On the bowline-on-the-bight

[xarax]

   The bowline on-the-bight is a great, beautiful knot. TIB and easily tied in-the-bight and in-the-end, simple, secure, either-end-loadable - what more can we demand from a loop ? It should occupy a place in every knot tyer s Pantheon of "old" knots - and it certainly would do in mine, if I allow 13 deities, and not only 12. 
   However, there are three things that I do not like in this knot :
   First, that feeling of weakness of the single-line collar. All the rest of the knot is made of a double line, and this offers strength and resistance against wear, except the tip of collar...
   Second, the two eyes are directly "communicating" with each other - which may not be what we need in an application, and which can also increase wear in the area of the collar, because, if/when the size of the two eyes changes, the rim of the collar is working like a wire saw on the pair of the ends, and vice versa.
   Third, the double line is fine when it runs along straight segments, but poses problems when it makes O- and U-turns. Exactly as it happens in all such "retraced" knots ( meaning knots which can be tied in-the-end by retracing the path of the first line with a second line ) - the fig.8 bend and fig.8 loop, for example. In each turn, each of the two lines can follow the inner or the outer track, and so the knot can be dressed in many ways, and the most regular, streamlined dressing of them requires attention from the knot tyer. That is the reason I had abandoned the "shrunk", single-eye version of this loop : too many dressing forms mean that the knot is not self-dressing, and it can settle in a not-optimal form (1).
   Transforming the simple collar into a "Bull" one, i.e., an one-and-a-half turn "doubled" collar, as shown in the attached pictures, may address all those "faults"/ shortcomings. The knot remains easy to tie, and it becomes more easy to dress and to inspect - furthermore, I have the feeling that it becomes more homogenous, because the line is doubled everywhere. As a drawback, the two eyes may initially be set up unequal ( but they can also remain unequal, after the end of the dressing ), so the one may be forced to bear more load than the other - but I think that this difference will be smoothed out very soon by some slippage of the eyeleg of the smaller/ more loaded eye through its other end = through the nipping turn. 
   ( To see a "vertical" picture, like the first one, in a larger size, download it, and then flip it horizontally ).

1.  http://igkt.net/sm/index.php?topic=4687.0

[knotsaver]

Now we have to try with the clove(two half hitches)-collar, with the ring-collar, with the constrictor-collar (?!)
s.

[xarax]

   And the Girth hitch collar...
   No, I just wanted to duplicate the line at the tip of the bight of the collar, because this point was the only one where the line was single - and I did this in the simplest possible way, as I should. Also, I wanted to make room under the one leg of the collar ( which passes "over" the "lower" rim ), for the continuation of the pair of the on-going eyelegs - having to fit in this space, and so forced to be twisted and oriented accordingly, the path of the pair is uniquely determined and streamlined, so both lines enter into and exit from the turns of the nipping loop in a very smooth way. In many such "retraced" knots, the one line may go "over" and then "under" the other, and thus we have the formation of unnecessary twists, "kinks" - which, besides being ugly, they may also be detrimental regarding the strength of the knot. Smooth, undisturbed/untwisted paths of adjacent/parallel double lines can also be inspected more easily, because any irregularity/mistake in the flow of such lines within the nub will be spotted instantly. 

[Tex]

The thing I don't like about the BoB, and why I don't use it if a fixed loop is critical, is that it easily capsizes into the slip knot form in some ropes.  It's hard to even call it capsizing, it's on the edge constantly and just a matter of how you pull it a little.  This seems to help that a little, but I'm not too impressed.  I DO like the pulled-up-loop version and I don't mind your criticism of it much at all (even if it's not in "perfect" dressing, it's still good)

The saw effect is some point, because it will not redistribute as easily (when it does, the saw effect will not be any better).  Do you really think the breaking strength of this knot is any higher than the standard BoB though? (sincere question) I guess it makes tension from the collar more spread out, but I would also guess a bowline isn't broken as much by the collar as by the nip.



Ok.. better (well...) way.. tie a normal single bowline,  with LOTs of extra rope of course, do a u-turn around the outgoing eye leg and back up into the nip, and then retrace the whole thing from there... including the collar.  Of course the u-turn around the eye leg is still single.  It's got to turn around somewhere, but there is NO tension happening there now. 

This is a BEAST of a bowline.  It actually looks pretty neat, and to my surprise it is TIB (good luck learning how).  Try it.

[Tex]

So here's something unusual for me:

Is the name "Double Bowline" taken?  Because this is as close as it gets.  Cut the dangly loop, and it really is just two bowlines.

[Tex]

Actually the TIB method is quite easy.  You start by folding a bight just like for a BoB, but turn the end of the fold into a big open cow hitch.

Making a nipping loop on both legs together above the cow hitch.  Pull a bight from the return legs of the cowhitch through the nipping loop, then "halter and collar" it.

Another varition

Now if you take the finished knot, and take the tail-end loop and pull it all the way around the main loops and up through the nips.. you get something else quite interesting:  a cow hitch collar. Of course the whole reason I looped around the returning leg was to prevent that from happening, but then, that's kind of interesting too, but kind of gets back closer to where this started, and shows a way to go back and forth between the two places.

[xarax]

...even if it's not in "perfect" dressing, it's still good

  Some dressings lead to a ("narrow") two-rope diameters first curve, and some to a ("wide(r)") almost three-rope diameters one. How smoothly and gradually increasing, and how wide the Standing Part s first curve is, is perhaps / IMHO, a most important factor regarding the strength of any knot, because, at this area, a great amount of the tensile forces running along the rope are "uploaded', absorbed and re-distributed to other segments of the nub. That is why I always prefer knots where the exact size and shape of this curve is determined uniquely, and almost automatically, by the knot itself ( i.e., knots which, regarding this aspect, are self-dressing ). However, exactly the same problem exists with the various dressings/forms of the fig.8 eyeknot, but nobody seems to acknowledge it or care about it - so I guess I will never convince anybody that the required additional attention during the dressing of a double-line knot should always be considered as a disadvantage...

Do you really think the breaking strength of this knot is any higher than the standard BoB though ?

   On the contrary, the standard double-eye bowline-on-the-bight, as a double-eye loop, where tensile forces can be easily equalized and evenly distributed, would be ( slightly ) stronger than either the single eye, shrunk / pulled out version, or the version presented in this thread. It was resistance to wear, and a general feeling that this double-collar version was more homogeneous, that led me to it, not any concern about strength.

to my surprise it is TIB

  Not only this ! It also acquires an always-three-rope diameters first curve ! 
( Although not in the form you show it - in the form shown in the attached picture )
   Good find ! A nice knot !

   P.S.
   Knotsaver found an easy and quick way to tie this bowline in-the-bight. See :
   http://igkt.net/sm/index.php?topic=5490.0

[Tex]

I refer to how how to tie the knot in the picture in a TIB way.  I then refer to how to transform the same knot by moving the tail loop around and back up to the collar (which maybe I didn't describe well, try it and see).  Two knots, but changeable from one to the other.

The knot in the picuture is easier to tie than I first suspected.  We can call it a joint a find.  I just followed the theme.  I had also kind of thought about the single collar before and realized that the BoB which I've seen called a retraced bowline, wasn't a fully retraced bowline, exactly because, as you say, it doesn't have two collars.  This (new, I think) knot is.

[xarax]

   The "tail loop", as you call it, can encircle either one of the two legs of the eye - or both of them. I had just tried to shorten its path within the nub as much as I could, so I made it encircle one leg of the nipping loop. All those variations are interesting - and TIB.

[Dan_Lehman]

to my surprise it is TIB

  Not only this ! It also acquires an always-three-rope diameters first curve ! 
( Although not in the form you show it - in the form shown in the attached picture )
   Good find ! A nice knot !

?!
Tex's twin-eye knot has four diameters in
the central nipping space, not three.  (But I like
the looks --comp.image so far, not yet rope in hand--
of your variation on this theme, which looks to be
slack-secure.)

As for the (traditional) BoB, I do NOT like it :
there is no good way to orient it to be kind to the
S.Part, and if an eye is unloaded and its corresponding
end IS loaded, the eye can collapse.
For rockclimbers' tie-in, there is a neater way to finish
the knot than the brain-dead tracing to achieve a BoB
--which trad. finish has nothing but an extant *thing*/name
to refer to, to commend it.


--dl*
====

[xarax]

  It ... acquires an always-three-rope diameters first curve ! 

  ?!
  Tex's twin-eye knot has four diameters in the central nipping space, not three.

  At this sentence, just as in the two previous paragraphs of this post, I compared it to the single eye/shrunk bowline-on-the-bight, presented at (1).
     
  1. http://igkt.net/sm/index.php?topic=4687

[alpineer]

Xarax, what is/are your reason/s for displaying identical images in both color and b&w? One or the other is fine IMO.   

[alpineer]

My question was - and still is - why do you display both color and b&w images? 

[alpineer]

   I am lazy !

Not lazy enough, as either is sufficient on its own. Regarding my question - which remains unanswered... 

[Tex]

I think xarax's version has 3 ropes in the nip, because he does the u-turn not around an eye leg, but around the nip itself, returning outside the nip.  This could put torque on the nip which might change how things act, so I reserve judgement, but it's certainly another interesting option. 

Actually, I probably won't find myself using any of these much, but if I need a double loop, there it is.

Is 3 ropes better than 4 in the nip?

I don't like doing the u-turn around the return-eye leg because if it's (relatively) too easy for it to slide through the nip.  even then.. it's still secure though.