Author Topic: KNOT TEST RIG DESIGN AND SETUP  (Read 7847 times)

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4278
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #15 on: November 25, 2018, 08:50:09 PM »
I will reiterate that although I was/am-yet? on the side of
thinking that evidence suggests a difference,
IMO it's not something to rise to great significance
and worth much worry.  (Might there be enough
extant test reporting to show a *hint*of bias?)
((Admitting that to "feel the force (first)" requires
there to be some resistance to the injected movement
which resistance is *feeling* it ... on & on & on.))


More significant, IMO, is the use of TWO knots
per test, simultaneously loaded.  I admit that there
are *impure* aspects to this vis-a-vis statistics,
but again I'll assert that their significance is small
--or that a conservative/more-2BSure lower value
is better to have in mind(!)-- and not to be worried about
(and esp. as stat.s would want prohibitively costly
testing).

Better is to have a way to extract good general
values and information about how/where knots
break.
(E.g., EvansS had some tests of HMPE in end-2-end
knots where he reached some strong force (maybe
still well shy of 50% tensile), the knot held A WHILE,
then ... slipped down to 50# force or so!!?  I'm thinking
like SS369 that perhaps there was such heating and
melting --low-threshold for HMPE!-- that its surfaces
at least in the knot became fluid and ... that
won't hold well!
What if he could repeat the test but pour icewater
or ... do it in sub-0^F temps?!  That might show
a different result?!  (HMPE does pretty well, cold
--better than PP, e.g..)

Heck, even if testing were done NOT TO RUPTURE,
but to say about 40-50% tensile --where likely a
knot had taken a highly compressed form unlikely
to change much, further (well, this would be something
we'd be able to check), just seeing what our dressed
& set knot had become, geometry-wise, would be
helpful insight.  "OHhhhhhh, so THAT explains it!"
could well be a reaction.  --seeing so much torsion
in the SPart, say, or a thought-to-be nice gradual
curve hauled dead straight into a hard turn.
... and so on.  AND presumably such insight would
not be something in need of repeat testing PER
ROPE TYPE, but only across various ropes.

E.g., Mark did some side-by-side presentations
of images of BWII under 3? increasing loads.
That tells much!  --and his rope lives to show again.

Thanks,
--dl*
====

agent_smith

  • Sr. Member
  • *****
  • Posts: 1470
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #16 on: November 26, 2018, 01:58:12 AM »
A key concept that needs to be reminded is that very few knot testers report details of their test rig design and on which side knots have been either breaking or jamming.
It is an area that is largely devoid of data.
Testers simply report their data - and it is just assumed that readers somehow understand the design of their test rig and what actually transpired.

Those testers that use video - in these instances we can catch a glimpse of their test rig - but again, details are normally not given - you just have to try to understand it all from the video footage.

It does appear that the default setup is to have force 'injected' from one side - that is, the force generating machine is installed on one side of the rig and it 'injects' (or introduces) force from one side.
Now this is not in conflict with Newtons laws - I am merely describing physical design.
In certified test labs, the test rig can be horizontal or vertically mounted - and normally the force generating machine is in the form of a ram.
The ram is installed on one side - that is, it is actuated from one side against a fixed anchor point (one anchor point is fixed in place and does not change its location). The opposite anchor moves - in order to induce tension which s-t-r-e-t-c-h-e-s the test article. All test rigs have a travel/stroke limit.
I haven't seen professional, certified test rigs which have 2 rams installed - each one actuating and 'injecting' force from opposite sides simultaneously.
Note: Please read carefully...I did not say this is 'bad' - I'm simply saying that I have never seen it.
The same goes for the word 'inject' - its simply a physical description of where the force generating machine is installed - and the point-of-origin of force. In my personal case, my force generating machine is a 'lever hoist' - which is hand operated from one side of the test rig.

As Scott has alluded to, each knot will behave differently when subjected to load - and different rope segments within the core will move at different rates - no doubt determined by its geometry. Although many like to assume that force is transmitted instantaneously from point to point in a test rig - the fact that there is a knot which compresses and has moving segments may confound that assumption. Now I used the word 'may' - I didn't say 'will' or 'does'.
I have a feeling that force is not transmitted instantaneously from point-to-point in a test rig...there is a 'point-of-origin' (ie the force generating machine) - and force must pass through the knot which itself has segments moving at different rates. There may be a millisecond delay as the knot absorbs initial force and rope segments are ratcheted out. It may be a millisecond or it may be more or less (I don't know). No one has actually investigated this.
This is one of the reasons why I prefer to test identical knots against each other (like against like)...rather than 2 different knots against each other.
I think there is very little data about actual test rig design and how this might affect test results.
The point being...if every tester uses a different design rig with different test parameters - it may be difficult to build a consistent database.
This may hint at why there is such a huge disparity in test results around the world (aside from the fact that everyone uses different ropes/cords and rarely reports on exactly how a knot specimen is tied).

As stated, none of this has really been explored or discussed in any detail before - its all just been 'assumed'.
And the phenomena of a knot breaking on one side of the test rig or the other...this hasn't been properly explored. I don't know of any knot tester who even bothers to report which side of the rig the knots broke on (or if its random). It appears to be rarely reported (if ever).

And if a tester does report a little more detail - it potentially opens pandoras box...so maybe that's another reason why some testers prefer to only report the results and not go into details about the test rig design and other observations (eg on which side the knots broke or jammed).
« Last Edit: November 26, 2018, 02:11:44 AM by agent_smith »

GlasafMT

  • New Member
  • *
  • Posts: 4
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #17 on: November 26, 2018, 12:50:42 PM »
Should you also be differentiating between "Strain generating machines" and "Force generating machines"?

What's being described in this setup is a strain generator. When the knot slips, you are likely losing some load in the system.

That is very different from a real-world situation where you have someone hanging off the end of the rope.  If the knot slips in that system, load is not released.

Is this worth considering?  I suspect that it's much easier to build a test rig with a ratchet than find enough heavy weights!

agent_smith

  • Sr. Member
  • *****
  • Posts: 1470
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #18 on: November 27, 2018, 05:19:52 AM »
Quote
I suspect that it's much easier to build a test rig with a ratchet than find enough heavy weights!

Thanks for your post GlasafMT.

I think you had already answered your own question with your last sentence..

The manual handling of a series of weights would create a number of problems - not to mention back strain from all the lifting.
Also, such a test rig would have to have a number of weights - which would need to be added sequentially - and each time a weight is manually handled and added, there would be a spike in the load as it is attached and then released from your grasp.
It isn't practicable to work like this.
Also, you would have to purchase a number of weights - likely to be circular disk weights that body builders use on a bench press machine (these are insanely expensive to purchase).
Then there is the storage of all those individual weights...

You mentioned that if a "knot slips, load is not released"...while a knot is 'slipping' - something causes it to slip and this usually means load. If zero load, then no slip.
I might also point out that the knots I test don't 'slip' - they just reach their yield point and then fail. It is not a case of 'slipping', rather, the knot compresses more and more and tries to resist the injected force. Once the force reaches a certain threshold - the knot yields (you can hear snapping/popping sound as this yield point is reached).

The ideal test rig is to have a permanent bench setup that uses a hydraulic ram that can reach 5 tons or more force.
Attached to the ram would be a tension load cell that polls at 50 cycles a second to capture data - that can be linked to a laptop computer.

The Rock Exotica 'Enforcer' load cell is a marvelous gadget but, it only goes to 20kN (about 2 metric tons).
I wish Rock Exotica would release a newer version that can reach at least 40kN (about 4 metric tons).
« Last Edit: November 27, 2018, 05:22:01 AM by agent_smith »

agent_smith

  • Sr. Member
  • *****
  • Posts: 1470
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #19 on: November 27, 2018, 07:30:30 AM »
Derogatory comment by Derek:
Quote
Certainly it was not important enough to justify Mark backing himself into a hole and throwing out invective to cover his embarrassment.
This type of comment is an example of a fertile imagination actuated by malice.
The real embarrassment ought to lie with Derek - for failing to read English properly.
I assume that Derek speaks and reads English as a first language - and if this is true - it only adds to the level of puzzlement over his inability to read properly.
I never attached the word 'direction' to injection of force - instead; merely describing a physical design where the force generating machine is installed on one side.

Derogatory comment by Derek:
Quote
and I doubt that Mark is of a mind to swallow his pride and come and play ball,
This type of comment is another example of a fertile imagination actuated by malice.
The person who should be 'swallowing his pride' is Derek - for failing to properly read and understand his native language (English?) - and to reach false conclusions based on a false premise.
The concept I had advanced being; 'injection of force' - was directionless. It was simply to describe a physical design where a force generating machine was installed on one side of the test rig. This was to distinguish it from a design where you have two (2) force generating machines - one at each end - both 'injecting' force simultaneously.

And further to that, I made no comment about a dual force generating machine setup (ie each one pulling at each end simultaneously) being 'bad' or 'no good' - merely hoping to encourage repeat testers to use a similar test rig setup to mine - so as to remove potential for bias in the tests results (compared to mine).

I had surmised that as force is 'injected' into the test rig - the knot begins to respond according to its geometry. Different rope segments will move at different rates within the knot core. I also surmise that there could be millisecond delays in the response of the knot to load - and that maybe end-point to end-point transmission of force is not instantaneous or precisely equal on account or ratcheting of the knot segments and heat buildup and heat loss by direct radiation and by convection. In other words, it seems possible that some energy pumped into the system is converted to heat and possibly vibration. I don't have all of the answers and I doubt anyone else does either - all I can do is report what I observe.

One way to measure potential losses is to use two (2) load cells - one installed at each end. The load cells would have to be very accurately calibrated and poll at 50hz - with data captured in real time on a laptop. If there is disagreement between the load cells and in millisecond delays - even by the tiniest amount - that would point to something worth following up on with a battery of further tests. I would say that this sort of real-time precision measurement would be beyond the capability of the hobbyist/enthusiast tester.

...

I have noticed that the people who make stuff up and seem willing to sling insults are also the same people who spend more time slinging insults than doing practical testing and publishing their results for others to see.

I also note that it is rare for a knot tester to go into any fine detail about their test rig and to report on observations about which side their knots jammed or broke on. Its simply never reported.
And yet, if you do go the extra mile and report these observations - you risk drawing negative attention for doing so.
I can understand why many on this forum choose to remain silent and not bother to report on their own tests..I mean, why bother when you have people who make stuff up, sling insults and construct meanings from their own imagination...?

It all seems not worth the effort.

SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 2021
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #20 on: November 27, 2018, 02:32:20 PM »
To All;

Healthy debate, intelligent banter, staying on topic and positive contributions are what we should stay doing. I feel that there is a situation that is degrading here and personalities are equaling knots under strain.
So, lets leave the invective, etc. out of here and put back the ideas and discussions.

What I have used in the past has been a ratcheting come a long that employs a  cable, winch like. I know there has got to be too many built in loses in my testing set up. And I don't have a load cell either. I use the stated MBS of the rope's manufacturer as  the breaking benchmark. And, sincerely, that matters little because I am trying to cause a break. I use the tug of war method between test specimens, sometimes with X vs Y different knots. Example; Standard bowline vs modified bowline or vs F8, etc.

My anchor points are generally a large tree and my truck's trailer hitch. I can adjust the distances that way. I loop a heavy duty, industrial sling ( way beyond the strength needed) around the tree with a screw shackle that is near 3/4 inch in diameter to the test sample. From there the other end of the sample, two eye knots,  generally 16 inches overall length, some locations marked with sharpie, to the come a long which is connected to my truck via large chain. I take out all the slack by driving, lock the vehicle down well then continue by adding load at the rig. This allows me to watch the test fairly close and I can even record, stills and video.

Not scientific in the least ( nor very safe), I know, but my testing is for me, to verify certain ideas I have before I offer them to anyone.

I would like to thank you Mark for what you bring to our table and I personally do understand how much you invest, life and otherwise.

Same goes out to Alan Lee.
Thank you both!

So, please, no more sword slapping, by anyone.

SS
« Last Edit: November 27, 2018, 02:33:14 PM by SS369 »

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1572
  • Knot Botherer
    • ALbion Alliance
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #21 on: November 30, 2018, 08:09:50 PM »
The amazing (and surprising) power of the shock wave --  relevant perhaps to the proposal that small shock waves induced by the Strain inducer ratchet action could be leading to a bias in the jamming observed in the knot nearest to the Strain inducer.

https://www.facebook.com/5min.crafts/videos/708318142644107/?t=41

Derek
« Last Edit: November 30, 2018, 10:17:26 PM by DerekSmith »

SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 2021
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #22 on: November 30, 2018, 08:59:26 PM »
Ahh, I see now.
Maybe edit your post with some explaining of the relevance.

I have personally noticed the incidence of tensile resonance. I suspect and believe it is from the grabbing and releasing of forces in the tight part(s) of the knot.
And I am of the opinion that the knot closest to the tensioning mechanism receives it first.
First come, first served...

SS

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1572
  • Knot Botherer
    • ALbion Alliance
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #23 on: November 30, 2018, 09:31:32 PM »
Ahh, I see now.
Maybe edit your post with some explaining of the relevance.

I have personally noticed the incidence of tensile resonance. I suspect and believe it is from the grabbing and releasing of forces in the tight part(s) of the knot.
And I am of the opinion that the knot closest to the tensioning mechanism receives it first.
First come, first served...

SS

RE 'post with some explaining' - Will do Scott,

Re 'First come, first served'  - as the shock wave enters the first knot, there is a localised Stress wave and the knot, verging on responding to an incremental increase in load, is tipped over its static frictional resistance threshold and tightens (slips internally).  This movement (force x distance) is work and the energy to induce it is taken from the shock wave, cancelling it out, thereby preventing it from getting to the second knot.

Derek

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1572
  • Knot Botherer
    • ALbion Alliance
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #24 on: November 30, 2018, 09:32:54 PM »
Should you also be differentiating between "Strain generating machines" and "Force generating machines"?

What's being described in this setup is a strain generator. When the knot slips, you are likely losing some load in the system.

That is very different from a real-world situation where you have someone hanging off the end of the rope.  If the knot slips in that system, load is not released.

Is this worth considering?  I suspect that it's much easier to build a test rig with a ratchet than find enough heavy weights!

Glasa, you are of course totally correct, we are using Strain generators, the strained cordage and associated knots respond by generating a reactant Stress force measured on the load cell device.

And again, I totally agree, in a gravitationally loaded system, as the knot slips the load remains constant.  But, just as we can use gravity to load a knot to a certain (constant) value, so with a Strain inducing system, we can induce strain until the requisite stress is achieved and the results can be observed.  For example, with my Instron screw driven system, I am able to program in a Stress force and record the resultant Strain as the test subject slips or distorts.

Oh, and welcome to the Forum / Topic...

Derek
« Last Edit: November 30, 2018, 10:39:38 PM by DerekSmith »

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1572
  • Knot Botherer
    • ALbion Alliance
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #25 on: November 30, 2018, 09:51:01 PM »
@ Mark,

Quote
I might also point out that the knots I test don't 'slip' - they just reach their yield point and then fail. It is not a case of 'slipping', rather, the knot compresses more and more and tries to resist the injected force.

A simple observation of the amount of 'slippage' that occurs within a knot may be made by sewing a small 'witness' thread close to the points the cordage enters the knot.  You will be able to observe the 'witness' marks moving away from the knot as slippage occurs inside the knot and cord is drawn out from it.  Some, poor, knots also 'slip' by distortion and consequential migration along the cord.  In low cF cordage, the great majority of known knots 'slip' badly.

All knots slip under load.

Derek

NB - Dan taught me this simple but valuable technique several years ago when I set up my 10 ton hydraulic test rig to run knot comparison tests.
« Last Edit: November 30, 2018, 10:44:38 PM by DerekSmith »

agent_smith

  • Sr. Member
  • *****
  • Posts: 1470
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #26 on: December 02, 2018, 12:47:00 PM »
per Derek:
Quote
All knots slip under load.]

Although off-topic and not directly related knot test rig design - I completely reject this statement.

However, my rejection originates from the strict English dictionary definition of 'slip'.

It is possible that Derek intends and has constructed a narrow meaning to the word 'slip'.

For example, I have conducted load testing on the Zeppelin bend and I observed no 'slippage' (by my definition of the word). What I did observe is response to load which caused the following:
[ ] core compression
[ ] extrusion of rope from the core
[ ] stretching

These events continue until core compression (collapse) is complete and the elastic limit of the knot is reached...at which point the knot reaches it yield point.

It is possible that Derek may choose to construct a meaning where core compression, extrusion of rope from the core and stretching is held to be 'slippage'.

However, the ordinary definition of 'slip' or 'to slip' could be taken to mean:
^to let loose from a restraining leash or grasp
^to cause to slip open : release, undo or slip from a lock
^to let go of
^to disengage from (an anchor)

In the context of knotting,'to slip' implies some form of insecurity. And this 'slippage' therefore indicates tail length reduction as the knot allegedly 'slips'.
In the case of the Zeppelin bend, I don't see tail length reduction - because it isn't 'slipping' (by my definition).

I see the same phenomena in #1425A Riggers bend and #1411 F8 bend. Any tail length reduction is simply a consequence of core compression - the tails partially draw into the core (but only a small amount) - until core compression inhibits any further movement and at which point the elastic limit of the rope is reached and fracture is imminent. This tail draw-in is not 'slippage' (in the ordinary sense).

Due to the complexities of the English language, it is likely that some will construct a meaning where tail draw-in due to core compression and extrusion of rope outwards from the core is defined to be 'slippage'. I would disagree with such a construction.

Edit note: Fixed quote function
« Last Edit: December 02, 2018, 02:09:09 PM by agent_smith »

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1572
  • Knot Botherer
    • ALbion Alliance
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #27 on: December 02, 2018, 05:55:02 PM »
the tails partially draw into the core (but only a small amount) - until core compression inhibits any further movement

With apologies for prolonging an off topic excursion Mark, could I ask how you see compression within a knot causes an otherwise unloaded tail to 'draw into the core'.  Or do you see it as some other mechanism that causes the tail to tail to be drawn into the core?

Derek

agent_smith

  • Sr. Member
  • *****
  • Posts: 1470
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #28 on: December 03, 2018, 01:17:02 AM »
Derek; start a new topic post titled:
"Proposition that all knots slip under load"
Debate your proposition there.
I would advise that you carefully define the word 'slip' - before embarking on expanding your theory.

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1572
  • Knot Botherer
    • ALbion Alliance
Re: KNOT TEST RIG DESIGN AND SETUP
« Reply #29 on: December 03, 2018, 01:51:03 AM »
@ Mark,

Thanks for your suggestion and your advice, I will consider both.

Meantime, I will take it that you have no rational answer to my question, which is good, because I could not think of one either.

Derek