Author Topic: Knot I have made requiring identification  (Read 8015 times)

cuesta

  • Jr. Member
  • **
  • Posts: 19
Re: Knot I have made requiring identification
« Reply #15 on: September 04, 2007, 12:41:37 PM »
Brian, Derek,

I am most curious to understand Derek's findings regarding the 4 dressings of the knot. Can you elaborate regarding how you came about distinguishing such differing characteristics in the knot? Was it by adjusting the pressure to tighten the knot or other adjustments or materials? Would it be possible to display your findings by way of photos or drawings.

The acronym USP, what do you mean by that?

Thanks for all your interest

Oscar



DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1530
  • Knot Botherer
    • ALbion Alliance
Re: Knot I have made requiring identification
« Reply #16 on: September 07, 2007, 02:37:14 PM »
Sorry Oscar,

I really should not use TLA's (three letter acronyms).

USP is a marketing term -- Unique Sales Point -- it is also an acronym that is starting to take on a meaning all of its own, vis "that which is special about the subject".

As for the dressings.  Your knot really tickled my interest, because, after discovering that it was simply a double cord slipped overhand with a twist, I realised that quite simply it should not be capable of doing what it was doing -- a slipped overhand should slip -- yet this little beastie had a definite ratchet built into it.

I first became interested in the mechanisms involved in knots gripping or slipping after having been introduced to the infamous Grief Knot AKA the Whatknot -- see roo-two's excellent site for details http://www.geocities.com/roo_two/reefknot.html  The Whatknot can either lock solid, or at the twist of the ends, it can flow like rope through a pulley.  From this knot, I discovered that components within a knot can act either as a ratchet or as a cog, the ratchet locks movement while the cog encourages it.  Tie this lovely example of knot mechanics for yourself and watch how one dressing sets the knot up with ratchets, while the alternative dressing creates two lovely cogs which run freely against one another.

When I come across a new knot, my first two exercises are usually to slacken out the knot in order to see the internal structures it contains, then to progressively 'deconstruct' it to seek alternative ways to tie it.  Doing this you immediately see that the 'Lyndy' is based on a slipped overhand with an additional twist.  This offered a simple way of tying the Lyndy 'in line'.  The trouble was, that when I tied the Lyndy this way it did exactly what a slipped overhand should do -- IT SLIPPED.  There must have been something special about the way you had tied it that I had missed which was conferring its ability to grip.

The Lyndy is 'in essence', four cores running parallel through the heart of the knot, surrounded by four gripping turns.  On one side of the knot, two of the cores emerge as the two load lines and the other two connect to the 'adjuster' loop.  On the opposite side of the knot, two of the cores connect to one of the load loops, the other two to the other load loop.  Dressing the knot allows the connections of one side to the other to be varied and which is connected to what makes all the difference in functionality.  Only two of the cores pass right through the knot connecting the adjuster loop to one side of each of the two load loops.  The other two cores are fixed to the knot and cannot slide under load.  When loaded, the four outer wraps are tensioned and squeeze the central cores and increases the friction between the cores and the wraps and between the cores themselves.

Consider for a moment the situation with only one of the cores being able to slide.



In this diagram, the grey ring represents the external wraps under load squeezing the four cores together.  If B,C and D are fixed to the knot and cannot slide and 'A' is the only core that is free to slide, then you will see that it is firmly gripped on all three sides by surfaces that do not move.  The surface contact of 'A' with the other cores and the ring is about 3.5 diameters and the Lyndy is about 4 diameters long, so we have a working frictional area of ca 14 diameters square.  If 'A' and 'C' are connected to a load loop holding say 100 lb, then 50lb would be transfered by 'C' into the knot and 50lb would be applied to 'A' which works out to about 3.5lb tension per square diameter of contact area.  If the intrinsic friction of the cord system is high enough it might not slip, but in most nylon braids it will slip easily.

In the case of the doubled slipped overhand (with a twist), there are two load loops, consider A-C is one loop and B-D is the other.



This time the load is shared across all four cords with A and B holding 50lb between them.  The frictional contact area has gone up so the tension on the contact surface has come down to about 2.5lb per square dia.  It should grip better than the single core -- but! A and B are effectively joined by the 'cogging' surface marked in green.  Because they are different cords, A and B will rarely be the same length, so the load will often be biased more onto one than the other.  If this causes one of the cords to be loaded above its slipping point, then some of its motion transfers a force into the B cord, inducing it to slip as well.  The two cord surfaces 'cog' one with the other to create a mutual inducement to move.  Despite the increase in frictional contact area, this knot is highly prone to slipping.

Now consider the 'resistant' Lyndy format.



In this configuration the adjuster loop is connected to cores A and D, with the load loops as before on A/C and B/D.

Three things have happened.  First, the frictional density has dropped still further to 1.7lb per square diameter.  Second, there are no 'cogging' surfaces to induce sympathetic movement and third, the two load loops are connected up 'opposite handed'.  Any slippage in the A/C loop will tend the rotate the load anticlockwise, feeding the force of this movement into the fixed 'B' leg of the other loop, while any slippage in the B/D loop rotates the load clockwise and this time the fixed 'C' leg resists the load.  This is 'ratcheting' where one part of the knot takes up and resists movement in another part.  Although in extremely low friction cord this knot will still slip, in 'normal' braids it easily holds to breaking strain.  The only 'tell' that you have tied this variant is the presence of one of the fixed cords entering the top of the knot between the two adjuster loop lines.  The 'tell' is shown in red in this image.



As this post is now quite large, I will cover the two intermediate variants in a new post.

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1530
  • Knot Botherer
    • ALbion Alliance
Re: Knot I have made requiring identification
« Reply #17 on: September 07, 2007, 06:39:58 PM »
In the previous post, I touched on the fact that knot performance may depend upon the configuration of the cords external to the knot itself.  This again opens the discussion that a knot may not end at the obvious boundaries of its structure and the ratchet effect of the two opposing loops is one such example.  We do not have a term for this concept of the whole interacting construct (or if we do then hopefully someone will educate me) but when the function of the knot depends on the dressing of the 'external' components (in this case, the loops) then we really need some way to recognise the 'extended' knot and have some terminology for such an entity.

In the case of the slip resistant Lyndy (shall we call it the A/D variant), the ratchet was effective without any deliberate dressing of the external load loops, because their connection to the cores A/B and C/D naturally conferred the counter rotating 'ratchet' effect.

In the two intermediary gripping variants, the intermediate effects has are achieved by 'dressing' one of the loops to generate the ratchet or cogging effects.

In the slipped overhand example above (let's call it the A/B variant), the cogging effect is present in the external loops in that any turning moment induced in the load through one cord slipping was passed directly to the other loop which further induced slippage.  If however one of the loops is twisted half a turn, then as slippage of A and B happens, one cord tends to turn the load clockwise, while the other tends to turn it anticlockwise.  The friction of the cords around the load and even against one another works as a ratchet to counteract the force causing the slippage and so a greater force is necessary to make this variant slip.  One half turn in one of the load loops is all it takes to make the slipped overhand variant behave much more like an intermediate A/D variant.

Finally, the ratchet effect working naturally in the A/D variant can be converted to a cogging effect by allowing a half turn to be incorporated into one of the load loops.  So for example if the A/B loop is twisted half a  turn and then the 'A' cord slips slightly, it allows the load to rotate clockwise and immediately transfers this load into the 'D' slipping cord.  This cogging effect in the working loops can weaken a previously strong grip of the A/D variant into an intermediate strength 'knot'.

Happy knotting.

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1530
  • Knot Botherer
    • ALbion Alliance
Re: Knot I have made requiring identification
« Reply #18 on: September 11, 2007, 01:56:29 PM »
Oscar,

You may be interested in some initial work I have done on the comparative strength of the Lyndy loop.

I have only compared it with four other loop knots at the moment -- Overhand Loopknot, Constrictor Loopknot, Bowline and Fig 8.

Guess what -- IT BEAT THEM ALL.

I was particularly surprised that it beat the Fig. 8 which had sat at the top of the list until I put it up against the Lyndy.

The testing protocol is not formalised yet, but at the moment I have to say congratulations for bringing us a simple, easy to tie, adjustable double loop knot that seems to be stronger than the Figure 8.

Well done.

Derek

cuesta

  • Jr. Member
  • **
  • Posts: 19
Re: Knot I have made requiring identification
« Reply #19 on: September 12, 2007, 12:58:59 AM »
Derek,

Thanks for your explainations regarding your findings in regard to the variants of the Lyndy knot. Your descriptions regarding knot dynamics have provided me with a much deeper appreciation of knotting in general. I have found your observations quite informative and surprising, I am surprised that it has brought about such interest and also pleasantly surprised that your comparative tests have shown the knot has such strength. Derek, you mentioned that you have a protocol in place for testing knot strength, When you formalise it, I would be interested to know more. Thanks for sharing this information with me and I'm glad that it has tweaked your interest.

Regards

Oscar

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 3956
Re: Knot I have made requiring identification
« Reply #20 on: September 12, 2007, 06:36:05 PM »
four other loop knots at the moment -- Overhand Loopknot, Constrictor Loopknot, Bowline and Fig 8.
And once again we get to wonder what exact knots these names denote.
(I think that C.L. has an Ashley number, #1045.)  Given that, as noted previously,
the Lyndy LK is a reverse Fig.9, that it might have some strong test results is to
be expected.  And then, the geometries at test time.
And the break points (stitch thread through the cord to calibrate it, if risk of
affecting strength w/Sharpie marker is to be avoided)!?

 :)

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1530
  • Knot Botherer
    • ALbion Alliance
Re: Knot I have made requiring identification
« Reply #21 on: September 13, 2007, 03:41:54 PM »
four other loop knots at the moment -- Overhand Loopknot, Constrictor Loopknot, Bowline and Fig 8.
And once again we get to wonder what exact knots these names denote.
(I think that C.L. has an Ashley number, #1045.)
 :)

Well said Dan, I think you have voiced the outstanding issue behind all testing to date vis -- what does the name denote.

Re the Constrictor Loop.  I do not believe the variant I use is depicted or described in Ashley.  The nearest is #409, the Poacher knot which is based on a single strangle or the #1120, the Scaffold which is based on the double strangle.  In the test I used the #1120 variant, but tied iwith the end slipped and not the line slipped as #409 and #1120 are.  The end is then finished with a gooseneck around the working line.

I prefer this loopknot to the Bwl because it is easily adjustable but once set it is self holding.  I have yet to see it fail with the ease of the Bwl. but this cursory testing does indicate that it could be weaker.  On balance, I still prefer a weaker knot that stays closed to a stronger one which is prone to undo itself.
« Last Edit: September 15, 2007, 08:11:44 AM by DerekSmith »

cuesta

  • Jr. Member
  • **
  • Posts: 19
Re: Knot I have made requiring identification
« Reply #22 on: September 14, 2007, 02:19:13 PM »
Derek,

I had a look at your post regarding your rig to deconstruct knots after the results you mentioned. Could you post what specifications you would test for. I am looking at what fields one could put into a database to record such findings.

By the way , I am going to post another knot which could be useful and looks good as well. I will start a new post for that one following this entry

Regards

Oscar

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1530
  • Knot Botherer
    • ALbion Alliance
Re: Knot I have made requiring identification
« Reply #23 on: September 15, 2007, 08:31:49 AM »
Hi Oscar,

I am about to post an outline protocol which hopefully will attract debate and go through the process of peer improvement.

I will then attempt to implement it and post the results to see if the protocol should be further refined.

Hopefully, by then we will have established if the Guild will agree to fund and supply reference test cord.

The results will feature diagrams and photographic images which will pose a challenge for your database.  Ideally, the database should be accessible via web pages -- is this what you had in mind or were you planning a private database?

I look forward to your input and your new knot.
Derek

cuesta

  • Jr. Member
  • **
  • Posts: 19
Re: Knot I have made requiring identification
« Reply #24 on: September 16, 2007, 02:28:38 PM »
Derek,

What i propose to do is put together a SQL server database to link up and reference knot encyclopedias, journals and other collections into a central server which is accessible on the web.
What i have gathered about the knotting world is that it is made up of collections and disparate references that contain knots. It also appears that there is a lack of concensus regarding what knot names mean. There does not appear to be a central resource to collate and reference known knots which can list a knot, related knots, existing provenance, nomenclature , articles, abok or other reference materials, pictures and links on the web. My day time occupation involves the planning and design of databases and I can put together without too much effort the structure that we would need to start such a repository that users can contribute to. I suppose you could look at knots in a similar way to the nameing and classification of organisms. Genus phylum , choradata etc. Obviously I don't propose to know how to classify knots , however, others with more of an idea should contribute their thoughts regarding this. All I can do is make relationships work on a database.

I have already started to lay out the tables and fields needed in the database to establish nice straight forward relationships. That is why I would like to look at your list of parameters to include it into the structure.

I am building this on a SQL server I have at my disposal. When i get to the point of being able to share the data, I will either set up a server myself or if the IGKT has a hosting ISP, then it can be set up there. I would then create the scripts and ftp them to the site.
Anyway, thats the jist of what i intend to do. I think this would be quite a large undertaking and I will need quite a bit of input and guidance on what would be acceptable and necessary to make it successful. The real challenge will be getting the data into the database, but it's early days yet. I am sure that others have thought of putting together something like this and I would be interested to know what they have done, where they have come unstuck and if any datasets currently exist to import them into a common datasource.

Regards,

Oscar


P.S
I think discussion regarding a database should commence on another posting ....




DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1530
  • Knot Botherer
    • ALbion Alliance
Re: Knot I have made requiring identification
« Reply #25 on: September 16, 2007, 04:12:04 PM »
Wow,

I am so glad you came across this forum, there are many opportunities ahead of you -- I hope your love of knots can tie you to the tasks.

I agree, a new thread is wholly appropriate, I look forward to posting on it.

In the meantime, have you come across the igkt.pbwiki.com site.  A few of us have been working on it to collect together knotting information.  It is a database of sorts in so far as the whole wiki can be searched and anything can be linked as required -- it is a very ad hoc database without any formal structure.  I would appreciate your thoughts on it.

Derek