Rolling Hitch on Snubber

[s/v Jedi]

Originally Posted by Andrew Troup

Ahem; I disagree, Nick

It's true the stopper knot sees only half the load, but not true (IMO) that this is changed to the full load when it's half way round the thimble.

(in fact, if the thimble is large, the theoretical load when it's exactly half way round will be relieved a tiny amount by the frictional transfer of load from the rope either side of the stopper to the metal of the thimble, but that's 'angels dancing on pinhead' territory.

I agree with your conclusion, that it would be preferable (probably much preferable) to move it away from the thimble, but I disagree with your stated reason.

ON EDIT: Thanks a lot for the links, and the effort you've gone to on this topic. I don't have time right now but will definitely take a thorough look, and no doubt learn a lot.

I did not test this Andrew, just repeat the experts on the topic who verified this with destructive testing.

[Andrew Troup]

<<I did not test this Andrew, just repeat the experts on the topic who verified this with destructive testing. >>

They can only verify the result, not the reasoning.

I can very well imagine the stopper would fail at a lower load when straddling the thimble, and can think of at least two reasons why, but the stated reason is a red herring which cannot explain the lower load, because the "halving of the load" tendency applies throughout the strop.

The radius of the bend causes the tensile stress in the outer fibres to increase, as it does around a sheave, but this has nothing to do with the "halving of the load": it's a function of the ratio of the radius of curvature to the diameter of the strop.

I think however that my other explanation is more likely to account for the lower failure load, but I won't inflict that on the thread.

It may be that the radius when their soft shackle was tested to destruction happened to cause it to fail at around half load, so that explanation came to mind.

If so, it's a classic example of how deviating from the scientific method can end in tears. It's a lot safer to form a hypothesis before carrying out the test, and then design the test in such a way as to try to prove the hypothesis wrong. (And be unable to do so)

It's as if I were to notice that, in the real world, 16/64 = 1/4

I might come up with an explanation that is so because you can cancel the sixes.

There are, rather surprisingly, an infinite number of examples where cancelling sixes works in practice (apparently "proving" the hypothesis correct), but it does not work in theory, and this turns out to matter.

[Dockhead]

Originally Posted by estarzinger

mmmmm . . . .does any reasonable person use Dacron for a mooring pennant? . . . . yes, essentially anyone who has studied the situation.

I am afraid you and I will not agree on this . . . that's fine with me.

You directly imply that I am not a reasonable person . . . . that's fine with me also . . . but a bit odd for a mod I think.

My apologies -- no such implication was intended, of course. My argument was badly formulated.

[wsmurdoch]

Originally Posted by s/v Jedi

Has nobody experimented with a snubber design that uses different diameter lines all spliced in parallel, with the thicker ones longer so that a thin one first stretches until the next thicker one gets tight and stretches together with the thinner one until the next thicker gets tight etc. You can splice that onto 5/8" Amsteel Blue which then functions as a stop when all the stretch is used up...

That's a good question. What spring constant do we want in a snubber? You know, how many pounds force per inch of stretch? Do we want it very low like a big rubber band or bungee cord or do we want it very high like a piece of polyester rope? Too stretchy may be bad because the bow blows off every time a gust strikes the boat presenting the side of the boat to the wind. Not stretchy enough may be bad because shock loads may break gear or tear the anchor out of the bottom. What is the right amount? What is the number? Should there just be a single spring constant, or like Jedi suggests, should the spring constant rise with increasing force? Maybe 100 lb/in for the first 100 lb of load and 500 lb/in for the 100 lb additional load between 1000 and 1100 lb and 1000 lb/in for the 100lb additional load between 2000lb and 2100lb.

[rebel heart]

I worked on a sport Fisher that used all nylon, we didn't get elastic-bounced at all. Lots of boats are using hundreds of feet of nylon without whipsawing about.

[Dockhead]

Originally Posted by Andrew Troup

Dockhead

I'm glad you've withdrawn one of your accusations about the broken snubber.

As to the other: every material has elasticity. You may be speaking shorthand when you claim that polyester rope does not, but I think it's potentially misleading. Even straight chain has elasticity.

Dacron (polyester) halyards are a lot more elastic than wire halyards, but even wire halyards stretch.

There are situations where a polyester rope's degree of elasticity would render it perfectly suitable for snubbing.

In some cases nylon is just too elastic. One problem with this has been touched on further up the thread: the heat buildup is too concentrated (a polyester snubber for a given situation is a lot longer, so it dissipates a given amount of energy without the same temperature rise). This largely explains why the dynamic climbing ropes of which you speak must be discarded after a VERY low number of load cycles. (ie, falls). Nylon does the job, but at a cost to itself.

The crucial point, I reckon: for any snubber not to be at risk: the elongation should be controlled by the amount the chain is eased. The chain should straighten before the snubber reaches a % elongation corresponding to a load which will, over time, cause it to fail.

It seems to me that, regardless of material, that failure load limit will need to be set lower if the snubber terminates in a rolling hitch than a sheet bend, because in a decent bend, the rope receives some clamping to get rid of tensile stress before it is required to perform a sharp turn.

I think that to dismiss a potential learning opportunity in the OP, albeit "learning by proxy", would be a shame.

Maybe I'm unfairly maligning rolling hitches and I may be maligning nylon. In the latter case, I think it's partly because I favour much heavier chain than is currently fashionable, so there's little extra energy left to 'sop up' by the time that's been hauled straight.

In fact for me, except in the most extreme conditions, or on a boat whose chain is light in relation to displacement, a snubber is a short affair rigged with minimal 'give', intended to keep the noise of chain rumble from telegraphing through the bow roller more than anything else. And pretty much any type of line will serve.

I realise that, in theory, heavy chain is a total waste of time, but I console myself with how well it works in practice.

Well, first of all, I never said that Dacron has no elasticity. Naturally, like everything, it does. But it is a fraction of that of nylon -- 1/4, or 1/5. I will stick with my assertion that Dacron is a poor choice for a snubber for most people, because of this: when you are "tuning" your snubber, you are walking a fine line between elasticity and strength, as Nick implied. For a given amount of elasticity, nylon is going to be several times stronger than Dacron. So unless the amount of elasticity your particular setup requires is extremely little, your "tuned" Dacron snubber is going to be extremely weak, or else it's just not going to be very elastic at all, and will not absorb any shocks. I would bet any amount of money that Evans' snubber failure is because of the inelasticity of that snubber, jerking up and ultimately breaking it. Stuff that jerks, breaks. That's the whole reason why we introduce elasticity into jerking systems like anchor rodes.

As to heavy chain -- this is thread drift but interesting. Choosing chain is another one of those big tradeoffs without any clearly good solution. My choice is actually the same as yours -- I have 100 meters of big, heavy, 12mm chain, G40, weighing about 330kg. This is a disastrous choice in terms of what it does to my boat's sailing performance, since my chain is kept right in the bow. There are plenty of other disadvantages, too. But I am conservative about everything having to do with ground tackle, and for all the obvious advantages of light, high tensile chain, I just don't like the idea of it. It seems to me cutting it too fine, relaying on such little bits of metal to keep my boat off the rocks. Just that too much corrosion, a little overheating while regalvanizing, and that finely engineered extra strength is gone. So I stick with old-fashioned, heavy, full size, G40, and sleep better at anchor for it. One of the blessings of heavy chain is that the catenary damps the snatch loads, but only to a point -- I don't agree that having heavy chain obviates the need for a proper elastic snubber. Once the catenary pulls out, it doesn't matter how heavy or light your chain is, and it's not just high wind loads, but wave action, which can do that (I speak from unfortunate experience).

As to knots -- I don't speak about them with any authority at all. My knowledge of knots is pretty much limited to what I was taught, and which is partly misremembered, probably, and the dozen or so knots I know in muscle memory. There was a recent test of knots in Practical Boat Owner which I believe repeated what I had learned about sheet bends, and which added the interesting information that dyneema and other exotic ropes with a polyester cover perform very poorly when knotted, because the cover rips off, essentially neutralizing their added strength unless they are spliced. If I'm mistaken about the strength of sheet bends, I will be very happy -- I use them all the time. On the other hand, I don't give all that much thought about it, because I never use ropes anywhere near the limit of their strength anyway, which seems to me to be fundamentally bad practice.

For tying on snubbers, I don't think the choice of knot is very important -- we read that someone even just ties a couple of half hitches with good success. Evans' is the only case I have ever heard where a snubber broke at the knot (and as I said, I don't think it was because of the knot at all). Especially with nylon, the rope works into the chain anyway, especially a big chain. So I think everyone should experiment and just use what feels right. I think the simplicity of tying and untying is a key value in the choice of knots for this application. If you have to futz around with it, this is a big downside. That's one of the several reasons why I like rolling hitches for this -- two seconds and without a single brain cell of thought wasted on it, and it's either on or off.

[Kettlewell]

I think the simplicity of tying and untying is a key value in the choice of knots for this application. If you have to futz around with it, this is a big downside. That's one of the several reasons why I like rolling hitches for this -- two seconds and without a single brain cell of thought wasted on it, and it's either on or off.

^^This^^ You need to be able to do whatever it is you do in the middle of the night, having just woken up to the howl of wind, after putting out additional chain. For me, the modified rolling hitch can be done and undone in seconds in the pitch black by feel alone if necessary. Another thing I like about using an ordinary line with a knot is that I always have plenty of alternative lines around if needed--if I have to let a snubber go over the side, or if I decide I need more length than I have, or if I need to move up a size, or if one should start to look iffy with chafe.

[estarzinger]

Soft shackles:

So, I deployed one 'for real' this morning. I picked a shorter one, since it does not need to be long for the 'thru the chain link' method (it does need to be long for the klemherst hitch method). It does seem to meet all the requirements I was looking for - dead easy, never jam or be at all hard to untie, can be done on deck and run in and out over the roller, never slip, cheap and you can build your own . . . so I will be using it for the rest of the summer unless some unexpected issues develops with it.



It also struck me while I was deploying it that you can also very easily add a back up second soft shackle to another chain link. This should never be necessary, but might be nice if you get a storm forecast just for reassurance.

[estarzinger]

Regarding the rolling hitch:

I was at a party last night, and knew a Hall Spars rigger would be there, so I took the broken end to show him. He said that if rope breaks due to shock loading/too little elasticity it will break first on the outside of a bend (Where the fibers are most stretched), but if it breaks due to motion/chafe/cutting itself it will break first on the inside of the bend. Looking at my broken piece he said it was pretty clear it broke at the inside of the bend.



I had the mantus hook on last night, with a spectra sling and a sheet bend (actually a 'triple sheet bend') to the same Dacron line that broke with the rolling hitch. The conditions last night were quite similar to when the rolling hitch broke. And the sheet bend did not break or show any wear at all.

Here it is as I lifted it out:



And here it is with the sheet bend untied . . . it had been loaded up and took a little fiddling to 'break the back' and get it started but it was not too bad.



So, my conclusion is that while the rolling hitch does have the advantage of being easy to tie - we all know how to do it, but it is probably not the absolutely ideal knot struicture for this cyclic loading application, and if you wanted to learn one of the other knots (like the icicle) it would be 'better'. But obviously it's not a huge problem as some of you have used the rolling hitch a lot.

But the soft shackle seems to me to beat either the rolling or icicle hitches.

[Andrew Troup]

Originally Posted by Dockhead

In order for a snubber to do any snubbing, it must have elasticity, so only nylon will do. ....

... the main problem is that Dacron is really entirely unsuitable for a snubber -- it won't do any snubbing at all. ...

Maybe I misunderstood; it seemed to me that the logical import of your statement 1 is that lines other than nylon do not have elasticity.

and it further seemed to me that your statement 2, that such lines could not "do any snubbing at all" is absolute and unqualified, and accordingly misleading.

In comparison with chain, ropes other than nylon are still immensely elastic. If the chain is heavy enough to do the lion's share, the snubber can be a lamb ... except of course in "bows going under" conditions, or anything approaching that (or in shallow anchorages, or other special cases)

You describe your "heavy" chain as not providing enough snubbing to get away with a minimal snubber: perhaps that's because, given the size of your boat, in my terms it's actually medium, rather than heavy.

To me, 5/8" / 16mm is "about right" for a boat your size, and therefore qualifies as "heavy". It was pretty standard on the expedition yachts in that size range I've sailed aboard, but admittedly less so in recent times: the fashions have certainly changed.


In the more general case, my feeling is that designers simply put the anchor locker too far forward, particularly on boats intended for the mass market ....

[Andrew Troup]

Originally Posted by Dockhead

...
I've never tried a dynamic climbing rope, but since these are specifically designed to stretch and absorb shocks, I can imagine that this is the best material of all for a snubber. Shame the largest size is 11mm, which is too small for my boat.
They are also made out of nylon by the way.

They may indeed be ideal in practice, but I think when making the argument on theoretical grounds, it is important to remember that they're not intended for cyclical loading, so it does seem to me that drawing a straight line between the two applications does not seem prudent.

I have a problem with nylon which is quite narrow and situation-specific: if the boat is to be left unattended, I would rather not risk the snubber failing from internal heat combined with weakening due to water absorption.

As you have pointed out, a broken snubber can put the boat at risk.

Faced with this situation in future, I plan to try something more tailored to enhance the catenary action of the chain, involving alternating kellets and floats along its length. It seems to me this would be particularly effective in shallow water.

ON EDIT:
getting back to the climbing rope, and attended anchoring:

Has anyone with a boat in the Dockhead/Evans size range thought about using two 11mm climbing ropes, ie a loop? One tail could be set tighter than the other if you wanted to experiment with a variable rate spring as suggested further up the thread ...

[Andrew Troup]

Evans:

You asked what sort of webbing people were using for slings. It used to be possible to get really stiff climbing tape (webbing) from European climbing equipment vendors, almost like a flat belt.

It seems to me there must either be a sewing machine built integral with the loom and sewing the tube longitudinally in several places, or they've got some crafty way of arranging the heddle, because the tube cannot be opened.

Fantastic for jacklines, prevents the dreaded 'rolling underfoot' problem *.... Last time I was in France I found a marine supplier who had some, but I've subsequently used it all up... I've hawked a sample around all the local outfits but nobody has ever seen anything like it.


For the kleimheisted sling I mentioned further up, I use conventional softish tubular webbing; I would prefer something somewhat stiffer, but it seems to work just fine in practice, and I think it would be very difficult to tie the beer knot in stiffer webbing.

* to counteract this, I recommend buying the tubular webbing from a sailmaker and getting them to run a couple of rows of stitching along it

[Kettlewell]

I have a problem with nylon which is quite narrow and situation-specific: if the boat is to be left unattended, I would rather not risk the snubber failing from internal heat combined with weakening due to water absorption.

If you're worried about your snubber failing due to heat and water absorption you're already in survival conditions--has never happened to me up to and including hurricane conditions. The whole melting thing is way overblown if you have a properly sized bit of nylon on there. Sure, it can happen, but chain breaks too, and dacron apparently can snap (see the OP) that isn't right for the purpose.

[JPA Cate]

FWIW, I saw an article that suggested that the nylon is more likely to overheat if it is inside hose chafing gear; and that frequent changing of softer, water absorbent chafing gear would be much more secure...

Ann

[Andrew Troup]

Originally Posted by Kettlewell

If you're worried about your snubber failing due to heat and water absorption you're already in survival conditions--has never happened to me up to and including hurricane conditions. The whole melting thing is way overblown if you have a properly sized bit of nylon on there. Sure, it can happen, but chain breaks too, and dacron apparently can snap (see the OP) that isn't right for the purpose.

Have you left your boat unattended for a week on a nylon snubber in a windy location?

Regardless of your answer to that: a single person's experience is a small sample to generalise from; I too have not have problems with nylon in the past, but the fact that other people have done is sufficient to make me leery, specifically in the 'abandoned ship' scenario.

In addition to the cumulative effects of cyclical loading on nylon, when there's nobody there to freshen the nip at the first sign of chafe etc etc, the problem is altogether different.