Chainplate refurbishment

[Roger Long]

Rich,

You covered this briefly in the tang thread but, for the benefit of anyone doing a search down the road, would you please repeat and expand on exactly what you think should be done with these chainplates?

They have pretty deep wire brush marks from when they were made. There is no visible sign of surface penetration at the rust line where the deck bedding ended.

The chainplate slots through the angle as shown here so the weld at the base is secondary:

In addition to the surface metaleurgy, I would also like to double the thickness so as to be a closer fit for the toggle jaws and eliminate the bending stress on the clevis pin. Plan A is to have a disk made with a bevel that fits the top of the chainplate and a hole that is slightly larger. This would be welded all around with enough weld build up in the hole that it can be redrilled out to the proper size to fully bear on the pin.

Plan B is to simply double the chainplate down to the base in a similar fashion. The strength isn't needed but, why not? A constant section would also make it easier to install slotted finish plates to cover and protect the bedding seal at the deck. The new piece would have a full penetration weld to the angle.
[FONT=&quot] [/FONT]

 

[Witzend]

Which plate is this?
So you believe the C-pin will bend or fail due to the length of the pin v the thickness of the tang? Is it bent now? How much wider is the toggle, pic please?

Any chance the glass will fail at the corner first, seem like a small surface area.

 

[Roger Long]

Which plate is this? So you believe the C-pin will bend or fail due to the length of the pin v the thickness of the tang? Is it bent now? How much wider is the toggle, pic please?Any chance the glass will fail at the corner first, seem like a small surface area.

That's the upper chain plate. I'm not concerned about the immediate strength of the cotter pin but standard practice is for at least a 5/16" thick chainplate with 1/2" pins that have 1/2" toggle width. Over time, the bending loads are working on fatiguing the pins so I'd like a closer fit. My current Nav Tech turnbuckle toggle ends have even wider jaws so the turnbuckle moves around when the lee rigging is slack and the jib sheets then get caught under them. I put seizing wire on to hold them in place but it's kludgy. I'd like to solve both problems by adding 1/8" thickness to the chainplate. May as well make it a bit stronger while I'm at it.

With so much slack in the jaws, there is also an often overlooked failure mode where the pin is held in the hole by the rigging tension and a sideways force applied to the fitting sheers off the cotter pin by sliding the toggle jaw over the clevis pin. I found one of my cotter pins broken this way when I took the rig down. There are some new clevis pins on the market with threaded heads, expensive, but I’m going to take a look at them.

This boat has more glass in the solid hull and deck than most built in the last 20 years. The hull deck joint is backed up by a full length aluminum toe rail which is distributing any strains well out from the chain plates. They were rather poorly installed with a thick layer of bedding compound but three decades of sailing, including five seasons of my hard driving, hadn't disturbed it a bit. I think you could lift the boat easily with just these two upper chainplates.

 

[Bill Roosa]

Consider the failure modes

If you beef up the chain plate to shroud attachment strength will that allow a demasting to carry the deck away? Better, INHO, is to have the shrouds break first and leave the deck and chainplate inplace.

 

[Roger Long]

Better, INHO, is to have the shrouds break first and leave the deck and chainplate inplace.

Yes, yes. We heard you the first three times I agree but that would be a crazy rational for not increasing the strength of my chainplates. Their attachment into the boat structure is a different matter and figuring that out on a 1980 vintage boat with all the difficult to quantify issues of fiberglass characteristics and multiple structural components makes that strength calculation simply impractical. As I said, you are fooling yourself if you think any real engineering went into this boat or most production craft.

In industry practice (see: http://forum.woodenboat.com/showthread.php?123813-Rigging-Design-Data&p=2809241#post2809241 ) the chainplates always have a higher factor of safety than the shrouds.

When I compare my chainplates to those of many boats where they are simply bolted to a fiberglassed in bulkhead, I don't expect to have them be one of the many things gnawing at the back of my mind when I'm out in the dark with waves crashing into the boat.

 

[Witzend]

Bill, I like the forward thinking, but too many variables to try to predict failure scenarios. Might better just overkill the design and not have it fail in the first place, and if it does have the bolt cutters ready.

 

[RichH]

Assuming all the deck, etc. to plate loads are in order ... I 'like' this plate geometry design - the stress bearing element 'pierces' the angle and is attached to the vertical face of the angle at the bottom, gives rigidity (I^3) to the whole base assembly, 'triangular' shape and the main axial stress component 'goes all the way to the bottom' !!!!
Full penetration welds - definite plus.
Piercing the angle for the main load bearing plate/bar and then full penetration weld is my 'ideal' vs. ductile + brittle failure; a lot of 'cowboys' would simply weld the butt end of the plate/bar to the 'top' surface of the angle .... then all the load 'hangs' solely on the weld integrity, a real bad idea in my mind. I see a lot of cranse-iron tabs butt welded on New England boats ... ????????.


My suggestions for 'improvement' to prevent potential premature 2-phase simultaneous crevice corrosion and fatigue failure would be: 'surface finish' (its become almost 'religious' for me)
The current industry preference is for all 300 series stainless re: fatigue & corrosion resistance is:
• No 'mill finish' in high stress areas but all surface flaws, All sharp internal/internal 'corners' removed/chamfered or 'blended' - I would deem this critical for the 'bar' portion and the bar-to-angle 'top' weld of the assembly.
• Surface finish to 10 micro-inches Ra (root mean square - average) ... one can get close to that by belt sanding then mechanical mirror-polishing; then, Electropolish if available and economical.
• Weld (bar-to-top-of-angle) - perfect weld: no gas blow-outs, inclusions, continuous pass with no 'laps', no weld 'splatter', etc. ... FULL penetration welds with no 'surface' intersticies ... then ground flush (before final mirror polishing where possible), ground welds also to be 'mirror-polished' (where possible). You typically need a 'code welder' to do this type of 'quality' - Code 3 or 8, doesnt matter although a code 3 pressure vessel (pharmaceutical, chem-process, microelectronics -- lots in NH) welder usually will have a better 'handle' on this. The ground and smooth polished welds and surface will be the key against developing crevice corrosion & fatigue - IMO / experience.
Fatigue can already be 'started' in the metal from the mill from rolling, etc. and the macromolecular cracks are the pathways for crevice corrosion to begin ..... surface finish will prevent/retard a lot of this 'duplex brittle failure' from happening.

• plates made from scrap from mill cert'd lot would be a plus - better and consistent "chemicals and physicals". Industrial Hx / Pressure vessel shop stuff.

Yes, definitely 'beef up' the area of the clevis connection. A clevis pin bearing load at the midspan is .... a joke; too much flexure and too much 'point to point' contact when the pin bends (ductile egging or fretting of the clevis bore/pin). Better to 'fill' the clevis space so that the pin is in mostly in shear and most of the stress is transmitted more near the 'ends' of the pin. But be sure to 'seal weld' the sister plates to the 'main bar' to prevent moisture intrusion 'dead leg' traps.
Changing the wire terminals to a correct span clevis would be another way and requires no 'fancy' metal work. Got Norseman or Sta-locks?

My material choice would be 316L as I have no experience with 2000 series .... and it usually takes some time for any 'new' alloy to eventually show it's 'warts'. Thats what happened in the 'early days' of 304/316, etc. nobody realized the fatigue and crevice corr. etc. problems until significant field failure began to happen many years later .... and geiger counters began to go 'haywire' :-o.

Other consideration - proper sized 'shoulder bolts' (shoulder length) for the attachments. Avoid or minimize having bolt threads inside any 'bore' - re.: crevice corrosion starting in the thread 'root'. Polish the bolt 'shoulders' if possible to remove mill type finish, etc. You'll appreciate this when your 'caulking' eventually leaks. (If they used properly sized shoulder bolts on keels there probably wouldnt be so many failures ... its nearly always the threaded areas that 'goes'.)
.................................................................
High latitude sailing is not something that I'd want to do in a boat that was assembled by 'guesses and gosh's' .... I can appreciate your concern to evaluate and refit all the 'important stuff' for this planned trip. 'The Rock' is not a kind place for an 'unprepared' boat.
I used to travel to 'The Rock' a lot in my late teens, and always questioned "why the ratio of women to men was so high". The same answer always was ... "well, the men mostly all drown". My retort ... "well doesnt anyone know how to swim?" was always followed by: "well Mr. Smartypants, it does one no good to learn to swim up here because once your in the water, in 15 minutes yer gonna be dead anyway, so why to go to all that bother to learn to swim in the first place". What I learned from this is that you need a 'totally bombproof' way to stay on top of the water, 'down' there.

 

[MSter]

Roger, Rich H & Maine Sail,
I try to follow along on the discussions because I find them informative and fascinating, however the recent posting on this forum titled Tragedy in Beagle Channel puts all these into very clear perspective !
To summarize, " you can not be too carefull or take too many precautions when it may mean the difference between making it out in 1 piece or not ".
Thanks for sharing.
Mster

 

[Roger Long]

Here is how I plan to increase the thickness of the upper shroud chainplates to reduce the bending stress on the pins and the annoyance of the toggle jaws sliding to one side so that the jib sheets snag under them. (Slide show so wait for it to start)

After welding the chainplates will be polished and passivated as Rich describes above.

Good links on passivating:

http://www.mmsonline.com/articles/how-to-passivate-stainless-steel-parts

http://www.imagineering-inc.com/white_paper/passivation.pdf

 

[Ross]

That ought to work as well as making new ones to those scantlings.

 

[jibes138]

I don't understand or I don't like it

Roger,
If your chainplate is already all the way out to the rail of your boat as depicted by the chainplate being at the deck to hull joint (correct me if I am wrong) why go to all this trouble with welding and gussets and stress risers? Why not just add chainplates like they have on the Pacific Seacrafts where they are just bolted directly to the hull and the shrouds come down right to the rail? Your hull is a thick solid layup so I really can't think of a better or stronger way with minimal concerns with assembly. When you bolt up an assembly like that with suitable backing plates the stress of the rig loads is basically through the friction of the joint that is in compression from the bolts through the hull. The load is not actually in shear through the bolts but they will take the load in event the torque on the screws somehow loosens. I really don't like the complexity of what should be a simple load bearing structure. Another option would be to have a machine shop make a block of steel that can be bolted to a structural member that is attached directly to the hull on the inside.

 

[Roger Long]

I really can't think of a better or stronger way with minimal concerns with assembly.

I can and it's just what I've got. This is the way it looks above deck:

See above in the thread for the below deck view. The chainplates would be virtually as strong with all bolts removed. There is a large flat bearing area under the hull flange and deck joint.

The chainplates themselves are fine. Stress lifting the entire boat by just two of them would be only 16,700 psi. The problem is the strain they put on the clevis pins by being too thin. Bending stress on the pin is added to the sheer stress. I could just weld on a disk to spread the load out over the pin but I then wouldn't be able to put on cover plates to seal them at the deck. If the additional pieces alone were taking the full weight of the boat, stress would be only be 22,222 psi. The two pieces together drop the stress to 9,500 psi so even a stress riser wouldn't be a concern. It won't actually be a stress riser because the large horizontal plate is the primary load transfer to the hull and everything pretty much ends there. The chainplate running through to the leg of the angle just stiffens things up and the two bolts through the hull side are suspenders backing up the belt.

The primary purpose of the bolts is to hold the units tightly in place and prevent movement that would break the bedding seal in the through deck slot. There was some leakage after 30 years as you might expect but very minimal evidence of movement.

This is a great design.

 

[jibes138]

What is that triangular shaped object in the first picture coming in from the right?
The design still seems to depend on the integrity of the welds as all the stress is loading the welds. Rich is correct about full penetration welds, this is hard to do getting under the right angle bracket where the chainplate intersects at an acute angle. It will be very hard to inspect and validate. Also the finished assembly should be heat treated after welding to normalize it.
I still think attaching the chainplates to the hull directly would be preferable. We are entiotled to politely disagree on this. Think about a Jordan Series Drogue. The preferred attachment is to bolt a metal strip equivalent to a chainplate to the hull on each side and then attach the drogue with a shackle. The chainplates on the Pacific Seacrafts and several other "bluewater" boats are set up like this. The drawback is the sheeting angle is reduced but you can always run a sheet inside the shrouds if necessary.

 

[tommays]

Roger


Looking at how you want to weld in the pin holes and then drill to size i am just not seeing that as good practice to get a true round hole through the complete part

If anything at work we would replace or drill and bush and weld in the bushing to ensure a smooth hole

 

[Roger Long]

What is that triangular shaped object in the first picture coming in from the right?

I'm not sure what you are talking about but a picture is worth 10,000 words.

The chainplate is bent and runs right through the flange with welds top and bottom of the flange and at the lower end. That's three redundant welds that need to break and two of them are where they are protected from water and the elements. Just the weld bead needing to pull thorough the hole in the angle provides a lot of strength. This is a very robust design.

 

[Roger Long]

If anything at work we would replace or drill and bush and weld in the bushing to ensure a smooth hole

Sure, that might be best in a highly loaded part but we are talking about pretty low stress levels here in a part where the original metal is already of sufficient area to do the job. With the original hole as a guide, there is no reason why cleaning up the weld bead won't leave a hole that provides proper mating with the clevis pin which is the whole object of the exercise.

Since the existing chainplate is in good shape and of sufficient cross section, I see more of a downside in enlarging the hole and increasing the amount of disturbed metal and welding by inserting a bushing. The welding in the hole is primarily to provide a water seal.

If this chainplate were a part that could get loaded up anywhere near it's fatigue limit of 30,000 psi, which might be the case if it was sized like many modern production boats, I might do things differently but the boat simply doesn't have the stability to do that. Even breaking the wire, say by wiping the rig off in a rollover, will only put about 23,300 psi of stress on this fitting in the portion either side of the pin hole after the modification.

 

[catsailor]

Amoung other things, I am an API certified alloy pressure vessel inspector.

Certainly, the suggested repairs could be done, and the suggestions are sound. The chance of getting everything just so and not introducing stress, a pocket, or a crack, aren't that good. Welding in the heat affected zone near an old weld that has seen high-strain is a bit dodgy; you'll have steel that has been stretched, and steel that has not been. Heck, it would certainly be easier to refab from scratch.

Get new toggles or replace the pins every year. Easy.

 

[Roger Long]

Welding in the heat affected zone near an old weld that has seen high-strain is a bit dodgy

Unless the boat has had a full knockdown that I don't know about (unlikely given her history), maximum stress that these fittings have ever seen in the weld effected zone is about 3000 psi so I'm not overly concerned about it.

Future stresses, even with hard sailing, will be so low that even adding any locked up stresses caused by the modification to the rig strain will be unlikely to get them into any critical range.

These and the other suggestions above would be quite valid for highly engineered and minimum weight fittings on a racing craft.

 

[jibes138]

Roger,
Look at the first picture in your original post. There is a triangular piece of stock on the right hand side of the picture. Is that used for something or was it just in the picture by accident? I was wondering what it was for in relation to the discussion.
This design is still putting the full load of the rig directly into the fillet radius of the hull layup at the hull to deck joint where the hull is flanged to accept the deck. This is a stress riser, I have no idea the thickness of the layup at that point but I still suggest it is less desirable than a plate bolted directly to the hull surface with a whole bunch of surface area and shear area before you'd see a failure. Most production boats run the chainplates to the bulkhead to spread the load, I know Island Packet laminates them to the hull on the inside with fibers and resin, it just doesn't feel right to me to run the chainplate loads into the corner like that.

 

[catsailor]

Unless the boat has had a full knockdown that I don't know about (unlikely given her history), maximum stress that these fittings have ever seen in the weld effected zone is about 3000 psi so I'm not overly concerned about it.

Future stresses, even with hard sailing, will be so low that even adding any locked up stresses caused by the modification to the rig strain will be unlikely to get them into any critical range.

These and the other suggestions above would be quite valid for highly engineered and minimum weight fittings on a racing craft.

Absolutly correct.

Which is why simply replacing the pins more often or replacing the toggles will do. I think you don't really have a problem. If you did have a problem, because the stresses were very high, a new fitting would be the correct answer.