Knot jamming test

[Harold Kahl]

I tested a number of bend and loop knots for jamming and thought I would share the results here. I used 3/16 inch braided polypropylene rope rated at 60 lb working load limit. I cut the rope into short pieces and tied a series of knots, loaded the rope to 180 lb. then attempted to untie them. Here are the results.

Knots that could be untied without tools:
bowline required 15 seconds to untie, bowline on bight, bunny ears, yosemite bowline, about 30 seconds, zeppelin loop 40 seconds and zeppelin bend about a minute.
The bunny ears (double figure was a bit of a pleasant surprise because the regular figure 8 is not a jam resistant knot.

Knots that required a tool (pliers) to untie:
butterfly loop (eye loaded), butterfly bend, and butterfly bend loop all required about a minute to untie with tool. The eye loaded butterfly loop was actually a little quicker to untie than the bend, contrary to what I expected. The Hunter's bend took me over 2 minutes to untie, and I almost gave up on it.

Knots that were jammed and I could not untie: Directional figure 8, double dragon, double fisherman, double harness bend, double uni, edk, eskimo bowline, figure 8 loop, figure 9 loop, honda, perfection loop, reever bend, sheet bend, simple simon over.
The eskimo bowline and sheet bend were surprising to me as I expected them to be jam resistant based on similarity to the bowline. Figure 9 didn't seem to have any advantage over Figure 8. Double dragon was a mild surprise since I have seen it touted as jam resistant.

This appears to have been a pretty harsh test, so some of the jammed knots may well have survived less extreme loading.

[Dan_Lehman]

Harold, thanks for the testing.

It would help immensely to be able to see your test knots
before and after loading.  There are many ambiguities in
the knots you list (e.g., the butterfly is asymmetric, so
it matters which end is the SPart; and its dressing is most
often shown in a way that is inferior to resisting jamming
(to what e.g. Wright & Magowan presented, with the eye
legs crossed within the knot).

I'm surprised irrespective of knot-geometry clarity about
the zeppelin end-2-end joint, as there the versions IMO
all should yield a non-jamming knot!  And, like you, also
for the Eskimo bowline --and the butterfly end-2-end
vs. eye knot (though, again, what version/dressing?!).

This is the sort of "backyard" testing that we all should
be more engaged in performing; we likely all have some
rope for which ease of untying can be lost with a wrong
knot and some serious loading, which can be generated
with 100# or so, and maybe some dynamic application
(e.g., body weight amplified by standing in the end of
a 2:1 pulley arrangement --even inferior ones w/carabiner
for the block-- on 9mm rope.


--dl*
====

[Harold Kahl]

Thanks, Dan. I made some video of the knots being untied, but I don't know if they would help answer your questions. Not sure what you are asking in regard to the zeppelin, as both the bend and loop were untied without tools. I think I will redo the alpine butterfly tests for you, and post either pics or video. It's easy enough to do.

[Harold Kahl]

I redid the test for the Alpine butterfly knots and the Eskimo bowline. This time, the butterfly bend and butterfly bend loop came untied without tools. The eye loaded butterfly loop jammed as did the Eskimo bowline. I'm not sure what the difference was compared to the last test.
Tying video:
https://youtu.be/gyxdacyqqBw

Untying video:
https://youtu.be/_ts1UNx86FM

[Dan_Lehman]

Not sure what you are asking in regard to the zeppelin, as both the bend and loop were untied without tools.

Mainly, I was just expressing surprise that the ends joint
wasn't easily untied --I might've guessed that it would be
the easiest (tho' there are various factors ...) !

(IMO, Ashley's #1408 is most like the zeppelin
in should-be-non-jamming.)

(-;

[Harold Kahl]

Mainly, I was just expressing surprise that the ends joint
wasn't easily untied --I might've guessed that it would be
the easiest (tho' there are various factors ...) !

The zeppelin loop was at the end of a string of knots, so it was probably looped over the hook of the chain hoist or whatever it was looped over at the other end. This may have kept the knot from bunching up as tight as it might have otherwise. Just another variable to consider.

[Dan_Lehman]

I redid the test for the Alpine butterfly knots and the Eskimo bowline. This time, the butterfly bend and butterfly bend loop came untied without tools.
...
Tying video:
https://youtu.be/gyxdacyqqBw

As I see it, the butterfly end-2-end joint is tied in the
way recommended by Wright & Magowan.  I try
to remember this as the pretzel form'd overhand's
SPart bears against its own tail, and so --tails crossing
within knot-- the opposite end (timber hitch form)
doesn't bear directly upon its tail (but also upon
the other tail).

[Z]

1 minute to untie a Zeppelin??? I have loaded a Zeppelin harder than that in the same material. I got the knot loose in about 5 seconds.

Double Dragon has a trick to untie that practically makes it impossible to jam. To untie, you ring load it strongly. It should be obvious to identify which strand has no choice but to loosen. Then, you go from there.

By the way, I want to note that jamming is not always undesirable. I had an application just yesterday where I wanted only jamming knots throughout. And, of course, I want a fishing knot to jam.

[Harold Kahl]

I retested some of the knots where previous results were unexpected or inconsistent. This time, the bunny ears jammed, as I had originally expected, though it came untied rather easily in the first test. The end-to-end butterfly came untied in a little under a minute, though the butterfly bend was jammed. Not sure why this would be different because it is the same knot if you join the tails of the butterfly bend. Likewise, the butterfly bend loop was also jammed as was the eye loaded butterfly. The zeppelin bend (2 examples) and zeppelin loop all came untied in around 45 seconds or so. Two examples of Hunter's bend were jammed. Other than the bowline, the zeppelin bend (including zep loop) is the only knot I've tested that could consistently be untied after loading.
Video
https://youtu.be/ZUISW-t0Diw

[Z]

I watched your video, thanks.  Regarding the times to get unjammed, the time comparisons would be more useful if you to stop the clock the moment you get a strand loose in the knot.  At the moment you get a strand loose, it is pretty much guaranteed the rest of the know can be untied. So, there is no need to continue timing after you get a strand loose.

The knot complexity or the particular knot's dressing may cause the remainder of the untying to take longer or shorter.  Including that extra time (beyond the loose first strand) makes the time comparisons mostly useless. So, the untie times you have posted here are not that useful (to me). However, again, thanks for your video.

[Harold Kahl]

Z, I agree, you shouldn't read too much into the time to untie the knots in these tests. There are just too many variables like how strong one's fingers are, the skill of the person doing the untying, etc. and it isn't even consistent from one trial to another. My objective was to compare the knots relative to one another, and I feel like I learned a lot. Mainly, if I ever need to pull a car out of a ditch with a rope, I'll be using bowlines, zeppelin loops, and zeppelin bends. I learned to avoid sheet bends and eskimo bowlines in situations where jamming may occur. As far as hitches are concerned, that may be a subject for
future testing.

As to ending the timing after getting a strand loose, I'm not so sure I agree. There were some knots where one part came loose right away but the rest of the knot was still jammed.

As a footnote to the last test, I later worked at the jammed knots using tools and was able to get them all untied eventually except for the eye loaded butterfly.

 

[agent_smith]

I think we need to keep in mind that all knot testing generally falls into one of the following categories:
1. Backyard testing
2. Pseudo lab testing
3. Certified testing done in an accredited test lab (which is typically a purpose built facility with expensive calibrated equipment).

Expectations of quality, repeatability and statistical accuracy scale accordingly.

When thought of in this manner, some of the criticisms arise from having too high expectations.
For example, NautiKnots strongly argues for statistically valid data points - which requires some knowledge of math and reasonably accurate tensile force generating equipment to capture data.

This may be placing too high an expectation on the 'backyard' style testers.

Harold is certainly a 'backyard' tester - and so he doesn't have access to thousands of dollars worth of calibrated force generating equipment and sophisticated software to sample and capture data points. And, he likely isn't an expert in statistical mathematics.

Be that as it may, 'backyard' testers can still make a valuable contribution and should be able to follow some simple rules.

Some simple rules:
1. Take reasonable quality photos/video
2. Use a 'control'
3. Have a specific objective in mind
4. Dont just follow the default and mind numbing 'pull-to-failure' line of thinking (but if you do, give valid reasons why)
5. Use ABoK reference numbers (where they exist) to aid in positive identification of knots
6. Report on the type of cord/rope material used in the test - including its diameter (ie if testing an end-to-end joining knot, report whether equal rope diameters were used or unequal diameters).
7. Write a conclusion that summarizes your findings.

Good to see that Harold was investigating something other than the default and mind numbing pull-it-till-it-breaks mentality. He investigated vulnerability to jamming. So well done

Z's comment re untying was not entirely correct. Harold correctly pointed out that in some cases, he could loosen one rope segment only, but, the remaining structure was jammed.
As long as Harold is bench-marking against a 'control' - his timings should be valid. I think this is where NautiKnots could chime in with advising what type of control would be valid for Harold to measure against.

I would like to see Harold perform a test of the #1425A derived Riggers X bend...using #1425A Riggers bend as a control.
Is #1425A Riggers X bend resistant to jamming?
And, if it does have vulnerability to jamming, what is the load threshold at which jamming is triggered?
Refer attached photo.

[Mobius]

The "#1425A derived Riggers X bend" is one I found for myself several years ago. I liked it enough I did trials in various rope materials (not dynamic rope at the time). I never trialled/documented it properly then (even from a home-tester perspective) and the material I used in the display was once described (unfairly) as "grandma's thread" With those things in mind, take what you can from the link below if you wish.

https://docs.google.com/presentation/d/1D_hVQSps-IBUx_TJ8y5mWkw7mjueCoOa4dQWkz4mmo8/edit?usp=sharing

I have started looking at this bend again, looking for something that can be adapted to work in Dyneema. I am not happy with anything yet.

[agent_smith]

per Mobius:

The "#1425A derived Riggers X bend" is one I found for myself several years ago.

The term 'Riggers X bend' (derived from #1425A) is one that Xarax I think prefers...and I believe that he was first to investigate the effect of crossing the tails in some known bends (I'll check with him...no doubt knotsaver will also confirm/refute shortly). Although I think Dan Lehman had also been aware of this 'X' modification for some time too...

So I think your naming the structure 'infinity bend' may be in the same situation as your alleged 'M' bend...which Xarax also predates (did you check the link he gave?).

I think 'Riggers X bend' is more in tune with its structure and heritage - in that it is simply a transposition of the tails within the structure.

Good to see your prior work on the Riggers X bend - and thank you.

...

I am hoping that Harold could re-examine the Riggers X bend and conduct further homebrew style testing against #1425A Riggers bend as the control. Maybe also look at the load threshold which triggers jamming in #1425A? And is the Riggers X bend immune from jamming? More data points would be nice so we can build a case and reach a solid conclusion.

[Harold Kahl]

Thanks for the suggestion, Agent Smith. I'll do that test soon. For some reason I like that whole series of interlocking overhand knots, and now I have another one to learn. I also want to test the Kalmyk loop (slipped version of the Cossack knot) and also some hitches.

My testing is definitely in the class of backyard tests, but I find it interesting and useful anyway. I apologize for the quality of the videos. I have gained a lot of respect for people who can make good videos. it's harder than it looks.