```html
Based on your detailed description of an unconventional trimaran-like seastead with large, partially submerged NACA foil-shaped legs, here's a comprehensive comparison of open-source simulation tools that can handle wave interactions with complex geometries.
| Software | Method | GPU Support | Accuracy for Your Design | Setup Time (with Claude Code) | Visualization | Python Integration |
|---|---|---|---|---|---|---|
| DualSPHysics | Smoothed Particle Hydrodynamics (SPH) | ✅ Excellent (CUDA/OpenCL) | ★★★★★ | 2-4 weeks | ✅ Good built-in | ⚠️ Limited (post-processing) |
| OpenFOAM | Computational Fluid Dynamics (CFD) | ⚠️ Via external libraries | ★★★★★ | 4-8 weeks | ✅ Excellent (ParaView) | ⚠️ Via wrappers |
| Capytaine + MoorDyn | BEM + Mooring Dynamics | ❌ No | ★★★☆☆ (linear limitations) | 1-2 weeks | ⚠️ Requires custom work | ✅ Native Python |
| Project Chrono (FSI-SPH) | Multi-physics with SPH | ✅ Yes | ★★★★☆ | 3-5 weeks | ✅ Good (Irrlicht/VTK) | ⚠️ C++/Python bindings |
| Blender (Mantaflow/Physics) | Visual physics simulation | ✅ Yes (via Cycles/EEVEE) | ★☆☆☆☆ | 1-2 weeks | ★★★★★ | ⚠️ Python scripting |
| WEC-Sim + MoorDyn | BEM + Multi-body dynamics | ❌ No | ★★★☆☆ (linear limitations) | 2-3 weeks | ✅ Good (MATLAB) | ❌ MATLAB required |
Setup Time: 2-4 weeks with Claude Code assistance. The main challenges will be creating the complex geometry of your seastead and setting up appropriate boundary conditions.
Best for: High-fidelity simulation of your unconventional design, especially for investigating non-linear effects and large amplitude motions.
Setup Time: 4-8 weeks with Claude Code. Mesh generation for your complex geometry will be the most time-consuming part.
Best for: Ultimate accuracy if you have the patience and computational resources. Consider if you need publication-quality results.
Setup Time: 1-2 weeks with Claude Code. Fastest path to initial results.
Best for: Quick parametric studies and initial design screening. Use results with caution for extreme conditions.
Setup Time: 3-5 weeks with Claude Code. Good balance between accuracy and development time.
Setup Time: 1-2 weeks, but only for visualization purposes, not accurate simulation.
Recommendation: Use Blender only for final visualization of results from a proper physics simulator, not as the simulation engine itself.
While you correctly identified the MATLAB/Simulink cost issues, it's worth noting that WEC-Sim would likely struggle with your design due to the same BEM limitations as Capytaine. The commercial software requirements make it unsuitable for your "open source" criteria.
Start with Capytaine + MoorDyn for quick parametric studies. Use it to:
This will help you quickly iterate through design variations.
For critical design validation and extreme wave conditions, use DualSPHysics. This will:
Use ParaView (for DualSPHysics/OpenFOAM) or custom Python scripts to create compelling visualizations that show how your design behaves in different sea states.
You are correct that Chrono::FSI-SPH uses Smoothed Particle Hydrodynamics, not BEM. This makes it suitable for your non-linear requirements. However, for pure hydrodynamic simulation of your specific design, DualSPHysics is likely a better specialized tool.
Capytaine + MoorDyn offers the best Python integration, making design iteration extremely fast. You could create parametric scripts that automatically test different leg positions, sizes, and wave conditions.
Once you have a simulation working for one design:
All recommended tools work well on Linux. DualSPHysics and OpenFOAM actually perform best on Linux. You can develop on Windows if preferred, but consider Linux for production runs.
With Claude Code assistance:
For your unconventional seastead design with large, partially submerged foils, I recommend a two-pronged approach:
This combination gives you both speed of iteration and engineering accuracy where it matters most. Your GPU will be particularly valuable for the SPH simulations, allowing you to run multiple wave conditions relatively quickly.
The key advantage of this approach is that you can quickly eliminate poor designs using the faster BEM method, then focus computational resources on validating the most promising candidates with high-fidelity SPH.
```