Mountain Project Logo

Equalizing the Middle Anchor Leg

David Coley · · UK · Joined Oct 2013 · Points: 70

Joe, the answer is simple: don't worry about it. The pull is not often vertically down the cliff when someone falls off or weights the anchor. It will be in one direction when the leader is hanging from it, in another when she reaches for her water bottle which hanging from the top piece, another when the second falls when he hasn't removed the last runner, another if he has removed it. If you have two seconds the pull can be in various directions, if the leader falls of the next pitch it will be in one direction, then another if using double ropes and she has a piece to the side on the other rope.

This means that one or more of the arms will be slack, so no point in trying to set it "just so".

As keeps getting said. If you really think the anchor poor, use the rope not the cordellete.

jktinst · · Unknown Hometown · Joined Apr 2012 · Points: 55
rgold wrote:... Every test I've seen suggests that with three anchor points sliding systems can actually be worse than fixed-arm ones...
Sorry for compounding the thread drift but I’ve been looking for years for solid test data comparing dynamically-equalizing and statically-equalized three point anchors and have yet to find them. So please, please, provide references. The McKently tests were done on multiple points but in the context of rescue anchors and using only static materials. The comparisons that I’ve seen used only two-point anchors (mainly the Long/Gaines/Ewing tests and DAV test) and I did not see anything in those to suggest that three-point sliding systems based on similar principles would be worse than fixed-arm ones.
rgold · · Poughkeepsie, NY · Joined Feb 2008 · Points: 526

Jim Titt has some results a while back, but I don't have a link.

Other results were from several years ago, I'll see if I can dig them up.

Jim Titt · · Germany · Joined Nov 2009 · Points: 490
jktinst wrote: Sorry for compounding the thread drift but I’ve been looking for years for solid test data comparing dynamically-equalizing and statically-equalized three point anchors and have yet to find them. So please, please, provide references. The McKently tests were done on multiple points but in the context of rescue anchors and using only static materials. The comparisons that I’ve seen used only two-point anchors (mainly the Long/Gaines/Ewing tests and DAV test) and I did not see anything in those to suggest that three-point sliding systems based on similar principles would be worse than fixed-arm ones.
Marc Beverly also did tests on multi-point equalised anchors with some pretty miserable results and we also did some random climber tester in the field to see what was actually being tied outside, not in the lab. Very poor!
The sliding systems are not particularly better and sometimes worse, they all range from bad to fairly bad to bad at actually distributing the force equally. The better ones gave a load distibution of 15%/30%/55% and the worst one 0%/17%/83%.
To get equal load distribution it is a matter of adjusting the leg lengths within millimeters and is hard enough to do when you can read the force on each piece, impossible in normal circumstances.
Healyje · · PDX · Joined Jan 2006 · Points: 422

The one and only original over-the-top more-than-anyone-ought-to-ever-want-to-know about anchor equalization thread

rgold · · Poughkeepsie, NY · Joined Feb 2008 · Points: 526

Haha, Nerds Gone Wild. Yeah, I said some things there I'd rather not have to defend any more.

jktinst · · Unknown Hometown · Joined Apr 2012 · Points: 55

I had not seen the Beverly paper but looking at it know (if it is the “multi-point pre-equalized anchors” paper available on his web site), I see that it only looked at different configurations of statically-equalized cordelettes.

It does state the following in its introduction: “the ‘magic x’ has fallen out of favor for several reasons. It has been shown that a two point equalized anchor does not share the load as equitable (sic) as a two-point non-equalized anchor (ref1). Only a specialized anodized figure 8 device showed any reasonable equalization in the ‘magic x’ configuration (ref2)”. However, while the paper quoted in ref1 does indeed come to that conclusion, it was not from testing the sliding X but some other sliding system that I’ve never seen debated or used in the climbing community (possibly because I haven't been on the forums long enough) and the second reference is some unavailable personal testing results.

Regarding your own testing, Jim, I have read your claims of poor-to-absolutely dreadful distribution ratios for sliding anchors in forum posts going back 5 or 6 years. I don’t doubt in the least that you’ve obtained the results you mention for certain rigging set-ups in your particular test system but, again, I’ve not seen the kind of detailed test report that would allow me to understand exactly what was tested and how, and to attempt to reconcile the results with the other detailed test reports I’ve seen.

Jim Titt · · Germany · Joined Nov 2009 · Points: 490

That the sliding X doesn´t equalise was shown back in 1984 if I remember correctly so nobody bothers to give any particular references.
Once you accept that the friction over the karabiner will cause a load split of around 1.6:1 on the pieces (the karabiner factor of 1.6.1 for nylon rope over a karabiner was presented in a paper to the UIAA in 1964) then it follows that this will occur in a dynamic belay around each karabiner and the common setups display this perfectly. Various materials and wrap angles alter the actual values but the commonly used cords are generally higher than this around 1.7:1 and this is an unalterable physical effect despite claims by some to "perfectly equalise".
I´ve tested all the common system for 3 point anchors, the sliding X with two karabiners, the Moosalette, ACR and the Alpine Equaliser. None achieve better load distribution than 15%/30%/55%.
Unfortunately the sheer number of tests that have to be done on the various combinations of materials and systems and then the dynamic tests of the various failure modes mean the write-up I originally did is over 120 pages long, the condensed version is 36 sides and offers little more information than saying:-
"Equalising systems don´t work very well and the penalty for a piece failing can be extremely high."

Buff Johnson · · Unknown Hometown · Joined Dec 2005 · Points: 1,145

Even though what you're looking at might be perceived as "miserable" in terms of 'perfect equalization,' the rigging is still adequate to handle the load distribution needed to maintain itself in a hit.

Much ado over not much, again.

JoeLars41 · · Unknown Hometown · Joined Mar 2015 · Points: 60

This is a terrific thread! Thank you everyone for contributing!

I must be a nerd gone wild because I love this discussion.

rgold · · Poughkeepsie, NY · Joined Feb 2008 · Points: 526
Buff Johnson wrote:Even though what you're looking at might be perceived as "miserable" in terms of 'perfect equalization,' the rigging is still adequate to handle the load distribution needed to maintain itself in a hit.
True dat, except for those approximately once-every-ten-years total anchor failures I mentioned upthread. As long as you're not one of those unfortunate souls, you're good to go.

But I think the discussion is really about whether we are going to live in the real world or some fantasy, and how our language determines that. This particular tangent got going because of the claim/troll,

Greg D wrote: [sliding systems with three anchor points] work quite well with an equalette and someone competent at this technique.
The equalette-based three-anchor systems I've seen in anchor books guarantee worse than 50-25-25 distribution before any deleterious friction effects happen and so are as bad and probably worse than what I always thought (hoped) the best candidate was, the ACR. And the idea that being "competent at the technique" can overcome the intrinsic limitations imposed by real-world friction is, sadly, a fantasy. Friction never sleeps.

Rigging systems can never be counted on to equalize, and it is too bad we can't get rid of that word forever and substitute "distribute." Still, there are reasons to try for as good a distribution as possible, which brings us back to discussions like this, relevant or not.

I think a practical take-away from the testing is that no matter how we rig multi-point anchors, we should anticipate that one of the anchor points will end up taking at least half of the total load. It makes sense to try for equalization in distributed rigging---just don't confuse attempting with succeeding.
jktinst · · Unknown Hometown · Joined Apr 2012 · Points: 55

For my part, I‘m glad that Long/Ewing and the DAV did not pay too much attention to those old well-known facts when they considered whether to embark on their tests.

To try and make up for the thread drift somewhat, I would suggest to the OP to check out the following thread on RC.com

rockclimbing.com/cgi-bin/fo…;post=2589801;page=1;sb=post_latest_reply;so=ASC;mh=25;

Mathias · · Loveland, CO · Joined Jun 2014 · Points: 306

I had always wondered if a 3 piece equalized system could be achieve by doing this:

Equalized and limited sling between point A and point B. Same setup with a separate sling from point B to point C. Then and a third equalized and limited sling from the central points of the first two slings to a master point in the middle.

Overly complicated? Maybe, considering the limiters would need to be tied or adjusted for each anchor. But it would be an equalized 3 point anchor wouldn't it?

Greg D · · Here · Joined Apr 2006 · Points: 883
rgold wrote: True dat, except for those approximately once-every-ten-years total anchor failures I mentioned upthread. As long as you're not one of those unfortunate souls, you're good to go. But I think the discussion is really about whether we are going to live in the real world or some fantasy, and how our language determines that. This particular tangent got going because of the claim/troll, The equalette-based three-anchor systems I've seen in anchor books guarantee worse than 50-25-25 distribution before any deleterious friction effects happen and so are as bad and probably worse than what I always thought (hoped) the best candidate was, the ACR. And the idea that being "competent at the technique" can overcome the intrinsic limitations imposed by real-world friction is, sadly, a fantasy. Friction never sleeps. Rigging systems can never be counted on to equalize, and it is too bad we can't get rid of that word forever and substitute "distribute." Still, there are reasons to try for as good a distribution as possible, which brings us back to discussions like this, relevant or not. I think a practical take-away from the testing is that no matter how we rig multi-point anchors, we should anticipate that one of the anchor points will end up taking at least half of the total load. It makes sense to try for equalization in distributed rigging---just don't confuse attempting with succeeding.
Thank you for appearing to disagree with me again when actually supporting my points. In another thread I said thousands to be conservative and you said millions. Minutia I say. I'm sorry you have resorted to name calling. Everyone knows I'm not a troll. Although, my sarcasm can be a bit much for the more sensitive guys.

I don't have time this morning cause I have a major work day balancing large loads circuits and controls.But, much worse than 50-25-25 is vague. But this distribution can be a benefit when one piece appears quite solid compared to the other two. Yet vertical cordolette numbers are even worse. So what are we actually comparing. Add in three clove hitches and a nylon cord common in cordolettes and you have addition energy absorption. Everyone knows how a clove and nylon absorb. See DMM's video breaking material and how the clove reduces load. You see, when someone makes a blanket statement about one component they are not taking in all aspects of what is truly going on.

The first line of defense is to avoid worst case scenarios on your anchor such as factor 2's.
The second line of defense is to ensure you anchor is "good enough" to handle worst case scenarios.
Its the system as a whole that keeps us alive, not a single link or technique.

The truth is, a very, very small percentage of anchors are put to the test which is why most people find these discussions pointless. "Well my anchors have worked just fine". "I've been using bla bla bla technique for 30 years and I'm still alive". You know these claims. So folks, ask yourself, have yo done everything possible to avoid worst case scenarios? is your anchor "good enough" for worst case scenarios? And hope you never have to test it.

But, its kinda like "my air bag in my car works just fine, I'm sure". But, how many have you actual tested it. Maybe that's a bad analogy. But, you get what I'm saying. Bring on some more wind. All things are not equal for all people
csproul · · Pittsboro...sort of, NC · Joined Dec 2009 · Points: 330

At least RGold has some comprehensible evidence to support his "wind"...you OTOH...

Jim Titt · · Germany · Joined Nov 2009 · Points: 490
Mathias wrote:I had always wondered if a 3 piece equalized system could be achieve by doing this: Equalized and limited sling between point A and point B. Same setup with a separate sling from point B to point C. Then and a third equalized and limited sling from the central points of the first two slings to a master point in the middle. Overly complicated? Maybe, considering the limiters would need to be tied or adjusted for each anchor. But it would be an equalized 3 point anchor wouldn't it?
Are you talking about an equalised anchor (one that you used your skills to get the tensions in each leg the same) or a dynamically equalising anchor?
Either way it´s unlikely to produce equal loads on the pieces. The lowest sling splits the load 50/50 (if you are lucky/skilled) and the upper slings 50/50 again so 25% on each leg. The outer pieces are thus loaded 25% each and the middle one 50%. This simple principle is death to nearly all the multi-piece equalising systems promoted.
Mathias · · Loveland, CO · Joined Jun 2014 · Points: 306
Jim Titt wrote: Are you talking about an equalised anchor (one that you used your skills to get the tensions in each leg the same) or a dynamically equalising anchor? Either way it´s unlikely to produce equal loads on the pieces. The lowest sling splits the load 50/50 (if you are lucky/skilled) and the upper slings 50/50 again so 25% on each leg. The outer pieces are thus loaded 25% each and the middle one 50%. This simple principle is death to nearly all the multi-piece equalising systems promoted.
Jim, I should have been more careful in my writing. I was thinking of a self equalizing anchor that could deal with the load moving from side to side. Applications would include TR anchor (though I'd probably use 4 pieces for that) and maybe for the start of a traversing section.

But as you mention equal loads I have a question about that: Given real world placements, is it ever actually possible (factoring in cord or sling angles and the extra forces they create) to get a system that will actually distribute the load truly equally over all placements?

I'm not arguing your assessment of the anchor I described. It makes sense. How about an equalized and limited sling from point A to point B, and then an equalized sling from that center point to point C? Spreading the load more equally onto the placements with that method?
Marty C · · Herndon, VA · Joined Aug 2008 · Points: 70

At the 2012 International Technical Rescue Symposium, Mike Gibbs (from Rigging for Rescue) presented a paper "Considerations for equalizing multi-point anchor systems."

His basic claim was that sometimes, equalization was not the goal, but to better match the strength of the arms of the rigging to the strength of the pro (i.e. The arm that takes the largest percentage of the load goes to the strongest piece of gear, and the arms that take the lesser % of load goes to the weaker pieces of gear).

He presented several techniques to accomplish this:
- using both single and double strands in the rigging
- various techniques to "shorten" a leg so as to force more load to other strands
- etc.

I found it an interesting read.

jktinst · · Unknown Hometown · Joined Apr 2012 · Points: 55

I’ll take the bait and proceed on the assumption that you’re not trolling. I don’t think that Jim will mind my answering for him. This:

Mathias wrote:...How about an equalized and limited sling from point A to point B, and then an equalized sling from that center point to point C? Spreading the load more equally onto the placements with that method?
is exactly what rgold was referring to as giving a 25:25:50 theoretical distribution, which is why it was ironic that your first question was regarding a system that would be more complicated to set up only to achieve in theory 25:50:25.

As for this question:

Mathias wrote: ... Given real world placements, is it ever actually possible (factoring in cord or sling angles and the extra forces they create) to get a system that will actually distribute the load truly equally over all placements?...
Jim ’s already given his answer in his earlier posts on this very thread (and in a good number of other threads before that). He's obtained some of those piss-poor actual distributions by testing systems whose theoretical distribution was 33:33:33 or pretty close to that (like 37.5:25:37.5).
jktinst · · Unknown Hometown · Joined Apr 2012 · Points: 55
Marty C wrote:... sometimes, equalization was not the goal, but to better match the strength of the arms of the rigging to the strength of the pro (i.e. The arm that takes the largest percentage of the load goes to the strongest piece of gear,...
That's a fun concept to play with but remember that estimating the actual relative strength of the different pros you’re proposing to use for your anchor is a crapshoot. The stronger-looking one could very well turn out to be the weakest and vice-versa.

Even if you do have a clear strongest (like a single bolt that you want to back up with a couple of solid-looking clean pro pieces), you should be very careful when adapting rescue anchor rigging techniques for climbing use. For rescue, the direction of pull is pretty well determined and generally won’t change (unless a pro blows). For climbing, even if the most likely direction of pull appears to be pretty obvious, the actual one could be different. And the direction of pull will have a major impact on which pro gets loaded first and most.
Guideline #1: Don't be a jerk.

General Climbing
Post a Reply to "Equalizing the Middle Anchor Leg"

Log In to Reply

Join the Community

Create your FREE account today!
Already have an account? Login to close this notice.

Get Started.