Discussion Topic |
|
This thread has been locked |
Gobie
Trad climber
Northern, Ca.
|
|
|
Apr 15, 2008 - 02:44am PT
|
I saw a post a few days ago on the triplette and Karl mentioned the same thing. (using screamer to equalize) There use to be a slow mo video around of what happens when a screamer is active. It was common practice to place lockers on screamers and It would seam practical for the belays as well. When a screamer is active it actually causes the gates of the carabiner to flutter. This is not really a problem until the last bar tack rips. At this point the gate could be open at the same time that the sling is fully loaded. The open gate strength of some biners is low and what force that was absorbed could be negated by the gate being open. I will state ahead of time that this is rare, but if your concern is a beefing up weak belay then all links in the system should be considered. I wonder how many people still use lockers on their screamers?? That would be interesting to know. I have had the displeasure of ripping one all the way open and its pretty violent.
|
|
Karl Baba
Trad climber
Yosemite, Ca
|
|
|
Apr 15, 2008 - 03:08am PT
|
One of the alleged advantages of wire gate biners is less gate flutter
FWIW
PEace
Karl
PS Great minds link alike eh?
|
|
Gobie
Trad climber
Northern, Ca.
|
|
|
Apr 15, 2008 - 04:00am PT
|
Im with you Karl. Most wire gates also have an I beam spline which also makes them stronger when open. It is hard to get a wire gate to open by smacking it on its backside with your hand so I would venture to say they would perform well with a screamer.
Kudos
|
|
saho
Ice climber
Anaheim, CA
|
|
|
Apr 15, 2008 - 07:56am PT
|
"This is not really a problem until the last bar tack rips."
"I have had the displeasure of ripping one all the way open and its pretty violent."
How long ago was that violent bar tack rip Gobie?
Modern Screamers are not bar tacked. They are sewn in 3 rows along the length of the webbing, not across, so they tear. My experience falling on them in recent years was that they are quite smooth. I have not seen people using lockers on them, maybe I just did not notice.
-Steve
|
|
AbeFrohman
Trad climber
new york, NY
|
|
|
Apr 15, 2008 - 08:03am PT
|
yea, i think ive heard the gate flutter thing is a non-issue at this point, right?
|
|
Moof
Big Wall climber
A cube at my soul sucking job in Oregon
|
|
|
Apr 15, 2008 - 11:31am PT
|
Nothing worse than having gate flutter when you factor 2.5 onto your microcracked biner.
|
|
Russ Walling
Social climber
Out on the sand.... man.....
|
|
|
Apr 15, 2008 - 11:40am PT
|
Couldn't read all the "higher learning" stuff above... but.... if you have 3 screamers all equalized in your anchor, wouldn't the activation force go up by that same factor? If the screamer goes off at say 600 lbs, would it Knott™ be 1800lbs now with 3 in the system?
|
|
Gobie
Trad climber
Northern, Ca.
|
|
|
Apr 15, 2008 - 11:52am PT
|
Its been a while (90's) isnce I ripped one apart. It was an early yates screamer and it went the whole way. Eddie Joe mite still have it. Bar tacking could be the wrong use of words. Gate flutter is an inherient issue and should be considered even with out screamers. If you are at the point of using screamers at your belay then a lot is going on. In the cordalette thread I think Karl mentioned finding out the stats on belay failure. In the rare cases it happens (alpine and ice mostly) it is probably attributed to somethign that equalization or screamers would not of helped on. I will second my gratitude for being able to climb on Sierra granite where this si generally not a problem. Its all good data, and I like knowing it, but in the real world it just doesnt get applied all that much.
|
|
piquaclimber
Trad climber
Durango
|
|
|
Apr 15, 2008 - 12:10pm PT
|
I just add a screamer to the power point with two lockers if I think there may be a factor 2 fall. (or if my anchor is suspect)
Other than that, I don't use lockers on my screamers.
|
|
Nefarius
Big Wall climber
Fresno, CA
|
|
|
Apr 15, 2008 - 12:17pm PT
|
Gate flutter was only an issue in the older style Screamers. The stitching in the newer (not really so new now) Screamers doesn't have this issue. Some of the other brands still suffer from this - ie. Mammut. But, then again, why would you use another brand?
Come on guys - you know this. We've talked about this a ton here!
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Apr 15, 2008 - 02:08pm PT
|
Steve, your copy of Eric's mathematical analysis is missing some things that would be needed by anyone who would want to invest the time in reading it. Things start out rather badly with
How It Works
= Fc*(2*cos(theta)+1).
There is nothing on the left side of the equal sign, although further reading indicates the quantity should probably be Fa. There is also no description of what Fa, Fc and theta represent. Later on, we encounter the undefined variables Et, Er, Es, Sr, eLoad, and Is. Some of this can be inferred from the text, but frankly, few people (and I am not among them) are going to make the effort it takes to read such accounts if the author or provider hasn't put in the minimal amount of time needed to make the account readable.
Although the account as posted is basically unfit for public consumption, it is still possible to notice, at least at the very beginning, some potentially fatal errors. It appears that Fa denotes the total load applied to the cordelette and that Eric has decided that each strand has equal tension Fc. (This is not simply guesswork; if these assumptions are true, the equation relating Fa and Fc is Fa=Fc*(2*cos(theta) + 1), as contained in the post.) So an analysis that is supposed to deal with the potential inequality of strand tensions begins by assuming all strand tenstions to be equal. If this was true, we wouldn't be putting screamers in there to equalize what is already equalized!
The combination of undefined terms and a beginning assumption that is contrary to the central question suggest that it will not be profitable to try to read on unless and until Eric has a chance to fix up what's posted here.
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Apr 15, 2008 - 02:20pm PT
|
Russ wrote: if you have 3 screamers all equalized in your anchor, wouldn't the activation force go up by that same factor? If the screamer goes off at say 600 lbs, would it Knott™ be 1800lbs now with 3 in the system?
Well, yes, but the assumption for this contraption is that the cordelette doesn't equalize the load to each piece. The point (or one of the points) of having the screamers there is to provide equalization (and at the same time keep the loads on each piece down).
|
|
EB
Trad climber
|
|
|
Apr 15, 2008 - 02:51pm PT
|
Hi, I'm the Eric referred to in saho's original post. The description in saho's post was garbled and unfortunately was missing a lot of important information. I've reposted it below. Note that it is rather long and detailed, however the detail is there only if you want it. The first few paragraphs give a reasonable description of what's going on and detail is presented to allow you to make calculations if you want to. An example caluclation is given as well. If you don't care about the details of the math and just want to see some numerical results, skip to the end. (Note added later: due to several errors, the mathematical part has been removed - a corrected version will be posted later.)
Scream-o-let description:
Introduction
The proposed anchor has one Yates-style screamer (load-limiter) in each of 3 legs. We assume that the 3 legs are close to being equalized, but not quite. Since the legs themselves (not the rope) initially fairly static and have only small stretch, the load in each leg rises rapidly during a fall. One of the legs is shortest. This one will take the load first and the load will increase in this leg faster than in the others until it reaches the screamer-determined load limit and begins to deploy. Since the 3 legs were close to being equalized, almost immediately after the short leg screamer begins to deploy (at constant force), load builds up more rapidly in the other legs (their screamers haven't deployed yet) and moves towards self-equalization. Eventually all 3 screamers will be deploying and the load (by definition) is equalized. This continues until either the fall energy is absorbed, or the screamers reach the end of their fully deployed length, in which case, the anchor returns to its original, not quite equalized state. However, as will be shown below, this is an extreme case. In any case, if the fall energy is so large as to reach this point, much of the fall energy will have already been absorbed.
How It Works
The key point here is the following. In a system that is almost but not quite equalized, individual legs that obey Hooke's law (force is directly proportional to extension) will never self-equalize - the shorter leg will always take the largest force and will therefore extend further, increasing the force in this leg still more. What is needed is something that doesn't obey Hooke's law, but rather is able to extend at constant force. Such an item is readily available in the form of load limiters such as Yates Screamers. Once a load limiter reaches its deployment length, it will continue to extend (here "unzip"), but the force remains relatively constant (for example, see http://www.yatesgear.com/climbing/screamer/index.htm#1 ). Note that once the force on the rope is constant, even though the climber continues to fall the rope is no longer absorbing more energy. Using load limiters in all three legs of an anchor will allow the longer legs to "catch up" with the shorter leg once the latter's deployment begins, until the loads in all 3 legs are equal (this idea is not entirely new - it has been mentioned in one form or another by others in various web forums).
Evolution of Forces
The time evolution of the force in each leg of a 3-leg anchor system is complex and depends critically on both the relative lengths of the legs (as has been pointed out), and on whether each leg is taught or loose to start with. The only thing that is certain is that there exists one leg that will initially take the largest load. The screamer in this leg will be the first to begin deployment. Once deployment starts it will lengthen at approximately constant force (it's actually somewhat noisy as can be seen from data on the Yates web site mentioned above ) until the other legs take up the load, eventually the screamers in all legs will be deploying. By definition, at this point the force in each leg is the same (self-equilization). The important point here is that before screamer deployment, since the spring constant of the static webbing or cord is so high, force builds up with very little change in length. Thus from the time when the first screamer begins to deploy to the time when the last one starts to deploy, the change in length of all arms is small compared to their overall length and for the purpose of estimation it can be ignored. Now, once all screamers are deploying, the net upward force supplied by the anchor to balance the downward force of the rope is related to the angle between the legs. If the single screamer load limit is Fc and the angle between two adjacent legs is theta (assume for simplicity that both such angles are the same - that the anchor setup is symmetric), then the total upward force that the anchor can supply (during deployment) is:
Fa = Fc*(2*cos(theta)+1).
For example, if all legs were in a straight line (for instance, 3 anchor pieces in a vertical crack), theta would be 0 and Fa = 3*Fc. If instead the center leg points straight up and the side legs are each pointing off at 45 degrees from the vertical (theta = 45 degrees), then Fa = 2.4*Fc.
The evolution of force on the rope can be modeled in 4 stages. The first stage goes from the start of the fall to when the force in the rope has built up to the point where the first screamer begins to deploy. In this stage, the force on the rope starts from 0 and goes up to the load limiting value of one screamer.
The second stage goes from the end of the first stage to when the last screamer begins to deploy. In this stage the force on the rope increases in a complex way that depends in detail on the difference in the legs mentioned above. Stage 2 can be quite long. In fact, the fall can end in stage 2, which is good since it means that the load force never got high enough to activate all 3 screamers. However we are most interested in the case where the load rapidly progresses to stage 3 (below). In the most severe falls, stage 2 will be short. We analyze the case where stage 2 is short and argue that this is the worst-case. If the anchor was "equalized" by eye initially then the first screamer will probably have deployed less than 1 cm in stage 2 by the time the last screamer begins to deploy. If the total screamer deployment length is 33 cm (approx.), the energy absorbed by the anchor during this stage is small compared to the total energy absorbed by all of the screamers in a factor two fall (where all screamers fully deploy).
The third stage goes from the end of stage 2 to the end of screamer deployment. In this stage, the screamers in all legs are deploying, the force in all legs is the same and the force on the rope is given by the above equation for Fa.
Stage 4 goes from the end of stage 3 to the end of the fall if the fall energy exceeds the design energy absorption capacity of the screamer system. In this case, the force in the rope will start from its value at the end of stage 3 and increase (smoothly). The screamers are now static and the system is returned to its not-quite-equalized state. However by the time stage 4 is reached a lot more energy has been absorbed as compared to a static anchor just reaching the same load value so the peak load on the anchor should be lower as well.
Estimating Total Energy Absorbed In A Fall
When estimating the total energy absorbed by the system we note that when the fall is such that the system rapidly proceeds to stage 3, to a good approximation we can ignore the energy absorbed by the anchor during stage 2 and consider only the energy absorbed by the rope during this stage. To estimate the total energy absorbed by the system, we estimate the energy absorbed by the rope in stages 1 and 2 and add to that the energy absorbed by the anchor in stage 3. The design goal of the system is to absorb all of the energy in a factor-2 fall of height h above the anchor such that stage 4 is never reached. However it is important to stress that in addition to the self-equlaization property of this anchor system, it has the additional advantage that even if stage 4 is entered, the total load is spread out over a longer amount of time. Thus the maximum load on each protection piece is reduced from that which would be present in even a perfectly equalized, but non-load-limited anchor.
**
EDIT:
The mathematical analysis that was included here has been removed because of several errors that were pointed out by readers. A corrected version will be reposted later.
**
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Apr 15, 2008 - 03:15pm PT
|
Thanks, Eric, that's much better. Sorry if I came off as rather grouchy. And for those not interested in reading Eric's expanded account, let me just mention that the objection I raised does not occur when the equation I complained about occurs in the proper context.
|
|
Karl Baba
Trad climber
Yosemite, Ca
|
|
|
Apr 15, 2008 - 03:37pm PT
|
After carefully pursuing the paragraph spaces in Eric's detailed post, I'd like to heartily concur with his eminent analysis of the pertinent factors.
carry on chaps!
Peace
Karl
|
|
mark miller
Social climber
Reno
|
|
|
Apr 15, 2008 - 06:54pm PT
|
I was hoping the scream a let thread was tied into the sade De marquis thread, what's wrong with you people......
|
|
tolman_paul
Trad climber
Anchorage, AK
|
|
|
Apr 15, 2008 - 07:32pm PT
|
Do you people actually climb, or just calculate theoretical anchors and loads? Do you climb with a calibrated load cell? I swear I work with some of your long lost cousins.
You can chase the idealized anchor to the nth degree, and end up with a tie in point that is at your at your toes, so if your second pulls you off, you'll place a much higher "perfectly equalized" load on your anchor than if you'd used a much simpler but real world practical anchor that you'd taken out all the slack so that a falling second wouldn't pull you off to incrase the load on the anchors.
Not to mention one can outclimb the incoming thunderheads if they are actually climbing vs. building and tearing down complex anchors at the top of each pitch.
Am I missing something?
|
|
Karl Baba
Trad climber
Yosemite, Ca
|
|
|
Apr 15, 2008 - 07:44pm PT
|
"Am I missing something?"
Yeah, I say the same stuff but we have to remember, geekhood is fun and the technology we develop as climbers may actually help the space program someday.
Actually, that was a joke until i remembered that screamers where used by Nasa in some testing awhile back.
Peace
Karl
|
|
tolman_paul
Trad climber
Anchorage, AK
|
|
|
Apr 15, 2008 - 08:11pm PT
|
The funny thing is, I'm an engineer. Fortunately I was exposed to folks that didn't let theories and calculations get in the way of good engineering practices, i.e. you don't need to calculate whether you should use a 1/4" or 5/16" bolt for a load if experiences shows 3/8" is more than enough. I also have worked with many fine "engineers" that never received a degree, but truly were engineers in every sense of the word.
Sorry to spoil the fun, never mind me and carry on 8~)
|
|
Al_T.Tude
Trad climber
Monterey, CA
|
|
|
Apr 15, 2008 - 08:24pm PT
|
It appears that after years of misinformation some climbers are beginning to realize that a standard cordelette can be set up to be EITHER equalized OR non-extending - not both.
The much publicized SERENE anchor (Secure, Equalizing, Redundant and Non-Extending) is a climbing anchor instructional book author fantasy and does not exist in the standard cordelette set-up.
Most commonly promoted is a figure 8 tied in the cordelette which achieves non-extension status, but kills the equalization feature. Adding 3 screamers is a viable solution in high risk situations (probably not worth the time and weight 99% of the time.)
A lighter and simpler solution that I frequently employ is to leave off the figure 8 knot and create a SERE anchor. In MOST situations (where I have 3 decent pieces in), sharing the load more equally between all pieces greatly decreases the chance of any of the pieces pulling. This makes for a much stronger anchor and nullifies the benefit of a non-extending design.
Worst case scenario: Unexpectedly a piece pulls and the (3 pieces of pro) system extends roughly half of the length of the loop attached to the failed piece - typically around 1.2-1.5 feet. This suddenly lowers the power point and drops the falling climber an additional 1.2-1.5 feet plus another few inches of rope stretch. As this is a small portion of the typical fall distance and highly unlikely, I see omitting the Figure 8 as a safer (and faster)choice than the commonly advertised set-up in many situations.
I will confess that I do miss the handy anchor shelf that the figure 8 provides.
|
|
|
SuperTopo on the Web
|
Let us know!