How do Horizontally placed Cams Work?

I don't know complicated math but out of curiosity I plugged some pro in the pockets of a set of Rock Rings. Oddly with BD dual axles, I could pull them pretty easily out when pulling straight against them but they didn't budge when mimicking a horizontal fall. I repeated with single stem Metolius master cams (the new ones) and they were solid from all directions of pull. Like I said I only know basic math so I can't explain, just what I experienced.

eli poss wrote: With cams you run the risk of damaging the cable if it runs over an edge and is whipped on. Damage isn't failure, but it costs money repair/replace.

A small price to pay if on the RARE occasion damage occurs. It has never occured to me so I'm not worried.

eli poss wrote: Cams can walk and become undercammed or overcammed, which doesn't really happen to tri-cams.

Movement should be minimised whatever your placement. Tri-cams are far more likely to have problems if they do experience movement.

eli poss wrote:Also, when you're dealing with less-than-perfect placements such as pockets, flares, funky pods, etc tri-cams tend to fit better than cams. They are also less likely to slip out of a placement in slick, greasy rock than SLCDs. And they're lighter and cheaper.

True all that but how does that relate to horizontals?

Christian Fracchia wrote: Ok, we were out again today and at the end of the day we decided to do some more yanking. This time we used body weight of a short 2-3ft fall on 2-3ft of rope and we found a mildly flared horizontal and placed the cams at the edge of where they would sit without falling out on their own. I must say that doing this made me wonder how anyone pulls cams out of horizontals when the rock quality is good because we did a terrible job placing them...and they still held!

Thanks for that! Though in my opinion the first came was a long way from being terrible. Horizontal crack, good rock whats not to like. (Differential lobe expansion isn't a big deal.)

Chris, I'm not personally concerned about cams in horizontals, although the math still interesting.

Not to be a nudge, but in your tests the backup is clearly sharing the load each time---you can see the stretch in the backup strand in the video in both falls.

As for FF1, you aren't anywhere near that. You'd have to be falling with your waist at the level of the pro (and not have the backup strand involved...)

AndySalo wrote: Unless I missed something, you didn't actually try the slot on Carbs.

Whichever slot was used, my memory (from quite a ways back) is that at that point the next piece down is also considerably to the left. In this situation, the top piece will rotate towards the lower piece, and if the placement is shallow the high cam can rotate clean out of the slot. This seems to me to be as likely a failure mechanism as any asymmetry in lobe loading coming from the horizontal orientation of the cam placement, meaning that fretting over the intrinsic stability of horizontal placements might be looking for answers in the wrong place.

On diagonaling routes, one typically tries to place cams in horizontals not straight in, but angled towards the anticipated direction of pull. If a slot is too shallow to place a cam this way, remember that it is likely to end up like that in a fall anyway.

patto wrote: A small price to pay if on the RARE occasion damage occurs. It has never occured to me so I'm not worried. Movement should be minimised whatever your placement. Tri-cams are far more likely to have problems if they do experience movement. True all that but how does that relate to horizontals? Thanks for that! Though in my opinion the first came was a long way from being terrible. Horizontal crack, good rock whats not to like. (Differential lobe expansion isn't a big deal.)

Tri-cams can be set which pretty much negates any chances of movement, provided you sling it appropriately. I dunno about you but I've come across many horizontals that were less than perfectly parallel cracks. Sometimes you will have a pod that is too narrow to fit 4 cam lobes, or even 3 lobes, but a tri-cam just barely fits. Most of the time cams will work just fine. Unless you're climbing in the southeast, where the rock really lends itself to tri-cams.

I never trust horizontal cams in a perfectly parallel crack. In the real world you normally can find a spot with a lip and it isn't perfectly parallel and would prevent alot of what you are testing with.

Where I live I normally go for tri-cams over normal cams for horizontals anyway.

I would think most people do a simple pull test to test their placement to make sure they wouldn't pull like your test. Your gear looks like it is poping from not so much a fall but the fact that the pull is causing the cam to turn, no different than a bad placement where your cams aren't even pushing into the wall. If it holds a simple pull than it is orientated in such a wall that the force is probably going to hold a full fall.

If you can change your top piece of metal to slop down like 2% and see if that prevents your pull out issue. Even from your video it looks like that placement isn't perfectly parallel and has a little lip coming down.

Could we have negated the need for the metal crack, youtube vid, and force diagrams by having the OP just do a simple bounce test on the piece in question? What makes a horizontal placement so unique that we cant do the same testing climbers do for other placements to test their reliability? Does the theoretical add value over the practical?

jason.cre wrote:Could we have negated the need for the metal crack, youtube vid, and force diagrams by having the OP just do a simple bounce test on the piece in question? What makes a horizontal placement so unique that we cant do the same testing climbers do for other placements to test their reliability? Does the theoretical add value over the practical?

We may or may not have "needed" all that stuff, depending on what you think "need" means. Simple bounce tests can tell you something but have little or no explanatory value.

What makes a horizontal placement "so unique" is that the load on the cams is asymmetric---the stem does not lie the plane of symmetry of the cam that contains the axle. The potential for asymmetry increases as the cam size increases. Whether this could, in theory, result in unexpected cam behavior is a perfectly reasonable question whose answer might help to think about how to place cams in shallow horizontals, and could shed light on the assumption that jerk testing is as reliable for horizontal placements as it is for vertical ones. Or, the answer could be that it doesn't matter at all in theory.

Does the theoretical add value over the practical? The less practical or available effective testing is, the more theory matters. We don't send astronauts to the moon by shooting up 1,000 rockets at random to see if any of them make it into a moon orbit suitable for landing. Back on earth, let's not forget that cam invention and design depended entirely on theoretical considerations and calculations, so if you think cams are a good thing than the answer about the value of theory is a resounding yes.

More generally, the value of theory depends on many things. It certainly adds nothing if the user either doesn't understand or misunderstands what the theory proposes. In some cases, practical variations constitute a kind of "noise" that overwhelms the theoretical predictions. In other cases the theory is based on a model that misses too many aspects of the real situation.

Many people seem to think that testing is somehow more conclusive, but forgetting about the rocket conundrum, the challenges of achieving statistical significance (as opposed to confirming pre-existing prejudices) are substantial, a good many of the tests we see in the climbing world are worth almost nothing, and in the relatively rare cases when the tests are good, one is typically left with probabilities but no explanations. And often, only those with some theoretical knowledge are capable of spotting design flaws in the tests.

The title of the thread is unfortunate. Most experienced climbers already know that "pro really works in horizontals." A question theory might address is whether or not cams work better in vertical placements than in horizontal placements, particularly shallow horizontal placements with medium to large cams, where the asymmetric loading is has the most chance to be significant.

Although the following tests aren't in horizontals, they do test another situation in which the cam is not loaded in a way anticipated by its design. Note that not all cams stay put. (Interesting but unanswerable question: could the failures have been detected by jerk-testing?)

youtube.com/watch?v=Wl3bVx4…

rgold wrote:We don't send astronauts to the moon by shooting up 1,000 rockets at random to see if any of them make it into a moon orbit suitable for landing.

Nice one!

Likewise we don't build our bridges and buildings by 'testing' them beforehand.

jason.cre wrote:Does the theoretical add value over the practical?

Our industrialized world is build on scientific and engineering theories. Even sporting ability can be improved by understanding the background science.

rgold wrote:The title of the thread is unfortunate. Most experienced climbers already know that "pro really works in horizontals."

Including myself, otherwise I wouldn't be trad climbing in the Gunks.

The title of the thread is a question, I think one worth asking regardless if my approach to answering it is right or wrong. Fortunately, there are many people here, yourself included, that can provide a better explanation/analysis of what's happening and I think that's info anyone could afford to hear.

Randomdiscourse wrote: Including myself, otherwise I wouldn't be trad climbing in the Gunks. The title of the thread is a question, I think one worth asking regardless if my approach to answering it is right or wrong. Fortunately, there are many people here, yourself included, that can provide a better explanation/analysis of what's happening and I think that's info anyone could afford to hear.

Yep this. Goto the gunks on a busy late spring Saturday and tell the 400+ people there " sorry brah cams dun stop workin in dem horizontals since evah"

I have lived at the gunks and been leading there for over a decade. I dont even use tricams (sometimes pink or blue). I have took more whips on cams in horizontals than I can literally remember. Never had a bad experience.

Thanks everyone for this good discussion.

Randomdiscourse, I just want to add that I've done Carbs & Caffeine a few times and that Number Two placement after the crux bolt has not seemed iffy to me. But later, at the beginning of the final traverse crux, there is a placement for a truly shallow, tipped-out Number One Camalot. (I watched your vid and you placed nothing here.) It is a horrible placement, and I've hung on it! Not that I was happy or proud to do so, but it held just fine.

You can actually reach further out along the rail before you begin the traverse and place a perfect blue or green Alien (where again I was surprised to see you placed nothing at all!). The last time I did the route I skipped the crappy Number One placement, but it is an option if you really want to test a bad placement in a horizontal.

This has been a very interesting thread to follow and read.

And now I'm recalling that as part of my own ground practice placement on real rock, specifically to learn to trust cam placements, I placed, weighted, and bounce tested them on both vertical and horizontals. For the horizontals the options I had happened to be shallow, so I only opted to do this with two cams as I was concerned of stem bending and possible damage. I did it with a Black Diamond C4 Camalot and a Metolius Ultralight Master Cam. Both cams held, which was very inspiring, although as I predicted the C4 stem bent quite a bit. Not a problem for one test as it bent back easily. The Metolius cam the bottom trigger wire that parallels the stem and connects to the plate that pulls the kevlar trigger wires got dinged and bent quite a bit. The main stem bent slightly. The main stem was no issue to straighten, the trigger wire part I bent back as well as I could, but I could see with repeated falls in horizontals with this cam that you'd probably destroy this wire.

For the long term you'd need to be concerned of inspecting the wire stem where it enters the stem. See the following thread paragraphs 5 through 8 with the two pictures in between:

thebmc.co.uk/have-you-check…

This is part of what eli poss has been talking about with Tricams being a better option for horizontals in terms of the sling taking minimal to no damage on a fall. Wire stems of cams will eventually break with repeated off axis falls. There is no doubt that it's quicker and easier to place a cam in a horizontal if it places well. I guess the question is whether the cost of the cam and how long the stem will last with shallow horizontal falls is worth it to you.

And thank you rgold for mentioning the direction of pull on diagonal routes. That sounds rather terrifying for that particular route if the cam pulls on a fall.

SethG: Thanks for the beta! The reason I never placed anything before beginning the traverse at the end was over confidence on my part that the traverse wasn't going to be an issue and, ironically considering the thread, I knew the #1 (backed up with a #2 TCU I believe) I had placed earlier was bomber.

anotherclimber: It has, though admittedly frustrating (for myself) at times. But that's the nature of communication, I'm not claiming I'm any good at it.

I annotated the beginning of the video with a link to this thread because I think it's a worthwhile read.

rgold, well said as usual! Testing and theoretical analysis go hand-in-hand; neither one is worth much without the other.

Ted Pinson wrote: Ha, wow! That's wild. Wouldn't work with modern cams though, unfortunately. I can't help but feel like the perpendicular loading is problematic, though I suppose Rich et. al will run the numbers.

Interesting point (I'm not sure this is exactly what you're talking about, but I'm going to assume it is, because I apparently have nothing better to do with my time :)

Edited to clarify: this only pertains to rigid-stem cams in horizontal cracks, and addresses a different (theoretically possible but, as it turns out, unrealistic) failure mode than the one discussed earlier upthread for flexible-stem cams.

The rigid stem means that the force on the axle from the stem is no longer necessarily directed along the stem axis; the stem can act as a lever around the bottom edge of the crack, pushing the top lobe vertically against the upper wall. Now, it is still true that the forces between lobe and rock are directed along the line from contact point to axle, so the normal analysis applies, and the lobes shouldn't slip. But, there is the possibility that the net force on the upper lobe will be closer to vertical than the cam angle, which will cause it to collapse (into the completely cammed position), possibly tipping out the bottom lobe. This is the same failure criteria that I alluded to in an earlier post, and that patto verified with his computer program, but this time applied to the top lobe rather than the bottom one.

For the purposes of calculation, I assumed the stem is horizontal (this should be worse than having it angled downward, although if you manage to have it angled upward that's a whole other can of worms) with a vertical fall force (again, worse than a force with an outward component). The upshot of the calculation is that the upper lobe collapses if:

(r/R + 1)*mu<tan(theta), where

r = distance from axle to edge
R = distance from edge to sling where force is applied
mu = coefficient of friction between rock and stem
theta = cam angle

(I can upload a scan of the derivation if anyone cares to see it)

Note that, again, the result is independent of force magnitude. For theta = 14 degrees and r/R = .25, the coefficient of friction must be less than 0.2 for the upper lobe to collapse, which is pretty low (think plastic on steel). Put another way, with a coefficient of 0.5, which is more realistic, there is no value of r/R where the lobe can collapse. In short, this probably isn't something to worry about. Of course, since the theoretical analysis shows that the result should be independent of force, any problems should show up in a jerk test performed in the field. Breaking the stem, however, is a much more real concern, hence the short tie-off.

chris24graham wrote:For the purposes of calculation, I assumed the stem is horizontal (this should be worse than having it angled downward

Huh? So we are back to having the stem parallel with the crack? That is basic cam 101. And certainly not worse for cam operation than having it angled downwards!

The thing about flexible stems is that the force is directed along them. A horizontal stem in a horizontal placement would mean that most of the fall load is loaded on the rock lip. Sure not great for your stem like has been pointed out earlier but just find for cam operation!

patto wrote: Huh? So we are back to having the stem parallel with the crack? That is basic cam 101. And certainly not worse for cam operation than having it angled downwards! The thing about flexible stems is that the force is directed along them. A horizontal stem in a horizontal placement would mean that most of the fall load is loaded on the rock lip. Sure not great for your stem like has been pointed out earlier but just find for cam operation!

This is for a rigid stem cam in a horizontal crack, addressing a possible failure mode in which the stem acts like a lever around the bottom edge of the crack, compressing the top lobe (I don't think this is a real failure mode, as demonstrated by real-world experience, I was just doing the calculation to show why that is the case). Different situation than was being discussed before, in response to Ted Pinson's comment.

Christian's posts are genius here. As demonstrated in CF's videos, the realities of horizontal placements are far from nuanced and so very far from a steel plate mounted to a 2x4.

The initial video and analysis are not only entirely uncontrolled and unrealistic but a mild mockery of science poised to scare budding (or gullible) climbers. This is not to say that I don't respect the time and effort expended in putting this together, but rather to suggest a bit more thoughtfulness be put into these posts. They can dramatically affect the psyche of climbers without the anecdotal experience to see through the false suggestions. In a pool of knowledge that is constantly being re-written, anything but the most stringently-studied and concretely-backed information muddles the water.

Yeah, I was picturing the levering action when I posited that flexible stems would be better. I also would wonder if the cam would have the same rated breaking strength if loaded like this. Not as big of a concern with a red C4, but possibly if you're comparing a green C3 (more rigid stem) to a green alien/basic (flexible stem). I would prefer a flexible stem irregardless, because of the failure mode Rich mentioned (lateral rotation if the route wanders/traverses).

Stephen, I think the original poster could have presented the video slightly differently, but it nonetheless raised interesting questions. There would be a lot of missed opportunity for productive or entertaining discussion if people felt they could only post hypotheses for which they had scientific "proof". That being said, people should, perhaps, be more careful to present their ideas as hypotheses rather than as facts. This video was certainly a little too sure of its conclusions, such that it might be misleading to some people; but who hasn't been guilty of being a bit too confident from time to time? I think it was posted in the spirit of pursuing deeper understanding, rather than making concrete recommendations, even if it didn't necessarily come off that way.

Ted, I'm not sure about breaking strengths, that's another matter entirely. Certainly a rigid-stem cam that is not tied off short will be fairly weak due to excessive bending stress, but as for one properly tied off vs a flexible stem kinked over an edge, I'm not sure. Maybe rgold or someone has some data?