3D printed boats

The below link may be of interest. The appeal I see in 3D printing boats as one pieces, is the ability to construct in any way printable. For example with multiple buoyancy chambers. Or storage lockers, conduits etc, depending on design printability.

Of course, this is all subject to getting the material to have the correct physical properties, but still it's very exciting.

UMaine 3D prints two new large boats for U.S. Marines, breaking previous world record - 3D Printing Industry

 

I've seen the Thermwood process demonstrated and I believe it has tremendous potential.Like almost any other 3D printing I've ever seen,it produces a surface that isn't of the same high gloss type that we associate with production moulds.Given what has been posted earlier in this thread about the quality of adhesion of epoxy it wouldn't be such a big job to give it a coat of high build epoxy,which could then be the basis for a highly finished part.Due to the fine ends of a canoe hull and the tight confines,it wouldn't be my first choice of hull form but it is achievable and would be much faster than the alternatives.You certainly would have to put a lot of work into direct machining a similar mould from foam or tooling block and then coating and polishing the surface.Not only that,but you would very likely have to make the mould in two halves to get the space for the head of the machine to get close to the surface.I think there will prove to be practical size limits for this type of technology but for canoe size boats it appears viable and I can think of any number of smaller components that would be well suited to the process.I'm thinking consoles,instrument pods and even furniture-if a decorative veneer could be bonded to the surface.After all good quality wood is getting harder to find and 3D printing could produce a substrate of shapes that are not limited by the forming characteristics of sheet material.I'd like to thank Jan for sharing his results with us and will be watching with interest.

 

Thank you guys I will share some basic hull renders of projects I want to further test/improve this technology on, I would love any feedback/input:
First and least complicated is something I have been thinking about for a long time, it's a very fast paddle-craft with pretty much zero static stability, instead relying on dynamic stability from 2 high-aspect ratio elliptic bilge-boards (not yet show in the render).
It's all about speed, I know that foiling craft would be faster, but foiling has some disadvantages and this is an experiment in the fastest possible displacement paddle-craft.

500cm length
30cm width
~200l total hull volume

Basic hull renders, I will enhance it soon with fittings (seat, pedal-steering, rudder, bilgeboards) + basic human figurine in order to show paddling position:




This hull with use just single outer skin in order to test chisel (my gcode generator) infill structure for tubular-like cross-sections:

 

@Jan Herich , I agree with your observations about that paper, the flaws you pointed out are really there. Still, I have to say that I was somewhat surprised to see that the core strength has so little effect on the overall strength of the part. This also implies that perhaps plastics other than PLA could be used, even if they have significantly weaker mechanical properties, if they offer other advantages, such as thermal resistance. Then again, PLA is so cheap, accessible and easy to print that I would probably go with it as well.
Do you happen to still have the data on these shear/compression tests that you made, and also any info on how you made predictions of strength of the structures you designed?

Also, I understand that the slicer you created is now closed-source, but perhaps you would be willing to share a rough description of how the sine wave warren truss infill generator logic works? I am a coder myself, and it would be a fun challenge to try to replicate something similar. I tried modeling this kind of infil in Solidworks CAD and succeeded, but I have to admit it was a lot of work and difficult to make it parametric. Still working on it.

As for the hull you designed, is 30cm width really enough for a human to sit in it? (I'll have to measure my *** now lol). The plan view shows a very nice curve, I think it should do nicely for your goal of going fast. However, side view reveals that the hull is box-like and has no rocker (bottom curve). I would very much like to suggest that you add some, which will increase buoyancy, volume and height where the paddler sits, and help hydrodynamics as well. Additionally, I would add a fillet at least to the forefoot and aftfoot, as these sharp points will easily take damage when they come into contact with anything harder than water. Difficult to say more without seeing sections of this hull. I am no expert on canoe hulls anyway, so take this with a grain of salt, perhaps others will chime in as well.

 

I keep wondering how the longitudinal strength of that hull, or torsional strength, is guaranteed.

 

Well, to my understanding, longitudinal and torsional strength is provided almost entirely by the GFRP / CFRP skin. Core is only responsible for placing some distance between the inner and outer skin layers, and having sufficient compressional/shear strength, which it apparently does more than well enough. Jan, correct me if I'm wrong.

 

Sorry but as long as I know, shear in plane, between core transversal rings has not been checked. I may be wrong.

 

@TANSL , I think Jan tested this:

But I might be wrong too. Best he confirms this, just in case.

 

@laukejas When we take the experiments done in the paper to the (absurd) extreme, imagine somebody would test rectangular steel plate in tension by pulling on its ends while bonding different (much weaker) materials on one side of the steel plate -> it would provide zero insights into behaviour of the weaker material, as everything measured (strength/E-modulus) is dominated by the steel plate.
Yes, they do use 2 plates with core sandwiched between them, but load direction is the same, they were just testing tensile attributes of the skin -> TLDR I don't think it provides any insights into properties of core material (3d printed in this case).

From what I remember PLA samples with density of 100kg/m3 did have around 1MPa compression strength which is around 2x stronger then core PET foams with similar density, shear strength in favourable direction was even better (~3x of the PET foam).

Predictions of structure strengths were using doing hand calculations taking into account tensile/compressive strength of the GFRP skin, numerically integrating its cross-sectional area along the hull (effectively calculating second moment of inertia) and using the know formulas for flexural strength.
It was very basic, I oversimplified many things (I treated the sandwich structure as single skin and just disregarded the core altogether) and only calculated critical load in bending, it could be way off from real limit, the only way to know would be destructive experiment -> I will probably do it with some sub-scale model.

Regarding the slicer and infill, the main thing is to treat sandwich perimeters as parametric surfaces where you can evaluate [[0 -> 1] [0 -> 1]] relative 2d coordinates.
Calling `patch.getPoint([0,0])` will get you lower left point (absolute x,y,z coords) of the patch, while `patch.getPoint([1,1])` will get you upper right corner of the patch.
Once you have this, it's actually easy to implement all kind of interesting infill/connection structures which naturally follow the contours of the patches (there could be even more then 2 patches, some infill types are applicable to any number of control surfaces).

I design all my structures as tensor product patches, where "longitudinal" parametric control curves (could be bezier, nurbs, or polyline-interpolation...) grow in Z direction and evaluation of all "longitudinal" curves at particular parameter (in the [0 -> 1] range) provides inputs for cross-sectional curve (could be again anything satisfying parametric curve interface).
Therefore it's very easy to iterate from 0 to 1 parameter in primary "longitudinal" control curves in order to obtain layer-data and again 0 to 1 in cross-sectional curve(s) to obtain perimeter polyline(s) for that particular layer.

Of course that's super simplified and there was huge amount of issues to solve, for example the need to normalize parametric curves in order to obtain uniform evaluation steps, etc.

Thanks for the hull inputs, I opted for no rocker in order to maximise waterline length in all conditions, but slight rocker would probably be OK and even look better

The hull is 30cm wide ~ on the waterline, significantly less then that on the (flat deck), the depression on deck is just to put your feet into (for ergonomic paddling position), there will be separate seat behind that on flat deck (probably rotating as in K1 boats):

This boat would have no static stability whatsoever and would need forward motion with rudder/bilgeboards providing stability

 

@TANSL As already mentioned, outer skin would provide flexural strength in bending/torsion, probably best realised as 0/45/-45 deg plies, with most weight in 0 deg ply and much less in 45/-45 deg ply, as cross-section is almost circular, naturally strong in torsion.
Structure of the core should stabilise the composite skin in place, but I never actually built/tested structure with ~30cm thick 3D printed core (which is essentially this).
So far I only built/tested structures which were "classic" 2 skinned sandwiches in cross-section with core up-to ~2cm thick.
First thing I will do is to 3D print ~20cm long section of the core sample (like the green-one I posted), laminate the outer perimeter and test it's crush-strength by placing it on some soft support in lying direction and progressively placing more weight on it.

 

It is possible that the laminates provide sufficient resistance to bending, it would have to be demonstrated, but not in terms of overall torsion. A hull with such a high L/B ratio, without a deck, has global torsion problems that should at least be quantified. On the other hand, two contiguous 3D printed rings are two independent elements that slide transversally relative to each other, depending on the inplane shear between them. From what you say, I believe that this phenomenon is not that it has not been studied, but rather that it was not known to exist. You have solutions for all this but do not trust the +-45º, etc, etc.

 

Hi @TANSL thanks for the input, but I have to say I'm very confused:

Why would deck (eq one thicker flattened side) help with torsional stiffness/strength of something with tubular cross-section ? The optimal shape for resisting torsional loads is tube with material uniformly distributed around perimeter, that's why every driveshaft in the world has this shape - Free Moment of Inertia & Centroid Calculator | SkyCiv https://skyciv.com/free-moment-of-inertia-calculator/

3D printed ring was just sample cross-section -> real core won't consist of independent parts free to slide against each other, they are bonded together with adhesive so that shear strength of the joint is actually superior to shear strength in the middle of the part (because of the much bigger bonding surface area, the hollow space filled with triangles is fully filled at the start/end of the core section).

 

I suspect that perhaps what TANSL meant that if the hull has no deck at all (C-shaped), even it's cross section being more or less circular wouldn't help torsion resistance. That kind of makes sense: if we take a tube and make a longitudinal slit through it's whole length, it would clearly lose a lot of it's torsional strength, since as the torsional force is applied, the edges along the slit would slide longitudinally in relation to one another. However, your canoe clearly has a deck, so I'm not entirely sure how we arrived at this question. In any case, I would consider this a design/structural consideration more than a core material consideration. I might be wrong though.

 

That the strength is provided by the outer skins can perhaps be visualised more easily by considering a laminate with a Nomex honeycomb core.It is a material that can be manipulated very easily until it is locked into shape by the skins.At which point the mechanical properties of the whole laminate will utterly eclipse those of the core.Obviously the shape will only be maintained if the constraining elements of the whole structure are in place and in the case of the prototype presented,the deck has a major part to play.For more complex boats the various bulkheads would also contribute.The basic principle of not having excessively large unsupported panels will always have to be a consideration,but with this technique it may become possible to produce complex shapes with minimal tooling costs and if they need a decorative surface then there are ways to achieve this also.

 

@wet feet Exactly, although there was probably misunderstanding in cross-section of my latest plan ("torpedo" canoe) which was assumed to be C-section with open deck, while it will be closed tubular section (except for depression for feet which will be relatively short/shallow).
While the prototype on first page (405x46cm) does have partial deck (spanning less then half of the total boat length) providing some torsional strength, previous prototype printed from clear high-temperature CPE filament (TG over 100C) was fully open C-profile without any bracing/coaming:

It's a super small boat with 270x44cm dimensions (for my small daughter) so obviously I don't want to draw too much conclusions from that, but I was surprised how very stiff it was, probably property of up to 2cm thick core (tapering to 1cm towards bow/stern).