Discussion Topic |
|
This thread has been locked |
NutAgain!
Trad climber
South Pasadena, CA
|
|
|
Aug 24, 2015 - 02:01pm PT
|
Two days ago I was at the top of a pitch with a tied dyneema powerpoint between two bolts, and ATC in guide mode belaying up my follower. I did actually have a backup to that so the dyneema sling wouldn't be a single point of failure...
So what's the cliff notes version here?
Belaying at the top with ATC guide means no pulley effect doubling the force on the anchor. Is the force of a follower fall enough to jeopardize a tied dyneema sling?
Seems like it would be a bad idea to let the follower lead through and factor-2 before placing anything else- at least rig up the belay back off the harness for a more dynamic catch....
It would be good for the top rope tests to simulate real conditions with rope drag or inattentive belayer where there is a loop of slack out.
|
|
High Fructose Corn Spirit
Gym climber
|
|
|
Aug 24, 2015 - 03:41pm PT
|
Banquo, thanks for those details. It's all very interesting.
Werner, can I have it? :)
So say we moved away from (a) dynamic climbing rope and climber and belayer to (b) a ten foot static truckers rope and ten pound weightlifter's weight and dropped the weight from anchor with this load cell in place. Would this be too much for it in any regard? Apparently not in load as Werner's load cell load max is 20kn but how about speed / shock values? In above example OP's max load takes place on the order of half second. Max load in weight drop on trucker's rope is likely tenth second (100ms) or faster. Could it handle this speed? How about a 10lb drop on a steel wire where speed /shock might be 10ms at very high lbf value?
Will check into it more later.
Could it handle this speed?
I mean in regard to accurate measurement, graph data, on the order of milliseconds... 2ms, 20ms, 200ms, etc..
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Aug 24, 2015 - 03:51pm PT
|
Banquo, care to speculate about how R&I did a very similar test (fall factors are 0.11 and 0.14) and got nearly double the readings? I don't think offhand that much increase is explained by having the belay device on a bolt rather than a human and by having a 200 lb faller rather than a 180 lb faller.
|
|
20_kN
Sport climber
20 kN Land
|
|
|
Topic Author's Reply - Aug 24, 2015 - 04:04pm PT
|
//Good to think about when using a "guide-style" anchor while belaying a second up to your belay spot. That would be pretty similar to belaying off a bolt, no?
//
When belaying directly off the anchor you do not have a a 2:1 pulley like you do when top roping, which unless there is a bunch of slack in the system or the climber is near the anchor, means the loads should be even lower when belaying right off the anchor on a multipitch climb.
//Not that I'm much of a safety freak but wouldn't it be a good idea to back up that load cell in the anchor with another, loose, QD/sling?
Nice test btw...
Be interesting to see a "static" line test as well.//
IT was backed up with a 4' sling, I just removed the sling to take the photo. Anyway, load cells are designed for overhead lifting and the one I was using was stronger than the biners attached to it.
|
|
CCT
Trad climber
|
|
|
Aug 24, 2015 - 04:21pm PT
|
When belaying directly off the anchor you do not have a a 2:1 pulley like you do when top roping, which unless there is a bunch of slack in the system or the climber is near the anchor, means the loads should be even lower when belaying right off the anchor on a multipitch climb.
Thanks for the info, 20_kN! Really great discussion here. Since it's not a pulley, does that mean that the maximum force would be closer to 3.5kN (ie half of 7kN)?
Also, does this have any bearing on the common practice of clipping the top-most piece in the anchor as a first piece of pro as the leader takes off? My understanding is that people do this to protect the belayer. I'm thinking that getting jerked towards the anchor with 7kN could be damaging too, or even rip the piece out (small cam, marginal placement).
|
|
20_kN
Sport climber
20 kN Land
|
|
|
Topic Author's Reply - Aug 24, 2015 - 04:23pm PT
|
Banquo, thanks for those details. It's all very interesting.
Werner, can I have it? :)
So say we moved away from (a) dynamic climbing rope and climber and belayer to (b) a ten foot static truckers rope and ten pound weightlifter's weight and dropped the weight from anchor with this load cell in place. Would this be too much for it in any regard? Apparently not in load as Werner's load cell load max is 20kn but how about speed / shock values? In above example OP's max load takes place on the order of half second. Max load in weight drop on trucker's rope is likely tenth second (100ms) or faster. Could it handle this speed? How about a 10lb drop on a steel wire where speed /shock might be 10ms at very high lbf value?
Will check into it more later.
I have conducted static drop tests in a number of different types of material: dyneema, nylon, via ferrata lanyards, steel cable, and other items. In addition, this thread I created talks quite a bit about sample rates and load cells as well:
http://www.mountainproject.com/v/quick-study-involving-falls-on-static-materials/108262937
In short, it depends on your load cell and signal conditioner. Even cheap load cells can effectively respond down to about 0.5mS. Below that you'll need a better one. However, the limiting factor is usually the conditioner. In my case, 520Hz. is as fast as I can go. The Rock Exotica Enforcer pictured above can do 1,500Hz. but it can only log at 500Hz. I found 520Hz. to be more than fast enough to measure anything climbing related except drop tests on very static material, but even then I can still get fairly close. Further, it's easy to determine if the sample rate was fast enough or not once I look at the data.
Here is a test in which I compared 10Hz. vs 520Hz. on a dynamic load on the same event. You can see the results are not that different, although there is a bit of a difference.
Here is an example in which the sample rate was not quite fast enough:
|
|
High Fructose Corn Spirit
Gym climber
|
|
|
Aug 24, 2015 - 05:16pm PT
|
Thanks 20kn, I'll be checking it out.
a "gibbon butt bounce," hmmm :)
|
|
Banquo
climber
Amerricka
|
|
|
Aug 24, 2015 - 05:50pm PT
|
Banquo, care to speculate about how R&I did a very similar test (fall factors are 0.11 and 0.14) and got nearly double the readings?
Different ropes will have different stiffness, even lead ropes. A fat rope might be stiffer than a skinny. Also, isn't 20_kN's test a fall factor 1 while the R&I sounds like it is close to 2? Lifting the belayer will certainly reduce the force. However, the lower load readings could be in part due to filtering in the instrument. If it was designed for static weighing, it might be heavily filtered. Many scales have filtering because if the truck or whatever you are weighing is bouncing or something you want to filter that out. I have a cheap platform scale in the garage that is very accurate weighing boxes but cannot weigh me no matter how still I try to stand - it probably doesn't have filtering.
Werner's tool is very cool with bluetooth and all but I tried to look up the specs and couldn't find anything about frequency response. It seems to have two sample rates which is different. Since it is purpose specific, it probably has high enough frequency response but I'd want to know before I bought one. I guess I wouldn't buy one anyway since I have a box of load cells in the garage already.
I have some signal conditioners in the garage (CALEX Model 162MK Bridgesensor) that you can vary the cutoff filter from 10Hz to 1000Hz by changing dip switch settings. I always want specifics on filtering if something dynamic is being measured. I usually filter less, sample faster and then filter the resulting data by smoothing or FFT. I like to sample at least 4x the maximum frequency I am interested in. Low sample rates without filtering results in aliasing and your data is junk. Best solution is to filter to some reasonable frequency and sample 4x that frequency. So, for a climbing rope drop test I might analog filter to 100Hz and sample at 400 Hz.
If the signal conditioner gives flat response to 520Hz, it is plenty fast enough. I'd still be interested in manufacturer and model number for the equipment. Just because it samples at 500Hz doesn't mean there isn't filtering to something lower. Calibration is probably right because the static load at the end of the test appears to be the combined weight of the climber and belayer.
|
|
20_kN
Sport climber
20 kN Land
|
|
|
Topic Author's Reply - Aug 24, 2015 - 06:08pm PT
|
However, the lower load readings could be in part due to filtering in the instrument.
Werner's tool is very cool with bluetooth and all but I tried to look up the specs and couldn't find anything about frequency response. It seems to have two sample rates which is different. Since it is purpose specific, it probably has high enough frequency response but I'd want to know before I bought one. I guess I wouldn't buy one anyway since I have a box of load cells in the garage already.
I have some signal conditioners in the garage (CALEX Model 162MK Bridgesensor) that you can vary the cutoff filter from 10Hz to 1000Hz by changing dip switch settings. I always want specifics on filtering if something dynamic is being measured. I usually filter less, sample faster and then filter the resulting data by smoothing or FFT. I like to sample at least 4x the maximum frequency I am interested in. Low sample rates without filtering results in aliasing and your data is junk. Best solution is to filter to some reasonable frequency and sample 4x that frequency. So, for a climbing rope drop test I might analog filter to 100Hz and sample at 400 Hz.
If the signal conditioner gives flat response to 520Hz, it is plenty fast enough. I'd still be interested in manufacturer and model number for the equipment. Just because it samples at 500Hz doesn't mean there isn't filtering to something lower. Calibration is probably right because the static load at the end of the test appears to be the combined weight of the climber and belayer.
There is no filtering enabled. I always keep filtering disabled. I am not using some crane scale or truck scale of sorts. I am using a conditioner designed for R&D analysis.
The conditioner is VERY user programmable. You can write your own filter, choosing cut-off frequencies, a range of freqs, or specific freqs, and you can set the steps per division filtering level as well, or in my case completely disable it. You can also set dynamic filtering where you write a custom filter to suit your needs, and have the device momentarily disable filtering if the load spiked above a value of your choosing over a duration of your choosing (e.g. to measure a dynamic load when measuring something that is normally static). I always keep filtering disabled because the resolution of the conditioner is good enough that even at 500 Hz. the readings are still quite stable thus rendering the filtering quite unnecessary unless I am going to try to weigh a 100 lb object with a 50,000 lb load cell or something silly like that.
The unit also allows for temperature calibration linearization, sensitivity linearization across 8 steps/ readings, as well as a number of other linearization features if you want to use them.
If you want a conditioner with a TON of features and a ton of user customization for a reasonable price, it’s a great option. The drawback is because it’s so customizable, it’s not that user friendly if you do not understand how signal conditioners work, and you need a very basic understanding of how command prompts work in Hyper Terminal or Putty if you want to unlock the most advanced features. But for those who want all the advanced features and are willing to put in the work to learn, this conditioner is a great buy for the money.
This is the conditioner if you want to look it over: http://www.mantracourt.com/products/signal-converters/usb-load-cell-converter
As far as the sample rate of the Enforcer, it runs at 1500 Hz. if put on peak hold or 500 Hz when logging. So if you want to log you're limited to 500Hz. The Enforcer is a good option for something simple and compact that you can take in the field, but if you are "really" into test and analysis, my conditioner offers far more features at the cost of being less compact and requiring a laptop to run.
|
|
Banquo
climber
Amerricka
|
|
|
Aug 24, 2015 - 06:59pm PT
|
That's a cool signal conditioner. Things sure have changed in instrumentation during the past 35 years. I'm an Analog type guy so I don't know much about recursive filters and should probably study up. When I google recursive filter and look at images, I see some interesting data plots. But, if you have left the filter inactive, it probably isn't a factor.
Anyway, I'm not dissing what you've done since I doubt there is anything happening above 2 Hz which is pretty much static in the instrumentation world. Two sacks of meat hanging on the ends of an elastic rope are not going to be a high frequency system. Your results seem entirely reasonable to me.
Out of idle curiosity I'd be interested in what data your system produces for a high frequency impact. It would be fun to see what you get with a hammer and funkness device or something like that. Even more fun to see what the recursive filter does to it. I know I can generate 10 kips with a hammer on a solid surface. No need to do this, it's just what I'd do if I was goofing off.
Like I think you are, I am suspicious of the R&I data and we know nothing about their equipment. I don't read or get R&I so don't really know what the article says. The last issue on my shelf after I dropped my subscription is November 2002.
|
|
Paul Martzen
Trad climber
Fresno
|
|
|
Aug 24, 2015 - 07:09pm PT
|
Nice report.
Would be interesting to see the results of a direct replication. Attaching the belay device directly to a static bolt. Seems odd that the forces would jump from 2 KN all the way to 10 KN.
|
|
20_kN
Sport climber
20 kN Land
|
|
|
Topic Author's Reply - Aug 24, 2015 - 07:10pm PT
|
That's a cool signal conditioner. Things sure have changed in instrumentation during the past 35 years. I'm an Analog type guy so I don't know much about recursive filters and should probably study up. When I google recursive filter and look at images, I see some interesting data plots. But, if you have left the filter inactive, it probably isn't a factor.
Anyway, I'm not dissing what you've done since I doubt there is anything happening above 2 Hz which is pretty much static in the instrumentation world. Two sacks of meat hanging on the ends of an elastic rope are not going to be a high frequency system. Your results seem entirely reasonable to me.
Out of idle curiosity I'd be interested in what data your system produces for a high frequency impact. It would be fun to see what you get with a hammer and funkness device or something like that. Even more fun to see what the recursive filter does to it. I know I can generate 10 kips with a hammer on a solid surface. No need to do this, it's just what I'd do if I was goofing off.
Like I think you are, I am suspicious of the R&I data and we know nothing about their equipment. I don't read or get R&I so don't really know what the article says. The last issue on my shelf after I dropped my subscription is November 2002.
I included a link to a copy of the article in my study under the reference section. The conditioner is not fast enough to measure the peak force of a funkiness device. You would probably want a minimum of 3,000Hz. for something like that, although 10,000 would be better.
|
|
mikllaw-Oz
Sport climber
tempe Australia
|
|
|
Aug 24, 2015 - 09:04pm PT
|
Backing up (but not guaranteeing) the accuracy of this is the fact that the graph showing the 766.3 lbf / 3.41 kN goes to 340 lbs at the end, which is the combined weight of the 160 and 180 lb climbers.
|
|
rgold
Trad climber
Poughkeepsie, NY
|
|
|
Aug 24, 2015 - 10:18pm PT
|
Different ropes will have different stiffness, even lead ropes. A fat rope might be stiffer than a skinny.
Yes, but 20 kN was using one of the stiffest dynamic ropes around.
Also, isn't 20_kN's test a fall factor 1 while the R&I sounds like it is close to 2?
OMG no---they were using real people! According to 20 kN's numbers, his falls had fall factors about 0.13, whereas the R&I fall factor was about 0.10. So 20kN's climbers were a bit lighter than the R&I guy, but not enough to halve the maximum tension!
Lifting the belayer will certainly reduce the force. However, the lower load readings could be in part due to filtering in the instrument.
Yes for lifting. But 20 kN had no filtering implemented for the tests.
So it remains mysterious: a slightly heavier faller with a slightly smaller fall factor and a more static belay nearly doubles the maximum impact force?
|
|
i'm gumby dammit
Sport climber
da ow
|
|
|
Aug 25, 2015 - 01:12am PT
|
Not necessarily. A lot of people build anchors out of nylon or dyneema slings. I don't know how well dyneema can absorb impulse forces, but I thought it wasn't very good. You are correct that some do. They shouldn't for exactly the reason you are worried about. I use to to lead solo by clipping longish dyneema slings to bolts or gear and then unclip the low one as i moved up until the error of my ways was pointed out to me. I still had the dynamic absorption of my harness but that may have not been enough. I rarely do it these days (got a girlfriend) but when I do I use 10' segments of dynamic rope double fishermaned into 4' loops with loosely tied figure 8 bights on each end so I get a softer catch from the dynamic rope, harness, and the tightening of the bights. Now I'm slightly less likely to experience the 'yer gonna die' syndrome if i fall.
|
|
CCT
Trad climber
|
|
|
Aug 25, 2015 - 09:44am PT
|
I wonder about a guide belay onto a cam anchor via a dyneema sling (knotted or not).
I also wonder about clipping an anchor piece as the first piece of gear using a dyneema sling in case of an early fall.
These two scenarios are very common, and you don't really hear of accidents related to them. But from the numbers alone, it feels like they can't be very safe.
Anyone understand the physics better than me and can set my worries to rest?
|
|
20_kN
Sport climber
20 kN Land
|
|
|
Topic Author's Reply - Aug 25, 2015 - 05:09pm PT
|
I wonder about a guide belay onto a cam anchor via a dyneema sling (knotted or not).
I also wonder about clipping an anchor piece as the first piece of gear using a dyneema sling in case of an early fall.
When a dynamic rope is in the system, there is little consequence between using a dyneema or nylon sling. The sling is not the primary energy absorbing element, the rope is, and it will have little to no measurable effect in the field. In the case of the second falling on TR with a dyneema-made anchor vs nylon-made, there would be absolutely zero measurable difference. With regard to taking a fall directly into the belay on a dyneema sling vs a nylon, there would also be little difference. BD tested this quite awhile ago and found about a 2-5% lower impact force with a nylon sling vs dyneema.
However, they used a rigid weight, static belay and a drop tower. In the field, with a flexible mass and movable belayer, the differences would drop even more.
|
|
i'm gumby dammit
Sport climber
da ow
|
|
|
Aug 27, 2015 - 01:27am PT
|
I wonder about a guide belay onto a cam anchor via a dyneema sling (knotted or not). There is a huge difference between knotted and knot with Dyneema. Watch this video.
http://dmmclimbing.com/knowledge/slings-at-anchors/
As you can see here the dyneema should be ok at 7kN, but I wouldn't be overly comfortable with the margin since the overhand knotted dyneema failed at 10.4 kN in the example. Although a 10kN fall is likely gonna have you pretty messed up even if you don't deck.
|
|
|
SuperTopo on the Web
|
Let us know!