Discussion Topic |
|
This thread has been locked |
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 12:14am PT
|
The way that Richard's Geekqualizer works is that the red sling is free to slip and extend under the force of the fall. The slipping is necessary to equalize the forces on the blue slings.
If the red sling is not free to slip then it does not equalize.
The friction can be a large issue. For a sling looped around a ring, the forces can be unbalanced because of the friction. If F0 is the larger force, and F1 the smaller, on each strand of the sling, then the sling will not slip when:
F1 = F0/exp(2*mu*pi)
where mu is the coefficient of static friction and pi = 3.1415...
If mu = 0.4 then F1 = 0.08*F0 or roughly 1/10th the force... put an other way, a 1 lb weight on one side can hold a 10 lb weight on the other.
This is not at all surprising, since that is the way a belay device works.
For multiple bends, these forces multiply, which is GOclimb's point. For a sliding-x there are actually 4 bends, one at each anchor and two at the clip-in 'biner, since the webbing is doubled there. In this case it is a bit more complicated since the normal forces on the inner webbing is increased by the outer webbing. It is possible that this large friction is the reason why the sliding-x did poorly in the Largo/wootles test.
I'll try to measure the static coefficience tomorrow (it's raining in California) to get an idea of how big the problem is...
...my intuition is that the friction is the big problem here.
Of course, one could coat the rings in teflon (or use teflon tape) which would considerably reduce the coefficient. Since this is exponential, it should make a big difference.
|
|
Paul Raphaelson
Ice climber
Brooklyn, NY
|
|
|
Feb 10, 2007 - 12:15am PT
|
I'm looking at variations on the Sliding W.
Besides extension, and lack of redundancy, are there problems with the strands binding? If so, are they catastrophic (rope burning itself) or minor (imperfect equalization due to friction)? If the binding isn't a serious issue, consider this ...
I just did a couple of experiments at home and found that the amount of extension from a typical sliding W setup is actually pretty minor. And it can be greatly reduced by tying one of the strands short (with an overhand or an 8) in order minimize the length of the rigging. You can get some (imperfect) redundancy by tying the rope into the power point biner with a clove hitch, and then backing it up with a figure 8 onto the strongest piece of pro.
Not a perfect setup, but it has some things going for it:
-simple. uses no extra gear, easy to evaluate, faster even than a cordelette.
-excellent equalization, assuming binding of the strands isn't an issue
-extension can be limited to 1 or 2 feet in most cases
-some redundancy is available with minimum effort by using the climbing rope.
On another note, I tried the variation on the sliding W that has overhand knots tied in the outside strands. It didn't work properly. It looked as if it would, but it did not equalize. All the load went onto the outside pieces, and was born by single strands on each side. This is difficult to explain ... I suggest trying it to see what the issue is.
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Feb 10, 2007 - 02:14am PT
|
"Question: Are you going to use this yourself as a default anchoring tool?"
Yes, for a while anyway. I want to see how much trouble it is in real life.
I don't see myself using the other three-placement anchor systems; they seem tricky to set up, they consume too many biners, I don't understand how they'll perform in all circumstances (there seems to me to be a wrong way to clip the bottom loops of the Mooselette, for example), I'm not sure I can reliably tell from visual inspection when everything is ok, and I don't trust myself to think about any of these things clearly when I'm tired, cold, and in a rush to get down before dark. I don't think any of the ones that purport to equalize three anchor points have fewer friction problems than the Geekqualizer. They may be fine for other people, but I'm a very slow thinker and I make a lot of mistakes on my way to understanding things.
I have very rarely carried a cordelette---mostly I tie in with the rope, so using a special-purpose anchoring gadget is two giant steps in a different direction. But I'll try it out for a season anyway and see what I think after that.
"did you get Wootles to test it?"
There's a chance that Wootles might have a job and a life, and this makes me just a wee bit reticent to ask him to test stuff, especially when he'd need three load cells to see the effect of friction. First I'm going to see if I can stand using the Geekqualizer, 'cause if I can't, why test it? (I'm not concerned about it being worse than the methods I typically use.) In any case, at present I only have two Geekqualizers (under the extremely optimistic assumption that I can convince a partner to carry and use one).
Friction:
I've already said that friction could kill the whole show, and that finding ways to reduce it like Ed's teflon coated rings could be the key to any reasonable success. I'm gonna look into his teflon tape idea....
As for comparisons, I don't think that, for example, the Mooselette fairs any better than the Geekqualizer. In my possibly mistaken reading of Gabe's diagram, the Mooselette appears to have seven bends around biners compared to the five bends around rings in the Geekqualizer. Two of the Mooselette bends are at relatively large angles, in some cases (e.g. the indoor picture posted earlier as opposed to the sketch) so large as to be insignificant, in which case the number of bends could in some cases be effectively as low as 5, the same as the Geekqualizer. However, the Mooselette has three strands interacting and possibly binding at the power point biner, whereas the Geekqualizer has no such interactions. I don't see any a priori way of predicting which might be better, and even if they had the same average performance, the binding potential in the Mooselette and other designs raises the possibilty of bad worst-case performances.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 02:26am PT
|
nylon-nylon coefficient is reported to be 0.15 to 0.25, which reduces some of the webbing on webbing binding, but the real problem I think is the increased normal force caused by one strand of the webbing being on top of another over a 'biner or ring.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 02:14pm PT
|
The tests for measuring the friction of slingage on 'biners is complete and the conclusions are rather interesting. I believe that they may even be informative.
"Executive Summary"
I'll tell the story here with the technical details below for the geeks out there.
There wasn't much difference in using a big pear 'biner or a regular D... hanging a weight on one side and counter balancing it with a weight on the other you find that a 10lb weight can support a 22lb weight because of the friction. This can complicate a system which depends on the sling to slip around to equalize because 2-to-1 forces have to develop to do it... add more 'biners and you add more imbalance.
The sliding-X was really interesting. Using 9/16" webbing you find one strand sitting on the other. A 10 lb weight held a 65 lb weight statically in this configuration. The perlon sliding-X had the strands sitting next to each other, and the frictional forces were greatly reduced, 10 lbs held 25 lbs, slightly higher than just looping the strands.
If you are going to use the sliding-X, use materials that sit next to each other, not on top of each other! I think the common usage of slings to build a sliding-X should probably NOT be recommended as the equalization properties are defeated.
The tests
The test was conducted by hanging the sling over a 'biner so that the loop made 180º contact with the 'biner. Two weights were hung on either side of the sling. One of the weights was increased until the sling moved. The coefficient of static friction can be calculated by using the formula:
mu = ln(F2/F1)/pi
where ln(F2/F1) is the natural logarithm of the ratio of the greater weight to the lesser weight and pi = 3.1415...
The slings used were a 9/16" wide tubular nylon web, and a 7mm perlon cord.
They were looped both in just a single strand, and in the "sliding-X" configuration.
Two 'biners were used, a large pear 'biner that I use for belaying, rappeling, etc, and a "D" 'biner that I use for "normally" for climbing ropes to anchors, etc. I wrapped Teflon tape on the large pear 'biner to see if it had any effect, it did not.
What I find is that the coefficient of friction is in the range of 0.22 to 0.31 for all slings and 'biner configurations. This gives a force ratio of anywhere from 2 to 2.7 on the 'biner, that is, one strand can have as much as 2 to 3 times the force on it than the other. This would be an issue in a belay rig that is stressed close to the breaking strength of the sling, one side might blow first because the sling isn't able to equalize the force by sliding around the 'biner.
The sliding-X configuration was very interesting. The coefficients were much higher in this case, with the 9/16" sling being in the range from 0.5 to 0.6, able to maintain a force ratio in the range of 5 to 7. This is primarily caused by the fact that one strand of the webbing sits on the other.
The 7mm perlon sliding-X did a better job, the strands sit next to each other. Here the coefficients ranged from 0.29 to 0.4 with force ratios from 2.5 to 3.5.
The data
sling,carabiner,configuration,static,F1 (lbs),F2 (lbs),F2/F1,mu
9/16,large pear,looped,yes,3,8,2.67,0.31
9/16,large pear,looped,no,10,25,2.50,0.29
9/16,D,looped,yes,10,22,2.20,0.25
9/16,D,looped,no,10,25,2.50,0.29
9/17,large pear-Teflon,looped,yes,10,22,2.20,0.25
9/18,large pear-Teflon,looped,no,10,25,2.50,0.29
7 mm,large pear,looped,yes,10,20,2.00,0.22
7 mm,large pear,looped,no,10,22,2.20,0.25
7 mm,D,looped,yes,10,22,2.20,0.25
7 mm,D,looped,no,10,25,2.50,0.29
9/16,large pear,sliding X,yes,10,47,4.70,0.49
9/17,large pear,sliding X,no,10,50,5.00,0.51
9/18,D,sliding X,yes,10,65,6.50,0.60
9/19,D,sliding X,no,10,70,7.00,0.62
7mm,large pear,sliding X,yes,10,25,2.50,0.29
7mm,large pear,sliding X,no,10,30,3.00,0.35
7mm,D,sliding X,yes,10,30,3.00,0.35
7mm,D,sliding X,no,10,35,3.50,0.40
|
|
Paul Raphaelson
Ice climber
Brooklyn, NY
|
|
|
Feb 10, 2007 - 03:02pm PT
|
Has anyone looked at the friction properties of the sliding W?
There's definitely a lot of friction going on. I'm actually less concerned about perfect equalization than I am about what happens if a piece blows and there's extension. I worry that that much cord sliding over itself could cause some burning/shredding ... especially troublesome since the configuration doesn't have built-in redundancy.
If this isn't likely to be an issue, then I'm going to pursue variations on it as a solution.
|
|
Ain't no flatlander
climber
|
|
|
Feb 10, 2007 - 03:31pm PT
|
First, it's "pear" not "pair" though it would be interesting to see how a pair of pears performs. Largo suggests that anodized biners slide better too. Please try your tests with the new generations of 8mm Spectra webbing, which is becoming much more common on many climber's racks.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 04:43pm PT
|
my guess is that you can't tell the difference between the spectra and the nylon tube...
I did change the spelling....somedays are just like that though...
the friction is dominated by the schmutz on the surfaces... we all know that a dirty rope has more friction moving in a belay device.
and the 'biner used to be anodized, but if you use them a reasonable amount they aren't pretty quickly.
multiple 'biners also shouldn't change things, the way they work with a belay or a rap is to increase the contact angle, in this set up it probably wouldn't do that (angle would still be pi).
|
|
jstan
climber
|
|
|
Feb 10, 2007 - 05:23pm PT
|
Sloppy, but hey. Good enough for Topo work.
sling---carabiner--------config.----static---F1(#)---F2(#)---F2/F1--mu
9/16-----lg pear--------------looped-----yes-----3----------8-----=2.67-----0.31
9/16-----lg pear------------- looped-----no------10--------25-----2.50-----0.29
9/16-----D---------------------looped-----yes-----10-------22-----2.20-----0.25
9/16-----D---------------------looped-----no------10--------25-----2.50-----0.29
9/17----- lg pear,Teflon---looped-----yes----10--------22-----2.20-----0.25
9/18----- lg pear -Teflon--looped-----no-----10---------25-----2.50-----0.29
7 mm----- lg pear --------- looped-----yes---10--------20-----2.00-----0.22
7 mm----- lg pear ----------looped-----no-----10--------22-----2.20-----0.25
7 mm-----D-------------------looped-----yes----10-------22-----2.20-----0.25
7 mm-----D-------------------looped-----no-----10--------25-----2.50-----0.29
9/16----- lg pear -----------sliding X-----yes--10-------47-----4.70-----0.49
9/17----- lg pear -----------sliding X-----no----10------50-----5.00-----0.51
9/18-----D-------------------- sliding X-----yes--10------65-----6.50-----0.60
9/19-----D---------------------sliding X-----no----10------70-----7.00-----0.62
7mm----- lg pear ---------- sliding X-----yes--10------25-----2.50-----0.29
7mm----- lg pear -----------sliding X-----no----10-----30-----3.00-----0.35
7mm-----D---------------------sliding X-----yes---10----30-----3.00-----0.35
7mm-----D-------------------- sliding X-----no-----10----35-----3.50-----0.40
Does "No" in the static column mean it is a test of sliding friction? If so, I remember coef. static friction as generally being larger than sliding friction. ?
|
|
Jingy
Social climber
Flatland, Ca
|
|
|
Feb 10, 2007 - 05:31pm PT
|
I never do it. I boulder. No fuss no muss.
|
|
Ain't no flatlander
climber
|
|
|
Feb 10, 2007 - 05:33pm PT
|
I suspect the 8mm spectra will test closer to your 7mm cord due to slipperiness and a shape that is nearer to round.
As for anodizing, if it's shown to make a significant difference then that may be an argument for using hard anodizing on locking biners instead of the soft sulfuric anodizing that is currently used. Nobody makes a hard anodized biner...yet.
c'mon AAC, get your feces cohesive and do something useful for a change.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 07:10pm PT
|
"No" means it did not hold statically, so the actual value of sliding is somewhere in between those two. Since the range is small, probably smaller than the variations in the coeff. I just bracket the values (usually within about 10% of the total weight)...
I also got a BD dynema sling to see what that's all about, but now I'll have to go back and get a new pear 'biner with anodization... which I don't think is work while. The schmutz factor will dominate any surface preparation.
You used to be able to get a nice nylon pully that fit in the 'biner. Something like that could be made out of teflon (or some other slippery material) and used in this "special" case...
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 09:17pm PT
|
sorry, the coeff. is calculated assuming static... this gives the range. Usually when I have it slipping, it is just slipping... so I take the range of static mu to be between the "yes" and the "no" values.
It can't be larger than the "no" value.
|
|
cintune
climber
Penn's Woods
|
|
|
Feb 10, 2007 - 09:21pm PT
|
Petzl still makes those pulley wheels to go on ovals. And DMM has their Revolver biners that are rated to something like 10kN (the pulley part, that is.) More gear, yeah, but if you want to reduce friction.....
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Feb 10, 2007 - 09:43pm PT
|
"Ed, do you think the dynamic coefficient of friction will be low enough compared to the static coefficient to make a difference?
The reason I ask is that the force ratio needed to overcome static friction should occur well before the peak force on the anchor is achieved. Now wouldn't that imply that the cord/slings will be sliding when the peak is reached?"
I think this is a critical point. Somewhere I read a study of the sliding X done by static loading that concluded it was worse than the cordelette at distributing the load. Then Wootle's dynamic tests reached the opposite conclusion, due, I suspect, to precisely the effect DC mentions here.
Here's a reference to the study I mentioned
http://www.t-rescue.com/articles/self_equalising_anchors/index.pdf
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Feb 10, 2007 - 10:11pm PT
|
I should have mentioned that the Geekqualizer
http://www.supertopo.com/climbing/thread.html?topic_id=307091&tn=285
was subjected to testing, although separate load cells on the three anchor points were not available to quantify the presence or absence of equalization.
The rig withstood 11 falls with an 80 kg weight, at which point the climbing rope failed. the first two falls were factor 1 falls and the remaining nine were factor 1.45 falls. The same piece of rope was used, and so the peak force kept going up, from 6.1 kN on the first factor 1 fall to 8.9 kN on the third fall, which was the first factor 1.45 fall. The rest of the falls, with one exception around 10.4 kN, were all around 11 kN.
On the eighth fall, one arm of the Geekqualizer was purposely cut 50%, on the ninth fall the cut was increased to 80%, and on the tenth fall the cut was increased to 90%. The total load on the ninth and tenth falls was 11.4 kN each time. On the eleventh fall a "fuse" was inserted, but the rope began to cut and extend, ending the test.
Once the loads got up to around 10 kN, the clove hitches tended to slip about 3/4 of an inch.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Feb 10, 2007 - 10:31pm PT
|
what kind of detector does Juan use for his seismometers? maybe they could be used instead of load cells?
I also though that you could use a stress guage on a steel rod that is used as the anchors to measure the dynamic loading.
|
|
WBraun
climber
|
|
|
Feb 11, 2007 - 12:04am PT
|
No those won't work Ed (Juans detectors).
Ed if you need load cells come see me next time you're here in the Valley and I'll lend you mine.
|
|
GOclimb
Trad climber
Boston, MA
|
|
|
Feb 11, 2007 - 12:41am PT
|
rgold said: In my possibly mistaken reading of Gabe's diagram, the Mooselette appears to have seven bends around biners compared to the five bends around rings in the Geekqualizer. Two of the Mooselette bends are at relatively large angles, in some cases (e.g. the indoor picture posted earlier as opposed to the sketch) so large as to be insignificant, in which case the number of bends could in some cases be effectively as low as 5, the same as the Geekqualizer. However, the Mooselette has three strands interacting and possibly binding at the power point biner, whereas the Geekqualizer has no such interactions.
You're quite right that in the Mooselette, (or any configuration with multiple strands through a power-point) the strands at the power-point biner might bind in a hard fall. I've never seen it happen from bodyweight, perhaps because cord strands tend to sit next to rather than on top of each other, or perhaps because all the strands at the power-point are moving in the same direction, or because only one is moving more than millimiters. But with a harder fall, or with a piece blowing - I don't know, that's beyond the scope of my testing.
As for the number of biners with tight bends, take another look. Of the 6 biners, as you say, two of them are at angles that are probably insignificant, and one of them is completely fixed (due to the limiter knots on the middle strand). So there are really only three biner interactions at high angle. Of those three, there's only one strand/biner that moves much. Do folks think it matters how much cord must move through a biner, or is any movement equivalent?
By the way, the CharlesJMM and the Equalette look like they should be big winners on the friction front, unless the tiny-movement theory redeems the Mooselette.
GO
|
|
climbingjones
Trad climber
grass valley,ca
|
|
|
Topic Author's Reply - Feb 11, 2007 - 01:08am PT
|
WOW! Thanks for all of the posts to this thread. I knew there was some discussion to be had here. Though I must admit that I dont understand all of the equations that have been posted. I still feel that the sliding "W" is the best, for a regular free climbing anchor anyway. Tying a know in the system anywhere just limits the distribution of the load. We know that the belay is always shifting if you, as the belayer, are hanging from it. I feel the slider works constantly to evenly distibute the load. The shock loading of one piece blowing is minimal, in my opinion, since we routinely rely on one piece to arrest a fall of much more distance than the resulting "fall" in the event that one piece blows in the anchor. But, I still respect the opinion of you much more knowledgeable folks in the physics of this situation. Thanks again. Keep it coming to make this the most responded to post I have seen. ;)
|
|
|
SuperTopo on the Web
|
Let us know!