'Soft Eye' in climbing anchor construction

Practical Knots
1

Big fan of all the discussions here, but (evidently) first time poster.
I would like your expert opinions on the ‘Soft Eye’ used in anchor construction as taught mainly in Europe.
Allow me to introduce the concept from the climbers perspective. When one climbs a route and comes upon the place where the anchor needs to be ‘build’, and finds there two bolts which appear to be bomber (as in extremely strong with zero doubt that they should fail), the method of anchor construction advised by climbing associations (e.g. Dutch, German, Swiss and Austrian) is the use of sling (circular webbing) in which a knot creates a ‘Soft Eye’ which is attached with a carabiner to a bolt. The remainder of the sling is attached to the second bolt and serves as a backup. See for examples http://www.alpenverein.de/chameleon/public/07ace84a-bc33-615a-62a3-151e7b8b859a/Standplatzbau-Juli-2012_19947.pdf (figures 17 and 19) or the dutch translations: http://www.nkbv.nl/fileupload/Kenniscentrum/Sportklimmen/StandplaatsNKBV2010_1.pdf (figures 2 and 3).

http://www.w-hillmer.de/stand-neu-02.jpg

The common practice teaches that an overhand knot in the sling works and is acceptable, but since it is hard to remove after loading, a bowline on a bight is preferred (which incidentally is also significantly stronger), referred to in dutch as ‘dubbele paalsteek’ or ‘bulin-knoop’ and in german as ‘doppelte palstek’ or ‘doppelte bulin’. Now standard anchor building practice taught all over the world uses acronyms such as ‘SERENE’, ‘ERNEST’ and ‘STRADS’, where in each instance the ‘R’ stands for redundant. If one then takes a look at the ‘Soft Eye’ method, one might question the redundancy of it. In case of a single overhand knot, there is only 1 strand of webbing material that forms the eye, hence, a tear in that single loop would mean total failure (and serious injury or death). I have been told that this danger is taken away by the bowline on a bight which provides redundancy by having 2 strands of webbing in the ‘eye’. However, on this forum it has been discussed that a bowline on a bight readily fails if only a single strand is loaded (I can’t find the link for where this was discussed, possibly here: http://igkt.net/sm/index.php?topic=19.0 ) which would mean that a tear of a single loop could still mean total failure. I was hoping someone could confirm or disprove this ‘feature’.

On a second note, would you agree that the eskimo variant of the BoaB is more appropriate for the application, since the loop will practically only be ring loaded. (http://igkt.net/sm/index.php?topic=4168.0)

Lastly, if you find that the BoaB is unsuitable when redundancy is a requirement, do you have a better alternative. Remember that due to the nature of the ‘sling’ (circular webbing) you do not have access to any tails, however, it does not necessarily mean that it has to be ‘on a bight’ since you do have easy access to the ‘other’ end of the sling. I’ll post a suggestion of my own for your consideration.

2

So for my own suggestion, I followed these criteria:
-2 strands in the loop
-rupture of a single strand does not compromise the second loop in any way
-loop should be biased towards ring-loading capacity
-smaller = better (least amount of material used for the knot)

I came up with a design based on the water-knot (ABOK 296), since this knot has been the knot of choice for years for the creation of a loop with webbing material.
Wiki:
https://upload.wikimedia.org/wikipedia/commons/thumb/b/bb/Water-knot-webbing-loose-ABOK-296.jpg/320px-Water-knot-webbing-loose-ABOK-296.jpg

A single loop version of this knot is also used as a not well known tie-in method for climbing under the name ‘competition knot’ or ‘ring-bend’ (https://www.youtube.com/watch?v=Au8KiUS4RCc and http://igkt.net/sm/index.php?topic=3747.0 and http://www.ukclimbing.com/forums/t.php?t=585172). To draw the parallel with the water-knot, I’d prefer to call this knot a water-loop. See example below.

The double version of this knot can easily be tied ‘on a bight’ and the loop sticking out of the final knot can be folded around to sit around the standing part of the loop. This knot might be dubbed ‘double water-loop on a bight’. See example below.

When the water-loop or competition knot is used as a tie-in for climbing the tail is left long enough that it can’t slip through when the knot is tightened (no serious slippage is observed when using this knot as a tie-in method). Might slippage occur in webbing material, which is usually slicker than rope, then the standing part of the rope is not the most secure location for the looped tail in the double water-loop on a bight (remember that I’m considering ring-loading to be the main use of the loop). If the looped tail can be formed around the loop itself it is fully secured, and while that cannot be achieved on a bight, it can be achieved using the ‘free end’ of a sling, which makes me name this loop a ‘double water-loop on a sling’, see example below. If you think there is a more appropriate name for all of these, please let me know. See also the example of the double water-loop on a sling tied in a sling.

3
a bowline on a bight is preferred (which incidentally is also significantly stronger)

Be cautious with alleged test results in relation to MBS yields of knots.

Most of the test claims made around the world are largely parroted or unverified by other testers.

Knot MBS yield (ie Strength) is irrelevant. You will not reach the load threshold required to cause a knot to yield. Even highlines (slacklines) do not break knots - and these are tensioned significantly.

Note that climbing tie-in knots do not fail - I am unaware of any mass reporting of knots breaking in lead climbing falls - because its a myth.

…

With specific regard to your questions:

I assume you are referring to a multi-pitch lead climbing route? Or are you only referring to single-pitch ‘sport’ routes where you lower off at the top of the pitch?

  1. Always best to cord - webbing/slings are not as robust.
  2. It is useful to carry a length of 6mm or 7mm diameter ‘accessory cord’ so you can build a ‘cordalette’ anchor.
  3. Keep the included angle at 60 degrees. 60 degrees is the optimum angle of separation. Note that angles less than 60 degrees also pose a risk - since any load shift left or right will cause one of the legs of the anchor to become slack. 60 degrees gives you some swing room…
  4. Nylon cord has better stretch/resiliency than other stiffer cords - including webbing - you want the material to be able to stretch to absorb some of the energy of a fall.

I have been experimenting with ‘Purcells’ in lead climbing belay anchors. The Purcells allow fine adjustments and will also yield a little to further assistant in energy absorption (refer to photo). Sterling USA make and sell 6.0mm diameter sewn Purcell loops - they are excellent.

You can also just use the accessory cord to build an anchor with an overhand knot (or F8)…refer photos.

To reinforce my points…

  1. I do not recommend the use of tape/webbing slings.
  2. I only recommend using accessory cord.
  3. Anchor system must be built so that there is ‘no single point of failure’.
  4. No single point of failure means that if any one point fails, it will not trigger catastrophic collapse of the entire anchor system.

With regard to #1080 Bowline on the Bight - there is a potential failure mode with the collar - if any part of the collar should fail, it could trigger catastrophic failure of the entire anchor system. The risk is small, but it is nevertheless present.

When rigging an anchor system using a ‘cordalette’ - by tying an overhand knot or F8 you effective isolate each leg of the anchor and therefore achieve the aim of ‘no single point of failure’.

EDIT: Added anchor photos showing ring at focal point…these is my personal favourites. The ring is permanently attached to the purcells - it is super easy and quick to rig belay anchors with this rig. The ring also ‘engineers out’ the need to tie any further knots to create a soft eye.

Mark Gommers

Anchor_Purcell_3point_Ring.JPG

4

Hey Mark,

I love your document on bowlines, great work!

Thanks for your input, I do believe however that you do not specifically answer my question.
Yes, I was referring to anchors for multi-pitches, and yes, I am very aware of other, possibly better, options for anchor building. And I would find that an extremely interesting topic by itself if we approach the accepted methods from a knotting perspective (climbing forums have plenty of general anchor building discussions). However, here I’m specifically interested in this method, because this has become a general method taught to hundreds or thousands of climbers (I have no specific numbers on how many climbers are taught multi-pitch climbing by the german/austrian/swiss alpine associations).

So let me address your first remark separately. I know often knot-strengths do not matter much for climbing, however, this is a different case. Dynema slings have become increasingly popular and it is noted by the manufacturers of such slings that knots reduce their strength much more than they do in Nylon slings (or rope for that matter), reductions of over 50% are noted. I agree fully with you that tie-in knots do not fail (at least I’ve only once heard of such an incident, but that was in a rope-swing accident with a 500 meter drop). And I have neither heard of incidents where this particular setup (using the ‘soft eye’ method) has resulted in a failure. Though it is this particular setup that receives the largest load if a lead makes a factor 2 fall when leaving for a pitch, and if the breaking strength is lower than that of the tie-in knot, rope, carabiner or bolt, it becomes the weakest link in the system. This is in my opinion unnecessary since it can be avoided by having 2 strands and a ‘stronger’ knot in the sling. Anyway, this was not my main issue with the setup.

My main issue is the redundancy of it all. If so many people are taught this method, I think it is important to take a critical approach and assess the knots involved. A single overhand does not offer any redundancy, and since slings are much easier cut or torn than carabiners are, I would say the anchor does not adhere to serene/ernest/strads. If a bowline on a bight would hold after a single strand is cut, I would say the two loops are redundant and therefore safe.
So I’m mainly looking for comments on this last issue, and a possible solution if one would want to keep using the ‘soft eye’.

Thanks

5

Off-topic. If this should be discussed somewhere else please let me know.
Very interesting setup using Purcell loops, haven’t seen that before. How would you judge the danger is of cross-loading on your biner? Do you use a steel biner there to prevent such issues?
So far my personal favorite for two anchor point setups is a tied off sliding x.

6
Thanks for your input, I do believe however that you do not specifically answer my question.

I did but in a indirect way.

So to be more clear - I updated my reply post.

To repeat:

  1. I do not recommend the use of tape/webbing to build an anchor system for a climbing belay.
  2. Accessory cord is the preferred material (6.0mm or 7.0mm diameter).
  3. I use and recommend a ‘cordalette’ type anchor construction.
  4. Check the photos above - by tying a simple overhand knot or F8, you effectively isolate each leg of the anchor which achieves ‘no single point of failure’.
  5. The isolation knot also provides some capacity to absorb the energy of a fall - as the knot compresses in a dynamic loading event, it absorbs energy.
  6. There is no fall that can generate enough force to reach the MBS yield point of a knot - even highlines (which are maximally tensioned) do not cause knots to fail.
  7. There is a potential failure mode with #1080 Bowline on the bight…the collar is the segment of the knot structure that is ‘vulnerable’. The risk is very small - but, nevertheless it is a residual risk. A simple overhand knot or F8 is effective and isolates each leg of the anchor (as already stated above).

Note: All sorts of fancy architecture has been advanced for building belay anchor systems for lead climbing (eg ‘sliding X’s’, etc). None of them really offer anything revolutionary. Equalization still remains problematic - and the various solutions offered only add layers of complexity. Best to keep things simple - avoid complexity - and try to achieve a 60 degree included angle.

Mark G

7

Thanks for the additional figures and clarification.

I see you don’t recommend using webbing, though clearly many climbing associations do. Just telling them not to do so won’t probably get anybody very far, but pointing out a redundancy issue with the method they teach new climbers might. So for the sake of the discussion, you might even consider this knot to be tied in regular rope and not meant for an anchor but just for a hypothetical situation where you want to ring-load a loop which has redundancy built in.

8
I see you don't recommend using webbing, [b]though clearly many climbing associations do[/b]. Just telling them [b]not to do so[/b] won't probably get [b]anybody very far[/b]

All truth passes through three stages. First, it is ridiculed. Second, it is violently opposed. Third, it is accepted as being self-evident.

    attributed to Arthur Schopenhauer (1788 - 1860)

A truth I have found with the climbing community in general is that many beliefs are parroted or copied from one person (or book) to the next without actually verifying the facts. For example, we can see a lot of material published by climbing knot book authors as being derived from second hand sources - without being verified as correct by the author (particularly MBS knot yield figures). You have to really dig deep to find new ways of thinking about old materials. Slings have many pitfalls - and this includes daisy-chain slings, and falls directly on to slings has been shown to generate very high forces.

One can look to the industrial rope access industry as an example where attempts have been made to test various systems and then publish robust codes of practice.
For example, you wont see many rope access operators using web/tape for their ‘lanyards’. They mostly use EN892 dynamic rope (single) to make their twin-tail lanyards which are integrated into their harness. And I applaud this industry for doing so.

I often see outdoor recreation operators using systems and techniques that simply wont stand up to serious analysis and testing. For example, on challenge ropes courses (eg high elements) - many Guides still use web/tape to make their twin-tail ‘lanyard’ so they can climb poles and move through ‘elements’. The smarter ones (who know better) make their twin-tail (double hook) lanyards from dynamic rope.

Mark G

9

This is just random deviation from the topic.
The use of material in lanyards by ‘outdoor recreation operators’ has nothing to do with this issue.
Except if you are proposing to ban all sorts of webbing from climbing, including those used in quickdraws.
I have not seen any tests showing that nylon slings are dangerous in constructing an anchor and I do not want to tell others not to use it just based on a personal preference. I’m not parroting any beliefs, I’m actually trying to critically look at a method.

10
The use of material in lanyards by 'outdoor recreation operators' has nothing to do with this issue.

It has everything to do with the topic!
Lanyards are exposed to high impact forces in a fall. Without a tear-web’ energy absorber - the lanyard material must absorb all of the fall energy.
Nylon rope/cord is better able to absorb energy than webbing/tape.

Except if you are proposing to ban all sorts of webbing from climbing, including those used in quick-draws.
Not so fast...nobody is suggesting or proposing to ban anything. Quick-draws are one component in a fall-arrest system. They are the links between the removable protection device (eg a wired nut or cam) / bolt and the dynamic climbing rope. The rope is the principal energy absorption material - not the quick-draw.
I have not seen any tests showing that nylon slings are dangerous in constructing an anchor
I would suggest that you might not be looking in the right places?

Here is one quick link to see how web/tape sling react to fall energy: http://dmmclimbing.com/knowledge/slings-at-anchors

Here is another link: http://www.fishproducts.com/tech/High_Strength_Cord.pdf

There are many other links…

EDIT: Here is another reasonably good report: http://www.bethandevans.com/pdf/lanyard_tests_v6.pdf
it is worth reading!

…

Keep in mind that I am providing you with a detailed reply - and in good faith. How you perceive this information is entirely up to you.

Mark G

11

Thank you for your effort Mark. I do appreciate it as well as your care for safe practice in climbing.

Let me please one more time explain the question and present it in a manner which has nothing to do with climbing so we can avoid answers to questions I’m not trying to find answer to at the moment.

I’m purely interested in the knot/knots, that’s why I came to this forum. I wanted to know whether the integrity of a double bowline on a bight, when ring loaded only, would severely suffer from the rupture/tear/cut of a single strand. I understand it matters which material is used, and I should not have given any background before, since I do not seek advice in anchor building, I’m just interested in the specific question. So for arguments sake, lets say we are using regular rope

Mark indicated:

Does anybody have a definitive answer to this.

12
present it in a manner which has nothing to do with climbing so we can avoid answers to questions I'm not trying to find answer to

Interesting… the whole premise of your original question was clearly in relation to anchor systems used in recreational lead climbing.
Context is key when discussing knots and the material that knots are tied in.
You cannot escape context when discussing knots and their applications.

I wanted to know whether the integrity of a double bowline on a bight, when ring loaded only, would severely suffer from the rupture/tear/cut of a single strand
And the answer had been given already. #1080 Bowline on a Bight does have a vulnerability - and it is with its 'collar'. The risk is small but, it is a residual risk. The loss of one leg of the anchor will not trigger catastrophic failure. However, a rupture of any segment of the collar will result in catastrophic failure.
I do not seek advice in anchor building
And yet the whole premise behind your questioning is indeed anchor systems. You may not like the answers you are receiving - and so this may be a source of potential irritation to you - which you now seek to avoid.

…

The video link you gave is a good example of incorrect information.
The presenter refers to the knot as a ‘ring bend’. This is incorrect.
It is an eye knot.

A ‘ring bend’ can be found in ABoK at entry #1412 (and also at #296).

A bend is an entirely different structure - where 2 separate rope ends are united.
The presented also tries to provide supporting comments that ‘it is dead easy to tie’ - and; ‘useful all-round knot’… but nothing of any further substance is given beyond these limited remarks.

A point of fact is that Bowlines can be tied in a one-stage tying process. The eye knot in the video cannot be tied in a one stage process - it is not PET (post eye tiable).
This fact weakens the presenters case that his eye knot is easy to tie.. Furthermore, the concept of ‘usefulness’ is not substantiated to any degree. Bowlines are jam resistant on account of their nipping loop - and this property is useful. Bowlines are PET - and that property also makes them useful.

The loading profile of the core of the eye knot is significantly offset but, since both legs of the eye contribute tension force, it is unclear as to how well the structure will resist jamming in some of the newer sub 9.0mm EN892 dynamic ropes.

Does anybody have a definitive answer to this.
Keeping mind the crucial importance of context - it is hard to build a substantial case to support webbing/tape in lieu of accessory cord. Given that accessory cord is cheap, lightweight and readily available, there is no reason in principle why webbing/tape should be used in preference. Also, many independent testers have demonstrated the advantage of cord over webbing/tape to absorb energy. Given that the premise behind your original post was 'belay anchor systems' - it would seem illogical to avoid using accessory cord to build a solid and reliable anchor system (with no single point of failure). Given the ease with which a 'cordalette' anchor can be built - and the redundancy it offers (ie the 2 legs of the cordalette can be easily and quickly isolated with a simple overhand knot), it would appear to be the best overall choice. Cord also has a thicker cross-section compared to webbing - and so it is inherently more robust. This is one of the reasons why climbers use a rope - and not flat webbing - as their primary fall-arrest element. Imagine lead climbing with flat webbing instead of rope - and taking a fall...

#1080 Bowline on the Bight was never intended to be tied in flat webbing/tape material. The Bowline was intended to be tied in rope (a round cross-section material). That said, it is possible to tie knots in flat webbing - however the behavior of the knot under loading events will be unpredictable. Knot testing is typically done with rope/cord, not flat webbing material…and so it is extremely difficult to find any reliable test data. My point here is, why step outside of proven methods? What is the real reason behind wanting to use flat webbing/tape instead of cord?

In one of the links you provided on your original post: http://www.alpenverein.de/chameleon/public/07ace84a-bc33-615a-62a3-151e7b8b859a/Standplatzbau-Juli-2012_19947.pdf
Figure 25 (page 21) - left image… this is the preferred rigging setup using cord.
The image at right shows webbing/tape… which is not recommended for the reasons stated - chief amongst which is that cord is a better material for absorbing energy than flat webbing/tape.
One has to keep in mind that in multi-pitch lead climbing situations, there is a risk of a factor 2 fall right off the belay if the lead climber does not place the first piece of protection early and actually clip the rope into it (or that protection could pop and hence fail). This will result in a fall directly onto the belay - with severe impact forces. Cord will offer a better margin of safety than webbing/tape - because it will absorb some of the energy and you might live to climb another day.

There are several important reasons why you just want to create one main soft eye attachment point. The most compelling reasons are:

  1. It simplifies rescue - note that you should always rig for rescue (in other words, you should always assume the worst and be prepared).
  2. Building a cordalette anchor keeps your anchor rigging separated from your climbing rope - this ensures that you have the maximum length of rope available for the task at hand.

The soft eye created by an F8 or a simple overhand knot effectively isolates each leg of the anchor system thereby creating redundancy (no single point of failure).

Is there a compelling reason not to use cord?

Mark G

13

A Pile Hitch is simple, easy, and tough to get wrong:

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

You’ll have to excuse me if I don’t go over the whole thread. I have to go. Cheers.

14
A Pile Hitch is simple, easy, and tough to get wrong:

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

Indeed it is simple. But unfortunately it is not relevant in the context of multi-pitch lead climbing.
As with all knots, one must have a clear understanding of the context in which they will be applied.

The belay anchor system will be: A configuration where 2 (or ideally 3) anchor points are interlinked, and a fixed soft eye is created as a singular main attachment point.
The entire system will be aligned with respect to the fall risk - with load distributed evenly (equalised) across each of the individual anchor points (note that it is hard to impossible to achieve perfect equalisation).
There will be no bollards or posts on the cliff - just cracks in which removable protection devices can be inserted (eg wired nuts, cams, etc) or fixed bolts.

15

Hi Roo,

thanks for the suggestion, but I was interested in loops not hitches.

I’m was looking for a 2 stranded loop knot of which the cut/rip/tear of a single loop does not affect the integrity of the knot.

The interest was purely from a theoretical point of view, and I was just interested in knots.
I gave a background of how this interest arose but should not have, since I’m not interested in climbing advice, but just in the knot expertise on this forum.

16
I gave a background of how this interest arose but should not have, ... [just interested] in the knot expertise on this forum
Some knot experts on this forum may find this remark possibly derogatory.

Roo for example is a fine fellow and extremely knowledgeable about knots. If you stop and think about this for a moment - and look at the response Roo gave - you can see that he needs to understand your context. He gave you the example of a pile hitch - and this is indeed a simple structure. However, it is not relevant to your needs - it is in the wrong context.

You will find most of the knotting experts on this forum will need some context so they can properly analyze the question and then provide a considered response.

I'm not interested in climbing advice
A smart man once told me that the day you die is the day you stop learning.

Being ignorant is not so much a shame, as being unwilling to learn.
Benjamin Franklin

You question concerning #1080 Bowline on the Bight has already been answered in depth.

What you need is a knot that will isolate each leg of the belay anchor system. The idea being that if any single anchor point fails, it will not trigger a catastrophic failure of the entire anchor system. Knots operate best when the material is rope (not webbing - apart from the ‘tape knot’ #1412). Although it is possible to tie common knots using webbing, the performance of such knots will be unpredictable. Very little test data is available on knots that are intended for rope which are instead tied in webbing. Dan Lehman may have something to say about this matter.

The isolating knot must create a ‘fixed eye’. This eye will serve as the focal point of the anchor system - and this is where everyone will attach to.
It also permits rescues to be performed (keeping in mind that whenever you rig a belay anchor system - you should always rig for rescue).

With your insistence on using webbing/tape - this is of course being valid in situations where you have no alternative (ie - you have no accessory cord) - reasons might be; you forgot to carry cord, or you dropped it, or you lost it. The ability to improvise is important in recreational lead climbing. But that is a caveat…in that cord should be your first choice - and you only resort to improvising if faced with no other alternative. Given the low cost of accessory cord, its low weight, ability to be carried in compact form, and its ability to absorb energy, makes it hard to ignore as a first choice material for building belay anchor systems.

For me personally, another compelling reason to use cord is that it w i d e n s the scope of possible knots and hitches that can be tied - in other words, you have more choices available. With webbing/tape, it actually narrows your choices.

Dan Lehman may want to weigh in on this discussion - I believe he has climbing experience so he will understand the context very well. Again - context is crucial to providing advice that is relevant.

It is possible that you may dislike many of my comments I have tendered…

Mark G

17

There are carabiners. The OP’s picture shows his loop tied into a carbiner. A Pile Hitch can replace that loop that he shows in his first picture.

I would like to see the picture a little bigger so I don’t have to squint.

18

There are some, but I wonder if even in your theoretical search, there are some-real world limitations or application parameters that may narrow the search.

19

A close-up of the setup as proposed by the german national climbing association (DAV) below. They intend to use it only in situations where there are 2 bomber bolts.

one limitation would be no access to tail ends. a preference would be ease of removing the knot afterwards.

close-up.jpg

20

Hi all!
(Here comes knotsaver 8) :wink: )

First of all, I agree with agent_smith about the importance of the context,
second, I haven’t climbing experience, so I can’t provide climbing advice…

(IIRC)

  • What about a 2 strands standard bowline ( or better ABoK 1034 1/2 for ring loading?) tied with a bight?
    you could use the bight as a third eye, or, if you don’t like it, you could “back-flip” the bight (you can choose the side and the position).

  • What about ABoK #1085? (but here the back-flipped bight could have the same problem of the collar in the BoaB (ABoK#1080), as noticed above)

I’m only in doubt about the security of the knots if only a single SPart-strand is loaded… :-\

hope this helps,

ciao,
s.