[Dan Lehman]

Originally Posted by Brion Toss

The most recent one was on some 5/8" Vectran, as I recall, and it broke in the high 90's

Actually, this raises another issue: Cordage Institute standards required that conforming
vendors use splice strength AS the material's tensile strength; they made some bit of noise
about this in contrast with methods of calculation used internationally. Naturally, this begs
the question as to real rope strength, what with some particular splice being as it were
defined to be 100%!
(This is mentioned e.g. in that Practical Sailor Sept 2001 issue carrying their purported
re-testing of your Sail article's alarm about knots being so weak in hi-mod cordage.
--that great testing with Aramid Rigging, seemingly somewhat associated w/Yale, who "knew
a trick or two" about dealing with the new-fangled ropes, but whose eye splice(s) in Yale
Light pulled out before a Bowline broke !! And then they reported THAT value in the table!)
So one question is whether vendors are following the CI recommendations on advertised
strength.

You ask how invisible stitching could hold well.

Rather, I asked how such simple back'n'forth few strands of stitching such as I've seen
recommended could hold well. Re "invisible", I simply asked What's that? The Grizzly
stitch splice I think has many bindings of a braided core, and is somewhat more of a
seizing between two rope parts. I don't see a single yarn going through cordage
as finding much purchase with which to hold.

And I believe you are mistaken in saying that stitching puts less material into the rope; a whipping has 6 or 8 frapping tucks through the rope, and sometimes only 2, whereas one can stitch as many times as one wants -- though I usually say 6 to 8 is fine. Not only that, but those stitches can be made at a wide variety of angles, and can travel linearly and radially, intersecting the core in more places.

Actually, I don't see a whipping putting ANY material through the rope, but as tightly
compressing it. Again, I mostly whip small stuff (3-12mm), and prefer to haul tight an
Extended Strangle, or French (and was just playing around with a sort of doubled
half-hitch in that--a structure presented by Geoffrey as a decorative hitch, but which
has some appeal qua whipping). This ExtStrangle has an extra crossing of its ends, and
enough wraps to cover, and usually a finish of one end with a sort of Common Whipping
or Blood Knot binding (the end being a tucked bight, initially done to not have to size and
waste the whipping thread!). Give a few squeezes with pliers during tightening, which my
belief (hope?) is helps distribute tension at the hauled-tight ends into inner wraps.
And, again, all this for a low-load squeeze when finger-trap sheath squeeze hasn't risen
to the task.

As for inspectability, again, why would you need to inspect it, any more than you would need to inspect the "invisible" buried rope tails?

Because this stitching is only a few bits, and breaking it somewhere makes for a big loss,
percentage-wise. Assuming that the splice gets beat up enough there, you might not so
readily see that the stitching took a hit; no way can something be happening to the core,
and if whipping suffers greatly, that will be obvious. Again, for use by arborists where the
splice is in plain, frequent view.

And yes, you could say that the splice is stronger than the rope, but then you could say the same thing about a Bowline or any other knot

And this could be tested by ... ?! The claim has been made; e.g., On Rope 1st ed.
carried some caver's claim for the Triple Fisherman's Knot. But this surely resulted from
a naive testing & interpretation of results. You have to transfer load from the rope to the
additional parts of a knot or splice, and that takes friction and isn't going to multiply the
load bearing.
Sometimes angler's knots are given high claims; this comes from using rated vs. actual
tensile strength (of which there can be a huge discrepancy in esp. gel-spun lines!!)

For instance, for years I was emphatic about the need to taper any splice. And it almost always does make a significant difference in rope strength, as well as being less likely to chafe than a square-shouldered finish. But destruction tests on 3-strand ropes consistently showed that untapered splices were stronger.

I'm interested in exactly how the competing splices were made. E.g., if it's the case that
the taper began at a point where the untapered splice simply ended, one might speculate
that the tapering simply aggravates a build up of torsion in the strands. As opposed to an
engineering principle of If it breaks, take some material away from it (well, that's a bit
of the rule in swaging bicycle spokes.) So one might thus back out one strand 1, another
2, tucks, and yield a by-strand aka West Coast/California taper!? (some of the eyes on the
Cape May - Lewes ferries have 3-4-5 tucks in this way, with a common start; using that
Pro-Splice start I think puts one 1 tuck ahead in one strand and you'd have 3-3-4, etc.)

Thanks,
(-;
[now to see if the system wants to ID me yet again (yep)]