ATTENTION: The primary respository for cuIBM shall henceforth be hosted on GitHub. However, we will continue to push changes to both Github and the old Bitbucket repository and keep the code synchronised.

Currently, cuIBM runs only on Unix-based systems and has been tested on Ubuntu 12.04. It is not supported on Windows.

Please ensure that the following dependencies are installed before compiling cuIBM:

• Git distributed version control system (git)
• GNU C++ Compiler(g++)
• NVIDIA's CUDA Compiler (nvcc)
• CUSP Library (available here)

Install git. On Ubuntu, this can be done via the Terminal using the following command:

> sudo apt-get install git-core

Install g++ using the following command:

> sudo apt-get install g++

Check the version of G++ installed:

> g++ --version

Other development and version control tools can be installed by following the instructions under Step 1 in the CompilingEasyHowTo page on the Ubuntu Community Help Wiki. Software developers will find it useful to install all of them.

Download and install the CUDA Toolkit.

Check the version of NVCC installed:

> nvcc --version

cuIBM has been compiled and tested with NVCC versions 4.1 to 6.0.

IMPORTANT: nvcc-4.1 is compatible only with G++ version 4.5 (g++-4.5) or below. nvcc-4.2 and above are compatible with g++-4.6 and below.

CUSP is a library that provides routines to perform sparse linear algebra computations on Graphics Processing Units. It is written using the CUDA programming language and runs code on compatible NVIDIA GPUs.

CUSP is currently hosted on GitHub. cuIBM has been tested and works with version 0.4.0, available for download here.

The instructions here assume that the CUSP library is to be installed in the folder $HOME/lib, but any other folder with write permissions can be used. Create a local copy of the CUSP library using the following commands:

> mkdir -p $HOME/lib
> cd $HOME/lib
> wget https://github.com/cusplibrary/cusplibrary/archive/0.4.0.zip
> unzip 0.4.0.zip

The folder $HOME/lib/cusplibrary-0.4.0 is now created.

If you wish to use to latest version of CUSP, it can be cloned from the GitHub repository.

> cd $HOME/lib
> git clone https://github.com/cusplibrary/cusplibrary.git

which creates the folder $HOME/lib/cusplibrary.

cuIBM is currently hosted on both Github and BitBucket. The Github repository is the official one, but both will reflect the latest changes.

Run the following commands to create a local copy of the repository in the folder $HOME/src (or any other folder with appropriate read/write/execute permissions):

> mkdir -p $HOME/src
> cd $HOME/src
> git clone https://github.com/barbagroup/cuIBM.git

To compile, set the environment variable CUSP_DIR to point to the directory with the CUSP library. For a bash shell, add the following line to the file ~/.bashrc or ~/.bash_profile:

export CUSP_DIR=/path/to/cusp/folder

which for the present case would be $HOME/lib/cusplibrary-0.4.0.

We also recommend setting the environment variable CUIBM_DIR to point to the location of the cuIBM folder. While the code can be compiled and run without setting this variable, some of the validation scripts provided make use of it.

export CUIBM_DIR=/path/to/cuIBM/folder

which is $HOME/src/cuIBM, as per the above instructions.

Reload the file:

> source ~/.bashrc

Switch to the cuIBM directory:

> cd $HOME/src/cuIBM

or if you've set the environment variable,

> cd $CUIBM_DIR

Compile all the files:

> make

Run the test:

> bin/cuIBM

IMPORTANT: If your NVIDIA card supports only compute capability 1.3, then edit Line 13 of the file Makefile.inc in the cuIBM root directory before compiling: replace compute_20 with compute_13.

The following schemes have been tested for the available solvers:

• Convection

  • EULER_EXPLICIT: First-order Explicit Euler
  • ADAMS_BASHFORTH_2: Second-order Adams-Bashforth

• Diffusion

  • EULER_EXPLICIT: First-order explicit Euler
  • EULER_IMPLICIT: First order implicit Euler
  • CRANK_NICOLSON: Crank-Nicolson

The convection terms are calculated using a conservative symmetric finite-difference scheme, and the diffusion terms are calculated using a second-order central difference scheme.

The following are available in the default installation:

• lidDrivenCavityRe100: Flow in a lid-driven cavity with Reynolds number

• lidDrivenCavityRe1000: Flow in a lid-driven cavity with Reynolds number

• cylinderRe40: Flow over a circular cylinder at Reynolds number 40. The flow eventually reaches a steady state.
• cylinderRe75: Flow over a circular cylinder at Reynolds number 75. The initial flow field has an asymmetric perturbation that triggers instability in the flow and vortex shedding is observed in the wake.
• cylinderRe100: Flow over a circular cylinder at Reynolds number 100. The initial flow field has an asymmetric perturbation that triggers instability in the flow and vortex shedding is observed in the wake.
• cylinderRe150: Flow over a circular cylinder at Reynolds number 150. The initial flow field has an asymmetric perturbation that triggers instability in the flow and vortex shedding is observed in the wake.
• cylinderRe550: Initial flow over an impulsively started cylinder at Reynolds number 550.
• cylinderRe3000: Initial flow over an impulsively started cylinder at Reynolds number 3000.
• flappingRe75: Flow around a flapping airfoil.
• oscillatingCylinders: Flow across two oscillating cylinders.

To run any of the examples:

> make <examplename>

The biggest case (cylinderRe3000) requires a graphics card with 2GB of memory.

The only currently available post-processing script is $CUIBM_DIR/scripts/python/plotVorticity.py. It plots the vorticity field of the flow at all the saved time steps. To display a list of all the command line options (which include the case folder, and the coordinates of the corners of the region of interest), run:

> python $CUIBM_DIR/scripts/python/plotVorticity.py --help

To obtain the vorticity plots, navigate to a case folder (or specify it using the command line option) and run the script:

> python $CUIBM_DIR/scripts/python/plotVorticity.py

• CPU routines do not work.
• Cannot specify which Krylov solver to use for solving the linear systems.
• TairaColoniusSolver and DirectForcingSolver fail if no body is present.
• DirectForcingSolver has not been tested for cases with multiple or moving bodies.

Please e-mail Anush Krishnan if you have any questions, suggestions or feedback.