Skip to content

christophmschaefer/miluphcuda

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

miluphcuda

miluphcuda is a 3D Smoothed Particle Hydrodynamics (SPH) code, mainly developed for modeling astrophysical collision and impact processes. It supports several rheology models for elasto-plastic solid bodies, as well as granular-like materials.

  • miluphcuda is the CUDA port of the original miluph code.
  • miluphcuda runs on single Nvidia GPUs with compute capability 5.0 and higher.

For an overview and quick start guide continue reading, for further details see the documentation.

Main features

  • hydro and solid body modeling
  • arbitrary number of interacting materials
  • several material strength models for elasto-plastic flow
  • fragmentation model for brittle-failure (Grady-Kipp)
  • porosity models (P-alpha, epsilon-alpha, Sirono)
  • self-gravity (via Barnes-Hut tree)
  • various equations of state including tabulated (e.g., ANEOS)

Quick start

  1. Install dependencies and compile (see below).
  2. Run one or more of the scenarios in examples/. You find detailed instructions there in the respective USAGE.md.

For generic simulations, based on your own initial conditions and/or material model choices, see Usage below.

Compilation

Download the latest version of the code or clone the github repo.

After installation of the required dependencies (see below), compile via the Makefile:

  1. run configure.sh which performs some tests and system-dependent configuration
  2. run make to produce the miluphcuda executable

It may be necessary to make some changes to the Makefile even if configure.sh succeeds. Compiler errors will let you know if that happens.

Note: Many important options for the material model and numerics are set at compile time in parameter.h, i.e., you have to run the Makefile (but not configure.sh) again once you make changes there.

Dependencies

CUDA
Install the CUDA Toolkit to compile and run code on the GPU.

libconfig
Used for parsing the material config file.
To ensure best functionality, install the latest version:

  1. unpack the .tar.gz archive
  2. sudo ./configure
  3. sudo make
  4. sudo make install

HDF5
If you want to use HDF5 for input and/or output files, install libhdf5 and libhdf5-dev.

Usage

The basic usage philosophy for miluphcuda is:

  • Basic simulation-wide settings, mainly for the material model and numerics, are specified in parameter.h at compile-time.
    See comments there for details.
  • Material-specific parameters, mainly for the material model and equation of state, are set in a material config file (typically material.cfg) and processed at run-time.
    See material-config/CREATE-MATERIAL-CONFIG.md for details on the file structure and available settings.
    A collection of parameters for various materials is available in material-config/material_data/.
  • All other main options are controlled via cmd-line flags. Check ./miluphcuda -h once compiled.

For each simulation, you need:

  • an input file which specifies the initial conditions (SPH particle distribution and their properties), check ./miluphcuda --format for the required format
  • a material config file, typically material.cfg, see above
  • suitable cmd-line options for miluphcuda, check ./miluphcuda -h

Specifically the examples/, but also the test_cases/ provide various scenarios and use-cases.

Developers

in somewhat arbitrary order:

Christoph M. Schaefer, Sven Riecker, Oliver Wandel, Thomas I. Maindl, Samuel Scherrer, Janka Werner, Christoph Burger, Marius Morlock, Evita Vavilina, Michael Staneker, Maximilian Rutz.

References

The code is also listed in the ACSL.

About

smoothed particle hydrodynamics code

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •