Who's the Daddy

General Discussion
1

Only preliminary testing (which is another way of saying I am playing while the Council decides whether or not to formally support the testing project), but the Lyndy adjustable loop knot has just been knocked out of the top spot of the strongest loop knot.

I have been noticing that failures in the 2mm cored polyester braid I have been testing with was generally occurring at the first hard load shedding turn within the knot, so a knot that had a well padded turn such as the Lyndy would do well, but better still, a knot that did not have any load shedding turns should be better still.

The knot that has strolled into the top position has long been a favorite of mine as a nice easy tie adjustable loop knot.

What is it?

Drum roll if you please …

Well, it’s the Adjustable Grip Hitch http://www.layhands.com/knots/Knots_Hitches.htm#AdjustableGripHitch but tied with a generous loop instead of up close and personal to the fixing spar.

Just one thing to note, after the test, it wouldn’t slip anymore!!

2

In my book, this is strictly a noose, not a loopknot per se (for precisely the aspect that
led you to try it–no turn in the SPart); a knot ties to the rope to make a noose–purely
a structural vs. behavioral distinction (though certainly after loading, it was at least
a “fixed noose”!).

In such a knot, I’d suspect the break to occur in the structure’s/noose’s SPart, not that of
the knot, which makes a hard turn but has theoretically 50% of the load. And so I’d think
that a similar structure using Blake’s H. (mind you: I’m not counting wraps closely–add
as needed) might prove stronger, as the more loaded point of joining occurs beyond the
point where the away line enters the structure, i.e., on the eye side of it, and so there might
be some off-loading and graduating of compression to the noose’s SPart thus!?

In cases of such “noose” structures, one might have concern that the rate of loading
is (more) significant–that rapid (“shock”) loading might cause slippage that at least
leads to increased friction & weakening by heat than show loading will show.
(Life-on-a-Line asserts seeing differences in Overhand-Fig.8-Fig.9 LKs this way.)

Thanks,
–dl*

3

I think Dan, there is reason to take exception with your chosen classification of ‘noose’.

While the Adjustable Grip Hitch has the semblance of a noose in the four round turns and indeed it can slide when the knot is held, it totally fails to demonstrate the key attribute of a noose – the ability to tighten the loop (noose) easily.

Indeed there are a number of round turn constructions which slide when the knot is forced, yet grip ferociously when loaded in the designed manner, the Prusik is a typical example of four round turns designed NOT to slip like a noose, and while you could configure the Prusik into a noose (i.e. loop configuration) there is no way we should classify it as a noose and likewise, I do not feel that we should brand the Adjustable Grip Hitch in this manner.

As for breaking. I set it against itself and this is the first example I have seen in this cord of what I think can be called pure compressive induced failure (i.e. no radius element). At ca 80% loading, the four turns constrict devilishly and a knot which starts out 6mm in diameter shrinks to a little over 4mm across. The knot failed right at the nip of the very first round turn with some of the cord material pulling out from inside of the knot.

I have arranged the broken pieces in line with their position at 80% extension for the attached photo

http://knotbox1.pbwiki.com/f/griphitch.jpg

Derek

4

If you’re going to take exception with my classification, you need to object to my
definition, not take the common one and defend it! The reason for my classification
is that I don’t want to have knots jumping class depending upon factors of behavior
which follow from material, load, load rate, and setting. As I said, I define “noose”
based upon structural aspects, not behavioral ones–which vary (witness
the various discussions of friction hitches and how many wraps are needed). And it
also recognizes that, in a sense, the noose’s SPart plays no role–i.e., does no
twisting/entangling–in the knot, which is all formed by the end around it; so I
decided to view this as a compound structure. (And, yes, e.g., such a view makes
the venerable Two Half-Hitches a Clove H. noose, so to speak.)

Now, you might want me to go run along and get another word; but “noose” does
get me the knot-on-line aspect I want, and I’ve not been clever enough to come up
with something better. And one could want “noose” precisely for behavior (“the
Prusik acts qua noose in 12-strand HMPE rope”). But you brought to the
discussion of loopknots this structure that I think worth noting has a (as you
noted) significant characteristic; and “noose” was my way to indicate that, class-wide.

Indeed there are a number of round turn constructions which slide when the knot is forced, yet grip ferociously when loaded in the designed manner, the Prusik is a typical example of four round turns designed NOT to slip like a noose,
Your view is shortsighted here, both of some current opinions and of course of the Tarbuck hitch's reason d'etre: that of sliding [i]until[/i] some lower force obtains, and only then locking. (--not saying this theory worked all so well in shock absorption, but that was it.) Again, my decision on defining "noose" was motivated by wanting surety, freedom of variation by material.
The knot failed right at the nip of the very first round turn with some of the cord material pulling out from inside of the knot.

Exactly. Now consider the Prohgrip (Blake’s Hitch): in it, the knot (my distinction vis-a-vis “noose”)
delivers its initial grip to the EYE side of itself against the noose-SPart, and coils away from the eye;
it thus also makes less of a severe initial grip, but surely is lessening that compression at the away
side, where the Rolling/Adjustable/Hedden hitches (e.g.) concentrate it. Might this then prove a stronger
attachment? In B-v-B you’ll get a sure break in the given hitch B(lake’s); then A-v-B you can check the
two and see if Blake’s betters Adjustable.

Your break w/A btw is like that of the Strangle Noose (aka “Scaffold Knot”), where presumably that
knot hasn’t so well gripped to hold its position, but is snug to the secured ring (it’s a ring H.) and
can go no farther. --there, too, with similar thinking I’ve given above speculating better strength
in the ProhGrip, I think that the Dbl.Oh. would be stronger if oriented into a Fisherman’s Knot
vs. Strangle knot around the noose-SPart, as it seems that the concentration of force is less.

–dl*

5

In my testing both of these grip hitches acted as nooses, sliding down to the ring.

Blake’s is stronger. In multiple tests the Adjustable Grip Hitch broke as illustrated, where the standing end is first gripped. In a couple of cases it was the gripping line that broke, in others, it was the standing end. In the latter case I was left with a very tight cone of coils sticking out perpendicular to the ring. As predicted Blake’s is better able to spread the grip over the length of standing end.

In fact, with 10 turns, Blake’s hitch is stronger than the hitches that in other tests proved to be very strong (Ossel Knot and Lighterman’s). With this many turns, Blake’s becomes something close to a half blood knot or bimini twist.

But in small line, Blake’s is harder to tie than the Adjustable Grip Hitch.

paulj

6

PaulJ, did you make much effort to set the friction hitches? --sometimes it seems
to be a balancing act of finding the threshold of pre-tensioning at which then the
application of force to the structure will see the knot hold and in holding build
tension-grip in its wraps.

With the ProhGrip/Blake’s, one has the variability of version in tucking the end
beneath 1-n number of wraps–where tucking the end under more wraps is seen
as a way to redress stiff rope which will resist tightening snug around a single
part (but which thus reduces the surface of the hitched-to part that is contacted).
(I suspect that in both of the materials in use by you & Derek flexibility is great.)

–dl*

7

I don’t mean to ruin your fun, but getting hard numbers will give you better, more meaningful results. This has been discussed here:

http://tinyurl.com/yp2kc5

8

Interesting results PaulJ,

I seems that monofilament has a totally different slippage characteristic to 2mm Polyester braid. I tied the knots with Spectra kite line and yes they ‘noosed’ up on heavy loading. I guess that I am going to have to revise my stance on these knots being ‘Nooses’. Although I do not agree with Dan’s redefinition of Noose as a structure, these knots are definitely capable of exhibiting Noose characteristics in some circumstances. However, when I use the AGH as an adjustable bend, then it fails to be a noose, but must be recognised that under some circumstances, it can slip before it fails.

I put the Adjustable Grip Hitch against the Blake and the loading was going way up when the Blakes began to slip in a series of jumps, each accompanied with a bang. After the Blakes had slipped twice, the Adjustable Grip Hitch slipped with a big bang and a puff of ‘smoke’ which smelled just like cordite. Finally after two more slips in the Blake, the AGH failed in an identical manner to its previous failure against itself.

On examination of the cord where the AGH had slipped, there was a region where the surface was ‘pucked’ with hard melted thread ends sticking out from the braid. It looks as if at the moment of transition from grip to slip, significant energy was dumped into the knot, causing localised melting. This was not caused by the friction of movement, because further down the cord, just before the point where the knot finally stopped sliding, there was no melting and puckering - only in the zone where the knot was originally gripping and where it started to slide. Can anyone explain the physics behind this observation of apparent near instantaneous energy release?

9

Yes Dan, that is indeed what I was doing. Although I fully understand the need for a term which describes the structure of a simple straight cord passing unobstructed through the body of the knot, I do not believe we should hijack a term with an already well established meaning.

Our field is already littered with so much esoteric terminology as to threaten us with the appellation of ‘Cult’ and if we genuinely want to widen interest, then we need to be considering how to clarify our terminology, not worsen it. After all, how many of us have seen that glaze that comes over someone’s eyes when you try to tell them that they ‘bend’ their shoelaces together?

Now, you might want me to go run along and get another word; but "noose" does get me the knot-on-line aspect I want, and I've not been clever enough to come up with something better.

Again, yes indeed , I do think a more appropriate word might be used which describes the structure without confusion over function.

Could we use – Line knot or straight knot or linear knot or inline knot or collar knot or compression knot?

Derek

10

No arguments here Rue, only one problem (or more correctly, several tens of thousands of them) – cash. We as yet have not found a rich benefactor to fund the equipment necessary to perform the measurement of stress and strain of breakage. In its absence we can either elect to wait and do nothing, or use a very simple comparative system and use our wits to tease as much information out of the results as possible.

Using intelligent guesses, we can implement controls in the testing protocol which we believe should reduce the width of the bell curves.

Using replicate tests, we can deduce the degree of overlap of the curves – i.e. if the tests come out 50:50, then we can be reasonably sure the two bells overlap equally. If the results come out 2:1 or some other value, then again we can calculate the probable degree of separation of the bells and where we have a number of bells overlapping, we can use each to improve the probability that we have estimated the separations correctly. Hubble has an imperfect mirror, but thanks to ingenuity it is sending back fabulous images. In our case, the ‘noise’ of the bell curves could actually be to our advantage and indicate the degree of separation of any two knots without measuring a single pound of force – we just need to be a bit clever in how we set up the tests and how we interpret the results.

The work you mentioned by murlle is very interesting, he clearly has done a massive amount of testing. I had not come across it before and the link to the data given in the rec.crafts post has unfortunately expired. Do you still have the results file and could you send it to me please, also, do you have contact details for murlle, it would be valuable to be able to pick his brains and perhaps get his views on the current project.

Derek

11

There was a post on that usenet thread on Feb 22 by ‘ben’ that appears to summarize the tests.

FISHERMAN'S BEND GROUP:

ashley equals hunters (3 win, 4 lose)

zeppelin equals hunters (5 tests) and almost ashleys (one won, 5 lost)
zeppeling lost from bulldog (4x)

butterfly: loses from ashley’s (5x), equal agaist hunter’s (3 tests)
(and butterfly is tested once agaist: zeppelin (lost) and sheetbend
(won))

double sheetbend tested equal with hunters (3tests) and ashley (2
tests)
double sheetbend (ABOK 1434): lost from bulldog (4x) (and once from
zeppelin)

fisherman’s (ABOK 1414) tested equal with ashley (in 3 tests)
fisherman’s (ABOK 1414): lost single tests from several knots
(bulldog, hunters, double fisherman’s, twin bowline)

some bends might fit in this group too:
=> viceversa: equal score after two tests against hunters (and one
lost from ashleys)
=> double figure 8 (abok 1416) lost from bulldog, equal with hunter
(2tests)
=> ring (ABOK 1412) won (once) from ashley and fisherman’s; ring lost
(twice) from flemish

DOUBLE FISHERMAN’S BEND GROUP
Twin bowline, twin ring, double fisherman’s, flemish bend

flemish (ABOK 1411): equal with bulldog (2 won, 7 lost)
flemish equal after 2 tests with: ashley, twin bowline
(lost once from double fig 8 and fisherman’s double)

double fisherman’s (ABOK 1415) lost twice from bulldog (and once from
twin bowline)
(and double fisherman’s won once from flemish and fisherman’s single)

twin bowline won from ashley(4x) and hunters (2x)
(twin bowline won once from double fig 8 and fisherman’s double)
twin bowline equals twin ring (won 3 out of 4), and almost lost from
bulldog (1 won, 7 lost)

twin ring wins from hunters (2x) equals twin bowline
(and loses from twin flemish, blood B UEK)

TRIPLE FISHERMAN’S BEND GROUP
Triple fisherman’s, bulldog, double overhand, twin flemish

Triple fisherman’s equal with bulldog (3 tests) (and won once from
twin ring)

double overhand, ABOK 1417: tested against bulldog, equal score

bulldog wins from ashley and hunters (other results are mentioned
already)

twin flemish won from twin ring (3x)
(twin flemish won once from bulldog and from twin bowline)

BLOODKNOT GROUP? Or they might fit in this group too:
=> bloodknot 2 turn symetric: won from bulldog (2x)
=> bloodknot (Buddworth UEK): won from ashley (2x), twin ring (2x),
bulldog (1x)

there are several knots that cannot be taken into account:
=> square knot, surgeon knot (they pulled apart; they are not bends)
=> carrickbend, sheetbend (in one test they pulled apart; they are not
reliable?)
=> lap, lap double, tumbling thief tucked, ashley hawser, linfit,
simple simon under, fisherman’s quad, blood triple turn, blood 4 turn
sym, blood 5 turn sym, blood 7 turn sym, blood ABOK 1413, ABOK 1417
symetrical, ABOK 1422, ABOK 1424
(there is only one test, and that is against a high scoring bend, so
no ranking is possible)
=> ABOK 1408, bowline bend (Multiple tests lost against bulldog bend,
so no ranking is possible)
=> benson, shake hands (only one test (lose) against bulldog, ashley,
zeppelin (and hunters), so a ranking is not yet possible)
--------------------

12

Thanks paulj,

Yes, I had picked that post up and doubtless it meant something to Ben because he was publishing his analysis against the background of the original report. However, in isolation I could only guess at what Ben means

FISHERMAN’S BEND GROUP:
ashley equals hunters (3 win, 4 loose) ???

Perhaps it will make sense when we find the original report or manage to contact the author.

13

So a tight turn (bowline) is weaker than a slow turn (Lyndy) and
a slow turn is weaker than no turn (Sliding Grip hitch) and
an immediately choked no turn is weaker than a middle choked (Blakes) hitch

So what should be stronger than a Blakes hitch?

I have posted before (http://igkt.net/sm/index.php?topic=863.msg5714#msg5714) that I thought the strongest knot must be one with virtually no bends or strangles, something like the plaited double

http://knotbox1.pbwiki.com/f/strongloop03sml.jpg

So, did it beat the Blakes.

In the test, the Blakes made six snap closures (but did not close on the shaft) and just before the test ended I could hear the cord making a high crackling noise. I feel that this pair weer perhaps approaching the limit of cord strength (oh I wish I had a tensiometer).

Then the bang came and this was all that was left of the Blakes.

http://knotbox1.pbwiki.com/f/Blakes%20hitch.jpg

Any suggestions as to what might be stronger than a plaited double? Perhaps a splice? But then we are leaving the realm of knots behind.

At the other end of the spectrum, what structure is going to be the weakest?

Derek

14

What is a knot?

Derek, the ProhGrip (Blake’s) that you picture is set much more snugly than
I’d advise for getting strong breaks: add at least one more turn on the (as
you picture it) left (away from hitch-SPart) side, and do not snug up the knot
so much on setting. What one aims for is some more gradual curvature in
the hitch-SPart, and also maybe some bend in the end’s reach over the wraps,
hence less tight nip at the hitch-loop’s (term vice “noose” (or “fixed noose”?)) SPart,
where you got the break. (There appears to be more frayed-ruptured material
at the knot-end of the Adjustable H. than at the same end of the ProhGrip!?)

Thanks for the sharp photos!

–dl*