Securing a sheepshank in dyneema synthetic line.

[X1]

   Yes, I know, but the purpose is to half the cost of the tested/wasted rope !  I believe that one should test each knot, or pair of knots, at least 20 times, so that the results would be statistically reliable ( and a 5% difference would correspond to 1 whole knot/experiment !  )
   There are cheap, strong hydraulic bottles available nowadays, which can do the job, I guess.  I plan to use only 12.5mm-13 mm / 1/2" inch nylon ropes, not Spectra-Dyneema ( which require more convoluted bends than the ones we use in everyday life).
   I believe that we cannot achieve any further progress in the field of practical knots from the stage we are now, without detailed - and, I would add, exhaustive - measurments/tests.

[Dan_Lehman]

... each knot will only get 50% (actually less) of the load.

?!  I don't understand this assertion.  Each knot in theory should
see half the load; variance on this could come by there being some
difference in the way the knots took the load and compressed --e.g.,
one might do some inevitable slippage & compression ahead of the
other at a point common for such a knot (but hardly predictable to
the exact same force per token!).  However, the 100% of the load
must be shared by the knots --there is nothing else to take it along
their span!

(Which differs from the more common case in which the
loaded round sling has but one knot, joining its ends --and
where the knotted side can drop under 50% via compression
(and the difference held by virtue of friction at the pins) and
so the unknotted part actually takes a greater load (which is
sort of an ideal, each side thus more equal in % of capability!).)


--dl*
====

[Cactus Red]

I should have phrased that differently - the breaking strength of the line will be reduced due to the pulleys by a small factor, so the breaking strength will not be 50% of the actual strength for each line.

[X1]

the breaking strength of the line will be reduced due to the pulleys by a small factor

  It will depend upon the diameter of the pulleys, I suppose, as well as of the shape of the grooves. But how much ? Would a drum of a diameter, say, 8 times the diameter of the rope would be insignificant/negligible ?

[Cactus Red]

You might look at a commonly used winching item, a high strength pulley block. I use these from Viking Offroad since they meet the capacity of my 15000 lb winch. http://www.vikingoffroad.com/products/ARB-9000-Ultra-Light-Snatch-Block.html The smaller less costly pulley blocks would be sufficient for what you have in mind, and still provide a substantial working load limit for safety.

[Dan_Lehman]

I should have phrased that differently - the breaking strength of the line will be reduced due to the pulleys by a small factor, so the breaking strength will not be 50% of the actual strength for each line.

But this is (1) unlikely to amount to much --YMMV per line
and pulley diameter--, AND (2) it's only relevant were you
testing to determine line strength (which, though, is a good
thing to do, vs. using rated strength, if you're concerned about
making some assertions per the line rather than per knots
relative to each other).  At the point of information we now
have before us, much can be done in just getting good data
on the breakage, leaving the basis for "%" figures able to be
adjusted upon better info as to the accuracy of tensile-strength
rating.

I recall one fellow testing angling knots who found, in testing
himself for line tensile strength, that rated strengths could
be as much as HALF the actual strength (!!) --which, you see,
would give knotted strengths a high value compared to the
rating (50% of line rated 50% of its actual strength yields
a "100%" knot).


--dl*
====

[Dan_Lehman]

I plan to measure the absolute and/or relative strengths of various bends,
by connecting the four ends of two short pieces of rope ( around 1,25 m / 4 feet each)
by the two bends I want to compare, pass the sling around two large diameter drums,
and then pull those drums to opposite directions by a hydraulic car/truck jack.
This way one consumes less material, while kreeping the symmetry of the configuration,
so the friction on the surface of the drums will play no role in the measurment.

Is there anything flawed in this procedure ?

That way you will have a survivor knot to examine;
it should've seen upwards of 90% of the force that
--knotted-- it can bear, which should be informative.

But the flaw as I see it is in the stroke-length of the
jack vs. the elasticity of the material : I surmise that
you will run out of push before the knots are anywhere
near rupture.  Perhaps there could be some other means
to applying early tension, and then the power of the
jack brought to bear at that point, where there is no
longer so much elasticity remaining!?

(One will want to guard against significant whiplash
effects of ruptured line.  Could it be done simply by
tying, in finer line hitched to either side of both knots
and that line secured to small weights?  --the flying-apart
line shouldn't have much mass in its force.  In some
laid rope, the break will come --under slow loading--
in steps, a strand or two first, with resulting drop in
tension; the pulling can be arrested at this point, even.)


--dl*
====

ps : This topic should go under Knotting Explorations, IMO;
anyone is welcome to capture this (my) post and begin it
there.  Topic : "Break Testing" might be best; I can see it
being worthwhile to pursue other sorts of testing, such as
non-break force testing (since, in general, we do hope that
our knots don't break, though they might receive considerable
force; knots for abseiling, e.g., should never get close to rupture),
and security testing (some means of continuous shaking!?).

[X1]

The smaller less costly pulley blocks would be sufficient for what you have in mind, and still provide a substantial working load limit for safety.

  The problem is that the pulley should have a deep groove, of the diameter of the rope, so the flattening of the material would be minimized. On the other hand, some will argue that this flattening might be beneficial, after all - because the relative differences between the inner and the outer track fibers, as they rurn around the curve, would be smaller...

But the flaw as I see it is in the stroke-length of the jack vs. the elasticity of the material : I surmise that you will run out of push before the knots are anywhere near rupture.

   Right - if one uses common car jacks. Moreover, if one would like to test nylon-made ropes, as I do , where the elongation is excessive, he will have to solve this problem. I plan to use two-, or even three-stage hydraulic  bottles, and small diameter slings, so that even a 30% elongation of the rope+knot(s) sling would be within the limits, I guess. Another solution is the use of "High-Lift" jacks - a solution that I think is easy, cheap and reliable, for every knot tyer to test his knots - at least for destractive tests of small diameter ropes.

One will want to guard against significant whiplash ffects of ruptured line.

I believe one should place a protective tube around the pair of the tensioned lines.
The problem is it should be transparent  - because it will be nice to be able to take pictures of the deforming and rupturing knot...