Scotts locked Bowline (failure mode)

[agent_smith]

I started this new topic in response to a post by KnotLikely.
Reference link: https://igkt.net/sm/index.php?topic=6689.msg44495#msg44495

Alleged failure mode:

Scott's Lock just scared the crap out of me.  I was playing around with snags on the wall and I caught the collar (try sticking your thumb through the returning leg side).  My knot was suddenly 3 feet away from me and out of reach, with a hoop of eye legs big enough to crawl through.  If my (imaginary, thanks covid) belayer had slack in the line, this would take it all in to expand the tie-in loop.  This seems to be an issue with all single nipping loop bowlines that I have tested.

and this:

I snagged the collar and fell.  The knot did not fall apart, but the standing end pulled through like it was made of ice.  The knot was slightly looser than when initially set.

and this:

The extra friction introduced by the lock is not much in even slightly slicker rope.  Even in ideal conditions, I was unable to recreate this issue with my red (fuzzy) rope.  But it certainly Did happen.

and...

I'm surprised that this comes as a surprise to you.  ...  The extra friction introduced by the lock is not much in even slightly slicker rope.  Even in ideal conditions, I was unable to recreate this issue with my red (fuzzy) rope.  But it certainly Did happen.

In reply to refute to this alleged failure mode.
1. Scotts locked Bowline is an inherently secure Bowline.
2. It is intended to be tied in EN892 climbing rope
3. As with all tie-in knots to a harness, it must be tied diligently and accurately - in a tight and compact dressing state.
4. The size of the 'eye' of the Bowline must be small - not larger than 100mm diameter.

The alleged failure mode is entirely dependent on "snagging the collar".
KnotLikely asserts that he was able to "snag the collar" - which sets in motion a cascading series of events.

My contention is the 'test' conducted by KnotLikely is flawed.
In order to "snag the collar" - a very specific set of circumstances must be achieved.
As a thought experiment - even if such a test was carefully designed and could successfully "snag the collar" - this by itself does not mean catastrophic core failure leading to ejection from the climbing harness.

Any test which could 'successfully' snag the collar would be counter-balanced by the body weight of the climber. There would be a force pulling upwards which is counter-balanced by the weight of the climber pulling downwards.
Such a test would require pinpoint accuracy - to target the single rope diameter of the collar and lift it upwards from the core.

There needs to be a downwards trajectory - ie the climber would need to be free-falling (ie an uncontrolled fall) or be lowered (which is conducted slowly and deliberately).

I am of the view that this alleged failure mode is being overstated - and is conditional on the following:
1. Pinpoint accuracy to target the collar
2. Relaxation of the counter-balancing force of the climbers body weight (which acts in a downward direction).
3. A loose initial dressing state - so as to improve the odds of successfully snagging the collar during a fall or while being lowered.
4. The climbers body would have to come into very close proximity to the cliff (ie rock face) - implying a slab rather than a vertical or overhanging section of rock. If a slab, this suggests the climbers waist is scraping down the slab - feet pulled back.

I do not believe that a test could be designed in which others could reproduce the same failure mode.

I'm therefore calling this a false positive.

EDIT NOTE: Image added

[Dan_Lehman]

Mr. WordsWork, here,
and would like to see a sketch of the reported
failure / problem state.

I don't understand how snagging the collar
doesn't spill the knot but results in a large loop?!

... staying tuned ...

Meanwhile, I like the middle version  of A_S's trio
of versions, where the tail turns around 2dia (and
becomes TIB though that's irrelevant for tie-in).


--dl*
====

[agent_smith]

I don't understand how snagging the collar
doesn't spill the knot but results in a large loop?!

I think the first step is to design a test that can be reproduced by others.
See attached for a design proposal...

Note:
A problem will arise with the design of the protuberance.
The intent is to 'snag' and displace the collar 'upwards'.
There needs to be agreement of the shape of the protuberance - which will obviously have significant impact on the experimental results.
In the first instance, the 'snag' must successfully catch (ie 'snag') the collar.
Once snagged - it must remain in effective contact with the collar - so as to displace it.
While this is occurring, a continuous counterbalancing force must be maintained in the 'downward' direction (ie the climbers body weight).
I am of the view that 100kg would be a fair and reasonable weight (ie approx 1.0kN).

The tester would need to guide the collar to accurately snag the protuberance.
The intent is to simulate a climber who is free-falling or being lowered.
Free-fall is sudden shock loading to the collar.
A climber being lowered is a deliberate and slow process.

[SS369]

So, the conversation about the Scotts locked Bowline snagging and coming undone, failing, sounds like someone is trying very hard to discredit the simple lock. 

I have personally tried the tying of this in many different media and it has worked to lock the #1010 Bowline, and it hasn't failed.
This is my tie in and go to eye knot and I use it without a doubt multiple times during my week. It works and it is simple.

After reading the other post in the forum about the snagging and finding the knot some distance away from the particular climber it peaked my interest.

Using some BlueWaterII (7/16 inch), I tied the knot with a Sharpie (even tried it using a tapered spike) in the collar area between the SP and WE. I dressed and set it the same as I do normally whether climbing tie in or otherwise.

Putting my foot in the eye, then pulling up on the Sharpie/toggle using both hands, one on each side of the knot. Slow, hard pull - no untying.
Hard yanks, no untying.
Doing the same pulls with the SP in tension, no untying.

IMO, the reality of something inserting into a well dressed knot just boggles my mind. I can't visualize it and I am a climber.

Yes, the collar stretches open some, but no untying. The nip grips as it should.

I've repeated this "test" in accessory cord (6mm), paracord and twisted nylon rope, even slick poly throw cord, no untying.

I think if someone doesn't want to use it, doesn't like it for whatever reason, (real or not) then by all means don't. Please don't disparage it in a public forum as the truth according to you. Show some proof, not hypothetical imaginings..

SS

[Dan_Lehman]

I don't understand how snagging the collar
doesn't spill the knot but results in a large loop?!

I think the first step is to design a test that can be reproduced by others.
See attached for a design proposal...

Okay, thanks for the good quick images.
And the assertion is that such a snag
rather than spilling the knot
drew in more & MORE SPart so to enlarge
the eye hugely?!  --a reverse-direction slippage
to what Brion Toss showed for a dbl.BWL tied
in HMPE cord, where the eye collapsed.

Well, my, this would be mighty disconcerting
to a tied-in climber!

So, the conversation about the Scott's locked Bowline snagging and coming undone,
failing, sounds like someone is trying very hard to discredit the simple lock.

But, rather, I think it's more general than any
particular BWL : it would apply to most any
that have a simple nipping-loop base (vs. two
such loops or the water bwl and so on).
It stayed Scott-locked, just had a growing spurt!
And in the "it's not a bug, it's a feature" spin of
it all, it showed a way of semi-soft-belaying
oneself, surpassing they infamous Tarbuck h.!


--dl*
====

[SS369]

I don't understand how snagging the collar
doesn't spill the knot but results in a large loop?!

I think the first step is to design a test that can be reproduced by others.
See attached for a design proposal...

Okay, thanks for the good quick images.
And the assertion is that such a snag
rather than spilling the knot
drew in more & MORE SPart so to enlarge
the eye hugely?!  --a reverse-direction slippage
to what Brion Toss showed for a dbl.BWL tied
in HMPE cord, where the eye collapsed.

Well, my, this would be mighty disconcerting
to a tied-in climber!

So, the conversation about the Scott's locked Bowline snagging and coming undone,
failing, sounds like someone is trying very hard to discredit the simple lock.

But, rather, I think it's more general than any
particular BWL : it would apply to most any
that have a simple nipping-loop base (vs. two
such loops or the water bwl and so on).
It stayed Scott-locked, just had a growing spurt!
And in the "it's not a bug, it's a feature" spin of
it all, it showed a way of semi-soft-belaying
oneself, surpassing they infamous Tarbuck h.!


--dl*
====

Hello Dan.
Why are you combining quotes from two different authors in the same post without differentiating who wrote them?
You've quoted Agent Smith first and then the first line of my post in your input.

And I don't follow what you've written about my sentence, please elaborate. Is there some tongue in cheek in there?

SS

[KnotLikely]

Okay, I deleted those posts as agent_smith was right.  It was way off topic.  As for everyone demanding proof, I'm sorry that I don't have video of me falling.  I'm sure it would be amusing.

Scott, I was not disparaging the knot.  I was simply reporting what happened to me and that I thought it presented a possible danger that I had not considered until it happened.  Yes, it is probably a fringe case, but most accidents are.

The why:

I have a friend who is worried about knot size and stability on hard slab.  He explained that a tie-in that I had discussed with him was large enough to push him off a route that he had been trying (I think in Zion, but don't quote me) and he was looking for something almost as small as a bowline backed up with a double overhand on the returning eye-leg (what he said he switched to when his retraced 8 pushed him off the wall).  Scott's lock is one of the smallest knots and was the third I "tested" his scenario on (I was just messing around).

I simply posted my experience.  I'm extremely surprised that no one is able to reproduce it when trying to reproduce it, considering that it happened to me when I was just feeling what different knots felt like on slab when my foot slipped.  I've never climbed anything where a retraced figure 8 could push me away from the wall or even bother me.  That is all that I was doing.  I was not attempting to bring any knot to failure.

The how:  https://imgur.com/a/O4ZSFC6

This imgur album is as close as I can come to explaining what probably happened to me.

I was tied into the left pictured knot in agent_smith's picture.  I can not reproduce it on the middle pictured cowboy bowline version that also grabs the returning eye-leg.  I've never tied the cowboy bowline with the lock outside of the eye-legs until just now, to add it to my naming system text document.

The protuberance pictured in agent_smith's photo caught the opposite side of the collar as presented in his failure mode test picture.  The snag was limestone and was not nearly as pointy as presented.  It was basically the corner of a cube.  I may have fallen to the opposite side, allowing a better snag.  I don't know, as it took a total of maybe 0.5 seconds and was completely unexpected.

The snag would only pull in slack left by the belayer, though this is often enough to pull the knot out of reach, especially when clipping.  The collar was not significantly elongated by the time I landed (a few feet, and no injury).

I had already run my hips into the wall a couple of times and the collar was not as tight as when initially set.  Nothing had been done to it, intentionally, other than climbing a slab where my knot hit the wall a couple of times.

The first picture is the knot set as tightly as possible.  Even fully set, the simple bouncing of walking 10 feet allows some of the returning eye-leg to feed into the collar and give it the collar hole in the pictures.  The knot does not lock down into an immovable locked knot in my rope, though the tail stays put nicely, despite that.  The next picture is a little jiggle.  I'd guess that tiny hole or simply the width of the rope of the collar is all that could have been snagged.  I do not believe the knot would have opened further than the zoomed in picture by the time that I fell.

This is used rope with no core or sheath issues.  I do not know the brand or model, but it was originally purchased by a climbing gym before being given to me to practice tying with so I didn't have to pull out 60m or use the small nylon I had laying around.

My theory is that my position and the orientation of the knot to my harness allowed the ongoing eye-leg to receive a small tug before the returning eye-leg was set and added friction to the standing/nipping parts.  From how I usually feed my harness loops, this would mean that my leg loops pulled first, as I usually feed my harness bottom to top.  Once the standing end, the nipping loop and the ongoing eye-leg overcome the initial friction, there isn't much to stop it from feeding through, at least in this rope.

I do have trouble reproducing this in my stiffer, fuzzier rope.

Anyway, that's all I've got to say on the matter.  I just thought that since it happened, I should tell someone.

[roo]

Okay, I deleted those posts as agent_smith was right.  It was way off topic.  As for everyone demanding proof, I'm sorry that I don't have video of me falling.  I'm sure it would be amusing.

Scott, I was not disparaging the knot.  I was simply reporting what happened to me and that I thought it presented a possible danger that I had not considered until it happened.  Yes, it is probably a fringe case, but most accidents are.

The why:

I have a friend who is worried about knot size and stability on hard slab.  He explained that a tie-in that I had discussed with him was large enough to push him off a route that he had been trying (I think in Zion, but don't quote me) and he was looking for something almost as small as a bowline backed up with a double overhand on the returning eye-leg (what he said he switched to when his retraced 8 pushed him off the wall).  Scott's lock is one of the smallest knots and was the third I "tested" his scenario on (I was just messing around).

I simply posted my experience.  I'm extremely surprised that no one is able to reproduce it when trying to reproduce it, considering that it happened to me when I was just feeling what different knots felt like on slab when my foot slipped.  I've never climbed anything where a retraced figure 8 could push me away from the wall or even bother me.  That is all that I was doing.  I was not attempting to bring any knot to failure.

The how:  https://imgur.com/a/O4ZSFC6

This imgur album is as close as I can come to explaining what probably happened to me.

I was tied into the left pictured knot in agent_smith's picture.  I can not reproduce it on the middle pictured cowboy bowline version that also grabs the returning eye-leg.  I've never tied the cowboy bowline with the lock outside of the eye-legs until just now, to add it to my naming system text document.

The protuberance pictured in agent_smith's photo caught the opposite side of the collar as presented in his failure mode test picture.  The snag was limestone and was not nearly as pointy as presented.  It was basically the corner of a cube.  I may have fallen to the opposite side, allowing a better snag.  I don't know, as it took a total of maybe 0.5 seconds and was completely unexpected.

The snag would only pull in slack left by the belayer, though this is often enough to pull the knot out of reach, especially when clipping.  The collar was not significantly elongated by the time I landed (a few feet, and no injury).

I had already run my hips into the wall a couple of times and the collar was not as tight as when initially set.  Nothing had been done to it, intentionally, other than climbing a slab where my knot hit the wall a couple of times.

The first picture is the knot set as tightly as possible.  Even fully set, the simple bouncing of walking 10 feet allows some of the returning eye-leg to feed into the collar and give it the collar hole in the pictures.  The knot does not lock down into an immovable locked knot in my rope, though the tail stays put nicely, despite that.  The next picture is a little jiggle.  I'd guess that tiny hole or simply the width of the rope of the collar is all that could have been snagged.  I do not believe the knot would have opened further than the zoomed in picture by the time that I fell.

This is used rope with no core or sheath issues.  I do not know the brand or model, but it was originally purchased by a climbing gym before being given to me to practice tying with so I didn't have to pull out 60m or use the small nylon I had laying around.

My theory is that my position and the orientation of the knot to my harness allowed the ongoing eye-leg to receive a small tug before the returning eye-leg was set and added friction to the standing/nipping parts.  From how I usually feed my harness loops, this would mean that my leg loops pulled first, as I usually feed my harness bottom to top.  Once the standing end, the nipping loop and the ongoing eye-leg overcome the initial friction, there isn't much to stop it from feeding through, at least in this rope.

I do have trouble reproducing this in my stiffer, fuzzier rope.

Anyway, that's all I've got to say on the matter.  I just thought that since it happened, I should tell someone.

While I've been a fan of the Scott's Lock general slack shaking security, I have been wary of recommending it as it can easily be dressed improperly and become prone to assuming an unstable form. (cf. https://igkt.net/sm/index.php?topic=5995.msg40270#msg40270)

An accidental pull of the free end without loop loading can also cause the knot to capsize.

I haven't been following this discussion closely, but I think I was able to reproduce what you are talking about in snagging so thank you for bringing it up.  When that large, wide upper collar gets snagged, the clamping forces let loose and allow sliding.  I think if it's something that happened in real life, it's worth considering.

Upon revisiting the knot, I also see that the shiftiness of the form can allow a snag of the tail to capsize the knot while pulling the standing part.  I believe it is the same capsized form I first saw by accident while doing cyclical strain tests many years ago on the specimen with different degrees of snugness.  I think the collar snagging vulnerability you found is the greater risk with a more hazardous outcome.  Good find.  I'm glad you were not harmed.

P.S.  This seems like a thread that should be moved out of the Concepts Forum and into the Practical Knots Forum.

[Dan_Lehman]

But, rather, I think it's more general than any
particular BWL : it would apply to most any
that have a simple nipping-loop base (vs. two
such loops or the water bwl and so on).
It stayed Scott-locked, just had a growing spurt!
And in the "it's not a bug, it's a feature" spin of
it all, it showed a way of semi-soft-belaying
oneself, surpassing they infamous Tarbuck h.!


--dl*
====

Hello Dan.
Why are you combining quotes from two different authors in the same post without differentiating who wrote them?
You've quoted Agent Smith first and then the first line of my post in your input.

Because in this few-posts thread anyone considering
the points should have that clear enuff!

And I don't follow what you've written about my sentence, please elaborate. Is there some tongue in cheek in there?

Simply as I said, the slippage is of the SPart
and not any un"locking" of the knot --that the
hooked collar supposedly just arrested the knot
(nub) in place and then the loaded eye pulled
material through the knot on via the outgoing
eye leg (which leads to the SPart).

(And that all this movement likely had some
decelerating effect on the suspended load.)

--as contrasted with something that might've
pulled apart the lock.

--dl*
====

[agent_smith]

Scotts locked Bowline (failure mode)

In reply to all posters who subscribe to, and support the notional concept that Scotts locked Bowline has a potential vulnerability in that the collar could be 'snagged' which leads to catastrophic failure.

Answer: No - there is no such vulnerability and no such failure mode in real-world climbing conditions.

Assumptions:
1. That EN 892 dynamic rope was used.
2. That testing was carried out on a real natural rock surface while climbing.
3. That the tie-in knot (ie Scotts locked Bowline) was diligently dressed and cinched tight with an eye not exceeding 100mm).
4. That the tie-in knot was maintained under continuous body weight at all times during the test.

On Monday 04 May 2020, I conducted a series of tests while climbing on a real natural rock surface.
I climbed using a 'top rope' climbing system - using Scotts locked Bowline as my tie-in knot to my harness.
I made vigorous efforts to snag and disrupt the collar on the rock surface - each attempt was unsuccessful.
I tried scraping the front of my body down the rock surface in an effort to catch a snag and disrupt the collar - again unsuccessful.
I deliberately positioned the collar on a 'snag' and then tried to vigorously bounce on the snag to disrupt the collar - again unsuccessful.

I gave up after 30 minutes because I was becoming fatigued from the effort - and my belay person was growing impatient and bored.

I would also comment that Scotts locked Bowline is oriented so that the collar faces either to the left or right...it never faces directly toward the rock surface. This is due to the design of all climbing harnesses. Given this orientation - it is even harder to snag the collar. Despite this orientation, I deliberately rotated the knot in an effort to forcibly direct the collar on to the 'snag'.
All unsuccessful.

I would also comment that while the knot is under continuous full body weight, the knot core is under compression - and so it is very tight.

I therefore conclude that KnotLikely is deliberately doing 'something' which introduces bias into his test results.
I would surmise that he has deliberately loosened the knot to improve the odds of achieving a snag or, removed his body weight from the knot.
Else, he has precisely targeted a snag at an acute angle with some deliberate earlier loosening of the knot in order to disrupt the collar (which would invalidate the test).

Another issue is the notional concept of reproducibility.
I am of the view that other testers would not be able to reproduce KnotLikely's test result in the field (ie on a real rock surface).

[roo]

Scotts locked Bowline (failure mode)

In reply to all posters who subscribe to, and support the notional concept that Scotts locked Bowline has a potential vulnerability in that the collar could be 'snagged' which leads to catastrophic failure.

Answer: No - there is no such vulnerability and no such failure mode in real-world climbing conditions.

Assumptions:

4. That the tie-in knot was maintained under continuous body weight at all times during the test.

That's not a very realistic assumption.  In real-world use, knots are sometimes slack.

[KnotLikely]

Really, agent Smith?  All those words just to call me a liar?!

If you *conclude* that what happened to me didn't happen, then your conclusions are worthless.

Also, what is going on with you not just moving the goal posts, but creating new ones?  "Oh, he removed his body weight from the knot and it was deliberate!"  Yes, it is impossible to pull in standing line when the belayer is keeping you tight on top rope!  Imagine that.

I stated very clearly that belay slack is what will be pulled in to extend the eye-loop.  You can't keep your bodyweight on a knot that isn't weighted.  I have no idea what your issue is, but I'm done, here.  It happened.  I fell.  The eye-loop extended with slack.  Get over it.  I am.  Bye.

[agent_smith]

per KnotLikely:

Really, agent Smith?  All those words just to call me a liar?!

? Never called you a liar.
You should tender an apology for such a remark.

If you *conclude* that what happened to me didn't happen, then your conclusions are worthless.

Wrong.
Read more carefully before posting such comments.
Scotts locked Bowline is intended a climbing tie-in knot (right?).
Life critical applications is its scope of use.

Also, what is going on with you not just moving the goal posts, but creating new ones?

?  I report real world results under climbing conditions.
Scotts locked Bowline is intended as a tie-in knot for climbing (and can also be used to build anchors).
No goal posts have been shifted.
In the context of a tie-in knot - your test results cannot be duplicated in the field uder real-world conditions.

It is you who is skewing the goal posts with your test regime (not me).

Get over it.  I am.  Bye.

There is nothing to 'get over'.
I simply report the facts.
You appear to dislike the facts?

You might want to reconsider your attitude.

[agent_smith]

per roo:

That's not a very realistic assumption.  In real-world use, knots are sometimes slack.

Incorrect.
Tie-in knots to a climbing harness are life critical.
Climbers make sure that their tie-in knots are diligently dressed and cinched up tight (because their life depends on it).
For a tie-in knot to be 'sometimes slack' (implying loose) - indicates incompetence.
Incompetence can be a contributing factor to climbing accidents - but then you would have to attack an entire cohort of skills that climbers must possess.

I think your comment; "real world use" is referring to situations other than rock climbing/mountaineering.

The context of this failure mode debate is tie-in knots for climbing - specifically; Scotts locked Bowline.

I think this point is getting lost in translation.

Also, if a person was to call into question the reliability of Scotts locked Bowline - that person has the burden of proof.
That is, the onus of proof lies with the person making a claim that Scotts locked Bowline is not fit for purpose on account of a vulnerability triggered by snagging the collar.
And this is the heart of KnotLikely's proposition - he announced to the world via a public forum that Scotts locked Bowline has a vulnerability. Such an announcement would raise alarm in the minds of some climbers. Casual readers of IGKT forum posts (such as from KnotLikely) might develop concerns based on unrealistic test conditions to induce failure.

There are no liars in this debate.
The issue is lack of understanding of key fundamental factors to induce the purported failure mode under real-world climbing conditions.
Scotts locked Bowline's principal application is as a tie-in knot for climbing (it can also be used to build anchor systems).

Extraordinary claims require extraordinary evidence.
Link: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3114207/

[roo]

per roo:

That's not a very realistic assumption.  In real-world use, knots are sometimes slack.

Incorrect.
Tie-in knots to a climbing harness are life critical.
Climbers make sure that their tie-in knots are diligently dressed and cinched up tight (because their life depends on it).
For a tie-in knot to be 'sometimes slack' (implying loose) - indicates incompetence.

No, a "slack" or non-taut line (including the knot) indicates that it doesn't always follow your unrealistic assumption that:

"4. That the tie-in knot was maintained under continuous body weight at all times during the test."

...which you decided to snip out of the context of my reply.  It does not indicate that the knot is sloppily dressed. 

One wonders how you test knot security if you assume that knots are only allowed to exist in taut line.