In reply to Adrian Berry:
0 deg camming angle would mean the cam would be a circle, and the force applied to the rock wall would be at 90 degs to the load. Of course it wouldn't work though... On this diagram you've got sf which is the force vector from the cam contact spot to the axle. you've got H, the resultant horizontal force, i.e. the force exerted outwards on the rock, and you've got F, the vertical component of the load force pulling down on the lobe, usually 1/4 of the total load. The geometry of this grafically shows the size of the resultant forces - as alpha the angle between them all increases, if F stays the same, the H leg of the diagram gets shorter, i.e. the force gets smaller.
http://www.bigwalls.net/climb/camf/Camfigs/Camf3.gif
This diagram is the more relevent one with respect to what we are talking about:
http://www.waveequation.com/surfboard_wax_friction_defs.gif
Basically take this diagram and turn it on it's side and stick it where the camlobe contact spot is. Which is why the H component is important. Instead of an object and mass, replace it with the cam lobe and a force pressing down on the surface. The frictional force (and therefore the pull out force required to make the cam fail) is the coefficient of friction (variable on surface roughnesses and granular structure of the material (but not surface area)) multiplied by the normal (i.e. perpendicular to the surface) force exerted by the cam lobe on the surface.
So if you increase the normal force, you get higher frictional force and higher pull out loadings are possible. This is entirely separate to the force exerted outwards during that loading of the walls by the cam lobes. All the Totem does is use a lever effect by virtue of where the load cables are attached to push the cam lobes into the rock more firmly and therefore increases the normal force.
The final component which is more of a real world issue is that of rock pulverisation. Generally the way a real world placement will fail is that the force pressing outwards on the rock ( the SF component) will make the rock fail, either by flexing the rock or moving it (behind a flake for example - like on the Sloth) or by pulverising the rock. The former you can do absolutely nothing about. The latter you can. Previously I mentioned surface area being independent of friction - which it is. However a materials strength is based on the stress it can withstand. Stress is a function of force and surface area - if the force remains constant, the stress induced in the material can be reduced by spreading that force over a larger area. I.e. make the cam lobe wider, and increase surface area and in theory you will see a reduced stress in the rock walls. This is important where stone is relatively soft e.g. sand stone as they tend to be able to withstand less stress. With a thin cam lobe, the stress is higher and the rock pulverises creating a layer of sand and dust between the rock and the cam lobe and this reduces the frictional force we were talking about before resulting in a pull out.
So, you have to blend several factors:
Cam material - a softer material deflects more and causes greater mechanical interferance between rock grain structure and the metal surface - it increases the coefficient of friction.
Cam angle (friction) - lower cam angle means higher normal force to provide greater friction.
Cam width - reduces induced stress in the rock surface which is particularly relevant in soft rock
Number of cams - 4 cams spread the load between 4 lobes, i.e. lower induced force in each contact spot - hence why 3 cam units have a wider middle lobe.
Cam angle (range) - the higher the cam angle, the tighter the exponential spiral which governs the cam surface becomes, and the result is a greater differential there is as the cam moves from fully open to fully closed - i.e. more range.
Cam angle (outward load) - the lower the cam angle, the greater the H component becomes andthe more likely you are to pulverise the rock or break a flake
Cam angle (lobe strength) - the lower the cam angle, the more stress is induced in the cam lobe. This is less of a problem with small cams, but a big problem with large cams when the cam lobe will simply buckle with higher force. hence a large cam lobe needs to be inherently stronger to cope with this stress.
Cam angle (loss of friction) - as you get to very high cam angles, the cam lobe lose friction - hence the magic 13.75 number. Jardine figured this was the best compromise before you start losing too much friction and the pull out force starts to drop dramatically.
Aliens use 16ish - they use a soft metal to increase friction - it's the only reason they stick well - 6061T6 is as soft as butter. If he had used 13.75 they would stick even better. But you can't use the material for large lobes as you compromise strength too much.
Camalots are in the region of 15 degs and use 6082 t6 (I think - that's from memory!) as it's much strongerbut still fairly soft. In addition 6000 series is easier to extrude. 15 allows them to fully maximise the range of the unit to reap th rewards of their double axle design.
DMM Dragons and WC cams use 13.75 and 6082 t6 - theoretically a higher holding power than BD. The dragons you will notice have a marginally smaller range than BD's, but we're talking a gnats chuff - that's the cam angle taking effect.
Metolius, last but not least use I believe 13.25 and 7075-t6. The material is super hard and durable, but less grippy as a consequence. 13.25 allows for friction to be increased again.
As for totem totems - don't know, for sure - but you get the drift?