Carbon spreader wall thickness

[Ja guar]

I was looking at the specs of carbon spreaders (Forte) and noticed the wall thickness for all sections was 0.12 in.

Using the following numbers:

E = 19,000 ksi
Wall thickness = 0.12 in
Iyy = .36 in4
Spreader L = 54 in
K = 0.7

Based on Euler buckling, this works out to about 47,000 lbs max. i haven't worked out compressive resistance since this is a highly unlikely mode of failure.

For a shroud load of 2,000 lb and SF of 6, this works out to 12,000 lb factored load, so it looks like the carbon spreaders are much thicker than required. By my calculations, wall thickness could be as thin as 0.025 (1/40) in thick and still have sufficient resistance.

My question then is: Is this thickness due to thin wall (local) buckling issues or is it because that is the minimum practical layup thickness? Is there another failure mode I haven't accounted for? Or have I just messed up in my calculations?

 

[Eric Sponberg]

It is difficult to comment on your observations because we don't know the overall dimensions of the spreader. Typically, there should be a certain wall thickness in relation to the major and/or minor diameters so that the spreader does not buckle due to improper laminate schedule. That is to say, while you might be able to design a laminate that is indeed 0.025" thick and made only of unidirectional fiber (that would be about 2 layers of 300 gsm UDR), it won't hold together for very long because there is no cross fiber or fabric. You do need some off-axis fiber in the laminate both sides of UDR to keep the UDR in column. I would caution you about the modulus of elasticity = 19x10^6 psi--that is awfully high. That might be achievable in the lab with specially prepared samples of UDR only, but it is not a realistic number for a finished production laminate incorporating a mix if fabrics. A finished laminate modulus on the order of 10x10^6 to 12x10^6 psi.

A suitable fabric might be just a 0°/90° cloth fabric with the warp (0°) fibers running axially and the weft (90°) running circumferentially. A 200 gsm cloth would be about 0.008" to 0.010", and you might need 2 layers so that there aren't any gaps or voids in the laminate. Three layers would likely be better, so the wall thickness would be 0.024" to 0.030". A 0°/90° cloth fabric accomplishes axial and circumferential lay-up all in one go. A cloth only laminate may have a modulus of 6x10^6 to 8x10^6 psi. Figure your column buckling behavior accordingly.

So if a simple cloth fabric does the trick at least as far as overall laminate stability is concerned, then what is a suitable minimum thickness in relation to the overall dimensions of the spreader? For a 2" x 4" spreader (to pick a size out of thin air), a wall thickness of 0.120" does not seem out of sight, all things considered.

Eric

 

[Ja guar]

Thanks for the response Eric.

I guess the real intent of my question was: By my calculations carbon spreaders are considerably overbuilt with regards to wall thickness, so why not use FRP for spreaders rather than carbon?

The main advantages of FRP I would think is that one could side step the corrosive and ductility issues associated with the use of carbon.

I’ve searched the internet and cannot find reference to FRP, GRP, or fiberglass spreaders. I assume people much smarter than I have resolved that it is not a suitable material for use in spreader construction, but I am unable to come up with reasons for why it is not.

In my calculations, using ¼ in SS wire (breaking strength of 8,585 lb) with a 10 degree shroud angle, 5 degree spreader upsweep and SF of 6, the factored shroud load would be 4,460 lb. An FRP spreader using a conservative E of 2x10^6 and dimensions listed below has an Euler buckling strength of 5,000 lb. So with regards to Euler buckling, FRP seems adequate.

If it's a matter of bending stiffness, people still use wood spreaders. While I haven't compared the bending stiffness, I assume a hollow FRP section would be superior to solid wood. Does it have to do with eccentric loading, P-Delta effect? Again, just trying to figure out why FRP doesn't appear to be used in spreader construction.

For the record, the dimensions of the spreader in question were: length = 54 in, width = 3.75, depth = 1.36 in, wall thickness = 0.12 with an ellipsoid section.

 

[Eric Sponberg]

Actually, you are probably correct that fiberglass would be a suitable material alternative for spreader construction, given the numbers that you are using. But the reasons that you don't see fiberglass spreaders are, in my opinion:

1) Aluminum mast builders do not like gooey fiberglass, and they already have extrusions suitable for aluminum spreaders. They'll stick with aluminum as they are already tooled up for it.

2) All composite mast builders use carbon fiber and they minimize the use of fiberglass. They use a lot of carbon and hardly any fiberglass as all. So they are going to go with what they know and design and build everything out of carbon fiber. It would be more of a bother--another set of materials to contend with--to use fiberglass for relatively small structures. Carbon is stronger and stiffer, so they'll go with that.

That's my take.

Eric

 

[Ja guar]

Thanks again Eric,

I'm sure you are right about the preferences of mast builders mast builders. I was just wondering if there were structural issues I hadn't considered.

In my case, I'm converting the frac rig on my 34' boat to a sweptback configuration. Moving the shrouds and chain plates aft and outboard to almost the toe rails. Of course this will require considerable interior rework.

I've measured and can use the lower spreaders as new uppers, but I will need to buy or fabricate new lowers. If I choose to fabricate, carbon will have the issues I mentioned as the mast is aluminum. So I was thinking of fabricating new ones with FRP as an alternative, if it was feasible.

Now I have the answer - yes, it is feasible.

My only remaining question is how to attach the things - I haven't given it too much thought yet. I had thought about bonding, but apparently epoxy doesn't bond well to aluminum in the long term - it eventually delaminates.

Many thanks again Eric.

 

[Eric Sponberg]

Roger that. If in your redesign of the rigging to swept-back spreaders, are you also thinking of removing the backstay? If you are, then you will need reverse diagonals to help stabilize the mast. Just a thought in case you are considering eliminating the backstay.

As for spreader attachments to the mast, you should have double-pin attachments for each spreader. These would be pins oriented vertically, one pin at the leading edge of the spreader, the other pin at the trailing edge. These would pin the spreader through new base brackets on the side of the mast that would be bolted, screwed, or riveted thereto. So when you make the new spreaders, you'll need to make or buy new spreader attachments. You are correct that you should not bond the new spreaders to the mast.

Eric

 

[Ja guar]

That is the second time someone has mentioned going with a B&R type rig, so there must be something to this.

The boat in question has a double spreader rig with discontinuous wire rigging.

My goal is to make the boat easier to short hand so You are correct that my plan is to ditch all of the following: check stays, runners, and top mast runners.

I`ll do some research on the B&R rig, but I`m trying to figure out why the reverse diagonals increase mast support.

There is no sign that I can see of reverse diagonals on the sail plan of a 1D35 which has a similar setup to what I envision. My previous boat was a lightweight 27 ft frac with permanent, runners, and about 20 degrees of sweep. An owner (sailmaker) of a sistership went with a similar setup to what I`m contemplating on this boat. He did not install reverse diagonals, but threw in a mechanically adjustable forestay as well (which I think is good idea).

I have considered jumpers to support the top mast and maintain forestay tension under mainsail loading, but I`m leaning towards an adjustable forestay.

My concern with an adjustable forestay is I think that with a fair amount of forestay tension dialed in it would load up the shrouds something fierce (possibly on the order of 2-3 times forestay tension, assuming leeward shrouds are slack).

 

[Eric Sponberg]

In a normally stayed rig, the spreader bases and their rig attachments are considered pinned connections. That is, they can allow the mast to rotate slightly in the plane of the spreaders at each connection. The spreaders move in concert with the mast wall--ever so slightly, but they all still rotate a bit. By putting in reverse diagonals, the spreaders are constrained from moving practically at all because they attach at the underneath side of the spreader tips to balance the regular diagonals above them. The reverse diagonal wires also inhibit rotation of the mast where they are attached at the spreader bases. This effectively makes the spreader bases fixed points rather than pinned. In an ideal engineering world, fixity increases the stiffness of the mast by a factor of 4. In reality it is probably more like a factor of 2 or 3. But that is really significant, and in most cases worth the extra cost of a few wires and fittings in favor of a lost backstay.

Eric