Based on your requirements—simulating cable snap loads, handling variable leg immersion (from 1/3 to full), and achieving high engineering accuracy—you have a tricky trade-off between computational cost and physical fidelity. Below is a comparison of the best open-source options.
Since you explicitly stated that linear BEM (like Capycante) is "not accurate enough" for your variable-immersion legs, and you want to simulate "snap" (snatch) loads, the best solution is Project Chrono (Multibody Dynamics) combined with a custom Hydrodynamic Load calculation using the Morison Equation.
This approach uses your Python skills and massive CPU to simulate the exact physical forces (drag & inertia) on the angled cylinders as they move in and out of the water, rather than relying on pre-calculated linear coefficients.
| Software | Physics Approach | Accuracy for Your Design | Time to 1st Sim (Est.) | Resource Usage | YouTube/Resource Link |
|---|---|---|---|---|---|
| Chrono (MBD) + Custom Hydro (Morison) | Multibody Dynamics + Empirical Forces | High (Accounts for variable immersion & cable snap) | 2 - 3 Weeks (Coding required) | High (Uses 64 Cores) | Chrono Multi-Body Demo |
| Capycante + MoorDyn | Linear Potential Flow (BEM) + Lump Mass Cables | Low/Medium (Assumes fixed geometry/wetted surface) | 3 - 5 Days | Low | Capycante Demo |
| OpenFAST | Linear Potential Flow (Fossen theory) + MoorDyn | Medium (Better hydro theory, but hard to mesh weird shapes) | 2 - 4 Weeks (Steep learning curve) | Medium | OpenFAST Floating Platform |
| Chrono::FSI (SPH) | Meshless Fluid (SPH) + MBD | Very High (Real fluid physics) | 1 - 2 Months (Very complex setup) | Extremely High (Needs GPU) | Chrono FSI Demo |
What it is: Chrono is an MBD engine (C++/Python). You will build the platform legs, pivots, and cables in Chrono. Instead of using a built-in "fluid" which is either too simple or too heavy (SPH), you will write a Python script (a loop) that calculates the hydrodynamic force on each leg every time step using the Morison Equation.
Pros: Extremely accurate for your specific geometry. Fully customizable. Uses your massive CPU/RAM efficiently.
Cons: Requires writing custom code to apply forces (a "drag law" script). You are essentially building your own solver, but it is the only way to get the physics right without CFD.
Time Estimate: 2 Weeks. (1 week to set up geometry/constraints, 1 week to tune the drag/mass parameters).
Video: While specific seastead videos are rare, this Chrono simulation of floating structures shows the capability.
What it is: Capycante calculates added mass and damping (BEM). MoorDyn simulates the cables.
The Problem: Capycante assumes the "wet" surface is fixed. In your design, the legs pivot and change how much is underwater. The linear BEM coefficients from Capycante will be an average. It will likely underestimate the snap load because it won't capture the "whipping" effect of the leg hitting the water or leaving it.
Verdict: Good for the cables (MoorDyn is great), but the fluid forces will be " textbook" approximations, which you specifically said were not accurate enough.
What it is: Using Smoothed Particle Hydrodynamics within Chrono.
The Reality: This is the "Brute Force" accurate method. It models water as thousands of particles hitting the leg.
Why not: To simulate a seastead, you need a massive tank of water. Your NVIDIA A6000 is good, but simulating enough particles to represent ocean waves for a 70ft wide structure for more than a few seconds will take years of compute time or require a supercomputer. It is overkill and likely won't simulate the cable dynamics stably.
What it is: The industry-standard open-source tool for offshore wind turbines. It handles waves and moorings.
The Challenge: It expects standard platform shapes. Modeling a "tensegrity" rectangle with pivoting legs is very difficult to mesh in OpenFAST. You would have to fight the tool to make it understand your geometry.
ChLinkTSDA (Tension-Spring-Damper-Actuator) in Chrono. Set a very low damping coefficient so they go slack easily.Once you have this Python script working for one model (e.g., the 40x16ft design), changing the design is easy. You just change the variables for the sizes of the cylinders in the script, re-run, and you have the new simulation.