Break testing 3 knots

[moray]

By now nearly everyone knows that knots weaken a rope, often severely. In those ropes that can be spliced, a splice tends to be superior to any knot, preserving 90% to 100 % of the native rope strength. A completely undistorted rope is strongest; any distortion or bend in the rope will weaken it. Knots involve very sharp bends (which get worse under load), so it is no surprise they weaken the rope. Splices, in contrast, have only very gradual bends. There is a significant disturbance to a spliced rope where the buried core penetrates the cover, but the resulting weakness is mostly or completely nullified by the fact that the disturbed part of the rope is only carrying half the load; the undisturbed buried leg of the splice is almost perfectly straight and nearly full strength.

Now that I have a rig for break-testing ropes and splices, I have not rushed to test a bunch of different knots because knots are much more complicated than splices. Whereas I can design meaningful experiments with splices, it would be much harder to do that with knots as there are so many more variables. You don't have to tie, dress, and set a splice. Also, splices keep their shape, whereas knots slip and contract and significantly change shape as the load is increased.

Nevertheless, having just received a new batch of rope for testing (5/16 in. Tenex Tec), I decided to test 3 knots while also determining a baseline for native rope strength.

For the baseline tests I made two slings: one with two normal eyes, and one with one normal eye and a second eye protected with a locked Brummel.

Three knotted slings were tested. The first had two normal eyes and an overhand knot in the middle. The next sling had one normal eye and a figure 8 on a bight aka rethreaded figure 8 aka figure 8 follow through. This is a big favorite with rock climbers for their harness attachment.
The last sling had one normal eye and a directional figure 8. This has been a favorite of mine for anchoring my SRT line to the base of the tree. It is easier to tie and easier to explain than an alpine butterfly, my other standard anchor loop.

 

[moray]

Results

The picture shows all the slings prior to testing.

Here are the breaking strengths, in pounds:

Baseline tests:
Normal eyes--- 4524
Brummel eye--- 4596

Knot tests:
Overhand------ 1676
Directional 8- 1840
8 on a bight-- 2100

The Brummel, as expected, had no effect on strength. The two baseline tests (no knots) give an average rope strength of about 4550 lbs.

The overhand knot was extremely weak, about 37% of rope strength, and the others were not much better.

The picture shows the 3 knotted ropes after breaking.

While it is not apparent in the picture, the knots, or at least parts of them, became extremely tight and constricted under load. They all broke where the rope entered the knot, and none of them severed cleanly as a splice will do. As I said above, knots involve a lot of variables, rope type being just one. Perhaps the very same knots tested in another rope would give different results.

 

[scotclayshooter]

Great work again!
I didnt realise that knots would make that much weakness!

 

[Philbert]

Great post. Interesting and informative.

Tried to rep you for it, but it says that you are still in my rep holding pattern . . .

Philbert

 

[Nailsbeats]

I agree on the 8mm Beeline, I would like to know that one with a double fishermans on it. Great work!

 

[outofmytree]

Moray......that's just way to simple......awesome idea. How about a teaser of 5/16 beeline, inquiring minds wanna know...lolopcorn:

Moray could you please please please test a dfl termination and dfl stopper with this rig. I will gladly contribute to your costs in whatever way you request.

This work is outstanding in its benefit to climbers. It is one thing to say you "think" a knot weakens a rope, it is another to supply proven evidence using pre tested equipment.

Repped for this great effort.

 

[moray]

A number of people have asked about double fisherman's and Beeline, which seems to be all the rage these days. I am working on an upgrade to my testing rig so I can test non-spliceable ropes and determine full rope strength. This will also allow me to test some spliceable double braids without forcing me to make TWO splices for every test.

One of the members here on AS has offered to send me some Beeline for testing, and I look forward to working with that as soon as I get it. Stay tuned...

 

[Philbert]

Love to see photos of the test rig too, if possible.

Thanks.

Philbert

 

[scotclayshooter]

What about rope thats been wet with petrol or other chemicals?

 

[pdqdl]

Once more, Moray, this is great stuff.

Here is a thought: rather than testing with splices on one end, you could make comparison pulls with two different termination knots.

I would love to see bowline vs 8 on a bight. Then you know for sure which knot was stronger.

Perhaps double fisherman vs an anchor hitch? Oops. That would require introduction of a chunk of steel to the system...better make sure your safeties are hanging on real good.

I would REALLY, REALLY like to see a 1/2 hitch holding a timber hitch vs a running bowline, but then you would have to break the rope while it was attached to a tree. Not sure how well that would work.

 

[moray]

...Here is a thought: rather than testing with splices on one end, you could make comparison pulls with two different termination knots.

I would love to see bowline vs 8 on a bight. Then you know for sure which knot was stronger...

This is true, you could just compare two knots directly and skip the heavy work of making a tough splice. But here's the problem. If you were to measure two knots, in separate experiments, at 3000 and 3500 lbs. break strength, this means one thing if the actual rope breaks at 3600 lbs. but another thing altogether if the rope breaks at 10,000 lbs. The straight comparison isn't meaningless, but knowing the rope strength completes the picture.

But there is a way out, and it needs a couple of chunks of steel. With some heavy pipe and a pair of heavy shackles I intend to make a couple of bollards I can fit to my chains. Then I can terminate ropes with several wraps around the bollards instead of a splice.

 

[fsfcks]

The Alpine Butterfly (or Lineman's Knot) is meant to be one of the simplest and strongest knots, supposedly retaining about 80% of the rope's original strength according to some other tests I've seen on the web. The load can be safely applied: from the loop to either end of the rope; between the two ends with the loop hanging free; or to the loop with the load spread between the two ends.
http://www.animatedknots.com/alpinebutterfly/index.php

Moray - I wonder if your tests would confirm this?

 

[moray]

I really like the Alpine Butterfly and have used it a lot. I am about to do a batch of knot tests, and the Butterfly is high on the list, along with bowline, double fisherman's, and several others. Dollar signs swarm in front of my eyes when I think of all the rope I'm about to destroy.

 

[pdqdl]

Well then post your mailing address, and see if some money doesn't show up!

Shucks, if everybody that read your splicing and testing threads sent you $1.00, you could probably retire.

Would the moderators step in and stop that?

 

[pdqdl]

With some heavy pipe and a pair of heavy shackles I intend to make a couple of bollards I can fit to my chains. Then I can terminate ropes with several wraps around the bollards instead of a splice.

You might increase your safety and speed of testing by digging a 4' deep hole and setting a 6" steel post in it. Then it is it's own bollard, and no flying chains or other metal is involved. You might set two of them for gateposts or some other genuinely useful purpose, then use them occasionally as an anchor point for your tests.

You're up in Maine aren't you? Do you ever need a sturdy post to anchor a winch for pulling yourself out of the snow?

 

[moray]

You guys are very funny.opcorn:

As for a bollard design for my floating cylinder pulling rig, I took some drawings and big shackles over to show them to my friend the local arborist. This guy is very old-time, but he has shown a lot of interest in my various schemes and loves to argue with me. After patiently listening to my design ideas and looking at my drawings, he went to the back room and returned with two brand new pulling bollards he had welded up for me! And painted bright red to boot. They were nearly identical to one of my own designs, and I can't wait to try them out. Tomorrow!

To get out of the snow in Maine no one ever uses a post anchor. It is impossible in Maine no matter where you go to be more than 10 feet from a tree.

 

[046]

how's about posting a paypal address? much easier than mailing...

I really like the Alpine Butterfly and have used it a lot. I am about to do a batch of knot tests, and the Butterfly is high on the list, along with bowline, double fisherman's, and several others. Dollar signs swarm in front of my eyes when I think of all the rope I'm about to destroy.

 

[eljefe]

Re: Break testing knots

Hi moray,
Your information is very very good, very very useful. And more than a bit scary. I very much hope you continue to test various knots and ropes and publish the data here. Thank you very much for taking your time and money to do this. I agree we should all chip in for expenses.
In all the years, in all the books, etc I've heard this knot or that knot maintains alot of the rope strength. I would not call less than 40% of the native rope strength alot.
it does get one to wondering just how much the type of rope -3 lay, 12 strand, etc or what it is made from manila, polyester, nylon, etc plays in the strength of a particular knot.
eljefe

 

[moray]

... it does get one to wondering just how much the type of rope -3 lay, 12 strand, etc or what it is made from manila, polyester, nylon, etc plays in the strength of a particular knot.
eljefe

Absolutely! It's going to be quite a job trying to untangle all these variables to come up with some useful general rules.

When I break tested a double braid splice for the first time a few days ago, I discovered there had been fiber melting right at the tip of the bury. Everywhere else the fibers seemed normal. There must have been relative motion of the fibers at that location under very high tension, and the heat of friction was enough to cause local melting. Your comments about different rope types got me thinking about knot tests in general, and it occurred to me that knots should be far worse than splices when it comes to relative motion, and thus friction heat. The knot shrinks, changes shape, and can even roll out under heavy tension, whereas a splice pretty much just sits there.

So I went back to my bucket of broken ropes and found the test rope with the overhand knot. Sure enough, under the dissecting microscope you could see lots and lots of melting in the region of the knot. An inch or so from the knot, all the fibers seemed normal. The picture was taken through the eyepiece of the scope, and the arrow points to some molten globs at the ends of fibers. Sometimes I saw several droplets arranged along the length of a fiber like dewdrops on a spider web. Sometimes bundles of hundreds of fibers were glued together and resembled a paint brush in which the paint had dried. All this melting at only 1700 lbs. tension! Wait till we start breaking knots with 4 or 5 times as much tension.

If local melting plays a role in the actual process by which a knot fails, as seems likely, then knots in non-melting fibers like vectran or technora should behave differently than knots in the more common fibers like nylon or polyester. Knots in polyethylene (spectra, amsteel, etc.) should be especially bad because of the low melting point.

 

[TheTreeSpyder]

Fantastic job!

i think the point should be maid that how a knot is dressed out can make a lot of differance; so someone else might not get the same results with different final "seating" of the mechanics. Also, this Tenex is flat on mount, so would have less deformed dimension/ no height at the bend on that axis(for less strength/efficiency loss there) perhaps affecting results. Global staemeants are hard to make, due to variances of dressing, materials, braid etc.

A Butterfly test should perhaps include a test of just 1 leg to eye; due to some of the ways it is used, that don't conform to leg to leg with no or less load to eye and pulled perpendicualr to line.

Thank-you very much.