Bend radius for ropes

[Ron]

Hmmm, something isn't making any sense at all. If a 6000# rope is eye spliced and clipped into a biner and we pull on the other end, you're saying that when we apply just over 3000# to the rope, it will break at 3000# in the bight of the eye? Yet rope manufacturers claim an eye splice retains nearly 100% of the strength of the rope and don't mention bend ratio. I have to believe their wise enough to know that in nearly 100% of the applications of their eye splices, that eye is gonna be attached to a biner or similar sized device. How could they claim near 100% strength of the rope if the eye breaks in the bight, at the biner at 50%?

But if it's true that the rope breaks at half its strenght in the bight, at the bend radius, then all loop knots and eye splices attached in a 1:1 bend ratio configuration will fail at half the rated strength of the rope. So the knots would be as strong as the splice over a biner.

I've done a bunch of pull tests with knots in 1/4" nylon rope. Out of some 30 knots, I never once saw a knot break in the bight. Now I may not have had a 1:1 bend radius, but it was close, but tomorrow I'll do some with exactly a 1:1 or worse and see where they fail.

 

[TheTreeSpyder]

Flair has to be right/ gradual; for low impacct of the change in forces for strongest splice. This does have the 'weakness' of creating a thicker and stiffer portion of line over a longer distance than short splice in trade(besides taking more time and care to make). This can make the area less useable around any type of bight or bend size and leverageable wise.

Regular knots, would only bend their Standing Part's full force with 1 Turn. Destabilized by several factors suddenly. A Scaffold; gives several stacked Turns; to give more gradual destabilization/impact. It still has the same amount of force on the the bottom Turn of this stack as a single Turn of regular knot. But, a Scaffold's upper Turns can more gradually make the changes and kinda firm/stabilize(?) the same key area. To me, kinda like a short or long taper of a splice; and having the same tradeoffs of weaknesses.

Going to wider things to preserve soft bend radius; also requires watching the slant of the legs and their leveraging; as well as the force that is best an inline pull on the traps of these machines; now instead recieving a sidewards pull on their mchanisms; that they were not designed to capitalize on i think.

 

[Ron]

I don't know about traps on machines, what I can't understand is how different knots have different breaking strengths if the breaking strength is due to the bend radius in the bight instead of the strength of the knot.

If the bend radius is the weakest point, and reduces rope strength by 50% the why would the knot break? The bight over the bend radius would break before the knot. I have never seen that happen.

But, to redo some of the tests I have already done, I just bought 50' of 1/4" nylon rope and I've got two 1/4" diameter screw links. Tomorrow I'm gonna tie a F8 with a single loop in both ends of a one foot piece of rope, hook the bights in the screw links, and pull against the screw links until one of the knots fail. A figure 8 is rated at at least at 75%. I'm going to repeat this 10 times. If I see the same thing I have every other time I've done this, the rope will break all ten times on the single line side of the knot very near where it goes into the knot, and the bight will still be in tact.

Now, if it is true that the 1:1 bend weakens the rope by 50% of the single strand strength, the bight should fail before the knot at about 50% of the single strand strength of the rope.

I'll make pics of the failed knots. But I gotta say, based on doing this before, and the people I've talked to about this, Bruce Smith for one, I'm under the impression that the bight in the 1:1 bend is not the weak point, the knot is.

Well, we'll see. I wish I could measure the strength of the 1/4" rope; I can't do that yet, but it's in the works.

 

[TheTreeSpyder]

The knot is the machine; the trap is the way the mechanics locks itself with it's own forces; usually including bending the fully loaded Standing Part.

To this imagery; the bend is how the Standing is leveraged by it's own stiffness, bent. A slack line cannont be leveraged; for it doesn't resist the bending; like a wrench that flexes. But, once loaded, the stiffness, does have resistance to bending, therefore is leverageable by perpendicular force. A tight bight is when most of the line isn't bearing the load; only a portion. Thinning is how the line is stretched; a knot can stop this suddenly; giving a change in taper. All changes and their suddeness matter.

The Standing is the fully loaded part; so it's leveraging etc. has the highest potential multipliers in these situations; stiffness/resistance to bend for one.

Here, the bend stands against ye; but understanding sweating or swigging a line to gain more purchase; shows how the same forces can help. i think rope, pulley systems give clues to the happenings inside the microcosm inside of knots; being subject to the same forces, in the same materials.

 

[Ron]

Be that as it may, the 50% value for a 1:1 bend ratio reflects the effect of all those mechanisms at work in the rope.

The question remains, with a spliced eye, does the 1:1 bend radius the bight experiences reduce the strength of the eye to half the single strand strength of the rope?

I can readily see that if you try to lift 3000# by attaching one end of a rope to the 3000#, run the rope through a fixed overhead biner with a 1:1 bend ratio and then exert 3000# on the other side of the rope to lift the 3000#, the rope would be at it's breaking strength. In other words you're using the biner like a simple 1:1 pulley system, where the biner acts as the pulley. I can clearly see in this example how the rope would only be capable of supporting half it's single strand strength.

But that's very different than an eye or loop knot attached to a biner with a 1:1 bend ratio, with a single strand attached to the 3000#. In this case, the rope in the bight of the rope would not be near it's breaking strength because the each of the strands in the bight only have to hold 1500# each. Since the rope is a 6000# test rope, each strand should be able to hold a maximum of 3000#, i.e. 50% of 6000# per strand. If each strand can hold 3000#, half the rope's breaking strength, then it would take a 6000# load to bring the rope to it's breaking point in this example.

 

[Norm_Hall]

Every rope construction and fiber will react differently in a break test (3 strand, 12 stand, 16 strand, double braid). All will break at different loads depending on construction, pick ratio, fiber twist, fiber, AND how much rope is in the system.

 

[Leon]

Norm,

I understand what you say about different ropes breaking at different strengths. But how about where they break? If several different rope types are tested with the same knot will they break at the same part of the knot, or in different areas?

 

[Ron]

Well I promised pics and here they are. I pull tested 15 knots in two different materials, polypropylene and nylon and two different rope constructions, laid and braided.

Figure 8 loops were used in all pulls because they are one of the stronger loop knots.

I used 1/4" rope and a 1/4" diameter screw link and a 3/16" screw link.

The tests were performed by tying F8s in each end of a short piece of rope. I did five tests for laid polypropylene, five tests with laid nylon, and 5 tests with braided nylon.

I tested the laid polypropylene first and after four pulls, I switched from a 1/4" screw link to a 3/16" screw link. The 3/16" link with the 1/4" rope gave a bend ratio of 0.75:1. That's 25% worse than a 1:1. But no matter, the failure was at the very same place in all 15 tries. As you will see from the pics every knot failed where the working end of the rope enters the knot. There were no failures of the bight. In fact in every case, as you can see, the bights are still totally in tact.

Oh, BTW, you can't see it in the pics but there's a piece of rope tied to every piece of chain, hook, and any other piece of metal that WILL go flying throught the air when the rope being tested breaks. If you ever do pull tests, be aware stuff is gonna be flying fast when the rope breaks - could be very dangerous if not restrained adequately.

Here's the set up; force was applied via a 4000 lb chain hoist.

Here's the first set of pulls. This is the 1/4" laid polypropylene rope. The four top ones were pulled with a 1/4" link and the bottom one with a 3/16" link for a 0.75:1 bend ratio. There's no difference in where the failure was.

Here's the next set, the 1/4" braided nylon, all pulled with the 3/16" link. All broke at the same place and in the same place as the first batch.

And here's the last batch, all with the 0.75:1 bend ratio, failed in the exact same place as all the other tests.

So, even with a bend ratio of 0.75:1, the bights in loop knots don't fail. The knots fail where the working strand of the rope crosses the first strand in the knot. I learned that from Bruce Smith of On Rope 1 a year ago and after testing about 45 loop knots, every one I have pulled or seen the results of a pull have failed exactly where Bruce said they would.

Bruce however was talking specifically about climbing rope. I have seen from the testing I have done, it holds true for smaller diameter ropes as well.

So placing a loop knot over a biner and forming a 1:1 bend ratio does not weaken the rope by 50% of the rope's strength.

But if you tried to use the same rope and biner as a simple pulley system, the load the rope could support with the 1:1 bend ratio would indeed be half the strength of the rope.

That's my story and I'm stickin' to it.

 

[Leon]

Thanks for doing all that and posting it Ron. Very interesting.

 

[Lazarus2]

Quote from Ron:

"So placing a loop knot over a biner and forming a 1:1 bend ratio does not weaken the rope by 50% of the rope's strength.

But if you tried to use the same rope and biner as a simple pulley system, the load the rope could support with the 1:1 bend ratio would indeed be half the strength of the rope.

That's my story and I'm stickin' to it."




I see where you are coming from now Ron - you're trying to prove that the 50% strength reduction doesn't apply to loop knots. That MAY be down to the mechanics of the knot. I think doing the same test with splices of various types would be more conclusive by removing the knot tying variables.

Seems from the quote above, that you were arguing a point of the first para, where as I was (and I think most others) were discussing the second : )

Also, as explained earlier, the 50% isn't a 'Rule', just a concept that is more relavent to cycles to failure.

I think your reasoning with knots is in keeping with strength increase/reduction concepts of sling configurations. But these too are related to the strength loss from tight bends or even too large a bend (leg angles to the load again). e.g. I have pull tested 1700Kg cord, tied in a loop with a double fishermans, and it tested out at 2,700Kg - 60% increase rather than 70% from a splice, which is surprisingly good for a knotted sling.

Maybe its a simple case that a 1:1 bend accounts for 30% strength loss rather than 50%, regardless of where it is?

Also interesting, is the strength of the fishermans type knots - seem as strong as some splices for sure in tests I've done. A Scaffold knot is certainly stronger than a Fig 8. This must also be bend related?

Interesting stuff.

 

[Ron]

[ QUOTE ]
Quote from Ron:

"So placing a loop knot over a biner and forming a 1:1 bend ratio does not weaken the rope by 50% of the rope's strength.

But if you tried to use the same rope and biner as a simple pulley system, the load the rope could support with the 1:1 bend ratio would indeed be half the strength of the rope.

That's my story and I'm stickin' to it."




I see where you are coming from now Ron - you're trying to prove that the 50% strength reduction doesn't apply to loop knots. That MAY be down to the mechanics of the knot. I think doing the same test with splices of various types would be more conclusive by removing the knot tying variables.

[/ QUOTE ]

You might want to run that by Nick. He's had a number of splices tested and as he said it, and I quote, "...I would never expect the eye to tear....i'd expect them to tear right where they did. [he was referring to my tests]

Even if those spliced eyes, the eye wouldn't be what tore...it would be at the weakest part of the splice, where the sharpest angle was...and that would break at WAY higher than 50% of the ABS."

[ QUOTE ]

Seems from the quote above, that you were arguing a point of the first para, where as I was (and I think most others) were discussing the second : )

[/ QUOTE ]

All my remarks have been to address the original question, and I quote, "...This would mainly be referring to the use of hardware and the attachment to hardware. The carabiners and rope snaps, is there a certain diameter that is the standard for such hardware? I know at tight bend radius is not good on the ropes but what is too tight?...

[ QUOTE ]
Maybe its a simple case that a 1:1 bend accounts for 30% strength loss rather than 50%, regardless of where it is?

[/ QUOTE ]
From about 45 knot pull tests, and Nick's splice tests, I can't see that a 1:1 or even a 0.75:1 ratio has any effect on the <u>single strand</u> strength of the rope. That doesn't mean there isn't an effect, but I'll get to that later. It is the knot that is consistently breaking at the point where the working part of the rope enters the knot. If the bend ratio generated the weak point, then all knots and splices would fail at let's say at 50% of the single strand strength.

But we see some knots are rated at 75%, some 65%, etc. - where the percentage is the percentage of the <u>single strand</u> strength of the rope. E.g. A bowline has about a 65% rating. BTW, in the pdf paper you reference in an earlier post, the initial control pull tests shows that a bowline failed at 67% in one case and 66% in the other. Anyway, if we tied a bowline in a 6000 lb rope, the resultant strength would be 65% * 6000 lbs = 3900 lbs. But if the bend ratio produced a 50% reduction, the rope would fail in the bight at 3000 lbs.

I did see two pics in the study you referenced where the bowlines broke in the bight. However, this seems to be a rare exception, rather than the rule and makes one wonder if some other factors were in play. In any event, the ropes were not pulled against a 1:1 bend ratio. The ropes were 1/2" (12.7mm) and were attached to a 16mm screw link. That would provide a 1.26:1 ratio.

[ QUOTE ]
I think your reasoning with knots is in keeping with strength increase/reduction concepts of sling configurations. But these too are related to the strength loss from tight bends or even too large a bend (leg angles to the load again). e.g. I have pull tested 1700Kg cord, tied in a loop with a double fishermans, and it tested out at 2,700Kg - 60% increase rather than 70% from a splice, which is surprisingly good for a knotted sling.

[/ QUOTE ]

You're makin' me think too much. I won't be worth a hoot the rest of the day. Actually this is the point I said I'd get back to. I'm assuming you have formed a loop by joining the ends of the cord with doublefishermans bends. Then you pull against the loop and it failed at 2700kG. Since this configuration consists of two strands of rope supporting the 2700kG pull, each strand would have to hold half the load or 1350kG each. So the rope broke at the knot with 1350kG applied to it. But 1350kG / 1700kG = 79%. So the rope broke at 79% of the <u>single strand</u> strength. And 79% is right on target for the breaking strength of a double fisherman's bend. I've read that a triple fisherman's bend is right at 100%!

[ QUOTE ]
Also interesting, is the strength of the fishermans type knots - seem as strong as some splices for sure in tests I've done. A Scaffold knot is certainly stronger than a Fig 8. This must also be bend related?

Interesting stuff.

[/ QUOTE ]
That is interesting, but how are you comparing an eye to a fisherman's knot?

Edit: Ahhhh, a fisherman's loop to an eye splice - my brain must not be fully engaged yet, if ever. That is interesting. Testing I have done has indicated fisherman's and F8s were pretty close to the same, but I'm wondering if I tested a full double fishermans, i.e. a fixed loop formed by two locking fishermans or a 'half' fisherman that cinches down.

Hmmmm, scaffold knot, that sounds interesting too, I need to take a look at that. Wait a minute, isn't a scaffold knot just a fishermans with an extra turn?

 

[TheTreeSpyder]

Very nice effort-Thanx!

i'd think a triple anchor to self; to form eye(Scaffold, fisherman's etc. naming) would run more efficient than fig.8. i look at fig.8 class as a linear round turn, locked. Each, though has a 'stash' of extra line to pull from to ease dynamic loading forces though.

The eyes you tested would be loaded with each leg half load x secant(secant=1/cosine) of the angle from inline lay of the line in my book. Still the full loaded Standing Part bend would create a larger force agaisnt the line (and other changes outlined).

i loosely categorize a Bowline as a SheetBend to self to form eye; except that the Sheet's Hitch is turned into a Half Hitch. A Bowline's inefficiency; would be ruled by it's Half Hitch bend from the Stahding; before choking around the bight in it's own tail/Bitters. A Double Bowline's half hitch would be softened by giving longer for the same bend, so stronger (but must be dressed better to not use the extra length to leverage inversion-as per KnudeNoggin).

But, a Running Bowline around a post, would have an inefficiency of the eye's bend on the Standing Part. This as/and an eye to eye sling would be the more common applications of coming closer to a fixed pull breaking at the krab, rather than the 'knot'; would be an eye to eye 'sling' choked. Whereby, there would only be 1 support leg; but once again, the bend of Standing would probably create greater force, than the compromise of the bight around krab. These situations would be most dependant on the teepee point. But, also, any frictions from the choke slip; would kinda lend towards 2 legs of support on krab (but not fully 2 legs of support).

The, next step in trying to break at the bight without movement/cranking fibers around the bight; would be a "Friction Less Hitch" of tight bight or even less commonly a Mayhem Puzzle type bight ; that like Frictionless wouldn't have a bend in the Standing; but have a tight bight; and then double, rather than half load the bight(in a low to Zero friction bight)...

As all ways; we thank you for your co-arboration

!

 

[Mahk_Adams]

Thanks Ron, very interesting.


[ QUOTE ]
So placing a loop knot over a biner and forming a 1:1 bend ratio does not weaken the rope by 50% of the rope's strength.

But if you tried to use the same rope and biner as a simple pulley system, the load the rope could support with the 1:1 bend ratio would indeed be half the strength of the rope.

[/ QUOTE ]


Have you tested the latter statement? Photos? Info??

 

[Ron]

[ QUOTE ]
...The eyes you tested would be loaded with each leg half load x secant(secant=1/cosine) of the angle from inline lay of the line in my book...

[/ QUOTE ]

I like your book! Absolutely, I agree completely. However because I used F8's with small radius screw links, and a relatively long bight, the angles were quite small. But, yes, the angles could be significant in common applications.

I wonder about those really tight eye splices we like to use on, oh all right, krabs, I learn somethin' new every day. I guess you'll have me sippin' hot tea in the middle of the afternoon next. Anyway, that tight eye does increase the angle significantly which would lower the strength of the eye. But unless the eye was gonna be loaded almost to the maximum strength of the rope, it probably wouldn't matter.

Interesting stuff isn't it?

 

[scott_baker]

I have been using some Amsteel ropes lately as I have been learning to do the Statics Integrated Method "tree pulling tests".

You can't use knots in this super strong rope...have to use a special thimbel with the right radius and a splice.

On one job we decided to save soe time and a thimbel by using a knot and then planing on sacrificing that part of the rope. The knot slipped as it was not dressewd tight and the rope broke at the knot. We had a load of about 1.5 metric tons on the system.

Zeb Haney was helping us and he tied the knot...he thought that just the slip as the knot tightened generated enough friction to cause the failure. This rope has a startlingly low melt point.

I will be very careful with this rope from now on.

Any comments from the rope guru's?

Aslo what is the pulling test that someone mentioned in the thread? As far as I know only one person (not me) in North America has the gear and know how for the SIM.

Scott

 

[scott_baker]

A picture of Amsteel rope getting loaded!

 

[Ron]

Scott,

I'd never consider myself a rope guru, so I hope I'm not out of place commenting here. I don't know what the SIM is you referred to, but pull tests on knots and rope are pretty common tests these days. Even Sherrill Tree has done (or had done as the case may be) rope tests to measure stretch in ropes.

The pull tests I did were knot against knot tests to see where the rope would break. I had no way to measure the force in the rope, but I hope that's about to change.

It is interesting that you mentioned friction and low melting point. In a number of tests I've done with nylon rope which has a higher temperature rating than Amsteel, I've seen the nylon actually melted and fused together in the knot. And this was with slow pulling. So the friction and pressure within the knots are high enough to generate high temps. That maybe one reason you can't tie knots in Amsteel.

Wow, 1.5 metric tons, that's only about 3300 lbs. I wonder if it would have still broken if the knot was tighter to begin with. It quite possibly would have. You didn't say what size Amsteel you were using but even 5/16" Amsteel has a tensile strength of 13,700 lbs. That's just interesting!

Oh - great pic!

 

[knudeNoggin]

Ron, did you save the test specimens?!
I'd love to see those things up close &amp; personal.

[ QUOTE ]
Well I promised pics and here they are. I pull tested 15 knots in two different materials, polypropylene and nylon and two different rope constructions, laid and braided.

[/ QUOTE ]
You da Man(tm)!

[ QUOTE ]
Figure 8 loops were used in all pulls because they are one of the stronger loop knots.

[/ QUOTE ]
As best I can discern from the photos, many of the Fig.8 loopknots were
of what has been called the "weak" "perfect form"--one can see an obviously
untensioned twin outer part, and the knots are more aligned w/axis of
tension than scrunched to a nearly 45deg diagonal of it (though I think
I see some like that: I think that the bottom one for 1st white rope (2nd
pic set), the 1/4" nylon braided rope, has one of each (and indeed the "weak"
one broke).

[ QUOTE ]
As you will see from the pics every knot failed where the working end of the rope enters the knot.

[/ QUOTE ]

The break in the 4th (1 up from bottom of set) PP laid yellow rope looks
especially interesting &amp; potentially informative: it seems that two of the
strands broke in one knot, and the third in the other. (I've seen results
of tests where the pulling stops upon the rupture of 1 or 2 of the strands.)
It's sometimes possible to figure out where the broken strand was at the
rupture point; from my inspections, it seems to be on the inside of the
bend, contrary many arm-chair theorists' speculations!

(Hmm, and the break in the bottom laid nylon looks similarly separated!)

[ QUOTE ]
Oh, BTW, you can't see it in the pics but there's a piece of rope tied to every piece of chain, hook, and any other piece of metal that WILL go flying throught the air when the rope being tested breaks.

[/ QUOTE ]
Good warning, although I'd think that the chain wouldn't have so much
stored energy (inelastic) to fly (though I've heard of steel cable doing
impressive shearing damage ...)!? Seems that in one of the pics of testing
on Tom Moyer's site there is a heavy blanket/carpet(?) draped over the
danger zone. (And don't we in fact see some of your safety cord in the
top photo--thin, white, upper left?)

[ QUOTE ]
The knots fail where the working strand of the rope crosses the first strand in the knot. I learned that from Bruce Smith of On Rope 1 a year ago and after testing about 45 loop knots, every one I have pulled or seen the results of a pull have failed exactly where Bruce said they would.

[/ QUOTE ]
Hmmm, that might hinge on what is meant by "crosses the first strand":
how would you (or Bruce) interpret that for the Bowline? --where the main
line enters crossing a (bight) collar before forming the nipping turn (loop)?
(It should break in the turn, but this can depend upon how tightly one sets
the knot--shouldn't be all so tight.) It appears that most of your Fig.8 LKs
broke where the loaded main line emerges from the twin collars and
curves around over a diagonal part (which is a weak point I don't think
much obtains in the "strong form" of the Fig.8).

[ QUOTE ]
So placing a loop knot over a biner and forming a 1:1 bend ratio does not weaken the rope by 50% of the rope's strength.

[/ QUOTE ]
There have been reports of testing slings made by joining two ends of the
rope with a Grapevine bend (Dbl. Fish. knot) where the break came at the
'biner-sized pins; there was conjecture that this resulted from movement
of rope resulting from material feeding out into one side as a result of
knot compression. One could check this case, too; and to anticipate the
conjecture by putting the knot in both sides (gratuitous to making a sling,
but key in trying balance material feed between the sides of the sling),
and by pre-loading the knot so that there's not much more material to
give out. Then does the break (still?) occur at the pin, or now in the knot!?

[ QUOTE ]
But if you tried to use the same rope and biner as a simple pulley system, the load the rope could support with the 1:1 bend ratio would indeed be half the strength of the rope.

[/ QUOTE ]
How come?
I think that the results of some of Kolin Powick's testing a Black Diamond
reported at [url="http://www.bdel.com/scene/beta/qc_kp_archive.php#011907"]www.bdel.com/scene/beta/qc_kp_archive.php#011907[/url][/url]
will dispell this belief, or mitigate it some degree. (If I'm reading one of the
tests of "pulled over a 'biner" [sorry, krab shortage stateside--big dead zone
in the Chesapeake] correctly, a 4500# rope generating 6600# implies some
3300#, or about 75%; and in the drop tests, max. impact forces rose to
nearly 4000# (this is on the dropped end, and the rope runs over a bar
to simulate a 'biner).

Further, there was a test of the Strangle noose hitch (what is shamelessly
called many with tree sappy nomenclature a "(dbl.) fisherman's) tied in 8mm
low-elongation rope (BW) around a 'biner and breaking at some high %age
(tester didn't bother figuring tensile), and at the point in the mainline where
the Strangle knot's outer reach severely nips it--not at the bend around
the 'biner.

*knudeNoggin*

postscript: In the August (?) issue of Sport Fishing, Doug Olander (editor?)
tests a slew of mono fishline with three knots (Bimini Twist, Palomar, and
some odd-named bight hitch like the Albright); he has some surprising
results--ranges from 45% to 110%, and indeed repeated strong-than-line
results (w/the Bimini, IIRC, but I only scanned it). He DID test for tensile
(and notes the discrepancies between his &amp; rated strengths--which can
be huge (esp. in some earlier tests of gel-spun (HMPE (Spectra/Dyneema))
lines), and he did not believe the &gt;100% results so focused on them.
But I don't think he gives details of his test configuration, alas.
now, wait a minute, what forum have I wandered into, ...

 

[Ron]

[ QUOTE ]
Ron, did you save the test specimens?!
I'd love to see those things up close &amp; personal.


[/ QUOTE ]I did.

[ QUOTE ]
As best I can discern from the photos, many of the Fig.8 loopknots were of what has been called the "weak" "perfect form"--one can see an obviously
untensioned twin outer part, and the knots are more aligned w/axis of tension than scrunched to a nearly 45deg diagonal of it (though I think I see some like that: I think that the bottom one for 1st white rope (2nd pic set), the 1/4" nylon braided rope, has one of each (and indeed the "weak"
one broke).

[/ QUOTE ]

I'm not familiar with the "weak" terminology but there are two ways to tie a figure 8 loop. One with the tail strand ‘on top’, the second way is with the working strand ‘on top’. In the following pic, the working strand is the top loop, i.e. the one nearest the eye.

.
.
In this pic, the tail strand is the top loop, or the one nearest the eye.

If memory serves, from previous tests, pulling one form against the other, the form in the lower pic always won - gosh I hope I'm remembering that right. I'm pretty sure though, because I was surprised at which one was stronger and I started using that form for all my F8s.

It was my intent to tie all the F8s the same way, but I tied over 30 of them and I could have drifted off and tied one differently, but I tried to tie them every one the same. I really think that if I looked at them, assuming they aren't too distorted to tell, they would every one be tied with the tail strand 'over the top'.

[ QUOTE ]
The break in the 4th (1 up from bottom of set) PP laid yellow rope looks especially interesting &amp; potentially informative: it seems that two of the strands broke in one knot, and the third in the other. (I've seen results of tests where the pulling stops upon the rupture of 1 or 2 of the strands.)It's sometimes possible to figure out where the broken strand was at the rupture point; from my inspections, it seems to be on the inside of the bend, contrary many arm-chair theorists' speculations!
(Hmm, and the break in the bottom laid nylon looks similarly separated!)

[/ QUOTE ]

Yep I noticed that, but since I was more interested in whether the rope breaks at the bend around the screwlink or somewhere else, I didn't give it a lot of thought but it was interesting.

The knots become very distorted. As I casually looked at them after failure, I noticed both knots were pretty much a mess. But, again, my observations were focused on 'did the bight break or did the rope fail at or in the knot'. The first thing I looked at on both ends of the rope was the bights/loops and knots to see if they were still in tact. Though distorted and fused, the bights and knots were still looped around the screw links and in all cases I had to open the screw links to remove the loops. I did observe that I could not see ANY distortion, fusing, or hardening of the rope where it passed over the screw link. I remember thinking that if I didn't know, there would be no indication at all that this bight/loop had been involved in a pull to failure test.

[ QUOTE ]
Good warning, although I'd think that the chain wouldn't have so much stored energy (inelastic) to fly though I've heard of steel cable doing impressive shearing damage ...)!?

[/ QUOTE ]
It wasn't the chain that was the problem. I put an 8000#, 3 inch wide tie down strap around each tree I was using as anchors. When the rope failed the chain would go all the way back to the tree it was anchored to. My 30# winch would jump forcibly rearward - toward me, as far as the safety line would let it go - which BTW, was carefully measured to limit travel. Recoil in both directions of the anchors was significant.

[ QUOTE ]
Seems that in one of the pics of testing on Tom Moyer's site there is a heavy blanket/carpet(?) draped over the danger zone.

[/ QUOTE ] Yep, I remember seeing that.

[ QUOTE ]
(And don't we in fact see some of your safety cord in the top photo--thin, white, upper left?)

[/ QUOTE ]
Yes, but I wasn't sure that the safety line would be recognized in the pic as a safety line, and I wanted to be sure everyone understood the need for safety lines and that I did indeed use them.

[ QUOTE ]
Hmmm, that might hinge on what is meant by "crosses the first strand": how would you (or Bruce) interpret that for the Bowline? --where the main line enters crossing a (bight) collar before forming the nipping turn (loop)? (It should break in the turn, but this can depend upon how tightly one sets the knot--shouldn't be all so tight.) It appears that most of your Fig.8 LKs broke where the loaded main line emerges from the twin collars and curves around over a diagonal part (which is a weak point I don't think
much obtains in the "strong form" of the Fig.8).

[/ QUOTE ]

Let us not forget that it was over a year ago that I talked with Bruce about this. It would be very easy to over-simplify his statements about this. The principle here is that the working line is what breaks, and it breaks at the highest point of stress IN THE KNOT. I presume that would be as you described it (quoted above) if I understand what you mean by 'twin collars' and 'diagonal'.

[ QUOTE ]
...But if you tried to use the same rope and biner as a simple pulley system, the load the rope could support with the 1:1 bend ratio would indeed be half the strength of the rope....

[/ QUOTE ]

[ QUOTE ]
How come? I think that the results of some of Kolin Powick's testing a Black Diamond reported at [url="http://www.bdel.com/scene/beta/qc_kp_archive.php#011907"]www.bdel.com/scene/beta/qc_kp_archive.php#011907[/url][/url]
will dispell this belief, or mitigate it some degree. (If I'm reading one of the tests of "pulled over a 'biner" [sorry, krab shortage stateside--big dead zone
in the Chesapeake] correctly, a 4500# rope generating 6600# implies some 3300#, or about 75%; and in the drop tests, max. impact forces rose to nearly 4000# (this is on the dropped end, and the rope runs over a bar to simulate a 'biner).

[/ QUOTE ]

Here's how come. First the statement I made (quoted just above your quote) was made in reference to the notion that a 1:1 bend ratio would produce a 50% reduction in rope strength at the bend point. None of us were necessarily looking at the 50% thing as an absolute rule, but accepting that as a number for discussion sake.

So in context with that, there is a huge difference in doubling a rope over a biner and connecting both ends to the load, than attaching one end of the rope to a load and pulling on the other end of the rope.

In the first case, both strands (theoretically) share the load equally. So if we use a 6000# rope then theoretically each strand can support 6000#. The theoretical maximum load the two strands could support is 12,000#. But for discussion sake, if the bend over the biner weakens the rope by 50% then both strands are weakened by 50% so each strand can only support 3000#, but there are two strands so together they can theoretically support 6000#.

But in the pulley configuration, we would have a 6000# line attached to a load on one side, passed over a biner, and an equal and opposite force applied on the other side. So from the start, a 6000# load would produce a 6000# load in both strands of the rope - the strand holding the load and the the strand the force applied to. So the rope is right at it's breaking strength with a load of 6000#. But, again, if the rope's strength is reduced by 50% by running over the biner, then each strand is reduced by 50% for a maximum of 3000#. So 50% strength reduction, the 6000# rope will break at about 3000# or half it's strength.

I looked over the link you supplied. I couldn't find the incident you described?????

 

[knudeNoggin]

Ron, your Fig.8 lk pics are good BUT for the top one having the tail being
above the loaded end when outside of the knot; how such a dressing
will draw up will vary on circumstance, but it amounts to the loaded end
crossing atop the tail, and how set. (To see it another way, one should
be looking at a SINGLE photo that has the ends labelled and then talking
of "end-x" vs. "end-y"--note that your two Fig.8s are NOT the same,
and the ends' crossing over/under immediately inside of the twin collars
occurs (rightly) only in the lower photo; make the same crossing in the
upper knot and the loaded line comes out above the tail.) IIRC, though,
that misformed (by my standard) top form is (for that *half* of it) what
<u>Life on a Line</u> recommend--also for easing untying.)

For the upper form to be stronger, I'd set it by loading the tail vs.
the eye, to put in the shape of the knot that gives such nice bending to
the loaded line (whose loading in the case of manual setting will far
exceed setting forces, of course). I'd thought that the form you favor
to be stronger, as it appears to make the loaded strand take a wider
u-turn; but a rockclimbing safety manual was explicit--with GOOD
illustrations, not the usual nonsense traced-w/o-crossing ones such
as <u>On Rope, 1st ed.</u> and many others use--on this point,
I tried to figure out how/why, and hence my conjecture. Lyon Equip.
concluded, testing both but w/few cases, that it didn't matter. The
<u>Life on a Line</u> e-book asserts yet a different symmetric
form, more awkward to tie, of having 10% improvement--which is
I suspect the item aimed at by current "OnRope1"'s Mythbuster on
Fig.8 variability (showing, alas, that nonsense form but at least
admitting that force will make significant alterations to it).
Set as I'd do, the loaded line will see some of the collar's
constriction on it but w/only minor deflection, and then pass
over the twin diagonal parts while bearing slightly against its own
parallel twin--line leaning into a diehedral groove, almost--; by the time
it makes the u-turn around the eye legs, it's had a change of direction
and I surmise off-loaded some force across much of the rest of the
knot. --but this is so far my own conjecture of things.

As for that 1:1 dia. object reducing strength to/by 50%, I don't accept
it, and as Mahk asks would prefer to see some testing of the assertion
(which I suggested some ways to do with a closed loop/sling).

------

Ah, I see about the anchors &amp; recoil; I recall on one of these forums
someone telling of a trailor-hitch ball being ripped off by a tow strap
in some off-road fun--and that ball being shot well into the pulling
vehicle's engine (trying to pull free another vehicle stuck on a rock) !!!
One would only want to see this if at all from a distance!

Thanks,
*kN*