Mountain Project Logo

Another Accident due to mis-use of the Gri-gri

reboot · · . · Joined Jul 2006 · Points: 125
rgold wrote:In the case that started this, the belayer's hand was badly burned, for whatever you want to do with that piece of information.
You can touch a running rope & burn your hand. You can hold your brake hand above the device, right against the device and burn your hand without much meaningful brake force.

The Grigri is just a dumb mechanical device, and was found to be functional after the accident. It does not know whether it's catching a slab fall or a big one but softened by the belayer grabbing the lead end. So which is more likely? That there was something magically interfering with the cam or the belayer fucked up?
csproul · · Pittsboro...sort of, NC · Joined Dec 2009 · Points: 330
reboot wrote: You can touch a running rope & burn your hand. You can hold your brake hand above the device, right against the device and burn your hand without much meaningful brake force. The Grigri is just a dumb mechanical device, and was found to be functional after the accident. It does not know whether it's catching a slab fall or a big one but softened by the belayer grabbing the lead end. So which is more likely? That there was something magically interfering with the cam or the belayer fucked up?
Who's arguing that the belayer did NOT fuck up? I've not see RG do so! What he's arguing is that the mechanism up this fuck up (ie pulling on climber side) is a possible mode of overriding the camming action of the Grigri. And that this mechanism is not only possible, but that you may not have enough friction in the uncammed device to control a fall, even with a properly applied brake hand. Why is that so hard to believe?
reboot · · . · Joined Jul 2006 · Points: 125
csproul wrote:I'd rather err on the side of caution and assume that under certain circumstances, the Grigri cam CAN be overridden, and that if it is, there is not adequate friction to arrest a fall.
That statement makes absolutely no sense from a mechanical stand point. Applying enough brake force will provide enough frictional force to engage the cam unless there's something interfering it. Not applying enough brake force will NOT, as we can feed the rope through the device by just gently pulling on the climber side (and in fact, as long as you keep the acceleration under a threshold, you can keep speeding up the rope through the device. Those are pretty simple physics. But I guess people like to believe in voodoo magic.
csproul · · Pittsboro...sort of, NC · Joined Dec 2009 · Points: 330
reboot wrote: That statement makes absolutely no sense from a mechanical stand point. Applying enough brake force will provide enough frictional force to engage the cam unless there's something interfering it. Not applying enough brake force will NOT, as we can feed the rope through the device by just gently pulling on the climber side (and in fact, as long as you keep the acceleration under a threshold, you can keep speeding up the rope through the device. Those are pretty simple physics. But I guess people like to believe in voodoo magic.
Let me say this again...the cam CAN be overridden. IF you deny that, you clearly aren't paying attention. The mechanism may be in question, but if the cam were not overridden somehow, then nobody would have hit the ground. It would have engaged. Period. IF the cam is kept from working, there is little to no evidence that the belayer can provide enough friction to arrest a fall.
reboot · · . · Joined Jul 2006 · Points: 125
reboot wrote: You can hold your brake hand above the device, right against the device and burn your hand without much meaningful brake force.
I should add that getting pulled up while grabbing the climber's end increases the likelihood of this happening if you didn't brake hard enough initially, but that does not mean grabbing the climber's end is the actual problem.
reboot · · . · Joined Jul 2006 · Points: 125
csproul wrote: Let me say this again...the cam CAN be overridden.
Yes, by blocking it, not by pulling down the climber's end (that only makes the cam a bit harder to engage, same with a slab fall). And you don't know the cam was overridden or simply did not engage. And you CAN burn both of your hands in the latter scenario.
bearbreeder · · Unknown Hometown · Joined Mar 2009 · Points: 3,065

folks the grigri is a sound device ... and perhaps the "safest" device if used PROPERLY and with ATTENTION

the problem is of course that many folks dont use it properly and dont understand the failure modes ... and they get lazy belaying with it as it "always locks up"

as long as theres nothing interfering with the cam (including the climbers left hand pulling hard, you pull down HARD on the brake, and the rope is within spec ... the grigri will lock up every time

however the moment something interferes with the cam all bets are off ... more than once ive fallen several past more bolts than needed because the belayers were using the "fast feed" method and werent quick enough with the reaction time

its amazing when you have time to think how many months being dropped will put you out of climbing for as you hear the rope whiz through the grigri... fortunately they recovered in time in each case by eventually pulling own HARD on the brake

the grigri is a great device ... but its requires skill, experience, practice and vigilance ..

its not a device for the beginner or the inattentive ... those folks need to work on their brake hand

many folks even on MP believe its a "hands free" device ... nor do they understand the failure modes ... a simple search through the forums and some posts at the beginner of this thread show this

heres another one that happened recently and just got posted this past week ...

This happened at my local gym, where I had just started lead climbing. I was reaching for the 4th draw when I lost my grip and fell. My belayer was new to using a GriGri, he was holding it open to pay out slack and didn't get his hand on the brake end until it was too late. I decked hard and right away knew it was serious. My foot was twisted sideways in an unnatural looking way, and I couldn't move it at all.
I was taken by ambulance to the nearest hospital, where they attempted to reduce the dislocation. They pulled and tugged on my foot, but it wouldn't budge. Then I was taken for x-rays where they discovered fragments of my talus were blocking the bone from moving back into place.
The on-call orthopedic surgeon arrived and took a look at my case. She told me it was an unusual break, and not something she was comfortable attempting. I was transferred to another hospital where they found a surgeon willing to attempt putting my ankle back together.
It took 3 surgeries over 3 weeks to get everything back where it should be, but I'm now on the road to recovery. I should be attempting to walk in another six weeks.


reddit.com/r/climbing/comme…

csproul · · Pittsboro...sort of, NC · Joined Dec 2009 · Points: 330
reboot wrote: Yes, by blocking it, not by pulling down the climber's end (that only makes the cam a bit harder to engage, same with a slab fall). And you don't know the cam was overridden or simply did not engage. And you CAN burn both of your hands in the latter scenario.
Overridden or not engaged is kind of semantics. Either way the cam is not working. There may have been any number of things contributing to this. Blocked Grigri due to physical interference, new slick rope, slow fall/rope drag (although stated to not be the case), grabbing the climber side of the rope...We'll never know. What I'm challenging is the notion that the Grigri will behave like an ATC should the cam be overridden. In the absence of proof, I don't believe that it will.
eli poss · · Durango, CO · Joined May 2014 · Points: 525

of course it's not enough friction to arrest a fall without the cam engaging. a tube style device uses the brake force to pinch the rope against the locking carabiner. this is why you get different amounts of friction with different shaped carabiners (roundstock vs I beam). with a gri-gri, there is no pinch on the carabiner if the cam doesn't engage so you essentially get the friction that you would if you just wrapped the rope around a metal bar. If they don't understand the concept then it isn't worth arguing about, natural selection will do its job.

Healyje · · PDX · Joined Jan 2006 · Points: 422
bearbreeder wrote:folks the grigri is a sound device ... and perhaps the "safest" device if used PROPERLY and with ATTENTION
Really? Good to know as I've been getting the feeling that someone somewhere on is being dropped at least once per hour 24x7x365. That would be about 24 people a day, 8760 per year planet-wide. Doesn't seem all that unlikely given there are probably a few million grigris out there at this point.

In fact, if anything, that feels like it's a tad bit on the low side.
PRRose · · Boulder · Joined Feb 2006 · Points: 0

It would seem to me that if (1) the cam failed to actuate,(2) the belayer had control of the brake strand, and (3) there was insufficient friction in the system to allow the belayer to catch a fall despite the cam failing to actuate, then the belayer's hand would be pulled into the device. The characteristic injury would therefore not be rope burns only, but rope burns plus a severe pinching injury (to the web between thumb and forefinger, or the fleshy part below the pinky, depending on which way the belayer was holding the brake strand).

Are such pinching injuries actually observed?

Greg D · · Here · Joined Apr 2006 · Points: 883

^^^ I think you are confused. The mere fact that rope burns occurred on either hand demonstrates the belayer is not in control of the rope strand, breakfast strand, climber strand, guitar strand, tomato strand. They all burn when slipping.

Jim Titt · · Germany · Joined Nov 2009 · Points: 490
rgold wrote:Some open-cam Grigri tests would be interesting. But in addition to using a new slick thin rope, make sure to use a 100 lb belayer (under the hypothesis that grip strength is very roughly proportional to body weight). Climbing is a strength-to-weight game, but when it comes to hanging on to a loaded rope, strength is all that matters, lighter people will need more help from their device, and one of the factors influencing the choice of a Grigri over an ATC is precisely the concern that the belayer might have trouble with control. Before lobbing leader falls (which sounds dangerous to me---what if the belayer can't hang on) I'd try some static tests. The climber hangs on a single strand on their device (atc, Grigri with cam taped open) as if rappelling, and weight is added to the braking strand until the climber is held in position without slipping. The climber's feet can be inches above the ground so no risk involved. I'm not able to do this myself right now or I'd do that test. The ratio of the climbers weight to the holding weight would give an estimate of the "force multiplying factor" for the belay device (a linear concept that has been used in a number of papers in spite of appearing, according to some of Jim Titt's tests, to be not at all linear). This could at least settle the question of whether or not an open Grigri provides at least ATC-level braking.
I tested the GriGri 1 threaded backwards and with the cam immobilised some years back and as reported by Dobson it´s a piss-poor device like that, I couldn´t get better than 70kg resistance with one hand. If I remember the discussion at that time was lowering a climber back down after you discovered it was threaded wrongly and through a draw and both hands it was o.k but to stop a fall for sure not.
It´s actully not that easy to duplicate most of the scenarios described in this thread, with most of the ropes the thing locks up no matter how hard one tries to stop it by cable-tieing the cam down etc. I haven´t tested the GriGri 2 much since I prefer to use the GriGri 1.
Some ropes however give loads of problems, thinner ones with a fine sheath weave and more extensive dry treatment are particularly a hassle and I´m not the only researcher who doesn´t use certain brands of rope for belay device testing for this reason (I use a GriGri as the rope lock on the other end when testing belay plates). Since even basic ropes come with some sort of dry treatment these days I give them a good dose of dust before I climb with them, the rope companies efforts to reduce friction are not universally appreciated:-)
slim · · Unknown Hometown · Joined Dec 2004 · Points: 1,103
This post violated Rule #1. It has been removed by Mountain Project.
slim · · Unknown Hometown · Joined Dec 2004 · Points: 1,103
csproul wrote: Overridden or not engaged is kind of semantics. Either way the cam is not working. There may have been any number of things contributing to this. Blocked Grigri due to physical interference, new slick rope, slow fall/rope drag (although stated to not be the case), grabbing the climber side of the rope...We'll never know. What I'm challenging is the notion that the Grigri will behave like an ATC should the cam be overridden. In the absence of proof, I don't believe that it will.
why would it not behave like a an atc? it basically comes down to the belt friction equation.
Mort · · Unknown Hometown · Joined May 2004 · Points: 20
csproul wrote: Who's arguing that the belayer did NOT fuck up?
Hey Folks,

I know it's a lot to ask, but try to have a little compassion. We are talking about a 13 or 14 year old girl whose father is in the hospital for a long stay.
Mort · · Unknown Hometown · Joined May 2004 · Points: 20
slim wrote: why would it not behave like a an atc? it basically comes down to the belt friction equation.
Oh yeah, the trusty old "belt friction equation." Known to every MountainProject rocket scientist. Listen Slim, go grab your Grigri and your ATC and run a rope through them. I have mine right here. The ATC over a fat locker has a radius of about 3/16" and when engaged PINCHES the rope between biner and ATC. This is NOTHING like an open Grigri, which has a radius of about 11/16" and most certainly does not pinch the rope against anything. The biner isn't even involved.

Shove those facts into your trusty "belt friction equation."
csproul · · Pittsboro...sort of, NC · Joined Dec 2009 · Points: 330
slim wrote: why would it not behave like a an atc? it basically comes down to the belt friction equation.
Test it out and let us know. So far three separate people on this thread alone have tested a backwards threaded Grigri or one with the cam held closed and found it inadequate. Let us know if you find something different.
rgold · · Poughkeepsie, NY · Joined Feb 2008 · Points: 526

I apologize in advance for actually taking claims involving math and physics seriously. I hope that folks not interested in technical discussions will skip directly to the last paragraph.

Slim’s comment about a disengaged Grigri having the same belay characteristics as an ATC because of the capstan equation suffers from two fatal flaws.

If all you had to do was chant "belt friction equation," then all ATC's would have precisely the same braking behavior; something I assume everyone knows to be false. This means that the phrase "behaves like an ATC" isn't meaningful---which ATC with what geometry and which additional features, like grooves—which are not accounted for by the capstan equation—are we speaking of?

So the first error is in asserting that because the capstan equation applies, ATC’s and Grigri’s must have the same belay characteristics. But it turns out that there is another error, which is whether the capstan equation actually applies. If you look at the capstan equation and the assumptions underlying its derivation, several things emerge. 1. It is assumed that the rope is not slipping. 2. It is assumed that the rope is not elastic. 3. It is assumed that frictional forces are linear with respect to the normal force. 4. It is assumed that no forces are required to bend the rope. For an analysis taking all these items into account, see for example ningpan.net/publications/10…. You’ll need a math, physics, or solid engineering background to read this however, and it will help to realize early on that what they refer to as a "rod" is what we would call the rope.

I don’t know much about the significance of Items 1—3, but it turns out that Item 4 is significant. It is well-known in mechanical engineering that pulleys with large diameter sheaves are more efficient that pulleys with small-diameter sheaves, a fact that you can find mentioned in Petzl’s ad copy for its rescue pulleys. This particular failure of efficiency is not related to friction or the capstan equation, but comes instead from the force required to bend the rope around the sheave. Since efficiency reduction is significant for pulleys, where virtually no friction is present, it follows that bending forces ought to play a significant role in the amount of friction supplied by ATC’s and Grigris. If you’ve read this far but your eyes are still glazing over, the practical take-away is that the bending radii are important and not just the bending angles, which is all the capstan equation deals with.

If memory serves, Jim has done tests that indicate that work done in bending the rope around the “capstans” in a belay device accounts for more than half (I think close to 60%) of the braking effect. This means that rope path and capstan effects are roughly equal in producing braking results, and since there is a significant rope path difference between Grigris and ATC’s, it is inevitable, as Dobson and Jim have already indicated, that a disengaged Grigri will have less braking power than an ATC and can in no way be considered equivalent. In fact, both have found barely enough effect to hold a climber’s body weight, depending on the amount of friction in the system, the actual weight of the climber being held, and the grip strength of the belayer.

The point I’ve been trying to raise from the beginning is that it is possible to envision a set of circumstances in which a particular belayer might be physically incapable of holding a particular leader. Actually, it is well-known that this is true for ATC-type devices and is a feature of the warning panels in most instruction sheets. This thread is about an unusual set of circumstances in which the same might be true for a Grigri. The heat is generated by those who say that there are no such circumstances if the belayer behaves correctly. It is hard to know whether this claim has content or is just a definition of correct belayer behavior; one would have to be able to test the situation with real belayers, and it is very difficult to reproduce the unusual Grigri behavior at will.

I can report on one aspect of the situation that comes from bygone days when we actually used test the belayer with high fall-factor falls using weights, and that is this: for a fifty-foot fall you have about 1.8 seconds to stop the leader before they hit the ground. If for any reason you momentarily lose control in a high-load situation, it is almost impossible to get it back in that time frame.

Let’s suppose that I’m wrong (hardly an unusual circumstance) and there is no such thing as situation that would physically overwhelm a belayer even though the belayer does everything correctly. Then we can still say that the problem brought to our attention by John at the beginning of this thread is one that will, at the very least, greatly increase the demands on the belayer, way past anything that they are used to dealing with when using a Grigri. Forgetting all the arguments about unconditional belayer responsibility, isn’t the prospect of a sudden huge increase in the performance demands on the belayer enough, all by itself, to merit serious consideration of John’s warning? And isn't it right, in terms of the actual tragedies that have happened, to let it go at that?

Ray Pinpillage · · West Egg · Joined Jul 2010 · Points: 180
rgold wrote:I apologize in advance for actually taking claims involving math and physics seriously. I hope that folks not interested in technical discussions will skip directly to the last paragraph. Slim’s comment about a disengaged Grigri having the same belay characteristics as an ATC because of the capstan equation suffers from two fatal flaws. If all you had to do was chant "belt friction equation," then all ATC's would have precisely the same braking behavior; something I assume everyone knows to be false. This means that the phrase "behaves like an ATC" isn't meaningful---which ATC with what geometry and which additional features, like grooves—which are not accounted for by the capstan equation—are we speaking of? So the first error is in asserting that because the capstan equation applies, ATC’s and Grigri’s must have the same belay characteristics. But it turns out that there is another error, which is whether the capstan equation actually applies. If you look at the capstan equation and the assumptions underlying its derivation, several things emerge. 1. It is assumed that the rope is not slipping. 2. It is assumed that the rope is not elastic. 3. It is assumed that frictional forces are linear with respect to the normal force. 4. It is assumed that no forces are required to bend the rope. For an analysis taking all these items into account, see for example ningpan.net/publications/10…. You’ll need a math, physics, or solid engineering background to read this however, and it will help to realize early on that what they refer to as a "rod" is what we would call the rope. I don’t know much about the significance of Items 1—3, but it turns out that Item 4 is significant. It is well-known in mechanical engineering that pulleys with large diameter sheaves are more efficient that pulleys with small-diameter sheaves, a fact that you can find mentioned in Petzl’s ad copy for its rescue pulleys. This particular failure of efficiency is not related to friction or the capstan equation, but comes instead from the force required to bend the rope around the sheave. Since efficiency reduction is significant for pulleys, where virtually no friction is present, it follows that bending forces ought to play a significant role in the amount of friction supplied by ATC’s and Grigris. If you’ve read this far but your eyes are still glazing over, the practical take-away is that the bending radii are important and not just the bending angles, which is all the capstan equation deals with. If memory serves, Jim has done tests that indicate that work done in bending the rope around the “capstans” in a belay device accounts for more than half (I think close to 60%) of the braking effect. This means that rope path and capstan effects are roughly equal in producing braking results, and since there is a significant rope path difference between Grigris and ATC’s, it is inevitable, as Dobson and Jim have already indicated, that a disengaged Grigri will have less braking power than an ATC and can in no way be considered equivalent. In fact, both have found barely enough effect to hold a climber’s body weight, depending on the amount of friction in the system, the actual weight of the climber being held, and the grip strength of the belayer. The point I’ve been trying to raise from the beginning is that it is possible to envision a set of circumstances in which a particular belayer might be physically incapable of holding a particular leader. Actually, it is well-known that this is true for ATC-type devices and is a feature of the warning panels in most instruction sheets. This thread is about an unusual set of circumstances in which the same might be true for a Grigri. The heat is generated by those who say that there are no such circumstances if the belayer behaves correctly. It is hard to know whether this claim has content or is just a definition of correct belayer behavior; one would have to be able to test the situation with real belayers, and it is very difficult to reproduce the unusual Grigri behavior at will. I can report on one aspect of the situation that comes from bygone days when we actually used test the belayer with high fall-factor falls using weights, and that is this: for a fifty-foot fall you have about 1.8 seconds to stop the leader before they hit the ground. If for any reason you momentarily lose control in a high-load situation, it is almost impossible to get it back in that time frame. Let’s suppose that I’m wrong (hardly an unusual circumstance) and there is no such thing as situation that would physically overwhelm a belayer even though the belayer does everything correctly. Then we can still say that the problem brought to our attention by John at the beginning of this thread is one that will, at the very least, greatly increase the demands on the belayer, way past anything that they are used to dealing with when using a Grigri. Forgetting all the arguments about unconditional belayer responsibility, isn’t the prospect of a sudden huge increase in the performance demands on the belayer enough, all by itself, to merit serious consideration of John’s warning?
Correct me if I'm wrong but the layman's take away from the engineering is that the GriGri has lower bend angles and greater contact surface radii contributing to reduce imparted friction on a braking rope when the cam is not engaged.

It should be obvious just by looking at a GriGri that it does not function like an ATC without the cam.
Guideline #1: Don't be a jerk.

General Climbing
Post a Reply to "Another Accident due to mis-use of the Gri-gri"

Log In to Reply

Join the Community

Create your FREE account today!
Already have an account? Login to close this notice.

Get Started.