Author Topic: Towards a Science of Knotting  (Read 2945 times)

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3764
Towards a Science of Knotting
« on: August 01, 2009, 07:10:29 AM »
Knots exist as often minor components of many activities.  Those who are engaged
in such activities more often than not have limited interest in knotting, except as
is necessitated by the activity.

Taking knotting out of a world of hearsay and story telliing, typically with
some nautical flavor, should be one of the goals of serious study of knotting.
Elsewhere I've talked of cleaning out the rubbish.

As I've noted elsewhere, we live in a time of increasingly easy information capture
(photography) and transmittal (the WWWeb; and international mails -- still good!);
with which capabilities there is the potential for accumulating data from a broad
variety of firsthand sources -- users or in-use cordage structures.

Among the attributes ascribed to knots is strength -- and it is a seductive
pursuit to learn the supposed strengths of knots.  Objective testing stands
in need of a rigorous review, to better understand its current/historical
limits (much is in vagueness of process), and to draw up standards for
making some headway in understanding knot behavior.  (E.g., after how
many centuries of tying and sometimes breaking Sheet bends has
anyone testified to which of the two differently formed ropes in that
bend (when "S.B" is understood to be an ends-joining knot) breaks,
usually?  --or that there is no generally weaker half?  And how should
one gauge the strength of the knot where differently sized ropes are joined?
In general:  where do particular knots break?  What are the differences
between slow-tensioning and repeated tensioning and rapid tensioning?

Of course, a science of knotting should have some good nomenclature
for unambiguous communication.  What exactly is a "standing part"?
-- seems to me it is something extant at the time of tying, and
not so much defined afterwards; yet it is the completed state of a knot
that I usually want to address.  "loop" is a term so overloaded one almost
always must qualify it to fit the need -- is it now that knot structure, or
a grommet-like ring, or an eye ...?

And knots classification.
There is an old thread here where some of us made some efforts to get
a grip on what a knot is, and so on.  That went not so far, but it can
be resumed.  igkt.net/sm/index.php?topic=168.0
(Even some old voices might rise again.)

And so on.

--dl*
====


squarerigger

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 729
  • IGKTPAB Immediate Past President
    • The Knot Guy
Re: Towards a Science of Knotting
« Reply #1 on: August 01, 2009, 07:57:54 PM »
Eloquently put Dan!

Yes, a science of knotting would be a good.  Yes, it would be an ideal to have authors who have their audience's best interests at heart and produce, even against their publisher's wishes, knot-books of great and credible accuracy.  For the science to be good would take the efforts of many hundreds of people, or one person for many dozens of years perhaps.  What group could afford the cost of such an endeavor?  Science must have credibility and no individual's opinion, no matter how credible they may appear to themselves, would be supported by the scientific community without having a sufficiently broad base of repeatable testing performed and recorded with accurate means which could be independently verified and tested against a standard.  Testing also involves rigs, machinery perhaps, time and of course lots of rope of many different types and sizes.

The bottom line for such testing, apart from the time it would take, is money.

Let us even suppose that time and money are available through some generous donor - to what useful purpose could the results be applied?  Certainly the testing would have to be repeated say a dozen or more times with the same type of line and then again with a different thickness of line.  It would then have to be repeated dozens more times with a different tyer, or three or four or more different tyers, each perhaps working on the basis of a standardized testing instruction on the basis of a standardized method of tying instruction.

It is an elephant and it cannot be consumed other than one bight at a time.  Certainly the IGKT could offer such testing were there any standard accepted by a worldwide community.  However, as we have seen here in this small forum, there are differing opinions in the matter of which knot is best for which application - how do we set a standard with so many dissenting opinions?  There are opinions that (broadly) agree, so perhaps a start could be made on the thousand-mile journey with those agreements.  I will try to offer what I can, but even then we have dissenting opinion as to the value of any individual's worth as a knot-tyer.

I support the notion that we should have testing and that perhaps the IGKT or some other body would be the best for the matter at hand.  I offer my own hand for the limited time that we are each available - what then? ???

SR

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3764
Re: Towards a Science of Knotting
« Reply #2 on: August 01, 2009, 09:03:32 PM »
For the science to be good would take the efforts of many hundreds of people, or one person
for many dozens of years perhaps.  What group could afford the cost of such an endeavor?
Science must have credibility and ... a sufficiently broad base of repeatable testing performed and recorded
with accurate means which could be independently verified and tested against a standard.  Testing also
involves rigs, machinery perhaps, time and of course lots of rope of many different types and sizes.
The bottom line for such testing, apart from the time it would take, is money.
There is something to be gained w/o so much cost.  Agent_Smith on this forum
has already done more with some home-brew testing than was available to us
prior to his good efforts, aided by comments from this community.  So, at last,
we got some marker threads in knots and remnant in broken specimens and thus
an indication of rupture points for some knots in some cordage -- a start.  This
technique (marking the rope; photographing said marked specimen during and
after testing) obviates the drive for renting some super-high-speed camera
(which in the case of the (Mclaren? Milne?) report concluded that even that
was too slow for good identification of break points).
And I hope to use the technique in submitting some specimens for testing.

Quote
Certainly the testing would have to be repeated say a dozen or more times with the same type of line
and then again with a different thickness of line.  It would then have to be repeated dozens more times with
a different tyer, or three or four or more different tyers, each perhaps working on the basis of a standardized
testing instruction on the basis of a standardized method of tying instruction.
Currently we see use of data that has not even the beginning of these considerations.
Surely even just explicitly recorded & reported testing will be an advance -- recall,
one usually cannot know even which end of a Fig.8 eyeknot is loaded!!!  --or which
rope in a Sheet Bend breaks?!!  JUST KNOWING THAT would be progress, albeit small.

As for the full implication of considering all the various factors, I have hopes that (as
I once wrote here) some statistical reasoning can greatly reduce the need for test
cases.  Esp. e.g. in some things that should be non-factors, such as rope diameter
-- or might be seen to have some drift in some cases of full testing (let's say
that a thickness rises, maybe % strength slightly diminishes), and then can be
assumed and spot-checked in further cases (and if something falls out of expected
range, THEN do more testing and re-thinking).

Heck, it might be that one benefit of some limited set of rigorous testing amounts
to a clear advisory:  ONLY USE TEST RESULTS OF EXACT MATERIALS ... .  That
would be progress over relying on some supposed "knot" strength which is supposed
to come with the structure irrespective of material.

Quote
Certainly the IGKT could offer such testing were there any standard accepted by a worldwide community.
I think that the prior step is this:  to collect test data/reports as they are available,
and to examine them for relevance & shortcomings -- to test their adequacy against
a battery of questions.  I think we'll find them all coming up well short of thoroughness
(and here I mean especially in detail of the report -- that it will be easy to demonstrate
that one cannot perform repeated, verifying testing, because of so many unknowns).

From this beginning, in an iterative build-&-critique process, a test-standard can be
developed.  The IGKT can apply this to new testing that comes to our attention,
and maybe proactively encourage some places to adopt it for future testing.
If that happens (adoption), then future test reports will be more useful.

Quote
... there are differing opinions in the matter of which knot is best for which application
 - how do we set a standard with so many dissenting opinions?
Well, noting the differences is only the first pass; then the differing POVs are pressed
for a rationale, for backing, and that gets examined.  As you have done, asking for
evidence/rationale against the Blackwall Hitch which you have used w/o problem
-- your usage in evidence of its (at least sometimes) effectiveness.

Unfortunately, I lost some list of misc. test reports in a dead computer,
but that can be re-collected (and maybe from the computer somehow).

--------------
Another cheap-testing example.
LCurious et al. have disparaged the Offset Ring Bend (aka "EDK (Euro.DeathKnot)", "Thumb
Bend", "#1410", "Dbl.Overhand Knot", ...).  (I like "ORB" because it uses "offset" against
an understood (sometimes) knot name "Ring Bend", and works to establish "offset"
as such a handy prefix/modifier -- "Offset Grapevine Bend", "Offset Fig.8 Bend", etc..)
What do they cite?  -- rumors of failure (I know of only one dubious accident report
for the ORB; of a fatal one for the OF8g.8), and testing that shows it can "roll" (flype).
Yes, well, it has inverted (into itself but nearer its ends), but at loads WAY above
what abseilers should generate -- say, 800#, when this knot is used in but ONE of
the two lines supporting the load of maybe 350#.

So, relevant testing for this knot is something that each climber should be able
to perform using a few 'biners for pulley sheaves (if lacking something better)
and body weight:  they will be thus (1) applying much greater force in such
a test than is expected in practice, and (2) in materials of their choosing (which
should be similar to any they'll find when climbing with associates) -- such as
with 7mm haul line tied to 10.5mm dynamic rope.  What can aid them in this
is a well-presented guide on tying the knot, on orienting it so that the thinner
rope is <here>, and so on.  They can even test the wrong way (which
in some cases of back-up (stoppered tail) should still work, albeit with less
impediment to flyping.
But to the point:  the above is very cheap/easy home testing that is also
immediately relevant -- more so than costly break testing on a device.
But, alas, folks are more likely to heed fearmongering on the WWWeb,
which comes in 1- or 2-sentence dribbles and is echoed likewise,
and where a few paragraphs of explanatory guidance is mocked
(too much for the cell-phone display, HELP!).

--dl*
====

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3764
Re: Towards a Science of Knotting
« Reply #3 on: August 01, 2009, 09:09:02 PM »
One need I see is a way to give credibility to some "new knot".
In the case of rockclimbing abseil-ropes-joining knots, the Offset Ring Bend
and some others have a history of usage supporting them.  Now, suppose
one develops/discovers/"invents" a knot to give the advantages of an
offset knot w/o the apparent risk of flyping?  How can you sell this
"new" knot to the community?  What battery of testing will be (should be,
is better thinking) accepted as supportive?  (More likely it would be What
personality should market it?!)

Of course, it is more than just good behavior during tests that will move
one to accept a knot; it needs to be easy to learn/remember/employ.

--dl*
====