Euro Death Knot, Flat Overhand, Barrel Knot evolution

Please see my updated post explaining what I have learned about the best practices and limitations of the flat overhand knot:
topherdonahue.com/blog/2016…

My apologies to everyone for the misinformation. I was, happily, wrong.

@topher: Good on ya! Thanks for posting an update and having the humility to correct a prior error. Lots of respect for that.

Thanks for having the humility to correct your mistake. One thing that didn't seem right in your new article was the figure of 3kn for failure of the EDK. I'd like to see the testing behind that one because I've read results from other tests that the knot broke at 11kn. 3kn seems on par with what I'd expect to see out of 3mm or 4mm cord, not the typical 7-10mm cords people are rapping on.

eli poss wrote:Thanks for having the humility to correct your mistake. One thing that didn't seem right in your new article was the figure of 3kn for failure of the EDK. I'd like to see the testing behind that one because I've read results from other tests that the knot broke at 11kn. 3kn seems on par with what I'd expect to see out of 3mm or 4mm cord, not the typical 7-10mm cords people are rapping on.

Ditto. 3kN (or even 6kN with the doubling) seems low.

Topher's blog post says "flat overhand fails at about 3Kn"

Does failure mean the rope breaks at 3kN or does failure mean the knot starts to roll/capsize at 3 kN?

I agree though, 3kN seems a bit on the low side compared to other testing results.

user.xmission.com/~tmoyer/t…

This is what I've looked at in the past, and for most cases a dressed EDK was holding a 1000-2000lbf, which works out 5-10 kN

To the best of my knowledge there is know such knot as a Euro Death Knot so shouldn't be referred to as such.

cragrat wrote:To the best of my knowledge there is know such knot as a Euro Death Knot so shouldn't be referred to as such.

Beg to differ. The flat in-line overhand or whatever else you'd like to refer to it as is commonly called the EDK ala Euro Death Knot.

Commonly used as a knot for tying two ropes together for rappel.

There...you've expanded your knowledge.

"OFFSET", rap-knots are preferred to be *offset* from
the axis of tension, to flow over rough surfaces and
edges easily. They are not "flat". (This was Clyde
Soles's clear-thinking (_The Outdoors Knots Book_),
and I'm happy to see it now echoed in current _Freedom of
the Hills_.) It simply is the right adjective : there is
an axis of tension, and the knot is offset from this.

> As the "barrel knot" is already another name
> for a dbl. fish. depending on convention, ...

So, one takes the orientation indicator "OFFSET",
and prefixes it to the known "water knot", "fig.8",
"grapevine", and whatever --to get "offset water knot"
(aka "EDK"), "offset grapevine bend", and so on.
With the "barrel", though, there's not so ready a name
(I'm forgetting what anglers call it --"surgeon's knot"
IIRC, and that's a bit of a confusion to that name's
more common denotation --a beefed-up "square knot");
I'd call it "offset strangle bend",
as the structure is of the strangle knot (a binder).

The reason that the offset strangle works arguably
better (= higher capsizing force) is that the choking
strand makes a FULL TURN/wrap around the two
loaded parts. (One actually need only make the
double overhand with this full wrap in the rope
that is doing the choking --the other can be a (mere,
single) overhand, but that's trickier to tie.)

> Speaking of ropes of unequal diameter,
> there is a right and wrong way to tie the EDK for such situations,
> so reports ... that don't reference the tying orientation of the knot
> are worthless, and one has to wonder if the testers are even aware
> of the fact that there are two distinct ways to tie the knot.

Indeed. And beyond that, RG, there are other orientations of
the offset knot body which determine how the loaded strands
bend on their entry :: one loops backwards while its opposite
arcs forwards (these are twinned parts); vice versa; and some
middle-ground. The particular orientation here will show in
the projection of the tails (perpendicular to axis of tension
for loop-&-arc, parallel w/middle orientation.

One MIGHT think that all of this has been covered in the decades
of broad usage; I'd rather see specific cases given some testing
(which need not be to the point of rupture, but just, say, double
or triple expected forces).

But, to say again, such fine distinctions aren't going to sell
for the typical user. IMO, the simple expedient of backing
up the base offset water knot with a 2nd (which loads not
as offset but as a stopper knot abutting the first),
has a lot to offer for accommodating (mis)-tying variances!
As Brian_in_SLC highlighted, page 1, the offset water knot
has the great benefit of simplicity. Doing the simple tying
a 2nd time is still quite simple --simpler than going for the
offset strangle or even offset grapevine (or offset 9-Oh),
or just tying of the one "right" tail.

• kN*

> The thing we all should have learned from the flat-8 issues
> is that just because a knot seems as if it ought to be better
> doesn't mean it is better or even acceptable---failure modes
> might occur that are unanticipated.
> So recommending a new knot without accompanying pull tests,
> done at least to the Moyers standard, is verging on irresponsible.

Both more & less, here :: Tom's testing falls way short of being
comprehensive (note that he, too, doesn't note orientation (and
he thought that the additional orientation aspect I described above
wasn't real --but it is (private knots chatter lonnng ago :o) )).\
That's the "more" part --it would take much more testing.

> All the testing data show ...

All is far from comprehensive; usage, though, is getting there.

The "less" is only meant to point out that individual users
CAN DO THEIR OWN TESTING --they can achieve meaningfull
loads and on the actual ropes that they will use. Since there's
no point to break-testing, but only to see what happens as
expected and --for safety's sake-- roughly double-expected
loads, users can rig themselves simple testing of these knots.
And decision criteria can have "stays solid as tied" vs. anything
dubious, even --no need to insist on failure as rolling.

> Go halfway down to see the pics [re RG's orientation of thick/thin].

Thanks much for finding that!
Note that in the case of the offset grapevine,
or offset strangle,
one could do with *just* an overhand in the
blue rope as oriented by RGold in those pics
--as it is the red that chokes, and would benefit
from the double overhand's (grapevine) full wrap.
(And compare the "wrong" (dubious) orientation of
thick/thin ropes occurs in the offset grapevine's
mages at TopherDonahue's blog.)

But back to the point of simplicity :: getting this right
has two can-get-wrong parts --the thick/thin orientation,
and the which-tail-to-tie-off choice. The "EDK-backed EDK"
(for a shorthand, readily understood moniker :-)
I think will accommodate botching all, and even the
crossing-strands issue, and ... more. YES, this needs
to be tested; and it can be by all the readers of this
thread, with their collected various combinations of
cordage. (Stand in a simple 2:1 pulley'd arrangement
for force boosting, already up if loading the knot in
single strand where rappels load double strands).

> the offset overhand is adequate, but with that particular
> [Petzl RAD] rope the offset 8 is actually stronger.

But the important "actual" --Real Rockclimber Rappelling(tm)--
cases, they're equal --they both hold.

These results are surprising! --putting the offset fig.8
as stronger than the (usually strongest!) grapevine ("dbl.fish").
I'd really like to see video of the testing, to try to see
what is happening here. (In a great many tests with HMPE
cordage, the knot simply spills --doesn't hold. And we know
that often a triple fish / dbl.grapevine is rec'd for HMPE.)

• kN*

knudeNoggin wrote:> The thing we all should have learned from the flat-8 issues > is that just because a knot seems as if it ought to be better > doesn't mean it is better or even acceptable---failure modes > might occur that are unanticipated. > So recommending a new knot without accompanying pull tests, > done at least to the Moyers standard, is verging on irresponsible. Both more & less, here :: Tom's testing falls way short of being comprehensive (note that he, too, doesn't note orientation (and he thought that the additional orientation aspect I described above wasn't real --but it is (private knots chatter lonnng ago :o) )).\ That's the "more" part --it would take much more testing. > All the testing data show ... All is far from comprehensive; usage, though, is getting there. The "less" is only meant to point out that individual users CAN DO THEIR OWN TESTING --they can achieve meaningfull loads and on the actual ropes that they will use. Since there's no point to break-testing, but only to see what happens as expected and --for safety's sake-- roughly double-expected loads, users can rig themselves simple testing of these knots. And decision criteria can have "stays solid as tied" vs. anything dubious, even --no need to insist on failure as rolling. > Go halfway down to see the pics [re RG's orientation of thick/thin]. Thanks much for finding that! Note that in the case of the offset grapevine, or offset strangle, one could do with *just* an overhand in the blue rope as oriented by RGold in those pics --as it is the red that chokes, and would benefit from the double overhand's (grapevine) full wrap. (And compare the "wrong" (dubious) orientation of thick/thin ropes occurs in the offset grapevine's mages at TopherDonahue's blog.) But back to the point of simplicity :: getting this right has two can-get-wrong parts --the thick/thin orientation, and the which-tail-to-tie-off choice. The "EDK-backed EDK" (for a shorthand, readily understood moniker :-) I think will accommodate botching all, and even the crossing-strands issue, and ... more. YES, this needs to be tested; and it can be by all the readers of this thread, with their collected various combinations of cordage. (Stand in a simple 2:1 pulley'd arrangement for force boosting, already up if loading the knot in single strand where rappels load double strands). > the offset overhand is adequate, but with that particular > [Petzl RAD] rope the offset 8 is actually stronger. But the important "actual" --Real Rockclimber Rappelling(tm)-- cases, they're equal --they both hold. These results are surprising! --putting the offset fig.8 as stronger than the (usually strongest!) grapevine ("dbl.fish"). I'd really like to see video of the testing, to try to see what is happening here. (In a great many tests with HMPE cordage, the knot simply spills --doesn't hold. And we know that often a triple fish / dbl.grapevine is rec'd for HMPE.) *kN*

This reads like a bottle of Dr Bronners

Brian Abram wrote: This reads like a bottle of Dr Bronners

Yeah, esp. in the age of life-by-Tweets --as though reality is so simply explained.

• kN*

A rigging for rescue instructor was teaching this:
rockandice.com/climbing-gea…

This week in a class I was taking. 3" tail. He said they have pull tested it and it's good to go. His argument against a regular EDK - why do I want to use a knot that requires a 12" tail.

In 46 years of climbing primarily trad climbs with thousands of rappels, I've never had a grapevine get stuck.

The rappel ropes which did get stuck got stuck on a flake or the dreaded bight in a crack which is as bad as a knot. The rope around a flake can sometimes be flipped off the flake. However, the bight in a crack always required climbing up to free it.

With a little thought, you can avoid the problems with the grapevine snagging during the pull. An effective method to avoid problems is to have the first climber down give the rope a test pull. If there was a problem it could be addressed by the second climber before starting a rappel.

Every knot or technique has advantages and disadvantages. Like most discussion of items like this, it comes down to what you are most comfortable with.

sfotex wrote:His argument against a regular EDK - why do I want to use a knot that requires a 12" tail.

I've been making that argument about the EDK as well. Glad to see I'm not the only one.

aikibujin wrote: I've been making that argument about the EDK as well. Glad to see I'm not the only one.

Don't you put a tail on your 8 as well? Maybe not 12 inches. But, most every knot has some tail. 12 inches is just a recommendation. Any test I've seen show that at fairly high loads (1200 lbf) the EDK may roll once consuming a few inches of the tail. In order for it to roll a second time, the load needs to be even greater. Even with a simul rap, 1200 lbf would be hard to achieve.

Greg D wrote: Don't you put a tail on your 8 as well? Maybe not 12 inches.

Exactly. I know all the argument for the flat overhand, and I've used it myself for many years without any issue. But it doesn't prevent me from switching to a knot that doesn't require a 12-in tail and roll at 1200lbf or whatever force. Obviously, some people like to stick to whatever worked for them.

Obviously people stick with what works and is as safe for intended use as with any other mentioned rigging configs.