Author Topic: Figure-8 vs Figure-8-1/2  (Read 6502 times)

xarax

  • Sr. Member
  • *****
  • Posts: 2781
Re: Figure-8 vs Figure-8-1/2
« Reply #15 on: June 04, 2011, 02:08:54 AM »
other ideas about how best to respond to the Fig-8 jammability and still remain with that style of knot?

   In fact, as regards jammability at least, there is really not 1 symmetrc figure 8 bend, but  5, or 17, or... :) (See (1)). The particular dressing of each variation changes the characteristics of the knot, the phase of loading in which it starts to jam, and its first curve radius (so, presumably, the knot s streangth as well). I am espessialy interested in the "fifth", odd one...(see attached picture) .Try it and tell us, please, your findings.

1) http://igkt.net/sm/index.php?topic=2198.0
« Last Edit: June 04, 2011, 02:12:27 AM by xarax »
This is not a knot.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3764
Re: Figure-8 vs Figure-8-1/2
« Reply #16 on: June 04, 2011, 02:42:55 AM »
Gallinooks, the short answer is that you're taking "one turn"
to mean more than is intended --the intent being approx. 180deg
wrap, not 360.  So, your "half turn" = "turn".  Beyond that,
there are various definitional nomenclature problems coming
from natural language.

As for state of the practice, "fig.8" is best known (though not
well known at all in its other state), and "fig.9" is based on the
simple "adding 1 turn", yes, not here on shape (though that
shape could be forces) ; "fig.10" comes now by addition, too,
and I'll not try to see shaping for this, and is less well known or
frequented.  Beyond "10" there are no takers but for those looking
to running the series out for the fun of it, esp. in exploring the
symmetric forms I've referred to above.

And, back to form, we should keep in mind the differences in
dressing the fig.9 which you show, which might go some
way to helping with jamming or not.

--dl*
====

Bob Thrun

  • Full Member
  • ***
  • Posts: 82
Re: Figure-8 vs Figure-8-1/2
« Reply #17 on: June 04, 2011, 09:17:09 AM »
The earliest depiction of the Figure Nine I can find is in the second French edition (1980) of Techniques de la Speleologie Alpine.  It is called "Noeud en neuf" there.
Alpine Caving Techniques, from which I took the drawing, is the first English edition translated from the third French edition.  I do not have the first French edition (1973).  I would like to get it.

galilnoks

  • Jr. Member
  • **
  • Posts: 15
Re: Figure-8 vs Figure-8-1/2
« Reply #18 on: June 05, 2011, 04:23:14 AM »
Attached is the wikipedia 'figure of nine' entry I forgot in my previous post. I agree, someone with greater knowledge than myself should correct the entry...


===QUOTE Dan_Lehman: you're taking "one turn" to mean more than is intended --the intent being approx. 180deg wrap, not 360. So, your "half turn" = "turn".

No, I'm not. Wikipedia's 'Figure-9' pairs a Fig-10 picture with "...a figure-of-eight loop but with an extra turn".  The author clearly means a 360 by "an extra turn", because that fits an ordinary knotting interpretation and fits the picture.  (The author's mistake is in calling it a Figure-9). If the author meant a 180, as you suggest, the author used the both the wrong description and the wrong picture.

CSCA's 'Figure-9' pairs a Fig-9 picture with "...the Figure of eight but with an additional turn".  The author does not clearly mean a 180 by "an additional turn", because that does not fit an ordinary knotting interpretation (or CSCA's own usage of 'turn').  But the description doesn't fit the picture, so either the author should have pictured a Fig-10, or should have said "an additional half-turn".

With their nearly identical wording, I wouldn't be surprised if CSCA's description is behind the wikipedia entry's (which tries to 'fix' the CSCA's mistake by changing the picture!). 
 
 
In retrospect, I see how the nomenclature digression is mostly of my own making.  I apologize for assuming that ambiguities in the wild have much bearing on the practices in the forum, or that those ambiguities implied that the names were an open question, or that my preferences were obvious or attractive.  Actually, I'm happy to realign myself accordingly.  So, for one thing, here is a revision of my bumbling graphic which I hope reflects at least some of Dan_Lehman's explanation.  For another, let me try to reintroduce my questions more clearly:

The Fig-8 is a well-known, widely-used bend and loop, considered very strong and secure.  But it is also considered prone to jamming.  This criticism is often voiced as a good enough reason to not recommend the knots.  Supporters of the knots offer no better response than to say that the knots are so strong, secure, simple, etc., that, at least when a life is at stake, the jamming problem is tolerable (apparently granting that, when a life is not at stake, it might not be such a good knot, that proneness to jamming is without remedy.)   I like this kind of knot, so I ask:

(1) open question: any ideas about ways to reduce the Fig-8 jammability and still remain with this style or kind of knot (leaving "this style or kind of knot" somewhat vague, but excluding, say, the zepplin as a remedy)?

(2) narrow question:  is the Figure-9 a good remedy?  It is obviously 'this kind of knot' and seems to have all of the good qualities the Fig-8 has.  Is it less prone to jamming with light loads? and at the limit with heavy loads?  My view is that casual comments here are not reliable, and may not even be relevant because of the confusion between the Fig-9 and Fig-10.  My experience with light loads on one brand of 5mm kernmantle is that the Fig-9 is much less jammable than the Fig-8.  Has anyone tested or know of any testing with different ropes, and especially, with heavy loads?

(3) tangential question:  I really like the Fig-9 itself: it is simple, attractive, has a more supportive wrap, is easier to tie without crossing lines, etc. Does anyone else like it?  (I do expect the Fig-9 will turn out to be at least no worse than the Fig-8 with regard to jamming.) I admit that retracing the bend and the loop around an object takes a little practice to find a quick way for yourself, and that enjoying retracing it, as I do, is an acquired taste.
 
I hope this provides a better focus for responding! Thanks for your patience....

knot4u

  • Sr. Member
  • *****
  • Posts: 1076
Re: Figure-8 vs Figure-8-1/2
« Reply #19 on: June 05, 2011, 05:24:45 AM »
Hey buddy, it might be good to edit Post #1 if you're going to change your naming midway through.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3764
Re: Figure-8 vs Figure-8-1/2
« Reply #20 on: June 05, 2011, 05:39:37 AM »
Hey buddy, it might be good to edit Post #1 if you're going to change your naming midway through.

... says he who used to have a .sig whining that he'd not be posting
until this forum was *fixed* so that changes to posts couldn't be made!!
(but who after took the lead on changing posts)   ::)

Rather, an edit to any post to note a change FROM the original text
(which might be quoted and replied to by others) and adoption of
the conventional terminology could help (e.g., "Okay, I'll use the
names 'fig.8-9-10' as seems to be done by the greater community,
and drop the "half" term).

--dl*
====

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3764
Re: Figure-8 vs Figure-8-1/2
« Reply #21 on: June 05, 2011, 06:24:49 AM »
Attached is the wikipedia 'figure of nine' entry I forgot in my previous post.
I agree, someone with greater knowledge than myself should correct the entry...

I wonder what is going on with knots on Wikipedia?  They HAD
been getting a good clean-up & treatment, but this is now another
display of things going quite amiss!?  For that matter, the entire
format of the snapshot image that you've included in your post
looks foreign to any Wikipedia entry I've seen --font, organization,
display of image, and text!?!?

Quote
===QUOTE Dan_Lehman: you're taking "one turn" to mean more than is intended --the intent being approx. 180deg wrap, not 360. So, your "half turn" = "turn".

No, I'm not. Wikipedia's 'Figure-9' pairs a Fig-10 picture with "...a figure-of-eight loop but with an extra turn".  The author clearly means ...

I'll step clear of that author, but note that indeed there is
this ambiguity with the terms "turn" & "round turn" and
we'll get nowhere arguing about them here.  Ashley defines
"turn" as a U-turn, and his references take examples where
the line is ultimately U-turned by however many wraps,
which leaves one to wonder where in fact a pure 360-degree
"turn" would fit!

Let's look at THIS series of knots as ones where an initial *turn*
closes a "loop" through which the end CAN be tucked, and the
series **steps** from its minimal overhand knot by a
turn of the rope one (half-)rotation until the end can again
be tucked out --call it however you may, the *steps* are
from one possibility to the very next one, and the direction
through the initially formed "loop" alternates in so stepping.
Thus, odd-*numbered/named* knots send the end out in
the same direction as the overhand and even-numbered
the opposite.

(For one of the symmetric forms, look at this fig.9++ as though
a ballet dancer with two hand joined (the "loop") around a leg
held high; to get symmetry, lower the joined hands to the crotch
with the leg brought down beside the other end --and then even
out the one part's wrapping ("turn"ing around) the other so that
they evenly *twist* around each other.  You can see from this
that one can keep adding an additional *twist*, alternating the
general direction of legs, building an ever longer twisted section
within the arc of the loop.)

Quote
The Fig-8 is a well-known, widely-used bend and loop, considered very strong and secure.

NB:  "well known" must be understood as "widely (somewhat) known"
in contradistinction from being truly understood!!

I say this with note of some obvious yet seldom remarked facts:

1. the knot (eyeknot or end-2-end joint) is often shown in an
impossible to exist "flat" form, easy to illustrate (much copied),
where the only crossings are of the twin lines with some other
part of twin lines, not of these individual lines with each other;

2. the knots are shown without indication of which end(s) is(are)
to be loaded;

3. test data show a range of strength for some things named "fig.8"
and typically fall to the above criticisms --so we don't really know
what is so strong and what might not be, of possible orientations
(in general, though, I think we can believe that the eyeknots are
stronger than the end-2-end joints);

4. the other symmetric form of the knot is so-far-as-I-can-tell almost
UNknown as a knot component --but it can serve in bowline-like
knots.

5. Xarax above cites some post(s) of his that present various ways
fig.8s can be joined.


Quote
(1) open question: any ideas about ways to reduce the Fig-8 jammability and still remain with this style or kind of knot (leaving "this style or kind of knot" somewhat vague, but excluding, say, the zepplin as a remedy)?

You need to be helpful re what your highlighted phrase means
--some *trace* knot?  --or just something that had a usual fig.8
knot as the base?  --or something *similar* in some other way?
(Well, I just built a knot to order for you, in the last case!)
((Dang, another *new* knot to record --toss it on the heap ... :-\  ))

Quote
(2) narrow question:  is the Figure-9 a good remedy?
It is obviously 'this kind of knot'
interjection : maybe obvious to you, but ...
 and seems to have all of the good qualities the Fig-8 has. ...
My experience with light loads on one brand of 5mm kernmantle is that the Fig-9 is much less jammable than the Fig-8.  Has anyone tested or know of any testing with different ropes, and especially, with heavy loads?

Dave Merchant's e-book Life on a Line bases its presentation
on his testing, and he liked the fig.9 pretty much (and has a
special way to dress the fig.8 not seen elsewhere).

I find the "symmetric fig.9" eyeknot, and what I regard as "the
same sort" of knot, Ashley's #1425, as a good end-2-end joint.
I've posted images for the latter on this forum (in the thread entitled
"#1452 & its ilk" or something similar), but not the latter (I think not).


--dl*
====