For the climbers out there, this is an analysis of cams, or "Spring Loaded Camming Devices" as the article and Wikipedia calls them. I have never heard of a cam called an "anchor" in contrast to the article title and the HN title. Wikipedia does not use the word "anchor" in the entire article.
Specifically, this is an analysis of the strength of anchors built with cams. 'Anchor' in the title is relevant, but the title here leaves out critical information, because the overwhelming majority of anchors in rock climbs are bolted.
"An Elastic Model of the Holding Power of Gear Anchors in Rock Climbing" would be much more apt (though gear anchors can include several other types of protection, "Cam Anchors" is not something I've ever heard)
edit: upon further reading, you're right. There's no mention of anchors in the article, other than referring to a cam placement as 'a type of anchor'. It seems that the article was not written by climbers, which is surprising.
In fall arrest this is a common term, and it's super accurate, regardless whether it's common with rock climbers, these are definitely anchors. So it's not surprising that a MIT paper would use it.
A more specific example, Black Diamond's camalot z4 are described in their manual as 'frictional anchors'
Edit: of course there is a standard hah - EN 12276:2013 Mountaineering Equipment – Frictional Anchors
An anchor would usually refer to the collection of gear placements equalised together, not an individual placement. (Although rarely you might have a 1 piece anchor). The term anchor is more a description of the purpose than what it's made of.
This article looks to be 15+ years old. The cam pictured in Fig. 1 looks look one of the earliest models of SLCDs, from sometime in the 80s when cams were all rigid, if I remember my climbing history correctly.
If you follow the link to the home page, you can end up on a page that refers to more recent work having been done in 2005.
I'm always amazed that these kind of devices don't just make the crack in the cliff wider and fail...
A climber falling, even on an elasticated rope, might briefly put 800kg into the cam. The cam itself is effectively a 10:1 lever, so exerts 8000kg in each direction on the rock.
I'm just surprised that something that is already a crack (ie. Has already moved in the past) doesn't sometimes move when 8000kg is applied to it... And if it moves, the climber likely dies.
I think 800KG is exceedingly rare, and also the upper limit of what a climber can tolerate even if their protection (the camming device) holds. It does happen with massive lead falls, especially when not much rope has been paid out. 1000 KG is likely to cause injury.
And of course, the cams are typically rated from 600KG to 1600KG of force, depending on manufacturer, size, and material. But it's rare for them to experience these loads.
Even with smaller falls though, say 200-300 KG of force, they do occasionally break the rock they're set in, or cause the rock to 'flex', causing it to come out (this is more common with a type of feature called a 'flake', which is a semi-detached piece of rock, rather than a typical crack which is unlikely to move).
Ideally, you have more protection beneath the one you placed. Then if one piece blows, you'll fall a little bit further to the next one, which is hopefully placed in better rock.
That is an important thing to keep in mind, but keep in mind that many of these climbs follow cracks in the middle of large cliffs. I can't imagine the force it would take to move a cliff like the own shown in the link below.
As others have said, you do need to assess how solid a feature like a flake is. I live in a part of southeast Alaska that has almost no rock climbing because most of our mountains are uplifted accretion formations, so a fall on a cam likely would move the pieces of the mountain and lead to a nasty fall, with pieces of the mountain falling on you as well.
Isn't it common to use multiple anchors? Kind of happens automatically in the process of climbing in a team of two. That's not to say that an anchor failing is a pleasant experience...
The anchors are often far apart, especially in trad climbing. If the first one failed the next one will have to absorb a bigger hit... so, it can be more than unpleasant...
The common advice when you don't trust your anchor is don't fall ;)
I agree, its alot of trust to put into your only safety. A good comparison, the Rope Access industry (windmills, arenas, etc) has a zero-fatality rate due to training, but also because they are always double roped.
Maybe a rope is more like the wings of a plane than a discrete component. You could have two sets of wings. But is that really a good idea?
And climbing is not expected to be 100% safe. In the same way that the single engine in a GA plane can fail and cause a crash. You have to balance the risks with the enjoyment.
There's redundancy in modern climbing ropes too. The part that does most of the support is the core. The sheath surrounds and protects the core, but is plenty strong on its own. If the sheath is cut through, the rope won't fail. Even if the core (which is composed of 10-30 strands each capable of supporting your weight) is partially cut through, the rope still won't fail. Of course, you are expected to inspect your rope regularly, and retire it if the core is exposed at all, or if there are signs of core damage.
The reason for retiring a rope when the core is exposed is not because it won't hold your weight, it's because there's no longer as much redundancy.
Climbing ropes (almost) never fail. There have been very, very, few accidents caused by this. You're more likely to get hit in the head by a falling rock while standing near a rock structure (something that non-climbers regularly do without thinking twice), than for the climbing rope to fail
There are probably more 'climbs' where a climber trusts the rope to hold their weight on descent in a single day than there are flights in a year, so I think the integrity of a regularly inspected climbing rope is the least of your concerns
(for some additional anecdata, in a 2 hour session at my climbing gym, I might participate in 10 climbs and witness 500)
I feel like there's different reasons to use a half rope or a single rope. I mean you aren't usually clipping in the same draw on each half rope anyways because a) putting both ropes in the same draw can cause them to cut each other from frictional shearing on a fall, and b) the fall will be really hard, and perhaps could injure you, since you fall on two springs instead of one. I assume you know this because you are so safety considerate.
https://en.wikipedia.org/wiki/Spring-loaded_camming_device