Let's leave that different ropes can have different behaviors.
There are some kernmantle ropes (PMI's standard (No-Flex, my name) "pit rope", e.g.)
that really don't want to bend to even a 1-diameter radius; but there are some laid
ropes --hard laid-- with similar intractability (and forget about splicing them!).
But there are some more flexible kernmantle ropes (PMI's "EZ-Flex"), though
these tend still to be moderately firm & non-bending.
FYI, my evolving definition of "kernmantle" (and I favor "mantle" vice "mantel"
for its English (at least) significance) is {{a rope composed of an outer sheath
around a core that itself is a multiplicity of fibres or cords, not able to be used
qua single rope}}. This contrasts with some definitions that I've seen in which
it appears that the relative novelty of climbing/caving/SAR ropes is only lately
being accommodated. Granted, probably most of what is called "k." is in fact
meeting these narrow definitions, but it seems better to define the term in a way
that is more useful vis-a-vis rope construction. If the core cannot stand by itself,
then ... "kernmantle". (A "two-in-2" double braid thus doesn't qualify; and there
was one (Rocca?, Spanish?) rockclimbing rope that had a braided core, which
would make it non-k. by my definition; I don't loose sleep over this.
There are some rather big ropes made with quite soft lay that bend over
backwards with no complaint.
With a firm cross section, one has smaller contact patches (theoretically,
a round cross section would yield sequence-of-tangent-points, linear
contact areas, yes?); with compressive rope, the contact patches get
planar. The so-called (by me) "reverse groundline hitch" is a common
commercial-fisherman's knot that depends on relatively flat (as thoug
compressed) cordage, which is even often used in adjacent pairs,
thereby exaggerating the width dimension vs. thickness, and getting
relatively large contact patches.
I will revisit my brief test of a mid-line eyeknot tied in 12-strand Spectra
--a very slick multifilament material, in a very smooth construction--
which surprised me by holding, for some small but pulley'd extra loading
of maybe 200#? And then the same knot (approximated, we might
say, re pure geometry) in firmer cross-section kernmantle accessory
cord (6mm, nylon, not new, though hardly much used but for age),
seemed to not hold even in my manual loading (IIRC) !! --which I see
as an effect of material firmness, with a result contrary to material
slickness.
And beyond flexibility is another factor, *springiness* !? I found some
soft-laid but rather hard cross-sectionally 4mm? diameter polypropylene
cord to refuse to honor the EBDB's security, holding only briefly
before gently loosening all 'round (!!); yet the Janus bowline with
its sharper turns of the bight did not loosen so much --the bights
tried to open like scissor legs and the turNip resisted this easily,
and that was all (whereas in the EBDB all of the turns just
widened in diameter!). --this was quite a surprise. I still think think
that the EBDB is one of the most secure bowlines for many
ropes, but not for that cord.
--dl*
====