Two by Four knot (or what knot is this)

[X1]

   Thanks. It would probably be more useful to many people to have the detailed results in an Excel or similar form, so they can do their own studies on them. Do not forget to include this "Slow loading" somewhere in the label. Perhaps in the future you will perform the same experiments with an abrupt loading.

[X1]

   Although the ("added"- not actual ) strength of the corresponding two straight, parallel lines rope system is almost 50% stronger that the two linked bights system, if we would test those same systems with ropes of a different structure/cross section, this difference could be even bigger. Why ?
   We must not forget that J.P. tested 3-strand laid rope. So, at each cross section of the rope, there are 3 partial, circular cross sections, one circular cross section for each one strand. That means that, at a turn of this rope, two strands can/will occupy the "inside track" of the turn, leaving only one "outsider". So, the differences in the lengths of the paths of the individual threads, which over-loads the out-siders and under-loads the insiders, and which is responsible for the reduction of the strength of a rope when and where it is bent, should be considered cautiously - and differently, in the case of 3-stands rope, than in the sase of a more uniform rope, with one circular overall cross section. In fact, the 3-strands rope is able to be "flattened" at a U turn, more than a rope with a circular overall cross section.
   So, of the 100% of the area of the cross section of the 3-strand rope, only the 33% belong to the outside threads, and 66% belong to the inside threads. This might have helped the threads of the 3-strands laid rope J.P. had tested, to withstand the load more efficienly than the treads of a 7 strands rope, for example.
   Just a theory - but, when we are confronted with experimental data different than the ones we had expected, we must try to re-inforce and re-adjust our theories - before we remove/throw them out of the frame/window. 

[James Petersen]

   
   We must not forget that J.P. tested 3-strand laid rope. , for example.
 

I should clarify: the string used in my last set of tests -- the bight-to-bight tests was braided, size "B" (one of the smallest sizes) braided Chinese knotting string. I regret any confusion this may have caused. I am including a photo with this post.

[X1]

   Oh ! You destroyed my last hope for a decent explanation !  Gone is the 2+1 strands "flattable" 3-strand rope theory, which sounded plausible to me ... Anyway, as we said, we should not confuse hopes with expectations. The linked bight-to-bight system, although weak, is stronger than I had expected / predicted. However, I am glad on its behalf ! 
   Now we will have to wait for Mr. Lehman, who is a specialist in figuring out "loopholes" and peculiar "variations"  .

[X1]

  J.P., as the knots shown in the present thread are two-round-turns-to-two-round-turns bends, it would be probably useful if you also test the twin-bights-to-twin-bights system (= double-line-bights-to-double-line-bights system )( similar to the one you have already tested, but where the single lines would be replaced by twin/double lines ). I expect that the results would be a little better, in relation to the corresponding 4 straight, parallel lines unknotted system, or to the single ( x 4 ) unknotted line - because, the 2 ropes of each bight, the one next to the other ( no twisting involved ) would function as a kind of a strap, of a webbing, and the 2 strands of each double line bight would be flattened a little bid when they make their U turn around the 2 strands of the other double-line-bight. So, the U turns of the strands at this double line system would not be as sharp as they were at the single-line-bight-to-single-line-bight system, and this should spell even better news for it re strength.

[Dan_Lehman]

I decided to sacrifice it to the question of how the strength of two interlinked bights
differs from the strenght of the line that they are made in.

It would be of interest to me to know how bight-2-bight
compared with their being interconnected in a granny knot
structure; one could also try the square.  The former
has a nice look to it.  And unlike pure bight-in-bight, there
is no ultimate point of high-pressure contact, so an interesting
case.


--dl*
====

[James Petersen]

I decided to sacrifice it to the question of how the strength of two interlinked bights
differs from the strenght of the line that they are made in.

It would be of interest to me to know how bight-2-bight
compared with their being interconnected in a granny knot
structure; one could also try the square.  The former
has a nice look to it.  And unlike pure bight-in-bight, there
is no ultimate point of high-pressure contact, so an interesting
case.

--dl*
====

I had thought of the same thing, but up to this point have been trying to keep things as simple/straightforward as possible -- which is not always simple or straightforward. I have presently completed about half of a set of tests involving some of interlocked EOL-loop-bends  against their bend and loop counterparts. It is worthwhile to note that to date, the loop-to-loop bends have yet to fail within a loop -- they have all failed where the standing end enters the knot.

[Dan_Lehman]

   Oh ! You destroyed my last hope for a decent explanation ! 
Gone is the 2+1 strands "flattable" 3-strand rope theory, which sounded plausible to me ...
Anyway, as we said, we should not confuse hopes with expectations.

The linked bight-to-bight system, although weak, is stronger than I had expected / predicted.
However, I am glad on its behalf ! 
   Now we will have to wait for Mr. Lehman,
who is a specialist in figuring out "loopholes" and peculiar "variations"  .

I suppose I had in mind the general rule-of-thumb "50%" in mind,
too, in seeing why interlocked eyes should be strong : that there
would need to be nearly 100% on the SPart in order to reach
the relatively (individually!) weak 50% threshold and cause trouble
at the eyes.

But going to about 150% and J.P.'s own guessed 70% per
bight leg is well beyond what we should have guessed,
if based on knots-strengths alone (where one will often
see figures on either side of 70%, and with greater values
deemed good).

Now, I am called upon to offer up "loop"holes for this!?

Let first note that we're looking at just one set of data
and all the factors associated w/that --i.p., measuring.
Beyond that, perhaps the "flattening" effect has some
merit in that at such thin material there is little difference
of fibres in relation to the bending (which harkens to the
thread on "thin lines" strength)?

Also, in the tension being delivered on both bight legs
--in contrast to a SPart's coming to a similar sharp turn
and pulling in one direction only--, there is elimination
of any much movement at the bend and so some
reduction of frictional heat weakening !?  !?

Still, 70% vs 50% is a big gap.  (And now one must
wonder what one sh/could expect of a granny structure
there : what of the weakening effect surmised/guessed
for the bight-2-bight joint could it redress?  --having
pressure more distributed?  (There should still be the
lack of movement, given the bight loading.)


--dl*
====

[X1]

perhaps the "flattening" effect has some merit in that at such thin material there is little difference of fibres in relation to the bending

  Fewer fibres means fewer insiders and fewer outsiders, that is true. However, for those fewer fibres, the U turns are sharper, because the diameter of the strand around which they are forced to bend is smaller. I can not imagine how one can relate the benefits of the former to the shortcomings of the later...

the tension being delivered on both bight legs --in contrast to a SPart's coming to a similar sharp turn
and pulling in one direction only--

  Would this tension have already been equalized - or would it remain equalized - at the final moments of the rupture ? Can we say that there will be no slippage, however small, of the one leg of each bight relatively to the other ? We have to watch "markers" attached to the legs, to be sure about that.
   I believe that a movement in the knot s nub would be less detrimental to the strength of the knot, than a movement in the legs. Inside the knot s nub it can be dissipated in a greater area, while inside the legs it would enhance the weakening "saw effect" on the dangerous eyetip-to-eyetip contact point/area.

   I am still not convinced that the final numbers of the eye knots-to-eye knots compound bends would come out of the actual experiments in such a perfect shape...
   The "golden standards" of bends that J.P. is testing are not meant to be "golden" in relation to strength...( except the retraced fig.8 bend ). Perhaps we should include some bends that would have fewer general merits than the Zeppelin bend, for example, but would be stronger ( I think of the "doubled" versions of those bends - like the Double Zeppelin bend ). There are hundreds of possible simple practical bends out there - I would like to know the strengths of at least 100 well known and 100 "new" of them, to be 100% sure that the common bends are old hat, and we should focus all our attention to the loop-to-loop compound bends.
   

[James Petersen]

   I finished a series of tests involving some bends, loops and two-loop-bends. The line used was the same as for my previous set of tests -- Size "B", Braided, Flourescent Orange Chinese Knotting String.

   The order of the tests and the knots tested were as follows:

Overhand Loop---line---Water Knot---line---2 Overhand Loops Bend---line---Figure 8 Bend---line---2 Figure 8 Loops Bend---line---Zeppelin Bend---line


   At certain points (when the length of line between knots permitted) I tested the line more than once between knots. I tried to test the line after each test of a knot, which slowed things down considerably, but hopefully avoided potential differences in line strength when a series of knots is tested after each line test. I later realized that this caused a complication in calculating the results. Should I judge each knot against the average of all the line tests or just against adjacent line tests. I have decided to report both results.

	Average		Minimum		Strength		Standard Deviation
LINE	22.60 kg (100%)		18.80 kg		N.A.		0.91
Overhand Loop
Including All Data:	19.10 kg		16.1 kg		84.50%		1.15
Excluding 1 Outlier:	19.25 kg		17.5 kg		85.17%		0.94
Water Knot	17.67 kg		15.2 kg		78.16%		1.14
2 OH Loops Bend	18.61 kg		17.2 kg		82.33%		0.77
Figure 8 Bend
Including All Data:	19.35 kg		17.3 kg		85.62%		1.15
Excluding 1 Outlier:	19.45 kg		18.2 kg		86.08%		0.94
2 Figure 8 Loops Bend
Including All Data:	20.05 kg		14.4 kg		88.69%		1.84
Excluding 2 Outliers:	20.49 kg		18.2 kg		90.65%		1.15
Zeppelin Bend	16.90 kg		15.7 kg		74.75%		0.97

   In previous tests, I noticed that some versions of the 1x2 Lazy Dog Loop is comperable in strength to the figure 8 loop, so I also did a series of tests on two 1x2 variations of the Lazy Dog/round turns loop. The order for these were:

   1x2s Lazy Dog Loop, 2 1x2 Lazy Dog Loops Bend (Two Lazy Dogs Bend??), 1x2ssl Lazy Dog Loop, and 2 1x2ssl Lazy Dog Loops Bend.

   Unlike the above series of tests I did not test the line after testing each knot, but the line was tested a total of 59 times. I only calculated results based on the average of all the line strength tests, and those are the results that have been posted. Also, please note that with these results, I have included results with and without outliers included. There was one outlier in the line tests, so when they are included, the results reflect the inclusion of any outliers in the knot tests as well as the one in the line tests.
   
   The results for these tests follow.

	Average		Minimum		Strength		Standard Deviation
LINE
Including All Data:	22.57 kg (100%)		13.0 kg		N.A.		1.15
Excluding 1 Outlier:	22.74 kg (100%)		18.8 kg		N.A.		0.94
1x2s Lazy Dog Loop
Including All Data:	19.98 kg		16.3 kg		88.53%		1.64
Excluding 1 Outlier:	20.14 kg		17.3 kg		88.59%		1.47
1x2s 2 Lazy Dogs Bend
Including All Data:	19.64 kg		14.2 kg		88.02%		1.82
Excluding 1 Outlier:	19.91 kg		17.3 kg		88.58%		1.36
1x2ssl 2 Lazy Dog Loop
Including All Data:	20.68 kg		18.80 kg		91.60%		0.92
Excluding 1 Outlier in line tests:	20.68 kg		18.80 kg		90.93%		0.92
1x2ssl 2 Lazy Dogs Bend
Including All Data:	20.05 kg		17.5 kg		88.82%		1.82
Excluding 1 Outlier:	20.18 kg		18.6 kg		88.77%		1.36

[SS369]

Hello James,

I for one appreciate your keen interest, dedication and good work!
Interesting results and some fall in line with my expectations.

I am also interested in where the failures occurred. I don't suppose that you have any photos that are suitable for inspection?

Any observations during the tests that might shed some light on the knot(s) movement that would influence the breakage if at the knot itself?

Thank you.

SS

[X1]

   I this a typo ? Exluding the outlier / minimum of 13.00 kg. the average should go higher, not lower ( 22.57 - 22,54 ).



     

[James Petersen]

   I this a typo ? Exluding the outlier / minimum of 13.00 kg. the average should go higher, not lower ( 22.57 - 22,54 ).
 

Indeed it was a typo. It should have been 22.57 - 22.74. You caught me once again. Thanks for your attention to detail.

[James Petersen]

I am also interested in where the failures occurred. I don't suppose that you have any photos that are suitable for inspection?

My feeling is that with most of the knots, failure occurred where the standing part enters the knot, but that is not based on detailed observations. I will look through some of the knots and see if I can find any that can be photographed -- yes, I still have all of the knots, stored in plastic zipper bags with the other knots from each iteration?/circuit?. Was there any knot in particular you were interested in?

Any observations during the tests that might shed some light on the knot(s) movement that would influence the breakage if at the knot itself?

Unfortunately, my scale doesn't lock on the highest reading, so my eyes are glued to the display (which is between my eyes and the knots) during testing. In a perfect world, I would have made videos of the knots during testing.

[SS369]

Of course I am interested in all the knots. Any of the specimens that you have suitable to photograph would suffice. Closeups of the broken area, perhaps naming the type of knot please.

Yes, video would be great and slow motion at that, but it's probably a tall order and I would expect some protection to the camera would be in order.

If you're incline to repeat any of the tests, maybe these could be with watching instead of recording.
Safely.

Thank you.

SS