Cordelette for Climbing: What Are the Issues?
Thanks Steph, yet another informative piece. I have one problem though, with something that was added to the comments at the end of the posts…
“Remember, when you create a loop of the 7mm cord with a fisherman’s knot or similar, you will always weight at least two strands of the cord, effectively doubling the total strength. So a loop of 7mm 13KN cord will hold a total of 26KN. A loop of 6mm 7.5KN will hold 15KN.”
This doesn’t sound right, and if it is wrong (which I think it is), people may rely on it and get hurt. While you no doubt have 2 strands making up each leg coming down from each biner (and joining up at the master point), there is in fact only 1 strand at each biner. So I don’t believe you get double the strength, as Olle suggests. I would have posted this comment directly, rather than bother you with an email, but did not think it was my place to do so.
Regards,
Chris
Hi Chris,
Thanks very much for writing and for raising this topic: as with all things involving loads and forces, there is never going to be a single, easy answer. Recently I’ve seen a few videos put out about both climbing and base jumping gear, addressing safety issues with equipment that are being “proved” with drop tests–the thing is, drop tests will never ever duplicate the same angles and degree of forces as we actually create in real life scenarios. As dramatic and interesting as those videos are, they just don’t tell us anything about the gear they are demonstrating when being used in the field. Nothing involving gear that handles force in variable environments is ever as easy and clean as we would like, but that’s physics for you 🙂
When I have questions about things involving physics and climbing gear, which require actual mathematical answers beyond me going “well, this is what I’ve always done and it’s always worked, so that’s what I can tell you,” I go immediately to Dave Furman, the hardgoods category manager at Mammut USA. He’s written some extremely informative guest posts here, and he is a great source of information.
To discuss this cordelette topic more thoroughly, I asked Dave what he could tell us. Thanks Chris for a great question, and thanks Dave for a helpful discussion of the issues for us to know:
Hey Steph—this is a complicated answer, actually both of your readers are correct in a way. I actually made a diagram and sent it to the engineer I work with to make sure I had my ducks in a row. He did some testing to see if we could arrive at a specific rule for you—unfortunately that would take a lot more testing because the answer is complicated by a lot of variables that are almost never the same between any two “real” anchors, so it’s difficult to be too specific–but nevertheless there are some general rules of thumb we can keep in mind. Because of this, I’ve tried to generalize my own thinking on this without getting too caught up in numbers. Hopefully it’s helpful for your readers.
It is correct that using a loop increases the strength, and in theory it could even be doubled when compared to the strength of a single strand. The caveat is that in a climbing anchor the strength of this cord loop is never doubled, not even close. As part of discussing this, we did a semi-scientific experiment to illustrate the point. We took a loop of 7mm nylon cord (13kn tested breaking strength) and tied it into a loop with a double fishermans knot. When clipped into a carabiner at each end and pull-tested, this broke at the carabiner (picture 1) at roughly 35% higher load than the stated cord strength—not double, but still significant.
HOWEVER, when you pull test like this it is important to remember that the carabiners and knots are the weak links, and changing those will dramatically affect the ultimate strength—it is hard to believe how much variability there is in this when you test in slightly different configurations and with different but similar equipment, but as a general rule any carabiner or knot is going to weaken whatever it’s clipped to. Add in a carabiner with a small rope-bearing surface or a sharp edge (like a burr on the inside of the biner from being weighted on a bolt-hanger?) and you further change the equation. In reality, given the vagaries of the above in addition to simple wear and tear, I would be nervous of relying on any one arm of a knotted cordellette anchor to hold too much more of a load than the tested cord strength—I like to plan for worst cases with the assumption that whatever can go wrong probably will at some point (and so far I’ve been right 100% of the time!).
Many people at this point are probably thinking that a cordelette anchor causes the various points within the anchor to be loaded differently—that’s also pointed out in the comments to your blog post. That’s true–as a for instance, in a knotted cordellette used in a 3-point anchor, if everything were “theoretically perfect”, each arm of the anchor would only be subjected to 1/3 of the total load on the power-point—so even the hardest fall allowed in a UIAA fall test (12kn) would only generate 4kn on each arm of the anchor. In reality, we NEVER have theoretically perfect anchors—it’s not possible. Among other factors, as we spread the arms of the anchor apart even a little, this theory goes out the window. However, what it means is that the total load placed on the power point IS divided between the 3 arms of the anchor to a significant degree, as long as the anchor is truly equalized. When we tested a 3-point knotted cordelette anchor we were able to place a load on the power point that was almost triple the cord strength before one of the arms broke, at which point a second arm immediately broke (picture 2) …but divide that load among the three arms and we are once again talking about a load on each arm of the anchor that is approximately the same as the tested cord strength (also note that these breaks are either at a carabiner or a knot).
This means that you can count on the knotted cordellette anchor as a whole to hold a larger load than the tested strength of the cord, but remember that ONLY works if the equalization of the anchor stays in place—pull that anchor even a tiny bit to the side so only one arm is loaded, and you are right back to the single loop above. In this case the knotted cordelette doesn’t serve so much as a way to increase the STRENGTH of your anchor, it serves to add REDUNDANCY to your anchor. An anchor that allows the equalization to remain in play as it is pulled in various directions will be far superior in this case, albeit with the possible issue of extending and shock-loading the remaining arms if one anchor-point were to fail, or for the entire anchor to fail if a single open loop is used and the cord breaks.
Just remember that a knotted cordelette is a great anchor—it’s fairly simple, very versatile, and in some situations it can be pretty fast and easy to rig, and it can serve to increase the ultimate strength of an anchor system, but it is not at all foolproof because you can very easily lose the equalization. In that case you’re back to square one which is the strength of one anchor point and one loop of cord at a time. Best to know several different ways to create an anchor and what the advantages and disadvantages of each are, and use them appropriately—a lot of guides use knotted cordelettes because they make anchor management with 2 or 3 clients or self-rescue very easy, and they seem to have caught on and are great in many situations—Steph mentioned walls and rigging where organization is key–but due to all of these complications, no matter what the “theory” is, if you aren’t careful you CAN wind up in trouble if you rely on a cord loop to always hold much more than its tested strength. At the very least, if you do this realize that you are pushing things into the realm where if any factors align in the wrong way you really don’t have much safety margin.
Dave Furman
Hardgoods Category Manager
Mammut Sports Group, Inc (USA and Canada)
Awesome info, very important. I have a question: i ask you about your bad ankle and we talked about taping, since i have a crushed heel and bad ankle too; how tightly do you tape and what brand tape do you use. Thanks
OMG I better buy a new something or other from someone
Dave- What did you find regarding distribution of load to each anchor leg? I noticed the leg with the double overhand bend didn’t brake. To what extent was the DOB pre-loaded before the pull to failure? I am also wondering about the geometry of the anchor. From the photo, it appears the leg lengths increase from left to right. Thanks for the input.
Hi, I forwarded your question to Dave:
Steph, whoever that is knows whats up! If you dont tighten a knot in a quantifiable way, it adds a big element of uncertainty into a test, in this case because if the knot were to tighten during the pull test it would essentially reduce the load on that arm, while increasing the load on the other two arms of the anchor. Similarly, if the arms are different lengths it is likely the anchors are not evenly loaded, again possibly causing some of the arms of the anchor to carry a higher load than the rest. Its actually not easy to construct an anchor like this where the load is evenly spread between the three arms–this was not really the angle I took, but it further illustrates the point that relying on smaller-diameter cord with the reasoning that an anchor like this will spread the load out isnt necessarily a good idea. In this case the knot was aggressively hand-tightened the way it might be if you made a new cordellette. The anchor was constructed as carefully as possible to make the arms even, but it was not designed to be a “scientific” experiment really, more an illustration of whats likely to happen.
Sent from my iphone with teeny tiny keyboard and often-problematic auto-correct.
with ankles I tape pretty tight, because you’re trying to cast it. I like to use bad tape because it’s not as sticky, so all the tape you have lying around that’s not good for hand taping (johnson’s and johnson’s, or whatever)
Thanks Steph
hi there,
i think it is in fact very difficult to create a truly pre-equalized 3-point anchor (using a cordelette tied into a loop and creating a master point). One way to at least eliminate the extra stretch created by the knot in one of the legs is to NOT tie the cordelette into a loop, but use a frost knot for creating the master point. in this knot both tail ends of the cord follow through the know, and this doesn’t take much more cord than you’d use when tying the cord into a loop.
Hi Steph. Thank you for this post and I appreciate Dave Furman’s statement. Informative and interesting!
I was shocked to see the strength of a 7mm cord sling is only 135% of the original single-strand (100%), providing the figure was the average or something significant. Theoretically speaking, the doubled-up strand should have the doubled strength, as you rightly pointed out originally, if both the strands were pulled equally and if there was no other weaker point. In reality, as Dave suggested, the assumption is never the case. Here, we can safely ignore the strength reduction with the double-fisherman, as it was allegedly not a break point in Dave’s tests – Figure 1 shows the sling breaks at the point of a karabiner. It is shocking to me as I never expected the cord/sling at the point of karabiner fails at such a low strength – I’d rather expect the point of knot to fail first, say, at 150%.
Now, may I ask you a few questions? In short how trustable is the figure of 135%? How many times has Dave conducted this tests and what was the standard deviation or variance? To get anything statistically significant, 10 trials must be minimum – preferably 100. Is the value of 35% of increase the median or average or what of the trials? Has the single-strand strength of the original cord used in the tests been verified? (The catalogued value can be somewhat different from the batch of cord used in the tests.) And, what were the karabiners used in the tests? A small-diameter one? Had the double-fisherman been tightened well, before the tests were conducted (otherwise it could create the asymmetry of the tension between the two strands when pulled and the knot is tightened, even though the added dynamic feature could be a good thing in the real situation – but not in the static pull test)?
Sorry, if these questions sound nerdy, however I honestly believe it is essential to clarify in order to have any credibility for the figure quoted.
Thank you in advance,
Masa
Hi Masa, here’s what Dave had to say:
We did very few pulls—in this case one of each! That’s nowhere near enough to quantify this reliably—your reader is entirely correct that it would take a lot more testing to really arrive at the solid, concrete figure he or she wants for a general rule of thumb. However, that was not the goal here really, the goal was to have one illustration of how the doubled cord does increase the strength as compared to a single strand, but that it is not double as theory might suggest. As far as how “trustable” the 135% figure is—it is not intended to be “credible” beyond being a realistically possible outcome, and people should NOT use it as any sort of a guideline—the tests used average non-locking biners and new cord with an aggressively hand-tightened knot. This single-loop should have a lot fewer variables than a knotted cordelette, but there are still enough moving pieces (age and abrasion to the cord, how tight the knot is, how well “dressed” the knot is, skinny/fat biners, sharper/rounder biners, etc) that I’m not at all comfortable relying on the stated figure as real “data”, my intention was only to illustrate the principle using what we perceived as a pretty ideal, textbook example and spur people to think about their own equipment and anchors with a healthy dose of conservatism.
Dave Furman
Wow, what a prompt, thorough and honest reply! Thank you very much, Steph and Dave! The things are clarified now, much appreciated.
“with a healthy dose of conservatism” is a wise word we should keep in mind.
Masa
Awesome article! I was looking for an answer to this question and as always it seems your site has an answer Steph!