Build My Own FeedQC Lab: To Screamer or not to Screamer?
The following post comes direct from QA Manager Corey LaForge, who returned from a recent ascent of the Eiger Norwand with some questions about Yates Screamers that he promptly put to the test here in our lab.
A couple of months ago I climbed the North Face of the Eiger with some friends. On one particular pitch above the Waterfall chimney, I was climbing some harder mixed moves right above the belay. My gloved hands groped a rounded rail, while my crampons scraped on marginal limestone edges and a thin layer of verglasse. The single piece between me and my belayer was a half-pulled-out, bent over piton. As I moved further above this piece, questions formed in the back of my mind: "Will that manky pin hold?" "Should I have a Yates Screamer on that thing?" "Do Screamers even help?" Luckily, I didn't have to find out the answer to any of these the hard way. As soon as I got back, however, I decided to do some drop tests in the lab to give myself some rough answers to these questions.
Test Setup
I roughly re-created the climbing scenario I described above in the Drop Tower, as shown in Figure 1 below.
- The "climber"-an 80 kg (176 lb) mass which runs on frictionless tracks-was tied into a 9-foot (2.74 m) section of 10.1 mm dynamic rope with a figure-8 knot which was pulled hand tight
- The mass was raised 4.5 feet (1.37 m) above the first "piece of gear"-a quick draw clipped in to the load cell --for measuring the forces generated in the fall
- The other end of the rope was tied to the "belayer"-a fixed point 4.5 feet (1.37 m) below the piece of gear-with a figure-8 knot which was pulled hand tight
- The "climber " mass was dropped from this height on to the single piece of gear resulting in Factor 1 fall of 9 ft (2.74 m) with 9 ft (2.74 m) of rope in the system
- A new rope was used for each drop, but it should be noted that this set up is designed for repeatability and simplicity, so it does not account for all dynamics in the real world belay
- Figure 1: 9-foot (2.74 m) Factor 1 Drop Test Setup Schematic
Test Variables and Hypothesis:
We decided to do this same drop using an example of all of the possible "quickdraw" configurations while holding all other variables (knots, biners, new rope, drop height) the same for all drops to test the effect of the dogbone on the force that the piece of gear sees. We did the same 9 ft, Factor 1 drop on the following samples with the same Hotwire on the rope end and a Quicksilver on the gear end:
- Steel Chain
- Dynex Dogbone
- Econo Nylon Dogbone
- Variwidth Nylon Sport Dogbone
- New Standard Screamer
- Old-Version Standard Screamer
We hypothesized that the steel chain would not stretch at all, resulting in the highest force to the piece of gear. We presumed that the Dynex runner would behave about like steel, while the two types of nylon runners would stretch absorb slightly more energy, and ultimately that the Screamers would absorb the most energy to reduce the peak force by upwards of 50% when compared to the baseline peak force established in the drop with the chain.
Results: 9-foot Factor 1 Fall on to 1st Piece of Gear
Analysis
From the data above, it became clear that we were not too far off in our hypothesis: the Screamers did reduce the peak force on the first piece of gear, while nylon absorbed a bit of energy, and Dynex behaved similarly to steel chain. The Screamer did reduce the peak force by 14-17%, or a reduction of 1.4-1.8 kN.
We figured that the very severe nature of a Factor 1 fall very near the anchor is perhaps not showing us the difference between Screamers and other dogbones as well as it could because the forces generated in such a fall are WAY above the activation force for the Screamer. We decided to repeat the experiment with a lower fall factor to get peak forces that are less severe and more typical of a real climbing scenario.
Revised Test Setup-Lower Fall Factor
I repeated the experiment on a new sample of each type of dogbone and another new Standard Screamer. Again, a new rope with hand-tied Figure 8 knots at each end was used for each drop.
This time, I set up the fall scenario with the piece of gear the same 4.5 feet (1.37 m) above the belay, while the climber's waist was 1 foot (0.30 m) above the piece of gear as shown in Figure 2 below. This created a 2 foot (0.60 m) fall on 5.5 feet (1.68 m) of rope, or Fall Factor 0.36.
Results: 2-foot Factor 0.36 Fall on to 1st Piece of Gear
Analysis
With the Factor 0.36 drop, the force generated was generally around 40% lower than the Factor 1 drop. These lower peak forces allowed us to see the difference between each dogbone type and the benefits of the Screamer more clearly. In this case, the dynamic stretch of each dogbone was noticeable over the static steel chain. Again, the Screamer was able to reduce the peak force by 1.6 kN, but this contribution was very considerable as a percentage, reducing the peak force that the piece of gear is subjected to by 26%.
Conclusion
The forces generated in falls right above the belay, where very little dynamic rope is in the system are very high. Forces in the neighborhood of 10 kN applied to your piece of gear will be very hard on small stoppers, micro cams, and screws. You should do everything in you power to prevent falls like this when possible! Our limited testing does show, however, that a Screamer (or similar energy absorbing device) could reduce the peak force that the piece is subjected to by up to 26%. This may be enough to keep that manky pin or marginal nut in place. Remember, our testing is not the real world, and using knots instead of a real belay device, as well as a rigid mass instead of a squishy human body are factors to consider which may make our results different than the manufacturer's claims. The bottom line is: they do work to reduce the peak force applied to the piece of gear in the system.
Other considerations for scenarios such as this should also be to always give a dynamic belay, and if you should take that whipper, come back to the belay and switch ends of the rope! The rope is the most key piece of energy absorbing equipment. If I would have done these falls in succession on the same rope, you would see forces go through the roof-with a Screamer or otherwise. That's a set of tests for another day...stay tuned!
Corey LaForge
United States / English 
26 Oct 2012, 1:39AM
Always good to get this kind of beta, I´m taking two to patagonia this season!!
31 Jul 2012, 3:59PM
Posting is almost a year old, so not sure if anyone looks at this anymore - but ....
Drop testing (at least the very little that I've seem as compared to someone who does it for a living) typically has the caveats that it is not "real world" due to rigid masses and tied-off knots instead of belay devices, etc.
Has anyone tried to make drop-testing more real-world? I am thinking :
1) instead of a hard tie-off for a "simulated belayer", use the drop-rope threaded through an actual belay device which is hooked to a reasonable mass (80 kg) which is free to lift. Allows the "belayer" to be jerked off the ground, reducing gear loading.
2) on free end of the rope which passes through the belay device, hook to a smaller mass (say 20-40kg) to simulate the force that can be exerted by the belayer's hand/arm holding onto the rope. Allows for reasonable amount of "rope slip" through the belay device, further reducing gear loading.
3) for dropped mass and belayer mass, instead of solid masses, use some kind of "deformable" mass, like a sack of potatoes strapped with bungie/etc. into a "non-natural" shape such that sudden acceleration will cause it to deform, but not change total mass.
It would be interesting introduce these changes one at a time in order to see how the results change, with say 5-10 datapoints per change.
Or you could just pony up have a real climber tie into rope and fall on the load cell with a real belayer and compare it to a similarly "tied off" lab drop-test :)
3 Nov 2011, 11:13AM
It would be nice to know how much the belayer has to do with reduction of forces. A good belayer should be able to contribute just as much of a reduction in force as any piece of gear. Communication is the key.
19 Oct 2011, 6:59PM
This is a huge area of discussion among ice climbing professionals. As a member of the AMGA ice instructor pool, I am often fielding questions about the applicability of screamers in an ice climbing scenario. My best evidence had often been anecdotal, and while this test is certainly not comprehensive it offers a good insight into how different links work in the protection chain. Thanks for making the effort to check it out, now decide if its worth it and give us a BD version. Is the technology already there in BD's via ferrata lanyards?
19 Oct 2011, 5:51PM
Interesting anecdotal data. I've heard people say that screamers increase the load in high fall factor situations because the climber falls further as the screamer rips. This data suggests otherwise.
1 Sep 2011, 4:22AM
Do you want to comment on the fact that in the second test the Dynex dogbone absorbed more energy than the Variwidth nylon one? This seems inconsistent with the previous test and your initial predictions.
28 Aug 2011, 7:50PM
I have some old screamers in good nick I'd be willing to donate.
19 Jul 2011, 10:02AM
Every lead climber cam benefit from the info here Thanks!
Need a BD load limiter...
13 Jul 2011, 3:21PM
Instead of tying the rope to the ground it would be a good idea tie it to another 80kg mass. The small amount of upward motion of the mass can reduce peak forces significantly. This may allow the screamer to absorb a greater percentage of the peak force before bottoming out. Its much more difficult to replicate a true dynamic belay the small slip of the rope through a belay device, stepping forward, or a small jump. I wonder if its possible to build a load limiting GriGri/ATC that would allow the rope to slip through the belay device if the forces were to great?
10 Jul 2011, 2:12PM
Nice to see some tests on this subject, just one question, what rope was used and/or what was it's impact force?
10 Jul 2011, 8:01AM
Thanks for an interesting test. I've always wondered why BD doesn't make a load-limiting draw. Any insights? Keep up the strong work!
9 Jul 2011, 11:13PM
you should try and test the alpine screamer!?( A shoulder length Dynex sling with multiple overhand knots tied between the carabiners.) much lighter than a few screamers.
9 Jul 2011, 12:17PM
What was the impact force of the rope you used? 10.1 diameter makes me think single rope.
It would be sensible to use a half rope with a screamer because they are softer(and better in the multi pitch environment, which this test is presumably aimed at). not many people would use screamers in conjunction with single ropes in the UK.
perhaps you should rerun these tests with a half rope, I would be very interested to see that.
I would also be interested in knowing how two screamers side by side worked with the hard falls. one would assume that activation load would be doubled, but would peak load halved? would dissipation be smoother or more pronounced?
it would be interesting to see the load / time graph
nice article and thanks for publishing, my comments are meant to be constructive criticism.
8 Jul 2011, 6:43PM
This is super useful and remember taking my first whipper on two rusty pitons in the Gunks, they held, but this more harrowing tale reminds me of it and will make me re-consider and tied-off pieces as well as ice-screws in the winter!
Alexandre, good to stumble into you. Were you thinking of going to Rocktober in Kentucky at the Red River Gorge in October?
8 Jul 2011, 6:24PM
Thanks for the analysis Corey! After meeting you in the Gunks in May its nice to see your work. Maybe a QC lab on 3 sigma would be enlightening? Anyway, thanks for justifying the screamer in my trad gear!
8 Jul 2011, 5:41PM
Alex-
The activation force of a screamer is only about 2kN, that is about double body weight which will only cause about 8% stretch in your rope. A dynamic rope will give around 30%.
It was really good to see this testing, hopefully this will quell some of the concerns. I personally use screamers particularly on pieces close to the ground. I also have first hand experience of one being deployed and it definitely saved my butt...
8 Jul 2011, 4:18PM
@Alex They are still two springs in series, so all energy absorbed prior to the pro (read: elongation from rest state), is from your fall. Your rope's elongation does not create additional energy, only you can do that.
8 Jul 2011, 3:57PM
Which still leaves me with the same question I've had for years: why don't more companies make more 'screamers?' Why no BD screamer?
8 Jul 2011, 3:44PM
Clay-
The reason that the reason that the old screamer was excluded from the test was because I only had one. It came from my personal stash of old stuff. I would have loved another data point for the old screamer too, but alas, I did not have any more to test. There is no compelling evidence saying the old is better than the new…
8 Jul 2011, 7:31AM
I would suggest that anyone reading the article above also considers the comments below before deciding on how reliable the results are (unless appropriately amended).
Was there any replication done with this test? It does not look like it from the results presented.
It is not possible to draw valid conclusions without reference to the sample size, appropriate standard error and statistical significance. It is misleading to present information with such limitations or without reference to the details. Percentage difference between single tests offers at best a distant indication of what may happen with an appropriate samples size for each treatment (e.g. n=21) and analysis (e.g. P
8 Jul 2011, 12:35AM
What about temperature dependence? Seems like a screamer would behave differently in cold alpine conditions, where plastic isn't as stretchy (and certainly if you get below the glass transition temperature...).
7 Jul 2011, 9:54PM
Thanks for the test and a chance for up date. I will consider taking one for the first runner out from the belay. I too feel strongly about placing lots of pieces if possible fresh out from the belay and not using a Gregory for any trad setting at all.
Thanks again
Paul Rogers.
7 Jul 2011, 8:55AM
Nice test. You guys should have used an alpine style draw too. It probabbly would hav been the same as the dyneema draw but the doubling up might have lowered the forces?
7 Jul 2011, 7:54AM
I Noticed that the second iteration of the experiment did not include a test for the "Old Screamer." What gives? Does the old screamer out-perform the new?
7 Jul 2011, 6:55AM
Great analysis, thanks!
I had heard that screamers were next to useless because the rope would lose its stretch during the stitch deployment, and be closer to a static line when the screamer bottoms out, generating very strong forces on the piece. I find it very interesting that your test doesn't seem to show this happening (or rather that even with this factor, there is still a substantial reduction in the force applied).