Eddy testing for chainplates

[tsell]

In all the research I have done on inspecting chainplates it seems that visual inspection and dye penetrant testing are the norms.

Has anyone had any experience using Eddy testing for chainplates instead of Dye testing? I brought the chainplates from a 1982 Catalina 38 that I am currently refitting to a machine shop that offers NDT testing. They suggested using Eddy testing to check for flaws, as opposed to Dye penetrant testing since if they saw something with the Dye test, they would investigate further using Eddy testing, so just starting with the Eddy test was more economical and should give a more definitive result.

The chainplates came back with a clean bill of health, but I wonder if there is anything else I should be looking at.

Any opinions welcomed.

[Gone Gypsy]

Dye penetrant is an excellent test for cracks intersecting the surface. Surface roughness could lead to a minor false indication of a crack. If the first test provided a questionable indication of a crack I would polish and retest. There is no denying a true crack highlighted by die penetrant. It is in your interest to not reinstall plates that have surface roughness. Have them electrolytically polished if they are rough and remove any future potential crack initiation sites. The fact the NDT firm was willing to set aside a dye indication using eddy current merely says they are more comfortable with technology providing a pass/fail signal rather than on experience with a very reliable test requiring human understanding. There is nothing wrong with eddy or dye. The difference in tests is like using a sextant versus GPS. Using a sextant you know why your observations got you to the islands. With GPS you arrive without understanding, but you do arrive. Your plates are good. I have personally performed dye penetrant, Magnaflux, eddy current and ultrasonic testing. I would sail the world on plates verified using die penetrant.

[Westcliffe01]

How about a photo of the surfaces of the chain plates ? If the surfaces are smooth and without pitting, Dye penetrant is an excellent test method. I used to work in the petrochemical industry and on many occasions inspected the welds between pressure vessel shell and external heat exchange tubes (often over 1000 tube ends from your knees to your waist). As I recall it was a 50 000 psi steam superheater, so the tiniest crack would lead to loss of system pressure. The refinery never had a false start after I did the job, so they used to call me every time that job came up... The way the job is prepared, the way the excess dye is removed (with the least use of solvent) in preparing to put the developer on can have a substantial impact on the outcome. Anyone who knows what they are doing can recognize in a few minutes if a rookie is at work. Refineries have strict documentation so if something fails and they pull the inspection report and it has your name on it, expect to get a subpoena.....

[BillKny]

By the time you have gone to all the effort to pull a chain plate, the cost of eddy testing, or even dye testing, seems totally silly.

In most cases we are talking about a simple piece of metal plate with a few holes in it. Just have a new one fabricated, it's not significantly more money than what you might spend on testing the old one and you don't have to worry about the effectiveness of the test method.

[Orion Jim]

We made use of eddy current inspection extensively in the aerospace industry. With the tolerances involved in chain plates and the cost of the finished part it seems like overkill. Buy your product from a reliable source and specify that it pass a visual and an F.P.I. inspection.

[RaymondR]

Originally Posted by BillKny

By the time you have gone to all the effort to pull a chain plate, the cost of eddy testing, or even dye testing, seems totally silly.

In most cases we are talking about a simple piece of metal plate with a few holes in it. Just have a new one fabricated, it's not significantly more money than what you might spend on testing the old one and you don't have to worry about the effectiveness of the test method.

If they are stainless steel, as most appear to be these days, renewing them is the best way to go and probably wont cost any more.

Fatigue in metals is cumulative. You could have a perfectly good test today, load it up a few times, and tomorrow have a crack.

[tsell]

Thanks for the responses. The testing wasn't expensive, as the owner of the shop gave me a huge break on the cost. $150 to inspect 7 chainplates and the stemhead fitting. The plates on the 38 are a bit more complex than simple bar stock with holes drilled in them, so I suspect the cost will be significant to have them fabricated. Catalina direct sells the stem fitting for $500. I will upload some photos after I pick them up.

The question of service life and fatigue cycles has been nagging me, and it seems that there is no good way to gauge that other than replacing after x number of years. Seeing that these are likely 38 years old, replacement might be the way to go regardless of test results.

[ArmyDaveNY]

Have you considered other materials for the chain plates? There have been a few threads on this forum that talk about it, and at least one member who makes them.

https://www.cruisersforum.com/forums...es-182007.html

This thread gives a pretty good idea of some of the different materials including Titanium, bronze, carbon fiber, 2205 duplex stainless, and others. There are some good comments that I think you will find interesting. Good luck with your selection!

[MJH]

Originally Posted by tsell

In all the research I have done on inspecting chainplates it seems that visual inspection and dye penetrant testing are the norms.

Has anyone had any experience using Eddy testing for chainplates instead of Dye testing? I brought the chainplates from a 1982 Catalina 38 that I am currently refitting to a machine shop that offers NDT testing. They suggested using Eddy testing to check for flaws, as opposed to Dye penetrant testing since if they saw something with the Dye test, they would investigate further using Eddy testing, so just starting with the Eddy test was more economical and should give a more definitive result.

The chainplates came back with a clean bill of health, but I wonder if there is anything else I should be looking at.

Any opinions welcomed.

This post got me thinking...has anyone heard of a way to test chain plates while they are in the boat? I'm wondering with all of this wonderful 21st century technology somebody would have come up with a more cost/time efficient way rather than having to pull them to test or automatically replacing them after a certain number of years/miles. What about a portable x-ray or sound refraction test? Any entrepreneurs out there?

Just wondering...

~ ~ _/) ~ ~ MJH

[Cheechako]

I don't think there is any cycle fatigue in chainplates to speak of. They really aren't cycled. they are just in tension, not flexed over and over.
I thought Eddy Current was mostly used in aluminum. ? Learn something every day I guess.
If the surface of the chainplate is clean and smooth enough, Dye penetrant should be good if done by a good reader. There is a lot of misinterpretation from DP otherwise. Any surface inclusion can be read as a defect or not.

[Gone Gypsy]

A good rule of thumb is that crack initiation is 90% of component total life and crack propagation (fatigue crack gradually growing with each load cycle) is 10% of the total life. Looking for cracks is really not the main driver in removing plates. Anaerobic corrosion of stainless steel trapped in a confined space that could serve as crack initiation sites and ultimate failure is what prompts most plate replacement. I cannot speak to the design goals or the duty cycle of your Cat 38 but unless you know of specific plate failure concerns in that design, many would consider no corrosion and no cracks as confidence to reinstall.

You can check for design issues on your craft and you have observed no corrosion. The wild card is past load experience. The earlier comment is right on. Fatigue failure is a cumulative process. Has the boat circumnavigated? Was it extensively raced? Is there evidence of poor past maintenance (loose rig)? For you, this unknown may cause you to lack confidence in your rig no matter what a design review and inspection tell you.

Being a responsible Captain is all about confidence. As a metallurgist with a career in materials, structures and failure analysis, my confidence may be different than yours. It is also possible my technical confidence could actually expose my crew to unreasonable risk.

You would always be confident with new plates, Captain.

[joelhemington]

Originally Posted by tsell

In all the research I have done on inspecting chainplates it seems that visual inspection and dye penetrant testing are the norms.

Has anyone had any experience using Eddy testing for chainplates instead of Dye testing? I brought the chainplates from a 1982 Catalina 38 that I am currently refitting to a machine shop that offers NDT testing. They suggested using Eddy testing to check for flaws, as opposed to Dye penetrant testing since if they saw something with the Dye test, they would investigate further using Eddy testing, so just starting with the Eddy test was more economical and should give a more definitive result.

The chainplates came back with a clean bill of health, but I wonder if there is anything else I should be looking at.

Any opinions welcomed.

Sounds like the perfect time to consider titanium chainplates - apparently much stronger than SS and not nearly as expensive as they used to be.

[RaymondR]

Originally Posted by tsell

Thanks for the responses. The testing wasn't expensive, as the owner of the shop gave me a huge break on the cost. $150 to inspect 7 chainplates and the stemhead fitting. The plates on the 38 are a bit more complex than simple bar stock with holes drilled in them, so I suspect the cost will be significant to have them fabricated. Catalina direct sells the stem fitting for $500. I will upload some photos after I pick them up.

The question of service life and fatigue cycles has been nagging me, and it seems that there is no good way to gauge that other than replacing after x number of years. Seeing that these are likely 38 years old, replacement might be the way to go regardless of test results.

One of the Q&D ways of setting your mind at rest is to calculate the section area under stress on the chain plates and the wire of the shroud, then derive the ratios of the areas. Lets say it comes out 10:1 and consequently the stress in the chain plates are 1/10th of those in the shroud wire line. Since your insurer is going to insist you replace the standing rigging every 10 years and your boat is 30 years old you'd have 7 more rigging replacements before you should have to replace the chain plates.

Is this a valid analysis? I dunno, but it's probably as logical as any other available.

[Wotname]

Originally Posted by RaymondR

One of the Q&D ways of setting your mind at rest is to calculate the section area under stress on the chain plates and the wire of the shroud, then derive the ratios of the areas. Lets say it comes out 10:1 and consequently the stress in the chain plates are 1/10th of those in the shroud wire line. Since your insurer is going to insist you replace the standing rigging every 10 years and your boat is 30 years old you'd have 7 more rigging replacements before you should have to replace the chain plates.

Is this a valid analysis? I dunno, but it's probably as logical as any other available.

While it sounds valid, I suspect it isn't.

IMO - Chain plates (at least SS ones) fail from anaerobic and crevice corrosion while the shrouds; not so much.

Only an opinion though!

[Jim Cate]

Only an opinion though!

One with which I agree, Wottie! chain plates don't have too much fatigue failure issue whilst standing rigging definitely does... added to the vagaries of corrosion to which all s/s bits are subject.

I will add my approval to the idea that for most chainplates replacement (once they are removed) would not be much dearer than serious NDT... and replacement really does reset the clock. Testing (even if 100% accurate) only says that the clock ain't run down yet!

Jim