Knots in Nylon Sling to Protect Marginal Placements

Previous discussions here on screamers have led me to the disappointing conclusion that they're probably not worthwhile as a means of reducing load on marginal placements. It's just too unclear whether they're helpful, and in some cases they might be counterproductive.

I was thinking this morning about knots in nylon slings and remembered seeing somewhere that loose knots (i.e. overhands) in nylon slings can provide some dynamic aspect, since force is absorbed by the tightening of the knot. It does reduce strength, but at 22kN starting sling strength, even large reductions in strength mean an ideal cam placement is still the weakest link in the chain. Multiple knots wouldn't reduce strength beyond the strength of the weakest knot. So it seems to me that a nylon sling with multiple (loose) knots in it could potentially be used to reduce the forces experienced by a marginal placement.

It's fairly clear to me that even experienced climbers are unable to look at a marginal placement and estimate how much weight it would hold, so we don't really have a target minimum force here. But I do have some pieces (white tricam) that have low ratings (3kN) when placed ideally, so maybe we could view getting the peak force on a placement during a fall down to 3kN as a target.

It's also prohibitively complex to test reductions of force on real life marginal placements to see if this would ultimately be effective in preventing placement failure, so sadly this discussion will be fairly speculative.

However, I'd like to hear some opinions from some engineering types (Jim Titt? rgold?) if they think this might work, and if there are any ways it might be counterproductive.

The specific scenario I'm proposing is that you get a nylon sling, rack it with some loose overhand knots pre-tied, and use it like you would a screamer.

EDIT: Forgot to mention the caveat that dyneema behaves very differently from nylon when knotted, and probably isn't appropriate for this application.

EDIT: To be clear; I'm aware of alternatives to this (Kong Kisa, doubling up placements, just running it out and tolerating the risk, just backing down, etc.). I'm not asking for other solutions to this problem, just wondering if this idea might be a helpful tool for marginal placement situations.

caughtinside wrote:

Seems like a lot more trouble than its worth. Plus you might glaze your nylon.

If we're talking about this as an alternative to screamers, these don't seem like problems. You'd just carry a pre-knotted nylon sling like you would a screamer. And if you fall on it, a screamer stitching will tear, leaving you with a regular sling, while a nylon sling with knots can probably be loosened up and reused--I agree it might get glazed, but I don't think it would get enough glazing from one fall to have to be retired. A screamer would typically not be useful as a screamer after even one fall.

Impact(kN) absorbing knots

Of course, since you appear to like experimenting - 

If you weren’t a boy/girl scout, a slip knot is essentially a loop of rope/string/thread passed through a bight (a u-shaped section of the same rope/string/thread). For a bit of extra flavor, you can throw in a few coils around the loop (see above). As you pull the slip knot tight, the loop tightens, dissipating the energy through friction, until the loop is tight enough that friction holds the knot in place. In theory, by pulling the loop through the bight (by tugging on the loose end), the knot can be undone — but in the case of Pugno and the world’s toughest material, it’s just the tying of slip knots that we care about.

https://www.extremetech.com/extreme/154471-worlds-toughest-material-created-by-tying-slip-knots-into-weak-commercial-thread

Thanks for the links amarius.

Feasibility of Knots to Reduce the Maximum Dynamic Arresting Load in Rope Systems

I see 2 problems:

1) Energy absorbed is the integral of force over distance.  Tightening 1 knot is maybe 1 inch of distance.  I suspect most of that distance is at low force, so the effective distance is even smaller.   You'd need a very long sling with a lot of knots to absorb similar energy to what a screamer absorbs.  A piece of climbing rope might work.

2) Knots aren't load-limiting devices.  A screamer tries to offer constant force the entire time it is being stretched.  That enables it to absorb more energy for a given peak force.  Knots are more like springs - the farther you stretch them, the more force is involved.  It's not clear that the force-vs-distance curve for knots is any more efficient (in terms of energy absorbed per peak force) than for climbing rope.  So it's possible knots offer no advantage over letting the rope do the energy absorption.

amarius wrote:

Feasibility of Knots to Reduce the Maximum Dynamic Arresting Load in Rope Systems

Those results seem very promising. If I'm understanding correctly, the technora-technora cord is fairly static, similar to sling material?

While they're lauding the results for the quadruple overhand on a bight as the end-line knot, the in-line quadruple overhand seems to perform nearly as well, and is more applicable to slings. However, it would be a glaring mistake to assume that a nylon sling will behave the same as the cord in the tests.

With screamers, the reason it ultimately doesn't work seems to be (although it's not conclusive) that once the screamer tears all the way, the resulting sling is shock loaded. It seems like a similar effect is observed with the slip-knots--once the slip knot is completely un-slipped, the fibers broke. So inline knots like the quadruple overhand which rely on tightening rather than slipping seem more promising to me.

Serge Smirnov wrote:

Knots are more like springs - the farther you stretch them, the more force is involved.  It's not clear that the force-vs-distance curve for knots is any more efficient (in terms of energy absorbed per peak force) than for climbing rope.  So it's possible knots offer no advantage over letting the rope do the energy absorption.

This is true: we can achieve the same result by having a rope with greater dynamic elongation. However, we can't increase the dynamic elongation of our rope mid-climb, but we can add a knotted sling to the system mid-climb. Although with credit to your other point, the amount of dynamic elongation added to the system might be negligible. The graphs amarius linked do seem to indicate that the difference isn't negligible with static cords, though.

EDIT: It's worth noting that you wouldn't want to simply buy a rope with a greater dynamic elongation, because this would have a major downside: longer fall distance. While longer fall distance in exchange for lower peak force on gear might be a worthwhile tradeoff if you're falling on marginal gear, it's not something you want all the time.

Serge Smirnov wrote:

I see 2 problems:

1) Energy absorbed is the integral of force over distance.  Tightening 1 knot is maybe 1 inch of distance.  I suspect most of that distance is at low force, so the effective distance is even smaller.   You'd need a very long sling with a lot of knots to absorb similar energy to what a screamer absorbs.  A piece of climbing rope might work.

2) Knots aren't load-limiting devices.  A screamer tries to offer constant force the entire time it is being stretched.  That enables it to absorb more energy for a given peak force.  Knots are more like springs - the farther you stretch them, the more force is involved.  It's not clear that the force-vs-distance curve for knots is any more efficient (in terms of energy absorbed per peak force) than for climbing rope.  So it's possible knots offer no advantage over letting the rope do the energy absorption.

i think you are over-thinking this.  for failure of a placement, peak load is more important than the total energy absorbed.  unless you are talking about a placement in thick mud or some plastic type medium.  also,  knots are a lot more like damping than they are like springs.  i am not sure how you come to the conclusion that it isn't clear that knots don't offer any advantage.  are you saying that the 'timing' of the two peaks may not be perfectly coincidental?  my guess is that there is some sort of additive effect and a slight advantage, but not a straight up sum of the peak differences.

David Kerkeslager wrote:

With screamers, the reason it ultimately doesn't work seems to be (although it's not conclusive) that once the screamer tears all the way, the resulting sling is shock loaded. 

If "shock loading" at the end were the only problem, screamers would still be an overall win.  Somebody pointed out that an ideal screamer should be a net win (in terms of peak force) as long as the force it tears it at least 2x the climber's weight, which is usually true.  I suspect this is a simple energy conservation argument.

Jim Titt mentioned on some previous thread that screamers don't actually always work the way they are supposed to - specifically, they provide less (EDIT: or more ?  My memory is fuzzy on this) resistance than they are designed to when the fall involves a high enough speed.

slim wrote:

i think you are over-thinking this.  for failure of a placement, peak load is more important than the total energy absorbed.  unless you are talking about a placement in thick mud or some plastic type medium.  also,  knots are a lot more like damping than they are like springs.  i am not sure how you come to the conclusion that it isn't clear that knots don't offer any advantage.  are you saying that the 'timing' of the two peaks may not be perfectly coincidental?  my guess is that there is some sort of additive effect and a slight advantage, but not a straight up sum of the peak differences.

I was comparing knots to using a lower-impact-force climbing rope (but I now realize that comparison is irrelevant for the scenario David had in mind).

Good point about damping - damping makes the element closer to a constant-force device.  So maybe knots are better than rope.

Kyle Tarry wrote:

These test results aren't with dynamic climbing rope, so I would suggest that they are nearly useless for the application you're looking at.  The difference in peak load will be very large for a nearly static rope, because the knot slip is a much greater percentage of overall system extension (i.e. the "dynamicness" of the system).

In my opinion, this is not why screamers don't work.  "Shock loading" is a popular term but not much rigor is used when climbers apply it; for example, in this case there is a dynamic rope between the climber and the screamer so it's not really "shock loaded" in a meaningful way.  The issue with screamers is that they add energy to the system (via extending the fall distance) and the amount of added energy isn't much less than the amount of energy absorbed.

(If you just place solid gear, you can ignore complicated systems designed to reduce peak load that may or may not work.)

not sure where you are getting that screamers don't work.  i don't think the tradeoff between the added extension versus energy absorbed is a 1:1 thing though, due to all of the non-linear aspects involved.

definitely agree on the whole "shock loading" thing though.  kind of drives me crazy when people describe something as "static" or "dynamic", as if there is some black and white difference between the two at a practical level.

here is some screamer info.

https://www.blackdiamondequipment.com/en_US/qc-lab-to-screamer-or-not-to-screamer.html

Kyle Tarry wrote:

These test results aren't with dynamic climbing rope, so I would suggest that they are nearly useless for the application you're looking at.  The difference in peak load will be very large for a nearly static rope, because the knot slip is a much greater percentage of overall system extension (i.e. the "dynamicness" of the system).

My impression is that the technora-nylon cord in the test was dynamic, but I can't be sure because the dynamic elongation properties of the ropes weren't given in the report.

To be clear, though, I'm proposing tying these energy-absorbing knots in a nylon sling, not in the climbing rope. The nylon sling has some dynamic elongation, but would be fairly static compared to a dynamic climbing rope.

slim wrote:

not sure where you are getting that screamers don't work.  i don't think the tradeoff between the added extension versus energy absorbed is a 1:1 thing though, due to all of the non-linear aspects involved.

definitely agree on the whole "shock loading" thing though.  kind of drives me crazy when people describe something as "static" or "dynamic", as if there is some black and white difference between the two at a practical level.

here is some screamer info.

https://www.blackdiamondequipment.com/en_US/qc-lab-to-screamer-or-not-to-screamer.html

There's a bunch of discussions on screamers on MP, i.e. here, where tests are discussed that show screamers increasing load (it's unknown exactly why). There's a better one than that one I just linked, but I can't seem to find it... But to quote Jim Titt from that thread (emphasis mine):

Some independent test shows some reduction, other independent testing shows little, none or even an increased anchor load. In theory, a screamer with a low enough activation force could increase the anchor load in a lead fall. This is because the leader drops twice the screamer elongation but the work done against stitch resistance is only for the screamer elongation. So in order to have any effect at all, the screamer activation force has to be more than double the climber's weight, otherwise there will be a net loss of energy and so an increase in anchor load. Ordinary screamers do activate at more than double most people's weight, so some fairly small amount of energy will be absorbed. One problem with screamers is that they can only remove a fixed amount of energy from the system, so that as the fall gets longer, the amount of energy removed is a smaller and smaller proportion of the total amount of energy to be absorbed, and the screamer fairly quickly becomes irrelevant, except for dropping the leader further. There is some hope that the screamer might reduce loads on a short high fall-factor fall, but the evidence is not conclusive. Nobody with any practical or theoretical experience believes Yates' numbers. One theory for how the anchor loads might actually be increased is that the screamer reduces the load to the belay device and prevents any slippage there. This might be a good thing, but on the other hand it is possible that slippage through the belay device might absorb more fall energy than the screamer would have. The problem with stitched-type screamers is the peak load to failure varies considerably with the speed the load is applied. To break the stitches first you need to tension them which means the threads have to slide through the parent material and frictional hysterises means this becomes harder and harder to do the faster you load them. The failure scan of a screamer is a jagged line with peaks as the stitches fail and valleys as the next stitches tension and the mean gives the rated failure load. Tested at higher speeds the peaks can be nearly twice the nominal rating. It is also the case that delaying the onset of the maximum force imposed on the rope/belay device may well (depending on how long the screamer takes to activate) cause the maximum force on the top piece to be increased, this can be seen by looking at the load curves from belay devices and time shifting the maximum belayer force to coincide with the maximum rope tension. The CAI tests show undoubtedly that screamers become less effective with larger falls and in the end counter-productive, whatever Black Diamond say based on a minimal amount of poor testing.

Kyle Tarry wrote:

Let me rephrase; if we don't know what the rope is doing, we can't make any assessment about how applicable those numbers are to a climbing scenario.

Agreed--I think the data is useful for informing an educated guess, but it's still well within the confidence range of a guess.

Screamers work reasonably well, but not in the scenarios most people use them for.  Here's the disconnect:

A screamer acts by converting a certain amount of energy to heat.  It activates at roughly 2 KN, and continues until all the stitches are blown, at which point it acts like a regular sling.  It does this just fine.  The trouble is the amount of energy it converts to heat during activation is relatively small compared to the amount of energy involved in a significant fall.  Basically, the effect is real but negligible.

The one time a screamer could have some significant effect on whether a piece pulls or not is when the fall is short, and the amount of rope out is small (like a fall right off the belay).  Just because the total energy is small, that doesn't mean the peak force on your gear might not be large.  So a screamer can actually lower the peak force a bit on your top piece in this scenario.

The question of whether a knotted sling would be any more effective really boils down to how much energy those knots can convert to heat in a significant fall.  I'd say the jury is out.  One interesting factor is that a stitch that blows should always "absorb" the same amount of energy.  However a knot should keep "absorbing" more and more, depending on the severity of the fall.  So each knot will do more in factor 1.5 fall than it will in a factor 1 fall.

GO

Ah, the quote from Jim Titt (posted at the same time as I was typing) said much the same as I did about total energy.  But I still think the question of energy-per-knot question is an important one.

GO

Interesting (to me, at least!) side note to this: having one's rope deviate through protection pieces side-to-side begins to remove rope from the system...judicious management of these "secondary pulls" or "rope drag" or however we term it leaves max rope in the system and in some situations is probably way more effective/critical than knots etc to dissipate energy...dealt with at length in the Mountain Guide Manual. #mtnguidemanual

Yes knots in a nylon sling will reduce peak force. The jury is still out, however, as to how much, and the majority opinion is that the effect will be negligible if you're climbing on a dynamic rope.

The study linked in the previous page talked about technora cord. In addition to being fundamentally different from webbing, you will also see less energy reduction because of reduced friction. Technora is a blend of aramid and polyester or nylon (depending on who's making it) and aramid fibers are rediculously abrasive, resulting in a lot more friction than nylon.

Also, if you do try this kind of thing out, using a single strand will be more beneficial than a loop. This is because it will stretch far, far more when loaded in a single strand configuration than a loop because stretch is proportionate to force and a loop will split the force between the 2 strands.

And also ditch the slipknot idea. Like a screamer, it will add fall distance which may result in a small net reduction or even a net gain in peak force, depending on the circumstance

there are these things... http://www.beal-planet.com/en/express-anneaux-sangles/412-longe-dynamique-dynaloop.html