Double loop knot recommendations....

[Dan_Lehman]

I have 'borrowed' Dave's image and coloured the weakness in blue.



I grant you that I am talking principle rather than practice.

No, you haven't : the serious problem with this knot (among perhaps
some others) is that the side of that eye that you have NOT colored,
which connects to the LEFT SPart, will likely pull out if that SPart is
loaded alone (!!).  I became aware of such things in e.g. a bowline tied
fully with a bight, and hence in *principle* fully redundant:  but,
load one SPart that feeds into the unloaded (cut is worst case) eye,
and --wow-- in rather silky new 1/2" Sta-Set yachting double-braid
polyester (and I think I tried some other...) it slips disappointingly
quickly.  (It IS a great knot IF BOTH eyes are used qua one, and
then one has from this two independent legs to anchor.)

--dl*
====

[Mike]

Would this work?

[Chad]

Wow, thanks everyone for the thoughtful reply's.  I'll try to answer everyone's questions.

knudeNoggin
You make good points about here being several non-redundant pieces in a climber's set-up (single rope, belay loop, belay carabiner, tie-in knot).  All of these pieces are very overbuild so as to not fail under normal use.  The cord that is often used for these lanyards, including the one pictured, is 7mm.  I've not had problems with the pictured lanyard but I also don't think it's completely kosher given the reduced load rating when compared to something like the belay loop (which I think are usually rated at 40kN).  

I retied the pictured lanyard once I had it off of my harness.  It was not girth hitched to my harness but rather tied to it via the loop that was formed from the overhand knot.  I hope this makes sense.  This is why I can't tie whatever knot that I'll end up using on a bight.  The lanyard that is currently on my harness was tied with an Alpine Butteryfly.  That was a pain to tie not on a bight.  

A figure eight double loop does not meet my criteria.  The loop comes undone with little resistance once the other loop is cut.

Transminator
The Triple Crown looks like a possibility.  I'll have to play with that.  It's a new one for me.

SS
Please see my comment above about trying the knot on a bight.  Unfortunately, the double loop butterfly has an interesting new problem in that one of the legs of the lanyard will pull though if a loop is cut.  It appears that the opposite loop would not be compromised though.

Derek
Your proposed knot looks good but you lost me on the acronyms.    Really though, I'm pretty new to this stuff, dumb it down for me.  I'll look into what you are suggesting and get back to you.  

Roo
Until I get my knot books from the library, I'm in the dark about the Midspan Sheet Bend.  The double splayed loop looks like a possibility.  I'll have to play with that.

Mike
Yes, substitute the quicklink with my belay loop and that would definitely work except the bulk is not so appealing.  I may not have a choice in the matter because I may not be able to figure out how to tie any of suggested knots not on a bight.  What is the term for this anyway?


Thanks again everyone.  Any more suggestions and input would be well taken.


Chad

[roo]

Roo
Until I get my knot books from the library, I'm in the dark about the Midspan Sheet Bend.

You won't find this double loop in library books.  Look at the center diagram on that page:

http://notableknotindex.webs.com/midspan.html

Notice that it will pinch off one loop around the rail, and a leftover loop will also be made.  That's your two loops.  But if on-the-bight loops don't fit the bill, this may not be a realistic solution.

P.S.  If you're certain that both legs of your system will always share that load evenly so as not to cause concern regarding overloading, then you could use two separate lines to connect to one carabiner or link.  The carabiner would replace the knot body and the associated double loops.  The bulk would be low as you could use standard single loop knots.  Single loop knots tend to have fewer vulnerabilities (since they are simpler) and tend to get much more testing in real life in comparison to double loops which rarely see use or testing.

If you only want to use one line, you could also use a Pile Hitch as an on-the-bight interface between a carabiner and the line:
http://notableknotindex.webs.com/pilehitch.html

Again, this would take the place of the knot body and the associated double loops.

If one loop of your double loop breaks, then your assumption about sizing and load sharing was wrong, and I'd imagine the second loop would almost certainly break once it is all alone (and it would happen in a microsecond after the first loop breaks).

Here are some jam resistant single loops with good security:
http://notableknotindex.webs.com/zeppelinloop.html
http://notableknotindex.webs.com/waterbowline.html

[DerekSmith]

Sorry Chad, I hate the use of acronyms and here I am using them like a seasoned Yank.

SP -- Standing Part - in this case I was referring to the line(s) going off to the 'bina(s).

BWL -- Bowline

MWK  -- Matthew Walker Knot

I will see if I can find a way to get you more detailed information.

Derek

[knudeNoggin]

knudeNoggin
You make good points about here being several non-redundant pieces in a climber's set-up (single rope, belay loop, belay carabiner, tie-in knot).  All of these pieces are very overbuild so as to not fail under normal use.  The cord that is often used for these lanyards, including the one pictured, is 7mm.  I've not had problems with the pictured lanyard but I also don't think it's completely kosher given the reduced load rating when compared to something like the belay loop (which I think are usually rated at 40kN).

The relevance of the load rating is to an expected load.  The lanyard can be
replaced if it gets worn, the belay loop is expected to endure longer, hence
its beefier nature.

I retied the pictured lanyard once I had it off of my harness.  It was not girth hitched to my harness but rather tied to it via the loop that was formed from the overhand knot.  I hope this makes sense.  This is why I can't tie whatever knot that I'll end up using on a bight.

One can tie a two-eyes knot through a ring (or a harness) in the bight.
E.g., a simple meeting of your criteria (beyond what I continue to say
meets it, the so-called "Bunny Ears 8") is tying a Fisherman's knot with
the bight; you can then finish by taking the independent ends out
through the bight-end of the knot (which finish now is clearly not in
the bight).

  The lanyard that is currently on my harness was tied with an Alpine Butteryfly.  That was a pain to tie not on a bight.

But that is because you have not realized that the knot can be tied
in the bight (i.e., as though the two ends were joined) in the usual
manner, and it will transform into the desired orientation with a little
coaxing -- legs of the eye become ends and vice versa, that sort of
transformation.  This is for the common knot, not a 2-eyes one, btw;
I have not investigated the latter for the property.

A figure eight double loop does not meet my criteria.  The loop comes undone with little resistance once the other loop is cut.

This statement I cannot understand; I can tie the knot and test it
with some force and see nothing to show me any coming undone,
and certainly not with "little resistance"!! ??  So, how badly are you
tying this knot or some pretence to it that you find your behavior?!

(I seem to recall some other climber's seeking of this questionable
ideal of redundancy to an even stricter demand than you make : he
wanted to guard against even the connecting strand of rope between
eyes being cut and leading to failure (it would take some skillful,
focused cutting to achieve this!).  And as I recall he found someone
who had tested the Bunny Ears and got the failure as feared, but at
a rather high force (I think it was higher than top-roping would
provide to an anchor).  So, that is testing of a more demanding
criterion than yours and results that strongly support the Fig.8
staying tied with a cut eye.)


*knudeNoggin*

[Mike]

Chad, I'm not sure what it's called.   It is just 2 figure 8 in a bight.   I got it out of a climbing/rappelling book called "Life on a line"  It's a very good book and the ebook can be downloaded pretty cheap.

[knot4u]

I may not have a choice in the matter because I may not be able to figure out how to tie any of suggested knots not on a bight.  What is the term for this anyway?

It's called tying a loop by using an end (as opposed to tying a loop "on a bight").

[Transminator]

I have 'borrowed' Dave's image and coloured the weakness in blue.



I grant you that I am talking principle rather than practice.

No, you haven't : the serious problem with this knot (among perhaps
some others) is that the side of that eye that you have NOT colored,
which connects to the LEFT SPart, will likely pull out if that SPart is
loaded alone (!!).  I became aware of such things in e.g. a bowline tied
fully with a bight, and hence in *principle* fully redundant:  but,
load one SPart that feeds into the unloaded (cut is worst case) eye,
and --wow-- in rather silky new 1/2" Sta-Set yachting double-braid
polyester (and I think I tried some other...) it slips disappointingly
quickly.  (It IS a great knot IF BOTH eyes are used qua one, and
then one has from this two independent legs to anchor.)

--dl*
====

I admit that the above picture looks a bit scary at first glance.
But when examining the mechanics closer, you find several points of friction
that explain the good test results of this knot.
First the line goes round the SP. When the loop is loaded, it will put downward
force on the SP, but the SP resist this by remaining stright.
Next the line presses down on itself and further down the two standing parts
put even more pressure on it. The left leg of the right loop loops around the first loop
which adds another friction point.

Also I could not reproduce the problem Dan found. Also on careful examination
of the workings of the knot, I find enough points of friction that should prevent an
easy failure of the knot even when only one standing part is loaded.
So in principle, it should perform well in all those scenarios and I put it to the test.

I used 5mm polyester string with a silk core I use for fancy knotting. It slips easily
and furthermore stretches slightly, which reduces the diameter of the string, which
increases the tendency to slip.

I tied one SP to a door handle and pulled on either loop without any slippage whatsoever.
For comparison I used the same material to tie a Karash Double Loop and when pulling only
one loop, it did not take much force to make it slip at the expense of the other loop.

Next I used a 12mm synthetic rope. It is a cripple of a rope as it is slippery and stiff but
very good for testing (and to demonstrate e.g how a bowline can fall apart in stiff rope when there is no load on it
or how well the zeppelin and butterfly bend work)
I tied it to the top of the landing with one SP and put my entire weight on each loop at a time
No slippage.
Based on the above, I would estimate the triple crown as one of the safest double loop knots.
And it also remained reasonably easy to untie.

[Dan_Lehman]

you find several points of friction that explain the good test results of this knot.

Actually, quite the contrary, for the specific loading I described:
the SPart loaded feeds into the broken eye and nips itself but one
time and there with a *protector* parallel strand to impede this.

Consider that in testing of kernmantle ropes done by Dave Richards
the Sheet Bend (& double) slipped.

Now, do you find the Fig.8 bunny ears slipping?  That I  cannot produce,
the connector bight placed farthest up along the SPart (rather than down
around the mid-part of the knot body).  In this case (Fig.8, i.e., either way),
the broken connector strand is nipped going into the knot from the broken
eye and then again going back out into the loaded eye (where it bears 50%
of the force, and will have --additionally-- the friction around whatever is
contained in the eye to reduce the ability to pull out.


As for Karash eye knot, I thought it was safe, but in some 3/16" solid braid
nylon(?), I did get it to slip, alas.

--dl*
====

[DerekSmith]

I admit that the above picture looks a bit scary at first glance.
But when examining the mechanics closer, you find several points of friction
that explain the good test results of this knot.
First the line goes round the SP. When the loop is loaded, it will put downward
force on the SP, but the SP resist this by remaining stright.
Next the line presses down on itself and further down the two standing parts
put even more pressure on it. The left leg of the right loop loops around the first loop
which adds another friction point.

In a good knot, friction alone is not enough, it must be coupled to a positive feedback structure.  A typical structure is demonstrated by the  'Simple Hitch (ABOK #49)



In this hitch, a small amount of tension presses the end against the crotch, gently pinning it in place.  The small friction of this pinned end, is amplified in the turn around the back of the crotch allowing a much greater force to be applied to the SP.  As load on the SP increases, so it pins the end progressively harder against the crotch creating an ever greater anchor force and in turn an ever greater amplified grip around the back of the crotch.

It is a positive feedback system that amplifies the friction as the load is increased.  Without this positive feedback, friction alone would eventually be overcome when the load exceeded the frictional grip.

In the case of the coloured example, if the RHS loop was cut just at the end of the blue marked section, then no force would be supplied into the knot from the RHS loop and there are no apparent feedback characteristics in the remaining knot.  If load is then applied to the LHS loop, it is just a matter of force before the cut end friction gives out and slides out.

Derek

[Dan_Lehman]

...  As load on the SP increases, so it pins the end progressively harder against the crotch creating an ever greater anchor force and in turn an ever greater amplified grip around the back of the crotch.

But evidently this is false -- we can see knots slipping at high forces that had
held up until then (e.g., a Prusik hitch on a line, or the Sheet Bend in climbing
kernmantle, and the entire darn Bowline in HMPE 12-strand line!).
There is something happening in which the increases of force give
increasingly more advantage to one vs. the other side of this struggle
--holding vs. slipping.

--dl*
====