Numerical Multi-Body Dynamics in C++ | A numerical simulation environment developed in C++ for the numerical code-generation and numerical simulation of the symbolic models developed in uraeus.smbd.
Note
Please visit uraeus.mbd for more information about audience , fields of applications and background.
- Code-generation for symbolic standalone topologies.
- Cython wrappers for selected C++ classes, exposing a minimal API that can be used from the python side to interact with the generated models.
- CMakeLists.txt for automated cross-platform build-systems generation using CMake, building simulation executable and python extension modules.
- Solver for Kinematically Driven Systems that can be used for forward and inverse kinematic analysis as well as inverse dynamic analysis.
- Solver for Dynamically Driven Systems that can be used for forward dynamic analysis and equilibrium analysis.
- Simple interaction with the generated models through a minimal API
Simulationclass, exposing only the needed functionalities.- Loading model configuration data from JSON file.
- Setting user-defined functions to control motion-actuators and generic force elements using normal C++ free-functions and lambda functions.
- Simulating the model for the desired time period and desired step-size.
- Saving the simulation results (positions, velocities, accelerations, reactions) as csv files, that can be used later for post-processing and visualizations in uraeus.visenv.babylon.
- Simple interaction with the generated models through python.
- Importing the model as a python module.
- Having access to the same minimal API of the Simulation class.
- Defining functions to control motion-actuators and generic force-elements using your custom python functions.
- Python 3.6+. If new to scientific computing in python, Anaconda is a recommended free python distribution from Continuum Analytics which provides a nice coherent platform with most of the tools needed.
- The uraeus.smbd python package.
- Git, for cloning the project repository.
- CMake, for build-systems generation. The project requires cmake 3.10 or higher.
- A modern C++ compiler supporting C++17 standards.
Note
The project is tested with the gcc 10 compiler on a Linux machine and the Microsoft C++ build tools MSVC on a Windows-10 machine.
You can clone or download the project on your machine. For cloning, make sure you have git installed on your machine and your system have access to the git commands through your terminal, command-prompt or power-shell.
Open your terminal then type the below command:
git clone https://github.com/khaledghobashy/uraeus_nmbd_cpp.gitThis will clone the project repository and download it on your machine in your current terminal directory.
On the same open terminal, go to the source-code directory via the command below:
$ cd uraeus_nmbd_cpp/uraeus/nmbd/cpp/enginethen create a /build directory by:
$ mkdir build
$ cd buildNow, we generate the build-system using `cmake`:
$ cmake .. -DCMAKE_BUILD_TYPE=Releasethen we invoke the build process via cmake --build. For multiple configuration build systems, i.e. msvc use:
$ cmake --build . --config ReleaseFor single configuration build systems, i.e. make, use:
$ cmake --build . This will build the project binaries and produce a shared library uraeuslib that will be used to link with the simulation executable and generated python extension modules.
Now, the project can be found via other cmake projects, where the paths for the headers and libraries binaries are automatically configured for projects using uraeus.nmbd.cpp.
t.b.c ...
The repository provides several examples in a form of jupyter notebooks that can be tested on the cloud using Google Colaboratory without the need for any setup on your local machine. The notebooks uses uraeus.smbd to model the system symbolically, then using uraeus.nmbd.cpp to perform the numerical simulation of the modelled system.
The implementation in this notebook can be broken down into six main steps as follows:
- Colab Machine Setup. We first starts by setting up the Colab machine environment by installing the needed tools and packages.
- Symbolic Model Creation. Here we create the symbolic topology of the system as well as a symbolic configuration.
- Numerical Environment Generation. We then pass our symbolic model to code-generators to generate the code files needed for numerical simulation.
- Numerical Simulation. We then use these code files to create our simulation instances and run the numerical simulation.
- Data Post-Processing. Now, we can use the raw results' data to evaluate the required characteristics and create plots.
- 3D Visualization. Finally, we use uraeus.visenv.babylon to visualize and animate our system in 3D.
Double Wishbone Direct Acting A kinematically driven system representing an independent suspension mechanism used for automotive.
Spatial Fourbar A dynamically driven system representing a closed-chain four-bar linkage driven by a torque at the crank.
Pendulum A dynamically driven system representing a simple pendulum subjected to the gravitational forces only .
To be discussed ...
The Implementation Details will be provided in a separate documentation and linked here
As the tool is developed and maintained by one developer for now, if you have any inquiries, do not hesitate to contact me at khaled.ghobashy@live.com or kh.ghobashy@gmail.com.
ToDo