When does it make sense to tie off a section of core-shot (sheath broken) rope with an alpine butterfly knot? First the test and numbers, then the tedious discussion.

I broke replicate samples of high-quality nylon type IV paracord (750 lbs advertised breaking strength), with 11 core strands. This paracord has 55% of its mass in the core, and 45% in the sheath; note most climbing ropes have 65-80% of the total mass in the core. (Canyoneering ropes may have up to 55% in the sheath, or may use special fibers, such as aramid, to increase the sheath abrasion-resistance.)

I used three configurations:

1) no knot in middle, but tied to shackles on each end with modified trilene knots (preserve most of the rope strength);

2) “core-shot” samples, where the sheath had been broken carefully, leaving just the core strands in the middle;

3) tied in the middle with an alpine butterfly.

Results:

1) 883 +/- 82 lbs (n=6) no knot (high quality type IV often breaks above 750 lbs)

2) 585 +/- 41 lbs (n=3) core-shot

3) 451 +/- 62 lbs (n=3) butterfly

So the butterfly would probably have “weakened” a core-shot rope – and consider that a typical climbing rope has more of the mass in the core. But it’s a lot more complicated than that, so read on.

Certainly, the butterfly tie-off will make people more comfortable in raps (though it requires everyone know how to pass a knot). I have rapped over a rope with completely broken sheath 30’ from the end, and the sheath moved down 2’ and scrunched up for another 10’ – and that was with a Munter hitch, not an aggressive rap device. If another person were to follow, the experience would be “unpleasant.”

But I’ve seen butterflies used to tie off core-shot sections in odder places, like on an old rope used as a static sling on a rock horn. There was a time when parsimonious climbers would cut sections out of old ropes and use them as slings, to save money.

At issue is what the sheath actually does. Lore about the strength contribution of the sheath varies a lot—some say it has “no strength” and is mainly to protect the core from abrasion. I think it’s really a lot more complicated than that.

I’ve done aggressive abrasion tests with 8mm ropes that had polyester cores and polyester sheaths, each taking about 50% of the total mass of the rope. The tests involved pulling the rope back and forth over landscaping bricks or coarse wood files, against a 10-lb weight. The sheath actually eroded rather quickly, then the core slid more smoothly back and forth over file, with a slower erosion; I was shocked and mainly wondered if my test were realistic. But it seemed pretty clear that a polyester sheath offered no magical protection.

I would argue that the sheath adds strength to the rope mainly by holding the core fibers tightly together, and keeps them from splaying out and getting tangled on every random projection. Really, I can’t imagine handling a bunch of fibers that were free to move everywhere. Any knot tied in just core fibers would become a mess. In a rap, a broken sheath that catches on the descender and either jams, or rides off the end of the rope, is just no fun.

But a rope that just sits on a horn as a sling? You could argue that the sheath provides more UV protection for a nylon core; but that protection is very hard to quantify. Nylon for ropes often has added UV adsorbers and stabilizers, and the dye may add some UV protection. It is hard enough to determine if the sheath is made from type 6 or type 6,6 nylon. Manufacturers of climbing ropes are usually mum about these additives, and if they are in the core at all. Certainly high-tenacity polypropylene cores are at the mercy of UV.

It isn’t really possible to test the strength of a kernmantle rope by testing just the sheath, and just the core, and adding the strengths together. When the sheath alone is pulled, the lay angle quickly decreases, so sheath is possibly stronger; and when the core has a sheath around it, the fibers are held close together, so they interact and are less likely to pop one at a time.

I’ve pulled sheaths (off kernmantle ropes) and tested them, and they broke at about the strength expected for their fraction of the fiber mass in the rope. So much for “no strength.” But they stretched a lot before breaking, reducing the lay angle from about 45 degrees to 30 degrees.

(Currently, my set-up won’t allow me to put more than 2200 lb force on a rope slowly and predictably, as I use a come-along to supply the force, and a crane scale for measurement. I can use my car to pull, but it is very quick and the crane scale doesn’t capture peak force well. Plus it uses up a lot of old dynamic rope as a buffer. Unknotted climbing ropes break at 4000-6000 lbs. Hence the kernmantle paracord.)