S/V Rainshadow

A few comments on the “hairpin” chain plate design on the Nic 38.  Looking at this as an engineer (well, actually I’m a physicist), there are some interesting design decisions that were made.  Basically, I think it’s a brilliant design with multiple redundancies built in.

Firstly, the fact it’s a hairpin shape means that each shroud is connected to two, semi-independent rods (pins). The most probable failure path for this design is for one pin of the chain plate to fail, followed by the deck plate lifting off the deck as the top of the hairpin moves upwards away from the broken pin, while remaining attached to the other pin, which likely bends some in response.  This upward movement is probably a cm or more, leading to a significant reduction in the tension on the shroud (the tension is taken up by the remaining pin, bend in the mast, and by the other shrouds attached to the mast).  This is great redundancy!  There are several reports of this kind of failure on the C&N Yahoo group.

To see why it’s unlikely that both fail simultaneously, think about how the process occurs. Consider a hairpin designed to have a working strength of 2,000 lbs and a failure point of 10,000 lbs. Per pin, this is 1,000 and 5,000 lbs. Suppose a material defect (manufacturing, corrosion, etc) causes one to become weakened so that its failure stress is only 1,000 lbs (20% of full failure strength).  A force applied to the hairpin of 2,000 lbs will then cause the weak pin to fail.  The entire force of 2,000 lbs is then held by the single remaining pin, which has a (defect free) breaking strain of 5,000 lbs, so should able to handle it.  As noted earlier, the actual strain after failure will be much less than the full strain because of the deformation of the hairpin.  As you can see, even if the remaining pin has lost almost 60% of its strength, in this example it still can be expected to hold the load. (OK, I’m ignoring shock loading, but the bending process of the hairpin should cushion that.)

It is possible that the bend at the top of the hairpin could fail, which would indeed be catastrophic.  However, this area is open to the air, easy to inspect and less likely to suffer from crevice corrosion than the hidden parts below deck, so that’s less likely. 

Another feature people have noted is the oversized holes in the deck – they are about 3 – 6mm oversized (the nominal diameter of the rod is about 9mm (12mm for the main upper shrouds), so this is a seemingly large error). People seem to see these as a manufacturing defect – some people even go so far as to fill in the holes with epoxy and redrill them to the “right size”, so the rods just slip in.  That’s a mistake – it begs for crevice corrsion to start, even if you goop it up liberally – water is going to get in.  The big advantage of the oversized holes is that it allows moisture to evaporate, and allows oxygen to get to what moisture there is on the surface.  A tight fitting hole will encourage moisture to remain there (capilliary action, etc) and the thin film of water trapped between the stainless and the hole will rapidly become oxygen starved – perfect conditions for crevice corrosion.  So, my advice is to keep those holes big!  Interestingly, on our chain plates, the dominant area for pitting was immediately below the underside of the deck (the deck is about 1-2 cm of solid GRP).  The previous owner had painted the rods up to this point.  I suspect the paint edge created a point for crevice corrosion to start – the water wicked in under the paint at the edges and started the corrosion there.  Moral of the story – don’t paint your rods.

I also think this is why the rods are only welded on top of the deck plates.  If the bottoms were also welded, it is more than likely you would get trapped spaces between the welds, which if (when!) a small (microns) gap occured would lead to water entry and terrible crevice corrosion.  In fact, if you think about it, the only purpose of the weld between the rods and the deck plate is to keep water from pouring in through the holes – they are certainly not structural. A good adhesive would have done as good a job (but looked crappy).  The tendency of the rods to move laterally (as the shrouds flex) is mostly resisted by the adhesive caulk used to seal the gap between the deck and the deck plates.  Those forces should be small.

It may well be that one of the weak links for these chain plates are the nuts used to secure them.  One CN38 owner reported that the nuts “crumbled”.  Worse, these nuts are an old British thread, incompatible with US threads, and no longer made/used in Britain.  So, they’re hard to find.

What can we do to prevent/reduce future crevice corrosion?  We’re going to try spraying liberal amounts of Boeshield T9 on the parts.  This repels water but leaves a film thin enough that oxygen should still be able to migrate through.  What we don’t know is how long this will last in an environment where the metal surfaces are left untouched.  The manufacturer claims it lasts “many months” and that it has no adverse effect on fiberglass, plastic, etc.  Unfortunately, not all of the chainplates are easy to get to, so regular inspection and recoating won’t be easy for those. That’s the worst flaw of the design in my opinion, and it’s a bad one.

Brion Toss did suggest replacing stainless chain plates with Titanium ones – he claimed it was cost effective nowadays.  We just might look into that.  I checked on McMaster Carr and the material costs are not that high – of order $30 per chainplate.  We have a bunch more chain plates to pull, and if any of them are scary enough to replace, we may well go down that route.