Edelrid MegaJul Belay Device

Brian wrote: How much rope slippage are we talking about? Will it slip a meter before locking up? Will it slip until the climber hits the deck without locking at all like an ATC that someone lets go of due to rope burn? Will it not slow down the climber due to the "rope squeeze" of the device? Has this been tested? If it slips a bit then locks that is a good thing. If it slows down the climber's fall enough to potentially recover control of the rope that is a good thing. Edelrid does suggests an "appropriate" carabiner (of course they sell it). I agree with you that testing the extreme low end diameter rope is not that useful, because as you state, a lot of devices do not perform as intended at the extreme ends. If you specifically use that skinny diameter rope than you should buy the more appropriate MicroJul which is specially designed to use skinny ropes.

You don´t seem to be understanding something. The device IS locked up and the rope slips through if the braking force is exceeded. Whether the climber slows down depends on friction in the system and their weight related to what the device and the belayer can hold. Once the slips reaches excessive levels (as previously discribed) the belayer will be in danger of suffering rope burns. Under the system we and the UIAA work to this counts as an injury and the belay device is a failure. What happens to the climber is another story, they may just go another couple of metres down in free air and think "a nice soft catch) whereas the belayer has three months off work and skin grafts.

I bought the Smart Alpine as it was specifically for that diameter rope. I was given the MegaJul by a dissatisfied user but it is also for that size of rope, it says so on it.

Jim Titt wrote: Jeez, we aren´t building spacecraft here!

What can I say, a statistician is suspicious of all statistics, and doubly so when they are done in Excel.

The process you use has a lot of loss of (potentially important) information, which could be resolved with more complete statistical modelling. In the very best scenario, there is no bias, but you are losing all information on stochasticity (i.e. the spread of the drop test loads). In other scenarios, this information loss could lead to substantial bias, yielding inappropriate conclusions. Without more information, there is absolutely no way to assess these issues.

Again, I am not an engineer, so grain of salt and all. I am volunteering to delve deeper into these potential issues mostly out of curiosity if you're willing to share, but otherwise no big loss to me. Who knows, we might all learn something interesting about the function of these belay devices.

Jim Titt wrote: You don´t seem to be understanding something. The device IS locked up and the rope slips through if the braking force is exceeded. Whether the climber slows down depends on friction in the system and their weight related to what the device and the belayer can hold. Once the slips reaches excessive levels (as previously discribed) the belayer will be in danger of suffering rope burns. Under the system we and the UIAA work to this counts as an injury and the belay device is a failure. What happens to the climber is another story, they may just go another couple of metres down in free air and think "a nice soft catch) whereas the belayer has three months off work and skin grafts. I bought the Smart Alpine as it was specifically for that diameter rope. I was given the MegaJul by a dissatisfied user but it is also for that size of rope, it says so on it.

Pardon me I'm not an engineer but I'll try to understand. What I'm trying to imagine is if a climber takes a (rare) factor 2 fall and the rope gets out of control of the belayer wouldn't the rope run more slowly through a MegaJul with the rope being squeezed against the biner than through an ATC (not being squeezed)? Seems reasonable to me that it would.

Brian wrote: Pardon me I'm not an engineer but I'll try to understand. What I'm trying to imagine is if a climber takes a (rare) factor 2 fall and the rope gets out of control of the belayer wouldn't the rope run more slowly through a MegaJul with the rope being squeezed against the biner than through an ATC (not being squeezed)? Seems reasonable to me that it would.

His tests & graphs are not based on situations where the belayer isn't holding the brake. No one here thinks an ATC will catch a fall without a hand on the brake.

His tests are saying that when the brake strand is "locked off" (belayer is in control) on both devices, the MegaJul slips sooner than the ATC XP. Why? Who knows, but those teeth on the ATC XP aren't just for show.

Squeezing isn't the only way to generate friction. The classic ATC generates plenty just through the sharp bend of the brake strand around the lip of the tube. Or, hell, have you ever used a brake rack? Those can generate tremendous friction without a squeeze in sight.

Patrick Shyvers wrote: His tests & graphs are not based on situations where the belayer isn't holding the brake. No one here thinks an ATC will catch a fall without a hand on the brake. His tests are saying that when the brake strand is "locked off" (belayer is in control) on both devices, the MegaJul slips sooner than the ATC XP. Why? Who knows, but those teeth on the ATC XP aren't just for show. Squeezing isn't the only way to generate friction. The classic ATC generates plenty just through the sharp bend of the brake strand around the lip of the tube. Or, hell, have you ever used a brake rack? Those can generate tremendous friction without a squeeze in sight.

Yes, I understand this. I know we at times are talking about different things. In our previous discussion of a factor 2 fall (another extreme use like the 7.8mm rope), I proposed that in most cases most belayers will not hold a factor 2 fall regardless of the belay device (with the exception of a GriGri). I think that it would be beneficial to have an assisted locking device like the MegaJul because at least it will slow down the fall by pinching the rope and the belayer has a better chance of recovering the rope with less rope burn.

So if my proposition that almost all factor 2 falls will not be held and the assisted locking device does slow down the fall than it is better to have an assisted locking device in a factor 2 fall. I'd love to see a (video) of dropping 180 pounds on a MegaJul in a factor 2 with no one holding on.

I contend that in normal use, outside of extreme use like factor 2 falls and rope diameters at the extreme ends, that the MegaJul is a safer device than a ATC especially with inexperienced belayers. I've held some big whippers by some big climbers and it locked up automatically every time with no slippage. People will say just train your belayers better or get new belayers but that is not always an option.

This thread is a shit-sandwich.

I'm still not sure why you climb places where you are at significant risk of a FF2, with an inexperienced belayer you don't trust to catch you.

almost all factor 2 falls will not be held

You realize this is 100% speculation with no proof?

Jim Titt wrote: And what´s wrong with YOUR national federation or do you expect us Euro´s do and pay for everything? :-)

because those anal germans test everything comprehensively and publish the results

unlike the AAC/ACC which just leave it to those corporations to tell us how "safe" their gear is

in north america we rely on intraweb and youtube "experts"

;)

The AAC doesn't have anything like the funds for a testing program. Unlike the big European organizations, it is just a climbing club, and for years a gentleman's climbing club at that. I've suggested over and over again that they at least make a serious effort to translate and make available the work done by the far wealthier European alpine clubs, to absolutely no avail.

rgold wrote: Neither the MegaJul nor the Smart seem to provide much in the way of hands-off resistance; the only device that appears to do that is the Alpine Up.

If I happen to have a finger in the yellow loop, giving slack on a mega jul when the leader falls, the impact will still cause lock off.

Syd wrote: If I happen to have a finger in the yellow loop, giving slack on a mega jul when the leader falls, the impact will still cause lock off.

for the smart this is totally and utterly untrue ....

even a SLIGHT pressure on the handle upwards can cause rope slip when a person falls before a "lock up", especially on thinner slicker ropes

folks have been dropped around here because they were giving out slack and the person fell and they tried to brake by gripping the rope with the hand still around the handle

the proper technique with the smart is that the hand comes OFF the handle down onto the rope fully

same with folks belaying with the grigri ... the thumb comes OFF the cam and pulls down on the rope entirely ...

feeding out slack while defeating the mechanism and having your climber fall ... and not being TRAINED in the proper braking technique ... thats one common way folks get dropped with these "assisted locking" devices

;)

Syd wrote: If I happen to have a finger in the yellow loop, giving slack on a mega jul when the leader falls, the impact will still cause lock off.

This doesn't seem to have anything to do with my comment, which was about what happens when the load to the device exceeds what the belayer can hold.

Fans of the assisted locking devices, including myself, assumed that once the device locked it would stay locked no matter what the load. But Jim's results show that once the loads are high enough---and they don't have to be very high, nowhere near factor-2 loads---all devices are going to slip. The MegaJul slips at a lower load than the Smart or Alpine Up, but they all slip eventually, as does the ATC XP.

One guess for how this happens is that at high loads the rope stretches more and gets thinner, thereby reducing the pinching effect that makes the locking action work. Once the locking action is no longer effective, the bending angles of the assisted locking devices provide less friction than those of the ATC XP because of the proximity of the carabiner to the top of the grooves in the assisted lockers. This is just a guess though, but it does suggest that there could be a plausible mechanism to explain the "unintuitive" results of Jim's graphs.

Anyway, back to Syd's (and Brian's) comment, if the belayer loses control of the belay because of the rope running, then it doesn't seem to me that the frictional resistance supplied by the assisted locker could be greater than what it is capable of when almost no hand force is applied---because in this situation almost no hand force will be applied. For the MegaJul and the 7.8mm twins, maybe you'd get 40 kgf or about 88 lbf resistance. With the Smart, maybe 70 kgf or about 154 lbf resistance, nearly twice as much as the MegaJul. With the Alpine Up, maybe about 230 kgf or about 500 lbf resistance, double the Smart and seven times the MegaJul. With the ATC XP, if you lose control you get maybe 10 kgf or 22 lbf of resistance, worse than all the assisted lockers of course.

So the assisted lockers provide more resistance than an ATC XP if the belayer loses control. But that is balanced by the fact that the belayer is going to lose control sooner with the assisted lockers and in particular much sooner with the MegaJul; the Smart and Alpine Up are closer at 16 kgf hand strength, and the belayer hangs on the longest with the ATC XP. And the longer the belayer hangs on, the more fall energy is absorbed by the rope, and the less has to be absorbed by work done against friction sliding through the belay device. So even if the belayer doesn't drop the rope, you can expect more rope to run through the Smart and Alpine Up and a lot more to run through the MegaJul compared to the ATC XP. This seems to suggest, somewhat paradoxically, that it is even more important to wear gloves with assisted lockers than with ordinary tubes, a practice I've already instituted for myself.

There is at least one flaw in my assumption about the assisted lockers supplying their near-zero hand grip resistance. If the rope really does get thinner, they they won't supply even the lowest level depicted in the graphs. But perhaps this is an oversimplification anyway; the rope dynamics and so possible diameter changes at the belay carabiner as the belayer brakes and then loses control are probably more complicated than anything I've imagined here.

shoo wrote: What can I say, a statistician is suspicious of all statistics, and doubly so when they are done in Excel. The process you use has a lot of loss of (potentially important) information, which could be resolved with more complete statistical modelling. In the very best scenario, there is no bias, but you are losing all information on stochasticity (i.e. the spread of the drop test loads). In other scenarios, this information loss could lead to substantial bias, yielding inappropriate conclusions. Without more information, there is absolutely no way to assess these issues. Again, I am not an engineer, so grain of salt and all. I am volunteering to delve deeper into these potential issues mostly out of curiosity if you're willing to share, but otherwise no big loss to me. Who knows, we might all learn something interesting about the function of these belay devices.

But I already know the sort of spread we get from pull-testing and that if you average 3 tests you´ll get somewhere in the ballpark. The range is about 5% which is sufficient to give a trend when the variation between the devices is large. The only vaguely important information you don´t see is the gradual fall-off in braking power as the rope slides further through but that goes for all devices and is fairly well understood, since we can´t change it anyway it isn´t of much interest really except to a tribologist.

rgold wrote: Fans of the assisted locking devices, including myself, assumed that once the device locked it would stay locked no matter what the load. But Jim's results show that once the loads are high enough---and they don't have to be very high, nowhere near factor-2 loads---all devices are going to slip.

They all slip in the end even the GriGri, in fact this is designed in to protect the rope.

It´s a bit confusing because by "locked" two different things could be meant, either the device has locked or the rope is prevented from moving.

As you note the problem is up to a certain level of fall severity the benefits lie with the Smart etc and then after that the conventional devices have the advantage. Where this level is and whether one needs to worry about it is a matter of what you are protecting yourself from and what you feel is important.

bearbreeder wrote: feeding out slack while defeating the mechanism and having your climber fall ... and not being TRAINED in the proper braking technique ... thats one common way folks get dropped with these "assisted locking" devices ;)

Perhaps with some devices but for the mega jul, fig5a of the manual edelrid.de/out/documents/do… describes the technique I mentioned. The locking is not "defeated" when holding the device in this manner.

Jim Titt wrote: But I already know the sort of spread we get from pull-testing and that if you average 3 tests you´ll get somewhere in the ballpark. The range is about 5% which is sufficient to give a trend when the variation between the devices is large. The only vaguely important information you don´t see is the gradual fall-off in braking power as the rope slides further through but that goes for all devices and is fairly well understood, since we can´t change it anyway it isn´t of much interest really except to a tribologist.

You might know (or think) that, but how am I supposed to know that from what you've provided?

Also, I am still failing to understand why you are using the average values in your regression lines, rather than just using all the data. Best case scenario is there is little/no difference, worst case is you've biased your results. Why not just use all three data points you've collected? You are wasting information.

Partial list of "vaguely important" information that we don't see based on your results, which could easily be examined through better statistical methods:
1) Are the force curves between these devices statistically significant, and to what degree/confidence level?
2) How "random" are the forces? In other words, how much would you expect the catching forces to vary for each device?
3) Is there a difference in the "randomness" across devices? How big of a difference?
4) Is the "randomness" consistent* across the drop force curve? In other words, would we see bigger variation in drop test forces at higher/lower forces?

Note: I am intentionally using layman (i.e. non-statistical) terms here. Yes, I am extremely aware that words like "consistent" have a very different technical statistical meaning than I am using them above, but I am not talking among statisticians.

Syd wrote: Perhaps with some devices but for the mega jul, fig5a of the manual edelrid.de/out/documents/do… describes the technique I mentioned. The locking is not "defeated" when holding the device in this manner.

thats a VERY interesting claim ... so you are saying that levering it open you can fall on it and the rope wont run through???

if anyone in squamish has a mega jul and willing to take whippers ill catch you ... well tie a knot about 10 feet down as a backup of course

thats a VERY important thing to test before making that claim ... as thats one of the most common ways folks get dropped with assisted locking devices ... smart and grigris included

;)

bearbreeder wrote: thats a VERY interesting claim ... so you are saying that levering it open you can fall on it and the rope wont run through??? if anyone in squamish has a mega jul and willing to take whippers ill catch you ... well tie a knot about 10 feet down as a backup of course thats a VERY important thing to test before making that claim ... as thats one of the most common ways folks get dropped with assisted locking devices ... smart and grigris included ;)

I'll be in Squamish at the end of August and enjoy shenanigans. Hmmmmmmmmmm.

shoo wrote: I'll be in Squamish at the end of August and enjoy shenanigans. Hmmmmmmmmmm.

Well even whip on my purple link cam at the crux

If that doesnt kill us, i dont know what will

If da link cams dun blow apart and the megajuls dun drop us ... The heads of MPers will explode !!!

;)

bearbreeder wrote: Well even whip on my purple link cam at the crux If that doesnt kill us, i dont know what will If da link cams dun blow apart and the megajuls dun drop us ... The heads of MPers will explode !!! ;)

I'll mail you mine if you email me your address.