Feeding slack with the GriGri, and understanding this catastrophe potential

See 3:58 in this video:

https://youtu.be/K-A2bVz9AAs?t=238

Here they describe how removing your hand from the brake strand and grabbing the GriGri in the usual way to overcome the camming mechanism is a recipe for catastrophe, since if the leader falls, the cam is no longer able to engage and brake the rope.

I am wondering, how could that possibly happen? I will never squeeze the device hard enough to prevent the cam opening in response to several kN from a leader fall.

In fact, this made me wonder if I am belaying incorrectly with the GriGri. For me, unless I am exactly in time to grab the device and start feeding slack when the leader is making a clip, the cam will engage and I'll be short-roping my poor friend. When this happens, I grab the lever on the device if I need to, but usually it is sufficient to pull out a couple of feet on the brake side, fold it into a loop and grab the device again, which is then easier to get loose with that extra slack on the brake side. I will even incorporate a little hop sometimes to introduce slack into both side of the rope (or equivalently, take a step toward the first piece), then quickly grab the cam with the load decreased, but of course that wouldn't work in a more restricted belay where no hopping is to be had. (this entire paragraph worded poorly, hopefully my dear readers understand)

My usual partner known to wait a few seconds before pulling the rope again if she feel the short-roping, but other climbers (usually those in a panic) will just keep yanking, making it harder to release the brake.

i've had ropes slip past the cam even when fully engaged (usually newer, thin ropes with dry treatment). keep your hand on the brake strand.

Joe Hollowed wrote:

Here they describe how removing your hand from the brake strand and grabbing the GriGri in the usual way to overcome the camming mechanism is a recipe for catastrophe, since if the leader falls, the cam is no longer able to engage and brake the rope.

I am wondering, how could that possibly happen? I will never squeeze the device hard enough to prevent the cam opening in response to several kN from a leader fall.

See this video for proof that it's very easy to override the Grigri cam if you let go of the brake strand:

https://youtu.be/jKe72j_mBlU

Don't let go of the brake strand.

curt86iroc wrote:

i've had ropes slip past the cam even when fully engaged (usually newer, thin ropes with dry treatment). keep your hand on the brake strand.

Of course I always do; I wasn't objecting to the mistake made in the video, I just wasn't sure about the physics of the consequence. All I meant to say was that, from my experience, the cam engages far easier than the video suggests.

How is it that in Kyle's video linked, the belayers hold the cam down with such ease, while a fall pulls meters of rope through the device, but I can't even keep it cammed when the leader is trying to pull rope up for a clip? As stated above, I am able to do this trivially as long as I grab the cam a moment before the leader tries to clip

You need to practice this so that you aren't short-roping the leader.  Part of it is simply anticipating how the rope will move as the climber moves upward, and trying to maintain that perfect amount of slack, as shown in the video.  Anticipating clips can be a bit tricky, but you should be able to tell when the climber reaches down for the rope, unless they are out of sight (in which case, you might discuss having them announce "clipping!" so you won't be taken by surprise.  If you have a comfortable enough stance, you can step closer to the rock as you start to give slack, adding a bit of reach, then step back a bit after the climber clips (while staying close to underneath the first bolt or piece, as shown in the video you posted).  

If you are in the middle of giving slack quickly and the climber pulls the rope enough to engage the cam, you should not have to take a hand off the rope to use the lever--simply pull down (just a bit!) with the guide hand to take the tension off of the device, and then give slack normally, or the quick way.  Unless the climber keeps yarding on the rope continuously, you should be able to unlock the device and throw more slack very fast this way, so the leader is barely inconvenienced.  

I could only make it through 6:30 minutes.

Was the point of this that someone could be stupid enough to hold the cam down until their partner took a groundfall? I believe that.

Joe Hollowed wrote:

Here they describe how removing your hand from the brake strand and grabbing the GriGri in the usual way to overcome the camming mechanism is a recipe for catastrophe,

Letting go of the brake strand is NOT the usual way. The current Petzl recommended technique for fast-feeding has the belayer hooking their index finger under the "lip" to hold the Grigri up, and only press the cam with their thumb if it locks up. The thumb should be released as soon as the rope is feeding again, so you're only depressing the cam for a split second. If you're constantly short roping your leader, your rope may be too fat and fuzzy and/or you're not managing the slack very well. A Grigri 1 works better with fatter ropes.

https://youtu.be/vbQ-sIBPbFU

Fun little video of someone belaying like this and dropping their climber from the second bolt to the ground. Non graphic, climber uninjured, but you can can see the belayer burn their hand on the rope as it zooms through the grigri.

See 1:06  https://www.petzl.com/US/en/Sport/Belaying-with-the-GRIGRI?ActivityName=rock-climbing

While giving slack with the index finger supporting / thumb on the cam should be used. If the leader falls one will not be clamping down on the cam. The rope stays in the brake at all times. All other techniques are not correct and potential for cratering. 

Joe, if I'm understanding you correctly (which is highly questionable) the difference is that in the scenario you're describing, you're trying to defeat the cam once there is already tension on the climber's strand from him pulling up rope. In the video they're talking about a fall happening while the cam is already defeated, so the rope just continues to move through the device.

Joe Hollowed wrote:

See 3:58 in this video:

https://youtu.be/K-A2bVz9AAs?t=238

Here they describe how removing your hand from the brake strand and grabbing the GriGri in the usual way to overcome the camming mechanism is a recipe for catastrophe, since if the leader falls, the cam is no longer able to engage and brake the rope.

Someone already beat me to this, but it doesn't hurt to say it again.

This is not the 'usual' or correct way to lead belay with a GriGri. The correct way is both easier and safer.

Joe Hollowed wrote:

 I can't even keep it cammed when the leader is trying to pull rope up for a clip? As stated above, I am able to do this trivially as long as I grab the cam a moment before the leader tries to clip

If you are having trouble paying out slack with the method shown about 10-20 seconds earlier in that video, there is something missing in your technique. Unless your rope is mega fuzzy and/or fat it's incredibly easy to pay out slack using that method.

In fact, this made me wonder if I am belaying incorrectly with the GriGri. For me, unless I am exactly in time to grab the device and start feeding slack when the leader is making a clip, the cam will engage and I'll be short-roping my poor friend. When this happens, I grab the lever on the device if I need to, but usually it is sufficient to pull out a couple of feet on the brake side, fold it into a loop and grab the device again, which is then easier to get loose with that extra slack on the brake side. I will even incorporate a little hop sometimes to introduce slack into both side of the rope (or equivalently, take a step toward the first piece), then quickly grab the cam with the load decreased, but of course that wouldn't work in a more restricted belay where no hopping is to be had. (this entire paragraph worded poorly, hopefully my dear readers understand) 

I'm not really sure if I understand you correctly, but what you are describing sounds pretty bad. The video you linked does a good job of explaining the correct way to do this. You should never need to grab the lever in order to pay out slack when lead belaying. I just did an 'experiment' with an old, fuzzy rope in my garage. I clipped an old, fuzzy rope into the GriGri then yanked  to engage the cam and held firm tension on the rope (as if I was a panicked leader, attempting to pull clipping slack). Even with that rope, releasing the cam to pay out slack using the correct method was trivially easy.

Joe Hollowed wrote:

Here they describe how removing your hand from the brake strand and grabbing the GriGri in the usual way to overcome the camming mechanism is a recipe for catastrophe, since if the leader falls, the cam is no longer able to engage and brake the rope.

I am wondering, how could that possibly happen? I will never squeeze the device hard enough to prevent the cam opening in response to several kN from a leader fall.

Your brake hand should never leave the brake strand. You can easily leave the rope through your middle, ring, and pinky fingers with your index finger under the lip. Your thumb wraps over the top to lightly press down on the cam. The lever is never used. Your thumb provides enough pressure to keep the cam from engaging but a surprise fall will easily move the grigri and/or overpower your thumb so that the cam can engage because you don't have much leverage on the cam. If you're firmly grasping the cam you're unlikely to release it reliably and your partner will be on the ground in a second or two. The only time your brake hand needs to leave the GriGri is to take up slack and then it should immediately return to that position. I prefer to PBUS as catching a fall is paramount and it keeps the motion consistent for when I use an ATC.

In fact, this made me wonder if I am belaying incorrectly with the GriGri. For me, unless I am exactly in time to grab the device and start feeding slack when the leader is making a clip, the cam will engage and I'll be short-roping my poor friend. When this happens, I grab the lever on the device if I need to, but usually it is sufficient to pull out a couple of feet on the brake side, fold it into a loop and grab the device again, which is then easier to get loose with that extra slack on the brake side. I will even incorporate a little hop sometimes to introduce slack into both side of the rope (or equivalently, take a step toward the first piece), then quickly grab the cam with the load decreased, but of course that wouldn't work in a more restricted belay where no hopping is to be had. (this entire paragraph worded poorly, hopefully my dear readers understand)

My usual partner known to wait a few seconds before pulling the rope again if she feel the short-roping, but other climbers (usually those in a panic) will just keep yanking, making it harder to release the brake.

You're right about belaying incorrectly. Watch some videos, hire a guide, or get more experienced friends to teach you. The lever should only need to be used to lower someone. Moving into the draw can be used to help defeat the cam with your thumb. I wouldn't jump as coming back down may now pull your leader off the wall. A leader with some experience should at worst be yanking cyclically instead of with a constant pull which should let you defeat the mechanism in between yanks. If they're panic pulling then you'll have more trouble. You're probably climbing with newer leaders too which makes things more panicky on both of your parts. Experienced leaders usually have a smooth pulling motion and will let off for a moment if they encounter resistance.

If they're clipping you should be feeding out some rope already. When they grab the rope after clipping a draw in I tend to feed out one armlength and reset my climber's side (left) hand in case I need to give more slack. If they drop the rope I can easily extend my brake side hand and take in almost all of that initial slack in one quick motion. The main thing that screws up belaying are twists in the rope and you can run your left hand down the rope a bit between clips to make sure there aren't any coming up soon.

This is one part of climbing that you really don't want to screw up. If you can't learn from 5-6 different very experienced belayers then recruit a guide. Messing this up can easily result in crippling injuries or death for the other party. I've only ejected one person from my climbing partners and it was from safety issues including bad belay technique.

Dan Gozdz wrote:

Your brake hand should never leave the brake strand. You can easily leave the rope through your middle, ring, and pinky fingers with your index finger under the lip. Your thumb wraps over the top to lightly press down on the cam. [...] The only time your brake hand needs to leave the GriGri is to take up slack and then it should immediately return to that position.

I agree with everything you said except the last part: the only time your brake hand should be in that position is when you're fast-feeding rope, the rest of the time your brake hand should be off the Grigri, holding the brake strand just like with an ATC.

You mean to tell me THIS (screen grab image) isn't the OFFICIAL, SAFE way??

Joe Hollowed wrote:

...I just wasn't sure about the physics of the consequence. All I meant to say was that, from my experience, the cam engages far easier than the video suggests...

The physics will take longer than this morning's coffee, but the summary is that when fall arrest does not happen promptly and correctly, bad things happen to climbers. 

Not the physics exactly, but as a consequence of physics, 3 "stool legs" of belaying; take out any leg and the belay is no longer a belay; from the German Alpine Club research:

1) The brake hand never leaves the rope (braking strand).

2) The belay system is always arranged so that it is in a braking configuration.

3) The "panic reaction" of the belayer produces effective braking.

If any of these elements are not in place, the climber is not on belay.

Speaking for myself, if I'm on-belay, I'd like to be on belay, always, even if the situation is outside of my belayer's experience. 

As a "fun" example: climber and belayer in the gym; climber falls; belayer pulled up into air; rope pulled hard into GriGri so it doesn't easily feed; belayer uses both hands to pull on the GriGri lever to unjam the rope; climber and belayer free-fall to ground. The example is "fun" because padded gym floor and no climber injury. But shows what happens when the brake hand is not on the rope, the system is not rigged for braking, and the panic reaction does not produce braking; and bonus point for situation outside of belayer experience. In this example, all three stool legs were absent, but as other posters note, there are plenty of videos on the internet that show what happens when just one is absent; and if you watch other climbers at the gym/crag, you'll see that missing stool legs is not uncommon; which in turn leads me to believe that we'll be seeing more of those videos and that Accidents in North American Mountaineering won't run out of content for future issues...

A V wrote:

If you give a dynamic belay (walk back and forth as you feed and take up slack) you don't even have to touch the cam most of the time

I see very few people doing this and it really does limit the amount you need to move/adjust your hands, therefore less room for error. Of course assuming you're in a position where its safe to step back/move around)

Joe Hollowed wrote:
I am wondering, how could that possibly happen? I will never squeeze the device hard enough to prevent the cam opening in response to several kN from a leader fall.

The tension on the rope is not the same as the force necessary to hold down the cam...

Not going too deep on this, but consider: the cam is activated by friction of the rope moving along it, not the tension in the rope. Static friction can have a greater magnitude than kinetic friction. The placement of your thumb on the cam (when using proper technique per Petzl) is at a greater radius from the pivot than the rope, so if they applied equal force your thumb would win due to geometry. 

J C wrote:

the cam is activated by friction of the rope moving along it, not the tension in the rope.

I don't think this is correct.  I think that:

Friction isn't what engages the cam.  If it was, it wouldn't matter if you were holding the brake strand or not, because your hand on the brake strand doesn't change the friction between the rope and the Grigri internally.

Rope tension engages the cam.  The rope is bent going through the Grigri, and under tension the rope "wants" to straighten out.  In addition, due to the vector of the rope exiting the Grigri relative to the pivot point, tension on the climber's side of the rope wants to rotate/lift the cam, which causes it to pinch the brake side.

I believe a quick test demonstrates that this is the case.  If you pull rope through slowly, the Grigri doesn't engage.  If you pull rope through at the same speed but also apply some tension to the brake side, it will engage and lock every time.  The friction hasn't changed, but the tension has.

This may be a nitpick, but I think it's important to properly understand and describe the way a Grigri works, if we're going to use that information to make decisions about how to use the device.

I agree that rope tension activates the cam; perhaps I was a bit sloppy with my wording, but what I mean't was that the rope and cam interact through friction, rather than say a rope tied off to an object. This was to illustrate that dropping your partner doesn't require holding their weight on your thumb.

Since the rope runs concentrically to the cam, I believe the only way it interacts to begin locking is through friction. Put another way, a GG with a totally frictionless cam would never lock. Do you agree?

J C wrote:

Since the rope runs concentrically to the cam, I believe the only way it interacts to begin locking is through friction. Put another way, a GG with a totally frictionless cam would never lock. Do you agree?

I was doing the same thought experiment earlier!  Actually, I think a frictionless rope would lock (if you were holding the brake strand, and didn't have the same "frictionless" situation there).  The rope basically does a 180-degree bend through the Grigri, so tension on the rope means it's both trying to straighten out and pulling "up" on the cam pivot.  I think the rope can be totally frictionless and static and it will still lift the cam into the braking position.

Here's a really crappy photo I marked up.  "F" shows the approximate forces due to rope tension.  Fn is the net force, which you'll note is offset to the left of the pivot, because the cam is offset.  This results in a torque (T) which will rotate the cam.  All of these forces are due to rope tension and do not require friction.

Interesting, I will give that some thought when I can get to my GriGri. It would be interesting to really break down the contributions from friction, rope bending, rope getting squished, etc. Makes me wish I had thought of doing something like that for my undergrad physics project.