Double sheet bend with double bight tuck

[RGB]

For the amusement of members, I present tying instructions for a double sheet bend with both tails tucked (a double, double-bight-tucked or DDBT sheet bend). My tinkering with sheet bends has been influenced by Mark Gommers' thinking about increased strength with more lines through the nipping loop or collar (in bowlines) and roo's point that (all?) sheet bends are susceptible to failure through bight-straightening.

Anyone who can tie a double sheet bend can tie a DDBT. There are just two extra tucks for much greater security. Apologies if all this has been documented before (I could not find it).

The starting bends are tied like sheet bends (direct tails after 1, oblique tails after 2 ). Asher prescribes 'simple simons' with tails like reef knots (direct tails after 2), which gives a more symmetrical appearance. Nowhere has the effect (on strength or stability) of tail orientation or loop chirality in 'simple simons' been documented. Tail orientation is moot after 3. However, tying steps 1 and 2 like Asher would change the shape of the final DDBT. To my eye it would put an undesirable sharper curve where the nipping line exerts strain in the knot. More importantly, it would create a flaw in structure mentioned below. A double simon over (even if tied to give oblique tails in 2) would give a less secure nipping line tail.

In step 3, there are 3 ways to tuck the tail of the bight back through the bight:

(a) Folded. Simply lift the tail and tuck it back down through the bight, as described by Asher in 'Knotting Matters 45'. This is rejected as it can still be bight-straightened by cyclic pulling on the ends of the bight (albeit hard work if the knot is started with a double sheet bend).

(b) Fig 8. Take the tail of the bight over and around its own stand, then down through the bight, as described by Warner in 'A Fresh Approach'. This matches the direction of slight movement of the bight tail under tension (normally a good thing) and is arguably the most handsome. Nevertheless, this knot is rejected because it is easily manipulated so that the nipping loop of the sheet bend (or the first loop of doubled sheet bend ABOK 1434) ends up around only the standing end of the tucked bight instead of both lines of the bight (Warner did not note this). This problem can be avoided by starting with ABOK 1435 (which Warner calls the Rivermans bend). Also the Fig 8 version loosens more than the overhand version with repeated shaking when made in a stiff, thick line. These flaws are most obvious when tying lines that differ substantially in thickness.

(c) Overhand. Take the tail of the bight behind its own stand, then bring it up and push it back down through the bight, in parallel with the other lines through the bight. This is preferred for security and avoidance of the flaws in (a) and (b). As pictured it is actually made counter-clockwise underhand, so perhaps it is better to use the generic descriptor 'thumb knot'. Caution: If tied incorrectly with oblique tails on the sheet bend (stage 1), it is the overhand version that is at risk of the flaw from manipulation of an inappropriate nipping loop, described for the Fig 8 version above. Be sure to tie the sheet bend (stage 1) with direct tails, and always use the double sheet bend with ABOK 1435 configuration as insurance.

There are also methods that weave tails through the knot (e.g. https://igkt.net/sm/index.php?topic=6100.0), but these get a bit more complicated, especially if both tails are to be tucked.

[RGB]

Whether there is still a 'nipping loop' after all this is open to debate. But it does not seem to matter for knot security. The lead is good when hauled by the thicker (bight) line, but poor (tails can snag) when hauled by the thinner (loop) line, unless the tails are seized to the stand of the thinner (loop) line. I think it is practically impossible to bight-straighten the DDBT (even without seizing), because it forms a thumb knot in the bight rope. I have given it 100 shakes from each stand, then another 100 shakes from the thicker stand, which did not cause it to fall apart even with the very different diameter and fairly smooth /stiff synthetic laid ropes pictured. The thumb knot in the stiff thicker line did loosen very slightly, but the whole knot tightened again as soon as strain was put on the two standing ends. Even with tail loading (for which it is not designed) I do not think it can capsize / roll like a bi-axially loaded offset Fig 8 bend or EDK.

Collar flare when pulling on the nipping stand and the tails is much more severe if the knot is commenced with a single (rather than double) sheet bend, and after this treatment the single DBT does not return to the normal collar form automatically when the stands are tensioned. Always make the small extra effort to tie the DDBT.

The 'dressing' or movement of the tails during tightening is slightly different compared to the previous slide, where the lines were more similar, but it is the same knot.

The DDBT might be considered a combination of an oblique double simon under (modified from Asher's double simon) with a thumbknot-bight-tucked sheetbend (modified from Warner's Fig 8 'loop-tucked' sheetbend).

[agent_smith]

Thanks for this RGB.

And the link you gave had some very nice photos presented by NautiKnots. The structures presented by NautiKnots (in my view) are worthy of follow up investigations.

per RGB:

My tinkering with sheet bends has been influenced by Mark Gommers' thinking about increased strength with more lines through the nipping loop or collar (in bowlines)

Actually, I think it was Dan Lehman who pointed out the potential benefits of 'padding out' the nipping loop. So knots with 3 or more rope diameters inside the nipping loop was thought to have certain advantages. And Xarax may also have been contemplating such matters...

I think a round of testing is needed to confirm (or refute) this proposition. Although logically, it appears that we should see some benefits from padding out the nipping loop.

Whether there is still a 'nipping loop' after all this is open to debate.

I have always held the view that a Sheet bend is devoid of a functional nipping loop.
I had advanced that a defining quality of a nipping loop is that it is loaded at both ends.
So in my view, there is no nipping loop in #1431 Sheet bend.

I would advance Scotts locked Sheet bend as worthy (and deserving) of further testing.

[RGB]

Thanks Mark,

The woven bight-tail-tucks pictured by NautiKnots sometimes secure against bight-straightening, and his idea of using the thinner line for the bight is interesting. Pull on the two ends of the bight line to check: in my hands Scotts locked Sheet bend straightens (for example).

In addition to ensuring that the knot is not subject to bight-straightening, perhaps it is wise to tuck both tails of sheet bends for security (see for example http://caves.org/section/vertical/nh/50/knotrope-hold.html). Then weaving gets a bit more complicated. Hence the DDBT.

I understand how the distinction of a dual-loaded nipping loop helped in your definition of bowlines (www.paci.com.au/Downloads/Bowlines_Analysis.pdf : a great analysis), but I think it is a losing crusade to try to convince people that a common sheet bend does not include a loop that nips (see for example KC's posts in https://igkt.net/sm/index.php?topic=5676.0).

There has been at least some careful analysis in support of the notion (advanced by Dan_Lehman, Xarax, you, and probably others) of increased strength from loop and collar 'padding', at least in your EBSB (http://www.bigel-labs.de/3.Physik/Knots/web).

I appreciate that you have already considered all of this, but the links may help others.

PS: If anyone else is initially visually confused by the knots illustrated in https://igkt.net/sm/index.php?topic=6100.0, note that NautiKnots ties his nipping loops (sorry Mark) with Z chirality, whereas in Z-laid rope I prefer to tie with them with S chirality. Once this is noted, translation is easier. Any effect of loop chirality on strength, stability or rope life, most likely to be relevant in knots in laid rope, is not well documented (Warner dismisses it as trivial, but Svennson seems to advocate otherwise). Once the step is made to a double sheet bend there are loops of opposite chirality (as in many other knots).

[agent_smith]

per RGB:

I think it is a losing crusade to try to convince people that a common sheet bend does not include a loop that nips

Not at all.

Very careful use of language is essential.

Terms:
[ ] Nipping loop = a closed helix that is loaded at both ends. It may have left-hand (S twist) or right-hand (Z twist) chirality.

In #1431 Sheet bend there is no structure that matches this definition.

You have constructed the phrase 'a loop that nips' - which is different to 'nipping loop'.

In a Sheet bend (which is an end-to-end joining knot) an SPart traps and crushes 'down' upon its own tail (ie 'down' referring to force acting from one side).  This action of trapping and crushing its own tail is a completely different operation compared to a 'nipping loop' within an 'eye knot'.
Your use of the phrase 'a loop that nips' is conditional and is one way of trying to describe the action of an SPart crushing 'down' upon and trapping its own tail.

In a #1010 Bowline (which is an eye knot), the SPart does not trap and crush a tail. The SPart morphs into a 'nipping loop' - which then fully encircles and crushes both legs of the collar (which collectively is a 'bight structure'). And in this case, the nipping loop is loaded at both ends.

NautiKnots ties his nipping loops (sorry Mark) with Z chirality, whereas in Z-laid rope I prefer to tie with them with S chirality

This 'preference' toward Z or S twist is often times linked to whether a person is left or right handed - and; how you originally learned to tie a Bowline or a Sheet bend. Right handed knot tyers will usually form a Z twist nipping loop (I say usually, but there are exceptions...and again, influenced from whom you learned how to tie it). Whether a person ties a common #1010 Bowline with a Z twist or an S twist makes no difference to performance of the eye knot.

As to differences between #1010 Bowline and #1034 1/2 Bowline - we know that the latter is resistant to 'ring loading'.
In a head-to-head MBS yield (ie pull-it-to-failure) test, I think the matter has not been investigated to any degree of scientific rigor by an accredited, certified test lab. I am not even sure of any 'backyard testers' who have tried to construct a repeatable test that has been properly peer reviewed.

In the case of comparing #1431 Sheet bend (tails same side) to #1432 (tails opposite) - we must keep in perspective that these are 'end-to-end' joining knots..they are not 'eye knots'.
It may be a useful analogy to look at how wire rope grips are oriented on steel cable to form an 'eye'. There is a right way and a wrong way of orienting a cable grip.
In the same analogy, it is always advanced that #1431 is more secure and stable compared to its cousin #1432.

[agent_smith]

Here are some images of response to loading in #1431 Sheet bend in comparison to #1010 Bowline.

Please look very carefully at how each structure responds to load.

Note the differences.

EDIT NOTE:
Both ropes are EN892 certified dynamic 'single' climbing ropes.
[ ] Blue rope is Beal 'Joker' 9.1mm
[ ] White rope is Edelrid 'Corbie' 8.6mm
Difference in diameter is 0.5mm (ie half a millimeter).

The last photo is showing both knots using same rope (Beal EN892 'Joker' 9.1mm dynamic rope).
Of immediate note is the differences in the collar.
In the #1010 Bowline, the collar is not heavily loaded and can be manipulated/moved).
In the #1431 Sheet bend, the collar is loaded and is quite tight - it cant be manipulated/moved.
The distribution of forces in the core (nub) of the #1010 Bowline is quite different compared to the core of #1431 Sheet bend. There is a functional nipping loop acting in the #1010 Bowline. There is no functional nipping loop in the #1431 Sheet bend,

[Dan_Lehman]

Here are some images of response to loading in #1431 Sheet bend in comparison to #1010 Bowline.

Please look very carefully at how each structure responds to load.

Note the differences.

But this is for some particular cordage; there are
different geometries present in some others, which
makes generalization problematic --and basing
definitions on behavior is something for that
reason I'm chary of.

I was impressed by a photo in CLDay's first book,
Sailors' Knots, in which a broken sheet bend (tested)
was pictured : the break came in the bight's SPart
where the Is-It-Or-Not-Loop's constriction of it
was so great --i.e., in a straight part of the knot
not the bent one (!).  IIRC, in that manilla cordage,
the sort of stretching out seen in the kernmantle
nylon hadn't occurred.

So, in THIS case, "the potential benefits of 'padding out'
the nipping loop" would come in protecting one of these
padding strands by giving it more united-we-stand
protection, I guess; and following the advice that
for ropes of different sizes the bight component
should be formed with the larger rope suggests that
an increase in strength arises by strengthening the
part that is most vulnerable --in that manila rope,
at least--, and possibly adding some strength to
the broader-turning loop part, as well.   !?


--dl*
====