Hull tuning with Octopus

HI EVERYONE AND HAPPY NEW YEAR!

========================

I gave a try at the new Octopus multi-objective solver.

5 objectives :

1 : Prismatic coefficient

2 : Position of main area section

3 : Wet surface

4 : Waterplane inertia

5 : "Finesse" f = WLL/(V^1/3) _ f stands for "frog" also, I think you use WLL^3/V in the English speaking world. I could have used WLL directly because the hull is set at the desired displacement before analysis - no use to test at random displacement indeed!

---First try on a 40' cruiser with jittery control points and bad volume repartition, a 30-seconds-work to a tee.

Red dots are original points, white mesh is current surface control polygon.

INPUT

After just a few steps I was able to find good candidates. Check the video.

OUTPUT

---Second try on an Open 60' hull. This was an rhino modeling exercise for students, so it's clean and realistic already, and I gave the same objective values as the original to see if it would fall on its feet. In this case I blocked the edges points.

I was amazed to see that It works!

INPUTOUTPUT

See the nice diagram! All points closest to the wet surface axis have good CP and position. On the left, very round hulls with poor stability (purple) and short WLL (big), on the right, wide hulls with higher water resistance (cyan) and longer WLL (small).

This is only one generation. with a few steps more it zooms in where the two sides meet. The two surfaces actually cross and I found very quickly an individual dominant on both wet surface and stability fronts.