The Offshore Voyaging Reference Site

Retrieval Problems With Unsheathed Dyneema (Spectra) Series Drogues

Update February 2020

This is now a solved problem. See this chapter.

This is an article I hate having to write:

  • It will provide a big headache to a large number of our readers, and two manufacturers I value and respect.
  • I’m sure most of you are thoroughly sick and tired of reading about series drogue details.

But, like the article we published about a dangerous dragging modality with Rocna anchors, I have to write it because it deals with an important safety issue and, in addition, this article corrects a mistake I made.

Here’s the problem:

It will be extremely difficult to retrieve a Series Drogue built to Don Jordan’s design if it is made from unsheathed Dyneema™ or Spectra™.

Randall Reeves’ experience shows there is no proven and reliable method to attach the retrieval (nipper) line to Dyneema. And Trevor’s retrieval method, the only one we know of that is tried and proven under real conditions, and workable for a single-hander, requires that (as does ours).

When this problem first came up, over a year ago, I suggested splicing loops into the drogue to attach the nipper line to. And, based on discussions I had with Andy Schell a couple of weeks ago, I thought this was a simple and relatively easy solution.

I was wrong. Trevor Robertson, who is a clever and deeply experienced splicer, has shown, after hours of actual experimentation, that there is no way to make said splice securely. Trevor’s conclusion:

Splices used in Dyneema seem to rely on both legs of the line being nearly equally loaded for the splice to lock tight. Only one leg of this splice is loaded, consequently much of its strength depends on the seizings/whippings. I don’t like this at all.

I have every faith in Trevor’s testing and conclusion but, having made one mistake already on this issue, my paranoia is running high, so I called my friend and all-things-rigging guru Jay Maloney, who confirmed that there’s simply no way to make a secure splice in braid rope unless both legs are loaded.

Bottom line, if neither Trevor nor Jay can do it, it ain’t happening.

So where does that leave us?

The Answer

For those who have not yet bought a series drogue but want the added strength and decreased weight of Dyneema, Trevor, Jay (Jay and I came up with the idea separately), and I all believe that the best answer is to use a rope like New England Endura Braid that has a Dyneema core and Dacron sheath.


And yesterday I tested that as follows:

  • 1/2″ (12mm) Yale Endura Braid to simulate the main line.
  • 1/2″ Dacron as the nipper line.
  • Attached the nipper with a double rolling hitch.
  • Loaded to the full pressure I could put on a double grip handle in the highest ratio on our three speed #65 sheet winch.
  • This is more load than we needed during our retrieval tests while motoring ahead at 3 knots, so a good simulation of real world retrieval loads.
  • I estimate at least 3000 lbs based on sheet loadings on the boat and the power ratio of the winch.


  • The nipper did not slip at all.
  • There was a little bunching of the sheath, but not bad (see photos).
  • Said bunching was easy to milk out.
  • Jay Maloney says he’s confident that the rolling hitch won’t damage the Dyneema core.


There’s a lot to like about using Dacron-sheathed Dyneema:

  1. When retrieving, we are doing something that has been tried and proven for years: attaching a nipper with a double rolling hitch. (I do this all the time for all sorts of things, many of them under high load, and I’m sure many of you do, too.)
  2. Still going to be lighter than Nylon since less water absorption and smaller diameter for the same strength. For example, a boat that requires 3/4″ (18 mm) Nylon double braid can get slightly better strength with 1/2″ (12mm) Dyneema with a Dacron sheath. Not a trivial benefit, particularly on larger boats.
  3. If used for the bridles, may reduce whipping around since the sheath will absorb water.
  4. The sheath is not structural (confirmed by Jay Maloney), so it can act as a built-in chafe protector, for things like joining cow hitches, without the issues of bunching of chafe gear that Randall noted.
  5. The nipper can be attached anywhere convenient.
  6. For retrieval techniques that require taking the drogue line itself to a winch, like ours and Hal Roth’s, the much less slippery Dacron sheath will result in far fewer turns on the drum and that, in turn, will lead to far fewer cone fouling issues.
  7. Is the simplest solution—see below for other options—and that almost always means the least chance of unintended consequences.


Of course, like anything around boats, there are downsides to this option:

  • More expensive than single-braid Dyneema.
  • Will be harder to work the cone attachments through, but I can’t see any harder than with Nylon double braid. (May be easier than Nylon since the core will be more slippery and probably looser braid.) See photo.
  • Heavier, particularly when wet, than single-braid Dyneema.

Trevor’s Take

Here’s Trevor again after I emailed him my testing results and conclusions:

My immediate reaction is to agree with all you have said about using sheathed line. It seems to address all the issues with acceptable downsides. That worries me a bit – singing from the same hymn book is fine in choirs but does little to solve problems. There should be something that niggles me or needs further investigation but I cannot see it now. I’ll think more on it and email you immediately if I come up with anything, but at present I can see no reason other than [cost] to prevent me changing over to Dyneema with a polyester sheath.

To date, I have not heard from Trevor with any “niggles”, but I really like that he is thinking about it. This is not about me being right (that ship has sailed), it’s about getting it right.

Don’t Forget Nylon

All that said about Dyneema, as commenter Brent pointed out, there’s also nothing wrong with sticking with good old proven Nylon, other than the weight and increased vulnerability to chafe.

Other Options

But what about those of you who have already bought series drogues made from single-braid Dyneema? I wish I had a good proven answer, but I don’t. That said, there are a few possibilities:


Climber and AAC commenter Alan Pottasch’s suggestion of using a climber’s ascender is probably the most readily available alternative for attaching a nipper line to unsheathed Dyneema, but it’s not without worries.

Ascenders are:

  • Not designed for this use since climbers don’t use Dyneema rope.
  • Not designed for these kinds of loads.
  • If it breaks we are then screwed with no way to fix the situation.
  • I can see a sudden surge on my boat loading it to say 5000 pounds, the break strength of the strongest ascender I could find, and many ascenders have break loads under 1000 pounds.
  • And even if it holds, I simply don’t like the idea of using gear that close to its break load.
  • Without deep destruction testing we can’t know whether or not using an ascender will weaken the series drogue.

Bottom line we are doing something totally new with a piece of gear not generally used for the task, so testing is required before I would rely on it. Trevor agrees:

This system has promise but I would like to see it tested more rigorously with a strain gauge in the system, the whole loaded to perhaps 20% of the vessel’s displacement and then the warp tested to breaking to see how much using the ascender has weakened it. If possible a range of ascender types should be tested.

Trevor also followed up the next day with:

The more I consider the use of a climber’s ascender to retrieve a Dyneema JSD the less I like it for exactly the reasons you give. I certainly would not recommend using an ascender for retrieval based only on a primitive cockpit test. My feeling is that this idea should be shelved until there is some solid work done by someone with access to a proper load test facility. It is very likely that such a test will find an ascender is inadequate for the loads met retrieving a drogue and it is unsuitable for that role (for which it was never intended).

Sounds like a job for Drew Frye over at Practical Sailor…I’m so subtle. But he’s not a fan of the idea:

I’ve broken several name brand accenders used in test rigs; you are at VERY high risk of ruining them over ~ 1200 pounds. Rope damage is also very likely, or I should say, probable. So make sure there is a load-limiting mechanism. Anything over 600 pounds is abuse.

There are belaying devices better suited for this, but nothing is designed for use with Dyneema, and they are optimized for specific rope types and diameters.

Modified Clutch

Richard Elder suggested a modified clutch, which might go a long way to solve the worries with an ascender since it is a device designed for the task but, unfortunately, actually building one would be a complex undertaking.

Also, as far as I know, clutches are generally not used on unsheathed loose plait Dyneema—yes, racers use unsheathed Dyneema but they usually accomplish this be stripping the sheath except in the areas where a clutch will be used.

That said, Richard’s suggestion might work, but it is not readily available and would also require extensive testing.

Other Gripper Knots

Is there a gripper (constrictor knot) that will reliably hold on bare Spectra? I think the answer is no, but I could be wrong, so if anyone wants to test that option, which should be easy using the same set up I tested double braid with, here’s a great starting point over at Practical Sailor.

Do take note of the problems that PC testers had remembering the more complex hitches—might be a deal breaker, particularly after a bad thrashing from a storm.

Also note that their criticism of the rolling hitch assumes a single. In their detailed testing Drew confirmed that a double holds very well , as my testing confirmed (on Dacron sheath).

Fatten The Line

Just before we published this article Drew came up with this idea:

what about inserting tapered line inside (a short section of core) to make fat spots? It should not weaken the rope and would be quite easy to do. A few stitches would lock it in place. You would have to test the holding power of a prusic, but it may be one of the very few ways to grab Dyneema 12-plait without damage. It would be smooth and not damage the rope.

I like this one a lot, because it’s simple. But like all the other’s, it will require testing.

Update 21st July

Trevor has tested this and it has fundamental problems that make it unlikely to work.


Our recommendation is not to buy or make a series drogue from unsheathed Dyneema, at least until one of the above alternative solutions has been adequately tested, preferably in several real world heavy wind and sea retrievals offshore.

I will leave the last word to Trevor:

That leaves polyester-sheathed Dyneema and Nylon double braid as the options.

Inline Feedbacks
View all comments
Drew Frye

The breaking strength of ascenders is typically 4.5-6 KN (1000-1300 pounds) depending on the rope size. The safe working load is 300-500 pounds. You can see this stamped on Petzel ascenders, for example, and I have tested this. It is NOT 5000 pounds. Ascenders are NOT rated for climbing falls, only rope climbing.

This time I’ll include a link.

There are stronger ascenders, up to 3500 pounds, but serious cover damamge is still likely at 1000 pounds. The cover of 1/2″ Endura Braid, for example, is probably only 4000 pounds or so ( I do not know–probably worth testing). Finally, not all ascenders will work with 1/2″ rope; like pulleys, when near the end of the stated range, they often get clunky. Few climbing ropes exceed 11 mm, so most are optimized for 9-11 mm.

I like the double rolling hitch idea. I might try prusiks, but I think they may be more prone to sticking at high load.

Yannick Piart

Hi all,

For those of us already equipped with a unsheathed dyneema JSD, what about sewing non slippery material around the core at the strategic distance that separates your retrieval winch from the pulley probably located somewhere in the front part of the boat?
With a distinctive color to help spotting it when it comes out.
Just an instinctive suggestion after reading the development of this important matter.



Rick Hearn

In reference to the ‘fat spot’ idea, I’m not clear on the step after making a ‘fat spot’. Would the hitch then be placed not on the fat spot but aft of the spot?

Edward Scharf

As I was reading the article I wondered about using a cable grip such as they use to pull power lines or tree removal ropes. I found this one. But think there may be others. Just an idea. May not work.

Stein Varjord

Hi John,
Important topic treated perfectly. Very interesting and it inspires creativity. Great!
I don’t have a JSD, but since I’m such a weight fanatic, the idea of not being able to use unsheathed Dyneema feels as painful as if I owned one. 🙂

I love the idea of fattening the rope to give the rolling hitch something to stop on. It’s cheap, seems very easy to do and I see no way it could create related problems. The only possible issue I see is that the hitch might still slide past the bump. I think it wouldn’t, though. Such fat spots could be put many places to give a lot of options. Perhaps such a fat spot could be at every cone, as part of the cone attachment procedure?

Another option I think of seems less perfect but still maybe possible: A perhaps 60cm / two foot (?) piece of sheath from a strong material around the JSD line, seized/whipped/sewn onto it at the aft end and at the forward end having a small loop for attaching the nipping line. The tension between the loop and the whipping will make the sheath stretch and compress hard around the JSD line, which should hold securely with a very large contact area. The Ronstan Constrictor clutches use this principle for high load lines to reduce rope wear and slippage.

I think this could work, but with some weaknesses compared to the rope fattening method. The sheaths need to be long enough to get sufficient grip. This might interfere with the cones. It has to be installed when the JSD is made, as the sheath pieces can’t pass over the cones. The loops on the sheath pieces need to be done right. I don’t know the details. One can’t realistically have as many nipping line attachment options.

Before this article, I thought spliced on loops was the way to go. Now I realise that’s not working. I accept that fully sheathed Dyneema is good and will work, but the fattened rope seems very attractive. It must be tested, of course, but I believe and really hope it holds the hitch well!

Alastair Currie

This method is used to haul wire drilling rope around a derrick’s block set and onto the drawworks (big winch). In drilling they are called rope snakes. An example can be seen by Googling LSG Snake Grips (I have no association with Lewis Oilfield). The snakes are slipped over both the new and old wire rope that is still on the winch. When the line tensions the snake grips both the new and old line allowing the new line to be reeved around the blocks and onto the winch.

We all know the trick of using a small section of polypropolen line pushed open and slid over a finger, where it become impossible to pull off.

I think there is milage in your idea and there is certainly real world experience of pulling lines this way in many industries, not just oil and gas drilling. Also Google SUPER ROD SUPER GRIP CABLE CONNECTORS

Stein Varjord

Hi John,

I agree that sheathed Dyneema is the right choice at the moment. I think that one could solve the mentioned weaknesses of other ideas to make un-sheathed Dynema work. However, potentially ingenuous solutions will probably simultaneously reduce the weight, bulk and especially the price benefits of un-sheathed Dyneema. There’s no point in making the effort if there’s little or no reward.

The Constrictor idea I mentioned above does need some attachment at one end to make it “constrict”, but it isn’t loaded at all. In the Constrictor clutches they use a shock cord for this. Probably the best in this case too. It only pulls the constrictor tight to start the friction. I’m convinced that this would hold well and be reliable, but it might need more space than available between cones and probably will cost a bit. I also have mixed feelings about having moving parts on the JSD.

Trevors comment about the “fattened rope” method makes sense. The method clearly needs development or is a fail. One thinkable solution: The insert could have two fattenings separated by perhaps 10 cm. The rolling hitch would be placed between the bumps. The aft bump would then be in the loaded part of the JSD rope, be fairly long and not too thick, and should have enough friction. The second bump should be bigger and will stop the hitch from sliding. The two parts of the insert must be connected by a part that is strong enough for the load. I suspect that this method will also challenge the cost benefit of un-sheathed Dyneema, unless the insert can be part of the standard cone assembly.

Anyway, these thoughts are mostly mentioned for the record. They would need testing. I think sheathed Dyneema is a perfectly good solution. If it ain’t broken, don’t fix it. 🙂

Richard Elder

Hi John
Great summary and analysis.

Trevor Robertson

Hi John and Drew
I don’t think the ‘fat spot’ idea is going to work. As I understand it a short piece of Dyneema is inserted into the main rode, the end of which forms a ledge in the outer of the warp that prevents the rolling hitch sliding along the slippery Dyneema. The problem is that the rolling hitch has to be BEHIND the insert and as strain comes on the hitch it bunches the outer up so it is loose and lets the fat spot insert slide along inside the warp. I tried variations of a fat spot using a tapered insert, an untapered one and one with a lump melted into the loaded end of the insert. None worked. The one with the blob melted in the end came closest to being effective but put an unacceptable hard lump in the main warp.

So far the best method of attaching a retrieval line using an insert that I have tried is to insert the tail of a Brummel-locked eye spice into the main warp and hold it in place with a sewn seizing/whipping. This relies entirely on the seizing for strength but at least the outer is not being bunched up and positively encouraged to move with respect to the insert. As this attachment is only used during retrieval when the loads are relatively low, and then only for a short period, it may be good enough for the job, but I don’t like it much.

Drew Frye

Thanks for testing that. I was just spitballing. Makes sense though. Dyneema is weird stuff.

For a new JSD one could splice on sections of cover, but again, I would be concerned it might slip. That said, this is done under jammers all the time.

Another option I just thought of might be a rope coating, such as Rope Dip (Flexidel) that stiffens the rope and increases friction. I’ve tested these a good deal on polyester and nylon, where they are quite impressive, but never on Dyneema. But for certain, there are thick antichafe coatings for Dyneema that do NOT weaken the line and are intended for very heavy use (mooring cables). It might be a viable option for retrofitting. Just a 1-foot area every few cones.

Drew Frye

OK, it would be good to know, if not for this, for some other application. Gimme a week, I’m traveling.

Bertil Klevner

Reading all about retrieving the JD and pros and cons for different ways.
While harbourbond I could not help but trying a “collapsed rolling hitch”.
I have not given it the 2ton test, but it looks promising.
Start making the rolling hitch with your pulling line. At least three to four roundturns. Finish with two half hitches or maybe even with a constrictor.
Then the tricky part: collapse your pulling line on the unloaded part of the dyneema so the dyneema wraps around the pulling line, (start from the unloaded direction of the dyneema) – – Then tighten up! The dyneema will tighten around the pulling line and can not slide past the clove hitch.
Can easily be undone when not under load.
How do I submit photos?
Bertil /sy SOL

Brent Cameron

One of the things I like best about this site is John’s willingness to admit he got it wrong sometimes and issue revisions and mea culpas on a very timely basis when necessary. This makes me trust the advice on this site that much more. I will stay tuned on this issue and see if sheathed Dynema works well in the real world before switching from Nylon. I am not fussed on the idea of adding speed bumps to the line as it will introduce new things to the equation that could only ever be fully tested over decades in the field. As for the existing nylon rode, if I ever needed to retrieve it, I would be using the big Lewmar #58 Electric Winches so weight isn’t that big of an issue for me anyway as much as love Dynema for other purposes.

Alastair Currie

I just see this information and idea progression, as a natural evolution of discussions and testing, not a right or wrong decision. I have said this before, it’s why I subscribe to this resource, I trust the facilitation by yourself to get the answers.

Alastair Currie

Apologies, I wasn’t meaning yourself facilitating a forum style discussion (nor do I want that at AAC). I meant how you go about arriving at a conclusion and offering advice. The facilitation comment was in the context of your interaction with those who provide information in your research to come to that conclusion i.e. engaging with those who have real life sailing experience or deep technical experience of the matter to hand.

Ben Logsdon

The length of the line could be broken up into sections with a spliced eye on each end such that they would connect eye to eye. The eyes would provide the necessary attachment point for the retrieval line while maintaining tension on both legs. Of course, the eyes would need to be made with the appropriate hardware. I think the sheathed line is the most elegant improvement over nylon suggested thus far.

James Peto

I hesitate to ask but would you please show a diagram of how you rig the whole unit as I can only see that firstly trying to reach over the stern to attach the retrieval as nearly impossible and then one has to keep doing this.
I have probably missed something!!

Richard Ritchie

John, this trail shows great example of collaborative problem solving, being open to even wild suggestions and input from diverse skills, background and talents, then testing them. Thank you.
May I return to the proposal of Dacron sheathed dyneema as the current preferred option. Given the inherent slippery-ness of dyneema core, it seems all the load will be taken on the Dacron sheath. Thus this has to have an enormous breaking strength. Or am I reading this wrong? This will be partially offset by extra friction from the grip forces, but upstream the sheath is being opened up so all forces seem to become tension on the sheath downstream. Or can the combined brains show me where I am wrong?

Rob Gill

A really valuable discussion John and others,
Our planned retrieval process uses a powered Lewmar 58 sheet winch, though I remain concerned about feeding the ACE cones around the winch on the front Dyneema section, as I expect we will need 5 turns to safely handle the load. If some Dacron cones rip in the recovery, so be it – we can probably repair them. I am considering ordering a few spare ones following this discussion.
I am also planning a test retrieval under different engine loads. We did a trial recovery originally before our Cat 1 certification, but not under any significant load. I will report results back with video footage if possible, though it won’t happen soon – it’s winter here!
Meanwhile, does anyone have real-life experience of directly retrieving a single plait Dyneema JSD, using a powered or manual sheet winch please? How did the cones fair? Nylon or Dacron?
Thanks, Rob

David Huck

Hi John, I built a JSD some years ago in Australia, and could not find a local supplier of the double braid nylon than was originally specified, but found a rope maker who made it for me. My understanding was the stretch characteristics of the nylon was critical to the performance of the JSD. I’m surprised to see JSD’s being made with spectra and other low stretch fibres given this, and could not find much discussion about the impact, other than on bulk, weight and cost. Maybe I’ve missed it, but interested in what engineering and testing has been done on the various alternatives to the original spec for the rode.

Bertil Klevner

Hej John!
You are absolutely right about the problem to get the knot to collapse when the drougeline is loaded.
A constrictor as stopknot may help to get that first inch in that gives enough slack to make the rolling part curl over. Some distance between the rolling part and stopknot also seems to help in the hitch transforming.
Of course, I have yet not tried this with heavy loads and under severe conditions.
Thanks for Yours and all fellow sailors efforts to make the ocean safer!
SOL / Bertil

Eric Klem

Hi John,

I am glad you are trying to chase this problem down, while it doesn’t directly affect me at this point in time, it is definitely important to people sailing offshore.

It seems to me that the base assumption that we are operating on is that a winch is the appropriate tool for retrieval. Your method uses the winch for its mechanical advantage while Trevor’s method just uses its ratcheting function. The other major difference is whether the drogue line is attached directly to the winch or via a secondary line. To me, this raises a few important questions which I have not seen answered here:
-With Trevor’s retrieve the slack method, why is the secondary nipper line necessary? I realize that he had problems with lying beam on when going direct to the winch but why couldn’t the same fairlead have been used or one set up with a decent lead? Is there a safety issue with handling cones or a concern about overrides on the winch?
-Have you or anyone you know done a test to try to prove whether or not uncovered dyneema in a JSD works on a winch at the loads associated with retrieval when conditions are still boistrous? Capstan effect is incredible and a non-linear effect so it might be doable (Stein’s idea with the cover relies on the same basic principle where the friction is not linear with distance but exponential). I could also see issues with too many turns meaning that cones foul.
-Does the solution change with boat size? Factors could be energy required to retrieve, retrieval forces that may mean different ways of gripping the drogue apply to different boats, etc.

For non-single handers, if you can use the drogue directly on a winch, the challenge just becomes getting everything to the winch. If leading the drogue line direct, the insertion of a single extra connection point in the drogue a little ways past the bridle connection should allow Trevor’s method to be used to get the main line on a winch if the line is too slippery for a rolling hitch. For single handers, the issue with a winch appears to be that you don’t have enough hands provided you are actually using the winching function. With a drill or winch handle, 1 person is tied up operating the drill/handle while the other tails. This would suggest that if the winch is in fact the solution, a powered winch with a control that can be operated on a rolling deck with something other than hands is necessary and there are tons of controls of this type out there from mouth to foot but not necessarily in ruggedized form. There is also the question of whether a suitable winch exists and I have not looked at the available options enough to know.

Trevor’s method of only retrieving slack is intriguing and if you could actually count on this always working and not needing to power things, it opens up a bunch of possibilities. For the handy people, it would become possible to build a reel for the drogue that had a couple of releasable pawls and have it be sturdy enough but still super simple whereas a powered one would be a real engineering project to build not dissimilar to building a powered winch. I am thinking like the old school manual anchor windlasses minus the teeter totter but with a wide, small diameter drum. You could use a retrieval line from where the bridles connect and just wind everything onto the drum. Of course, you would then be stuck unhooking the boat end of the bridle legs once you had retrieved enough as it needs to go around the drum too which means you would then be stuck repacking to a bag for an easy deployment. Even better would be to build it strong enough that you never unhook from it but you wouldn’t have the better steering effect of the bridle (another interesting question here which is, is a bridle truly necessary and have we overcomplicated life and should we just have a single attachment point solving another big issue). Probably too much stuff and too much hassle but it is interesting to think about.


Damian Heaney

HI John,
Could you just terminate at each required (loop) distance with a eye splice and attach the next section with another eye splice, so essentially making short sections which would then have a loop at each correct position?

Craig Burnside

Just had little play which seemed promising…
So first attempt-12mm d12 dynemma, made a short length of 5mm d12 with a soft eye one end, chocked the eye around the 12mm then fed the bare end into the 12mm away from where the cones would be, towards the boat if deployed. Then ran it inside the 12mm for about 300mm, out the side, circled the 5mm bare end around the 12mm a couple of times with a few half hitches along the way and back inside the 12mm back towards the cones. Pulling on this with a short bit of 5mm chocked round the 12mm worked. Though the very first loop had no tension on it so tried again, 5mm fed inside-out the side with a couple of wraps and half hitches then back inside. Seems to work with a few hundred kg, I’ll leave it overnight. I really need to get that camera on the phone working 😉

Craig Burnside

Maybe this shows it a little better >

Craig Burnside

Some more pics – one of the 5mm tails stuck out the side, just cos it was too long. There didn’t seem to be much tension down there so hopefully should be no different with both tails buried completely.
Close up of the 5mm chocked round the 12mm
Choke undone. the fluffy looking bits are the 5mm insert, like that before.

Knot undone

No visible damage to the 12mm afterwards, probably not far off 500Kg test load.

So hopeful this end anyway, pretty sure that will work for my setup, tidy with nothing to catch and cheap!

Stein Varjord

Hi Craig,

This method seems interesting. I don’t think I get the details of the first part of it right, but my understanding is that you attach a Dyneema rope fattening with an eye splice that goes around the JSD line, then it goes inside the JSD for 30cm (1 foot), then out again with some half hitches to form the thick part which the nipper line would hang on.

If that’s about right, it seems to be how to make an improved version of what I suggested earlier and which John named a barbell in the rope. Your method doesn’t depend on a lashing or any such weaker part. The insert is working inside the tensioned part of the JSD rope, giving good friction. The half hitches should give plenty of hold for the nipper line, with a rolling hitch or such.

This seems highly worthy of further testing! I’m not able to do any such testing for the time being, but when people are back from holidays, I guess there will be some creativity going on.

Terence Thatcher

I have a JSD made by Dave Pelissier of Ace Sailmakers. The bridle legs are dyneema, the drogue itself is nylon. I carry a “lazy” line to the attachment point to bring in the bridle. I would be interested to know if you have talked to Dave about the ease or difficulty of attaching the cones to sheathed dyneema. It seems a great approach, as long as the sailmaker can get through the sheath to make the attachments. If I go offshore again, Dave and I have discussed replacing my drogue with dyneema to reduce weight. I would certainly not do it with unsheathed dyneema now. Thanks, as usual, for all the great work.

Terence Thatcher

oh, one other thing. Have you tried the icicle hitch on the dyneema? I haven’t, but I know it holds better than a rolling hitch.

Andrew Climie

Two things:
Jumars and other ascenders have metal teeth and can damage climbing/caving ropes…after all that effort building the series rogue!
In my limited experience as rigger/climber/sailor, a nipper line significantly thinner than the main line will grip (bite) much better than a line of similar diameter. Could your nipper line have a tail of say 8mm spectra? For your 12 mm dyneema drogue.

Drew Frye

It took a while, but I finally did the pull tests using a prusik hitch that we were talking about. Part of the delay was waiting for a coating, that I though might improve friction, to dry completely.

On 5/16″ Amsteel (Dyneema single braid). This was lightly used line, so the texture was realistic. The prusik was tied using a sewn spectra climbing sling.

Bare. Slips at 5-8 pounds. I can do better with my hand…. but not much.
Yale Maxijacket, dried for 3 weeks. Slips at 30-40 pounds. Better, but hardly useful.
Tapered core inserted. Holding >1000 pounds is practical. However… the core will try to slide up inside the cover stating at ~ 150-400 pounds, depending on the rate of taper. Additionally, the Amsteel under the prusik will be flattened and distorted (I have no idea if this weakens it–maybe not). Finally, the prusik can be difficult to untie over about 500 pounds. It will come out, but it wouldn’t be fun on a bouncing boat.

What I did NOT test was pulling a prusik up behind a cone. When it slips it will eventually come to the cone webbing insert point. I do not believe a prusik can slide past that, not with the splicing and knots. In fact, the key might be to rethink how the webbing is inserted so that THIS becomes the prusik attachment spot. Any irregularity in the Dyneema, including chafed spots, greatly increases holding. Odd that no one brought this up.

On the other hand, polyester covered Dyneema would be more straight forward. But I don’t have a sample.

Prusiks generally hold ~ 1200 pounds on 3/8″ nylon DB and a bit more on polyester.
The cover-only of covered 1/4″ Dyneema (5/16″ total) is probably about 1500 pounds. Loading it much over 500 pounds is not a long-term strategy. There will be some friction with the core, but I wouldn’t bet on it. It will shift. There are things that can be done to stop this (such as in jammers), like Spinlock RP25 coating or stitching, but I’d hate to mess with the strength an fatigue properties of the line core. I don’t know that there has been any testing of that sort.

So my guess is that prusiks on Endurabraid or similar would work fine, so long as the load is not over about 500 pounds, which seems reasonable even in a seaway, if you wait until the wind is down. I’ve recovered lots of drogues with the engine in reverse slow; when doing repetitive testing, human fatigue is a problem even in moderate weather. But you really need to watch it and turn it off for the last bit, by which time you don’t really need it. In fact, a few times I got drogues tangled under the rudder without using the engine; the boat slid down a wave, toward the drogue, just as I was gathering it in. Untangling that mess in waves was not the most fun I’ve ever had. One was a chute, with all those strings. The next trial I brought it over the side!

But I wonder if this is a non-problem. A prusik can’t slide past the cones, can it? Does anyone know?

I would NOT use Jumars. I’ve ruined ropes that way. This is well known. They are only for body weight, IMO. At least if a prusik slips, you have not damaged anything, other than perhaps a cone or two.

Stein Varjord

Hi Drew,
Thanks for the great analysis. I can’t improve on anything, but just wanted to defend my honour a slight bit. 🙂 I did actually mention that the knot stopping insert should probably be part of the cone attachment procedure. How that would be done, however, I don’t know.

I do think that relying on the cone webbings alone, or more specifically, that they are fed through the spectra, would not alone be enough. I imagine that it would not normally spread the load well onto enough spectra fibres. I do however feel that there should be ways to get that done in a better way. Perhaps if the webbings are attached by a longer insert, which would be in the loaded part of the spectra, and the prusik knot then stops on the webbing but the load is carried by the insert….?

Steve Broom

John, I am NOT an experienced ocean sailor, yet I am doing it again. I am posting. So, take my question/suggesting for what it is worth coming from a novice of significant proportions.

Retrieving a series drogue. The drogue is designed to have a weight attached to the end of it so that it is held down in the water.

What if the weight attached was a fisherman’s lead weighted line that is about 100-200 feet longer than the drogue itself? It could be tied off on a regular cleat on the boat, and obviously be tied to end of the drogue. I imagine this lead weighted line would be flaked into the water first and then the drogue to follow so that it would not have a chance to get twisted in with the weighted line. After the drogue was flaked out and supplying the drag, the fisherman’s lead weighted line would simply and gently hang well below the drogue line in the shape of a ‘U’. It would never be under tension.

Retrieval would simply be grabbing the lead weighted line and pulling, turning the drogue around, making the retrieval and exercise of relative ease.

The predetermined weight that is attached to the end of the drogue is easily translated into a fisherman’s lead weighted line based on length required. (That was a clumsy sentence, but you get what I mean.) These lead lines come in varying diameters and weight per foot as you most certainly know.

What thinkest thou? Do I earn bonus stars or am I relegated to the corner with a dunce cap?

Stein Varjord

Hi Steve,
I’m not John, but I have a feeling he might have the same reasons for doubting your idea as I have: The idea is messing with the fundamental design principle of the JSD. That’s risky, since even small changes always have important consequences.

In this case, I see two problems, but there might be more:
1. An extra weighted line will probably most of the time hang under the JSD, but “most of the time” isn’t good enough. If hanging on the JSD for a couple of days, a pretty large number of waves will pass. Each of them being slightly different. Each of them testing a new way to get the two lines fouled.
2. A lead loaded line would need to be perhaps 200 meters. That’s a very long line with a lot of drag. It’s far less than the actual JSD, but since the weighted line is supposed to pull down on the actual end of the JSD with authoritative force, I find it hard to trust that the weighted line will manage that. Its drag will most likely make it move fairly close the the JSD and one would perhaps need significant extra weight at the end.

I think the conclusion seems to be that the idea is tempting, but that the added risk of fouling is too big, the deployment might be more demanding and the retrieval might not be noticeably easier.

Steve Broom

Thanks Stein and John! I truly appreciate the voices of experience on this site. Thank you for taking the time to reply.

Peter Adamson


Have you tried a third bridle for retrieval? I was thinking of tying a line to the splice at the bridle joint. This “recovery bridle” would be slacked while the JSD is deployed (which makes a knot OK since this 3rd line will have no tension until retrieval). When it comes to retrieval time, this bridle would be used to winch in the JSD on the primary winch until the bridle joint is in the cockpit. Then the bridles are disconnected from the shackles. One bridle is wrapped around the secondary winch and tensioned (too take load off recovery bridle) and the other bridle brought to the forward block then back to primary winch. Then the recovery bridle is slacked (all load now on 1st bridle on secondary winch) then the 2nd bridle is swapped with the recovery on the primary winch and winched in until the JSD bridle joint reaches the forward block. The recovery bridle is removed from the joint and used to hold tension on the JSD (w/ secondary winch) while JSD is wrapped around primary winch. Then either primary winch is used to pull the rest of the JSD in directly (assuming the chutes don’t get tangled on the winch hub) or the nipline-rolling hitch shuffle is done with the forward block. I would also add a 3-roller rolling fairlead on the transom to make things a little easier. This main everything happens within the lifelines, the retrieval is central so no change in boat direction.

Note: I haven’t purchased a JSD yet, but will soon and am reading your articles as a basis for how I set things up.

– Peter

Peter Adamson

Wow. I thought I read that article, clearly not. Although I am encouraged that I can up with the same idea. The inability for an electric witch to handle the job is informative. I will have to look into a Milwaukee I guess.



Peter Adamson

Yes I saw an Ad for one a while back. I was concerned that I couldn’t find any torque numbers on their site. With a cordless tool, the torque specs are always right there. I meant to follow up with the eWincher folks. I guess will put that on my to do list (with the other 10,000 things for the boat).

Andrew Ruff

What about a simple dyneema tuck eye splice, for example’s sake let’s say 1 meter total length,  where the tuck’s tail end comes out of the end of the sheath.  Then simply feed the drogue up through the center of the tuck and bring it out through both lines just before the eye.  Sort of a line (drogue) within a line (tucked tail) within a line (sheath of the eye splice line).  That way the Chinese finger trap forces are essentially being applied by two lines when you attach your recovery line to the eye and put it under load. 

Obviously you would need to remove the cones to put the eye spliced recovery points further down the drogue, but for someone building their drogue it would be a fairly simple process. Care would have to be taken when attaching the cones through one of these triple thick dyneema line sections to ensure the cone’s attachment webbing goes through the center drogue line.

Thoughts? What am I overlooking… besides how hard I think it would be to initially feed the drogue line through the recovery sections?