Ludwig: a lattice Boltzmann code for complex fluids

Overview

This introduction provides a brief overview of what the Ludwig code does. It is assumed that the reader has at least a general knowledge of Navier-Stokes hydrodynamics, complex fluids, and to some extent statistical physics. This knowledge will be required to make sense of the input and output involved in using the code. Those wanting to work on or develop the code itself will need knowledge of ANSI C, and perhaps message passing and CUDA.

Purpose of the code

The Ludwig code has been developed over a number of years to address specific problems in complex fluids. The underlying hydrodynamic model is based on the lattice Boltzmann equation (LBE, or just “LB”). This itself may be used to study simple (Newtonian) fluids in a number of different scenarios, including porous media and particle suspensions. However, the code is more generally suited to complex fluids, where a number of options are available, among others: symmetric binary fluids and Brazovskii smectics, polar gels, liquid crystals, or charged fluid via a Poisson-Boltzmann equation approach. These features are added in the framework of a free energy approach, where specific compositional or orientational order parameters are evolved according to the appropriate coarse-grained dynamics, but also interact with the fluid in a fully coupled fashion.

How the code works

The code is written in ANSI C, and can be used to perform serial and scalable parallel simulations of complex fluid systems based around hydrodynamics via the lattice Boltzmann method. Time evolution of modelled quantities takes place on a fixed regular discrete lattice. The preferred method of dealing with the corresponding order parameter equations is by using finite difference. However, for the case of a binary fluid, a two-distribution lattice Boltzmann approach is also maintained for historical reference.

Users control the operation of the code via a plain text input file; output for various data are available. These data may be visualised using appropriate third-party software (e.g., Paraview). Specific diagnostic output may require alterations to the code.

Potential users should note that the complex fluid simulations enabled by Ludwig can be time consuming, prone to instability, and provide results which are difficult to interpret.

License

The code is developed largely at The University of Edinburgh, and is freely available under a BSD-style license. See https://github.com/ludwig-cf/ludwig.

Contents

• 1. Building and Testing
  • 1.1. Configuration
  • 1.2. Compilation
  • 1.3. Testing
• 2. Input File
  • 2.1. The input file
  • 2.2. Simulation parameters
  • 2.3. The free energy
  • 2.4. Lees Edwards SPBCs
  • 2.5. Colloids
  • 2.6. Plane walls
  • 2.7. Porous Media
  • 2.8. Open boundaries
  • 2.9. Order Parameters
  • 2.10. Viscosity Models
  • 2.11. Force on the fluid
  • 2.12. Parallelism
• 3. Initial Conditions
  • 3.1. Fluid initial conditions
  • 3.2. Initial composition
  • 3.3. Liquid crystal order
  • 3.4. Three-phase model
• 4. Outputs
  • 4.1. Diagnostic Output
  • 4.2. Fluid Output
  • 4.3. Colloid Output
• 5. Tutorial guides
  • 5.1. Poisseuille Flow
  • 5.2. Spinodal Decomposition
  • 5.3. Liquid Crystal Output
  • 5.4. Colloid Postprocessing
  • 5.5. Lees Edwards BCs
  • 5.6. Electrokinetics

Index

• Index

• Search Page