Login with Facebook
 ADVANCED
Webbing & threaded highline analysis w/ strain gauge indicator
View Latest Posts in This Forum or All Forums
   Page 1 of 1.  
Follow replies to this topic? Notify me at the top of web site.
1

Email me.
 
 
By 20 kN
Administrator
From Hawaii
Mar 30, 2012
I want to start off by saying that I am not a highline expert, I have set up a few highlines, but there are others more qualified to comment on highline rigging. This thread is non-exhaustive and was written to express the data generated from my experiment.

In the world of highlining, slackliners normally use two separate systems to protect them should they fall. The first system is a tensioned mainline that they walk across, basically just a slackine. The second is a backup line should the mainline fail. One purpose of this thread is to focus on one specific type of back-up, threaded lines. According to slacklineexpress.com,

There are differing schools of thought on the fine points of all rigging, and how to make a backup main line is no different. Some people like to thread 11/16" webbing through the 1" webbing on the main line to create a Threaded Mainline and that's it, calling the 11/16" the backup system. Personally, I disagree with this practice. While a threaded main line will get tighter easier for some tensioning systems, it does not increase the strength of your system. From our testing, threaded lines are less than ideal because your backup is pre-tensioned and strength wise the core breaks independently and then the 1" breaks at standard breaking load. Since it doesn't yield any higher tensile strength, nor does it share the load if you ever managed to break your 1".

The purpose of this study is to determine three things. What is the strongest way to terminate a piece of webbing? Is slacklineexpress.com right? Does threading a mainline have no effect on the ultimate strength of the line? And lastly, how much force does a highline fall actually produce?

Chapter one – Termination of the webbing

I am going to test five ways to terminate webbing, a clove hitch, an overhand knot, a frost knot, a line locker, and a sewn loop. The frost knot is a variation of the overhand knot that introduces a second piece of webbing, which is supposed to reduce the strength loss from tying a knot.

First, we will start with some basic theory. The two most popular ways to terminate a highline or slackline is likely a line locker (or webbing locker) and an overhand knot. So I picked up some cheap pink webbing from the hardware store to test the strength differences between a line locker and an overhand knot.




Overhand knot: 4.55kN
Line locker: 5.54kN


I only tested one sample of each, but the initial results look promising, the line locker yielded a 17.7% higher strength than the overhand knot. Moving on, it is time to test the strength of the different types of termination options. Here are the specifications of the webbing I used:

- ABC 1” nylon tubular webbing
- Manufacturer rated strength: 17.78kN


I only tested two samples for each termination point. However, the test results were very repeatable. Below are the results of the test. The percentage value expressed in the parentheses is the average offset from the manufacturer rated strength of 17.78kN. In all tests, the material failed at the termination point (the knot, bar tack end, or line locker).




Clove hitch: 11.33kN & 11.18kN (-36.7%)
Overhand knot: 12.44kN & 12.65kN (-29.4%)
Frost knot: 13.46kN & 14.00kN (-22.8%)
Line Locker: 17.50kN & 17.46kN (-1.7%)
Sewn Loop: 19.46kN & 19.61kN (+8.6%)


As you can see, the only two termination methods that preserved the webbing’s rated strength were the line locker and sewn loop, with the sewn loop being the strongest of them all.






Chapter two – Analysis of threaded lines

In this section I am going to visit the theory behind whether threading a line has no effect on the lines ultimate strength, as stated by slacklineexpress.com, or whether it does increase the ultimate strength of the webbing. slacklineexpress.com said “From our testing, threaded lines are less than ideal because your backup is pre-tensioned and strength wise the core breaks independently and then the 1" breaks at standard breaking load.” Before jumping head first, I would like to determine whether the elongation of a narrower piece of webbing is different from the elongation of a wider piece. I want to start with a 3/4" mainline threaded with 1/2” webbing rather than going straight to a 1” threaded mainline.

My theory is that if the elongation of the different pieces of webbing is the same, they should share the load equally up until the breaking strength of one piece of webbing is reached, which would result in its failure, shifting 100% of the load to the remaining piece, which would then result in the immediate failure of the remaining piece. In other words, my theory is that if both the 3/4” mainline and the 1/2” thread stretched equally, the strength of the line should be twice that of the weaker piece (normally the 3/4” piece). If the elongation is different, one piece will see a greater load than the other, which greatly complicates things. I believe that if the elongation of the 1/2” is lower than that of the 3/4”, the 1/2” will fail prematurely shifting 100% of the load to the 3/4”. Likewise, if the 3/4” stretches less, it will fail first shifting all of the load to the remaining 1/2” piece. But enough guessing, let us just test it.

First, we need to determine how strong the 1/2” and 3/4” samples I am using are. I tested a few samples and came up with the average below. I tested the samples with line locker terminations.



1/2” tubular nylon: 4.95kN
3/4” tubular nylon: 13.8kN


Next we need to determine what the elongation of the material is. This process is a bit tricky because I cannot pull test a long sample. I am limited to a webbing sample length of about three feet so the accuracy of this test will be limited to some extent. I conducted two different tests: the average elongation of the samples that encompasses the stretch from 60 lbs. to material failure (ultimate elongation) and the average elongation of the samples from 60 lbs. to 300 lbs.



1/2” webbing 300 lb. elongation: 17%
3/4" webbing 300 lb. elongation: 9%
1/2” webbing ultimate elongation: 51%
3/4" webbing ultimate elongation: 47%


So as you can see, the two pieces of webbing do have different elongation parameters. So far this theory is looking in favor what slacklineexpress.com stated. Although the 1/2” webbing does elongate a bit more than the 3/4” does, the difference between the strength of the 1/2” and 3/4" is greater than the difference between the elongation of the 1/2" and 3/4" samples, meaning it looks like achieving a much higher ultimate strength with a threaded line is looking pretty unlikely.

So finally we are there, time to find out for sure if threading a line increases its strength. I tested the threaded webbing the same way I tested the 1/2" and 3/4” samples, with line locker terminations. The percentage values in the parentheses are the offset values from the 3/4" non-threaded sample.


1/2": 4.95kN (-64.1%)
3/4": 13.8kN (+/-0%)
3/4" threaded: 14.78kN (+7.1%)


So a 7% increase - this study is looking in favor of SlacklineExpress.com’s theory so far, but we still have a bit more testing to do.



In this next section we will examine 1” webbing threaded with 3/4” webbing. Both the 1” and 3/4” samples are from the same sample pool I have been using in the previous tests. However, before we start testing the breaking strength of threaded 1” we need to examine the elongation of the materials being used in the mainline. Please note that the elongation data presented earlier in this study is only relevant to that section of the study because the method I used to calculate the elongation data is far different from the method I am using below.

Below are some photos showing how I captured the elongation data. Basically, I just pulled on the webbing with a ratchet and recorded the elongation of the webbing every 120 lbs. The graphs below represent the elongation of the webbing alone. The graphs do not include the elongation of any outside sources, such as the attaching carabiners or line lockers.





Figure 7 shows the elongation data of a piece of 3/4” webbing, a piece of 1” webbing, and 1” webbing threaded with 3/4”. In figure 7, I tested the samples independently; they were not tested as a single threaded line and analyzed separately. Figure 10 differs from figure 7 in that I tested both the 3/4” and 1” strands together as a threaded line. Basically, figure 7 represents three separate samples and three separate tests in one graph. Figure 10 represents one sample and one test; a test of the threaded line. In figure 10 I measured the length of the threaded line as a single unit, but I measured the load on each strand individually with two load cells. Because I separated the 1” and 3/4” pieces and I terminated them each to a different load cell, it was critical I ensured that each strand was the same length. If they were different lengths, they would elongate differently in relation to each other in a method not analogous with its modulus of elasticity. So rather than trying to measure the length of each strand, I simply placed 80 lbs. of force on the sum of the two stands and adjusted each strand until both load cells read the same weight, that way I knew both strands were precisely the same length.




Here are some photos showing how I captured the load on each strand:




As you can see from the graphs above, the threaded mainline elongated about 4% less at 1400 lbs. than a 1” non-threaded mainline. However, there was only about a 2.5% difference between the 1” and 3/4” at 1400 lbs.

One very interesting feature of Microsoft Excel is its ability to predict future readings in a graph. Because I cannot generate 4,000+ lbs. of force with my ratchet, I am limited to elongation readings within the scope of the ratchet. This is where Excel comes into play. Because we are concerned about the ultimate strength of threaded line, we are also concerned with the ultimate elongation of the webbing. Figure 9 shows a 2nd order polynomial theorized treadline of what the elongation of the webbing would be at higher forces.



Before continuing, we need to take a quick detour. Because I am presenting data that has been theorized by Excel, I want to verify the accuracy of the theoretical treadline created by Excel. Figures 8a and 8b show the accuracy of the treadline in reference to the actual elongation of the 1” webbing. Figure 8a uses a 2nd order polynomial equation to calculate the treadline and figure 8b uses a 3rd order polynomial equation. I separated the three colored lines into separate readings so Excel could not use them all to calculate the treadline. I only used the dataset contained within the blue line to calculate the treadline. As you can see, both treadlines are reasonable accurate within a 250% forecast of the original dataset. Because the 2nd order polynomial seems to be marginally more accurate, I used that order to calculate the treadline in figure 9 instead of the 3rd order.

Please note that these treadlines will only be accurate if the mathematical function of the elongation dataset continues in its current pattern as the material is further strained. If the elongation pattern changes significantly, the treadline will not be accurate. In other words, it is just a theorized treadline generated by a computer.



Next we will examine the most important aspect of this set of tests, the difference in force between the two stands of our threaded 1” sample. Figure 11 shows the difference in force between the 1” and 3/4” strands when they are used together as a threaded line. This graph is very important because it tells us how strong the threaded strand “can” be. I say “can” because as we will find later in this chapter, the actual strength of a threaded line is dependent on more than just the strength of the webbing.



Figure 12 shows the theorized difference between the strands at higher forces. I used a linear treadline because the very inconsistent nature of the graph throws any other treadline options off the grid. So the liner treadline is probably of limited accuracy, but it does not matter, what we can take from this portion of the experiment is that there is not a huge difference between the forces experienced by the two strands of a threaded line.



Because we now know the difference in force between the strands, we can predict the breaking strength of the threaded line. We know a single strand of 1” fails at just shy of 4,000 lbs. and a single piece of 3/4” fails at around 3,300 lbs. Excel predicts the difference in force between the two strands at 4,500 lbs. is only about 265 lbs., with the greater load being carried by the less dynamic strand (the 1”). Consequently, we can theorize that a 1” threaded line will fail at around 6,600 lbs., which is double the strength of a piece of 3/4”.

Lastly I tested the breaking strength of 1” webbing threaded with 3/4”. The percentage values in the parentheses are the offset values from the 1" non-threaded sample.



1/2": 4.95kN
3/4": 13.8kN
3/4" threaded: 14.78kN
1”: 17.48kN (+/- 0%)
1” threaded: 21.09 kN (+20.7%)





So what happened? We theorized the 1” threaded line would fail at 6,600 lbs., but instead it failed at only 4,750. The answer to this question is under further research and I will update it when an answer is found. So far, experts in the field tell me the increased width of a threaded line is making the line locker termination significantly less efficient and the major loss in efficiency is the cause of the premature failure. That theory makes sense because there are only two possible reasons why the material would fail at a much lower force than what we predicted. First, the difference between the forces carried by each strand widens significantly as the load is further increased (unlikely). The second possibility is that the line locker is indeed less efficient when a threaded strand is passed through it.

CONCLUSION

So it appears that threading 3/4" webbing does not increase the strength by very much, only a few hundred pounds (under further examination). However threading 1” did yield nearly a 21% increase in strength. However, although the strength of the 1” was increased by 21% through threading, that is still only an additional 3.61 kN – nothing to write home about (under further examination). We learned that the load was almost equally shared between the 1” mainline and 3/4” thread. So what very likely happened in the case of the 3/4” threaded mainline was that the inner 1/2” core failed and then the 3/4” piece failed. The 1/2” core only held 1,112 lbs. in my tests, so a doubling of that yields 2,224 lbs. The likely reason the 3/4” threaded line held 3,323 lbs. instead of only 2,224 lbs. is because the difference in elongation between the 1/2” strand and the 3/4” strand was much greater than the difference between the 3/4” strand and 1” strand. Accordingly, the lower elongation strand (3/4” strand) likely held a greater amount of force then the 1/2” strand.

This study is under further research and will be updated within a few weeks.

SECTION 3 – Highline fall analysis with a strain gauge indicator

In the next chapter I am going to determine how much force a leased highline fall actually produces and how strong your webbing and anchor actually needs to be.

FLAG
By Eric Krantz
From Black Hills
Mar 30, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
Would you post a better photo of the setup of the strain-gage end?

FLAG
By 20 kN
Administrator
From Hawaii
Mar 30, 2012
Eric Krantz wrote:
Would you post a better photo of the setup of the strain-gage end?




The ropes dont really pull on the load cell. They are there so the load cell does not slam into the frame when the sample breaks. But under load, the ropes are loose enough not to affect the reading.

FLAG
By Eric Krantz
From Black Hills
Mar 30, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
20 kN wrote:
The ropes dont really pull on the load cell. They are there so the load cell does not slam into the frame when the sample breaks. But under load, the ropes are loose enough not to affect the reading.


OK, that is what I was wondering!

Great fun you have. What's it like when the 1" samples break?

Curious why the strength of the threaded line isn't the sum of the two. If they have roughly the same ultimate elongation, just before breaking they should both be stretched to nearly their respective rated loads, right? So, they both break at 50% elongation, the load necessary to take both at the same time to 48% elongation should be just less than the sum of their strengths. What am I missing? Is there some sort of pinching action happening at the line lockers that is snapping it?

I'd like to see a test of the 1" and the 3/4" together, but not threaded.

Cool experiment. Nice hulas.

FLAG
By 20 kN
Administrator
From Hawaii
Mar 30, 2012
Eric Krantz wrote:
OK, that is what I was wondering! Great fun you have. What's it like when the 1" samples break? Curious why the strength of the threaded line isn't the sum of the two. If they have roughly the same ultimate elongation, just before breaking they should both be stretched to nearly their respective rated loads, right? So, they both break at 50% elongation, the load necessary to take both at the same time to 48% elongation should be just less than the sum of their strengths. What am I missing? Is there some sort of pinching action happening at the line lockers that is snapping it? I'd like to see a test of the 1" and the 3/4" together, but not threaded. Cool experiment. Nice hulas.


Well my elongation specs are probably off a bit simply because its really hard to get accurate data from a sample that's only a few feet long. Anyway regarding the ultimate elongation, they MAY (remember, its hard to measure) have the same elongation separately, but when combined the elongation changes. Its like using twins ropes versus a single rope. If you look at the Beal Joker, it has a dynamic elongation of 37% as a single and 29% as a set of twins. So why doesent adding a second strand half the elongation? Because it changes the elasticity characteristics of the system when you add in more material. The elasticity curve of rope and webbing is non-linear. Take into account the non-linear elongation characteristic of the webbing, the already disproportional ratiometric elongation values between the webbing samples, and the load sharing between the two strands and well, you have a really complicated model of elasticity. It is because of the differences in elongation that you also see a difference in the strength increases between 3/4" threaded and 1" threaded. Ultimately, I dont know the precise answer to your question, I really wish I could figure it out, I just know the strength limitations are a function of the differences in elongation between the pieces of webbing and not some external force like the line lockers pinching the webbing.

It is also really hard to test the elongation because the material stretches slowly while under load. If you look at the line graph you will notice a saw tooth-like action of increasing force. That is because every time I pumped the cylinder up, the material would stretch some and reduce the tension on the webbing. This makes it difficult to figure out when to actually measure the length of the webbing while its under load. Because as soon as I stop moving the hydraulic piston, the weight on the webbing drops quickly as it elongates slowly. Eventually it tapers off and stops dropping, but it takes quite awhile and its ultra hard to figure out how much tension I need to place on the line to get it to drop down to the stable weight value I want to measure it at. So the question then comes in, does one measure the length of the webbing once they reach a stable reading, after five min or so, or as soon as the tension reaches the value you want to measure the webbing at? When I measured the elongation, I tried to let it stable out a bit, I let it sit for 30 seconds or so and waited until the reading dropped to 300 lbs.

I could test the 3/4" piece outside of the 1", but I doubt its going to make any difference. I will go grab some more 3/4" from the store and test it later. Anyway, the failure of the samples sounds something like swinging a 4' crow bar against a sheet of steel - it is pretty loud and annoying.

A real interesting experiment would be to test the two pieces simultaneously as a threaded line, but with a load cell on each strand of webbing so I could see precisely which strand sees what load. I do have two load cells that I could do this with, but the trick would be getting both of the strands the same length so I do not introduce any extra elongation in one strand from having a longer sample length. I am open to ideas. :)

FLAG
By Eric Krantz
From Black Hills
Apr 1, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
One very key observation here is that both the 3/4" and the 1" broke at the same time when threaded. It's key because [(1) it means they also broke at the same elongation AND (2) tested separately they break at slightly different elongations], OR (3) your machine has a very high machine softness, and they actually broke a few milliseconds apart and you don't know it.

I mentioned "pinching" above and I'd like to expound on that, but it will require a few paragraphs. We need to analyze the forces a bit more to get a better grasp of what's going on:

First, an assumption: Assuming each line breaks at exactly 50% elongation (which is true within 4-5%)... in that case, the situation could be drawn like this just before reaching 50%:

What's the load?
What's the load?


Answer: F(3/4") + F(1")

The forces in each respective line, at the same elongation, can not be any different if they are tied to the same member, than if they are tested separately. Together, the force experienced by the anchor is the sum of the two individual forces. Lower ultimate strengths can be caused by two things:

1. The assumption stated is wrong. Maybe the 3/4" breaks on average at 47% and every 1" at 51% (instead of both at 50%). When pulled together, the 3/4" would break when both were elongated at 47%. In an incompressible system (neglecting machine softness), the 1" would still be at 47% elongation after the 3/4" broke (because the machine doesn't move when the force is reduced). This is not the case in the threaded scenario because BOTH are breaking at the SAME time (same elongation), and THAT's the kicker here. And THAT is why the diagram I drew is way different than the threaded scenario. The reason they both break at the same elongation (an elongation which we don't know), is reason number 2:

2. Where the line wraps around locker, the outside loop of the 1" webbing is putting very high pressure on the 3"4" webbing, all of which are squishing the inside loop of the 1" webbing with a hell of lot of pressure into the steel, but only 3/4" wide of it. I should have drawn this out, too. Think of a sandwich (a ten foot sub is your threaded line) with 4 layers... the bread is the 1" and the meat & cheese are the 3/4". The top layer of bread and the meat layer and the cheese layer are each pushing down really hard on the bottom layer of bread, all of which are wrapped around a round steel bar. The bottom slice of bread gets cut. Right? All this force is squashing down on 3/4" strip of the 1" inside loop, and cuts it at a force less than the sum of the two separate strands breaking strength. Once IT fails, there goes the rest. This was my "pinching" comment.

For perspective, the pressure on the section of inside 1" loop at the line locker can be roughly computed as: [3323 + (0.5)(3920)] / (3/4" x 1/4") = 28,176 psi. A pressure like this may slice the inside section in half, even if the knife is as dull as a 1/4" diameter steel bar. And when it fails, the whole this fails... cuz the 1" fails first!

Eric

PS Final thoughts... The percent elongation is NOT linear with force applied (When at 44% elongation, stretching an extra 1% may require an extra 100 lbs force, whereas an extra 1% at 5% elongation may require only 32 lbs). We would like to know the relationship between percent elongation and force. You have plots of force vs time.... time is really not that relevant. Please plot force with percent elongation. If you did it with every 7% or so, you would have about 7 points before failure. Plotting them should make a nice logarithmic plot. (I'm guessing again).

Also, what was the percent elongation at failure of the threaded strand? Did I miss that somewhere?

FLAG
By Ryan N
From Bellingham, WA
Apr 1, 2012
RJN
I'm no genius but how would the data differ if you were testing a typical length used for a highline say 50-100ft of webbing? Would the data change? Seems like in the same way the amount of payout of rope in a leader fall would generate different forces the same would apply in this senario. I'd love to hear your input. I've set and walked several highlines and it's my expierence that shorter lines behave differently from longer lines.

FLAG
By Gunkiemike
Apr 1, 2012
Did you sew the end loops yourself? Can you provide any evidence to support my hypothesis that the strength of a sewn loop is quite linearly proportional to the number of stitches?

FLAG
 
By Eric Krantz
From Black Hills
Apr 1, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
Ryan N wrote:
I'm no genius but how would the data differ if you were testing a typical length used for a highline say 50-100ft of webbing? Would the data change? Seems like in the same way the amount of payout of rope in a leader fall would generate different forces the same would apply in this senario. I'd love to hear your input. I've set and walked several highlines and it's my expierence that shorter lines behave differently from longer lines.


Hi Ryan.

The amount of tension in a line necessary to maintain the same angle at the anchor stays the same regardless of length of line (when standing at the center of the line). In other words, if you are 100 lbs, standing in the center of a 20 foot line, the tension in the line will be about 502 lbs if you stretch it 1 foot down (giving an angle of about 5.7 degrees). IF the line is 100 feet, and you stand in teh middle and stretch it down 5 feet, you also have it tensioned at 502 lbs and creating an angle of about 5.7 degrees. But, you have dropped 5 feet instead of 1. (Also same percent elongation, which is a given because that is only dependent on tension).

Now, if you want to maintain the same vertical displacement, the tension necessary is linearly proportional to the length of line. In the first case, 20 feet of line, 1 foot of vertical displacement from a 100 lb person, gives 502 lbs tension. Causing only one foot of drop in a 100 foot line would require about 2500 lbs tension. So, of course the elongation in this case is much higher, the line is much tighter, and the forces when standing on it and falling on it are much greater. (Incidentally, the wave speed in the line is also higher).

So, those world-record-setting slacklines must be very tight to maintain enough elevation so you don't hit the ground when you get toward the middle! Or they have to start really damn high.

FLAG
By Eric Krantz
From Black Hills
Apr 1, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
To calculate tension in the line when you are standing the middle of it (which will cause maximum tension), use this:

w = your weight
v = vertical displacement when you're standing in the center
T = line tension
L = line length

T = w * sqrt [(L/2)^2 + v^2] / (2 * v)

It's a bit simplified, because it doesn't take into account weight of the line.

Oops: edited to fix equation!

FLAG
By Ryan N
From Bellingham, WA
Apr 1, 2012
RJN
@Eric you just blew my mind. Thanks for the info. So the forces being applied to a short and long are the same it's just the amount of tension to get it equal is the difference?

FLAG
By crewdoglm
From TAFB CA
Apr 1, 2012
78 degrees north at 40,000 bearing about 220. Five...
I saw you partially addressed this: what are the loads at a slackline's attachment points during actual use? The math suggests given my own limited knowledge, that said attachment points will experience MUCH higher loads than the mere weight of the person based on the large angle in the middle of the slackline. Stretch mitigates that load but would require some kind of spring equation to be included right?. It would interesting to see actual measured numbers when guys are bouncing on that thing. On the face of it, the mechanical disadvatage affecting the webbing is severe. Nice presentation BTW.

FLAG
By 20 kN
Administrator
From Hawaii
Apr 2, 2012
Eric Krantz wrote:
One very key observation here is that both the 3/4" and the 1" broke at the same time when threaded. It's key because [(1) it means they also broke at the same elongation AND (2) tested separately they break at slightly different elongations], OR (3) your machine has a very high machine softness, and they actually broke a few milliseconds apart and you don't know it. I mentioned "pinching" above and I'd like to expound on that, but it will require a few paragraphs. We need to analyze the forces a bit more to get a better grasp of what's going on: First, an assumption: Assuming each line breaks at exactly 50% elongation (which is true within 4-5%)... in that case, the situation could be drawn like this just before reaching 50%: Answer: F(3/4") + F(1") The forces in each respective line, at the same elongation, can not be any different if they are tied to the same member, than if they are tested separately. Together, the force experienced by the anchor is the sum of the two individual forces. Lower ultimate strengths can be caused by two things: 1. The assumption stated is wrong. Maybe the 3/4" breaks on average at 47% and every 1" at 51% (instead of both at 50%). When pulled together, the 3/4" would break when both were elongated at 47%. In an incompressible system (neglecting machine softness), the 1" would still be at 47% elongation after the 3/4" broke (because the machine doesn't move when the force is reduced). This is not the case in the threaded scenario because BOTH are breaking at the SAME time (same elongation), and THAT's the kicker here. And THAT is why the diagram I drew is way different than the threaded scenario. The reason they both break at the same elongation (an elongation which we don't know), is reason number 2: 2. Where the line wraps around locker, the outside loop of the 1" webbing is putting very high pressure on the 3"4" webbing, all of which are squishing the inside loop of the 1" webbing with a hell of lot of pressure into the steel, but only 3/4" wide of it. I should have drawn this out, too. Think of a sandwich (a ten foot sub is your threaded line) with 4 layers... the bread is the 1" and the meat & cheese are the 3/4". The top layer of bread and the meat layer and the cheese layer are each pushing down really hard on the bottom layer of bread, all of which are wrapped around a round steel bar. The bottom slice of bread gets cut. Right? All this force is squashing down on 3/4" strip of the 1" inside loop, and cuts it at a force less than the sum of the two separate strands breaking strength. Once IT fails, there goes the rest. This was my "pinching" comment. For perspective, the pressure on the section of inside 1" loop at the line locker can be roughly computed as: [3323 + (0.5)(3920)] / (3/4" x 1/4") = 28,176 psi. A pressure like this may slice the inside section in half, even if the knife is as dull as a 1/4" diameter steel bar. And when it fails, the whole this fails... cuz the 1" fails first! Eric PS Final thoughts... The percent elongation is NOT linear with force applied (When at 44% elongation, stretching an extra 1% may require an extra 100 lbs force, whereas an extra 1% at 5% elongation may require only 32 lbs). We would like to know the relationship between percent elongation and force. You have plots of force vs time.... time is really not that relevant. Please plot force with percent elongation. If you did it with every 7% or so, you would have about 7 points before failure. Plotting them should make a nice logarithmic plot. (I'm guessing again). Also, what was the percent elongation at failure of the threaded strand? Did I miss that somewhere?

Okay, now I understand what you are saying. Here is a picture of the failure mode of the 3/4" threaded:



As you can see, the material did indeed fail over the carabiner. However, the picture seems to imply the outside layer (the layer furthest from the biner) of the webbing failed before the inside. If you look at the picture, you will notice the rip on top of the webbing is larger than the rip on the inside of the webbing. My thoughts are that if the inside layer failed first, the inside of the webbing would have the longest tear because the middle and then the outside layer would subsequently fail in a zippering pattern. Furthermore, although the inner layer may be getting squished, I believe the outside layers see more tension. The "sandwich" is being warped in a 180 degree angle around the biner. Accordingly, the outside layer will have to stretch and the inside layer will have to compress in order for the sandwich to take that shape. When that happens, the outside layer of the sandwich is going to see a greater amount of tension than the inside layer.

The failure mode of all of the threaded lines I tested, 3/4" and 1", show that the layer furthest from the biner (the outside layer) has the longest tear and subsequently, the inside layer has the shortest tear.

I did not publish the elongation from the threaded line because it seems as if the readings were erroneous. I got 7.1% 300lb. elongation for the 1" threaded line which does not make sense, as there is no way 1" threaded could elongate more then 1" non-threaded. But the measurement method I was using was not accurate enough to properly gauge a one percent difference anyway. Really, I just need to revisit the elongation portion of the test and use a digital caliper this time around.

However first I want to test a threaded line with two separate load cells to see exactly what the weight on each piece is.

FLAG
By 20 kN
Administrator
From Hawaii
Apr 2, 2012
crewdoglm wrote:
I saw you partially addressed this: what are the loads at a slackline's attachment points during actual use? The math suggests given my own limited knowledge, that said attachment points will experience MUCH higher loads than the mere weight of the person based on the large angle in the middle of the slackline. Stretch mitigates that load but would require some kind of spring equation to be included right?. It would interesting to see actual measured numbers when guys are bouncing on that thing. On the face of it, the mechanical disadvatage affecting the webbing is severe. Nice presentation BTW.

Thats part of my next chapter, where I discuss how much force a leached highline fall actually produces. Before I present the data from the leashed falls I will use someone walking on a slackline as a small scale theory representation to try to explain the concept first. I just finished gathering all of the data, I just need to go through it all. But I can tell you that in my test, I had about 300 lbs. of tension on a slackline at idle, and while standing in the middle it increased to around 600 (I weigh 160 lbs.) In the next chapter I am also going to try to address the theory that standing, or falling, next to the anchor produces more force on the anchor than standing or falling in the middle. So far all of my data implies that is untrue. In every test I have done, I was able to produce more force at the anchors standing or falling in the middle of the line then I was able to standing or falling at the anchor, which is really interesting considering my perception when falling or standing near the anchor is the exact opposite.

Gunkiemike wrote:
Did you sew the end loops yourself? Can you provide any evidence to support my hypothesis that the strength of a sewn loop is quite linearly proportional to the number of stitches?

The material you use to stitch the webbing, and the thickness of the thread, will have a major effect on how strong the stitching is. I could try stitching a loop and doubling the bar tack quantity from one sample to the next to see if it doubles the strength. I made a mistake and threaded the bobbin of my sewing machine with cotton, so I had a 50% nylon, 50% cotton blend, and that really screwed me over, the stitching was failing at 1500 lbs with 10 bar tacks! I switched out to all nylon and that tripped the strength.

Ryan N wrote:
@Eric you just blew my mind. Thanks for the info. So the forces being applied to a short and long are the same it's just the amount of tension to get it equal is the difference?

I am not 100%, but no, I do not think that statement is ultimately correct. If the angle between the load (the slacker) and the anchor points remained the same, than yes your statement would be correct. However the angle produced by the load is a function of how tight the line is, how much the material elongates, and how heavy the load is. Tighter lines will resist a change in the angle more than looser lines will. Consider walking on an infinitely tight slackline. It would be like walking across an I-beam, it would not bend at all (0 degree deflection). Likewise, if the anchor points were right next to each other and there was no tension on the line, standing in the middle would essentially be like hanging from a rope (180 degree deflection). But someone else on here is probably more qualified to explain this than me.

FLAG
By Eric Krantz
From Black Hills
Apr 2, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
crewdoglm wrote:
I saw you partially addressed this: what are the loads at a slackline's attachment points during actual use? The math suggests given my own limited knowledge, that said attachment points will experience MUCH higher loads than the mere weight of the person based on the large angle in the middle of the slackline. Stretch mitigates that load but would require some kind of spring equation to be included right?. It would interesting to see actual measured numbers when guys are bouncing on that thing. On the face of it, the mechanical disadvatage affecting the webbing is severe. Nice presentation BTW.


Hi Crewdog,

The equation I wrote above will calculate the load (tension (T) IS the load). If you know the length (L) before you step on it, your weight (w), and the vertical displacement (v) after you step on it, you can calculate the load at each anchor:

Slackline tension
Slackline tension


To properly tension the slackline, determine it's length and figure out the maximum vertical displacement required in the center, and your weight. That will give you T (load at the anchor) when loaded. Now, if 20kN would post a graph showing % elongation vs tension, we could use that to determine the load at the anchor when you're NOT standing on the line, hook up the load cell, and tighten it until it's where you want it.

Edit: My diagram assumption 3 is wrong... rethinking this, the equation ONLY applies when standing (or hanging) on the line. It cannot be used to determine tension based on maximum vertical displacement during a fall. At maximum displacement (velocity = 0), the system is NOT in equilibrium. The equation requires equilibrium because it is based on the sum of forces in Y direction = 0.

FLAG
By Eric Krantz
From Black Hills
Apr 2, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
20 kN wrote:
I am not 100%, but no, I do not think that statement is ultimately correct. If the angle between the load (the slacker) and the anchor points remained the same, than yes your statement would be correct. However the angle produced by the load is a function of how tight the line is, how much the material elongates, and how heavy the load is. Tighter lines will resist a change in the angle more than looser lines will. Consider walking on an infinitely tight slackline. It would be like walking across an I-beam, it would not bend at all (0 degree deflection). Likewise, if the anchor points were right next to each other and there was no tension on the line, standing in the middle would essentially be like hanging from a rope (180 degree deflection). But someone else on here is probably more qualified to explain this than me.


You're right, it's only correct if the angle doesn't change. However, it's easier (and more accurate) to measure the vertical displacement due to the weight of the slacker than it is the angle, so use that.

FLAG
 
By kennoyce
From Layton, UT
Apr 2, 2012
Climbing at the Gallery in Red Rocks
Eric Krantz wrote:
Now, if 20kN would post a graph showing % elongation vs tension, we could use that to determine the load at the anchor when you're NOT standing on the line, hook up the load cell, and tighten it until it's where you want it.


The problem is that a slackline will relax after it is tensioned, meaning that you can tighten it until it's where you want it, but after a few minutes the tension will drop significantly and you'll have to retighten it.

FLAG
By Eric Krantz
From Black Hills
Apr 3, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
kennoyce wrote:
The problem is that a slackline will relax after it is tensioned, meaning that you can tighten it until it's where you want it, but after a few minutes the tension will drop significantly and you'll have to retighten it.


Good point. Also will relax with moisture and tighten back up when drying.

So, my question is, does this affect the ultimate strength?

20kN, are you planning any "slow-pull" tests? I'm thinking like, crank it up there to 40% elongation, record the tension, then wait a day, crank it back up to the same tension and record % elongation, wait another day... etc etc. Soon it will either break or stabilize. If it stabilizes, then find determine ultimate breaking strength. Fun!

FLAG
By 20 kN
Administrator
From Hawaii
Apr 3, 2012
Eric Krantz wrote:
Good point. Also will relax with moisture and tighten back up when drying. So, my question is, does this affect the ultimate strength? 20kN, are you planning any "slow-pull" tests? I'm thinking like, crank it up there to 40% elongation, record the tension, then wait a day, crank it back up to the same tension and record % elongation, wait another day... etc etc. Soon it will either break or stabilize. If it stabilizes, then find determine ultimate breaking strength. Fun!

I am working on a elongation vs tension graph now. I found a more accurate way of measuring elongation, so I spent all day redoing all of the testing for the elongation portion of my study. Basically I just use a 5:1 and a ratchet and pull the webbing tight. I stop every 100 or so pounds, wait a bit for the tension to stabilize, and record the length of the sample. This way I now have around 12 data points ranging from 100 lbs to 1000 lbs for the 3/4", 1" and 1" threaded samples. From this I hope I can input the data into an XY scatter chart in Excel and the computer and identify a logarithmic pattern between the samples and therefor predict the elongation of the webbing at any tension. However that assumes there is actually some type of identifiable pattern, it is possible the elongation of the webbing does not follow any identifiable pattern, but I doubt it. However I am less sure if Excel can identify logarithmic patterns. It may just plot an average bar using the data points and assume that the elongation increases linearly. Hopefully Excel is smart enough to recognize otherwise. I am also going to pull each portion of the threaded 1" and record the weight on each strand. That way I can compare the elongation of the threaded webbing to the difference between the weight distribution to see if the elongation difference has a linear effect on the weight distribution of the webbing. That is a key point of my study that I did not do last time.

I am not sure that I can accurately test webbing elongation over that much time. My hydraulic ram might leak a bit back into the tank if it is under load that long, there by artificially reducing the tension. Hydraulic fluid does not compress, so just a very small leak could make a noticeable impact in the tension of the webbing.

However I am pretty sure the webbing would never stabilize, every time I increased the length of the cylinder the material would just stretch more, eventually failing. I think the true ultimate breaking strength of the webbing is highly dependent on the peak load duration, just like it is with dynamic ropes. If I crank the webbing to its breaking strength as fast as possible, I think I am going to get a much higher value than if I were to just let it sit there for a day and stretch out.

But webbing is kind of weird, it seems as that it can elongate not only negatively, but also positively. I did some testing for my third chapter, highline falls analysis, yesterday and I noticed that after sitting on a slackline for awhile, when I got back off the line the tension increased slowly. So it seems if you stretch webbing out by increasing the load from one value to another and then release the load back to its previous value, the webbing will try to shrink back to its original length before the increased load was applied, there by increasing the tension slowly. The webbing will never reach the previous tension before the increased load was applied, but it wont rest at the tension present as soon as the load is dropped back to its original value. Strange.

FLAG
By Eric Krantz
From Black Hills
Apr 3, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
20 kN wrote:
However I am pretty sure the webbing would never stabilize, every time I increased the length of the cylinder the material would just stretch more, eventually failing. I think the true ultimate breaking strength of the webbing is highly dependent on the peak load duration, just like it is with dynamic ropes. If I crank the webbing to its breaking strength as fast as possible, I think I am going to get a much higher value than if I were to just let it sit there for a day and stretch out.


You are probably right, it would not stabilize. That would be plastic deformation, and would occur at high values of tension.

I found something, wish I had time to read it. Stress-strain curve for nylon. See figure 6:

ocw.mit.edu/courses/materials-...

In that graph, strain is the same as your % elongation. Stress is load per unit cross sectional area - you would have to divide your tension value measured by the load cell by the cross sectional area of the specimen. However we care more about the tension response (analyzing the system) and not the stress response (analyzing the material). Right? So, replace tension with stress and make the same graph, and it should be fairly similar shape.

Taking into account plastic deformation, you should actually measure the tension for each unit elongation you have (instead of the elongation for each 100-lb increase in tension). In other words, every time you get another inch of stretch, record the tension. If not (again, see figure 6), you will miss the entire plastic region of the curve (the point past "necking").

FLAG
By Eric Krantz
From Black Hills
Apr 3, 2012
smoke break, pitch 5 or 6 (or 7??) of Dark Shadows
20 kN wrote:
From this I hope I can input the data into an XY scatter chart in Excel and the computer and identify a logarithmic pattern between the samples and therefor predict the elongation of the webbing at any tension. However that assumes there is actually some type of identifiable pattern, it is possible the elongation of the webbing does not follow any identifiable pattern, but I doubt it. However I am less sure if Excel can identify logarithmic patterns. It may just plot an average bar using the data points and assume that the elongation increases linearly. Hopefully Excel is smart enough to recognize otherwise.


Excel scatter plot -- right click on the data set and add trendline, then format trendline and choose logarithmic as your model.

Looking at Figure 6, however, you won't need to do this. It won't be exactly the same because you don't have a solid sample, instead a woven pattern, but the material is the same (nylon).

FLAG
By Tom Grummon
From Golden, CO
Apr 6, 2012
Top of Montezuma's Tower
So here is some anecdotal evidence about threaded lines:

The other week we had an approx. 80' line set up fairly tight. The line was 1" tube threaded with 2 pieces 11/16" or 3/4", I forget, flat webbing, weird I know. The line was also kind of old, its not my line so I can't say how old. Guys (who are better than me) were doing the standard tricks on the thing (butt bounces, jumping, ect.) with out incident. I then got on the line and walked it upon turning around at the other side one, or both, of the pieces of flat webbing inside broke while I was walking on it (just walking, not bouncing).

The line did not fail catastrophically , but rather released some tension by stretching. The new state of the line was marginally looser and the webbing on the inside was all bunched up.

As this was not a highline, just a line in the park, we decided to get it a little tighter and see what happened. After re-tightening it we continued to use it including the same bouncing and jumping as before without incident.

I left before the line was taken down, so I'm not privy to what the webbing on the inside looked like when deconstructed.

What this says as to potential failure of threaded lines in general is hard to say, but anecdotally the line behaved in a way to prevent catastrophic failure due to one breakage.

-Tom

FLAG
By Tom Grummon
From Golden, CO
Apr 6, 2012
Top of Montezuma's Tower
Very interesting reading. Thanks for posting you tests to be rigorously reviewed by the scientific community of MP.com.

FLAG
By 20 kN
Administrator
From Hawaii
Apr 8, 2012
Eric Krantz wrote:
Excel scatter plot -- right click on the data set and add trendline, then format trendline and choose logarithmic as your model. Looking at Figure 6, however, you won't need to do this. It won't be exactly the same because you don't have a solid sample, instead a woven pattern, but the material is the same (nylon).

I wont need to do what? Choose logarithmic as the tread line model?

Eric Krantz wrote:
In that graph, strain is the same as your % elongation. Stress is load per unit cross sectional area - you would have to divide your tension value measured by the load cell by the cross sectional area of the specimen. However we care more about the tension response (analyzing the system) and not the stress response (analyzing the material). Right? So, replace tension with stress and make the same graph, and it should be fairly similar shape. Taking into account plastic deformation, you should actually measure the tension for each unit elongation you have (instead of the elongation for each 100-lb increase in tension). In other words, every time you get another inch of stretch, record the tension. If not (again, see figure 6), you will miss the entire plastic region of the curve (the point past "necking").

That is an interesting article, I read up to figure six. However, I am not quite sure how necking and the PSI/ elongation ratio is relevant to this study. As you said, we are concerned with analyzing the elongation of the system rather than the material.

The stress/ strain graph shows how much the material elongates in relation to the PSI value of force exerted on the webbing. Although an interesting graph, it would not be able to tell us how much the 3/4" elongates in relation to the 1" which is what we are ultimately concerned about. Regardless if necking takes place in webbing or not, the material will still increase in elongation logarithmically to the weight applied. Although the elongation/ tension ratio of the webbing is certainly not linear, its not polynomial either. I dont believe there would never be a case where the force applied increases and the elongation decreases. If the tension increases, the elongation will also increase.

Anyway, I just finished gathering the new and much improved elongation data so I hope to amend this thread shortly.

FLAG
 
By 20 kN
Administrator
From Hawaii
Apr 9, 2012
I have modified the original post and added quite a lot of new information! Hopefully this new elongation data will make my research a bit easier to understand.

FLAG
By 20 kN
Administrator
From Hawaii
Apr 26, 2012
More information and graphs added.

FLAG


Follow replies to this topic? Notify me at the top of web site.
1

Email me.
Page 1 of 1.