Jordan Series Drogue in Dyneema

[Benz]

Originally Posted by Breaking Waves

a couple of people asked me questions about the above - thought it worthwhile to post a short summary.

1. Why would you feel the need to 'test' the soft shackle as they are quite a proven solution?

Soft shackles are commonly used and tested in a two-point load situation. I have never seen one tested in a 3 point load situation. My best guess is that it will be fine. But it is striking how often we discover something a bit unexpected when we test dyneema in a new application. There are often a few known unknowns but it is the unknown unknowns that frequently produce surprises. So, in critical applications, it is almost always worthwhile to test.

In this particular case, in static testing - it would be worth confirming that 'hoop loading' does not cause a problem, that having one of the thimbles bare on either the loop or the noose does not cause a problem (it does not in 2 point loading, that is well known, but the strain angle will be different in 3 point loading), and personally, I am skeptical of the use of the wire tie to close the loop (I think it will either slip or break or if neither of those it could create a stress riser) so that is worth testing. In dynamic testing, one would want to know about wear/chafe from the thimbles, the behavior of the loop under cyclic loading (personally I don't think this is any sort of issue with a properly sized loop, we essentially see no accidental openings in 2 point loading applications) is worth looking at. And then in both static and dynamic cases - there is always the possibility of learning about some currently unknown unknowns. I guess it will be all ok, but on the other hand we might learn something new.

2. What is the 'more obvious textile way' (than the soft shackle) you refer to?

A multiple loop, with end-to-end bury, with a Dyneema chafe cover. Benz has talked about these in other threads. This is the highest strength (almost twice the strength of the soft shackle using the same amount of dyneema) most proven approach. It still is textile, and so is potentially vulnerable to some things, but eliminates the weaknesses of the loop/knot. However, it is permanent, I don't think that should be much of a problem unless you are planning to regularly remove your bridle legs from the rode (if for example, you might be planning to also use those logs for some other purpose).

3. You say this would not be your personal solution - what would be?

I like the solution Saltylass and Fixxty have shown. It is quite elegant - light, compact, no metal to clang around, strong, and well-proven. There is a small and well-defined weakness in that bridle connection (strop join), but that is very easy to compensate for with just one size bigger line at that point.

I have mentioned before if you want to do this in the most 'proven industrial' way, then using a delta plate is the solution. That unfortunately means a pretty large chunk of metal. If you want to spend a little money, you can get an engineer to FEA the delta plate and a machine shop to remove quite a bit of less stressed metal, and you can do it in Ti (I have personally done this for two high spec applications) or high tensile aluminum, all of which will make it elegant and lighter. This is perhaps the strongest and least vulnerable of any solution (which is why it is standard for heavy lifting) . . but it is still a decently sized chunk of metal. For most of our 'yachting applications,' I think the zero metal strop is probably preferred - easier to handle and certainly strong enough.

4. Care to comment on the quotes about knots and 1:1 bends?

The engineering is really quite well established and there really is zero controversy about this (among those who understand the engineering).

One needs to understand that there are two different cases here, and people sometimes get confused between them. Case one is a bend in a rope with two loaded/working ends - for example, a rope over a sheave. Case two is a bend inside a closed-loop (like a spliced loop).

In case one, the various bend ratio charts (like the HAMPIDJAN chart John F refers to) apply directly. They show that a 1:1 bend will weaken the rope locally by about 50%. That is clear and unambiguous. And we know from pull testing that knots will weaken Dyneema by 50-65% (depends on the knot and the particular flavor of Dyneema). That is also clear and unambiguous. Given that, the advice to avoid knots and 1:1 bends in a case one application is understandable because you are giving away a lot of strength. However, I will comment that the strength loss is well understood and if you have an application where you absolutely need a knot or a 1:1 then it is quite simple to adjust the rope strength to compensate - so it is just simply an engineering problem and not something evil.

In Case 2, a closed-loop (like a spliced loop), the local strength within the loop is 200% (roughly, a bit of complexity for a few percent about throat angle and hoop loading). That is well proven and understood (I will note that John F originally did NOT understand this until Evans and Brion explained it to him, and Evans then did 3 series of tests to show that it was true and that his three theoretical counter explanations did not in fact apply). So, in this case, if you have a 1:1 bend inside the loop, it is a 50% reduction of a local 200% strength, which gives you in the end 100% strength (it is a bit less than this, like 95% typically because of the marginal complexities). This has been proven and tested very extensively (if you are curious there is the old SA thread where Evans did multiple test runs to demonstrate it to John F.). So, in case 2 applications 1:1 bends are rather more acceptable (just upsize the rope by a few percent if you really need totally 100% spec strength) - generally, it is accepted that 1.5:1 bends are totally accepted here (there are industrial rope fittings that use 1.5:1 bends). This is true no matter someone's hand waving about Dyneema compressive strength or scaling or dynamic action. I personally would still rather avoid a 'knot' inside a loop, because almost always there is a solution to avoid that ugliness. A strop hitch however does not have the great weaknesses of a normal 'knot' - it is a bit worse than a straight 1:1 bend ratio (it is a strength loss of around 15%, depending on a number of factor), and most commercial riggers will recommend you just do a simple interlinked 1:1 splice to splice if you can, but the strop is quite acceptable and well understood and you can easily upside the rope to compensate for its loss if it is the best solution for the application (it does prevent some wear/chafe that directly interlinked splices may experience). So, in case 2 bend ratios higher that 1.5:1 don't, in fact, add much strength, and if the throat angle is high (which it has historically been in some thimbles which have been focused exclusively on pushing bend ratios up) then it can in the worse case in fact be weaker. I will comment that there is potential long-term value in thimbles reducing wear at the contact point, however, Dyneema is incredibly wear-resistant and this is not a huge issue for most of our yachting applications (it is a bigger issue in some commercial applications).

This is all pretty well-proven engineering and not controversial. But sometimes people get confused between case 1 and case 2 applications.

This is gold. Thanks for taking the time to write it all out. And for the time to test it all.

[Breaking Waves]

Originally Posted by wsmurdoch

The climber boys have been testing webbing strops linked together, and in that case the knot efficiency is not 100%. They show failures in the knot in both dynamic and in static testing. Now, webbing is not rope, but...

https://www.blackdiamondequipment.co...ings-together/.

I think we have discussed this test here before (some years ago).

Importantly this actually represents a 'case 3' - "bends and knots within a sling" - which is different than case 2 (splices end loop) because in a sling the whole system is baseline 200%, while in our case 2 the working part of the rope and thus the system strength is 100%.

Note that the strength %'s are the percent of rated sling strength and NOT the percent of the webbing strength. So they are starting with a baseline strength somewhere around twice the webbing strength. So, in order to be comparable to our case 2 numbers, you would about double the break percentages.

Unfortunately, a weakness in this particular test in my mind is that they used rated sling strength rather than actual sling strength. the slings are actually stronger than rating, so the breaking percentages are actually a bit overstated (eg the denominator should be larger).

finally, it is ofc webbing, which has edge issues with knots, which rope does not have, and has somewhat different bend behavior (because it is wide and flat you don't get the same stress risers on the outside of a bend).

[Seaworthy Lass]

Originally Posted by wsmurdoch

A couple of years ago I posted this after pulling a bit of spectra with my genoa winches. The knot I called a square knot is the way I saw riggers at work link slings together. We seem to now be calling it a strop hitch.

https://www.cruisersforum.com/forums...ml#post2883224

Although my Jordan drogue is is nylon, I have been following the UHWPE Jordan threads and keeping my eyes open.

The climber boys have been testing webbing strops linked together, and in that case the knot efficiency is not 100%. They show failures in the knot in both dynamic and in static testing. Now, webbing is not rope, but...

https://www.blackdiamondequipment.co...ings-together/

...gives some inkling of what might be needed for thorough testing.

Of course, we have the case 1 and case 2 examples.

Hi Bill
I wish to thank you yet again for proposing the method I eventually ended up adopting.

For all those who were not following that lengthy thread, I posted this mid 2019 (Note: I was incorrectly referring to a strop hitch as a cow hitch ):

Originally Posted by Seaworthy Lass

Hmmmmmmm.

That is the the sound of me thinking.

I think this is the same as the “cow hitch of the leader onto the bridle legs option” that Jedi suggested a while back, but rather than dressing it as a cow hitch it is left undressed like Dockhead’s “interconnecting loops” at the start of this discussion.

Reading the first paper Thinwater posted a link to a few posts back, I think this is called a Strop hitch when it is done with just two loops.

Of all the cow hitch options this is the one I like best as it is not so convoluted, for want of a better word. The line looks less tortured and what I have referred to as “nasty” looking.

This is a simple option and may in fact work the best.

My mind is truly spinning at the moment. I need to make a decision with what to do at the Y junction. Juho’s triple doubled loop eye splice would need testing before it could be used and despite the good bend ratio the compression would be sky high. I don’t like conventional cow hitches, but I think hitching on the leader is far preferable to hitching on the bridle legs, as is apparently currently done (Breaking Waves and Dockhead). I vastly prefer no knot but don’t know the impact of leaving the bridle leg under load free to move around. I don’t really like the interconnecting loop option as I don’t think it solves much and just complicates the plain eye splice option.

The only thing I am certain of is I don’t like the way the bridle legs are currently cow hitched onto the leader. Also, whatever method is used, the three eye splices need to be long to minimise tearing forces at the throat.

This is giving me a headache. I will mull over this a few days, but your suggestion, which as far as I can see is really a undressed version of a cow hitch of the leader onto the bridle legs, may have the least potential problems of all the above alternatives (for all the reasons you suggest) and is only a modification of the current method rather than going off on a new tangent. That appeals.

To everyone contributing here: thanks for all the thought put into this and the input. Having lots of ideas tossed around is extremely beneficial. It is by far the most effective way of arriving most quickly at the best solution in the absence of any data or additional testing.

SWL

PS The titanium ring is interesting. It has a MBS of 5098 kg-f though, which is unfortunately low.

Originally Posted by Seaworthy Lass

Another big bonus of WSmurdoch’s proposal is that there are no throat issues at all. People constructing a series drogue and using this form of cow hitch are not going to get themselves into serious trouble as they could with cow hitching the two bridle legs onto the leader and making the eye length of the leader too short.

I have just mocked it up. Another bonus is that the bridle legs end up on top of each other, not side by side relative to the leader. The 4 portions of the bridle eyes stack neatly in a 4 x 4 arrangement. This means they are permanently central relative to the apex of the leader, so zero shifting is required as load transfers from one leg to the other.

It helps to mull over it a while, but so far I think it is is looking more and more favourable. I think it is worth setting it up properly with 8 mm loops as I did with the alternatively dressed version of this a few days ago to see what happens when some decent load is applied using our winches.

SWL

Originally Posted by Seaworthy Lass

After ten days to think about it, the above solution that WSMurdoch proposed is simply looking better and better to me. I feel comfortable with it in this application. I need to stress though that it is untested. I have never seen a bridle connected this way.

SWL

PS Bill …. thanks for persisting with thinking about solutions. I knew if we put our thinking caps on, something good would come out of it .

I have now had a lot longer than “10 days to think about it” and I still think this is by far the best solution, and one I adopted, albeit untested. It deviates least from the method currently most commonly in use yet at the same time eliminates some of the drawbacks. Its simplicity appeals to me tremendously.

Bill, I could not find this knot described anywhere else. Referring to it as a “square knot” seemed incorrect, as there are no working ends, so I named it a “modified strop hitch”. It is tied almost the same way as the strop hitch that is commonly used to connect different sections in the tail of these drogues.

SWL

[Benz]

Have I missed something, or is it possible to eliminate some of the metalwork at the 3-way junction by splicing the two legs of the bridle directly to a SS ring? If they never need to separate, this eliminates two thimbles. I've used some big rings from a sailmaker's supply with a bar welded across them to splice bridle legs to. The crossbar keeps the legs from slipping around too far (not that it matters: everything aligns, on a ring, when pulled on).
That way only the long leg needs a thimble, and a soft shackle between thimble and ring would be pulled exactly how it's intended to.
If it were me, I'd use a "dogbone" loop (hate that name, but it's trying to stick in the industry) to attach the long leg to the bridle.
I've tried to attach a photo here of a mooring bridle I made by this method...hope it loads.

[Breaking Waves]

Originally Posted by Benz

splicing the two legs of the bridle directly to a SS ring?

we have talked a bit about rings (I remember DockH posting a picture of one also). Like with delta plates (rings w various crossbar configurations are essentially delta plates with a bunch of metal machined away), you need a pretty good chunk of metal for theoretical max series drogue loads (plus a safety factor).

There may be better examples for our case - but just from my bookmarks: Crosby rings

[oldbilbo]

The 'modified strop bend' may well be the best solution, and the cheapest. That would avoid any micro-aggressions about the merits of a rubber O-ring over a nylon cable-tie. I have lotsa both, and either would likely do the job.... Or both....

Here is a pic of a 'mock up' showing a MSB dressed as seems best. Note the tubular antichafe sleeve - which MAY prove beneficial - and the paired notional bridle legs.








It is appreciated that those with great big heavy boats such as the Bestevaer49 are committed to great big heavy rodes, bridles and warps - just as Don Jordan 40 years ago with his nylon warps - which justifies dividing the rode into cascading, thinner and easier-manhandled sections.

Those with rather lighter boats - e.g. myself - have no good reason to divide their rode into Leader and Section(s), thus introducing unnecessary cascading splices.

It is worth bearing in mind we are 'two peoples divided by a common language' and that it is far too easy to presume that we all understand the words we use in quite the same way. Thus we have ASTM, ANSI, ASME and ISO - all working to the same purpose, but with different units and definitions, which is why we had a Mars Lander capsule plough into the landscape at great speed as some engineer didn't quite understand the difference between metres/sec and feet/sec.

[Breaking Waves]

Originally Posted by oldbilbo

Note the tubular antichafe sleeve - which MAY prove beneficial

those do not look like dyneema sleeves - am i right?

if you use Dyneema sleeves, they are more chafe resistant (than nylon or polyester).

The cover does in fact get you a (slightly) better bend ratio.

If you tuck/bury the ends of the sleeves there is a theoretical potential local strength increase (it has to be done very smoothly/neatly milked onto the 'core' to get the strength increase), without the need for cascades. It is hard, with high modulus fibers to get the core and cover evenly balanced so they both contribute to initial static strength. I have seen a guy at the main test lab I use accomplish it, but I think it will be near impossible for most DIYers. But the Dyneema cover 'should' increase resilience even if not initial static strength.

Not a big deal but just seems proper to me to use Dyneema sleeves given all the effort you are putting in.

[Benz]

Originally Posted by Breaking Waves

we have talked a bit about rings (I remember DockH posting a picture of one also). Like with delta plates (rings w various crossbar configurations are essentially delta plates with a bunch of metal machined away), you need a pretty good chunk of metal for theoretical max series drogue loads (plus a safety factor).

There may be better examples for our case - but just from my bookmarks: Crosby rings

Perhaps there's a market here for a titanium ring made with 5/8" or so stock (to give the line a better than 1:1 bend radius), whether pear-shaped or not. I like the crossbar because it's like studded chain--it prevents the initial distortion that leads to failure.

[Breaking Waves]

BTW, on the questions of 'local strengthening' - we do have good data and test results showing that inserting Dyneema cord (into the hollow center of the main line) will strengthen it (significantly) even in a bend (note: the tested bends here were greater than 1:1, so might want to reexamine it to confirm with smaller bends).

This was demonstrated with the test series's (by Brion & Evans*) which developed the 'stronger than buried button' soft shackle designs - in which one design aspect was adding an internal insert into the soft shackle noose (since that is the failure point), and this did in fact greatly strength the loop area. (* note: SWL later/independently proposed the same idea).

Originally Posted by Benz

Perhaps there's a market here for a titanium ring made with 5/8".

It would be a nice piece. I somehow doubt the potential market is very large, as generally, the commercial world seems happy with the steel solutions. I know you can 3d print Ti (not DIY have to send away) but I don't know much about the quality of that process. Ti is pretty sensitive to small defects. But if that worked you could 3d print them, one at a time, or in very small batches as needed.

[fxykty]

Originally Posted by Breaking Waves

BTW, on the questions of 'local strengthening' - we do have good data and test results showing that inserting Dyneema cord (into the hollow center of the main line) will strengthen it (significantly) even in a bend (note: the tested bends here were greater than 1:1, so might want to reexamine it to confirm with smaller bends).

This was demonstrated with the test series's (by Brion & Evans*) which developed the 'stronger than buried button' soft shackle designs - in which one design aspect was adding an internal insert into the soft shackle noose (since that is the failure point), and this did in fact greatly strength the loop area. (* note: SWL later/independently proposed the same idea).
...

What proportion of size of the line for the insert? If I have a 10mm line, would a 3mm insert work? Or 5mm? And I assume the insert would need to be tapered at both ends?

[Breaking Waves]

Originally Posted by fxykty

What proportion of size of the line for the insert? If I have a 10mm line, would a 3mm insert work? Or 5mm? And I assume the insert would need to be tapered at both ends?

yes, nicely tapered.

I personally would do 5mm in your case. With the strop hitch you only need to make up (roughly) 15% to get back to 100% system, and 5mm will do that with the least risk of distorting the braid in the main line.

There is an alternate theory, which both Brion and SWL's (independently) espoused - that you should use the same size insert as the line (so 10mm in your case) - and that does work when you bury the tails in the stronger soft shackle. You need to do very nice tapers and very clean insert to not distort the braid - it can be done by a DIYer, just need to work careful.

We tested a range, all showed significant improvement over no insert, but not enough statistical significance to tell between them - unfortunately, we were mixing my shackles and Brion's which added more noise to the test than I expected. I should really circle back and answer the question of optimal size, just has not been high on my list.

practical note with this technique - you want quite long splices, long enough for like >30x diameters of insert buried. They need enough bury to carry some load, not just bulk it up.

[Benz]

Originally Posted by Breaking Waves

BTW, on the questions of 'local strengthening' - we do have good data and test results showing that inserting Dyneema cord (into the hollow center of the main line) will strengthen it (significantly) even in a bend (note: the tested bends here were greater than 1:1, so might want to reexamine it to confirm with smaller bends).

This was demonstrated with the test series's (by Brion & Evans*) which developed the 'stronger than buried button' soft shackle designs - in which one design aspect was adding an internal insert into the soft shackle noose (since that is the failure point), and this did in fact greatly strength the loop area. (* note: SWL later/independently proposed the same idea).



It would be a nice piece. I somehow doubt the potential market is very large, as generally, the commercial world seems happy with the steel solutions. I know you can 3d print Ti (not DIY have to send away) but I don't know much about the quality of that process. Ti is pretty sensitive to small defects. But if that worked you could 3d print them, one at a time, or in very small batches as needed.

There's a local guy who's a rockstar with Ti. I might see about getting something to test when I see him.

[Breaking Waves]

Originally Posted by Benz

There's a local guy who's a rockstar with Ti. I might see about getting something to test when I see him.

Ah, nice.

I used to know some guys at GE aircraft engines who could make me Ti pieces, but they have all retired now.

[Seaworthy Lass]

Originally Posted by Breaking Waves

….. You need to do very nice tapers and very clean insert to not distort the braid - it can be done by a DIYer, just need to work careful….

I have taken a series of photos showing how I achieve a “nice taper”.

Technique:

- Mark the length to be tapered and the halfway point clearly with a marker.
6 strands will be tapered over the first half and 6 over the second. I find it much easier to do half at a time.
Mark 3 sets of paired strands in the lower half (every second or third pair to achieve a reasonably even distribution).

- Pull out the three marked pairs, fan them out evenly and flatten them with a fingernail to spread the fibres out a little.

- Cut them evenly as one unit between the two marks.

- Splay out the next 6 and fan them out evenly as with the first lot.

- Again, cut them evenly as one unit on a long diagonal. I don’t bother much about trying to keep the weave intact in this portion, as it tends to become unbraided of its own accord.

The last image shows how the taper looks when smoothed.
It is easy to get a near perfect result.

SWL

[fxykty]

Originally Posted by Breaking Waves

yes, nicely tapered.

I personally would do 5mm in your case. With the strop hitch you only need to make up (roughly) 15% to get back to 100% system, and 5mm will do that with the least risk of distorting the braid in the main line.

There is an alternate theory, which both Brion and SWL's (independently) espoused - that you should use the same size insert as the line (so 10mm in your case) - and that does work when you bury the tails in the stronger soft shackle. You need to do very nice tapers and very clean insert to not distort the braid - it can be done by a DIYer, just need to work careful.

We tested a range, all showed significant improvement over no insert, but not enough statistical significance to tell between them - unfortunately, we were mixing my shackles and Brion's which added more noise to the test than I expected. I should really circle back and answer the question of optimal size, just has not been high on my list.

practical note with this technique - you want quite long splices, long enough for like >30x diameters of insert buried. They need enough bury to carry some load, not just bulk it up.

Next question - how long the taper?

Let’s say I’ve got a 10mm main line and I insert a 5mm line. If the insert needs to be 30x, then I need at least 150mm of 5mm. I would insert so it is centred on the loop.

Now, how long should each taper be? And is the length used for taper in addition to the 30x insert length, or is it part of the 30x? For my 30x long insert do I need 150mm + 2x taper?

For eye splices I have been tapering over the last 1/3 of the bury, so for a 60x bury with a Brummel lock I use the last 20x of the bury for the taper (with the bury being the same diameter of the line of course). But I think I’d lose too much line to the taper for a relatively short insert.