untying.

Not all usage will be slow pull, in fact I believe that most usage will not be of this kind. There has been some writing about the failure of knots in this material (and other materials as well) during repeated load cycling, and so this needs to be considered in any kind of test scenario that can possibly mimic reality.

SS

[xarax]

   There is a reason I deliberately avoid the "step by step" presentation of knots, which, most of the time, means offering just a tying recip : to my view, knots should first be understood, in their workings, and then they should be memorized and remembered as 3D geometric objects. The tying sequence is only secondary, and the images it is accompanied with are often misleading. People tend to identify a knot with one particular of the many ways of tying and dressing it, and this opens the door to the well-known tradition of knots parroting and knot tyers brain washing...
   A possible way is shown in the attached pictures. Start with two hooked bights, where the Working End of each goes over-under-over any segment it encounters. Then, tuck them through the central opening, so they will be crossed, continuing the alternating sequence of their over/under paths at each crossing : by now the working ends have gone over-under-over-under-over. Last, re-tuck the Working Ends, which are now the Tail ends, through the collars, so they become parallel with the Standing Ends.   
   Now, at the the end of the second step, what you have tied is a well-known bend, composed by two interlocked overhand knots : so, the last "retucking", through the collars, is, in fact an un-tucking : the overhand knots of each link become topologically un-knotted ( i.e., topologically equivalent to the un-knot ). We have two interlocked "Pretzel" shaped links, which are "simpler", topologically, to two interlocked overhand knots.
 

[xarax]

the tails continue to feed into the nub under tension.
When pulling on the tail ends ... the movement does not seem to be as great though

  What you say is that my reasonable suggestion for the Standing Ends / Tail Ends pair leads to a more slippery bend. So much for another knot theory ! 
  I have not seen any slippage with my "ordinary" ropes, but perhaps I have not loaded it as much as you. The consumption of some portion of the Tail Ends before a final "lock" is not a proof, nor even an indication, of a weak bend : more convoluted bends, as this, will probably need more time to absorb any remaining slack. However, you seem to say that this consumption goes forever ? ? What was the material of the 6mm accessory cord you have used ?

[SS369]

my reasonable suggestion for the Standing Ends / Tail Ends pair leads to a more slippery bend

Not necessarily. Dependence may be on the material used and the type of construction of the linear media.

My opinion is that for the slippery materials, the knots need to be big knots with many round turns in them, hopefully using a capstan effect to shed the load, or a braided structure to mate a large surface area. Even though 90 degree bends could/should bite into its neighboring parts, the material just doesn't like hard turns.

Some of these knots we've tinkered with, suggesting them to previous posters, will work for occasional knots, but I think that when we get down or up to it, the heavy tensile loads, the area with the sharp bend will be the undoing.

As for one of my "test" ropes, the one I used just a while ago that consumed the tails is a nylon/nylon construction with these specs: Elongation:
@ 300 lbf. = 9.4%
@ 600 lbf. = 12.8%
@ 1000 lbf. = 16.8%

Diameter: 6mm
Grams Per Meter: 25
Tensile Strength: 1,888 lbf. (8.4 kN)
Sheath Mass: 48%

As for continued tail consumption, that I can not test without using my truck, so I made an assumption (not cast in stone). It could eventually lock with this or other material and construction, but I feel the Dyneema will continue on its way.
Then there is the cyclic loading to consider , as I have mentioned prior.

Tying it: An exploded/loose view (?instead of?)should help where one doesn't necessarily have to get cross-eyed viewing from front to back. 

SS

[enhaut]

Thanks for this excellent quality photo sequence!
Now I can tie it without hesitation.
 There is a reason I deliberately avoid the "step by step" presentation of knots, which, most of the time, means offering just a tying receipt : to my view, knots should first be understood, in their workings, and then they should be memorized and remembered as 3D geometric objects. The tying sequence is only secondary, and the images it is accompanied with are often misleading.
So does that means you can figure out a knot just by looking?
Try to solve, if you wish, this 3D geometric shape and tell me how easy it was

[xarax]

  Two more pictures of the suggested by me ( perhaps without any good reason) loading pattern of this knot, after a moderate loading. The two ex-collars, which have been transformed into two oblique parallel "rings", are not yet entirely adjacent to each other - I suppose that, if the knot would had been loaded more, the two "rings" would be adjacent along their whole rim.

[enhaut]

Dear xarax,

  It becomes easier if you already know the knot - which, by the way, you have not dressed in the most symmetric way ! 

Believe me it's cannot be more symmetric than that!
In reality the bend that I showed is the real contender for the name Winter Bend
Care to know why?
 

[xarax]

   Trying to "feed", somehow, the core of the retraced fig. 8 knot ( the knot of the fig. 8 bend or the fig.8 loop ) with some bulk of the material, in an effort to make it jam less, I had re-tucked the tails through the collars. We can do this in a number of ways. There are two collars in each fig. 8 knot, and each of them has two openings, so, entering into them from the one side, the re-tucked ex-tail can form 4, in total, different re-tucked fig.8 knots - but only one is fully symmetric : the one shown at the attached picture. Then, we can re-trace those stoppers following the one or the other direction, i.e., starting from its Standing End or from its Tail End. This is how I had tied the retraced knot I believe it is shown as "Bend X". However, I was not satisfied with any of the results, and I had abandoned them, in favour of the knot based on a retraced ABoK#582 stopper - which gives a presumably non-jamming and most secure end-to-end bend, as well as an easy to tie TIB eyeknot. Dan Lehman thought that this quite convoluted knot is an overkill, and had tied "abbreviated" ( that is, once un-tucked) retraced end-to-end knots and TIB eyeknots based on the same stopper ( See (1)).
   "As we have seen in the case of the retraced fig.8 loop, all those retraced stoppers can be dressed in many ways - because we can twist the pair of adjacent and parallel lines running within their nub one or more times, at one or more areas of the knot." I remember that I had dressed the re-tucked fig.8 knots in ways which were more "smooth" and symmetric, relatively to the individual links. That is why I said that what I think is the "Bend X", as shown, is "not dressed in the most symmetric way".
 
  I would be glad to learn the true history of the "Bend X", even if I had tied it in a wrong way all the way !  There was only one winter on the Northern hemisphere in 2013-2014, so there can be only one Winter bend, I am afraid - so let us know which contender deserves this name most ! 

  1.  http://igkt.net/sm/index.php?topic=4715.0 

[xarax]

As for continued tail consumption, that I can not test without using my truck, so I made an assumption (not cast in stone). It could eventually lock with this or other material and construction, but I feel the Dyneema will continue on its way.

  At least, you have admitted that, despite all climbing exercises, your trained "arm and foot" are not yet at strong as your truck ! 
  I bet it will lock on any ordinary material - but I will not risk this bet for Dyneema, of course ! This sleek monster has slipped out of a triple fisherman s knot - so it is expected that this less convoluted bend will not be able to stand in its way, without some further enhancement.

[SS369]

At least, you have admitted that, despite all climbing exercises, your trained "arm and foot" are not yet at strong as your truck ! 

Of course.

  I bet it will lock on any ordinary material - but I will not risk this bet for Dyneema, of course ! This sleek monster has slipped out of a triple fisherman s knot - so it is expected that this less convoluted bend will not be able to stand in its way, without some further enhancement.

I feel it will probably lock too using it the way it was loaded as I mentioned previously (not the reverse, loading the tails). Using it reversed I believe it would capsize.
And there is the chance that it would damage the sheath as it dug in and bit tighter. But, that is a concern with many a knot under severe loads.

[xarax]

\Using it reversed I believe it would capsize.

  Very interesting ! I can not even imagine how... I believe it is time to USE this truck for this noble, for once, purpose ! 

[enhaut]

"so let us know which contender deserves this name most !"

Well you deserve to know since you have tried to solve it.

It's a pun.
Bend X is a re-threaded knot base on the Eskimo bowline.

One rope form the first bowline without closing the loop, than the second rope track back the first one.

See now? Eskimo and winter a mix made in the north!

By the way this thing is very solid, may be stronger than the re-threaded figure 8 knot, but again I have no way to verify my assumption!