Author Topic: Accurate loop in small kernmantle  (Read 9316 times)

peterbmack

  • New Member
  • *
  • Posts: 4
Accurate loop in small kernmantle
« on: April 19, 2008, 01:58:58 PM »
I need to tie a loop in both the end and the middle of some 3mm (1/8") spectra line to form braces for light (very light!) aircraft landing gear. The loops need to be accurately sized to slide onto 25mm (1") tube and also located accurately along the length of the line. Once tied they will never need to be untied. A small, neat knot is definitely preferable.

Several knots have been candidates, but so far the most likely appears to me to be the Fisherman's loop (or whatever pseudonym you prefer) at the end (I can tie the first overhand at an accurate location then the second after passing the loop around an appropriate diameter former, after snugging it up it should be reasonably accurate and repeatable).

Possibly the Fishermans could also be used in the middle, although there could be a small amount of movement and hence wear if the running leg gets loaded more than the standing.

I have two questions- does anyone have suggestions for alternative knots, and does anyone know of the approximate strength reduction of the Fishermans loop?

Thanks

Fairlead

  • Sr. Member
  • *****
  • Posts: 333
  • IGKT Member since 1984 - IGKT Librarian
Re: Accurate loop in small kernmantle
« Reply #1 on: April 20, 2008, 05:35:25 PM »
Hi Peter,
A few questions first -
1.  Must you use Spectra kernmantle?
2.  Is the Spectra core braided or strands?
3.  What knotting books do you have for reference?
4.  Will the end or middle knot be subjected to 'shock loading' (ie on landing)
5.  Where do you live/fly from?

You may not want to answer the last question on this forum so my email address is
IGKTLibrary@aol.com  -

Knots are known to weaken Spectra/Dyneema by as much as 70%, much greater than Polyamide or Polyester for example.  Thus splices and sew & seize eyes are generally a better option seeing as you will find it almost impossible to splice 3mm kernmantle.
As a pilot myself I would like to help and look forward to discussing this with you

Gordon

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1571
  • Knot Botherer
    • ALbion Alliance
Re: Accurate loop in small kernmantle
« Reply #2 on: April 20, 2008, 08:28:59 PM »
Hi Peter,

What an interesting challenge.

Spectra is utilised extensively in stunt kite lines because of (and in spite of) its special characteristics.  It is light and strong so very strong lines can have a small cross section to minimise wind drag, it has very little stretch so the control lines do not loose intimate and immediate contact with the kite, and it is super slick against itself, so literally hundreds of turns can be made in the flight lines and yet the braids flow over one another with undetectable friction even when loaded with several tens of pounds of kite pull.  Little surprise then that you have chosen Spectra as part of your loading structure, but is it the right choice ?

Each of those advantageous characteristics should be considered from the perspective of when they could in fact be a liability.

Light(thin) and strong -- the negative side of this is that masses of force is concentrated into a very narrow contact line.  Many a carbon fibre kite spar has been 'chopped' through by the point loading of a thin Spectra line  --  can your 1" tube take this 'point' pressure?

Little stretch  --  the negative side of this is that there is no advantageous elasticity to absorb shock loads -- will your landing gear be exposed to bumps and would a more relatively elastic brace line of say Polyester be more robust in landing and takeoff situations?

Super Slick  --  the knot tyers nightmare.  In testing knots to destruction in 100lb Spectra, at least half of the knots I have examined simply flow through themselves.  Knots which show the slightest sign of 'cogging' in their structures cannot hold themselves against the cord flowing through and out.  The knots that do hold exhibit extreme 'nips' to ramp up the friction and this effectively acts to scissor through the Spectra, weakening it and destroying its strength advantage.

Of course, Spectra is not super slick against all materials and sheathing Spectra with a braid which has a half decent grip is the principle route to achieving strong fixings in kiting.  Polyprop/Polyester 'grips' Spectra quite well and a kid flying a holiday kite on 50pence polyprop lines will have every stunt flyer running for cover in case the kid crosses lines with the Spectra, because the polyprop will 'grip' the Spectra and the high friction will rapidly melt straight through ?20 worth of stunt line.  So this is perhaps one trick you can take from the stunt kite fraternity- pull a polyprop/polyester braid sheath over the spectra before attempting to knot it.  Of course, the sheath has the added function of spreading the load from the Spectra a little before it bites into the tubes.

When it comes to knotting the spectra, one of the key lessons would be straight out of ABoK and that is "There is unsuspected virtue in a few turns of line".

This is especially true for Spectra, because it allows the loaded line to dump (transfer) some of its load into the spar (tube) before you try to make an end knot.  This way, the end knot does not have the full load on it attempting to make the spectra slip through itself.

So, if you stick with the Spectra, how would I fix it?

First I would slide a length of braid into the centre of the cord, sufficient to cover three turns around the central fixing spar and then I would fix this in place with a Constrictor (ABoK #1249).  The constrictor is self trapping, it incorporates two turns around the tube to shed load and the entry and exit points are inline with virtually no strength loosing nip points.  The ends of the covering braid will be fluffed by the wind buffetting, so you might consider a light whipping or a short length of elastic tube to prevent them from opening right up and becoming untidy.

As for the end loops.  Again, I would slide over ca 15" of polyester braid and start by taking three turns around the tube, then take the working end back under the three turns so that it is trapped under their tension and comes out alongside the SP going into the turns (opposite direction of course).  Take the end around the tube then fix it to the SP using a double Fishermans knot.  You will have to work the tension into the three turns and the double fishermans very carefully because this knot will grip itself as you make it.  You might have to pretension the tubes inwards as you make the knots then release the tension to tighten up the whole assembly.  Again, consider securing the ends of the braid against the wind.  The purpose of the three turns is to shed the tension load into the tube before the fixing knot so that minimum force is left to pull the Spectra through the knot.  In kiting, the ends are often further secured by melting the Spectra and the braid into a blob, tying an overhand knot and locking the blob and the OH stopper with superglue.

Alternatively use the same strength Polyester, use the same knots but forget the braid sheath, forget the pulling through, but remember you might need a lot more pretensioning because polyester really is very elastic.

Good luck

Derek




Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4188
Re: Accurate loop in small kernmantle
« Reply #3 on: April 21, 2008, 04:05:45 AM »
I need to tie a loop in both the end and the middle of some 3mm (1/8") spectra line to form braces for light (very light!) aircraft landing gear. The loops need to be accurately sized to slide onto 25mm (1") tube and also located accurately along the length of the line. Once tied they will never need to be untied. A small, neat knot is definitely preferable.
And where does the other (unmentioned) end go?  I.e., I don't understand what sort
of cordage structure you are trying to implement.  (You mention ONE end, & middle; that leaves an end.)

Why do you need "loops"?  --as opposed to hitches?  (--or perhaps even a bend, making a sling!?)

Quote
Possibly the Fishermans could also be used in the middle, although there could be a small
 amount of movement and hence wear if the running leg gets loaded more than the standing.
This implies that you are in fact loading both ends, making what sort of angle to the center knot?

Quote
I have two questions- does anyone have suggestions for alternative knots,
 and does anyone know of the approximate strength reduction of the Fishermans loop?

And your question is incomplete/mistaken:  the strength of Spectra line in a Fish.LK
is what is relevant here.  I'll be pretty safe in guessing that it's under 40%--maybe well less.

We can, with better information about the intended structure.
(I doubt I'll be recommending the Constrictor qua hitch, though!)

--dl*
====

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1571
  • Knot Botherer
    • ALbion Alliance
Re: Accurate loop in small kernmantle
« Reply #4 on: April 21, 2008, 07:35:06 AM »
Hi Peter,

Strange the way certain projects can fixate the attention.  I find myself returning to you landing gear rig over and over.

I have realised that I have made my usual mistake of jumping to conclusions.  In my minds eye I have a picture of your landing gear rig and this has coloured my thoughts about how I would set the assembly for maximum strength.  But in reality, I have no idea of the angular geometry you are planning so I have no idea if any of the approach I have taken above is in any way relevant.

So could you give us some idea of the physical layout of this rig.  I am plagued with questions like - 'Why do you have a central fixing?' and 'can the tubes take torsion?'.

A little more detail might lead to a 'better' structure.

Derek

peterbmack

  • New Member
  • *
  • Posts: 4
Re: Accurate loop in small kernmantle
« Reply #5 on: April 24, 2008, 01:17:16 AM »
After a problem logging in to the forum, I'm glad to be able to say thanks for your replies and offer more information.

Below is a picture (finally worked out how to get it to show inline) of the rigging. I've only included one leg and set of rope for clarity, although the complete structure has the same thing duplicated both sides. I think this should answer most of Dan's questions and also Dereks request for a layout.



As far as loops vs hitches vs bends, we prefer to make a batch of the rigging lines on a jig then assemble them onto frames as we build them rather than have to try to tie the knots on the frame and get everything the right length. Apart from that it doesn't really matter how the knots are done as long as they meet the criteria of being small, neat, reasonably secure (we can always add some heatshrink or some such if needed) and reasonably strong.

Gordons' questions I answer below:
1.  Must you use Spectra kernmantle?
I suppose not, but I want a low stretch, strong thin rope. Spectra seemed well suited and readily available and inexpensive.

2.  Is the Spectra core braided or strands?
The core is non-braided strands, the sheath polypropylene.

3.  What knotting books do you have for reference?
Various, including ABOK.

4.  Will the end or middle knot be subjected to 'shock loading' (ie on landing)
Not really. It will get a certain shock load, but the landing legs have nylon skids at their ends that act as a minimalist suspension, absorbing a bit of the shock.

Gordon also asked in an email, "What stops the knot from sliding along the spars?" - the answer to that is we have an M5 socket head cap screw (allen key bolt) through the tubes, the rope stops against the head and is retained there by a loop of webbing fixed under the bolt head.

Hollow braid vectran has been suggested, I'm looking into that but initially it looks like an expensive option.

I should mention that in the past the same rigging setup has been used with bowlines and I think an overhand knot at the middle, the knots passed through a 3/16" eyebolt to retain it on the tube. This worked OK in terms of strength, but the rope may have been a bit stronger than the Southern Ropes Dyneema Lite we are using now (175kg bs). Although that seemed to work, I'd like to ensure we have the best safety margin possible with the materials we are using.

For a bit more info on the application, look at http://www.airtimeproducts.com.au/atpexplorerld.html

Thanks again for all your suggestions. So far I am looking at adding turns around the tubes and re-considering stitching.
Peter
« Last Edit: April 28, 2008, 03:07:30 AM by peterbmack »

Fairlead

  • Sr. Member
  • *****
  • Posts: 333
  • IGKT Member since 1984 - IGKT Librarian
Re: Accurate loop in small kernmantle
« Reply #6 on: April 24, 2008, 07:54:36 AM »
Hi Peter,
Thanks for all the information and glad to see you have been 'allowed' back!

My choice of termination if you are sticking with the Spectra kernmantle would be to use an 'Alpine Butterfly knot' in the centre and to 'Sew and Seize (Flat or Racking Seizing) eyes in the ends.  Making eyes like this will certainly be stronger than using knots in Spectra.

You think Vectran is expensive - you should try PBO!

Gordon

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1571
  • Knot Botherer
    • ALbion Alliance
Re: Accurate loop in small kernmantle
« Reply #7 on: April 25, 2008, 12:01:28 PM »

The core is non-braided strands, the sheath polypropylene.

Peter

Well that dispenses with the need to pre sheath in order to tie knots, but raises the question of have you adequately sealed the core to the sheath.  The core is non stretch while the sheath is stretch, there is a tendency for the sheath to pull off of the core unless the ends are well joined.  Fuse the core and sheath by melting, then  seal behind the 'blob' with Superglue to make the sheath and core rigid and so restrict the tendency for the core to pull back into the sheath.

Strike the Constrictor for the middle spar.  I had imagined it as being central and so requiring the two arms to be 180 degrees from one another.  The Constrictor supplied this configuration with the leads running straight from the load into a round turn.  But in your diagram, the two load legs are not far off from being parallel to one another and leaving the spar at a fairly acute angle (30 degrees ??)

You also stated that the knots are held in place by an M5 Allen key bolt and a loop of webbing fixed under the bolt head.

The Constrictor would still be a strong fixing, provided it was tied 'at the back' of the spar, but it would fail your requirements because you could not pre-tie it, as without the spar it simply falls apart.  However, as you have the webbing holding the fixing in place, why not simply go for a round turn with the two ends and the webbing and the bolt head at the 'back' side of the spar.  The two load legs then would run out of the webbing and pass either side of the middle spar on their route backwards to the two skid legs.

However, you will probably discount this simplest fixing on the grounds that if either skid fixing fails, then the cable will simply pull through the webbing allowing the second skid to also fail.  Assuming that you want each skid to hold independently, then we need a knot at this point which is strong enough to take the double loading from two skids, yet hold if one skid line breaks.

Ideally the two load lines should enter the knot in parallel and proceed, without any sharp bends or constrictions, around the back of the spar in order to transfer their load into the spar.

Such a knot can be constructed fairly easily as a variant of the doubled overhand noose (doubled ABoK #8).



Middle the cord, then fold back a loop and tie the OH noose as in ABoK #8, but the working end will be a small loop.

Pass the noose loops through the small WE loop and fold the small WE loop back down behind and over the OH knot, then allow the OH knot to 'undo' by letting the now captured end loop pass back into the OH knot.  Tighten and dress the knot to remove excess crossings and twists and set the knot to give the correct loop diameter for the spar.  I do not know if this knot has a name, I remember it by calling it the Ying Yang noose.  If you look at the knot from the opposite side of the two SP's, then you will see a nicely balanced Ying Yang shape to the holding cords.  Although this is a noose, once set either of the SP's can be loaded without the other and the knot will still hold.  However, it still has one disadvantage - because both SP lines take the same direction around the spar, they torsion the spar in one direction and they have a 'cogging' effect on one another.

This can be overcome by having the two loaded lines pass in opposite directions around the spar.

Using the Knot Diagram Utility FCB41, https://knotcyphers.pbwiki.com/The+FCB+Cypher to draw the above diagram, you can then use it to change the direction of the second cord around the spar so they rotate in opposite directions like this:-



I haven't found an 'easy' way to tie this knot yet,  but again you can use the Knot Diagram tool to help.

Middle the cord and lay out the red shaded part of the knot.  Then take each cord and thread it through in the paths of the black and blue shaded loops respectively



As for the skid end loops, sew and seize seems the ideal route, but as I have seen many seized eyes simply pull through and making a sew and seize eye in cord this thin can be very demanding, I think perhaps a more doable route might be to use a tied knot - but which knot?  I will give it more thought.

Derek

peterbmack

  • New Member
  • *
  • Posts: 4
Re: Accurate loop in small kernmantle
« Reply #8 on: April 28, 2008, 02:25:13 AM »
Thanks very much for your time on this Derek.

I like the "Yin Yang", particularly the opposing rotations version. I'll certainly be trying that one out. The alpine butterfly is not quite as neat in this application, with the two legs being separated.

As for the end loops, I'm inclined to use two round turns and a bowline/double bowline/double dragon loop/alpine butterfly or perhaps water bowline. I'll do some trials to see what seems to work best. One problem with the 2 round turns is that by the time I terminate 2 ends on each tube there are now 4 round turns all trying to squeeze up hard against the stopper bolt, potentially causing them to roll over each other and so pass over the bolt head. More testing.

Peter

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1571
  • Knot Botherer
    • ALbion Alliance
Re: Accurate loop in small kernmantle
« Reply #9 on: April 29, 2008, 12:03:38 AM »
Hi Peter,

All the turns do not need to be below the stopper bolt, in fact there is merit in making the first turn - i.e. the heaviest loaded turn, above the stopper bolt in order to shed load into the spar before the cord gets into a compression situation against the stopper bolt.

Take the first turn around the spar above the stopper bolt, then the second turn around the spar below the stopper, then take the working end back to the loaded line to tie off the knot.  Ideally, this knot should hold the working end in place but it should have as little impact on the 'flow' of the loaded line as possible.

One of my favourite knots for this application would be the adjustable grip hitch because it allows the loaded line to pass straight through to the spar without any tight turns.



(except in your application it would have two round turns instead of the loop)

This knot is one of the strongest I have come across in testing, beaten only by knots such as the Lindy loop knot and the Plaited Double.  It is very easy to tie and has the final advantage of being adjustable - you can slide the knot by gripping the knot and pushing it up the tensioned line, when released it stays put.

After two round turns, the load will have been significantly transfered to the spar, so the residual load will not have much lateral force on the load line, so it won't pull the loaded line much out of straight.

Derek

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1571
  • Knot Botherer
    • ALbion Alliance
Re: Accurate loop in small kernmantle
« Reply #10 on: April 29, 2008, 03:53:49 PM »
Hi Peter,

Have you seen Roo's webpage on friction law of round turns?

See http://www.geocities.com/roo_two/friction.html

Derek

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4188
Re: Accurate loop in small kernmantle
« Reply #11 on: May 02, 2008, 10:19:15 PM »
Below is a picture (finally worked out how to get it to show inline) of the rigging.
 I've only included one leg and set of rope for clarity, although the complete structure has the same thing duplicated both sides.
 I think this should answer most of Dan's questions and also Dereks request for a layout.
Thanks for that, but I'm afraid it leaves me wondering, still.
For starters, I don't understand the seeming cylindrical object amind the legs,
which looks to be wheel-like, but then in that case is oriented perpendicularly
to the legs it lies upon/amid, and for taking movement on a surface (landing?)
also at right angle to the obvious curve of the leg ends!?  In your linked site's
photos, there is no such object, and apparently the guylines of light & dark hue
can be seen in deployment.

It seems that these lines should be expected to endure some bit of "shock" loading,
though the amount of load put on them can't be all that much (for the frame, for that
matter); some amount of dampening elasticity should be helpful in this, I'd think
(and I'm not agreeing w/Derek's assertion about polyester--not in general, at
least (for I think he has evidence re his particular stuff)--, and could see nylon even).

Quote
As far as loops vs hitches vs bends, we prefer to make a batch of the rigging lines on a jig
 then assemble them onto frames as we build them rather than have to try to tie the knots on the frame
 and get everything the right length.
Apart from that it doesn't really matter how the knots are done as long as they meet the criteria of being small,
 neat, reasonably secure (we can always add some heatshrink or some such if needed) and reasonably strong.

Hmmm, it would seem to me to be problematic to have very-static line so right-sized
as to be able to be so fit (or that all necessary tolerances lead to such a plan), versus
making attachment in a way that you individually set each line-pair correctly.  Now, maybe
the lines are the final determinant of the full-open position of the landing gear, and in that
can be within some comfortably variable range of lengths such that pre-tying is practical.

One could also see using the thing & high-strength but static HMPE line with small thimbles
& eye splices (or just a "soft" eye), and tied to poles for the final few inches by some elastic
cord, achieving attachment with the more knottably happy material and gaining some bit
of shock absorption via it!?  Now, maybe this would be for the ENDS, only, and the center
(between cord-legs, i.e.) loopknot would be just that; for that, something link a well-set
Fig.9 or Fig.10 LK should work.  (I can even imagine a sort of Blood Knot in the Bight.)

Quote
2.  Is the Spectra core braided or strands?  The core is non-braided strands, the sheath polypropylene.
So, very low stretch, high strength (unknotted!).

Quote
Hollow braid vectran has been suggested, I'm looking into that but initially it looks like an expensive option.
Can't see the rationale for Vectran (tm)--less UV resistant, and if anything I think less
shock resistant.

Quote
the Southern Ropes Dyneema Lite we are using now (175kg bs).
Although that seemed to work, I'd like to ensure we have the best safety margin possible with the materials we are using.
I doubt anyone here can attest to this--you'll have to just test whatever candidates there are.
It certainly is NOT the case that general, traditional "data" about knot strengths can be relied upon!

--dl*
====

peterbmack

  • New Member
  • *
  • Posts: 4
Re: Accurate loop in small kernmantle
« Reply #12 on: May 12, 2008, 01:04:11 PM »
Thanks for your reply Dan.

The "wheel-like" thing is a drive pulley on the end of the propshaft. The legs end in nylon skids (bottom right of pic). The picture shows a bare frame in roughly the landing position, ie the leg (only one is shown for clarity) is running down to the right, the two roughly horizontal tubes with upturned ends are the main frame structure and the propshaft with its pulley is in the centre of these tubes, above the leg (it is a pusher prop, so the prop would be at the far right if shown).

The nylon skids serve to take a lot of the shock out of a rough landing.

The legs are restrained in their landing position by the ropes only. There is a certain tolerance in the lengths of the ropes, perhaps +/- 3-5mm. Pre-tying is much easier to get the legs in the right relationship to the frame, and also the two legs of the ropes at the right length in relation to each other. Tying on the frame is awkward as by the time the ropes are installed the fabric harness is installed and access is somewhat restricted.

I don't think an un-sheathed rope would be suitable as I expect it would be more susceptible to wear and individual fibres snagging in transport and also UV damage. I could be wrong on this though. Certainly being able to easily eye splice an unsheathed rope is attractive!

Our rationale for selecting Spectra was : High strength in a thin cord and relatively inexpensive. Stretch was less of an issue, except in as far as we don't want the legs to gradually creep back over time and use. A polyester might be suitable, but the diameter would increase for the same breaking strain. Given the mentions in this thread about the significant reduction in strength knots cause in Spectra, perhaps a knotted 3mm polyester might be about the same strength as knotted 3mm spectra?

Peter

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1571
  • Knot Botherer
    • ALbion Alliance
Re: Accurate loop in small kernmantle
« Reply #13 on: May 12, 2008, 03:48:00 PM »
Peter,

'Strength' is all too small a word for the reality of the complexities of this subject.

We have test rigs which can perform 'static' comparisons of failure, but when it comes down to 3mm Polyester vs 3mm Spectra composite in a highly dynamic load situation, then, although we know some of the aspects of these systems, we just do not have sufficient knowledge to model these systems and create quantifiable comparative projections.  Although in straight static load terms, the Spectra composite will sustain a far higher breaking strain than the Polyester, you may well find that in practice the Polyester could way out perform the Spectra - i.e. surviving greater impact forces without damage to either the cord, the knots or the tubular structure (in the event of failure, which should be sacrificial?).

The big issue here is the fact that Polyester is markedly more 'elastic' than the Spectra composite, and this allows it to absorb a lot of energy long before its breaking point is reached - sort of like the plastic bumpers or crumple zones designed into a car, the car deforms so easing the transfer rate of force to the passenger.

In a 'heavy' landing, you have a large vertical velocity component which must be transferred to the ground ( energy = half mass times velocity squared).  If there is no elasticity in the landing gear, then that energy is transmitted through the landing gear the instant contact is made.  As Work(energy) = Force x Distance, then for any given energy dump, if the distance is short, then the force must be high.  However, if that same load is transmitted through a more flexible landing gear, then distance is greater and consequently the forces needing to be 'handled' are proportionally lower.

Of course, your gear has at least two sets of forces to contend with.  A flyer might be coming in relatively flat but with too much forward speed, they migh touch the ground and skip back up before slowing sufficient to land.  In that momentary contact with the land, you would not want the big instantaneous 'kick' to break the landing gear, so that next touch down, there was nothing there.  Much better would be to have a more elastic system which survived the first punch yet was still intact at landing time.  This of course is why conventional landing gear has suspension in order to reduce the shock loading that is transmitted to the airframe.

Of course, there is a limit as to how elastic the suspension can be before it becomes useless, but the point remains clear - strength is not just about breakproof knots and high breaking point cord, it is about dynamic load transfer, elasticity, pull out, padding and leverage (while at the same time being practical, look nice and of course, be the right colour).

It would need to be put to the test of course, but as a generallity, I would put elasticity way ahead of brute strength every time in a dynamic, energy transfer system.

Derek

 

anything