Engineers? Opinions on this please...

[barnakiel]

Yes. I think both transfer of the load and spreading the load play a role. However, as materials age, possibly the portions that are transferred and spread are different.

And structure ageing is one factor. Possibly another one is whether the structure is utilized in the designed-in context. Our boat was not designed for repeated, extended, ocean passages. I believe few small boats were.

We learn where the designer and/or actual use went astray when we monitor damage in similar boats put to similar use. It does not take an engineer to quickly discover rudders in some present day cruising designs are not up to the job of ocean sailing. But nothing stops an owner of such a design to call up Jeffa and order an "identical but better" rudder.

And the same thing applies, imho, to any older boat asked to do things that are not truly in line with her age and design. We can see where/if these designs have their weak points and try to improve/correct things before that rudder falls off or that beam snaps. And yes we are likely overbuilding things. But we are not racing.

We have noticed earlier safety margins are considerable. I do not quite care if the margin is x5 or x8. As long as the structure is not too heavy or too big, I am fine with the improvement. Yes, I have seen a cracked beam on an identical design.

In our boat, the extra beam is about 16 pounds total. About as much as we recover by getting rid of one old genoa sail from our way too large wardrobe.

Fine. Hopefully.

Cheers,
b.

[Don C L]

Well, Barn, it sure looks good! What boat is that? Sure is perty inside there. I'm crossing my fingers my little art project will get me out this summer, hopefully this weekend dang it! Then next fall, out come the tools and drawing board!

[barnakiel]

Ours is just a generic Swedish doubleneder. Designed in '67 and this sample built in 1981. 3 t displacement and 26x9ft. About 330 sq ft sail but up to 500 sq ft when we fly the optional kite. Nothing fancy and about very basic inside and out. Very much like anything else of the same age built in the UK or US, just lighter (less volume, less displacement, less hull thickness, same LOAxBeam otherwise). Swedish boats MUST sail upwind or else it is the rocks. Hence less displacement.

Cheers,
b.

[jimbunyard]

Hope no one minds a little more discussion on this...

It has been sort of mentioned, kinda sideways I guess, but to state it baldly, it's all well and good to set up static models and test them for load bearing capacity and stress points, but in reality, the only time the structure will see those forces will be in port with the sails down...though that's not to say that the information's not still good to have as a reference or benchmark.

While sailing, the forces will be much different, and cyclic. On one tack, the windward side will be pulled up, while the leeward side will be slack, and will tend to be pushed down resulting in an 's' shape being induced to the beam athwartships...and this is on a beam reach with both main and headsail up. On other tacks, with only a headsail or only a main, not only is an 's' shape induced, but twist or torsional forces come into play. And they're reversed everytime you come about...

The point being, of course, that it's not very prudent to isolate one structural element in a system as dynamic as this one is...pardon me for preaching to the choir...

But to discuss Don's 'beam' a little more specifically.

1. The 'grain' is running the wrong way.

I don't think so. If you look closely at the picture below, you can see that it appears Don used cabinet grade plywood (looks like oak), which has a very, very thin layer of veneer over the 'filler' plies. What I see in the picture is that the schedule is vertical veneer, horizontal ply, vert ply, h ply, v ply, h ply.

At any rate it doesn't really matter, used in this manner the plywood should be considered as a monolithic material, much as one would consider using a sheet of G10 without taking into account the 'grain' orientation.

I also considered that it may have been meant that the plywood should have been lain flat. That would be much weaker; even if the sheets were oriented 'correctly', bent to fit the cabintop and laminated properly, like a roof beam or a ship frame, 2/5's of the strength would be lost because the plies would lie across the grain.

2. Use or non-use of aluminum; whether it could contribute to the failure of the beam.

It is not clear to me why the use of materials with different strengths would automatically predispose the structure to failure, or early failure. Composite materials have been used successfully for thousands of years. Some clarification and/or explanatory information would be much appreciated.

While I feel that, as built, the structure is perfectly adequate for it's intended use, it could be even stronger, longer lived, and likely more resilient with a more nuanced use of the same, or nearly the same, materials. (Of course solid, clear teak or iroko or angelique or ipe or any of the other tropical superwoods would work alone, but that's another topic.)

I like your idea of being able to remove the beam (and am also aware of the difficulty you'd have trying to install and clamp the beam if you'd laminated it all in one piece) but with the aluminum on the inside you're not using one of the key strengthening features of your design, which is the stiffness produced by the clamping friction between the wooden beam and the wooden bulkhead. Perhaps a better solution would be, as others have also suggested, a plate on each side, with through-bolts clamping the 3 wooden (no interior aluminum plates) elements together (plywood is perfectly adequate in my opinion).

If you were to leave 1/2-3/4" of the lower perimeter of the wood exposed, you'd have adequate area to mechanically fasten a 3/4" fascia board, let in on the inside for the bolt heads and nuts, to dress things up a bit.

A couple illustrations, the red dots denote 1/4" fine thread bolts, the blue 5/16". Torque is important, I'd go with maximum allowable for the given size and check it after two weeks. The gray is the aluminum with an exposed area for fastening trim...

Obviously just some food for thought, I'd continue sailing it like it is...

[NevisDog]

I hope I've understood correctly. Taking Jimb's comments as listed:
0. There is no other way to analyse this structure, except isolate it and pop a load on top. If you can come up with any other analytical method I'd love to hear it, but it has to be do-able within our lifetimes, and I'm getting old.
1. The grain is the wrong way for a solid beam - might work for a hollow box beam, not this one.
2. FEA is an excellent tool (in the right hands of course) for highlighting stress-raisers within a complex structure. I think this is one area we should believe the computer, as it's not intuitively obvious to me (except that the plates are also oriented the wrong way).

So far as I'm aware, the most efficient (strongest) beam is still an I-beam. Composite is great, so long as we use the weaker material as the core, or web (balsa, ply, timber, etc) and the stronger material as the upper/lower skins, or flanges. Plating each side seems pointless, when the steel, or alu, or glassfibres, should be added to top and bottom horizontal faces of our plywood beam, to add strength where it is most needed - not in the vertical web, where it's pretty darned useless.

[jimbunyard]

Originally Posted by NevisDog

I hope I've understood correctly. Taking Jimb's comments as listed:
0. There is no other way to analyse this structure, except isolate it and pop a load on top. If you can come up with any other analytical method I'd love to hear it, but it has to be do-able within our lifetimes, and I'm getting old.
1. The grain is the wrong way for a solid beam - might work for a hollow box beam, not this one.
2. FEA is an excellent tool (in the right hands of course) for highlighting stress-raisers within a complex structure. I think this is one area we should believe the computer, as it's not intuitively obvious to me (except that the plates are also oriented the wrong way).

So far as I'm aware, the most efficient (strongest) beam is still an I-beam. Composite is great, so long as we use the weaker material as the core, or web (balsa, ply, timber, etc) and the stronger material as the upper/lower skins, or flanges. Plating each side seems pointless, when the steel, or alu, or glassfibres, should be added to top and bottom horizontal faces of our plywood beam, to add strength where it is most needed - not in the vertical web, where it's pretty darned useless.

0. Certainly. Actual destructive testing with a scaled model.

1. Grain is irrelevant in this use of plywood, as previously stated. Would be interesting to hear the reasoning behind "might work for a hollow box beam, not this one", and would like to know, precisely, what grain orientation is 'correct'.

2. No questions with that, but, intuition or not, it is obvious to me that the highest stress areas are going to be where the beam's support terminates, in this case on both ends, which, oddly enough, is exactly where E/V's FEA showed them to be. Not claiming any special insight though, just the results from designing and building, quite literally, thousands of beams (or should we say beam-like or beam-using structures?)

If by 'efficient', you mean 'strongest per unit mass', then I believe you're correct.

That, however, has nothing to do with the question in hand. The strength of any beam is determined by it's dimensions and construction; the upper, compression (if the load is directed downward) side, must be intimately(and strongly) connected to the lower (ditto) tension side for the beam to work. The strength of the beam is directly related to the depth (distance from the compression side to the tension side) of the web.

To not take all the components of the structure into account and analyze them as a whole is a mistake.

A homely example;

You've got two 2x6x16' planks and 12 3" screws, and you want to use them to cross the creek. Do you lay the boards flat and screw them together, or make a 'T' by screwing one board through the flat (6") to the edge (2"), thereby mysteriously increasing, I don't know, 5?, 10?, times the stiffness of the two boards? And with no lower 'flange' at all...


It is interesting that you ignore (disregard?) what is likely the most important point, that which could cause premature failure, which is the cycling and reversal of forces the structure will undergo when sailing...

[barnakiel]

How could an equivalent strength but minimum weight beam be built? Carbon/epoxy monostructure vacuum, or something else?

Or is there a way to get rid of the beam if we rebuild the cabin top in carbon laminates? Etc.

Or are we bound to accept the beam absolutely necessary?

b.

[Cadence]

I can't keep up with all the genius input. His original laminated plywood looked great and more than enough for that short span. Has any seen what is used if you want to span a two car garage without a center support, for example. It's a laminated beam of plywood not much thicker than his beam. I would leave out the aluminum it just interferes with the plywood bond. A nice adequate job however.

[NevisDog]

Originally Posted by jimbunyard

0. Certainly. Actual destructive testing with a scaled model.

Okay, go ahead and do it, let's see how long that takes, but remember you said it should not be isolated from the surrounding structure so you'll have to build a scale model of the whole foredeck, bulkheads and keel. Let us know when it's ready for testing.

We have a solid beam, quite a nice one. Why discuss a T-section made of two plywood planks? We have headroom to consider here so no deep T-section. A solid beam is the same as solid ply - if you had to lay solid ply and walk along the top, would you cut it into slivers the same width as the ply thickness then glue them together vertically, or just lay the ply as designed? Let's use some common sense here.

Too much theory - let's use the information Eigen has so expertly provided.

[NevisDog]

Originally Posted by jimbunyard

It is interesting that you ignore (disregard?) what is likely the most important point, that which could cause premature failure, which is the cycling and reversal of forces the structure will undergo when sailing...

Oh what absolute rubbish! What else is a safety factor for? Just for fun? Just so we over-build everything? This is too silly. SS wire I checked when assessing the loads has a safety factor of 6 applied - do you want it even larger????? We don't need theorising over this simple beam, we need answers to simple questions - if you cannot provide those simple solutions then why waste our time theorising?

[jimbunyard]

Originally Posted by barnakiel

How could an equivalent strength but minimum weight beam be built? Carbon/epoxy monostructure vacuum, or something else?

Or is there a way to get rid of the beam if we rebuild the cabin top in carbon laminates? Etc.

Or are we bound to accept the beam absolutely necessary?

b.

I'm sure something equally strong could be made with very minimum weight, using carbon fiber or other exotic materials, just look at the recent A/C cats. I'm not so sure it would be particularly cost effective...

An interesting question. Surely a solid fiberglass beam/skin could be built up (covering all fiber-reinforced plastics with that one phrase); I wonder, with such a short span, what the final thickness to reach a comparable strength would be. I'm guessing somewhere on the order of 1 to 1 1/2", the problems would probably include rigidity and integrating such an extreme thickness into a comparatively thin surrounding skin without creating hard spots and joints...not saying it couldn't be done but it certainly would be more involved that a relatively simple beam.

I'm assuming the primary goal is maximum headroom through the passage way. For strength, easy of execution and aesthetics I might try to emulate some of the ideas used by Robert Maillart in bridge design...he's done some remarkably beautiful and functional work with arches and beams, illustrating their necessity, but also the possibility of minimizing their ---obtrusiveness?---by a careful analysis and directing of the relevant forces.

It seems that some of these forces might be analogous to those encountered in a deck-stepped mast, and an adaptation of Maillart's marshalling of them might be apropro...

1933


1901, note the difference in clearance compared to the bridge in the background.


1933

[jimbunyard]

Originally Posted by NevisDog

Oh what absolute rubbish! What else is a safety factor for? Just for fun? Just so we over-build everything? This is too silly. SS wire I checked when assessing the loads has a safety factor of 6 applied - do you want it even larger????? We don't need theorising over this simple beam, we need answers to simple questions - if you cannot provide those simple solutions then why waste our time theorising?

Don't think anyone's forcing you to read them...

That you don't understand something doesn't magically turn it into absolute rubbish...

Other than suggesting you brush up on your reading comprehension skills, there's really nothing else to say ... I'll remember to not engage any further with you...whatever time management issues you have are your own responsibility.

[Don C L]

Well no need to ruffle feathers over my little arch. Eigenvector is plugging in some more numbers I gather, which is more than generous of him from my standpoint, but for all of us too, to get a bit of education how the materials, in the form I put them, will respond.

When I rebuild it (unless it comes back with spectacular numbers) I think I will go with even more of an arch, because, yes, headroom there was one of my primary considerations. I actually also considered an arched (to mach cabin top curve) SS box beam on the cabin top to span the bulkheads, bolted down through the laminated arch below.

Anyway, as usual I really appreciate all the input and discussion, I certainly have learned a lot! But no reason to come to blows yet!

[Don C L]

BTW, I am now not sure what the point load will be but I thought, or estimated, up to 10K# might be about right (now) which is why I began to question the thing. And that load is applied to a 4" x 6" pad on top, if that helps. I will say I am getting very tempted to build a copy and take it somewhere with a big press and see what happens. Maybe when summer is over.

The thing is, it's not just the arch sitting there, the bulkheads, the attachment to the bulkheads and the boat construction as well are taking the load, and figure into the strength as well, right? Additionally, unlike a simple I-beam, the chevrons will have to go out, pushing the bulkheads apart, in order to collapse, no? So now I am wondering, by cutting the original arch out, have I weakened the cabin top to the point that it might actually push the bulkheads apart a bit? I can't imagine it, but then I can't imagine 5 tons on that thing either. The cabin top is about 3/16 of layers of roving. I wonder what it would take to rip it apart? But perhaps a key part of my truss is missing, and there is no way to apply flying butresses!

[Cadence]

Originally Posted by jimbunyard

I'm sure something equally strong could be made with very minimum weight, using carbon fiber or other exotic materials, just look at the recent A/C cats. I'm not so sure it would be particularly cost effective...

An interesting question. Surely a solid fiberglass beam/skin could be built up (covering all fiber-reinforced plastics with that one phrase); I wonder, with such a short span, what the final thickness to reach a comparable strength would be. I'm guessing somewhere on the order of 1 to 1 1/2", the problems would probably include rigidity and integrating such an extreme thickness into a comparatively thin surrounding skin without creating hard spots and joints...not saying it couldn't be done but it certainly would be more involved that a relatively simple beam.

I'm assuming the primary goal is maximum headroom through the passage way. For strength, easy of execution and aesthetics I might try to emulate some of the ideas used by Robert Maillart in bridge design...he's done some remarkably beautiful and functional work with arches and beams, illustrating their necessity, but also the possibility of minimizing their ---obtrusiveness?---by a careful analysis and directing of the relevant forces.

It seems that some of these forces might be analogous to those encountered in a deck-stepped mast, and an adaptation of Maillart's marshalling of them might be apropro...

1933


1901, note the difference in clearance compared to the bridge in the background.


1933

They resemble the OP's original don't they. I think he did an aces job.