Looking for a knot that can be tensioned.

[[Inkanyezi] gone]

How about twisting?

Just the same way as you charge a rubber band motor.

This will cause an uneven stress through the rope, one set of strands, half of them, less tensioned than the other, although it may facilitate tensioning, if for example the anchor on one end is resting on an axial ball bearing, then rotated, maybe until the bearing collapses and locks from the compression load, otherwise locked by other means. Tension on the rope may be calculated from the torque and rope diameter.

[DerekSmith]

HI Injanyezi,

That is an interesting 'twist' to the solution.  And Amsteel has such a low elongation under load, a 5m length expands by only 5cm under 30% loading.  A double line shrinks by one rope diameter per turn, so in 11mm Amsteel, you would only have to give it five turns to tension it to 30%.  There would be ca 80 braid turns in a 5m length of Amsteel, so one group of fibres would be wound up to 85 turns, while the other six were 'unwound' to 75 turns.  So one group of fibres would be loaded to ca 60% while the other group remained essentially unloaded.

I wonder how a rope would perform under conditions like this, with essentially a whole load of redundancy built in against failure?  The rope would at first be stiff against load, yet as it failed, it would expand yet still retain the load...

Of course, in this application, the result would still lead to a failure of the concrete beam as the beam would only be prestressed so long as the cable did not elongate.

Derek

[SS369]

Looking at this challenge, considering the material to be used, I hate to state this, but I think knots are the wrong way to go, so far. 

I would consider a spliced sling where one end could go around the metal anchor bar (of as large a radius as possible) and the other end of the sling around a opposing wedge device that incorporates similar radius at the contact to rope face location.
The low slope wedges can generate a great deal of force. I have used them to lift houses that have settled foundations.

Attached is a crude drawing of what I am offering for your consideration.

SS

[DerekSmith]

I have to agree Scott,

But I would ask, what criteria would be taken into consideration in making the choice between spliced sling vs eyspliced braid?

I note that there are a huge array of uses for Amsteel Blue that use locked Brummel eyesplices.  Do these have some advantage over spliced slings?

One aspect of solution which came up in the discussions was the protection of the tensioning medium from damage when the beam is being located.  If the anchors sit on the end of the beam as you have depicted, then some sort of protective cap needs to be fitted.  Alternatively, one end at least could be recessed into the end of the beam, but it might be difficult to engineer a system of recessing both ends that also allows the use of wedges.  Any thoughts?

Derek

[SS369]

Hi Derek.

Actually I just thought that a spliced sling would only have one join whereas the other has two. Plus it is a double strength tendon.

Then there are splice-less slings and that would be even better. Round slings are generally of a continuous fiber wound around a sizing form and then outfitted with a braided sleeve to contain and protect the fibers. The fibers can then form to the load attachment needs and the shape of the item being hoisted.

Yes, the accessibility, both good and bad, to the tensioning method could be addressed by creative blockouts cast into the concrete forms, along with the thrust plates. Effectively making recesses where the ends and device will be located. 
More for the engineers to figure out. 

It just seems to me that knots aren't the right tool for this job. There are so many factors, safety and longevity, not to mention cost, to be considered and rigorously tested before implementing.
They are already using carbon fiber tendons for these type of applications and use specifically engineered anchorages.

Re: OP. Knots have been shown to severely weaken this material (Dyneema likes straight line tension,  it has creep and amazing slipperiness) and their trustworthiness (knots) are still pending further exploration. So, I believe there will have to be a marriage of sorts between the rope and mechanical fasteners. Such as wedge type and mortar/epoxy style.

This has been very educational and I am learning some very interesting things.

SS

[TheTreeSpyder]

Reverse Paradigm:
(Though small experience with amsteel/dyneema slip/low stretch, applying rope stuff i use to scenario):

Don't put knots in highest tension parts, use leveraged angle sweat/swig to tighten line not inline pulls/jig compressions.
Pillars, slick for slide, but right enough amount of friction to quickly call up; and keep purchases taken from tensioned line length.
Finalize with tensionless hitch type strategy that doesn't deform the loaded tension part with own rope part, perhaps even spiral around 1st turn around spar for many nip  points w/minimal deforming main tension from pure inline, then just keeper/low tension eye around tension line w/minimal deformity.



This is an adaptation of swig/sweat line, gives 2x1 leverage  to bend tension line, by "Muenter"/backhand Turn(?) on spar.

 Made to work when was in tree, could only work end on spar, but this "Muenter" has a 2/1 built in to help the bending of the main line.  Main line will wrap further around host spar, then need enough slip to slide back around spar, while not losing tension, and then bend main line again w/2:1 bend of "muenter" rinse/repeat ratcheting more line from tension /standing part to working end, can leg push off spar, impact etc.

Can convert to timber, cow, improved half etc. easily and back if careful for temp/perm lockoff.
Or  hold nip to spar tightly as lacing working end thru, around back of spar, to pull nip deep around spar at convex point.
Finish with no/minimal deformation to loaded line with tensionless hitch, types of clove, constrictor/ground hitch etc.

Can get more nips to lock by main tension (for slippery use) by sliding bight of working end under where standing part hits tree, or poking bight thru near previous nip, would then spread nips so don't bridge tension off of each other.