De Benedictis knot

[AndreaDB]

Hello everyone,
I'm happy to have discovered this forum.

I present to you the De Benedictis knot which I devised a few years ago.
The text and commentary items are in Italian.

[agent_smith]

Hello AndreaDB,

Thanks for presenting your end-to-end joining knot.
Most people on this forum speak English as their first language.

As for the originality of your presentation, I make no comment at this stage...but I am confident that others will quickly respond to such a claim very soon.

I think I found the English translations of the knots you tested:

[ ] Nodo De Benedictis: Your presentation (a type of inter-penetrating knot - similar in concept to #1415) 1280kg
[ ] Nodo Inglese Doppio:  #1415 Double fishermans 1201kg
[ ] Nodo Guida Inseguito: #1047 Figure 8  1130kg

You appear to have compared these knots solely on their MBS yield (ie ultimate strength).
The differences you have measured are not significant.
Note that #1047 F8 isn't an end-to-end joining knot (its a fixed eye knot).
#1415 Double fishermans is an end-to-end joining knot - and this is a useful knot to compare against.
Other end-to-end joining knots could have been the Zeppelin bend... and #1411 F8 bend.
I also am unclear what your experimental 'control' was in the test?

Given the type of rope you used - which appears to have been EN892 dynamic rope, perhaps your intended application is within the sport of climbing/mountaineering?
If yes, it is the considered view that MBS yield is not an important determinant of a knots efficiency or usefulness in climbing.

In life critical applications, what matters most are the properties of security, stability and resistance to jamming (not MBS yield).

In your pull-it-till-it-breaks type tests, it appears that the knot always failed on one side (rather than the other).
It appears that failure propagated from the side closest to the moving anchor and ram. In other words, failure appears to initiate and then propagate from the side where the force was injected from. Was this your observation too?

Was your knot resistant to jamming? Or, did it maximally jam?

If it did always break on one side rather than the other, what might be the reason for this?

[AndreaDB]

Thanks for the post Mark and I also take this opportunity to thank you for the excellent monographs published on the PACI website.

I have no problem speaking English even though I wanted to warn that the video is in Italian. However the translations of the nodes are correct, but the Figure 8 is the Figure 8 Bend (#1411, not the #1047)

I admit that I only performed the MBS performance test in a test center, although I personally did numerous craft tests without publishing them, including that of a user of my forum: https://www.avventurosamente.it/xf/threads/nodo-de-benedictis.6855/page-4#post-122994

What do you mean "experimental control"?

The knot was born for general use and initially I tried it with a dynamic rope because I am closer to the world of the mountains but I would also like to try it with static ropes.
Unfortunately for reasons of availability of certified structures I have not yet managed to do a serious drop test.

How can the scientifically be tested for the security, stability and resistance to jamming?

The failures in the first two tested nodes occurred from the fixed side of the test bench, in the last test (my node) the failure occurred from the mobile side of the test bench. Do you think this difference is relevant for some reason?

[Dan_Lehman]

the Figure 8 is the Figure 8 Bend (#1411, not the #1047)

For which there are several orientations.
It appears that you used the version in which
the interior (twin) strand is loaded
(which will have it pull away from its twin
into other parts of the knot).

I would like for you to re-test this knot,
AND load the exterior strands.
see image in Wikipedia
https://en.wikipedia.org/wiki/Flemish_bend
The green rope from right bears into red
twin strand on left side; vice versa on right.
(Whereas loaded on interior strands,
 the left half of knot would see the red
pulling away from green into other parts.)


AFTER setting the knot tightly in hand
by pulling on the tails (interior strands)
to build up a solid form into which those
loaded (exterior) strands will bear into,
pressing into their twin parts (i.e., into
the interior strands).  The point to the
hard, specific setting of the knot as I
described is to try to have the interior
strands (which will bear their twin's
strong force) tightened so that they
are mot merely quickly / easily pushed
aside!  (Dynamic rope will enable more
deformation than low-elongation rope,
but I think that we will see a difference.)

NB : in the Fig.8 broken strand --the piece
leading TO the knot (can't really see the part
IN the knot that is broken)--, the inner twisted
kern strands are broken IN TWO places :: I sense
that the loading coming around the U-turn
crushed these strands on the concave side
and then there was partial rupture and the
other kern strands broke.

In the two other knots, esp. the one you advance,
the kern strands appear to have broken nearly all
at once at one place, cleanly; I deduce that it came
at the pinch-point on entry (and in the grapevine
we can see that the surviving SPart is already
damaged at this point, the mantle torn).

So
70%, 65%, & 62%.  (And the Fig.8 I think
can be shown to be stronger, exterior-loaded.)

Also, it would be helpful to perhaps pause the test
machine so to photograph the heavily loaded knot.
It is sometimes hard to figure why knots break
where they do (or usually we have no information
about this); seeing them at high-load can be
helpful.  (E.g., Mark had some photos of kernmantle
rope at 3 different loads, and the shape changed
a good deal from a firm to a severe loading!)

Seeing the breaks at pinch points, esp. for the
grapevine bend leads me to wonder how the
behavior would be for a like knot (i.e., "pull-together"
(the Fig.8 is "traced", the Ashley's bend #1453/1408
"interlocked"))
to use the double overhand formed as
>>> anchor bends <<< (where the loaded
line will make the coil/turns to go away
from knot center,
tails reaching back to tuck into knot.
(So, loading like that for Blake's hitch, klemheist.)


--dl*
====

[agent_smith]

per AndreaDB:

What do you mean "experimental control"?

Every experiment must have a control group.
I realize that all you were testing was MBS yield and nothing else.
So for this very narrow test objective, your first test of an unknotted rope was the 'control' (which yielded at 1833kg).

If you were testing the effect of joining 2 different rope diameters, the control group would be equal diameters
Example:
[ ] control group 1 = 10.0mm joined to 10.0mm (EN892 dynamic)
[ ] control group 2 = 10.0mm joined to 10.0mm (EN1891 low stretch)
[ ] control group 3 = 6.0mm joined to 6.0mm (EN564)

[ ] Experimental group A = 10.0mm joined to 9.0mm (EN892)
[ ] Experimental group B = 10.0mm joined to 9.0mm (EN1891)
[ ] Experimental group C = 10.0mm joined to 8.0mm (EN892)
[ ] Experimental group D = 10.0mm joined to 8.0mm (EN1891)
[ ] Experimental group E = 6.0mm joined to 8.0mm (EN564)
[ ] Experimental group F = 6.0mm joined to 7.0mm (EN564)

Note: Variables such as rate of pull on tensile test machine would have to be identical in all tests. Also, one person ties all knots and applies same hand strength to set all knots.
...

How can the scientifically be tested for the security, stability and resistance to jamming?

There are ways to do this.
A test apparatus can be designed to induce various loading profiles such as; flogging, cyclic loading.
Slack shaking is probably best done by hand (same person performs each slack shaking test).

The failures in the first two tested nodes occurred from the fixed side of the test bench, in the last test (my node) the failure occurred from the mobile side of the test bench. Do you think this difference is relevant for some reason?

Yes - it is important.
It is extremely rare for a tester to report on this.
The break will always be from one side or the other.
The problem is that you have a statistical sample size of 1..which is a tiny sample (cant draw any conclusions from such a tiny sample size).
You need to perform a minimum of five (5) tests for each aspect you are examining.
You would then have sufficient data to draw a bell curve etc...

In my view, you should look at another round of tests that examine from which side the knots breaks.
Note: There are some test rigs that pull simultaneously from BOTH sides (ie, the test article is pulled simultaneously from both ends).
Nobody has actually investigated what effect various test configurations have on the test article.
It is all simply assumed...the default approach is to inject force from one side... and nobody gives it a moments thought. If the knot is breaking on one side or the other more often, why is this so?

...

I would encourage you to look at other aspects in your tests.
Jam resistance and the effect of different rope diameters would be a good start.

..

I have attached a couple of images to give you some ideas...

Note the duality of these test configurations.
Obviously, substitute your own knots for your test.
The advantage of testing in a mirror configuration is that it is easy to load the test article in your machine (ends don't need to be clamped).
Also, you will have a surviving knot specimen (one knot will always break first - they never break at the same instant in time).
This mean you will have a survivor specimen to examine.
The survivor will have been maximally loaded - so this is a great opportunity to see if it has jammed.

[KC]

Devil's Advocate:
What if (when) matching knots in either test don't break evenly?
.
We are making multi connections in solid structure;
1 can slip some to relieve as other tightens especially in parallel build
>>not exactly same as mono connection w/o other relief point.
.
Perhaps in some ways the best tests we have as shown;
but i think this Achilles's Heel should be observed suspiciously as compound data.

[Dan_Lehman]

Devil's Advocate:
What if (when) matching knots in either test don't break evenly?...
but i think this Achilles's Heel should be observed suspiciously as compound data.

While the testing might be less than perfect,
it will likely be well better than that typically
done, esp. if getting a **knot strength** value
is the goal.  (Whereas the trad. one-knot-formed-sling
test though corrupting the real strength of the knot
via imbalanced forces on each side due to knot
compression,
if that's the fact of one's SLING, then that is the
value relevant --in a way, it shows that one does
a bit better than "double the knot strength".

And with the two knots, one should have a survivor
to examine for near-rupture state.
BUT, yeah, I sense that the length of a test specimen
is at times rec'd to be longer than I'd like (if buying
the material!), the concern is for otherwise with a
short span exaggerating local effects of imbalanced
loading which in a longer specimen would be worked
out across that longer span.


--dl*
====

[KC]

To my imagery, a short line test for this 'cramps'  the wave form pulse thru
>>cosine pulse, then sine pulse waves; just like in vibration, electricity, sound, light the cosine wave pulse first, then sine later
>>which is only befitting if that is actual circumstance; and then only as an aspect.
.
But, as you say can't offer better test;
but caution observations should possibly be noted/weighted as test method;
for mono never handles same as multi/duals in other stuff seen.
>>especially more true in parallel, rather than series build i think.

[agent_smith]

per KC:

Devil's Advocate:
What if (when) matching knots in either test don't break evenly?

??

The 'matching' knots never break evenly!
One will always yield before the other.
It makes no difference if the knot specimens are aligned axially/linearly or if they are in parallel (within a round sling).
There will always be a 'survivor' knot specimen that has been maximally loaded.
I have personally conducted (or been involved with) well over 400 MBS yield tests of knots (linear and parallel configuration) and I've never observed both knots failing at the same instant in time.
This of course assumes the word 'evenly' means at the same time? Or do you mean the failed rope segments in both knot specimens are identical?
If you mean the same rope segments in both knots don't break at the precise same point - this is also something that never occurs in real world testing.
A human ties each knot specimen by hand. If the same human ties both knot specimens - no matter how hard that human tries, there will always be slight variations.

We are making multi connections in solid structure;
1 can slip some to relieve as other tightens especially in parallel build

Although according to Newtons laws, force should be the same at both ends of the test rig.
Both knots will be subjected to the same force.
However, as the knots tighten tension force is converted to heat energy. In all MBS yield tests, there is evidence of melting/fusing of synthetic rope fibres. This is clear evidence of heat (can also see this heat with thermal imagery).

>>not exactly same as mono connection w/o other relief point.

There has been endless debate about reference frames in relation to MBS yield testing of knots.
Test rigs generally have the ram located on one side.
No one has done a serious peer reviewed study if pulling from one side (or the other) has any measurable impact on the test results.
Also 'many' would argue that the fixed anchor post on a test rig is actually regarded as a 'moving reference frame' relative to the knot.
The ram end (which pulls to one side) is also a moving reference frame.

...

In any case, you don't appear to have offered up an alternative test rig configuration for AndreaDB.
Do you have a suggestion for an alternative knot test rig configuration that you could share?

[KC]

...But, as you say can't offer better test;
but caution observations should possibly be noted/weighted as test method;
for mono never handles same as multi/duals in other stuff seen.
>>especially more true in parallel, rather than series build i think.

.
Even tho can't myself offer better,
Still would like to offer to stare weakness of model straight on as use it.
>>to know well enough to realize what variances serving to perhaps more mute or accentuate outcome.
.
If 2 legs support to load, and one starts slipping, the other instantly takes more load and becomes more rigid.
So Arborists  on rope folded over support and 1end to saddle termination and running end thru friction hitch to saddle can descend quite well vs.  Friction hitch to saddle only, single line support.
(arbo on folded line also has 2:1 less support frictions over self/lifting self in these mechanix)

[agent_smith]

per KC:

If 2 legs support to load, and one starts slipping, the other instantly takes more load and becomes more rigid.
So Arborists  on rope folded over support and 1end to saddle termination and running end thru friction hitch to saddle can descend quite well vs.  Friction hitch to saddle only, single line support.
(arbo on folded line also has 2:1 less support frictions over self/lifting self in these mechanix)

How is this helpful to the original poster (AndreaDB)?

Are you now discussing a new line of theoretical argument that belongs in its own separate topic thread (in the realm of theory)?
Or - is your information providing something useful to AndreaDB?

EDIT NOTE: Refer this link: https://igkt.net/sm/index.php?topic=6332.msg42542#msg42542

[KC]

Just a different view/facet of this chase I try to lend in discussion form.
Honestly what I see as a working principle invoked into the mix that should be watched for I think
Because as active component could slant some results and comparisons.
Hopefully not mostly, but even as a minimized value , has promise of lesser and greater impact in chain ...
If not on guard watching for element, could sneak by as something else.
.
I say these things not exactly because can point to precise happening,
But rather in bigger scope of in such active component in mechanix ,
Always guard against such things working against target as a lifestyle I guess of dealing with forces.
Especially in 'sanitary' isolation of components to understand and define.
.
Would as well expect some pattern of differences at different points for serial or parallel loading of knots.

[Dan_Lehman]

per KC:

Devil's Advocate:
What if (when) matching knots in either test don't break evenly?

??

The 'matching' knots never break evenly!
One will always yield before the other.
It makes no difference if the knot specimens are aligned axially/linearly or if they are in parallel (within a round sling).
There will always be a 'survivor' knot specimen that has been maximally loaded.

Ah, but I saw a case where the survivor had started
to break --in a water knot in tape at the, um, tip of
the center triangular hump --whereas the full break
of the other seemed to be at the entry squeeze.

.:.  It could be the case that a partial break in one knot
eased force on that side (pin friction too much to equalize)
and then the other side might go !?

All in all, though, I think that one gets good info this way,
and indeed more & better info if one must limit testing!

--dl*
====

[SS369]

We are swinging wide of the OP's inquiry. Lets see if we can give AndreaDB some info on the knot presented.

SS

[KC]

To be totally clearer,
I'm not saying I can do better.
.
Devil's Advocate of turning paradigm backwards etc. to find Achilles's heel , for there must be one by 'law'.
>>Then play forward on the strengths, while guardedly watching the weakness(es)
>>just as in rigging etc.
I think is 'stable' at most points, but must slant things some
>>so therefore will slant things at some points more than others