Related to my suggestion about enhancing the Arc Length U Parameterization tool...

I applied ALUP to one side of the ship hull and copied the hull piece to a different layer for further work.

For whatever reason, I turned on Show Facet Edges and then realized the COPY did not retain the applied ALUP. Is this by intention? Since the ALUP does not seem to remain with a copy, I have copy the original curves to a new, "testing" layer recreate the surface, and then apply ALUP, then chop it, and resume whatever it was I was going to do. (which was attempt to "weld/edge-match" the fwd non-ALUP body with the after ALUP body and look for discontinuities. Since that distraction happened, I got scatterbrained and leapt into another alternative.

To be fair, and honest, I DO have another single-piece hull which is very smooth, and does not have the problems (or not so much in the visible/annoying realm), but that sonar dome is not optimized, and is a foreshortened one that does not resemble one in the real world. But, it has vastly superior fairness/smoothness. Lamentable that the unrealistic dome is more faired and visually smoother than the targetted/desired dome shape.

I'll hammer away another twelve hours then give up. I may have to release a shoddy model with some caveats. It's not for construction anyway, but it MIGHT cause negative light on SLT unless the end users of my drawings are better than I, or they adapt them without the rigors I'm trying to embed in the geometry.

I'm going to attempt to resolve the problem by rejoining the split curves of the futher-refined sonar dome and then reapply ALUP to the model. That's going to take 4-8 hours of tinkering just to decide how much more effort to keep pouring in.

Just as a heads up, this head-banging I'm doing in general CAD is not for naught. There are seasoned, serious naval architects and hull designers who for a few years now have been debating the merits of using general, 3D solid-model CAD as a huge cost-savings step compared to using high-end, highly-specific, non-generic ship design tools with all their intricate approaches to handling curves.

Apps like Shark, and it's recent competitors mentioned in this forum easily have the ability to displace SOME of the heavy-hitters in ship design. What Shark lacks (and this is not just about ship design, but parts design, airplanes, trains, autos, etc.) is a comprehensive or-API/plug-in capablle .dbf database interface to manage things. Editing a part destroys its parent part/entity ID number, rendering almost useless any attempt to meaningfully track geometry with an external database. And, such a database would need to track the coordinates of any curves, the commands and sequences of command application, and the mass properties information.


Shark needs a new brother/sister. IF it can cost less than $5,000 per seat, it would make deep inroads against other apps that only do a little bit more (aside from the nav arch calcs) in terms of geometry, but cost $15,000-$150,000 per seat. That new brother or sister app might rake in around 2,000 licenses if SLT can also attract eyes that AutoCad has not permanently taken away from considering alternative eyes.