Closed bochen2027 closed 9 months ago
here is what chatgpt had to say:
Based on the information gathered from the GitHub page of FluidX3D, it appears that this software has a variety of features that could potentially be used for a ship flooding sinking simulation:
FluidX3D uses the lattice Boltzmann method (LBM) for computational fluid dynamics (CFD), which is a numerical approach capable of simulating fluid flow and is frequently used for modeling complex fluid dynamics problems like ship flooding【14†source】.
The software is capable of computing force/torque on objects, which might be useful for simulating the forces acting on the ship as it floods and sinks【8†source】.
FluidX3D has added functions for voxelizing moving/rotating geometry, which could be applicable to a sinking ship scenario where the ship's position and orientation change over time【10†source】.
The software supports particles with the immersed-boundary method, a technique used to model fluid-structure interactions, which could be relevant for simulating the interaction between water and the ship's hull【11†source】.
FluidX3D provides visualization options, including slice visualization and surface pressure visualization, that could be useful for visualizing the flooding process and pressure distribution【12†source】【13†source】.
However, there are a few considerations to keep in mind:
The software is described as being free for non-commercial use. If you intend to use the simulation for commercial purposes, you may need to seek a different license or use different software【7†source】.
It's not clear from the provided information whether FluidX3D supports simulation of free surfaces, which are important in ship flooding simulations. Free surfaces refer to the interface between the air and the water, which can change shape and position as the ship floods.
It's also not explicitly stated whether FluidX3D can handle the overset mesh approach mentioned in the paper, or whether it can simulate the air flows in the ventilation pipes.
The paper mentions conducting the simulation with laminar flow. While the lattice Boltzmann method can handle both laminar and turbulent flows, it would be important to confirm that FluidX3D can accurately model the specific conditions you're interested in.
FluidX3D is a GPU-based software, which means it may require a powerful GPU for large-scale, high-fidelity simulations. The computational resources required for the simulation will depend on the specifics of the simulation, such as the resolution of the mesh and the complexity of the physics involved.
In summary, while FluidX3D has many features that could potentially be used for a ship flooding sinking simulation, it would be advisable to thoroughly test the software and confirm that it can accurately model all the relevant physical processes before relying on it for this purpose. It might also be helpful to reach out to the developers or the community around FluidX3D for further advice and guidance.
This is a guide on how to create a simple 3D voxel physics water based ship simulator using FluidX3D. FluidX3D is an open source project that simulates fluid dynamics on a voxel grid using the Lattice Boltzmann Method (LBM). You can find the documentation and source code here: https://github.com/ProjectPhysX/FluidX3D/blob/master/DOCUMENTATION.md
To use FluidX3D, you need to have a 3D model of the ship you want to simulate. In this example, we will use a Titanic model that was downloaded from the internet and converted to STL format. STL is a common file format for 3D printing that describes the surface geometry of a 3D object using triangles. You can use any 3D modeling software to convert your model to STL format.
To import your model into FluidX3D, you need to modify the setup.cpp file in the src folder. This file contains the main function that sets up the simulation parameters and initializes the voxel grid. You can use any text editor to edit this file.
The first thing you need to do is to specify the path to your STL file in line 14. For example, if your file is named titanic.stl and it is located in the same folder as the setup.cpp file, you can write:
std::string stl_file = "titanic.stl";
The next thing you need to do is to adjust the size and resolution of the voxel grid. The voxel grid is a 3D array of cells that represents the simulation domain. Each cell can be either solid, fluid, or empty. The size and resolution of the grid affect the accuracy and performance of the simulation.
You can change the size and resolution of the grid by modifying lines 16-18. For example, if you want to create a grid that is 100x100x100 cells, you can write:
int nx = 100; // number of cells in x direction int ny = 100; // number of cells in y direction int nz = 100; // number of cells in z direction
You also need to specify the physical dimensions of the grid in meters by modifying lines 20-22. For example, if you want to create a grid that is 10x10x10 meters, you can write:
float lx = 10.0f; // length of grid in x direction float ly = 10.0f; // length of grid in y direction float lz = 10.0f; // length of grid in z direction
The last thing you need to do is to set up the initial conditions for the simulation. You need to define which cells are solid, fluid, or empty at the beginning of the simulation. You can do this by using the setSolid and setFluid functions in lines 24-28.
The setSolid function takes four parameters: x1, x2, y1, y2. It sets all the cells within the specified range to solid. The setFluid function takes five parameters: x1, x2, y1, y2, z. It sets all the cells within the specified range and below the specified height to fluid.
For example, if you want to fill half of the grid with water and leave the other half empty, you can write:
setSolid(0,nx-1,0,ny-1); // set all cells to solid setFluid(0,nx-1,0,ny-1,nz/2); // set half of the cells to fluid
To import your STL model into the grid, you need to use the loadSTL function in line 30. This function takes three parameters: stl_file, scale, offset. It reads the STL file and scales and translates it according to the specified parameters. It then sets all the cells that intersect with the model surface to solid.
The scale parameter is a float value that determines how much to scale the model relative to its original size. The offset parameter is a vector3f value that determines how much to translate the model relative to its original position.
For example, if you want to scale your model by 0.01 and center it at (5,5,5) in the grid coordinates, you can write:
loadSTL(stl_file,0.01,vector3f(5,5,5)); // load STL model
After modifying the setup.cpp file, you need to compile and run FluidX3D. You can use any C++ compiler that supports OpenGL and GLUT libraries. For example, if you are using g++, you can write:
g++ -o fluidx3d *.cpp -lGL -lGLU -lglut
To run FluidX3D, you can write:
./fluidx3d
You should see a window that shows the simulation. You can use the mouse and keyboard to interact with the simulation. You can find the details of the controls in the documentation.
You have successfully created a simple 3D voxel physics water based ship simulator using FluidX3D. Have fun!
@bochen2027 Worth noting that those chatgpt hallucinations have very little in common with what the instructions README actually say so everyone who actually read all that has just been wasting their time.
Hi @bochen2027,
a ship hull bleach simulation could be somewhat possible. Main difficulties are that FluidX3D cannot model entrained air volumes/bubbles, and the movement of the ship with changing buoyancy cannot yet be modeled. Revoxization of moving geometry currently fails with the free surface extension as it becomes unstable. So you'd have to keep it simple with a static model.
Kind regards, Moritz
Is this suitable for ship sinking simulation? We all have heard of Sinking Simulator which is 2D particle sim, but there isn't a simple one for 3D etc...
For example get a autocad or 3ds max model of Titanic on turbosquid or whatever, export to .stl and only keep the 16 watertight compartments and do a 'large scale' simulation.
In the case of a sinking ship, you have a tiny breach in a large hull, and it's crucial to accurately model the water flowing through that breach while still efficiently simulating the larger-scale behavior of the ship and the surrounding water.
Does this program support extension for adaptive mesh refinement (AMR), or unstructured mesh etc?
https://www.sciencedirect.com/science/article/pii/S0141118722003327
https://github.com/sambolinski/FloodingSimulator
"""CFD results for the Case 3 with Star-CCM+ software. A mesh of about 3 million cells, using an overset mesh approach with a time step of 0.002 s was used. Model scale was used, and the results have been converted to full scale for comparison with the other codes. Also the ventilation pipes were modelled and calculations included the air flows. The model was free in all 6 degrees of freedom. The simulation was conducted with laminar flow and without prism layers to reduce computation time. Laminar flow was considered a reasonable assumption since the simulation was performed in model scale. Including prism layers would have improved the modeling of the water-wall friction, but this was believed to be of minor importance compared to the flooding dynamics."""