Chockstone Forum

      Sponsored By
      ROCK
   HARDWARE

Chockstone Photography

Australian Landscape Prints

Chockstone Forum - Accidents & Injuries

Report Accidents and Injuries

Topic Date User

fall at ton sai 28-Dec-2003 At 11:42:36 AM V

Message

On 27/12/2003 Mighty Mouse wrote:
>it was my understanding that a loose knot was a strong knot (i am not
>absolutely sure why- i think its because that way it can absorb more energy
>through the tightening process-its just what ive always been taught) so
>while it is obviously important to check and dress a knot, tightly cinching
>it may not be the best way

Not quite... When they say "loose knot" in this context, I think it means a knot where the various loops do not go through tight corners. As far as I know, a knot is pretty much always the weakest point in a rope. [OK, entering engineering statics theory here...] Strength is proportional to cross-sectional area, so the "flatter" the rope is squeezed, the less tensile strength that piece of rope has. For example, whenever a rope goes over an edge, like a belay device or a carabiner, it is flattened and thus weakened. The more acute the angle, the more flattening occurs, and greater reduction in tensile strength.

What this means is that in general, if the curves of the rope in the knot are wide, then there is less compression of the core, and the knot is stronger. So different knots weaken the tensile strength of a rope by different amounts, depending on their geometry. But leaving a knot loose does not help at all, seeing as it will cinch up as soon as you load it, and then you get the same squeezing effect. In fact a sudden tightening may cause friction and damage the sheath, if anything.

I use a figure-8 simply because it's easy to tie, and easy to inspect. I double it back sometimes to keep it neat. I don't know about the relative strength of bowline vs figure-8 as a tie-in, but if I ever fall far enough to find out I suspect my back will break before the knot does.

Disclaimer: my degree is in computer systems engineering, so if there are any "real" engineers here ;-) please feel free to correct me on the above.