Rope joining knot failure - death

[agent_smith]

Dan / Scott,

Can you guys please look into this recent fatality please?

Link: http://methowvalleynews.com/2016/05/04/climber-dies-in-fall-at-mazama

Tragic loss.

However, this is an opportunity to determine exactly what type of end-to-end joining knot was used for the abseil descent.

Anything you can dig up would be appreciated.

I am preparing to write a new paper on the analysis of end-to-end joining knots for joining 2 climbing ropes together.

So this is all very timely.

I think some of us here at the IGKT can help straighten out the facts from the myths surrounding 'offset' joining knots for abseiling.

Mark G

[Dan_Lehman]

Here's much of the info from the cited article :

?The team?s rappel [dl : previously by another member, & presumed (re-)done prior accident]consisted of a twin rope rappel which entailed tying the ends of two identical ropes together with a backed up figure eight knot. One of the ropes is then threaded through the anchor and a full-length double rope rappel is achieved,? Brown said in the press release.  ?This provides a much faster descent as they end up rappelling a full rope length of 70 meters per rappel station.?

Erps became fatigued from leading many of the pitches and then managing ropes on the rappel, Brown said. At the second-to-last anchor, Kautz took over and rigged the rappel at about 3:30 p.m., Brown said.
[dl : But why should this entail RE-tying the ARJoint knot?
 --vs. just re-threading/-clipping the already-joined-as-just-used
 abseil line(s) into the new, lower anchor!? ]

Jackson and Erps told sheriff?s deputies they trusted Kautz to rig the rappel because he had ?instructed this sort of thing with Outward Bound.?

A spokesperson for Outward Bound, however, said there is no record that Kautz worked for the organization.
[...]
Jackson was in position to see Kautz put himself on belay then pull down and check the function of the rappel device with a personal arrest system still in place, Brown?s release said.

Jackson watched as Kautz began his rappel. He had rappelled about 5 meters when Jackson and Erps said they heard a loud snap and Kautz and the rope fell. Jackson told deputies that Kautz struck two ledges before he fell into a gully out of sight of Jackson and Erps, Brown said. It was estimated that the fall was more than 100 feet.

I've highlighted key points, and interjected some notes
& questions (i.p., why should ropes be repeatedly tied
vs. just at the start, and then re-used on subsequent
pitches?!).

"a loud snap" isn't what I'd expect of a knot failing,
but it could be the consequence of a suddenly freed
pair of tails smacking rock AFTER the knot spilled!?

The ARJ knot of two nearly identical ropes (i.p., in length)
should be up with the waiting-to-abseil members, right
at the anchor.  They should've heard that snap, if from
this knot somehow spilling, close by --and so remark.

FYI, I just checked (eyeballing threads listing, then Searching
"rappel") supertopo.com's forum and found no hint of this.
Rockclimbing,com is similarly silent --though more so in being
now of challenged relevance (on account of the new owner
wanting to AVOID the too-common, abusive exchanges
--a sad commentary on the outlaw nature of climbers, alas!).
RC.com has just the sub-forum for this sort of thing, too.

Now, IMO w/o further knowledge, I'm baffled.
As previously noted, I'm puzzled at there even being some
need to RE-TIE the ropes together!?  And the cited knot
being not only one w/o known risk (if indeed it is the regular
and not "offset" fig.8 end-2-end knot) AND was backed up!?
Granted, the victim was to have re-set the abseil pitch on
which thing went tragic, but ... !?

Btw, there could/should be a remaining fig.8 single-strand
knot if the specified knot was used, and possibly also if the
offset one was.


--dl*
====

[roo]

I don't think we're going to find out anything until the investigation is wrapped up.  I'm not sure what the snapping noise would be (recoil? rocks?), but I think we can rule out rope breakage as that would be pretty obvious to anyone. 

I wouldn't be surprised if a knot was left partially tied.  An offset figure eight can turn into a time bomb if one end is backed out one tuck.

[Dan_Lehman]

Roo, of course, you're right at least in the sense that
we're left only to speculate, given the cursory information;
you might be wrong about their investigation getting it right
--but we can hope!  And such an investigation has so
little to go on : we almost can't hope.

Maybe the victim untied the ropes so as to avoid having
to thread 70 METRES entirely through the anchor, and therein
lies an unknown (what he tied --something that held for
a while).  The article's talk of knowing how "to rig the rappel"
sounds odd to ME vis-a-vis joining the ropes, which is the
only thing that seems to have been a problem (not a failed
anchor, which would leave ropes knotted in post-mortem).

Just as the infamous supposed only-known? failure of the EDK
where "Karen" fell to injury involved the knot tyer at first
thinking to re-tie the knot but then changing his mind,
and so where we might suppose that in trying to restore
the partially untied knot he got ... an offset fisherman's
or something less stable, even?

Apparently, the victim made claims of training that were false.
But there remains the question of how whatever training got
or not would come into play --just the knot-facts, ma'am.

I wrote to the sheriff's department w/pertinent questions;
perhaps they'll make a helpful reply.

Stay tuned,
--dl*
====

[Dan_Lehman]

I have now (Sat or Sunday, 5/08 latest) sent two e-msg.s
to the sheriff's office with questions; of course, they might
just ignore such extraneous queries ... .

There is some discussion of the accident on the Mountain Project
site --here:
www.mountainproject.com/v/rappelling-death-on-goat-wall/111831080__2

Notable on this site is a URLink to some YouTube video of an
offset water knot (EDK) tied in wet old climbing rope being
pulled (in a sling) to rupture.  A person present during this ad
hoc *testing* remarks that the knot didn't even start "roll"
until about 3_000 pounds force was reached (!!).

Maybe more notable is another video which I saw among the
handful that YouTube puts up as suggested further interests
in which the testing is of the "EDK & EDK-8" (OWK & offset fig.8).
What is so interesting in this is that
1) the offset fig.8 knot endures such great force,
and
2) ... it does so without showing the sort of large-jump
consumption of rope when it flypes --indeed, it seems to
more smoothly "roll" as does the owk !?
--which stands in contrast to what I believe Tom Moyer found,
and so what we should understand is that there can be some
rather different behaviors of knots in different situations!
(Not everyone ties knots just this way or that, in just
this material ... .)


--dl*
====

[agent_smith]

Here's the link for the Karen Turk accident report:  http://publications.americanalpineclub.org/articles/13199808000/Fall-on-Rock-Rappel-Ropes-Knot-Unraveled-Wyoming-Grand-Teton-Guides-Wall

With respect to the current Kautz accident... it would be helpful to get more detailed info on what the surviving members of the party witnessed.

However, as with the Karen Turk accident - there was shock/trauma so the survivors may have unreliable memories.

And since the rope 'came apart' in both instances - we are left with no hard evidence.

I really want to try to understand exactly what type of end-to-end joining knot was used... Perhaps contacting Jackson and Erps might be the only way - and them asking them to try to recreate the knot that they think Kautz tied? But given that they might be suffering from PTSD - they may decline to assist (I am of course speaking in terms of a formal request for information from the IGKT...which the IGKT may also be reluctant to do?).

Mark G

[Dan_Lehman]

Here's the link for the Karen Turk accident report:  http://publications.americanalpineclub.org/articles/13199808000/Fall-on-Rock-Rappel-Ropes-Knot-Unraveled-Wyoming-Grand-Teton-Guides-Wall

Thanks for making this URLink handy!
(memo to self :: copy this to computer knots file!)

IMO, the likely explanation of THIS failure is that the
first rappeller, who was HE who tied and then started
to UNtie but then RE-tied/restored --ALLEGEDLY-- the
original knot (offset water knot / EDK
ACTUALLY MIS-TIED ... and maybe got an offset fisherman's
knot instead, and this worked to failure.  Else, maybe
he got a poor dressing of intended knot, loosely set,
and it subsequently rolled off. ?!

[RE-READING URLink'd ARTICLE]
1) 13th Sept, but of WHAT year ?  (file access suggests "1998"?
2) WHOA, the original account --somewhere-- had quoted the two
gals as expressing doubts about the look of the knot --that it looked
odd/funny, but one gal replying "don't <mess> with it, we don't know...".
THAT note is missing from this account.

3) The account does suggest that my surmise above goes
too far, that the knot was merely "loosened" but then
re-set --so, no untucking ... .  But, then again, one can
both doubt the preciseness of expression, or even the
full honesty of a party who would be looking not-so-good
were he to have admitted to partially untying the knot
--and maybe one can wonder, too, if indeed the account
(or some other record of it) has been changed for this
archive, just as the missing testimony from the gals
shows a difference!?
If indeed the knot was merely loosened and then re-set
(and in any case, rappelled on by the guy, likely heavier
than the victim by 30# or more?),
I can only surmise that the used & slightly thicker rope
was, along with some issue of dressing, a contributing
factor in the knot rolling.  But that sure is a puzzling
picture, where a lighter (presumed) force --or did the gal
maybe do a jerkier, shock-laden descent?-- then makes
the knot not only roll but continue rolling!?

With respect to the current Kautz accident... it would be helpful to get more detailed info on what the surviving members of the party witnessed.

Sure would, but as of Friday post-noon (5/13) my query
to sheriff's dept. has had NO reply, alas.  I should think
that those members would've been right at the knot
and should've remarked at it were it much different
from a supposed --even THIS *fact* has yet to be affirmed--
backed-up fig.8 bend, and something that could fail
I'd think would need to be of noticeably less bulk than
that, and likely oriented differently --in short, OBVIOUSLY
new & different (and should've been questioned ; the one
testimony is of observing the victim pre-test the system,
after all).

And since the rope 'came apart' in both instances - we are left with no hard evidence.

This a non-sequitur.  Indeed, one might suggest
that of any knot entailing two *halves* each of which
itself is a knot (i.e., e.g. an overhand and not simply
a nipping turn or U-turn bight), it is more likely in the
case of spilling that one half will remain.
I remember my case of foot-vs-arm/torso tensioning
of #782 (or nearby) which CAN capsize into #1452
(and so I felt was mighty safe) where the knot spilled
--ONE side pulled out, mid-capsizing, the other overhand
remained intact.  (hard to loose both, simultaneously)

The case puzzles me in various aspects;
I'm not hopeful about getting clarity, but maybe
just sharpening/refining the troubling aspects.
(E.g., is it common practice to RE-tie rap. ropes?
One remark from veteran climber I read indicates
that it is NOT --that one ties once & re-uses on
later descents.)


--dl*
====

[agent_smith]

is it common practice to RE-tie rap. ropes?
One remark from veteran climber I read indicates
that it is NOT --that one ties once & re-uses on
later descents.

The answer to this question is: It depends.

If you are descending a pre-equipped (ie existing permanent anchors) vertical route - where there are no significant obstacles underneath each rap station - as you pull to retrieve the climbing ropes, you can immediately feed the end through the next rap station anchor and then just keep pulling.

As you pull, eventually you will reach the 'tipping point' - when the mass of retrieved rope perfectly counter-balances the opposite rope (you can feel when this happens).

The next pull or two initiates free-fall - and the point of no return in the retrieval process - as the remaining opposite end begins to free-fall down the cliff.

Basically, you are using gravity to your advantage - gravity aids in feeding the initial rope length through the rap station as mass increases - and then the free-falling opposite rope length automatically deploys. The location of the rap station affects how well (or how easily) this all unfolds.

The trick here is to avoid potential rope snags and foul-ups.

I have used this technique successfully many times... as it speeds up the descent and belay change-over process. And the rope joining knot remains intact and does not need to be untied and then re-tied. However, it does need to be re-checked (obviously).

I should emphasize that this whole procedure generally only works on pre-equiped (existing) routes. Where there are no existing permanent anchors (ie rap stations) - it is less likely that this procedure would be employed.

IMO, the likely explanation of THIS failure is that the
first rappeller, who was HE who tied and then started
to UNtie but then RE-tied/restored --ALLEGEDLY-- the
original knot (offset water knot / EDK
ACTUALLY MIS-TIED ... and maybe got an offset fisherman's
knot instead, and this worked to failure.  Else, maybe
he got a poor dressing of intended knot, loosely set,
and it subsequently rolled off. ?!

Yes indeed... the #1414 in its 'Offset' geometry is indeed unstable - it will yield at considerably lower loads than other offset joining knots. You know what?...I cant remember if any 'remnant' knot remains after #1414 yields... I had better double check this asap.
My point being that we could be left with some 'evidence' - a remaining #46 (overhand knot)?

Did your letter to the Sherriff's department enquire as to whether the rope ends were photographed? Or if there were any surviving knot fragments/remnants remaining intact?

[Dan_Lehman]

My letter could've asked ANYthing --the non-answer is unchanged.
But, no, I didn't ask for photos --just facts (the fact of remnant knot).
Whether a knot remains will depend on whether tails were the same
length --in some cases--, or how the knot failed (in my case with
the expected-to-capsize-into-#1462-offset-knot one rope was able
to pull free with the other still holding its overhand knot).

As for pulling ropes and tipping point, if one is pretty efficient in
sizing *pitches*, then very soon in the pulling you will have of
course the full length of the pitch on the pulled side and the
ever-shortening counter-balancing end going up; friction at
the top --more if run through just webbing-- will slow the
pull but not equal masses.

As for threading it while pulling, well, okay, but I don't see
that as making a big difference (and it would NOT work if
one needed to retain the knot on the same side --i.e., with
either haul line to pull or climbing line to pull.
 (I forget if there's a preference here, too : in the former case,
that tipping point comes later, and the knot would serve to
abut the anchor for any dissimilar slippage in the rappel
device; in the latter case, one would have the climbing
rope in hand (more of it, that is) should the tied lines
get stuck.).

In whatever case, checking the knot should be rather
forced just on account of having to position the knot
in the center by the anchor --not that one might even
care or want to check it for integrity, but the knot will
BE there anyway!


--dl*
====

[agent_smith]

Here is another abseil (rappel) accident where an 'offset' joining knot was allegedly involved:

Link: http://canyoncollective.com/threads/zion-fatality.3117   (scroll down to read report from Richard Connors)
Victim: Roeslan Tamin - aka 'Ross' (35 year old male from UK)
Date: May 21, 2002
Route: Descending the Leaning Wall from 'Spaceshot' - Zion, Utah USA
Knot used: Victims climbing partner reports that it was Offset F8

I've looked at the American Accident Journal but had no luck trying to find the full accident report (just for official complete records sake).

Dan - are there any full accident reports available on the AAJ site in relation to this? The report from Richard Conner does appear to confirm that it was Offset F8...

[Dan_Lehman]

I've looked at the American Accident Journal but had no luck trying to find the full accident report (just for official complete records sake).

Dan - are there any full accident reports available on the AAJ site in relation to this? The report from Richard Conner does appear to confirm that it was Offset F8...

Your search is likely as good as mine would be.
(And what if what were found turned out to LACK
some of the details given by the partner full account?!).

Meanwhile, I have received no reply regarding the recent
(May) death, alas.  (Though, after a week of nothing I'd
not expected anything.)
Nor do I find further discussion about it.


--dl*
====

[agent_smith]

If anyone can dig up any accident reports involving a situation where 2 ropes were joined (end-to-end) for the purpose of an abseil descent - this would be appreciated.

Am not just considering accidents where there was a fatality - also interested in accidents where the victim survived (eg the Karen Turk accident). Am trying to widen the scope of the search...

Important factors for me are:
1. Whether any remnant knot or tangle survived in the rope(s) - this will provide some indication of where to start looking for answers... eg if no remnant knot remained in the victims ropes, this helps to guide the analysis

2. The type of ropes used...eg low stretch ('static') or dynamic

3. Differences in rope diameter

4. Weather conditions - eg rain, ice, dry (eg wet/saturated/muddy or icy ropes may behave differently)

5. Knot geometry - eg an 'offset' structure, symmetric or asymmetric structure

6. Knot species (if known): The exact #ABoK number if known or precise structural details...although this is the most problematic - traumatized victims will have unreliable memories - and if the ropes have completely come apart with no surviving/remnant knot - this further complicates matters - again possibly only relying on the victims memory. A remote possibility is that the person most responsible for tying the joining knot may wish to obfuscate or alter the facts to protect him/herself from adverse investigation (eg to protect ones own personal reputation)??

...

Purpose: I am gathering information for my paper on abseil rope joining knots.


Mark G

[ogopogo]

Agent Smith,

What is the purpose and scope of the paper you are authoring?

I live and climb only a couple hours north of the Kautz accident. I'm an avid climber (although not a seasoned mountaineer), and what I would call passionate about knots (although not an authority all on my own).

The debate about the "Euro Death Knot" is an interesting one. The knot is an excellent knot to rappel with so long as:

1.  It is made clear that the flat figure eight is not a good alternative
2.  The knot is dressed with long tails left over
3.  The knot isn't used if the rope is not going to be subject to high forces

These seem like a lot of buts to me. The merits are of course that it is easy to tie and it may be less likely to snag.

Would it not fit with the IGKT constitution and duty to the public to definitively, exhaustively and scientifically look at the merits of the flat overhand bend and every other likely bend to find a best solution? To perhaps publish an authoritative study?

I like learning and understanding knots because I like knowing the very best knot for a given situation, I am also a fan of building good habits. As it stands now, someone who tied for example a zeppelin bend to join rappel ropes probably wouldn't be taken seriously, maybe that should change. People are dieing because of the knots that are or are not being tied in rappel ropes, not many, but maybe we have a chance to save even one future life.

Anyway, maybe this is what you are doing. Let me know how I can help. I'm not however aware of any unpublished rappel not failures.

[agent_smith]

Hello Mr/Miss ogopogo (nice name by the way),

Any assistance I can get (from anyone) would be greatly appreciated!

Would it not fit with the IGKT constitution and duty to the public to definitively, exhaustively and scientifically look at the merits of the flat overhand bend and every other likely bend to find a best solution? To perhaps publish an authoritative study?

You have captured the essence of my objective

Scott or Dan will most likely be able to summarise the IGKT position on this matter...my personal view is that the IGKT as an entity may not be able to publish such a paper. It appears to be left to individuals to do so.

Please download my Analysis of Bowlines paper (www.paci.com.au  then click on 'public downloads', and then 'Knots and knotting concepts'). This will provide you with an example of the work that has been done so far. This paper is not from the IGKT - but, brilliant and incredibly talented individuals like Dan Lehman, Scott Safier, Constant Xarax, Alan Lee, Knotsaver etc - generously gave their time to contribute and critique the paper. They gave their time willingly and without expectation of any reward - and this to me is the essence of what the IGKT is.

Although my name appears on that paper as the author - I will be the first to admit that my level of knowledge and understanding is insignificant compared to the genius of those individuals who I named. It is inconceivable to embark on such a project without the assistance and contribution of members of this IGKT forum.

...

Back to your post:

I would urge you not to use the term 'flat' to describe the structure of these rope joining knots. They are not 'flat'.
They are Offset.

Definition of Offset:
A knot is defined as offset if the core (nub) is displaced from the central axis of tension. A general consequence of the displaced core is that both Standing Parts (SParts) converge and then follow a parallel pathway that is perpendicular to the axis of alignment. A classic example of an 'offset' structure is #1410 (refer Ashley Book of Knots).

This definition is a work in progress...and I have no doubt that Dan Lehman will weigh in and correct me 

The 'analysis of rope joining knots' paper will examine in detail a range of offset and other end-to-end joining structures (eg the Zeppelin bend). It will also offer up some secure and stable solutions.

And there is nothing 'wrong' per se with the Zeppelin bend' - its typical of the climbing community to avoid or negative anything that is not within their normal paradigm. The key is practice. Practice breeds familiarity and familiarity breeds confidence.

I am hard at work on this paper right now... and will upload a first working draft for comment soon.

Some things I am looking at:
[ ] jamming threshold
[ ] instability threshold
[ ] force required to initiate translation around a 90 degree edge
[ ] remnant knots remaining after MBS yield point has been reached
[ ] effects of different rope diameters
[ ] effects of water (wet ropes)
[ ] dynamic compared to low stretch in (using same joining knot)

Maybe you can assist me?

Mark Gommers

EDIT: I have added an extract image from my up and coming paper on 'Analysis of rope joining knots' - hopefully it will meet with Dan Lehman's approval?

[ogopogo]

Thank you for the correction on nomenclature, I appreciate it. I will stay away from using the word flat.

I have started to read a couple of the publications on the PACI site, what a great resource. Do the papers listed under "Knots and Knotting Concepts" do a fairly good job of reflecting the current body of knowledge? At least in relation the the paper you would be working on now?

I am reading the papers "Knots" and "Bowline Analysis," and am very impressed. Very well done and thorough papers...I obviously have a lot to learn. No doubt a lot of time and effort went into creating those.

I would happily get involved with separating fact from myth and establishing an authoritative guide. I am looking forward to seeing what you have put together so far.

________
I would be particularly interested to see where you are at as far as developing a Quality Function Deployment (QFD) matrix, testing (and perhaps what kinda test standards are being used and if they are related to UIAA/EN or similar) and data collection as well as statistical analysis of the results. Then maybe I can find somewhere to jump in.