Gate flutter - what/how does that work and real-world analysis

From the YGD myths threads, "gate flutter" came up:

Cheiftan Mews wrote:

Gate flutter is real, although I think it's mainly only a concern on runouts. I carry a slider locker (or two) that I'll put on my last bomber piece before a runout or hard section. Very useful in Eldo.

speculated gate flutter incident

Didn't know about that, and here's a slow-mo (although not taken during a realistic event, just a simulation).

I don't know if I necessarily support the conclusions from that accident report, that being said. A pretty much as likely explanation to me would be that he bad-clipped 2 placements (although even then, with long runners, it's unclear to me how this would have mattered).

What I'd like to understand is this:

1. How can gate flutter lead to a rope unclipping in practice? Unless there is some other gate-fluttering mechanism other than that shown in the slow-mo video, there needs to be tension at the non-rope end of the rope carabiner (e.g. where the sling is clipped). With a long sling, I don't see how that could happen UNLESS the rope is pulling on the rope-end carabiner and the sling is taunt all the way to the placement. Then perhaps some swinging action would smack the rope-end carabiner and flutter the gate open, as in the slow-mo video. However, that gate opening action is fast, and for that to happen the system must be tensioned... how is the tensioned rope supposed to get out of the carabiner? Any realistic prospects for this? I could see how this could lead to carabiner failure (as the gate is opened, perhaps at peak load). But I'm not so convinced about unclipping the rope.
2. Vibrations? Meaning that some mechanism could hit the resonance frequency of a carabiner and open the gate, for example if the sliding of the rope through the carabiner or perhaps bumping of carabiner accross the rock as it swings? I have a harder time picturing this. I haven't looked into it in a long time, so perhaps I'm rusty, but from what I recall one way to get ride of resonance effects is to add dampening. Seems like most gate mechanism have a pretty good restorative action (E.g. the spring) which would make resonance unlikely. Or are those vibrations NOT about resonance? But then how is that supposed to work?

Think it is more about breaking the carabiner than unclipping.  Strength goes from 20kN+ to ~7kN when gate flutters open.  If massively loaded at same time, biner breaks, you die, etc.  High energy fall, spine of biner slams into wall at same time load comes on rope, gate is open at wrong moment.

Franck Vee wrote:

That's also my thinking. But the linked accident reports has no broken carabiners (just not clipped to rope anymore).

If you can contact Jeff Lea (CT climber) he may be able to tell you more. I did a freelance piece about carabiners, and he sent me some info. about gate flutter, testing that was done in Europe  25-30 (?) years back.

First, there are two types of things that seem to be called gate flutter.  One is the inertial bounce of a gate when the spine of the carabiner smacks something--you can demonstrate by just smacking a solid-gate carabiner against your palm.  Wiregate carabiners eliminate or at least mitigate this problem, which is the reason they were invented.  The other is a far more mysterious phenomenon involving oscillations induced by the rope stretching and moving over the carabiner.  This was reported in European tests years ago, but I don't think it has been consistently replicated and of course no one knows whether or not this happens in field conditions.

Either way, gate flutter is a momentary displacement of the gate and can't by itself explain unclipping.  It is a possible explanation for the carabiner breaking, because the gate-open strength of a carabiner is low enough for breakage to occur under high but realistic fall loads.

I've read about at least three unclipping incidents that point in another direction.  In all the incidents, the top piece failed and then the next piece down unclipped.  A study done by Mark Beverly and Steven Attaway, beverlymountainguides.com/w…, revealed that when a piece fails, there is a significant rope "snap-back" in the slack created, and the rope that snaps back curls chaotically.  These unpredictable twists and curls could account for the rope unclipping from the next piece down, and this effect is confounded with possible back-clipping of the piece. Less obviously but not entirely implausibly, the carabiner motions generated by the twisted slack might cause the next carabiner down to unclip from the pro. Here's an example of this phenomenon:  https://www.climbing.com/videos/weekend-whipper-big-trad-climbing-fall/

None of these has anything to do with gate flutter.

There is a take-away from this for situations in which one has high marginal gear and good gear below backing it up.  The good gear should be clipped with lockers so that it can resist the chaotic rope snap-back gyrations that will occur if the top piece fails.

rgold wrote:

A study done by Mark Beverly and Steven Attaway, beverlymountainguides.com/w…, revealed that when a piece fails, there is a significant rope "snap-back" in the slack created, and the rope that snaps back curls chaotically.  These unpredictable twists and curls could account for the rope unclipping from the next piece down, and this effect is confounded with possible back-clipping of the piece. Less obviously but not entirely implausibly, the carabiner motions generated by the twisted slack might cause the next carabiner down to unclip from the pro.

i basically agree with this.  i have witnessed several unclipping incidents. on on occasion i was at a belay about 6 feet above a steep fist crack pitch. a guy below me had just clipped a #4 friend and fell shortly after. the rope unclipped and he took a pretty good fall.  he looked up at me and asked me if had screwed anything up, and i responded no, everything looked ok.

so he starts climbing up again, clips the friend again (everything looks good), climbs a bit further and falls again. and again the rope unclipped from the biner and he took another pretty good one.  he was pretty freaked out so he lowered off.  i downclimbed, retrieved his friend, and returned it to him later.

i think what was happening was that as he fell the rope on the belayers side of the biner was rubbing on the rock or something to cause it to move more slowly downward than the rope on the climber's side, which created somwhat of an upside down "U" shape, which somehow unclipped. to this day i am still pretty shocked about it.

knowing wayne, and being somewhat familiar with his accident, i am still pretty wigged about that one as well.  he placed 7 pieces in the first 30 feet, slung out his other pieces, doubled up an iffy placement, used a screamer, etc.  wayne was definitely bold, but he was also a smart guy who had shit pretty dialed.

IIRC there was a similar incident on leaning tower a while later, where the rope unclipped from multiple slings.  with slings, i think it is a bit dicier in some ways because they have no torsional resistance and can twist really easily, which would be a major problem for the upside down "U" rope trick.

Some are quick to suggest that all unclipping incidents are due to user-error back-clipping, which is plausible for the rope unclipping from the biner, but seems less so when the biner unclips from the sling. Using a couple of lockers in certain situations seems prudent, and in some cases it seems that placing psychological pro might actually put you in more danger than not placing it at all. Weird shit happens. 

here's another weird one that happened to me.  my partner was leading a single pitch gear route at squamish and it had this kind-of crappy traversy/roof thing. hard to describe, but he couldn't get a multi-directional piece there. he managed to get an ~ok stopper that was good enough to get him up to the next piece.  then he got a few good ones in, no problem.  later he falls, and that one not-great stopper absolutely shot out of the crack.

this is the crazy part - the biner that was clipped to the rope came un-clipped, and the sling and stopper shot out over my head.  later that day i looked for it, and found it up in a tree probably 75 feet away from the route.  completely nuts! (no pun intended).

Franck Vee wrote:

From the YGD myths threads, "gate flutter" came up:

Didn't know about that, and here's a slow-mo (although not taken during a realistic event, just a simulation).

Not sure it’s a good comparison in the video as he’s holding the two biners in opposite orientations 

rgold wrote:

In all the incidents, the top piece failed and then the next piece down unclipped. 

I’ve witnessed several unclippings.  The top piece doesn’t need to fail in order to generate the cited effect - ie, I don’t think a supposed loop or wave traveling down the rope is by itself the problem, nor is the piece failing the only way to generate such a thing.  A hard fall is a rather violent disturbance to the entire system.

The most common denominator from what I’ve seen is an unconstrained biner on the rope end.  Crill’s accident, for example, was 2 24” slings with unconstrained Neutrino biners.

My experience and from what I've seen - distant memories of things that may/may not have happened, rumors, climbing  situations that I think I've seen or I may have heard of, technical expertise and knowledge I don't have, incomplete accident reports that I didn't read, random, freak outliers that happened to other people I never met or talked to - blah, blah, blah.

rgold wrote:

I've read about at least three unclipping incidents that point in another direction.  In all the incidents, the top piece failed and then the next piece down unclipped.  A study done by Mark Beverly and Steven Attaway, beverlymountainguides.com/w…, revealed that when a piece fails, there is a significant rope "snap-back" in the slack created, and the rope that snaps back curls chaotically.  These unpredictable twists and curls could account for the rope unclipping from the next piece down, and this effect is confounded with possible back-clipping of the piece. Less obviously but not entirely implausibly, the carabiner motions generated by the twisted slack might cause the next carabiner down to unclip from the pro. Here's an example of this phenomenon:  https://www.climbing.com/videos/weekend-whipper-big-trad-climbing-fall/

None of these has anything to do with gate flutter.

There is a take-away from this for situations in which one has high marginal gear and good gear below backing it up.  The good gear should be clipped with lockers so that it can resist the chaotic rope snap-back gyrations that will occur if the top piece fails.

I can add two data points. 

First one was a 20m aid fall where the penultimate piece was extruded through its placement (an RP4, becoming an RP2.5 in the process). I felt a big tug when this happened and the 'snap back' was probably substantial. The rope sustained damage to the sheath and two strands of the core in the fall. The carabiner attached to the piece that held had a piece of sheath in the notch in the nose. It seems that the rope had somehow jumped into the notch and then back to the basket again. Carabiner was a lightweight solid gate DMM. I returned it to DMM for inspection and they pronounced it was working fine.  

Second one was a 20m free climbing fall which stripped multiple pieces of gear. I was the belayer on this occasion. One wire + quickdraw managed to be pulled and also detach itself from the rope. Lightweight wire gate carabiners.  It landed in the sea where it was unfortunately unretrievable but I have no reason to suspect it was faulty in any way.

In long, gear-ripping, falls all kinds of odd things can happen. 

Here's a brief video that shows how a twist in a long runner could make the rope-side carabiner detach.  Chaotic rope motions could certainly put such a twist in a long runner...  

rgold wrote:

Here's a brief video that shows how a twist in a long runner could make the rope-side carabiner detach.  Chaotic rope motions could certainly put such a twist in a long runner...  

Great video, thanks. In know this issue can be addressed by using lockers on key placements, but what about clove hitching the rope-end carabiner to the sling? Is there any downside? 

Heather S wrote:

Great video, thanks. In know this issue can be addressed by using lockers on key placements, but what about clove hitching the rope-end carabiner to the sling? Is there any downside? 

From a pragmatic pespective, I suspect just the effort of tying the clove hitch.

Derek DeBruin wrote:

From a pragmatic pespective, I suspect just the effort of tying the clove hitch.

I was thinking you’d already have a subset of draws pre-rigged. 

Heather S wrote:

I was thinking you’d already have a subset of draws pre-rigged. 

Sure. I agree that prep on the front end would be the better move than mid-pitch. That begs the follow-up of why a subset and not all of them? Having a hard time thinking the chance of back clipping would increase in any meaningful way and can't think of any other downsides at the moment.

Derek DeBruin wrote:

Sure. I agree that prep on the front end would be the better move than mid-pitch. That begs the follow-up of why a subset and not all of them? Having a hard time thinking the chance of back clipping would increase in any meaningful way and can't think of any other downsides at the moment.

Yeah, not sure to be honest. I just started learning trad recently so I was just being conservative and assuming it might be better to have some standard/non-clove-hitched alpines but I honestly can’t think of why. Maybe in case you want to put a locker on a key piece? 

Franck Vee wrote:

From the YGD myths threads, "gate flutter" came up:

Didn't know about that, and here's a slow-mo (although not taken during a realistic event, just a simulation).

I don't know if I necessarily support the conclusions from that accident report, that being said. A pretty much as likely explanation to me would be that he bad-clipped 2 placements (although even then, with long runners, it's unclear to me how this would have mattered).

What I'd like to understand is this:

1. How can gate flutter lead to a rope unclipping in practice? Unless there is some other gate-fluttering mechanism other than that shown in the slow-mo video, there needs to be tension at the non-rope end of the rope carabiner (e.g. where the sling is clipped). With a long sling, I don't see how that could happen UNLESS the rope is pulling on the rope-end carabiner and the sling is taunt all the way to the placement. Then perhaps some swinging action would smack the rope-end carabiner and flutter the gate open, as in the slow-mo video. However, that gate opening action is fast, and for that to happen the system must be tensioned... how is the tensioned rope supposed to get out of the carabiner? Any realistic prospects for this? I could see how this could lead to carabiner failure (as the gate is opened, perhaps at peak load). But I'm not so convinced about unclipping the rope.
2. Vibrations? Meaning that some mechanism could hit the resonance frequency of a carabiner and open the gate, for example if the sliding of the rope through the carabiner or perhaps bumping of carabiner accross the rock as it swings? I have a harder time picturing this. I haven't looked into it in a long time, so perhaps I'm rusty, but from what I recall one way to get ride of resonance effects is to add dampening. Seems like most gate mechanism have a pretty good restorative action (E.g. the spring) which would make resonance unlikely. Or are those vibrations NOT about resonance? But then how is that supposed to work?

If the carabiner on the rope end  spins during the fall on a shoulder sling…. It may have been loaded on the gate, and unclipped.   I know the injured climber, and it was amazing I wasn’t the one belaying him that day.  He doesn’t bad clip placements, let alone two in a row.

Those videos that show the violence of the rope running through things should help explain a bunch.  

Heather S wrote:

Yeah, not sure to be honest. I just started learning trad recently so I was just being conservative and assuming it might be better to have some standard/non-clove-hitched alpines but I honestly can’t think of why. Maybe in case you want to put a locker on a key piece? 

Fair enough. I really can't think of much else but figured it was worth consideration.

I'm not sure I'll make the leap and clove hitch all my slings (it would commit you to a dedicated rope-end carabiner, which isn't something everyone does). However, I'm certainly gonna make sure to fix it anytime a sling makes a round turn around the carabiner to minimize the chance of this kind of unclipping.