i really feel that understanding a device that we use so often in so many ways (such as rope and knots) is very important. i've read over the years and fears hear; where folks favour this or that knot; never untying there termination or eye in knotless rigging, favouring splices strength over the other properties of knots (not wearing the same spot, not stiffening such a large part of the line; dampening properties etc.) that this study speaks directly to all. We also use knots in rigging that take more dynamic loading than just the climbing that this paper focusses on. We speak of SWL; but that is not the only measure of safety/ of putting as much science on your side as possible.
This is probably not perfect, but; hopefully the imagery is closer to center/truth.
My model imagery is:
In ropes that are straight; under dynamic conditions; we have immediately recoverable, later recoverable and non-recoverable elasticity properties. We macho tree guys only look at tensile, more is better. But to me the more delicate, fergiving and violatable 'feminine' propertease of hold/don't break (like tensile) but give room for distance/frictions to buffer/forgive-without breaking are the more elusive and easier lost, and many times in our operations more essential to life knot being lost).. The more tensile to loading ratio we have, the better we feel; but under dynamic conditions; more tensile to loading ratio is too macho and a loss of the elasticity; that works as a reciprocal-the lower ratio of lading to tensile gives mroe dampening!! So maybe; when the say to have a 4-10:1 ratio they really mean a 4-10:1 ratio and not 50:1!! Too much of a good thing... Maybe these decisions are maximum at much more brain thought than just flooding a system with strength; with any old knot...
Liken, this to a steel/non-dynamic response cable; that we can install different propertied rubber extensions on; to then emulate the aforementioned dynamic properties of ropes. Also, these properties would depend on the property/material, it's construction(to give some variance in properties in same materials) and the percentage of the tensile loaded. Whereby, a 5k tensile loaded with 1k will have less dynamic buffering than a 2k load on same device. This also means that; a 2k load on a 2:1 support will have more 'headroom' as far as strength; but each leg will be only loaded with 1k. Whereby, a pulley with a 2k load; will place 4k on support of pulley, but on a 2:1; it will place a 3k load on support; so will be safer...As a STATIC load. But, as a dynamic load, the 2:1 will offer less buffering of shock; so will give more shock to the pulley support!! Also, of curse the length of the buffering device (rope or rubber on end of steel cable) will give more buffering. So, if load is caught closer/higher to pulley, there is more shock etc.; like we do. So, this is serious stuff, with many variables.
Enter knots into the system. Now, several things change. The buffering we mentioned doesn't kill force off, but rather trades it for heat force/ internal frictions etc. (Energy forces are never created nor destroyed, only perverted) A knot can have have more frictions; to drag rope through from one side of the friction to the other. A double bowline or triple noose; can have a bigger bank of loops to pull the slack from one side threw the gauntlet of frictions of the knot; than a single bowline or double noose respectively. Also, each may try to use their turns to compress the host line part inside the turns. But (in my mind/imagery) the noose also provides another dampening/buffering agent. It can slide down to seat tighter, as the turns are also tightening, increasing it's friction tradeoffs etc. for energy conversions etc./buffering of dynamic forces.
But, with a knot, with these extra components, we also get into variable range recoverability; or until the line is relaxed/untied to get immediate and post recoveries. If immediately retying, the longer recoveries might not have a chance to 'reload' their dampening forces etc.
In our static/no dampening steel cable example; we can use more than 1 type of rubber dampner; to give different ranges, mixing of ranges and probably even multipliers of self adjusting ranges instantly; as Nature has more chances to choose from to assume Her all-mighty minimum loading scenarios. IFF("if and only if"; logical conjunctive); we lend Her these tools. Taken to the rope not cable scenarios, the rope has 1 buffering range/dynamics/properties; and the knots offer a complimentary set of same, per the theories of friction conversions of dragging, sliding and compressing.
So, would i depend on knot only for dampening, hell no!! But, by this imagery i keep; i do look for extra complimeanting properties; and the need to untie knots after loading and in storage; to recoup the properties into the knot/friction machines. Also, i do think that an undressed knot can be too loose and serve to impact 'close' and not serve properly.
The report bares these theories out i think. It is not the single properties but the orchestration of many devices/properties; working instantly at different times and combinations that make the difference. If these different devices were different pieces of rubber on a cable, or different types of rope strung together,the dampening devices could more immediately recover/reload their helpfull properties, than with a knot not untied. The specific knot properties for dampening perhaps wouldn't even be assumeable by straight line. Also, look at triple fisherman's bend, spaced some; so they could drag and do same; even with Square/ Thief in center.
Also, this is one of my 'cases' against splices; they are macho strong; but not feminine 'strong'. the other case against'em; is that a knot stiffens less line; which works to this topic too, but also give a larger area that shouldn't be bent, for it would then be leveraged more weaker, by virtue of it's increased resistance to said bend X the larger diameter of the bent axis as a multiplier. So, a short splice would give less area we needed to take such care of, but would probably be thicker, so give larger multiplier...
Or sumetin'like dat
[ QUOTE ] "Nature to be Commanded; Must be Obeyed"
-Sir Francis Bacon