Magic X?

I have recently noticed a large division when it comes to people's practices while setting up anchor systems with slings. I would like to hear some sound advice from the MP community on this. Basically, the questions is whether or not tying your master point on a bite vs the magic or sliding x is safer. The argument people use for the magic x is that it can re-equalize if one point fails, maintaining the direction of pull. The argument people use for the bit (overhand or figure eight or whatever) is that it won't shock load the system by not re-equalizing and that that is preferable to shock loading your sling. I also heard some debate about dynema versus nylon in terms of shock strength, that would also be nice if someone cleared up. , personally, was trained to use multiple slings or (whatever you are using) and tie a double figure eight on a bite to make a top rope safe, but I want to know what you guys do to be safe. Please share your experience and knowledge on this question. Happy climbing.

Last time I took magic X I woke up in the back of a nightclub with a massive headache. And their girl told me it was aspirin... bullshit.

bolted anchor = magical ex.
Trad anchor - overhand.

That's what I do.

For an analysis and graph of when a sliding x might help, go to multipitchclimbing.com then chapter 6, then section 8 "the sliding x"

If you want to know what we do in Europe given two high quality bolts in good rock, look at the section on Banshee belays.

I used to buy into the dynamic equalization bullcrap for anchors. Now I believe that the load should be "distributed" as best as possible and that should be NO extension if any piece blows.

Comments from the American Safe Climbing Association:
safeclimbing.org/education/…

Like everything in climbing it's really situational. If you have a climb that really and truly changes direction enough that it will become un-equalized during the course of climbing, a sliding x is probably to your advantage, however, generally I would say that the consensus is that a sliding x has more cons that pros, but again, situational.

A sliding x is all fine and good until a piece blows and you shockload the system. An easy way to minimize the extension in the case of failure is to tie limiter knots. This is a pretty good picture of that: i255.photobucket.com/albums…

Personally, I go for the pre-equalized cordelette anchors that won't have any extension in the case of failure.

" A sliding x is all fine and good until a piece blows and you shockload the system. "

Can you describe/define "shockload"?

Thanks!

This is a good primer for impact load calculations. You'll need an eighth grade math education to understand the equations.

roymech.co.uk/Useful_Tables…

David Coley wrote:For an analysis and graph of when a sliding x might help, go to multipitchclimbing.com then chapter 6, then section 8 "the sliding x"...

I've tried hard to stay out of the equalization vs no-extension debates in the last while but this has prompted me to jump back in. I'm afraid that I have yet to hear a convincing argument against the X with extension-limiting knots except the fact that it takes longer to build and the extension limitation also limits the range within which equalization works. I used to think that since outside of that range, the X will behave like a PP, nothing was lost by using the X. However, thinking about it some more I am realizing that this occurs only when the load is applied outside of the range, forcing the CP biner against the load-limiting knot on the long arm side. This will force the short arm to take the whole load, as in the PP situation. However, if the shift occurred in the other direction, the long arm will still be able to feed slack to the short arm, equalizing the load. In practical terms, what this means is that the extension-limiting knots could be placed judiciously on either side of the most likely direction of the load (selected the same way one selects the position of the PP) but with the knot quite close to that point on the short arm side and a bit further away on the longer arm side in order to maximize the range of equalization while still keeping potential extension to a minimum.

That multipitchclimbing.com theoretical graph of PP vs X really doesn't seem to be particurlarly useful because it compares the PP at it practical worst case (short arm takes entire load) with the X at its ideal best (perfect equalization). However, if you ignore that inconvenient fact, what the graph shows is that the X systematically has a better chance of holding than the PP (only a little better at low loads but much better at high loads) and that the PP systematically has a better chance of having one arm fail (again: only slightly better at low loads but much better at high loads).

Looking at the arguments made in the list of 5 points:

2. "If one piece can take the load independently, there is no point in trying to equalise, you should concentrate on redundancy, minimising the drop if the other piece blows and multi-directionality". But it is clear that one can never really know if one piece can actually take the whole load independently, even with bolts. If we did, we would only need to use that one pro for the anchor.

3. Equalisation plays no role if one piece has less than half the strength of another. Imagine one arm of a perfectly equalising belay can take 6 kN and the other 3 kN. It might be tempting to think this belay can take 9 kN. However, if a 6.1 kN load is applied each arm with receive 3.05 kN, the weakest arm will fail, and then the stronger as it receives the full 6.1 kN

4. This leads to the rule that the maximum theoretical strength of a two-piece sliding-X is NOT the sum of the strengths of the arms, but only twice the strength of the weakest arm. Understanding this is key (and most climber’s don’t): unless you know both arms are strong, a sliding-X is probably the last way you should be building a belay.

This rule is a useful reminder for people who think that they can just add up the pro strengths but it hardly makes an argument against the X. Linking only two pieces of widely different strengths is not a good idea with X or PP. This is where one really should have a 3rd pro but let's stay with this scenario for the sake of comparing PP & X. With the X and a 6.1kN load, point 3 states that the 6kN pro will end up with the entire 6.1kN load after blowing the 3kN pro but that doesn't take into account the energy absorbed by that pro before it blows.

Taking the same pro strength numbers with a PP anchor (3 & 6 kN) but a higher overall load, say 8.1kN: if the 3kN pro happens to have the shorter arm, it will get the whole load and blow. Will this 3 kN pro blowing at 8.1kN absorb enough energy to bring the remaining load below the 6 kN cut-off for the stronger pro to hold? I don't know but it seems pretty chancy. If that stronger pro has the shorter arm, it will take the whole load and blow as well. Will that 6kN pro blowing at 8.1kN absorb enough that the remaining 3kN pro will be an effective back-up after that? Again: don't know but seems pretty chancy.

With the X, keeping the same numbers but assuming a 50% transmission efficiency at the CP biner, resulting in the longer arm getting only 25% of the load (ie a 1:3 ratio, which is much less advantageous than the 1:2 ratio considered in the first point of the list). Since we're trying to be more realistic, let's also assume that fairly narrow extension-limiting knots essentially eliminate potential shock-loading. If the short arm is on the 3kN pro, it will get a 5.98kN load and blow. Will it absorb enough to bring the remaining load under the 6kN bar for the stronger pro to hold? Still seems pretty chancy but somewhat less so than the equivalent PP scenario. If the short arm is on the 6kN pro, it will get the 5.98kN load while the 3kN pro gets the 2.02 share and then the entire anchor will hold without anything blowing.

So it seems to me that working through these kinds of scenarios really makes the case in favour of the X!

The sliding X has it´s place in a very limited number of scenarios and should be the system of last resort and ONLY if you fully understand what you are doing and the implications. It won´t make a silk purse out of a sows ear.
Or use it for top-rope anchoring on two bolts where it doesn´t matter anyway what you do except you´ll look like a dork that doesn´t understand redundancy.

The work of failure of an individual piece is completely unknown and always will be until that particular piece fails, generally compared with the amount of energy in the system it will be negligeable anyway so we don´t consider it in working out the forces. It has anyway been shown that the effect of sequential failure tends towards higher ultimate impact forces as the system stiffness is increased for the second impact.
The balance of limiter knots and load shift against extension is a fine one, up to an angular shift of about 15° no sliding of the central karabiner will occur so deciding where the knots go is generally pure guesswork, it is anyway the case that the faller will be swung to under the belay rather than pull the belay karabiner across since the horizontal force component is too small to overcome the vertical component and corresponding friction.
Impact forces due to extension- these can be enormously high, I have recorded 32 times the initial load and others have similar results.

Thanks everyone for their replies. This has given me a much better sense of what to do and has hopefully helped others who have stumbled upon this thread. I appreciate everyone's input. Love seeing the MP community share their expertise. Also, shout out to jktinst for that super detailed response.

Here's a tidbit from the BD website...

blackdiamondequipment.com/e…

Has anybody here actually had a component of their anchor fail?

Edit: Except Tico... I'm pretty sure there's nothing that hasn't happened to Tico.

I just got in a facebook fight over this. No, I haven't had an anchor failure. But they seem to happen, the Warpy Moople and DNB incidents are one's tht come to the top of my old gray head.

jktinst wrote:4. This leads to the rule that the maximum theoretical strength of a two-piece sliding-X is NOT the sum of the strengths of the arms, but only twice the strength of the weakest arm.

The minimum strength of a two-piece sliding x is twice the strength of the weakest piece (assuming perfect equalization...). Think about an extreme example. One leg is 1kN, the other is 20kN. Clearly, this anchor can hold more than 2kN, even though the weak leg will blow. How much more is complicated, given that the amount of force delivered to the other leg after the weak one blows depends on the dynamics of the rest of the system (amount of extension, rope stretch).

Anyway, I used to use the sliding x for bolted anchors, but a sling with an overhand is just as easy to set up and more versatile as it has a shelf. I have also been using a variation on the banshee belay more recently when the anchor is two good bolts with chains: Clip a quickdraw between the bolts, clove in to one chain, belay off the other chain. Simple, quick, and clean.

I rarely use any anchoring method with the word "magic" in the name.

Mike Marmar wrote: The minimum strength of a two-piece sliding x is twice the strength of the weakest piece (assuming perfect equalization...). Think about an extreme example. One leg is 1kN, the other is 20kN. Clearly, this anchor can hold more than 2kN, even though the weak leg will blow. How much more is complicated, given that the amount of force delivered to the other leg after the weak one blows depends on the dynamics of the rest of the system (amount of extension, rope stretch). Anyway, I used to use the sliding x for bolted anchors, but a sling with an overhand is just as easy to set up and more versatile as it has a shelf. I have also been using a variation on the banshee belay more recently when the anchor is two good bolts with chains: Clip a quickdraw between the bolts, clove in to one chain, belay off the other chain. Simple, quick, and clean.

I am no expert in this but i would assume in the real world you will never have perfect balance so always assume your anchor is only as good as your single best piece... so if I have 2 points that only hold 1kN i would assume the anchor will only hold at best 1kN even though with some balancing it will likely hold a little better.

You will never balance it to the point that 2 1kN points will hold 2kN.

Mike Marmar wrote: The minimum strength of a two-piece sliding x is twice the strength of the weakest piece (assuming perfect equalization...). Think about an extreme example. One leg is 1kN, the other is 20kN. Clearly, this anchor can hold more than 2kN, even though the weak leg will blow. How much more is complicated, given that the amount of force delivered to the other leg after the weak one blows depends on the dynamics of the rest of the system (amount of extension, rope stretch).

Correct, originally I wrote this as "For a two-piece anchor true equalisation anyway performs no useful function when the strength of one piece is half or less than half of the other," which the authors quoted directly but then turned into "This leads to the rule that the maximum theoretical strength of a two-piece sliding-X is NOT the sum of the strengths of the arms, but only twice the strength of the weakest arm." which is clearly wrong.

Another problem with the sliding X: if I want redundancy in the sling material I'm using (not just the pair of bolts), then I need to setup two slings in sliding X formation -- this makes things far messier than just a single sling with power-point. And, it adds more friction for the auto-equalize correction making it even less likely to actually self-equalize.

My personal preference is a power-point setup using nylon (not dyneema) slings, tied as close to equalized as I can eyeball. I like nylon as it loses less strength when knotted than dyneema, and doesn't tend to weld as much when weighted. Also, in theory, I may get some load-balancing as one arm of the nylon stretches under load then the 2nd arm kicks in.

David Gibbs wrote:Another problem with the sliding X: if I want redundancy in the sling material I'm using (not just the pair of bolts), then I need to setup two slings in sliding X formation...

With extension-limiting knots around the Central biner and a clove at the end of each arm, you have a fully internally redundant system that will not be compromised by cutting any one strand. Of course, if two adjacent strands are cut at the same time, you could be in a bit of trouble if those two strands are those of one arm or in a lot if they are those holding the central biner, but you would be in trouble in this situation with the PP as well.

To guard against that you would indeed need a back-up system but, being a failsafe in case of a catastrophic and extremely unlikely failure of the primary system, this back-up system would not need to be an identical image of the primary one, clutter the central biner or hamper equalization. In fact it would be a much more effective back-up, much less likely to be messed up by the same event that did in the primary system if it does none of these things; ie, it should preferably hang a bit loose so its strands are not near those of the primary system, be made of a different material than the sling and be attached differently to the central biner (eg along the spine) in a way that does not interfere with its sliding/equalizing if the primary system is an equalizing one.

Jim Titt wrote: The work of failure of an individual piece is completely unknown and always will be until that particular piece fails, generally compared with the amount of energy in the system it will be negligeable anyway so we don´t consider it in working out the forces. It has anyway been shown that the effect of sequential failure tends towards higher ultimate impact forces as the system stiffness is increased for the second impact...

Are you saying not only that pros that blow do not reduce the final load but that they actually increase it?