The quad is the [best/worst]!

Connor Dobsonwrote:

What I was trying to get at is the 4KN of force is not consistently applied, the belayer + rope + climber can all be thought of as a single entity. 

Why is the gear so shit pieces are pulling with 0 energy absorbed, I think saying this energy being negligible is a huge oversimplification.

The energy required to break something is the area under a force/extension curve and called energy to failure. 

If you take a Dyneema sling and a nylon sling and pull test to failure in a hand-powered tester you suddenly notice that it takes vastly more work to break the sling compared to the rope because while the forces are identical the extension of nylon is far greater.

Brittle substances have extremely low energy to failure and rock is very brittle, that small bump or flake holding your nut may well hold 4kN but take virtually no energy to break off.

Jim Tittwrote:

The energy required to break something is the area under a force/extension curve and called energy to failure. 

If you take a Dyneema sling and a nylon sling and pull test to failure in a hand-powered tester you suddenly notice that it takes vastly more work to break the sling compared to the rope because while the forces are identical the extension of nylon is far greater.

Brittle substances have extremely low energy to failure and rock is very brittle, that small bump or flake holding your nut may well hold 4kN but take virtually no energy to break off.

Any ammount you are stretching the sling, stretching the rope, deforming the metal of the gear, friction of anything on anything else, deforming and fracturing the rock are all losses that are reducing the energy of the falling climber.

Also rock is very brittle in tension but much stronger in compression. If you are placing nuts behind a flake (causing the rock to go into tension) and then taking high factor falls onto it I think there are bigger fish to fry than whether you are using a quad or whatever other anchoring setup is of mode that day. 

Didn't Jordan Cannon take an 150' PDL whip on the nose a few years ago (and I doubt he was using a quad or a low extension or what have you anchor)

Also what I was alluding to it being about impulse and energy rather than force is one can't neatly add up the peak forces in this type of problem because they can happen at different times in a fall scenario. If the climber is putting a force on the belayer pulling him down, he is also putting a force pulling the climber up. This may be 4kN peak but it won't be consistently 4kN. Impulse is a change in momentum which conveniently is a force that varies (or doesn't) over a period of time. 

Energy comes in because it's easier to understand that it will take energy to do all the things I listed above, and that energy will be lost through those methods and not have to be transfered to the remaining pieces of the anchor. 

Reply to Jim because of the stupid post limit:

I have no intention to measure things because it's a waste of time. I agree that experimentally forces are easier to measure but energy is a more intuitive way to understand how pieces pulling does still slow people down. 

You still missed the point that I was making that oversimplifying adding peak forces together is an simplification that you can't always make (even though in many and even most it may work out to be super close enough). But that could be me being pedantic. 

Energy is also important to think about when thinking about the acceleration of arresting the belayer+climber with the remaining pieces. The kinetic energy of the climber is transfered into the piece as force over a time, an impulse. The shorter the impulse the higher the peak force, not that the length of the impulse changes much do to the stiff nature of most anchor materials.

If I was to actually spend time investigating and putting engineering effort into some means of providing greater safety to the climbing community I would do what the author is doing and look at both occurrence (how often it happens) and severity (what happens when it happens). While the author is correct that the severity is incredibly high with a total anchor failure, with the occurrence being so incredibly low (6 in 30 years or what have you), it is negligible in terms of total risk in climbing. You are looking at a very marginal improvement if your switch of anchor type would even have prevented any of them at all. 

There are much higher occurence and similar severity failure modes in climbing. See people rapping off the ends of their rope, pro pulling, free solo falls, people not knowing how to belay, avalanches and the Pareto goes on and on. I would rather spend time on those if I wanted to spend time on knocking failure modes off of the accidents list.

This is merely another quad thread, equally as (potentially un)useful at preventing injuries or deaths as the last 10 of these and doesn't really add any meaningful info imo other than summarizing (un)intentionally that anchor type doesn't statistically matter.

I have built thousands of anchors. The number of them that didnt have two solid pieces so the rigging method mattered… zero. If it was that way it wasn’t an anchor and I wouldn’t be trusting my life to it 

Gloweringwrote:

I have built thousands of anchors. The number of them that didnt have two solid pieces so the rigging method mattered… zero. If it was that way it wasn’t an anchor and I wouldn’t be trusting my life to it 

Exactly, if you are relying on the type of rigging to save you, you are screwed up. The most effective way of preventing total anchor failure is to place strong and redundant pieces, as well as to not consistently take factor 2 falls,  the rest is just convenience and mental masturbation. 

A quad is nice when climbing in a party of 3 so people can move independently without yanking on each other.

 A fixed point is nice to make with the rope when pieces are far apart relative to each other that would be inconvenient to have to extend to work with each other. 

If you have 2 solid bolts, clove, clove, boom you are done. Clip your device to one of the bolts or lower your device by clipping it to an 8 on a bight between the 2 bolts. 

Connor Dobsonwrote:

Any ammount you are stretching the sling, stretching the rope, deforming the metal of the gear, friction of anything on anything else, deforming and fracturing the rock are all losses that are reducing the energy of the falling climber.

Also rock is very brittle in tension but much stronger in compression. If you are placing nuts behind a flake (causing the rock to go into tension) and then taking high factor falls onto it I think there are bigger fish to fry than whether you are using a quad or whatever other anchoring setup is of mode that day. 

Didn't Jordan Cannon take an 150' PDL whip on the nose a few years ago (and I doubt he was using a quad or a low extension or what have you anchor)

Also what I was alluding to it being about impulse and energy rather than force is one can't neatly add up the peak forces in this type of problem because they can happen at different times in a fall scenario. If the climber is putting a force on the belayer pulling him down, he is also putting a force pulling the climber up. This may be 4kN peak but it won't be consistently 4kN. Impulse is a change in momentum which conveniently is a force that varies (or doesn't) over a period of time. 

Energy comes in because it's easier to understand that it will take energy to do all the things I listed above, and that energy will be lost through those methods and not have to be transfered to the remaining pieces of the anchor. 

Exactly, there are situations where failure of a piece may remove a negligable amount of energy from the system, not as you wrote "There is only so much energy to spread around, with some of it being absorbed in a piece pulling and the rope stretching before the piece pulls." The energy dissapated by a rope is well covered in 'The Physics of a Climbing Rope' by Ulrich Leuthäusser.

There is a slight delay in between the forces felt by the belayer and the faller due to frictional hysterises BUT they are of no relevance as the faller is still falling (and when the belay fails accelerating), unfortunately this also works in reverse when the belay fails and comes to maximum extension. We are anyway only talking of fractions of seconds anyway, investigation of screamers tells us that all is not what one hoped for.

You are welcome to investigate energy and we will be interested to see your calculations and data, I'll stick with forces because I can measure them and they are what break things.

To help you out I just pulled a Rock 1 to failure and the wire broke at 10.2kN (it's rated at 7kN) and measured the extension to failure which works out at 31J energy to failure. This is the energy from an 80kg climber falling about 2".

Gloweringwrote:

I have built thousands of anchors. The number of them that didnt have two solid pieces so the rigging method mattered… zero. If it was that way it wasn’t an anchor and I wouldn’t be trusting my life to it 

This, and Connor’s agreement and most of his analysis above are so patently foolish, it’s hard to fathom.  

You’re saying “I’m so worried about making a strong anchor, I don’t care about making it stronger”.    

Or “ I place my pro so strongly that I can afford to weaken them”  

Some of the posts seem to say “you’re stupid for trusting your life to a weak anchor, but I don’t worry about shaving a few kN off cuz it doesn’t matter”

How about instead, we just go with “since you’re trusting your life to it, just make the strongest anchor possible”?   

The amount of times my helmet saved my life? Zero.  The amount of times a knot saved me from actually rapping off the end? Zero.  But I don’t argue about how unnecessary they are in the forums. 

How bout we strive for the number of times One posts demonstrably false and mind numbingly foolish conclusions in threads to be zero?

And while I totally agree this ain’t the number one killer of climbers out there, no matter where on the risk Pareto chart you go, there is key information and techniques to reduce that risk.

A good mindset and logical framework is needed to maximize your safety in any risk category.  

Jim Tittwrote:

To help you out I just pulled a Rock 1 to failure and the wire broke at 10.2kN (it's rated at 7kN) and measured the extension to failure which works out at 31J energy to failure. This is the energy from an 80kg climber falling about 2".

And what energy would the rope have absorbed during that since it a less stiff spring and would see most of the extension?

Also Mark, you call me foolish but my logic is internally sound. The likelihood of total anchor failure is so incredibly rare that you could be more safe by doing almost anything else in a safer way (including buying a safer car to drive to the crag). The marginal benefit you might gain (AT MOST) according to the data is preventing an accident every 5 year, that is millions of climbing falls. And that is ONLY if that case of total anchor failure would have been prevented by the minor marginal benefit of a different anchor system.

That benefit is not worth the convenience of different anchor types in different situations or frankly the mental power of worrying about it I'm gonna A5 my anchor everytime I climb above the belay.

There are also better ways to mitigate a hard start to a pitch that I mentioned earlier such as lowering the belayer down to get more rope out for the start and making sure that the first piece (even if it part of the anchor) is a solid piece (or as solid as you think it is). These solve the problem at the root cause of there was too much load on the piece of gear when falling onto the anchor.

I am aware that it is the internal damping of the spring, my point was that the rope will stretch much more than the sling or the wire and will damp energy out of the system so if this is test was done with a sling it doesn't tell the whole story.

While yes there is not much internal energy to the rock, there is a fair amount of energy being dissipated into the rope as long as the piece can survive a decent amount of force, all of this energy dissipation means that the climber has less kinetic energy when the piece blows. I was never using energy as a predictor of failure but it is the same units as the work done by the rope damping the fall and sums to this work, however small this energy is doing something. 

I mean if the piece was right on the edge of not pulling when it does fail, the rope would have likely absorbed most of the climbers energy and then you are dealing with a much shorter "fall" where the climber and belayer have the length of the extension to accelerate vs. having a piece of shit first piece that barely absorbs any energy of the fall because it doesn't hold enough force for the rope to adequately damp the fall. 

This is more intuitive for me to think this way, but isn't neccessary to understanding the problem, just how I do. The cool thing about Newtonian physics is we can both be right, forces are easily measurable and cause things to fail but conservation of energy must also be kept. 

Connor Dobsonwrote:

Exactly, if you are relying on the type of rigging to save you, you are screwed up. The most effective way of preventing total anchor failure is to place strong and redundant pieces, as well as to not consistently take factor 2 falls,  the rest is just convenience and mental masturbation. 

A quad is nice when climbing in a party of 3 so people can move independently without yanking on each other.

 A fixed point is nice to make with the rope when pieces are far apart relative to each other that would be inconvenient to have to extend to work with each other. 

If you have 2 solid bolts, clove, clove, boom you are done. Clip your device to one of the bolts or lower your device by clipping it to an 8 on a bight between the 2 bolts. 

The discussion here is about a ceteris paribus analysis. You are shifting the goalposts. 

Derek DeBruinwrote:

Three people simply hanging on the anchor should not cause an anchor to fail. A useful rule of thumb for a hanging mass is to think of one person as approximately 1kN (again, hanging, not falling), so in this case three people might mean 3kN of force, which is not so different than slingshot top roping. I'd argue if the anchor failed under this load you don't really have an anchor.

Well I'm talking about 1 piece failing at 3kN.  The remaining piece(s) fail at whatever force the extension results in.  As a first approximation, I would take https://www.blackdiamondequipment.com/en_eu/stories/experience-story-qc-lab-personal-anchor-systems-explained/?cid=qc-lab-personal-anchor-systems-explained (table "dynamic tests"; FF in this case would be extension divided by (extension + anchor length + tether length)) and multiply by the number of climbers raised to a power between 0.5 and 1 (depending on how their individual tether lengths compare to the length of the shared anchor).

And yes the probability of even a single piece failing at 3kN seems low. But so is the probability of FF2. I wouldn't be surprised if the probability of a piece failing at 3kN turned out to be higher, which is why I think a piece popping under the static weight of the climbing party is something to consider in the extension discussion.

Serge Swrote:

As a first approximation, I would take https://www.blackdiamondequipment.com/en_eu/stories/experience-story-qc-lab-personal-anchor-systems-explained/?cid=qc-lab-personal-anchor-systems-explained (table "dynamic tests"; FF in this case would be extension divided by (extension + anchor length + tether length))

Any tests that are done with static or near-static materials, over relatively short fall lengths, in a fully static test environment (i.e. steel weights and rigid attachment) bear little resemblance to the real world.  Humans are squishy, so the force a human will put onto an anchor if they fall 12 inches onto a 60cm sling is going to be WAY less than the force a fully rigid steel test apparatus puts onto the same anchor.

(Caveats: A human pulled down at a high rate of acceleration, like Jim's example of a belayer arresting a hard fall, isn't the same situation.  And according to Derek's data, a dynamic rope may become less dynamic if it is significantly stretched after a hard fall.  But for simple cases of someone slipping off near an anchor and falling onto their tether, you need to take these steel weight tests with a grain of salt.)

Connor Dobsonwrote:

And what energy would the rope have absorbed during that since it a less stiff spring and would see most of the extensio

Err it's your theory, not mine. I've given a reference where you can investigate energy dissipation in a climbing rope so you can work it out for yourself, I've done half the work to show your idea is incorrect the rest is up to you.

Not Hobo Greg wrote:

For us mathematically challenged folks, what’s the take away? I would only use a quad on bolts, that seems to be ok? From a guide perspective I want to use em so my guests have an easier time breaking down and racking the anchor, and will speed up the day throughout the belay changeovers.

If on modern bolts in good repair, guiding seems a pretty reasonable application for the quad. I would advise against it for gear anchors or ice anchors, even if you're working.

JonasMRwrote:

When you read the paper's penultimate section, what do you take away from it? Again, the question isn't to use good or bad judgement. It's "can experience based judgement keep you safe from low probability/high consequence problems?" I'm not clear if folks didn't read that part, or just don't want to address it, or what the deal is exactly.  

Recognition primed decision making would tell you that judgment, intuition, etc. are useful for high probability events for which there has been consistent feedback tightly coupled to antecedent events (which allows for deliberate practice the the attendant development of expertise). This is how speed chess works. The context is always the same, the player gets immediate and appropriate feedback (was it a good move or bad move), and the chess games is a high probability event (since someone who wants to get good at playing chess plays a lot of it). Consequently, good speed chess players can analyze the pattern on the board and have a rapid reaction based on having seen hundreds or more similar or identical layouts of pieces. 

This doesn't always work for climbing situations. This is why you can get a really poor belay from someone who's never actually caught lead falls. Even if they've climbed hundreds of times, if all of that experience doesn't include the experience of actually catching lead falls, the belayer is going to perform poorly because they don't have the relevant practice.

Similarly, if you're not in the habit of taking or catching falls directly onto the belay, you don't really know for sure how good you are at 1) belaying such falls, 2) building anchors to withstand such falls, and 3) estimating the likelihood of such a fall occurring as the leader. So, why introduce a known failure mechanism when there are plenty of reasonable alternatives that eliminate that failure mechanism entirely?

As for the other folks who are playing the numbers game and saying that anchor failure is super rare anyway: to be clear, I agree in general. I'm not saying that the anchor rigging is the ultimate measure of safety for a climbing party; far from it. There are better ways to spend your time focusing on belaying well, placing good gear, route finding, etc. But why introduce a needless failure mechanism? If you can determine a rigging that will work every time regardless with no appreciable downside, why not just use that in perpetuity so you can forget about anchor rigging and spend time on those other more important things? It's akin to a seat belt in an airplane crash. Is the plane likely to crash? No, commercial crashes are very rare events. Not wearing your seat belt is probably more comfortable and convenient. But you sure would look stupid if you were the only person who managed to get killed in something like the Hudson River ditching because you decided not to wear your seat belt.

Serge Swrote:

Well I'm talking about 1 piece failing at 3kN.  The remaining piece(s) fail at whatever force the extension results in.  As a first approximation, I would take https://www.blackdiamondequipment.com/en_eu/stories/experience-story-qc-lab-personal-anchor-systems-explained/?cid=qc-lab-personal-anchor-systems-explained (table "dynamic tests"; FF in this case would be extension divided by (extension + anchor length + tether length)) and multiply by the number of climbers raised to a power between 0.5 and 1 (depending on how their individual tether lengths compare to the length of the shared anchor).

And yes the probability of even a single piece failing at 3kN seems low. But so is the probability of FF2. I wouldn't be surprised if the probability of a piece failing at 3kN turned out to be higher, which is why I think a piece popping under the static weight of the climbing party is something to consider in the extension discussion.

Got it. Thanks for clarifying, as I definitely misunderstood you the first time. Yes, I agree that extension even while hanging on the anchor could be problematic if a piece failed. I will note that the BD tests you linked were drop tests, so a real world comparison implies the party wouldn't simply be hanging on the anchor, but rather that someone would have a slack tether and then fall onto it. But if that happened and a piece failed, the resulting fall of 1 or 2 other people also getting yanked from their stance could end quite poorly (even if the anchor didn't fail completely).

And since I imagine I'll hit a post limit soon, for everyone's edification, here's those papers Jim mentioned in the discussion with Connor:

https://www.sigmadewe.com/fileadmin/user_upload/pdf-Dateien/SEILPHYSIK.pdf

https://www.sigmadewe.com/fileadmin/user_upload/pdf-Dateien/Physics_of_climbing_ropes_Part_2.pdf