Skin over frame sandwich, low cost high performance plywood carbon xps

Hello to all, and a huge "thank you" for all amazing contributions. I have been reading, lurking and absorbing as much of your shared knowledge as i could, for the past 2 years or so. I am currently tinkering an a franken-trinardo project with modifying a 707 Edwardsson into a typical trimaran main hull shape, and refurbishing old ply tornado hulls.
But to be honest, I get the most joy out of learning as much about design and construction, as possible. There for, I am also trying to design a few boats from scratch. Among other a 12' RIB and a 32' bridgedeck cat. And the most fun so far has been trying to get to grips with the construction bits.

To that end, I would love to get some feedback on a few thoughts/ideas, and maybe get a good discussion going. So... I've had an Idea (I know... very unique, haha ;-) Joking aside, it is very much just a rearrangement of existing ideas. And it is not for a specific project rather than as a possibility for a category boats. So here it goes.

In design considerations of relative small boats I have generally come to the conclusion, that some parameters tend to be extra important, relative to larger boats. 1) Weight, handling/trailering, 2) Cost, market for smaller boats/designs is generally much more price sensitive, and especially important for my self, 3) Performance, small design are often used more for sporty use, whereas larger more often for steady cruising or live aboard. 4) Ease of build, very important for DIY.
In summary, it is paramount with a design/construction method that is dirt cheap, is very low weight, and very high performance as well as being very easy to build! Hahaha... I know, I can hear the laughter ;-) But just wait a minute.... How about Skin "over" frame. Yes, I really meant that. And I am very sure, I am not the first. I just haven't heard it put in exactly those words.

The basic idea being: Start with a traditional plywood- or strip plank on frame, but only engineer that structure to say 1/2 the required mechanical properties. Then fill the voids with XPS, and sheet the whole shebang like a regular core, and dimension the skins and core to take the other 50% of the loads. Stacking sandwich panels isn't that uncommon. And I'm quite sure, that nesting 2 or more "semi separate" engineered layers/structures, isn't either.

Why oh why.... those unpractical, unrealistic ideas??? Well there could be quite a list of pros, and not that many cons.
Pros of plywood, or strip, plank on frame: is considered a quite practical method. Low jig waste, no plug needed, economical, fast, easy and forgiving. Also it is considered a very light option, despite being insensitive to dings and general ware and tear, compared to a thin skinned foam sandwich. Especially smaller boats seem to carry a skin thickness penalty for acceptable impact resistance.
Cons, would include 1) limited stiffness, with the use of slightly thicker than necessary ply to mitigate, but with added weight. 2) No insulation, troublesome both in hot climate, and in cold (like Sweden) with ridiculous amounts of condensation as a result. 3) Cluttered interior with lots of visible structural elements (stringers).

By combining the 2, you could get the best of two worlds.... sound impossible, but is it really?

Rough case outline. 30' cat. Usually about 9mm ply. Reduce to about 4-6mm. Modify frame to be stiffer, and level with 40mm XPS (finnfoam 300), perfect thickness for insulation purposes. Compressive 300kPa, shear 300kPa. (Same shear properties as 10mm Divinycell H60). Skin outside a tad heavier than usual abrasion layer. Maybe about 400gsm (12oz/sqy), and the inside with about the same. 300-400gsm. Impact resistance almost the same. Insulated, ultra stiff. Dirt cheap. XPS 40mm is about 6.5USD/m2. Uncluttered interior.

Result. Insulation weighs about the same, whether or not it is part of the load carrying structure. The outer skin is not that different, only now slightly heavier and part of the load carrying (not only abrasion protection and wood sealing). Save half the weight of the ply, (4-5mm okume = about 2kg/sqm), and add about 800gsm for inner-, and 200gsm for extra outer, fiber/epoxy skin . Resulting weight savings. Insulation +-0kg (32kg/m3), + 2kg/m2 ply, - 0,8kg extra glass/epoxy inside and - 0,2kg outside. Result roughly -1kg/m2 hull area, or 100-150kg per 30' cat. Usual ply about 7,25kg/m2 inc coating, sheating, frame and insulation. Skin over frame about 6,25kg/m2. That is quite a sizable difference!

The main challenge would probably match the stiffness of the two structures (Using E-glass epoxy in skins (sandwich component being to stiff), and a very stiff frame design (being to flexy). Also a challenge avoiding hard spots, and solve a few details in the load transferring between the two nested structures. Maybe a version of "egg crate plywood" like an over sized quadratic ply honeycomb filled with XPS. Weight difference to a standard frame I don't know, but I guess actually slightly lighter. With router, such ply and xps should be easy to cut to a near perfect puzzle.

Would love to hear your thoughts. Hit me hard! ;-) haha... Cheers!

 

Sounds good to me.
Carry on.

 

Aside from a somewhat komplex assembly, which I Believe could actally be manageable, i have 3 Main concerns. 1) Different stiffens characteristics of the two layers. 2) Asymetry in a sandwich is raerly of any good. 3) the whole idea is based on another nono, mixing different materials.

And I belive it is here a lot of boatdesignnet-contributors and engineers differ from ordinary people. We love a good chalenge ;-)

First, in using the "plank on frame" inside the sandwich, it automatically becommes a mould. Ply on a 40mm "primary frame" is built eg on a strongback, glassed on the outside, and then turned. Then the more detailed frame (eg ply eggcrate or some alternative) is layed inside and glassed. After that, bulkheads are installed, and hull halves, or hulls and decks, are joined. It is almost as if the sandwich is left inside a female mould, substituting the outer skin.

 

Questions I would love getting feedback on. 1) Could the ply on frame get similar or enough stiffnes from the egg-crate stringers working together with the XPS? By filling the space, the XPS should lock the stringers together, forming a semi-closed box shape. And the XPS would stiffen the ply on frame by taking up and transfer compression and tension loads.
1b) At the same time, the stringers would help with the shear forces. 1c-2) If need be, the stringers could be reinforced with battens, or carbon, on the inside, to increase stiffnes (and make the strenght more symetrical).
2) sandwich symetry.... hmmm, maybe a later discussion.
3) The most accute problem is probably mixing materials. I have tried my best, but still struggle to get clarity. To me it seems elongation at break (or rather at non reversable deformation, is the best measure if skin materials are compatible. But unfortunately it is very hard to find those numbers for different woods/plywoods. So far I have found about 0,8% for marine ply, okume, about 1,5% for bamboo ply. And at the top of my flawed memory, about 1,5-2% for douglas fir. For a carbon epoxy laminate, the numbers should be around 1,2% for T300 or T700 fibres (among the most common). For E-glass, the numbers seem to range between 4.5% to 6%.
Is there another usefull measurment, eg Youngs modulus or something else that could be used to determin elongation at break? Ply seem to be a good match to carbon. On the other hand, the 40mm sandwich stiffnes with carbon would be rediculous and a complete missmatch with the flexy ply on frame!
Is it possible to compensate for different stiffness characteristics, by using the position of the material relative to the panel-center, and relative to other materials in the same side. Eg, if ply has slightly lower elongation at break, placing carbon on the outside, compensates for the slightly higher 1,2%, to some extent syncronizing this charecteristic.

 

Another way of looking at it, could be as a more traditional sandwich, starting with 3mm to 5/6mm strip planking (eg Douglas Fir) to get the first outer layer alas the mould. Then adding +- 45deg Biax carbon (about 200-300gsm), and finishing with vertikal UD och 0-90deg aramid/kevlar. The Aramid and wood elongation at break would be similar, and the majority of the longitudinal loading would be taken care of by the wood. The carbon would ad some strength to the would, but mostly torsional strength and stiffness.
The big question I have here is. since carbon has almost half the elongation at break parallel to fiber, compared to Fir. If laying it only at +- 45 deg in epoxy (about 6%). Any ideas about the stiffness of the carbon skin compared to the wood along the wood grain axis? Spontaneously (but rarely correctly) I would guess that the stiffness would be somewhere in between the individual component values (1,2% carbon, and 6% epoxy resin), maybe even close to that of the wood and Aramid (2% approx).
the strip planking would be 3 times stronger longitudinally than the ply. Making it possible to go with a lighter outer laminate, saving even more weight. And to try and make ptoduction more rational, the subframe (egg crate ply) might be substituted for strips of E-glass filing the gaps between small XPS squares (somewhat like there is square cut Divinycell). The fiber strips would be infused, and form a sort of coarse square honeycomb filled with XPS. Adding shear strength for 30' and upwards designs, but slightly worsening the insulation properties of the foam.

The main question being: is it realistic to match materials with different stiffness, by using the relative angle of fibers, to match the direction specific stiffness?
And maybe strip planking is slightly less brittle compared to ply, making it a better choise for handling impact resistance?
I expect such a layup, might shave another 750-1250gsm or so, of the previous tally Fast approaching a "dirt cheap, low tech" super performance construction ;-) Well carbon may not be so low tech, but actually not much more expensive, considering 300gsm Biax at about 9USD/m2 comparing to about 500 gsm E-glas at about 5-6 USD/m2 but consuming 4-5USD more of resin. Inc vacuum consumables, I estimate it would be possible to build with carbon aramid outer, and all carbon inner skin, Fir strip plank E-glas web, Finnfoam f-300 40mm XPS core and epoxy resin for just over 40USD/m2 inc tax.

 

It sounds to me like your making a simple process as complex as you possibly can !
Then you want to use expensive materials like Carbon and Kevlar to compensate for the cheap **** XPS. And then expensive resin like epoxy so the cheap **** doesn’t dissolve.
By the way, Kevlar is a ***** to cut, a ***** to lay up and a bigger ***** to repair and, AND it’s hydroscopic.
Note that building the hull and deck is only 1/3 the job of a finished yacht. KISS Keep It Simple Stupid.

 

You're mostly right. Exept for that carbon realy can come close to the cost of E-glass because the lower resin consumption. And it's properties seem to match better with those of wood. Carbon, has other valuable properties, very low fatigue damage in long frequent repeated loads.

My reason for this thread was "to not do the wise thing and keep it simple" but to take an old idea, rephrase it, and ask for help in complicating the cr*p out of it. To explore the limits of 2 combined building methods, and a handfull of materials. Ignoring build complexity for a moment. Sadly I am lacking in engineering knowledge. But on the others hand, that gives me the excuse to start posting and take part in this great community

 

Since you're an explorer check out basalt fibre. Looks good for sheathing foams as it’s compression strength is better than Eglass and Kevlar and reasonably priced.
I’m no engineer just an ex glassie!
My personal view is if you like wood and working with it then use it together with epoxy and all will be well.
If you like foams and composites then unless your going for ultimate performance then pvc foams and vinyl ester resin is mor than adequate. If you want to play with carbon and epoxy then use it where you get max value- the mast and boom,prodder and maybe the beams.
That’ll be 2c please

 

Not being an engineer, you need to start making some samples.
You have so many variables it may be difficult getting useful opinions.

 

You must have a lot of money to spend.

Experimenting with cf is spendy.

 

Hey guys, thanks for the feedback! Maybe easier with a case and some illustrations. Thought I might share some of what I found in my research.
One of the ideas I'm working on is a bridgedeck cat, just shy of 33 feet. Good case for the proposed construction. Even if it is absolute lower end size for a cruising cat, it still has quite some accomodation. And thus insulation would be a must in northern Europe. I'll try and attach a file.

 

Design brief: Light enough to be very sporty (wicked fast). Trailerable if disassembled. Cheap. Safe. And still, a full fledged cruising cat with full standing headroom 1,90-1,95.
Now I know that dry weight specs differ greatly in how boats are equipped (spartan or extra everything), and also how much is calculated in the specified weight. But a few comparisons. Aventura 33 (ink 2x14hp inboards), 4500kg. Obviously slightly on the chubby side. A more sporty boat by Eric Lerouge, Ladyhawke 33, is 4100 inc 2x30hp inboards). That's better especially considering a nice trim interior as well as a slightly larger beam. 5,5m vs Aventura 5,3m. A production company known for weight consciousness is Fountaine Pajot. The quite recent Mahe 36 is larger but still weighs in at a respectable 4950kg. The earlier models tended to be more spartan and slightly less voluminous, as well as built in foam sandwich, even in the late 80's. The Maldives 32 was a meager 2250 dry weight. That's almost sensational! Comparing a few beautiful designs available at the moment. I like the Eclipse by Woods, listed at 2750kg Strip plank as well as foam sandwich. And then we have KHSD 30 half bridgedeck at 1360kg and another absolutely beautiful design, the 36' cylinder mould at 2450kg. A full bridgdeck 32' would probably come in at about 1800kg. It seams the KHSD are designed extreamly lightweight, and not only because they are somewhat spartan in their base configuration, and probably don't have motors included. But really a lot depending on the construction and design. Very interesting. Even a performance design like the Dazcat 995 is said to have 2600kg dry weight. It's no wonder ply or strip plank are so highly regarded and considered light weight. Would be a shame to throw that away, only because there is a need for insulation/foam. So.... 1500kg all up shouldn't be a problem, right ;-)

 

So what problems occur when going on a diet, (except from grumpiness and headaches? ;-) Well, obviously a starting point would be to incorporate foam into the construction. Otherwise it'll just be dead weight. And the good thing is, it makes wonders for the impact resistance. Well, obviously a 9mm glass sheeted ply isn't bad either. But going sandwich means the outer skin will be about 40% the thickness of a paneled non cored GRP layup. Hmmm.... sounds sensitive. Ever heard about dings and dent's in sandwich boats? The idea of leaving some of the ply the better impact resistance doesn't seem to far off. Especially since it's considered a lighter construction method in general for smaller boats. Ok, so what if there was some evidence of the merit to this concept.
I found 2 really interesting papers on the subject. The first I'll link looks at non cored panels. Some eye openers there. Later I'll post one on cored panels also.
https://www.vtt.fi/inf/pdf/publications/1997/P317.pdf

 

The other paper is at this link.
https://www.vtt.fi/inf/pdf/publications/1996/P281.pdf

All in all there are some really nice conclusions.
The PVC H80 foam has fantastic impact and energy absorption qualities. That's a known. Only bad thing. To get somewhat descent insulation, you'd have to add at least 24mm. That'll be 41€/m2 + tax, pleeeeaaaase :-O
Ouch! So what if XPS could be used. Actually I heard some where that some XPS was approved by some standard. Don't remember if it was Veritas, Lloyd's, ISO or if I just imagine things. Anyway as mentioned earlier. XPS f-300, would be about equal in shear to H60 PVC if about 4-5 times thicker. Well they differ quite a lot in density to. Which has implications for impact energy absorption (aside from other parameters). The denser PVC H60 and even more so H80 and H100 aso, support the sandwich skin better than the light XPS32kg/m3.
Here's something interesting. The above link contains info of both EPS35 (XPS) and PVC foam. What's even better, XPS sandwich with ply between outside skin and foam. And there are quite some conclusions.

 

I saw and at one time posted some test data showing the effect of glassing plywood and the gains in strength and stiffness, even with thin layers of glass are very impressive and the gain in performance on paper are very real. But when you read the actual test report you find that the mode of failure are very different once the plywood is glassed. In the testing the failures of the plywood were cracking and delamination. Once coated with glass the plywood (in this case 4mm) when overloaded literally exploded into pieces at the point of failure. While you gain a lot of strength and stiffness with glass over ply, you don't get as much energy absorption because the deflection at failure is much lower, and while the panels tested were was stiffer and stronger the amount of impact damage they could absorb was not increased as much and therefore the gain from glassing wasn't as good as you would initially think because really, for open water sailing you're designing for hitting that submerged log or almost floating lost container filled with tennis shoes...

In reality the thickness of glass necessary to prevent penetration could be helped by using a layer of plywood, you have to be careful since the resulting skin is strong but somewhat brittle. The folks who've designed boats using foam like Airex have guidelines for skin thickness and that's probably a lot safer place to start. If you go out on your own you're going to have to do some serious engineering and testing or you won't know what you have. It might be just fine, but do you really want to be miles out in the ocean, hit something and find out that the panel structure you cobbled up fails in a spectacular manner???

OTOH, if you design this boat as a plywood boat and skin it on the outside with some glass or other fiber and epoxy it will have good waterproofing and last a long time. If you bond a foam on the inside and then glass over that you're going to be even stronger, and have a true sandwich that is insulated, but knowing how much laminate to put on the inside is an issue. Remember that the inner skin of the sandwich takes the majority of the stress caused by bending on impact, and a thin inner layer of laminate doesn't buy you much before it fractures. This is where matching the penetration properties of the skin with the core and the inner layer are important and that's why it takes a lot of testing and engineering analysis to optimize a structure.

I'm partial to Airex as a core because it has very good impact properties as well as being a good insulator. If you design and use their design guidelines you'll likely be a lot better off than trying to design something from scratch.