The Offshore Voyaging Reference Site

A Better Chain Plate

chainplate2

While thinking about how to make the Adventure 40 an incredibly trouble free and reliable ocean voyaging boat, I have worried most about two areas: chain plates and the rudder.

Both for the same reason: they are the area in conventional fiberglass production boat construction where stainless steel and fiberglass come together in the presence of salt water to make an unhappy marriage.

Let’s leave the rudder aside for this post.

Stainless steel chain plates, particularly if they are set inboard and poke up through a slot cut in the deck, will, if the boat is sailed hard offshore, eventually leak, usually within a few years of launching, sometimes much sooner.

And once water gets in, the chain plate itself starts to corrode—a problem that is serious enough that any boat so built that is more than ten years old should have its chain plates carefully checked, and after 20 years replacement is often the only safe alternative.

Yes, I know, there are a lot of different ways to try and leak-proof a stainless chain plate. Some of them even work for a while. But the facts are that as long as you are mixing two different materials with radically different mechanical properties and trying to keep water from getting between them, things are probably not going to go well in the long term. Something to remember when refitting a boat—teak decks, or any wood trim for that matter, come to mind.

One of the very cool things about metal boats is that the chain plates are made of the same material as the hull and are bonded to it by welding, thereby avoiding leaks and other problems for many decades.

I want to do the same sort of thing with the Adventure 40, and Matt Marsh, engineer and all around smart guy, with a lot of hands on experience with fiberglass, makes a very convincing case that we can do just that by making the chain plates part of the hull.

One option would be to build the composite chain plates outside of the boat from pre-preg materials over a mold and then bond the resulting assembly to the hull and/or bulkheads, as well as the deck, with epoxy or Plexus adhesive.

I’m pretty sure this would work well since a friend of mine made a lot of complex fittings, including a goose neck, in this way, very quickly and inexpensively.

And I have another friend, one of the world’s most experienced composite boat builders, who assembles the high tech boats he builds with Plexus and says that the resulting bonds are stronger than the laminated assemblies they connect.

By the way, I’m not the only one who does not like stainless steel chain plates. The new Navy 44s have composite chain plates because of the intractable leaks that the older generation of boats suffered from.

A Campaign Against Stainless Steel?

And if you are beginning to think that I have a “tude” (Bermudian dialect for attitude) against stainless steel, you would be right. I think that we yachties put far too much trust in the material and often use it, just because it looks shiny and strong, when other materials like good quality bronze, galvanized steel, aluminum, or composites would be better and more reliable.

Comments

If you have solid engineering knowledge or first hand experience that will help us solve this problem in a cost effective way, both for the Adventure 40 and chain plate replacement in older boats, please leave a comment.

63 Comments
Oldest
Newest
Inline Feedbacks
View all comments
Steve Guy

How about titanium?

RDE (Richard Elder)

Hi John,
Bond pre-molded fiberglass/carbon chainplates into the boat with a secondary bond using Plexius or epoxy? Not on my boat! Look at a finite element analysis of the proper fiber orientation for load distribution and you will see why.

Plexus certainly has its place, and one of those is bonding an internal liner into a hull where the large surface area and imprecise fit makes use of the tenacious and somewhat flexible adhesive properties.

I’m no fan of stainless steel either— I’d always prefer silicon bronze for things like bobstay tangs that regularly get dunked in salt water.

Matt Marsh

Richard, while I agree with you that gluing a composite plate into the hull with Plexus is not the way to go, I respectfully disagree with your negative assessment of composite chainplates in general. We are talking here about long loops of continuous fibres, moulded right into the part they connect to; this is a tested and proven system that is well understood. Glued-on plates of flat stock are a very different animal and one that I do not recommend.

I have built similar components in both carbon and e-glass, by hand with minimal or no tooling, that took no longer to make than their metal equivalents would have. There are many well-documented successful examples of composite chainplates that are fully integrated into the hull.

Knowing how the fibres have to be oriented is one key to making it work; creating easy-to-use tooling is the other. The former is a solved problem, the latter can be a bit tricky to engineer but is certainly doable.

RDE (Richard Elder)

Hi Matt:
Actually I have a very positive attitude toward composite chainplates and would use them on any custom boat I was building for myself. My reservations are for a production run of perhaps 100 boats where the goal is to deliver function and reliability at a price point no other manufacturer currently meets. The valid comparison is not the time it takes to build a one-off item in composite vs. metal, but the difference in time and expense of building upwards of 1,000 semi-one-off fabricated items vs. calling up your favorite CNC equipped machine shop and having them delivered to the factory door.

John, I’d change your comment that metal chainplates always eventually leak to “improperly designed chainplates that protrude through fiberglass decks are a frequent cause of leakage. ” We could make the statement that all aluminum boats in salt water eventually disintegrate and be perfectly correct as well! (LOL) Time and the sea always win in the end.

A properly designed metal chainplate tang installed as D. Stevenson described in an earlier posting should easily outlast several sets of standing rigging, and need never leak if re-sealed when the standing rigging is replaced every ten or 12 years.

Matt Marsh

Six or eight years ago, Richard, I would have agreed with you 100%- doing this in composites, in production, would have been a royal headache of the worst kind.

In that time, though, composites manufacturing technology has been flying ahead at a crazy pace. In this case, you’d drape a handful of fibres through a set of CNC-milled moulds, clip that to the gunwale of the hull mould (or to the edge of a bulkhead that’s being infused as a flat plate on the floor), feather out the fibres, tape the mould seams, bag the rest like you normally would, and hook up a vacuum hose.

I wouldn’t have even considered doing this on a production basis without the incredibly precise tooling that modern CNC provides, or the excellent infusion resins that are now available, but it’s not nearly as difficult as it used to be.

Regarding leaks: I’m inclined to attribute the “leaking through-deck chainplates”) problem to structures that aren’t nearly stiff enough and to a generally poor understanding of how to properly seal complex joints.

RLW

Kurt Hughes recently sorta/kinda discussed this very thing on his blog and suggested a strong, leak-proof, inexpensive, and easily done on a production basis method…

Charles

Have we considered bonding a pair of composite load-distributing loops to the hull and passing SS bolts through the loops to be the sacrificial (maintenance) wear and shroud attachment point? An arrangement such as this seems like it would give the best of both worlds. Making and bonding loops should be pretty easy even in an assembly-line type mass production setting.

Matt Marsh

That’s essentially how they work, yes. Loops of continuous fibres form a thick, strong bundle with an eye above the deck, and the tails of those fibres then fan out into the hull or bulkhead. You then use a metal bolt or toggle through the loop to attach the rigging. This bolt (unlike the bolts that secure a traditional metal chainplate) is not subject to anaerobic / crevice conditions that would cause corrosion.

Myles Powers

Why aren’t we looking at the stick. Maybe a redesign here and we can forget about the chain plates ?

Mike

Great idea. A stayless Adventure 40 with carbon fibre mast ought to be an option.

Gary Mendenhall

I would not even consider using stain-less steel of any stripe for replacement chain plates when suitable titanium alloys are readily available in flat bar. Expensive? Depends on your frame of reference doesn’t it? How expensive do you imagine it could be to lose your rig in the middle of a passage?

Martin

It seems a key focus of this thread is how to avoid deck leaks in the long term.

We can easily avoid crevice corrosion by using bronze or another metal, but the potential for deck leaks would remain.

It may be useful to set a reference point against which we can compare – may I ask what is the “best practice” method we can use to install conventional metal chainplates in order to minimize leaks?

I assume it is to build a hard doughnut around the chainplate exit point, and fill this in with very low modulus butyl.

But if I am assuming wrong, can John or Matt or Richard correct me? It will be useful to have that reference point.

Tom Bohanon

Hi John,

I’ve looked for Dick’s ideas on chain plates, but no luck. Can you refer me to the article where they appear?

Thanks,

Tom Bohanon

Tom Bohanon

Thanks John. I’ve seen some descriptions of using SS or aluminum cover plates in a similar fashion to your use of the teak blocks, but it seems like better odds of stopping a leak with yours. I’ll lurk around to see if Dick chimes in…he’s got a lot of good ideas.

Nick Hallam

Way back in 1989, I installed some ‘hybrid’ chainplates on a series of 64 foot John Shuttleworth charter cats. For the main chainplates, John designed massive CNC-cut stainless plates, with broad horizontal slots below the usual holes for clevis-pins. The plates were installed through the glassfibre deck (solid at this point) and were a tight fit in their slots, so the deck played its part in stabilising the plates against rocking. We ran UD carbon through the slots and down onto the bulkhead or web below, in the obvious fan pattern, and that was that. 12 tons of cat with enormous righting-moment and as far as I know they are all still out there.

However, I still don’t have any problem with properly-installed stainless chainplates: totally agree about butyl bedding and also the use of good wide cover-plates, thick enough to be fixed down without bending, so as to compress the bedding nicely but gently.

Last thought: make sure the chainplate is thick enough to fill up the space between the toggle legs: there’s nothing nastier than a thin chainplate!

Roland

I recommend that you focus on a cost effective solution. Exotic and non-proven solutions means increased production cost.
Stainless steel chain plates are a proven method, and very cost effective. If installed properly they will not leak.
If you still have concerns about leaks you should perhaps focus on accessibility in case it leaks. If the chain plate is easy to access a leak is not the end of the world. Easy accessibility normally means less installation time in production.

Roland

Hmm. The most common thing to leak on a boat is portholes, skylights and
genoa tracks.

Eric Klem

John,

I thought that I would throw out an idea, I am not sure that I have sold myself on it but something to ponder. Certainly, composite chainplates would be the best but it would be an idea if you had to go metal.

My first thought was, is it really a problem? My conclusion is just like yours that it really is. The next thought was, can you mitigate the effects through good interior design but I don’t think that would work terribly well as you still have fresh and salt water entering through the deck. Finally, I thought about whether you could make it easily serviceable so that the people who actually sail offshore could reseal them easily once a year but again, I don’t think that this is the solution.

This got me to thinking about how we would have approached this problem when I worked as an engineer for large industrial air compressors and steam expanders. These machines are full of critical seals both from a safety and longevity standpoint and we had extremely few problems. The major difference in application was that in those machines, you could very tightly clamp two rigid surfaces together so that an o-ring or gasket (some of these are pretty advanced these days) would work. So why couldn’t you do the same thing on a deck? The problem with current designs is that they have too much relative movement between parts so we try to deal with this through using flexible goop which invariably looses its seal with one surface. There is no reason why you couldn’t design a chainplate with virtually no movement relative to the deck.

What I have in mind is a chainplate which sits flat to the deck and bolts through the deck to a massive backing plate. Both plates need to be designed to be rigid enough which should be easy and the deck needs to be designed to take a pretty high compressive load. The bolts would be torqued up pretty hard so that even at maximum rigging tension, the friction force of the plates to the deck prevent any movement. A little deflection and tolerance stackup is acceptable with a properly sized o-ring, this is standard stuff that any good mechanical engineer should be able to put numbers on.

From a production point of view, this would be doable at a reasonable cost. You could cast your chainplate and backing plate (including attachments to whatever internal structure are required) and do a bit of post machining to get the flatness tolerances. O-ring grooves are expensive on high run items as they take a little while machining but are probably cheap compared to many other options being discussed. The trick with the deck would be that you would need to get it flat and it could not be too far out of parallel.

Sealing head bolts have been around for quite a long time and they really work well. The simplest ones use an o-ring under the head and need to be torqued down hard enough so that there will be no relative movement between the head and the plate it contacts. Using these bolts in conjunction with an o-ring outside of all the bolt holes between the chainplate and deck would fully seal everything.

The biggest problem that jumps out at me is that you have created a really rigid point in the deck so you need to transition to the rest of the deck which is less rigid. I have also not run any numbers of the compressive strength of a solid fiberglass laminate versus the necessary bolt preload.

As I said at the beginning, I am not sold on this and have not done any real engineering on it but it might work. Thoughts?

Eric

Eric Klem

John,

I see that you feel composite chainplates will work which is great.

I did want to respond to the question about cast parts under high loads. Cast parts have gotten a bad reputation due to poor design and manufacturing over the years. Nowadays, there are 3D computer modeling tools which can be used to ensure that the mold fills and cools properly. Most problems are related to voids and areas where the metal cools too quickly and isn’t bonded to the surrounding metal.

From an engineering perspective, the first thing is to design a part which is compatible with casting (basically the part shape determines this). Then you need to do stress analysis on the part. When determining your allowable stress, it is standard to use knock down factors for castings as they are not as strong as billet material. In the end, you have to have a slightly larger cast part provided that your design and process control are good.

Spartan Marine Hardware casts all of their bronze chainplates. I am not aware of any failures and they have a lot of sea miles on them so it can be done.

I hope that this answers your question.

Eric

Magnus

High Eric,
This is exactly the type of chainplates which are fitted on my Nicholson 31, except without the o ring. I believe they are the originals ( built 1977 ) and I have just inspected them with the intention of replacing with nitronic 50 or silicon bronze. There is no evidence of leaking (even after all this time) so it would appear this is a well engineered design and a good alternative to composite.
Magnus

Daniel Penniall

Hi Eric,

What you describe here is basically what my own boat has.

The original rig was masthead, inline spreaders, deck-stepped. In the early 1980s the previous owner decided he wanted a bit more juice, so in went a fractional, swept spreader, keel-stepped rig from the next-bigger model boat from the same designer — all done with his approval.

Anyway, as you can imagine this necessitated new chainplates in new positions. The above deck part is just a flat plate with an upstand with the appropriate holes for the rigging, and is connected to the backing plate by six bolts. The backing plate is a litte different in that it’s actually a small “I” beam with end plates, and it spans the gap between the bulkhead and some very substantial FG knees that were added. But the above and through deck concept is as you describe, and it is a success.

In the 40 years since this work was done, the boat has done at least 30,000 miles — many of them offshore — and no leaks.

Roland

Eric,
You are absolutely right. It is all about preventing the movement between deck and the chainplate. This is a technique used on many boats, with success.

John,
You are right that leaks from skylights and ports are from poor techniques and material. But the reality is that quality ports and skylights are more expensive to buy. You have a very limited budget. You want to spend the money as wise as possible. Anyhow, Good Luck with the project.

David Head

Hi to all.
I love to embrace the new but surely a proper cruising boat should repairable worldwide with simple tools and low technology. If I were younger I would build a Fay 40. May even follow Paul Fays design to the letter and use a junk rig. Solves all the problems created by complex designs. This dialogue is very good reading, and maybe get us ‘die hards’ thinking of the alternatives.

murray

I would have to disagree on using bronze before Titanium. Titanium (if you were forced to use a metal) is superior in almost every single respect over almost any other metal. It does solve one issue and that is the corrosion and rust problem with commonly used stainless. if you could solve the leaking issue, titanium would certainly outlast the entire boat and you!

Paul

I may be mistaken but aren’t metal chainplates often part of the lightning dissipation system on many sailboats? If so, would some additional conductive material material need to be added to the shroud ends ?

Mike

Ive not seen the deck layout of your boat.
Composite or any chainplates located outboard, on the sheer clamp near Bmax , are vulnerable to collision damage at the dock . When the shrouds are outboard you are also exposing your standing rigging to the dock or boats next door during a blow on the beam. When damage occures expensive structural repairs and a new topside paint job are needed.
Additionally shear clamp chainplates make crew movement forward difficult because the diagonal and shroud block their way. Moving bags of sails forward at sea becomes troublesome .

When the chainplates inboard , thru the deck, crew can easily clip into the jackline with their safety harness tether then charge forward outside the shrouds with a steading hand on the lifelines.

I see no advantage to sheer mounted chainplates.

As for composite chainplates…ss chainplates are simple to visually inspect, composite is not.
When standing rigging is inspected for flaws… all components, chainplates, Hydraulic rams, mast step, partners… must be inspected .

Be careful when transferring lightweight race boat techniques to an all purpose cruising boat.

Mike

Dye testing rod is a fools game. Chop the head off and re head. Visual clues are corrosion at the terminal or loss of symmetry of the rod head when you rotate it with your fingers. Rod Standing rigging with more than seven years or 70,000 miles is normaly replaced.
I’m sailing a rod boat and have covered 300,ooo miles over 20 years. The rods have been replaced 4 times. Tip cups once and chain plates once. The cups and chain plates are condemned not for any visible decent, but because the rigger could not certify them. This will also happen with composites.
In a lightning event energy must rapidly escape. Rigging is a normal path. You should discuss this with your NA.

michael

“I think that the failure mode of rod is pretty much 100% cycle load driven, not time. ”

crevice corrosion at the stem ball fitting happens at the dock…no cycling. The stem balls of Turnbuckles and backstay insulators that face upwards and trap water are particularly prone.

Bruce Cuthbert

We have a steel boat , but the chainplates are all 2206 duplex s/s.
This is both stronger and much more corrosion resistant than 316.

Scott Dufour

I’m resurrecting this thread because I’m getting ready to replace the through-deck chainplates on my 1979 Pearson 10M. There’s a good conversation here about how to incorporate them into a new build, but only a few tantalizing references to retrofit. Is there anybody out there that’s done the math and testing?

Erik Bachmann

I know this is an old post but I was wondering what opinions were out there regarding using a material such as g10 which could be purchased premade and installed in much the same way as a conventional chainplate. Perhaps with fg reinforcements in addition to bolts. It would appear at least in tensile strength g10 is similar to SS. Perhaps a little less, (38,000 psi yield vs 42,000 psi) however, that could be compensated by going a little thicker on the chainplate. Especially as many chainplates to not fill up the toggle gap that attaches to them as they should. You would get the advantage of having similar materials all of which are corrosively inert.

Alwin Bucher

Hi John,

I am writing to ask about my chainplates (aluminium, welded to the deck and the stringers below), and this seems to be the best place. The clevis pins have not been removed for about 20 years, and are now quite stuck, having corroded themselves in place (i.e. the hole is clogged up with aluminium oxide around the pin). I’ve managed to remove one over several days with patience, lots of penetrating oil, heating the chainplate whilst cooling the pin, vinegar, and some larger swings of a hammer directed at the pin. The hammer did manage to do the trick, but there are 8 more, and they all have a lot less swinging room. Also, I am a bit nervous about damaging my chainplates in the process.

Do you have any suggestions for how to best proceed? More of any of the above? Or are there any specific tools I should be looking at? I could see for example that a strong rod of metal that just fits through the hole for the cotter pin could be used to apply a torque to unstick the pin by twisting.

If you have some ideas for how to prevent this in the future, that would also be great. I thought about gluing in a stainless bushing with some epoxy, but then I worry that this adhesion will create a weak point. Perhaps I will just coat the pin in tef-gel…
I’ve attached a photo for context. Many thanks!

IMG_3167.jpg
Alwin Bucher

Hi John,

Many thanks for your reply, and happy to hear of your solution. That sounds like it should work!

Best,

Alwin

Alwin Bucher

Hi again John,

Reporting back as after several rounds of penetrating oil, the press I bolted together is not moving anything. I suppose most likely the issue is the gauge of steel (3mm) – it is bending significantly. I will try doubling it up for now, and we’ll see where that gets us! Do you remember what thickness you went with?

IMG_3397.jpeg
Alwin Bucher

Hi John,

Thanks for the rapid response! I will go with that then.

Alwin Bucher

Hi John,

Still waiting for the bigger press. In the meantime, I’ve got a related issue to deal with: The screws holding the jib track onto the deck (blind threads into welded thickened sections) appear to be held down by the pressure of the surrounding expanded oxide more so than thread. Taking them out obviously breaks that bond, so I only see four possible solutions:

1. Leave them in place and postpone the problem until the track eventually has to be removed, hoping the pressure of the oxide and remains of thread provide a sufficient bond

2. Take them out and hope that the remains of thread will hold when putting them back (with fresh duralac/tef-gel of course)

3. Take them out, redrill and retap to a larger thread

4. Option 3, but fitting a helicoil.

What would you suggest? I’ve attached a photo of the bolt just to illustrate the degree of lack-of-thread.

IMG_3457
Alwin Bucher

In the end, we were able to extract the pins with relative ease with the tool shown in the image below (made from grinding out a fork from a wedge designed for smashing up bits of concrete). Turns out the trick is to pill instead of push – that way, you are elongating rather than compressing the pin, which very much helps it exit the hole.

7C30F1A4-261C-4209-875C-DF6BD1E80B99