Climbing Science - How many KN does a lead fall in a climbing gym generate?

I asked so many people how many kilonewtons the belayer, bolt and climber would see in a typical lead fall in a climbing gym and I got such a huge range of guesses.  Carabiners are rated for 24ish KN and ropes are rated for.... well, super good enough.  So we are going to more tests like this so tell us what else you want to see tested!  9:48 has the chart with the numbers if you want to skip the video.

So the z-fall took the highest impact on the protection point. Kind of logical, since it's more static. However I'd be curious to see how not perfectly aligned draws (as they tend to be in gyms) would fare. Would tend to think that between the z-fall and purely straight, you may get lower impact forces on the anchor, before it gets back up as you get closer and closer to a static fall.

Do you know what the fall factor would be on those? I bet you could get the measurements/approximates from the gym. Would be interesting to know that as well. Great job.

Tal Wanish wrote: Some literature on it from Petzl - great at always. Looks like you're not going above 5kN until you're close to a factor 1 fall, which a large proportion of people will never see.

https://www.petzl.com/US/en/Sport/Forces-at-work-in-a-real-fall

Cool idea Ryan, I'm just getting into trad and it's always reassuring to see more examples like this.

I just checked out this article.  It is pretty good and I like how they limited the variables and 3 tests.  It would have been nice to see the range of forces and not just averages but a good article.

Franck Vee wrote: So the z-fall took the highest impact on the protection point. Kind of logical, since it's more static. However I'd be curious to see how not perfectly aligned draws (as they tend to be in gyms) would fare. Would tend to think that between the z-fall and purely straight, you may get lower impact forces on the anchor, before it gets back up as you get closer and closer to a static fall.

Do you know what the fall factor would be on those? I bet you could get the measurements/approximates from the gym. Would be interesting to know that as well. Great job.

I think the last test was more like what you are describing with Ryan doing the overhang.  We plan on doing outdoor tests and will be able to test lots of very realistic scenarios.  It was 35 to 40 feet high and the fall was about 20 feet.  I'll bring my laser next time and measure it out so we can start building a model on our data.  

Swing by Denver, we can add some upper bound data for you (Mike 2.0 is still a super light human, from a certain point of view).

Also,

Great content as always, Ryan. I've never highlined, but I've become an avid fan of your video series.

Have you seen the testing on using a "direct belay" vs. harness belay? I've added a link below:

The results defy reason. Is a classic "dynamic belay" really so small a contributor to force reduction when compared to the rope slippage though a munter-hitch?
I'd love to see more data [wink wink].

The numbers don't add up because of the friction in the intermediate draws and even with none in the system they still don't due to the time delay created by the rope, doing belay device testing we know that the peak faller force and the peak belayer force occur at different times and the peak force at the top piece is when the sum of the forces is highest.
The karabiner standard is 20kN (ovals 18kN), not 24.

Jim Titt wrote: The numbers don't add up because of the friction in the intermediate draws and even with none in the system they still don't due to the time delay created by the rope, doing belay device testing we know that the peak faller force and the peak belayer force occur at different times and the peak force at the top piece is when the sum of the forces is highest.
The karabiner standard is 20kN (ovals 18kN), not 24.

The devices are not measuring at the same instant. so shouldn't each device record the peak despite the time delay created by the rope? Or are the devices recording the first peak and missing a second one (when the sum is maximal)?

Cool idea for the video!

Obviously shows in general terms what we knew already (e.i. highest number in the top piece, consistently slightly higher than the sum from belayer and climber)
The observation that was news to me: the belayer gets roughly half of what the climber gets. (Until you go into the static/drag falls, which seems obvious) 
​
but beyond these rough approximations, I wonder about reproducibility. Is the difference between numbers such as 2.72 and 2.84 real, or is it just noise? What’s the error here if you repeat the same fall with the same climber/belayer 5-6 times?

How much does the rope factor into this? (Personal experience, with no numbers, is that my rope is a lot “stretchier” — higher dynamic elongation, picked on purpose— and that makes the falls much softer than the gym rope falls, or some of my partner’s stiffer ropes, so I expect the numbers on climber’s side to change based on the type of rope used, but would like to know how much it actually changes)

Do you change the rope every time? If not, how different would the results be, if you do change it after every fall? And if you don’t change it, how different the results would be if you change the order of climbing tests? E.g. if the short fall happens after the long fall. Or the static/z-drag go first, and then you do the long fall.

Fran M wrote:

The devices are not measuring at the same instant. so shouldn't each device record the peak despite the time delay created by the rope? Or are the devices recording the first peak and missing a second one (when the sum is maximal)?

Watch the video again, it is questioned why the numbers don't add up. One would think the peak faller and belayer forces should equal the peak top piece force but they don't. The peak top force is when the sum of the forces on the faller strand and the belayer strand is highest.

Lena chita wrote: Cool idea for the video!

Obviously shows in general terms what we knew already (e.i. highest number in the top piece, consistently slightly higher than the sum from belayer and climber)
The observation that was news to me: the belayer gets roughly half of what the climber gets. (Until you go into the static/drag falls, which seems obvious) 

but beyond these rough approximations, I wonder about reproducibility. Is the difference between numbers such as 2.72 and 2.84 real, or is it just noise? What’s the error here if you repeat the same fall with the same climber/belayer 5-6 times?

How much does the rope factor into this? (Personal experience, with no numbers, is that my rope is a lot “stretchier” — higher dynamic elongation, picked on purpose— and that makes the falls much softer than the gym rope falls, or some of my partner’s stiffer ropes, so I expect the numbers on climber’s side to change based on the type of rope used, but would like to know how much it actually changes)

Do you change the rope every time? If not, how different would the results be, if you do change it after every fall? And if you don’t change it, how different the results would be if you change the order of climbing tests? E.g. if the short fall happens after the long fall. Or the static/z-drag go first, and then you do the long fall.

You have to use a completely new section of rope each time for the results to be vaguely accurate, even the Petzl are wise enough to round up to 0.5kN..

Jim Titt wrote:

Watch the video again, it is questioned why the numbers don't add up. One would think the peak faller and belayer forces should equal the peak top piece force but they don't.

Still don't understand why the delay between peaks would alter the sum of peak forces. The cells are not measuring at the same instant, but recording the peak at whatever instant they happen.

The peak top force is when the sum of the forces on the faller strand and the belayer strand is highest.

Aren't the peaks of those strands the ones recorded at climber and belayer respectively?

(I honestly don't get where that extra force is)

Fran M wrote: Still don't understand why the delay between peaks would alter the sum of peak forces. The cells are not measuring at the same instant, but recording the peak at whatever instant they happen.

Aren't the peaks of those strands the ones recorded at climber and belayer respectively?

(I honestly don't get where that extra force is)

If you just want to add the two peak forces together then fine BUT as I wrote above the peak force on the top piece is when the sum of the forces on the two strands is highest and generally ( probably always) this occurs at a different time to the measured faller/ belayer peaks. You lay the two force curves on each other (which are offset in time) and measure or calculate what the peak sum is. The peak force measured at the belayer and faller are obviously at different times to the peaks in the strands at the top piece as well but this is hard to measure and  unimportant so for convenience can be ignored.

Fran M wrote: Still don't understand why the delay between peaks would alter the sum of peak forces. 

Sum force diagram - all forces in the system add up to zero.

Very simplified
Force_top_Piece - Force_climber + Force_non_climber
Force_non_climber=Force_belayer+Friction_1st_draw+Friction_2nd_draw+Friction_3rd_draw+...

The numbers we see are - Force_top_Piece, Force_climber, and Force_ belayer, no numbers about friction losses on draws. So, can't expect them to add up.

Those friction forces add up, I suspect Friction_1st_draw is likely to be highest in gym setting were QD lines are pretty straight and the belayer is creating rope line angle change by standing to the side of 1st qd.  Additionally, there could be additional friction losses due to rope running over holds.

Doing the test only with the top piece would result in easier to interpret math.

Jim Titt wrote:

If you just want to add the two peak forces together then fine BUT as I wrote above the peak force on the top piece is when the sum of the forces on the two strands is highest and generally ( probably always) this occurs at a different time to the measured faller/ belayer peaks. You lay the two force curves on each other (which are offset in time) and measure or calculate what the peak sum is. The peak force measured at the belayer and faller are obviously at different times to the peaks in the strands at the top piece as well but this is hard to measure and  unimportant so for convenience can be ignored.

Thank you, I think I am starting to picture it. Could you mention the order in which the peaks happen for a non-tethered belayer?

From memory it's faller peak first and belayer last but I'd have to check.

Bump for awesome testing! I love seeing this type of climbing science :D

Dan Daugherty wrote: Neat experiment, but there are so many variables in play that aren't accounted for. This is more a real world empirical data gathering exercise vs. a true science experiment. If constant lengths of rope were used, attachment of the belay device to a static anchor, drop of a static weight from a constant height, and probably a couple other variables I'm missing were accounted for, these numbers would mean something from a scientific context.
Not bashing the work done here at all. Looks like a fun time, honestly, but I'd hate someone to come upon this thread, see the word science and think the numbers are accurate in that context.

I definitely use the word science loosely for sure, but what does limiting variables such as using a fresh, new rope every time we test something achieve?  The gear nerds that do these types of tests sterilize their tests so much that it almost doesn't become real life anymore.  I love throwing variables in there, not letting the rope rest, lifting the belayer up different every time, and even using a different belay device.  And static weight is not a human body as we are just bags of water, so when tests are done with blocks of steel we can graph a concept but don't actually know what would happen in our most common situations in real life.  We will do follow up videos and if we are trying to see do a ATC vs GriGri comparrison I will have to limit some variables in order to even see the difference but I like what you said about this being a real world empirical data gathering exercise.  This was kind of a summary of gym lead falls as a general concept since some were guessing 1kn and some 13 kn when I asked prior.  We will do more, with better dynos, and try to list out the variables that we have in the experiment next time!  Cheers

Dan Daugherty wrote: Neat experiment, but there are so many variables in play that aren't accounted for. This is more a real world empirical data gathering exercise vs. a true science experiment. If constant lengths of rope were used, attachment of the belay device to a static anchor, drop of a static weight from a constant height, and probably a couple other variables I'm missing were accounted for, these numbers would mean something from a scientific context.
Not bashing the work done here at all. Looks like a fun time, honestly, but I'd hate someone to come upon this thread, see the word science and think the numbers are accurate in that context.

All research studies have limitations and scope conditions...