sliding X

Thanks for the input. i think i'll go with the quad over the sliding x or fig 8 on a bight.

Pontoon wrote:rgold I didn't read everything but in the first paragraph you say some incorrect stuff. In the same book where long and Gaines present the equallette they show test results for the equallette with equal length arms as perfect and the unequal length arms being 1kn different on a factor 1 fall--the latter being on par with a cordellete of equal length arms (and of course better than a cordellete of unequal length arms by quite a bit). The sliding x equal length was actually the most consistent of all methods across multiple tests.

Subsequent testing has thrown some doubt on these results. The following is from Jim Titt indicates something of the scope of the testing...

Jim Titt wrote:To obtain a definitive answer (if there is one) and with help from others in the industry I decided to review all the available information and then do the testing required to give an accurate overview of the potential and pitfalls of every belay system. Currently I have performed over 400 hours of testing and have 700 drop and pull test tables (which normally are between 1500 and 2000 data sets each), 300 force diagrams, 80 graphics and 120 pages of notes, all of which have to be written up in a way the general public could understand. In my normal capacity as a climbing equipment designer and manufacturer I´ve done a number of similar projects but this is by far the worst and most complex.

mountainproject.com/v/acr-a…

...and this summarizes some of the results:

Jim Titt wrote:The problems is that the tests [in Gaines & Long] don´t seem to have been as extensive as they could have been and things were missed. Others including myself have subsequently gone further and the two main conclusions are: Sliding systems are much worse at sharing the force than was assumed (to the extent that some three-point systems give no force whatsoever on one piece). Extension is far more of a problem than was recognised and in the wrong situation will cause equipment failure. Adding extension limiting knots is effectively worthless since either you tie them so close that the anchor cannot equalise or so far apart so that the extension becomes a problem.

mountainproject.com/v/bowli…

I don't think anyone tested one-inch webbing, but I am positive it will lead to worse results in the sliding X configuration because of both the folding and material-to-material contact.

A problem mentioned but glossed over in the tests Long and Gaines report on is that the sliding X had some very bad outlier results. I can't tell which variation on the box-and-whisper plot they are using, so don't know if these outliers are included in the display, but I suspect not. The question of whether such outliers deserve critical attention as worst-case behavior for systems that people's lives depend on is one we could debate. My opinion is I don't want to use something to equalize if every now and then it won't come close, no matter how good its "average" behavior is.

I also find it problematic that the results in Long and Gaines are given in terms of the actual differences between leg tensions, rather than expressing those differences as percentages of the applied load. The absolute values are not very useful as a gauge of system behavior.

Perhaps it is worth mentioning another issue, which is when is equalization appropriate?

1. If you believe each anchor separately can take the load, then all you want is redundancy and there is not only no point in equalizing, the avoidance of higher loads resulting from arm failures in sliding systems makes non-sliding systems better when redundancy is the primary issue.

2. So equalization is only appropriate if you judge the individual pieces to be unable to withstand the load. You are already in sketchy territory if this is the case! Maybe on a multipitch climb you might be forced to deal with such a situation (I certainly have), but might I go so far as to say it is never appropriate for a top-rope situation?

If you soldier on anyway, and if the sliding X performs flawlessly and distributes half the load to each piece, then each of those pieces had better be able to withstand half the load, otherwise you get total anchor failure as the pieces blow sequentially. Of course, if you get one of those outlier situations with the sliding X that provides very unequal distribution, then all bets are off if the pieces aren't judged to be individually adequate, and even worse, the additional load resulting from an arm failure could even blow out a piece that would have held otherwise.

3. Forgetting about the outlier scenario, we can say equalization is appropriate only when your two pieces are individually unable to handle the load but both pieces are individually capable of handling half the load.

This type of judgement is way beyond any climber's estimating abilities, so one could reasonably argue that there is generally no point in trying to equalize with a sliding X.

I dunno, I kind of like the sliding X for sport routes. I trust both bolts, so it's just for redundancy... so you might say, "Use a system that has less extension". But the beautiful thing about the sliding X is how simple, quick, & light it is. 3 carabiners, one sling, and you can pre-tie the knots. Are there any good non-sliding systems that can compete in those regards? I suppose you could tie an overhand on a bight in the middle of the sling instead of using the X...?

Edit: that don't involve clove hitches & the rope, which are great systems for swinging leads on multipitch but not useful for single pitch sport!

Patrick Shyvers wrote:I dunno, I kind of like the sliding X for sport routes. I trust both bolts, so it's just for redundancy... so you might say, "Use a system that has less extension". But the beautiful thing about the sliding X is how simple, quick, & light it is. 3 carabiners, one sling, and you can pre-tie the knots. Are there any good non-sliding systems that can compete in those regards? I suppose you could tie an overhand on a bight in the middle of the sling instead of using the X...?

For a sport route -- are you setting a top-rope anchor or just lowering-off?

Quickest & Easiest: two draws, with the rope-side biners opposite & opposed.

If you're planning to top-rope through this anchor, so the two draws isn't really appropriate:

4 locking biners, one sling, over-hand knot on the double-bite. Two lockers through the power-point (opposite & opposed) for the rope. This can be pre-tied.

Problem with sliding X: if you use just one sling, the sling itself isn't redundant. (overhand, both arms are indepedent, so everything is redundant). Also, using only three biners, the rope-side biner isn't redundant. For a single lower-off and 2nd, this may be ok -- but if you're going to have a few people top-roping, I'd say that is not adequate.

Sliding X works just fine.

5 words. All you needed to know. 5 words.

Pontoon wrote:rgold I didn't read everything but in the first paragraph you say some incorrect stuff.

Rgold and Jimm Titt have some the best understanding of climbing physics out there.

John Long had some plainly flawed understanding in his book and some of his testing methodology and conclusions were outright wrong. Most notably his conclusion that "shock-loading" is a myth is dangerously wrong. As soon as there is mass involved at the belay (ie a belayer), then extension will result in drastically increased loads from (in layman terms) "shock-loading".

rgold wrote: Subsequent testing has thrown some doubt on these results.

Thank you rgold. This is the sort of researched, well-argued post that you so often provide, and that adds such value to these forum. Thank you.

And, then there is:

Buff Johnson wrote:Sliding X works just fine. 5 words. All you needed to know. 5 words.

Read what rgold says. He knows of what he speaks.

rgold thanks that's interesting info. I haven't seen that forum post before. I do find it interesting that he says the sliding X can see appalling results without seeing the "clutch" effect, but then he proceeds to recommend a cordalette (which has also been well demonstrated to have appalling results). I'd really be curious to see his testing methodology and a writeup before coming to any final conclusions.

When you try to break down situations where equalization is appropriate, I think you oversimplify a bit. It's not just about whether each individual piece will hold. First, since you don't actually know whether a piece will hold until you have the hindsight of a fall, I believe it's best to play it safe and attempt to equalize. Long and Gaines bring up what seems like a particularly reasonable situation where self equalization is important: an anchor which should withstand force from many directions. Let's say the leader has to traverse away from the belay. If the leader falls right away, it's a downward force. If the leader falls traversing, it's a sideways force followed by every direction until it's downwards (or a bit past downwards). Further, If it were me climbing and I took a whipper up high, I might pull my small belayer upwards, and she would cause an upward force on the anchor if her harness isn't anchored in. This is why I like the equalette (which I use like a sliding X when using a reverso to belay the follower) and the sliding X in scenarios like this.

I'm also really interested to know why he says that extension is more of a problem than previously thought, because it really seems intuitive to me that Long and Gaines' testing/assertion that it has a minimal effect is correct.

Also, I can't exactly take his recommendation of using the rope most of the time because when I'm climbing with my girlfriend we don't swing leads--I lead it all.

Pontoon says: "Also, I can't exactly take his recommendation of using the rope most of the time because when I'm climbing with my girlfriend we don't swing leads--I lead it all.


I have done a fair bit of climbing with John, he, we all learned what I call the "California Style" that boild down to this: very miminal gear.

Use some slings and the rope to tie in, almost Zero equilization because we climb on bomber granite and the thought was: one piece is the belay, one is the back up and the other one is the back-up to the back up.

Pretty ezy to hook the 2nd in when they get to you, restack the cord and your off.

And listen to rgold, he has been climbing longer than most people have been alive.

Buff has a point also.

so if sliding Xs are not great then what would you suggest using if you need the anchor to be multi-directional. i mean i am using the sliding x more for multi-directional forces than for general equalization as the trees appear to be bomber. it's just i was taught that, for top-roping, to use more than one tree if possible for the sake of redundancy. I'm thinking i will use a loop of 7mm cord tied together with a triple fisherman's for my sliding X as it seems to produce less friction than the 1" thick nylon sling, which is the thinnest i have in the length i need. would doubling the cord up make it redundant?

eli poss wrote:so if sliding Xs are not great then what would you suggest using if you need the anchor to be multi-directional. i mean i am using the sliding x more for multi-directional forces than for general equalization as the trees appear to be bomber. it's just i was taught that, for top-roping, to use more than one tree if possible for the sake of redundancy. I'm thinking i will use a loop of 7mm cord tied together with a triple fisherman's for my sliding X as it seems to produce less friction than the 1" thick nylon sling, which is the thinnest i have in the length i need. would doubling the cord up make it redundant?

Why do you need it to be multidirectional if the trees are solid?

From the American Safe Climbing Association:

"Note: The Sliding X

Many climbers use a "sliding X" to equalize two pieces - ususally beginner climbers with bolt anchors. You should NEVER use this except in two specialized cases (see below). While the sliding X does equalize the pieces, it assumes that neither could break, since if one does break, there is severe extension in the system - enough that it would likely cause the carabiners to break. Since it assumes neither piece would break, it's a stupid system - if neither would break, there's no need for equalization. If one might break, then there is WAY too much extension. This is why many call it the "death X." Instead, use one sling off of each bolt or piece. You can tie one shorter to approximately equalize the pieces if needed.

The two cases where the sliding X is used:

equalizing tenuous pieces in a larger anchor - for instance, two poor nuts in a large natural pro anchor. The nuts are equalized, then the sliding X is equalized with other pieces through a cordelette, webolette, or other non-extending method.
equalizing two very tenuous pieces in extreme aid - for instance, a hook and a bashie on A4 terrain."

Every climbing method I have read about usually gets WAY more flak than it probably warrants.

If you're going to climb the route only twice, then a sliding X will work in almost every situation. Sure it isn't as good as other anchors(which you should really learn to build), but most of the time you're not going to be climbing a route that has some serious pendulum problems(or actual equalization issues). If you are, you should recognize that and build the right anchor for the situation.

If you still really don't like the thought of a different anchor, just get a cordelette and tie it around the tree. Use a PAS or double length sling, or better yet you should you use your end of the rope with a clove hitch. Then belay your girlfriend up.

Pontoon wrote: I'm also really interested to know why he says that extension is more of a problem than previously thought, because it really seems intuitive to me that Long and Gaines' testing/assertion that it has a minimal effect is correct.

What happened is that they didn't test an important situation. They set up a leader fall of some size onto the anchor and then measured what happened if one arm of the anchor failed. So in other words, you have a fall with a certain H/L fall-factor ratio, you increase H by a few inches x by letting an arm fail, with x a relatively insignificant quantity compared to both H and L, and then the new (H+x)/L fall factor is virtually the same as the original one and you conclude, predictably, that extension has no effect.

But consider a perfectly plausible scenario in which the leader factor-2's onto the belay and pulls the belayer off. Forgetting about the interesting dynamics of having two fallers separated by a length of rope (coupled harmonic oscillators), you have two falling climbers whose fall energy now has to be absorbed by the belayer's tie-in. Now that same quantity x of extension is the total height of the fall, and the fall factor becomes approximately x/T (assuming the sling material doesn't stretch much), where T is the length of the belayer's tie-in. It is easy to see that this situation can lead to very high fall factors, even ones greater than 2 if the belay tie-in is very short (say 1") and the extension is rather long (say 3", so now a fall-factor of 3), and we now have the sum of the two climber's weights to contend with, not one. In this situation, there is no question that the effect of extension on the anchor load could be very dramatic.

But I don't think this is all. The DAV did tests in 2009 which were similar to the tests performed by Long and Gaines but concluded that extension had a significant effect. I recall a load "40% higher" but am not sure any more how the original baseline load was defined. Obviously, you would expect the load on a single anchor piece surviving an arm failure to be roughly double the load it would have experienced had the system remained intact and equalized properly. So the real question is how much (if at all) higher than double the non-failure load to a single piece is the load caused by extension. I just can't recall if the DAV tests really addressed this question, although it seems hard to imagine they would have considered anything else.

patto wrote: John Long had some plainly flawed understanding in his book and some of his testing methodology and conclusions were outright wrong. Most notably his conclusion that "shock-loading" is a myth is dangerously wrong. As soon as there is mass involved at the belay (ie a belayer), then extension will result in drastically increased loads from (in layman terms) "shock-loading".

No

redlude97 wrote: Why do you need it to be multidirectional if the trees are solid?

because there is a possibility that i may fall in a place that would cause it to load from one side or the other rather than directly below the anchor. the line i'm looking at traverses left for a little bit about 10 ft from the top and then goes up about 5 ft and then traverses back right.

are there any other anchor set-ups that are multi-directional because the sliding X is the only one that i know of

rgold wrote: What happened is that they didn't test an important situation. They set up a leader fall of some size onto the anchor and then measured what happened if one arm of the anchor failed. So in other words, you have a fall with a certain H/L fall-factor ratio, you increase H by a few inches x by letting an arm fail, with x a relatively insignificant quantity compared to both H and L, and then the new (H+x)/L fall factor is virtually the same as the original one and you conclude, predictably, that extension has no effect. But consider a perfectly plausible scenario in which the leader factor-2's onto the belay and pulls the belayer off. Forgetting about the interesting dynamics of having two fallers separated by a length of rope (coupled harmonic oscillators), you have two falling climbers whose fall energy now has to be absorbed by the belayer's tie-in. Now that same quantity x of extension is the total height of the fall, and the fall factor becomes approximately x/T (assuming the sling material doesn't stretch much), where T is the length of the belayer's tie-in. It is easy to see that this situation can lead to very high fall factors, even ones greater than 2 if the belay tie-in is very short (say 1") and the extension is rather long (say 3", so now a fall-factor of 3), and we now have the sum of the two climber's weights to contend with, not one.

Wow, you exhaust me. This doesn't add up. But, it is late. I'm happy to dissect tomorrow. But, I will give you time to restate it.

eli, I think your thought process is fine. Equalization is a good concept to implement. You have asked a much more complicated question than you thought. Actually, you asked several questions, some intentionally, and some inadvertently.

First, what knot to join webbing and cords. Water knot for webbing is best for short term use and are very secure when weighted, but can come loose over time when not weighted. Overhand or double fisherman's for cord.

Second, which webbing or cord gives best equalization. "Clutch effect" diminishes the effectiveness of sliding x with all materials. Thicker materials have greater clutch effects. Skinny and slick materials give less clutch effect and better equalization.

Third, should I use sliding x or not. This is a more complicated answer, as you have seen. I'll comment more later.

On two good bolts for multi the sliding x works just fine

Its very fast as theres no knots to untie latter

;)

Greg D wrote: This doesn't add up. I'm happy to dissect tomorrow. But, I will give you time to restate it.

I'm happy to try to clarify, but its on you to point why what I said "doesn't add up." Meanwhile, I'm not sure how to restate it to make it any clearer than it is. Ok---how about I get rid of the algebra?

The analysis in Long and Gaines didn't consider the loads involved if the belayer is pulled off the stance. In fact they didn't model a belayer presence at all, what they tested was the performance of the sliding X between two pieces that are holding a leader fall. If a belayer is pulled off the stance because of anchor elongation, the main rope is no longer the fall-absorbing agent, a task that passes to the relatively short piece of rope that is the belayer's tie-in. The critical factor in determining peak tension in the belay tie-in is the usual fall-factor adapted to this situation, which becomes the anchor extension (fall height) divided by the tie-in length (amount of rope in system). As this ratio can be quite large, and in view of the fact that the system now has to arrest the fall weight of two climbers, the possible loads resulting from anchor extension can be very substantial.

My second comment, but it was based on memory that might be faulty, is that the DAV essentially replicated the Long and Gaines set-up but concluded there was a significant load increase when anchor extension occurred. The rest of that comment was musing about how one ought to define "load increase" in such a situation.