Offset joining knots (abseil / rappel)

[agent_smith]

I am conducting research into knots for joining 2 abseil ropes together (eg for retrievable abseil system) - with the eventual aim of authoring a paper.

The Offset version of #1414 (Single fishermans bend)  I think has been presented before - in knotting matters perhaps?

I have also posted 2 other Offset structures which I believe are worthy of further examination.

#1410 (Offset overhand bend) is proposed as the control. I have included a photo of #1410 for comparison.

Offset version of #1414 has a smallest footprint...in fact, it would seem that it is the most compact (in terms of volume) method of joining 2 ropes together. Mobius has conducted some testing of this structure (refer to his post later in this thread).
Skeptics of this comment (re smallest footprint) should first examine this for themselves...that is, tie Offset #1414 and do a side-by-side comparison. Use vernier calipers to confirm if necessary.

I have not performed any rigorous testing of this structure (ie Offset #1414)...however it does appear to become unstable at quite low load (lower than 2.0kN). Do not trust it in mission critical applications!

I have largely formulated the data points that I wish to obtain as follows:
1. Tail movement/change as load increases (which should be characterised as 'compression induced tail draw-in' )
2. Load threshold where instability is triggered
3. Resistance to jamming (eg what is the load threshold where jamming occurs? )

With regard to #2 (instability) - I have theorised that it will be possible to establish a probabilistic threshold load beyond which, instability is likely. Note the use of the term 'likely'. This will not be a precise load value - rather, it will be expressed in terms of probability (sort of like calculus where the 'limit as X approaches Y' is a good analogy). We already know that #1410 (Offset overhand bend) is stable under nominal loads of 1 kN. What is not known is the probabilistic threshold load where instability is triggered. This probabilistic threshold will vary according to the particular Offset geometry tied, with other factors also playing a role (eg rope diameters and rope design type eg 'static' versus 'dynamic'). All of this needs to be examined.

NOTE: The tails in the photos are deliberately tied short - this was necessary so I could fit the knot structure within the camera frame. If I make the tails too long - I cant get close up 'macro' shots. I am trying to show detail of the core structure. Normally, the tails would be at least 200 mm in length (and maybe even more). With the testing that I have done over the years - there is no evidence that tail slippage is significant. I am not an advocate of long tails - eg some user groups advocate 1.0m (1000 mm) length tails! To me, this is unnecessary and counter-productive (since it increases risk of entanglement/entrapment).

...

EDIT NOTE: I have removed all of my additional posts in this topic/thread because I believe them to be redundant. They dealt largely with the design of test parameters for examining Offset joining knots. Mobius had already commenced his testing - which is to be applauded since there appear to be few who are moving beyond pure theory and progressing to actual testing.

The endless discussion of the design of a test has led nowhere and in my view has not been productive. It seems that its better to avoid engaging in debate and to simply proceed with practical testing and then publish results.

Further edit note re the images shown in this post:
1. Both rope diameters are virtually the same - the difference being less than 1.0 mm (ie the difference being insignificant)
2. I have not photographed Offset joining knots showing different rope diameters at this stage of the project (on my 'to do' list). All of my photos depict same diameter ropes joined.
3. Dan Lehman has repeatedly pointed out that the dressing of Offset joining knots with differing rope diameters will play an important role - and I agree. That is, the position of the smaller diameter rope relative to the larger diameter rope will have a significant effect on stability.
4. It is theorised that the smaller diameter rope should be oriented/positioned so that it is 'underneath' the larger diameter rope. That is, with reference to the last 3 images - the smaller diameter rope would be the blue rope.
5. With reference to the first image (Offset #1414) - the smaller diameter rope would be the white rope. The smaller diameter rope making the initial 'choking' turn. Again, it is theorised that the larger diameter rope ought to act as a barrier, with the _smaller diameter rope being less likely to roll around the larger diameter rope. I will photograph this concept so it is clear to readers...
6. Testing would of course be carried out to reverse the position/orientation of the smaller diameter rope to examine the effect this has on stability (ie resistance to structural capsizing). And this testing would be to probe what the probabilistic threshold load might be for each orientation of the smaller Vs larger diameter rope.

EDIT NOTE: Do not use offset #1414 for mission critical life support applications (eg to abseil with it). The instability threshold is very low...less than 2kN. This structure is shown purely for academic interest - please dont use it for mission critical applications!
Mark Gommers

[Dan_Lehman]

The Offset version of #1414  I think has been presented before - in knotting matters perhaps?

While Ashley #s might be credited with precision,
they are so by virtue of making a reference which
can be tedious or impossible to the reader; ergo,
please accompany your precision with helpful names
--esp. here, the commonly known "(single) fisherman's knot".

Can someone please confirm details (eg date presented and by whom).

Josh GUDELIUS presented what he (not-so-helpfully) named
the "triple fisherman's knot" (and which I see has now be re-named
something cumbersome by more puzzling than misleading), in which
the offset fisherman's knot is given an overhand stopper on the
tail of the *inner*/first-nipping overhand component of the base knot.
Cf. www.gudelius.de/spst.htm
and I think further searching by his name (and maybe by the name(s)
of the knot) will find a thread here, or ... (somewhere, I recall).
--with a "usual suspect" making comments.

I have included a photo of #1410 for comparison.
Offset version of #1414 has a smaller footprint than #1410.

Wow, one has to scrutinize the images to see a difference.
I beg to question the notion of ANY "smaller footprint",
and this especially challenges the significance of what
might be shown!  In any case, I'd not present either of
these knots as recommended : the infamous "EDK"
(i.e., the offset water knot / #1410) is usually presented
with the admonition to trust it but with "long tails";
and as you can see from Josh's presentation, he does
ad (essentially --one can divide the trio of overhands
into 2+1 or 1+2 (knot + a stopper, or "guard" + knot).

I've presented thoughts about offset abseil-rope joints in
various places.  Please NB: what you show for these knots
--for both, I think-- ARE BUT ONE PARTICULAR OUT OF MANY
ORIENTATIONS FOR THEM.  Cf. this presentation of extreme
differences --i.e., where on the left the orange cord makes
a U-turn/loop up into the nub; whereas on the right side
it makes a sort of upwards *arc* and it's the Goldline rope
that makes the loop.  SAME KNOT, just differently set in
hand and then tightened (and easily enough RE-set to the
other, or to a between-extremes, orientation.
  [edit to add (egadz!) the aforementioned presentation]
   www.postimage.org/image.php?v=PqAodEJ
THIS JUST IS IGNORED THROUGHOUT DISCUSSION OF THIS
COMMON KNOT --nobody is l00king intelligently at this !!!
(Though common, veteran climber-poster "RGold" has noted
the recommendation to make the smaller (orange, above) rope
have the "inner"/first nipping, the larger to thus be put into
stopper-position abutting this smaller cord.  (It would be
easier for large rope to roll around small cord than vice versa,
is the point!)


You and I have had our differences over your favored offset
double water knot (strangle?) knot, i.p. vs. my favored putting
in a full wrap/round-turn for the "EDK"/"OWK" and getting
what I name "offset 9-oh" (giving the foundation component
--what does initial nipping/choking of S.Parts-- prominence in
the name (here, "9").

I suppose it's worth some mental effort to see if one can
learn pros/cons of having the two components (in cases
where there are such --such above) arranged to be
more together (water knot) or separate (fisherman's knot) !?
(It's certainly easier to take the two tails and in one fell
swoop make an overhand ; it is less simple to do this
for the offset 9-Oh --one needs to put in the extra
turn for the 9.  But more important is the result.

(The EDK backed by an EDK is another simple & sure way
--to not must have "long tails' but to DO SOMETHING with'
those tails : tie the 2nd, stopper-EDK!)

a thought : the stopper knot --where used-- might enable
the base knot to be untied more easily, arresting the
deformation of the base knot!?

REPEAT AFTER ME ::
THIS JUST IS IGNORED THROUGHOUT DISCUSSION OF THIS
COMMON KNOT --nobody is l00king intelligently at this !!!

When you tie some offset knot,
as it is lightly tensioned (tie it off somehow),
grasp the nub and just turn it like doorknob,
this way and the other,
see how orientation changes.  What are the effects?!
(Ideally, we find a knot in which any orientation is fine,
as one doesn't want to prescribe something that comes
with a list of details to tend to!  And especially ARJ knots
can be called upon to be tied by tired, fatigued persons
in "exposed" conditions trying to beat darkness of,
missing that, work in darkness.)


--dl*
====

[Z]

Since we're talking about abseiling / rappelling, I vote that every description of a knot must be accompanied by a pic.

[agent_smith]

From Dan Lehman.....

Wow, one has to scrutinize the images to see a difference.
I beg to question the notion of ANY "smaller footprint",
and this especially challenges the significance of what
might be shown!

Nevertheless, the relative footprint (size) of the Offset #1414 (aka Offset single fishermans knot/bend) is smaller than #1410.
The core/nub of Offset #1414 is some 10-15% smaller in volume than #1410.
Admittedly, a 10-15% reduction in footprint isn't a major breakthrough - it is just an incremental improvement (in terms of achieving the smallest possible footprint).

I would recommend that you test this for yourself.. tie both knots and do a side-by-side comparison. You wont need verniers/calipers...it is noticeable by eye.

The EDK backed by an EDK is another simple & sure way
--to not must have "long tails' but to DO SOMETHING with'
those tails : tie the 2nd, stopper-EDK!

This is something I do not agree with.
The perceived need for [long] tails are a myth (one for the myth busters surely).
We should be aiming to achieve the smallest possible 'footprint' - leaving long tails and/or tying additional 'stopper knots' (to my mind) is increasing the risk of fouling or getting stuck somewhere.
Tail slippage in a firmly set and dressed #1410 (Offset Overhand bend / water knot) is not significant up to 4kN. As is typical in the climbing community in general, rumors spread fast and many authors simply parrot from each other. Even the moniker 'EDK' to me is absurd (I argue that the real 'EDK" is in fact an Offset F8 bend). Refer to the Karen Turk accident where the actual joining knot is unknown - I follow the evidence - #1410 is stable and secure up to about 4kN whereas certain dressings of the 'Offset F8' gives wide varying results. I think the Karen Turk accident was a prime catalyst for the 'EDK' moniker (refer link below):
http://publications.americanalpineclub.org/articles/13199808000/Fall-on-Rock-Rappel-Ropes-Knot-Unraveled-Wyoming-Grand-Teton-Guides-Wall

I have already long established my own 'locked' variant of #1410 - and it has been in use for several years. I had published my discovery previously on this forum. Its 'footprint' is only marginally larger than #1410 and it does not require 'long' tails. 200 mm length tails are sufficient.

Ideally, we find a knot in which any orientation is fine,
as one doesn't want to prescribe something that comes
with a list of details to tend to!  And especially ARJ knots
can be called upon to be tied by tired, fatigued persons
in "exposed" conditions trying to beat darkness of,
missing that, work in darkness.

Agreed.
We have a saying in Australia... KISS
And I'm sure the KISS principle applies elsewhere too

The Offset #1414 can be miss-tied - with 2 strangles done by the same cord - which would lead to a 'Darwin award'.
#1410 is simple and it works. However, there are different possible dressings - and some would behave differently under load.

In the climbing community, the term 'offset' has not been widely accepted - and is not understood. You often hear terms like 'Flat' or 'One-sided' or 'Same-sided'. I personally think these are terms created by laypeople who are just looking for something descriptive.

I think of the term 'flat' overhand differently to your Joe Average climber (see photo for illustrations of how I differentiate 'flat' from 'compact helical dressings')

A proper definition of 'Offset' is: Where the core of the knot is displaced from the axis of tension and both SParts converge and follow a parallel pathway.

NOTE: The tails in the photos are deliberately tied short - this was necessary so I could fit the knot structure within the camera frame. If I make the tails too long - I cant get close up 'macro' shots. I am trying to show detail of the core structure. Normally, the tails would be at least 200 mm in length. With the testing that I have done over the years - there is no evidence that tail slippage is significant. I am not an advocate of long tails - eg some user groups advocate 1.0m (1000 mm) length tails! To me, this is ludicrous.

please accompany your precision with helpful names

I might counter this by adding...please accompany your posts with clear photos to illustrate your propositions
And since this is the IGKT, I see no harm in simply quoting ABoK numbers amongst my peers  If I was posting in a different forum - I would certainly be more descriptive and careful with my terminology and use of acronyms...

EDIT NOTE: Do not trust offset #1414 with your life. it has an instability threshold less than 2.0kN.

Mark G

[Dan_Lehman]

From Dan Lehman.....

Wow, one has to scrutinize the images to see a difference.
I beg to question the notion of ANY "smaller footprint",
and this especially challenges the significance of what
might be shown!

Nevertheless, the relative footprint (size) of the Offset #1414
(aka Offset single fishermans knot/bend) is smaller than #1410.
The core/nub of Offset #1414 is some 10-15% smaller in volume than #1410.
Admittedly, a 10-15% reduction in footprint isn't a major breakthrough
--it is just an incremental improvement
(in terms of achieving the smallest possible footprint).

I would recommend that you test this for yourself..
tie both knots and do a side-by-side comparison.
You won't need verniers/calipers...it is noticeable by eye.

YOU can notice "10..15%" difference by eye?!?!
--i.e., that it is this much difference?!
--or did you just pull those figures out of the air?

But I'll humor you : I tied the knots in 5mm-ish cheap
kernmatle ("Home Depot rope' is the derisory term here
in The States) and set the knots jointly by lightly bouncing
a 15# weight on them after gentle dressing.  And then I
tied tea-bag strings (yes, I saved some for small whipping
and who-knows-what-else) around each knot's 4 ends
and measured.  1.08 / 92.5 were the results of dividing
3.125 / 3.375 as roughly measured (for each strand).
--in your favor, yes.
But I call this INsignificant, especially given tying methods
--the offset water knot is basic, one move, one overhand
with the two strands, and not one each in succession.
.:.  I stand my ground here.

The EDK backed by an EDK is another simple & sure way
--to not must have "long tails' but to DO SOMETHING with'
those tails : tie the 2nd, stopper-EDK!

This is something I do not agree with.
The perceived need for [long] tails [is] a myth (one for the myth busters surely).

Some tests have seen knots "roll"; users have reported
the tails being "not as long as set" at the end of rappelling.
Formal tests typically use (a) new & (b) usually ONE
combination of (c) same-size (maybe same, period!) ropes;
reaLife(tm) applications see various combinations of ropes
--in age, stiffness, compressibility, flexibility, & diameter.
And that goes without the variances I've emphasized in
my post above re orientation of the knot.

We should be aiming to achieve the smallest possible 'footprint'
--leaving long tails and/or tying additional 'stopper knots' (to my mind)
is increasing the risk of fouling or getting stuck somewhere.

One might seek a case history inventory of such supposed
problems (or join in your myth-land with "long tails").  One
doesn't hear all so much complaint of the much larger joints
--"square fisherman's", "grapevine", and so on.  The primary
goal is ease of use and reliable effectiveness.

Tail slippage in a firmly set and dressed #1410
(Offset [water knot]) is not significant up to 4kN.

Spoken from the Ivory Tower of unsoiled-by-use ropes,
and not even abseiling from that tower to ground.  It
is nearly like saying "slippage from an unslipping knot
is nil".

... rumors spread fast and ...

So, too, from elsewhere.  As I said before, I've NEVER
seen anybody not the orientation aspects of these offset
knots --which includes your favored one(s).  --let alone
do testing of it.

Even the moniker 'EDK' to me is absurd
(I argue that the real 'EDK" is in fact an Offset F8 bend).

You're spitting into the wind, here.  The moniker is
what it is, though, yes, one can argue that it would
better apply elsewhere : the fact is, it was applied
where it was, upon visual encounter & doubt, not
on actual failings.

Refer to the Karen Turk accident where the actual joining knot is unknown
 --I follow the evidence-- #1410 is stable and secure up to about 4kN
whereas certain dressings of the 'Offset F8' gives wide varying results

This is a mystery, with the only evidence that given by those
who might not know what they were doing or who have some
reason to not be honest (liability).  As I re-read this case, and
with this offset fisherman's knot in mind, I wonder if the guy's
partial looseing-with-intent-to-untie might have freed one tail
and then in changing course to "re-secure" the original knot,
he went wrong and got . . . the very knot you advocate here! !?
--or something went wrong and the original wasn't restored
(though he made a successful abseil on it).
(Citing that "up to 4kn" just smacks of Ivory Tower vision.)

I have already long established my own 'locked' variant of #1410
--and it has been in use for several years.

"locked fisherman's"?!  Aren't we talking about a strangle
formed with doubled ropes, with offset loading?  --which is what
I recall.

The Offset #1414 can be miss-tied --with 2 strangles
done by the same cord-- which would lead to a 'Darwin award'.

That's going out of your way to find a supposed flaw.
Any knot you propose can be mistied ... : that doesn't
help the deliberation any.  The components of a pull-together
knot must . . . pull together, and if tied in the same rope
they quite obviously won't.

#1410 is simple and it works. However, there are different
possible dressings - and some would behave differently under load.

One might want to test this assertion re behavior.  One can SEE
the differences (if looking, or heeding my emphatic notes!), but
it remains to be shown what effect they have on the knot.

In the climbing community, the term 'offset' has not been widely accepted

Thus it needs usage, reiteration, explanation (pro it, con the
competitors) --go for it.  Clyde Soles set it in motion.  (And I
concur w/you re "flat".  But your diagram needs arrows pointing
to TWO different images from the top one --e.g., of two possible
ways the top knot so formed might transform with force!

And since this is the IGKT, I see no harm in simply quoting ABoK numbers amongst my peers  If I was posting in a different forum - I would certainly be more descriptive and careful with my terminology and use of acronyms...

But other than just your peers might read this.
And this reader isn't always sure of the #s or near
ABOK to confirm --and we have pretty commonly
known names at issue here.
And gimme some credit for a URLink to images, this time.

--dl*
====

[agent_smith]

Dan, not sure where you are going with all this... essentially, I have exhausted myself in debating this subject matter endlessly with many others - and i have come to the conclusion that there may never be consensus.

I have reported my test results in this forum before - and I have applied the theory into actual practice.

Some things I will comment on:
1. Tail slippage in offset joining knots based on #1410 is not significant (provided the knot structure is dressed and cinched up tightly to begin with). Try it yourself, obtain some climbing rope of at least 8.5mm diameter and join it. Cinch the knot up and measure the tails pre-test load. Apply load and measure the tail lengths at various load milestones. You will see that the tails do draw in (and that is obvious and expected) - but, they dont draw in to the point of catastrophic failure. And this holds true in dissimilar diameter cords too. Again, please test this for yourself - why not post your test results here - we would welcome it

2. The dressing/geometry of any abseil end-to-end joining knot will play a significant role in its stability and security. For example, certain dressings of the Offset F8 will give alarming results (try this for yourself). In general, the 'flat parallel' dressings behave unpredictably under load. I have found that a flat-parallel dressing of the Offset F8 can lead to catastrophic failure earlier compared to a compact helical dressing. Again, try this yourself - and let me know your results.

3. The Offset #1414 does have a smaller footprint than #1410. I use cord diameters of 8.0mm and above. Effects/outcome scale according to diameter. Your use of 'thinner' cord wont appear to your eye as anything significant. Scale this up to larger diameter cord and you will see the effects more readily.

4. They key objective with any abseil rope joining knot is to obtain a shape / geometry that will translate easily around a 90 edge - this is self-evident. However, I might add that there are also other qualities worthy of pursuing...such as; ease of tying, ease of untying, and - to achieve a relatively small 'footprint'. Logically, the larger/bulkier the knot, the more likely it is to jam or get stuck somewhere (this is my personal observation - empirically derived by experience and observation in the field). If we can achieve all of the primary objectives - why not also try to achieve a smaller footprint? Do you believe this to be a bad quality to pursue?

...

I've uploaded a locked version of #1410. Tails are shown short (so as to fit within the macro view of the camera lens). It works - and I use it regularly with same and also dissimilar rope diameters. All abseil rope joining knots must be dressed and cinched up tightly before placing reliance on them. Perform your own tests to verify.

[Dan_Lehman]

Some things I will comment on:
1. Tail slippage in offset joining knots based on #1410 is not significant
(provided the knot structure is dressed and cinched up tightly to begin with).

You're missing my point, which was not to suggest
"slippage" but capsizing (with perhaps some bit of
slippage of the 1st-choking overhand component
enabling that).  In a universe of various combinations
or abseil ropes and possibly haul lines, I'm reluctant
to emphasize material efficiency over security.

And, again, "dressed and cinched up tightly" is a criterion
of variability, and a requirement that could be compromised
in some ARJ-knot circumstances such as I indicated;
and so one might prefer awkwardness & bulk if it
removes this (=> the EDK-backed EDK).
((And note, it's not whether the OWK deserves
the "EDK" moniker I'm arguing, just whether in it
in fact got it --and I believe that it did.))

3. ... Your use of 'thinner' cord won't appear to your eye as anything significant.
Scale this up to larger diameter cord and you will see the effects more readily.

Did you miss the report that I measured the difference
in rope consumption?  --nothing to do with eyeing it; ... measured.
And the difference IMO is insignificant (like making cameras
ever smaller : they've gotten TOO small, now (which Pany
in its GH3 & GX8 enlargements realized)).

Perform your own tests to verify.

THIS is the admonition to be emphasized with the
note that expected loading of such knots is well
within users' capability to apply --just make a crude
2:1 pulley and load the knot in a single strand,
and you have approached 3 times expected force
(theoretically "4" sans pulley friction, etc.; but say
"3" to be conservatively happy).  --and in the very
ropes (well, often) of issue, in the users' hands.

That said, one should heed others' observations
(such as the various turn-the-dial/-nub orientations),
which might show things that didn't come to mind
in one's own thinking --the orientation aspect was
something I realized well later than my initial
"knowing" of the knot.

--dl*
====

[Mobius]

4. They key objective with any abseil rope joining knot is to obtain a shape / geometry that will translate easily around a 90 edge - this is self-evident. However, I might add that there are also other qualities worthy of pursuing...such as; ease of tying, ease of untying, and - to achieve a relatively small 'footprint'. Logically, the larger/bulkier the knot, the more likely it is to jam or get stuck somewhere (this is my personal observation - empirically derived by experience and observation in the field). If we can achieve all of the primary objectives - why not also try to achieve a smaller footprint? Do you believe this to be a bad quality to pursue?

As a layperson to this topic, I will just go with this as a starting point.

The bulbous knots shown above do not excite me very much. They may be offset and not be likely to snag on their flat side, but what a monster the other side is, especially the Figure 8 ones. As an alternative the Zeppelin bend immediately comes to mind (all images at 500 kg in climbing quality rope), however the lip from one collar or the other is the potential problem I suppose. Otherwise, the knot is low profile and non-jamming.

What about what I call the Infinity Bend? Dan and Mark both know this knot http://tinyurl.com/infinitybend. It is low profile, does not jam and has a wonderfully smooth side which I do not see as snagging on anything. True, the second Infinity image is not what I think of as being 'offset', however the smooth knot mound it has is not going to have trouble with a 90^o bend either in my opinion.

Cheers,

mobius

[agent_smith]

This topic is nearly exhausted.

I have checked back through many of your posts Dan and the one common theme I find as that there are no photos to find. I typically try to include photos to illustrate concepts and/or various knot structures.

Why don't you do the same? Its not rocket science

Most of what you are saying i agree with. However, some of what you say is stating the obvious.

For example:

Please NB: what you show for these knots
--for both, I think-- ARE BUT ONE PARTICULAR OUT OF MANY
ORIENTATIONS FOR THEM.

This is obvious. I am not going to take scores of photos from different angles and then try to post them here and make them fit within a 100KB size limit. What would be the point?
I have chosen some of the obvious dressings and posted them. I have no intention of tying endless variations and photographing and them posting them here - this might be a job for you Dan?

The compact helical dressing I have shown is (more likely than not) the one most people would attempt. I showed the flat parallel dressing merely to illustrate my point that the use of the term 'flat overhand' isn't really a correct descriptor. Again, I could have photographed a different 'flat' dressing - but why bother. Its just to illustrate a concept.

And ref your comments re the size difference between the Offset #1414 and $1410... I did make the point that the size difference is not Earth shattering or deserving of a Nobel prize. I simply made the comment that there is a size difference - albeit only 10% or so (in other words, not significant by itself). However, I do make the point that achieving a small footprint is a worthy criteria.
For example, I could tie a gigantic knot that measures some several square inches in volume. Or, I could tie a knot that is only about a square inch or so in volume. Which would you say is more desirable - the large joining knot or the not so large joining knot? That is all I am really stating - its quite simple in concept - in that the smaller footprint is a desirable attribute.
I'll make my point again for clarity: "A desirable attribute is to achieve a joining knot that has a relatively small footprint".

Another desirable - and I might add very important attribute - is to achieve a knot geometry that will easily translate around a 90 degree edge. Unfortunately, the so-called Zeppelin bend does not easily translate around a 90 degree edge compared to the Offset geometries (eg #1410). In fact, in my experience - only the 'offset' geometries will translate relatively easy over a 90 degree edge.

And it goes without saying that whatever abseil rope joining knot you choose - it MUST be stable and secure (at least up to about 4kN or so). Keeping in mind that most retrievable abseil systems will be on a 'double strand' - ie, the load at the joining knot will only be 50% (friction around the anchorage point will alter that 50/50 balance somewhat).

And your point:

You're missing my point, which was not to suggest
"slippage" but capsizing (with perhaps some bit of
slippage of the 1st-choking overhand component
enabling that).

I don't think I am missing any point because I generally agree with you.
However, the capsizing you refer to isn't really an issue up to a certain loading.
Given that the joining knot may only be subjected to 50% load (unless the person descends on a single rope device such as a GriGri - in which case the load might be closer to 100% depending on how the retrievable system is configured) - capsizing is not going to occur on #1410 at the nominal load of a single person. Its all a question of reaching the load threshold at which point capsizing will occur. Little study/testing has been done to pin down these threshold values at various dressings/geometries.
There is a threshold loading where various Offset joining knots will capsize. For example, the Offset F8 will capsize before #1410 in a side-by-side test. And, the flat parallel dressings will tend to be unpredictable in their response to load (test this for yourself please - before making comment). Compact helical dressings in general, will behave more predictably and resist capsizing up to a higher load threshold (they will eventually capsize - but not as early as with flat parallel dressings).

The attached photos are simply to illustrate knot geometry - they are both the same knot (#1410) - just that one has a flat-parallel dressing while the other has a compact-helical dressing. My proposition is that the compact-helical dressing will reach a higher load threshold compared to the flat-parallel dressing. Its also to show that the term 'flat overhand bend' is not an appropriate or helpful descriptor. The term 'Offset' being a better term (in my opinion).

The top photo illustrates a typical retrievable abseil system - a double rope device is required. If a single rope device is used (eg a GriGri) - then the system must be configured differently (otherwise you would receive what I call a 'Darwin Award').

Note: The tails in all photos are deliberately tied short so as to fit within the macro view of the camera lens.

Mark G

Edited: Corrected some typos and fixed the grammar in places...
Also removed #1410 image - fingers going too fast on key board - thanks Dan (will re-upload correct image when I get home later..)

[Dan_Lehman]

4. They key objective with any abseil rope joining knot is to obtain a shape / geometry that will translate easily around a 90 edge - this is self-evident. However, I might add that there are also other qualities worthy of pursuing...such as; ease of tying, ease of untying, and - to achieve a relatively small 'footprint'. Logically, the larger/bulkier the knot, the more likely it is to jam or get stuck somewhere (this is my personal observation - empirically derived by experience and observation in the field). If we can achieve all of the primary objectives - why not also try to achieve a smaller footprint? Do you believe this to be a bad quality to pursue?

As a layperson to this topic, I will just go with this as a starting point.

The bulbous knots shown above do not excite me very much.
They may be offset and not be likely to snag on their flat side,
but what a monster the other side is, especially the Figure 8 ones.
As an alternative the Zeppelin bend immediately comes to mind
(all images at 500 kg in climbing quality rope),
however the lip from one collar or the other is the potential problem I suppose.
Otherwise, the knot is low profile and non-jamming.

In other applications, one will find users preferring knots
that place the tails where they are parallel to another part
and so easily bound to it (with tape or hog ring or seizing)
--quite unlike the knot-enthusiast's beloved zeppelin.

In this case of ARJ knots, the key aspect is as noted in
the above quote, and the simple test of that aspect is
pulling the knotted line over corners (of a desk or table,
e.g.) --non-offset knots need not apply!   

Here are images for what I call the offset 9-Oh / 8-Oh.
http://postimg.org/image/vawxt55kv/
(I'm skeptical about the latter, btw; I might favor the
former, but it's a contest with the offset water knot
+ stopper.


--dl*
====

[Dan_Lehman]

For example:

Please NB: what you show for these knots
--for both, I think-- ARE BUT ONE PARTICULAR OUT OF MANY
ORIENTATIONS FOR THEM.

This is obvious. I am not going to take scores of photos
from different angles and ...
I have chosen some of the obvious dressings and posted them.

As I have emphasized above, AND SHOWN (now
that I put in the URLink to the images I'd referred
to but forgotten), there is the "turning-the-dial"
variation to any given offset end-2-end knot (I'll
suppose --the "any"/every assertion) which should
be put into readers' attention.

The images I've linked to show a "side" view,
which should make clear the notion of what
I've called a "forward arc"/"backward loop".
A top-down view  (i.e., looking down at the
tails) could show the different relative angles
of *departure* of the tails --where both run
somewhat perpendicular to the axis of tension,
where they run roughly parallel (mid-range
dial-turning here), and then perpendicular
again, but opposite to first perp. direction.

This trio of approx. end-1, mid, & end-2 points
of the range of "dial-turning" orientations can
be achieved in a knot in hand through torquing
the line --simple (magician's!) twisting can work
amazing things, and drive home the point that
a given token knot can be so varied --it's not
a big matter of re-tying it some different way!

((Btw, could torquing have contributed to the failure
in the old mystery case of Karen Turk ??!!  --haven't
thought to try that and see ... . ))

>>> the term 'flat overhand' isn't really a correct descriptor.
Darn tootin' it's not; dang it to heck!!

Btw, you've got a regular water knot posing as "offset"
--IT is not.

And your comments re the size difference between the Offset #1414 and $1410.
I did make the point that the size difference is not Earth shattering or
deserving of a Nobel prize. I simply made the comment that there is a
size difference --albeit only 10% or so

You don't say how you arrived at your (er-hem) "10-15%"
assertion given previously (now meekly at "about 10");
my "about 8%" was got by using tea-bag strings,
so there!  Real actual-factual tea, which I imbibed
long prior to using the strings.  (Note that I don't ask
you to sew them into anything --but that would
be good penance ... .     )

However, I do make the point that achieving a small footprint is a worthy criteria.

And then we can wonder at what "small" means,
and when is it good enough, and what might be
going beyond need in size reduction : is "smalLER"
always betTER?!  --or is there "small enuff"!?

I'll make my point again for clarity: "A desirable feature is to
achieve a joining knot that has a relatively small footprint".

Which both have.

And it goes without saying that whatever abseil rope joining
knot you choose - it MUST be stable and secure (at least up
to about 4kN or so). Keeping in mind that most retrievable
abseil systems will be on a 'double strand' - the load at the
joining knot will only be 50% (friction around the anchorage
point will alter that 50/50 balance somewhat).

In a mood of caution/suspicion, we might be concerned
that such a knot might need to serve qua stopper
should mis-matched ropes running through a rappel
device lead to the rope moving through the anchorage
such that the knot abuts the rap rings and then
principally one line (the thicker, it would be) is loaded!?
Now one has a sort of stopper-hitch of this loaded
line pulling into the rings and tied to the thin line !


--dl*
====

[agent_smith]

Dan, I am considering authoring another paper on Joining Knots for retrievable abseil systems.

One area that has not been properly investigated is the threshold load at which certain 'Offset' joining knots will capsize / become unstable.
For example, testing to induce a capsize event on the following knot structures would be interesting:
[ ] #1410 (Offset overhand knot  aka Offset ring bend aka Offset Water knot and dare I say it... the 'EDK' )
[ ] Offset F8
[ ] #1410 (M Gommers locked variant)

The dressing would need to be controlled - exact geometry defined for the tests.
I am particularly interested in a test comparison of a flat-parallel dressing Versus a compact-helical dressing.

My proposition is that a flat-parallel dressing of the Offset F8 will give unpredictable and widely varying results.

Maybe Mobius might be interested to perform such tests?

No need for destructive (to point of failure) tests - as this is a pointless exercise. The aim is to measure and quantify instability. As load increases, the threshold load at which instability is triggered would be of great interest.

Some 'slack-shaking' tests could also be performed? And cyclic loading?

Mark G

[agent_smith]

Dan - incorrect image deleted. I will re-upload the correct image.

Its shown at the top of this thread anyways... but I will upload a high quality image later today.

Thanks for pointing out the non-offset image - you have a good eye!

Mark

[agent_smith]

such a knot might need to serve qua stopper
should mis-matched ropes running through a rappel
device lead to the rope moving through the anchorage
such that the knot abuts the rap rings and then
principally one line (the thicker, it would be) is loaded!

Interesting point.

I don't think anyone has performed a side-ways loading of an 'Offset' joining knot (ie the knot jammed up against an anchorage chain/ring and loading sideways as a 'stopper' knot? Do you know of any authored papers or test results?

In any case, I would surmise that such a loading profile would not be severe - some load balancing would ensue... but, there would be bias toward the thicker side - that is, the larger diameter rope will try to slide through the rap chains or whatever anchor rings are used - with the thinner diameter rope trailing.

Unclear as to what the effect of such a (abnormal) loading profile will have?

[agent_smith]

It is low profile, does not jam and has a wonderfully smooth side which I do not see as snagging on anything. True, the second Infinity image is not what I think of as being 'offset', however the smooth knot mound it has is not going to have trouble with a 90o bend either in my opinion.

Hi Mobius... unfortunately, only knots that have an 'offset' geometry will stand any chance of translating around a 90 degree edge from low anchors.

The Zeppelin and other structures based on interlocked overhand knots will likely jam (get stuck) at a 90 degree edge.

The problem is low anchors - that is - if the only available anchors are low-set, and there is a 90 degree edge to traverse, only the 'offset' geometries will translate.

It was this propensity to jam or get stuck up at an edge that led to the search for an alternative. And so this led us to #1410.

But there was a serious accident in the USA back in the late 1990's where Karen Turk took a nasty fall and suffered serious injuries (by a miracle she was not killed). And this (in my opinion) was a catalyst for the 'EDK' moniker to be attributed to #1410.

I follow the evidence...and in my experience - #1410 is stable and secure up to around 4kN.
However, join your ropes with an Offset F8 and its anyone's guess. The dressing of the offset F8 will be crucial.
My proposition is that Karen Turk was abseiling on an Offset F8 with poor dressing (possibly a flat-parallel dressing).

Mark G