3D printed boats

@Jan Herich, in the last picture uploaded by you (post #30), you can clearly see a longitudinal discontinuity. The core of the hull is formed by transverse rings joined together. These joints are the ones that I believe should be verified as sufficient to guarantee the longitudinal resistance of the hull (hull considered as a beam supported at its ends). On the other hand, all tubes with "C" sections can have problems of torsion along their longitudinal axis. I understand that these calculations may seem exaggerated for a hull of that size, but that does not mean that there may not be a problem worth studying.
I am curious to know how you were able to check the resistance that you say has been checked. But you don't need to answer me.
Thanks for your patience with me.
(By the way, a thickness of 2 cm in the core, by itself, does not guarantee anything.)

 

That is a very beautiful shot. I suspect you get a lot of questions when people see this thing on the water! Seeing this pic, two things come to mind:
1) Did you use some kind of UV-resistant varnish on top of the composite?
2) I wonder how hot can a hull get on a hot sunny day, especially if it's sitting on the shore rather than on the water. Could it get hot enough to soften PLA, even if it's annealed? This question assumes the hull is painted (and of course the color of the paint impacts this as well). Not sure how the thermal dynamics play when the hull is transparent, like yours. I wonder if there may be a greenhouse effect, contributing to a higher internal temperature when left under the sun for a significant amount of time.

 

I did use UV stabilising agent mixed into top epoxy layer for the translucent boat and durable PU coat for the white one.
Regarding the second question, I'm testing that by having the white boat placed belly-up under open-sky, supported just in the middle by narrow table (so there is some load at the ends of the boat), over the couple of last days, it was very hot here (33-34C in shadow), so far no problems/deformations and the white surface was warm, but not particularly hot to touch.
When fully annealed from amorphous to crystalline state, PLA is actually super heat-resistant, well over 150C, much more then PETG, ABS or almost any epoxy resin.
I have very little confidence that I achieved anything approaching that with my long low temperature annealing process, but so far it works reasonably well when hull is painted signal white (RAL9003), result would be probably different when painted in some darker color, I need to test it

For the translucent hull, I'm not worried over that at all, as said the CPE filament should be very temp resistant without any post processing (it's printed at ~270C without any cooling and according to datasheet has TG of ~105C).

Regarding the appearance, it's actually funny how much that affects the perception of structural qualities/strength -> the white boat looks like any other composite hull, while the translucent one is different, despite both using almost the same amount of glass reinforcement, epoxy & core density, therefore having very similar structural properties.
But occasionally I get very negative reactions about the translucent hull, with people saying it looks "toy-ish" and they would never trust such boat -> The only problem I have with translucent hulls is that unfortunately joints are very pronounced/visible.

 

Very interesting, I supposed temperature would be a bigger issue than it actually is. Building such a large annealing chamber is not something most would be able to do I suppose, but I think I'll make my own tests as well with un-annealed PLA and see what happens.
As for UV stabilizing, personally I had poor experience with UV-stabilized epoxy (it yellowed nonetheless), but I'm glad it worked out for you. As for the negative comments, these people don't know jack sh*t. Translucent hull with this wavy infill looks like something out of this world. I would probably go with translucent look as well, if it were not for my small printer, which means joints will be seen everywhere! I also wonder if perhaps there is no need to form flat "bulkheads" for gluing these different parts together. If the perimeter and infill of both joined parts matches, then perhaps that would be enough to form a solid joint. After all, it only needs to be as strong as layer adhesion next to the joint, and we already figured out that layer adhesion is not much of an issue when coating this core with fiberglass afterwards. This calls for another test - checking whether glued layer adhesion is as strong as printed layer adhesion, and if not - how extra big does the gluing surface area really needs to be to achieve a comparable strength. After all, the weight of these "bulkheads" does add up, especially if one has to print a big boat with a small printer.

 

I once had a 3 meter long kevlar kayak that weighed 4kg. I could barely use it, was designed for a kid. Anyway, when I paddled really hard it would get "on plane" and perhaps reach 5-6 knots exceeding its hull speed but with ridiculous effort.

If you want the fastest rowing displacement craft I think you need a rowing shell (much longer) What you have will generally be slow because of the waterline length.

I dont think you addressed water getting inside the infill and it will eventually. Unless you can fill it with closed cell foam, or perhaps somehow connect all of the inside sections (and you could leave small holes to do this in the infill with your custom slicer so it can be pumped and drained. Imagine compromising the outer shell and having the entire core full of water to paddle back.

As for open source, I would say it is essential. My main project (pypilot) would not be able to receive the user base and feedback without this. Furthermore, you will prevent people from hacking your code to improve it and send you the improvement. Without open source, your code will likely die in a few years as alone and with a small user base someone else can quickly superseed it (eg someone who does nearly the same thing but decides to open source it).

I still execute and use daily code free software was written in the 80s. Few people can utilize proprietary software written from this era: proprietary software has a short life and fewer users.

 

I think hull speed is kind out outdated concept which doesn't fit very well to ultra-slender vessels. I regularly paddle my 5.2*0.52 Nelo 520 surf-ski (which is not ultra-slender) over it's hull speed (which is ~10.3km/h) and it's certainly not in planing mode.
The fastest paddling (different from rowing) displacement crafts are K1 kayaks, which have to be exactly 5.2m long with waterline beam of ~0.38m, those could be paddled by top athletes over long distances at ~14-15km/h and exceed 22km/h top speed in short sprint efforts, which is more then double their hull-speed (still not on plane).
This concept is little bit shorter, but even narrower with less wetted surface area, the right balance of form/wave and friction drag needs to be found for top speed (certainly depends on total displacement ~ paddler weight), but I think that starting with ~K1 length is a good bet.

Regarding the infill, those ribs form closed channels along the hull (separated by "bulkheads" at each core part start/end), so in worst case, one ~2-3cm wide channel with length of ~1.3m will be filled with water - not ideal, but not end of the world.
Nomex honeycomb cores with very thin "paper" walls (which are not waterproof) are used on top-end kayak/surf-ski/canoes, those can also fill with water in case of hull breech and are usually quite hard to completely dry, but it's not regarded as a big issue - such accidents (hull puncture without more substantial damage) are quite rare, it never happend on my Nelo 520 for example (and I had may fair share of stupid and completely avoidable accidents, like banging the carbon under-stern rudder against submerged concrete barrier at almost full speed).

I completely agree with you on the open-source code, I'm big believer in open-source myself, contributed to many projects, my favourite text/code editor is ~50y old piece of amazing open-source software still being actively developed.
My modeller/slicer was originally open-source, but after getting negative feedback on that from multiple potential investors I closed-sourced it.

 

I believe I've read a paper somewhere that stated that once a hull approaches length:width ratio of 7:1 or greater, the exponential resistance increase associated with the hull speed flattens out significantly, meaning there is no more hard limit anymore. Probably that 7:1 ratio isn't fixed either, and this effect comes in gradually, I assume it's a ballpark figure. I can't seem to find that paper now, though, but in any case, this matches what you wrote here.