Author Topic: Offset joining knots (abseil / rappel)  (Read 44657 times)

Mobius

  • Sr. Member
  • *****
  • Posts: 338
Re: Offset joining knots (abseil / rappel)
« Reply #30 on: October 05, 2015, 11:23:27 AM »
I believe I learn a lot from knotting structures by observing how they load. I trialled the single fisherman's offset (#1414 offset) in 16ply, 3mm, stiff, smooth, poly braid.

I will share what I observed. The first image is the particular  knot. The rope kinks you see in the image are simply an indication of my frugality, the cord has been used before. I believe the braid used is similar in several ways to static climbing rope except it has no 'kern'. From my experience it will collapse more easily and and be harder to untie than Kernmantle. Is that a bad thing? No, since there have been many knots I have trialled that perform admirably in this material, and this gives me some confidence that those same knots will perform well in better materials. I am now trialling in climbing rope to 500kg and so far, from what I observe, the trials I have done in the past (and this one) are probably relevant.

The second image is at 10% MBS and the third is at 20% MBS. The knot collapsed at around 30% MBS and at some stage became jammed (last image).

My take on these results is that any offset bend is going to have trouble not collapsing and not jamming at some stage. Collapsing almost certainly means jamming I think, though in this case I think the single fisherman's might have jammed before it collapsed.

Cheers,

mobius

« Last Edit: October 05, 2015, 02:25:27 PM by mobius »

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4081
Re: Offset joining knots (abseil / rappel)
« Reply #31 on: October 08, 2015, 05:21:14 AM »
Dan,
In performing a series of tests with 'Offset end-to-end joining knots[,]
we need to be careful in how we design the test.

I am of the view that we need to include a flat-parallel dressing
in the test series.  I am using the term 'dressing' here.

I understand and acknowledge your contention that the flat-parallel geometry
is not a dressing per se.  Obviously, such a 'dressing' will not remain in that form
for long --as it will become unstable almost immediately as load is applied.

However, we do need some form of control in the test series
and we also need to compare the behaviour of various dressings.
"Before the load is applied,"
the knot should be (both) dressed AND set
--which requires sufficient load/force to put the parts
of the dressed knot into desired form.  Do you NOT see this?
Because, doing this --setting, i.e.-- with that flat-state knot
is going to result in some variety of who-knows-what,
and it is those set-by-accident forms that get tested,
and that strikes me as going far in the wrong way from the
sort of careful testing you seem to want.

.:. One can stop right there --with the set-by-accident knots--
and take stock of what you've got.  One MUST at least document
THIS state (the in-place-of-missed-setting, set-by-accident state).
It makes no sense to show the d#$*(! flat image  and then some
final result.  Frankly, I think that it's a mistake to risk validating
that mythical form by using it; but I can take a quick demonstration
of its "mythical"ness (it's ambiguity re geometry).

Beyond this, comes my point about the orientation of the knot
however set PRE-offset-loading : that of the "turning a dial"
to give one or the other S.Part a forwards arc, or back loop,
or the two sharing a sideways turn before these states.
Those are settable geometries.


Quote
I have postulated that we need to test at least 3 different dressings of the same knot species.
For example, in the case of the Offset Overhand Bend:
1. Flat-parallel dressing
2. Compact helical dressing
3. A randomised non-uniform dressing (a blind form - knot tied without any attention to uniformity / geometry)

I maintain that we need to test these 3 geometries for each Offset knot structure.
Again, as I've pointed out previously,
there are THREE DRESSINGS per offset knots that have so far
been mentioned here just by "dialing the knot" to give either
S.Part a forward arc / backward loop / equal-to-other side turn.
(to coin some shorthand terms --viz., arc, loop, & turn)

Quote
the use of the term 'Offset' to denote those geometries
where the Standing Parts (SParts) both enter the nub of the knot
from the same direction and are in parallel (rather than in opposition)
They might not be in opposition (for a while --for low forces) :
I developed some offset knots trying to get opposed "loops";
but these, I think, will tend to rotate into "turns" if not opposed
arcs.  Although the S.Parts do need to be adjacent and "choked"
at entry and ... that is sort of "parallel"-ish.   I'll need to go find
my illustrations to see whether to make much of this point.
(I became less enamored of some of my designs upon the
realization that the paradigmatic case is of dissimilar ropes
being joined, and an asymmetric knot likely was better.)


--dl*
====

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4081
Re: Offset joining knots (abseil / rappel)
« Reply #32 on: October 08, 2015, 05:52:39 AM »
Thanks Alpineer...I should have been more careful with my wording - all fixed now.

As promised, here are some photos of the 'Offset anti overhand bend'.

I still don't have complete historical facts about this particular structure.
//
One has to be careful with the dressing of this knot - the position of the tails
(either over-under or under-over)
is important to enhance the structures resistance to instability/capsizing.
???  What you show is a mis-tied offset 8-Oh, which
I've previously discussed and provided a URLink to a good
image of it.  You've misoriented the tails : they should
intergrasp, so to speak --each rising (as you present it)
up through/over the bend in the other,
not turning hard to run parallel on this ending tuck.
(There is nothing "anti-overhand" about this : there is one
overhand and one fig.8 --which (8) is in the
primary/choking position, ergo, "8-Oh".)

Here I'll point out, re ...
Quote
Again, as I've pointed out previously,
there are THREE DRESSINGS per offset knots that have so far
been mentioned here just by "dialing the knot" to give either
S.Part a forward arc / backward loop / equal-to-other side turn.
(to coin some shorthand terms --viz., arc, loop, & turn)
that you have the white S.Part arcing, the other looping.
And they could be otherwise oriented --in roughly three
distinct orientations (arc/loop, turn/turn, loop/arc).


--dl*
====

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4081
Re: Offset joining knots (abseil / rappel)
« Reply #33 on: October 08, 2015, 06:09:21 AM »
My take on these results is that
any offset bend is going to have trouble
not collapsing and not jamming at some stage.
:o
What a strange, unwarranted conclusion!
You test one of the dubious offset knots and
extrapolate from that to the entire genre!!

Please, try the offset 9-Oh --cut to the chase,
get the gold standard done and assured.

By the looks of your test bed, you should be able to
test two or three knots simultaneously --why spend
all that loading/recording effort on a single knot,
when there could be three in line?!

--dl*
====

Mobius

  • Sr. Member
  • *****
  • Posts: 338
Re: Offset joining knots (abseil / rappel)
« Reply #34 on: October 08, 2015, 09:41:33 AM »
My take on these results is that
any offset bend is going to have trouble
not collapsing and not jamming at some stage.
:o
What a strange, unwarranted conclusion!
You test one of the dubious offset knots and
extrapolate from that to the entire genre!!

Please, try the offset 9-Oh --cut to the chase,
get the gold standard done and assured.

By the looks of your test bed, you should be able to
test two or three knots simultaneously --why spend
all that loading/recording effort on a single knot,
when there could be three in line?!

--dl*
====

Thanks Dan,  I thought of editing my original  words a few days ago, however I am quite prepared to 'eat them' after a possible overstatement ☺ I actually hope I do, an offset knot that does not collapse and does not jam will be a very good find I think.

The 'unwarranted' part I will perhaps soften a little bit saying that my penchant for 'Yobo' style bowlines and that I have trialled many of them, many times in an end to end loading fashion has perhaps clouded my judgement. I realize that a knot designed as an offset bend and my poorly loaded bowlines are not the same thing, but that was where my statement was born.

My time for trials, and time for even posting has now evaporated, however I will trial a few more knots in a reasonable time frame and let you know what I find.

Cheers,

mobius.






alpineer

  • Sr. Member
  • *****
  • Posts: 513
Re: Offset joining knots (abseil / rappel)
« Reply #35 on: October 09, 2015, 05:09:22 AM »
Quote
Because, doing this --setting, i.e.-- with that flat-state knot
is going to result in some variety of who-knows-what,
and it is those set-by-accident forms that get tested,
and that strikes me as going far in the wrong way from the
sort of careful testing you seem to want.

Dan,

I maintain that we need to include the flat-parallel geometry in the test series. Obviously, as soon as load is applied, the structure will not remain in that flat-parallel form. It will become unstable.
To ensure that the test is comprehensive, we need to look at this behaviour and document it.
 
I can't see any justification for attempting what you're suggesting, Mark. Testing loosely tied knots and poorly tied knots, yes. From my perspective, testing a properly tied knot in same-natured material as well as odd-natured and transposing positions of the odd-natured strands are what interest me most about Off Axis Knots ;).     

Z

  • Full Member
  • ***
  • Posts: 59
Re: Offset joining knots (abseil / rappel)
« Reply #36 on: October 09, 2015, 12:43:34 PM »
I didn't read the whole thread yet.

Until I do, what is the preferred offset knot setup?

Also, what is the preferred stopper/backup knot to use?
If you're reading this, it's too late.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4081
Re: Offset joining knots (abseil / rappel)
« Reply #37 on: October 10, 2015, 04:35:07 PM »
Quote
I maintain that we need to include the flat-parallel geometry in the test series.
Obviously, as soon as load is applied, the structure will not remain in that flat-parallel form.
It will become unstable.
To ensure that the test is comprehensive,
we need to look at this behaviour and document it.
Quote
I can't see any justification for attempting what you're suggesting, Mark.
I can.

Its all about drawing comparisons between different geometries and how they behave under load.

To put your point a different way...
Can you see a justification for not performing a test series that covers the 3 geometries I listed above?

Certainly, little harm would come (other than loss of the testers personal time)
--and importantly, there is scope to increase the depth and breadth of the data that is captured.
I gave full reasons for NOT testing the flat geometry,
which I don't feel need to repeat in full here, but just
to point back to --a fews posts above/prior.  It's like
testing two dye at, say, equal numbers 4 & 4, and
2 & 6, and then this 3rd-for-some-notion-of-comprehensive
X & Y where the dice are dropped from and take who-knows-what
combination, but you "test" it and report an outcome.
--sketchy analog but the point is one of knowing the
structure vs. knowing little about it.

Moreover, it's really NOT a structure that is likely to occur
in practice, IMO :: books might give this illusory "flat" image
for instruction, but in hand it is NOT something that one
is likely to make, especially in tying the offset overhand
& fig.8
knots.
(Now, one thought re the latter is that the formation of an
8 for me entails turning the tails around the S.Parts held
in left hand and ... if I know I'm going for an 8 (vs Oh or
9 --i.e., a question of direction of tucking out tails),
I'll anticipate this by turning the tails behind the S.Parts
vs. over them, forming a sort of conical loop ready to
have the tails exit up & out vs. down & through --and that
is a possible orientation difference one could check.)

ON TOP OF WHICH COME THE TRIO OF ORIENTATIONS
I MUST KEEP REITERATING, since they seem to have
not struck home in some minds!!  --i.p., in one who keeps
wanting "comprehensive" testing and "three" orientations!
voici: www.postimage.org/image.php?v=PqAodEJ
 [warning : postimage site seems to have gone sour ::
  it gives me, below intended knots image, some clearly
  risque soft-porn images to follow,
  and today --first time-- it puts a pop-up saying that one
  cannot view as guest but must create an account !?
  Well, I AM seeing the intended knots --and will not create
  an account there!]
IN THIS IMAGE, the knots are the "same" except that
on the left knot, the orange 8mm knot (oriented to make
the "choke" of S.Parts) goes into the knot as a LOOP,
the yellow as an ARC; but on the right, it's vice versa.
DO YOU SEE THIS?!  (Between these extremes comes
the "TURN" these two S.Parts could make, before further
turning through the knot --and in this orientation their
tails would be roughly aligned/parallel w/axis of tension;
they are otherwise more nearly perpendicular to it.)

The whole bit about "THE helical way" and so on MISSES
that the definite article "the" misleads re orientation : there
are a range of them, marked by my presetation of this trio
of extremes & mid-range orientations.  And that's THREE
for what has so far been slid under the radar as "one" knot,
one orientation.   It's a blindness to exact geometry akin
to starting with the "flat" state and just looking at results.
(A survey of actual usage could indicate tendencies/frequencies;
but a test of the possible states covers all.)

Quote
It seems that there aren't all that many people willing
to step up to the plate and conduct testing?
Without data, there will only be ongoing endless debates
about offset joining knots, with no real definitive answers.
Alpineer and some others should be able to do some
testing of the sort without the fancy calibrated output
of load forces (that might give misleading information),
but which entail human-body forces (multiplied by the
pulley effect of a crude 2:1 MA load arrangement, and
further by loading a single strand (in contrast to the
typical abseil which is on two strands),
and reporting observed --even photographed-- deformations
of tested knots, perhaps even a capsizing!?

And I am thinking of maybe doing some hundreds of
click-ResetNo.-click-ResetNo.... repetitions so as to bump
the folder number and hence filename to "P300..." for
a newly got camera (where I have already two doing
"P200..." to extant 3rd camera's "P100...".  And, if so,
I might as well actually do something useful in part of
such an exercise (though I really want it to be quickly
done, dang --as done to get that P200...).

--dl*
====

« Last Edit: November 20, 2015, 06:00:43 PM by Dan_Lehman »

Mobius

  • Sr. Member
  • *****
  • Posts: 338
Re: Offset joining knots (abseil / rappel)
« Reply #38 on: October 12, 2015, 01:35:59 PM »
Here are Dan's 9-oh and 8-oh trialled in my knot material (see earlier post for more detail). The knots were tried in tandem.

I have several more images I can share at 10%, 20% ... etc. MBS, the ones shown are at 40% MBS.

Both knots held until 50% MBS+ and the knot shown image 4 survived a line breakage and was easy to untie. I didn't intend breaking either knot, however I learnt an important lesson in home safety after pushing the wrong button on my rig while having my hand in the wrong place... that really stung  :P The image 3 knot probably broke around 65% MBS.

I can write more about my impression of how both knots loaded, other images and the overall result if there is interest.

Cheers,

mobius
« Last Edit: October 12, 2015, 10:24:34 PM by mobius »

Mobius

  • Sr. Member
  • *****
  • Posts: 338
Re: Offset joining knots (abseil / rappel)
« Reply #39 on: October 13, 2015, 11:13:59 AM »
Here are 3 more images of the trial from the last post.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4081
Re: Offset joining knots (abseil / rappel)
« Reply #40 on: October 14, 2015, 05:06:20 PM »
Having lost my website
Very sorry about that, and I hope that NotMe's URLink
to an archived version is at least some help, and maybe
others have yet some of the material to help restore,
if better redress isn't available.
  >:( :(


Quote
(using my new Edelrid 'Corbie' dynamic rope).
A difference between new & old can be that the new enables
the S.Parts to slide over the choking turn(s) whereas the
increased friction of old rope can see the draw of the S.Parts
sooner pull at these turns, IMO --and this aggravates the
risk of flyping.  (But might friction elsewhere work to inhibit
movement of parts around parts, frustrating flyping?!)

Quote
The difference lies in the position of the choking turn
...
 in one structure the choke is performed lower where the SParts enter the nub/core
and in the other structure, the choke is performed last - just before both tails exit
IMO, this is a misuse of "choking" : choking is done at the
point where S.Parts enter the knot and are forced into
the offset nub --if they're not (well) choked, they pull apart
to more obvious separate entry points.  So, the latter turn
of your preferred knot is IMO a securing of it --what could
be done, with more material ("footprint"), by tying off this
tail around the other with an overhand.  And doing either
securing should work to strength the choking done by its
not-quite-full turn around the S.Parts, by inhibiting the pulling
out (slippage) of this tail.

Quote
I had previously theorized that the lower choking turn might
boost resistance to instability (testing is needed to prove or disprove).
I think that the "full turn" gives better choking, but the later securing
of the tail should strengthen the sub-full turn --two methods to preserve
the choke of S.Parts at the entry.  But I concur in the theory that
the full/round turn makes a better choke; in hand, now, it
seems that doing the latter, tail-securing leads to more
movement of the other rope to roll around this latter turn,
and that having the turns in the choke enable one
to more tightly "choke" the S.Parts and impede their prying
open the knot.  And maybe this strangle vs. fig.9 form
enables better tightening of this turn?!

In the case of exact orientation, you show the offset 9-Oh (with
slightly different dressing than I show) and then what could
be called "the offset Strangle-Oh" --a matter of whether the
tail makes the simple wrap & tuck out (former) or takes the
trickier-to-make turn *above* itself and out (which forms a
double overhand / strangle knot).

What differs in the two YOU present is that the "upper" case
could be lower --what I point to immediately preceding prg--
but is taken "higher/upper" and so becomes thus more a securing
of the tail than a doubling of the choke.  I.e., with the full
turn at the choke point, one has a dbl.oh. in the anchor bend
form, rather than strangle form.


--dl*
====

James Petersen

  • Exp. Member
  • ****
  • Posts: 217
Re: Offset joining knots (abseil / rappel)
« Reply #41 on: October 21, 2015, 02:31:48 PM »
Agent Smith:
There was a thread some time ago about this topic. At that time, I suggested what I have been calling the "life knot". It is basically a bowstring knot joining the ends of two ropes. There were pictures in the other post, but I am posting pics with this post, too (probably better looking). I don't climb but have done some very informal testing on this knot and it seems to perform very well for this purpose.

I would very much like to hear what others who test this knot out think.

JP

SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 1976
Re: Offset joining knots (abseil / rappel)
« Reply #42 on: October 21, 2015, 03:54:46 PM »
Hello James.

This looks to be the Ashley stopper knot as a bend.

I like it and did when I tried it some time back, but I have felt (without serious heavy load/cyclic testing) that the working ends could pull through the overhand portion. Slipping it does add more bulk to counter this possibility, although it makes the knot pretty huge.
Also, I did not find it to be all that easy to untie in the media I tried it with.

SS

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4081
Re: Offset joining knots (abseil / rappel)
« Reply #43 on: October 21, 2015, 05:41:18 PM »
This doesn't look like Ashley's stopper knot to me
(with or without corrective lenses   8) !
 ;)

This is one step wrong for an offset double overhand, the final
tuck being on the other side of itself for that.  And that position
if anything works to weaken the "choke" of the S.Parts, alas
(as it slightly limits the turn around the S.Parts).


--dl*
====

SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 1976
Re: Offset joining knots (abseil / rappel)
« Reply #44 on: October 21, 2015, 06:12:55 PM »
Nope, it does not look like an Ashley stopper knot.    Even with corrective lenses. ;-)
I incorrectly tied it in a hurry.
Sorry James.

SS

 

anything