Vacuum infusing wood that has been CNC'd to fit a mold

Can you vacuum infuse wood?
Could you CNC pieces out of lumber that fit the mold perfectly and has flow channels?

Hi, I'm new here and I've been dreaming of living on a big ship for a while now. Ideally a large powered cat. But the more I read I realize the enormity and cost of a build. I'm looking at a chine hull design by Richard Woods (Skoota 32) that allows the relatively simple construction of a mold using melanine plates for vacuum infusion. But foam core is expensive and requires a lot more expensive resin compared to plywood builds. And even marine grade plywood is rather expensive. Yes I'm a cheapskate

From what I understand the cheapest building method just from material costs should be strip planking. But for a large boat it takes thousands of hours, so I was wondering about methods to make this easier. I figure even if I have to spend 5k on a large CNC router that would save me more time than that. Plus I have a CNC afterwards

So very roughly this is a visualization of my idea:



You would buy suitable wood in the form of lumber planks (maybe 2cm / 1" tick), plane it and use the CNC to cut it into precisely fitting pieces. Along compound curves these would be honeycombs but along straight areas you would use longer rectangular pieces. The pieces include flow channels and a suitable number of holes so you don't need flow medium and the individual wood pieces are encased in epoxy. You would plane the lumber from one side, put it on the CNC and it cuts out the curved surface, flow channels and segments.

Then you would spray gelcoat in the mold and lay up fiberglass, and then lay up the pieces just like you would fiberglass. They would need to be numbered by the CNC so you don't have to play the worst puzzle in the world haha.

The above picture is just a quick example and isn't really accurate. The backside of the wood pieces would be flat from the planing. So you would end up with a flat "low poly" honeycomb on the inside. Also the pieces look very thick.

A second layer of wood with the grain in the other direction could be CNC'd too, essentially creating a two layer plywood. You could also have a second thin fiberglass cloth between the two wood pieces, which might allow for thinner wood to be used.

I would have to write a software to generate these 3D pieces and generate the files for the CAD/CAM software. Basically you would load a 3D mesh and put points along the surface depending on the curvature, then generate what's called a Voronoi diagram.

And you would need an engineer to see how thick the wood and the fiberglass has to be. I figure this could only be done accurately using some sort of finite element simulation software that can simulate multiple layers of oriented composite materials. Guessing would lead to overbuilding and unnecessary weight.

You could also coat the pieces in epoxy before infusion to avoid excessive resin uptake. But I think infusing the wood with epoxy would be a structural plus and you could use less fiberglass.

You would probably glue multiple pieces of lumber into a large sheet so the CNC can process a lot of pieces in one go.

You could also coat the individual pieces in epoxy before infusion to avoid excessive resin uptake. But I think infusing the wood with epoxy would be a structural plus and would lead to less fiberglass needed. You could probably also increase or decrease the resin uptake a bit depending on how much remaining humidity the wood has.

This is similar to end grain balsa core of course, but it would use less resin than grid scored balsa that creates larger caps. But I'd be very curious how thick a laminate you need for end grain balsa compared to plywood / core. I figure you definitely need more than for plywood because end grain has low tensile strength. Anyone have some numbers to compared to core / plywood?

The possible advantages would be (once it's set up)

• low cost of fiberglass and resin
  
• low cost for unprocessed wood (suitable wood might still be expensive though)
  
• low cost for labor (but more than vacuum infusing foam core),
  
• no consumables like peel ply and flow medium (if you use self separating vacuum bag)
• Less sanding / fairing

Disadvantages would probably be a lot of wasted wood and a lot of small pieces in areas of higher curvature (or thicker pieces).

So is this possible? Has this been tried?
I have zero experience in boat building. Nor with CNCing. All I have is silly ideas But I've seen tutorials for creating stabilized wood for wood blanks and similar. They basically use a very low viscosity thermosetting resin for vacuum infusion ("underwater" in resin). So it should work, it's just a question of making the flow channels the right size.

I'd love feedback or ideas. At the moment I'm just dreaming and reading and don't plan to start building a boat in the immediate future. So I don't have the tools to test this out, but I could start writing the software to generate the pieces.

 

Go with what you KNOW will work. If you're thinking you're going to save some money by re-inventing the wheel you're just kidding yourself. People have been trying to find the lowest cost way of making things for the last few hundred years. While technology has marched along there are really only a few good ways to make stuff and those methods have been refined and optimized for a good long time. Building anything like you are considering is an expensive and time consuming endeavor and if you go out on an unproven path you're likely to end up with a huge thing in your backyard that has to be taken to the dump to dispose of..

While buying foam and glass sounds expensive, if you don't design using tried and true methods you'll just have an expensive experiment. By cutting your wood into small pieces you're basically just making the wood into a core and relying on the skins for the strength. Much as end grain balsa or foam acts as a core, the design concept shown is really just using heavier wood for a core and that just jacks up the weight with no real advantage in strength, since the skins will need to carry the loads, the wood heavy core doesn't add much other than weight. From that aspect, glass over ply is probably the most effective means of building your own boat and that is tried and true, reasonable cost and will last a long time.

Probably the least expensive means to get what you want is to buy a used boat that is a good serviceable craft. The second way that is probably just as expensive or more is to refurbish a basket case. These are very cheap to buy, but will cost a lot to make useful. The advantage to that approach is that you can get something that is closer to what you eventually want. Third is to buy something that is gently used, and this is probably the best value proposition.

Building will get you exactly what you want, but you won't save as much as you think, and it will take years of your time and thousands of dollars to get it all done.

 

Strip planking can be a very fast way of building if you don't re-invent the wheel. Unless you want a varnish finish, there is no need for precise fits between the planks. Putty the gaps with medium density filling, grind the surface with a hard backed disc and laminate inside and out. It allows you to use low quality lumber, as long as the moisture is low before gluing and laminating.

 

But I figure it's still a lot of labor for larger boats? This idea would allow you to build wooden composite hulls inside a mold without having to sand and fair the surface afterwards. I know this whole idea is probably crazy especially coming from someone with no experience

Compared to bending wood (I've heard you actually compress the inner side since wood fibers can't be stretch) would you loose or gain strength by just milling slightly across the grain?

 

The rendering you show has bevels that only work for that particular curve. Any other section, or changing curves would require a different bevel for each piece. In other words, a nightmare of puzzle pieces. A long narrow hull, like multihulls have, are very fast to build with plywood for the relatively flat sections and strips on the rest. You can roughly figure the bevel of planks from the lines plan and cut the strips. Epoxy does not require tight fits. Wood fibers, like any other material, can be stretched. You can look at tables to see how much vs force applied. That is the stress/strain graph or table. I think the main problem with your plan, is that you are talking about cheap and a large ship in the same sentence. If you want to save money, buy a used boat. This is advice you will see repeated time and time again in this forum. Amateurs takes many years, sometimes a couple of decades, to build a boat. Usually the final product is mediocre. The storage during the build costs more than a used boat would've. To start with, what is your budget?

 

I think you have the wrong ideea about strip planking. As gonzo said it's only slow for small boats that use thin strips and want a natural finish. Thicker strips and painted finishes are fast. Two guys, one mixing glue and one laying strips with plastic nails or staples from a gun can make short work of any size hull. Then there is the so called "dry method" where you put on the strips without glue and trowel the glue into the gaps. The fiberglass can be infused over the strips if desired.
Strip planking does not need a mold, that's one of the key benefits.
For your method I see only disadvantages. You need a mold. You would use exactly the same amount of fiberglass and epoxy as end grain balsa, since your hexagons are doing the same thing, separating skins. They can not be structural, they are to small for that. Same amount of sanding and fairing is also the same. Labor would increase since they do not come atached to a grid and would have to be placed idividually by hand. You can also not make them to tight fitting or resin will not flow. There actually exists an infusion foam with a hexagonal pattern, the channels are large. Weight would increase over balsa.
Oh, and the self separating and reusable vacuum bag would have to be silicone rubber cured on the mold. That means unsable for any other boat, not to mention what fabricating it costs.

 

Yeah, each piece would need to be custom milled. Software would generate the shape of each puzzle piece and for larger flat areas the pieces would be larger. And you'd have to label the pieces on the CNC with coordinates. Like distance from bow and height and section.
And long curved beams for the chines and edges with perfectly fitting connectors to create precisely fitting joints. Only fully curved areas need the small pieces as shown in the image. The CNC would need to swap to a router bit to be able to undercut the seams to create precisely fitting joints. Like connectors that get pressed together by vacuum and infused.

I guess the only way to really evaluate this idea is to write the software, and after that build a small dinghy.

Oh and it's not a plan per se, it's more like a dream I've been thinking about the idea of living free on the oceans in a large (solar) powered cat for a while now. Not very soon, maybe in 4-5 years. But like you say the more you read the more you learn how difficult it is. And from what I've read large catamarans aren't cheap even second hand. And I think power cats are even rarer? I do have some money but I have to be stingy with it.

So I came up with this idea by thinking how I could possibly reduce material costs and hours spend. This technique wouldn't even be very practical for me for a one off - you need large molds and much more tools.

I just get these ideas sometimes and need to get rid of them And so far I couldn't think of anything why this technique wouldn't work, and wouldn't automate ship building and reduce material cost.

 

Thanks for the replies! Just to be clear, I'm probably not going to use this method and I know I have no experience boat building, but I'm still going to defend this idea as good as possible. Seems more fun and productive that way

But don't you still need to fill, glass and sand a very large area? Or are glossy surfaces on boats overrated?

But then you need consumables again like flow media, tape etc. And still need to sand.
Needing a mold is a disadvantage for people like me, but it also has the advantage of producing a good finish, doesn't it?

The idea of this is how to combine the faster production time of a mold with cheaper wood.

Well, what if you had a second layer of wood core elements going in the other direction and connecting the pieces with each other? Possibly it would get too thick of course.

I guess for areas with smaller pieces you could lay out more fiberglass, and for areas with longer panels less.

The question is if you need that much fiberglass for tensile strength, holding the pieces together. With foam you need it for impact strength.

But what you say is probably the biggest problem with this whole idea. Strip planking would probably lead to more strength.
Vacuum infusing wood pieces together might make up for that though.
With strip planking you join planks together as well, and I read with epoxy and some tape they are stronger than the surrounding wood?

Yeah conforming core material would be arguably less work. But I don't think that much since you still need to cut the foam or balsa wood sheet to the right size.

But why would you need the same amount of sanding and fairing?

You could have small spaces CNCd into the pieces and control the resin flow precisely by cutting in channels and drilling holes through them.

When I wrote this I thought that the gaps from the balsa core grid actually gets fully infused with resin. And I've seen the 3D core foam you mention, which is partly what gave me this idea.

PS: Yeah I looked at RTM light. Expensive. There is a company in the UK that offers a silicone gun starting at 2000 pounds though.
But I saw a "High-stretch polyolefine-copolymer film" vacuum bag that the store said is "in principle reusable".
But even if you have to throw away the bag, at least you could save the flow medium.

 

I have a different problem with the method. Puzzle joints are already used, but they are done on the ends of large panels in monocoque structures. If you use puzzle joins incorrectly; there would be a chance for structural deficit. There is a reason strip planking doesn’t change directions radically, for example, if the builder’s pieces are short; he doesn’t butt them all in the same place..

I see this idea as too much work and too much wastage with potential structural deficits and not really novel either.

Sorry.

Conventional methods were not borne of stupidity, but neccesity. Can they be improved upon? Sure. Not this way.

But the Skoota 32DM I am building did use a cnc for the frameset. And it is a flat or developable panel hull in both the fixed and dm versions. The purpose of the foam is to make the vessel lighter.

The skoota 30 is a smaller vessel, but far more affordable as an option in plywood and would offer faster build time.

 

Thanks fallguy for the feedback. Unfortunately I'm starting to agree.
I've just "discovered" cylinder molding which is obviously superior for a one off multihull build. Bending (developing?) wood is better because you maintain the fiber of the wood for large areas, while CNC cutting unbend pieces will cut across the grain which will weaken it's strength.

And if you do have a mold, I figure you could actually combine the method of laminating multiple thin plywood sheets together with what I proposed (Sorry I'm switching again to reinventing the wheel mode, so please don't be sorry to criticize!
Cut the thin 3mm plywood sheets on a CNC to fit precisely into that mold. You could have a flange on top of the mold to torture and "snap" the plywood sheets in position under it. It might not work because the sheets would move against each other or lift off the mold, or the scarf joints might not align, or the fiberglass below might move.
And if what I proposed for cutting flow channels into the wood with the CNC for vacuum bagging without flow media actually works, you could do that with thin plywood too.

You could also use the cylinder molding method to quickly create a plug to make a mold. Then you would only have to sand and fair one hull. But maybe I'm overestimating the amount of work it is to sand and fair a hull.

In any case the cylinder molding method sounds like it would save a lot of daunting work at least for the hull, which makes me very happy. Is there a book that covers it in more details? Is it harder to design an efficient hull for this method? Are there plans out there that use this method?

Kurt Hughes also suggests some cheap alternatives to flow media (bubble wrap) and just using polyethylene sheeting and an alternative to expensive tacky tape for sealing the bag.

 

3D printed boats are just around the corner )

 

They are? I'd be very curious!
I've actually thought about 3D printing core structures as curved honeycombs. You can 3D print in all sorts of materials like PET (3d core is also PET foam). But with the prices of filament that would probably not be cheaper than foam (except with a special extruder to use granulate). And you can't use vacuum infusion with not "closed cell" core materials. But the biggest problem is that you again have to use thicker expensive fiberglass laminate instead of cheaper wood as an active core. I could not find a way that would be better than using foam in a mold, but maybe they do a quite different approach. Or they are just smarter.

 

Glossy surfaces are very overrated for cruising boats. A large budget is necessary to keep them glossy. That is the reason workboats (cargo ships, trawlers, etc.) have a semi-gloss finish.

 

Ah I see, thanks! That is good news. Makes sense especially under water. You only see the surface above water which don't really matter for efficiency.

 

You don’t need to cylinder mold flat (developable) panels. With all these ideas of yours; it doesn’t seem like you understand round hulls vs flat panel hulls.

The Skoota 32 is a flat panel design, as is the Skoota 30. The budgets for these boats would vary from 50-100k US; perhaps a bit less for the Skoota 30...I made some vague assumptions budgeting for it as I leaned toward the bigger boat as there are habitable spaces in the hulls and no wood. I am not a fan of wood here mostly because I want a beachable boat and a single rock missed has the potential to cause a break in the glass and then wet plywood.

Laminating large sections of plywood together well is not simple and requires presses or vacuum table to do well. The delam risk is greater. Tortured ply or stitch and glue construction is not a Wood’s method. I haven’t seen any bigger cats proposed as s&g. It could be done, but haven’t seen it. But then a mould is not needed.