Containerization of Fluent#

Warning

You need a valid Ansys license to follow the steps in this section.

This document provides instructions and guidelines on how to containerize Fluent for efficient and secure deployment and management.

Prerequisites#

  1. A Linux machine with Docker installed.

  2. A valid Ansys license. Your Ansys reseller should have provided you with one.

  3. PyFluent source. You can clone PyFluent or download the zip from here.

Procedure#

1. Set current working directory#

Within the PyFluent source navigate to the docker directory.

2. Copy needed files#

Specify the pre-installed Ansys directory and docker/fluent_<version> directory for the particular Fluent release as a command line arguments and run this script to copy needed files from the Ansys installation directory to the particular Fluent release directory:

python copy_ansys_files.py <path to 'ansys_inc' directory> <path to 'docker/fluent_<version>' directory>

Note

For Fluent versions 23R1, 23R2, 24R1, 24R2, and 25R1, excludeFluentList.txt includes the path to the Fluent GPU capabilities folder (addons/gpuapp). This means GPU solver functionality is not copied into the container image by default. If your workflow requires GPU-accelerated solving, remove the addons/gpuapp entry from excludeFluentList.txt before running copy_ansys_files.py.

Starting from Fluent 25R2 and later (including 26R1), the GPU capabilities folder is no longer excluded and is included in the container image by default.

  1. Above excluded files are not needed to run typical Fluent workflows.

  2. If you find that some of the excluded files are needed to run your workflows then you can remove those files from the exclusion list.

Selecting solver-only, meshing-only, or both#

You can control which parts of Fluent are copied into the image by adding a third argument to copy_ansys_files.py:

  • Solver only:

    python copy_ansys_files.py <path to 'ansys_inc'> <path to 'docker/fluent_<version>'> solver

    Uses lists: ceiList.txt, cfdpostList.txt, fluentList.txt and excludes excludeCEIList.txt, excludeFluentList.txt.

  • Meshing only:

    python copy_ansys_files.py <path to 'ansys_inc'> <path to 'docker/fluent_<version>'> meshing

    Uses lists: cadList.txt, fluentList.txt and excludes excludeFluentList.txt.

  • Solver + Meshing (default):

    python copy_ansys_files.py <path to 'ansys_inc'> <path to 'docker/fluent_<version>'>

    Uses all lists: cadList.txt, ceiList.txt, cfdpostList.txt, fluentList.txt and excludes excludeCEIList.txt, excludeFluentList.txt.

3. Build the Docker image#

Specify docker/fluent_<version> directory for the particular Fluent release and execute this command to build the Docker image:

sudo docker build -t ansys_inc <path to 'docker/fluent_<version>' directory>

The Docker container configuration needed to build the container image is described in the docker/fluent_<version>/Dockerfile.

Run Docker container using the command line#

When you run the Docker container, you must specify the Ansys license file.

To launch the container in solution mode, use:

sudo docker run -it --name ansys-inc -e ANSYSLMD_LICENSE_FILE=<license file or server> ansys_inc 3ddp -gu

To launch the container in meshing mode, use:

sudo docker run -it --name ansys-inc -e ANSYSLMD_LICENSE_FILE=<license file or server> ansys_inc 3ddp -gu -meshing

Run Docker container using PyFluent#

  1. Install PyFluent.

  2. Use the following Python code to run the container:

import os
import ansys.fluent.core as pyfluent
os.environ["ANSYSLMD_LICENSE_FILE"] = "<license file or server>"
custom_config = {'fluent_image': 'ansys_inc:latest', 'mount_source': f"{os.getcwd()}", 'auto_remove': False}
solver_session = pyfluent.launch_fluent(container_dict=custom_config, user_docker_compose=True)

Run Podman container using the command line#

Follow these steps to pull and run a Fluent container using Podman.

  1. Pull the Docker image into Podman:

sudo podman pull docker-daemon:ansys-inc:latest

  1. Verify the image in the local Podman registry:

sudo podman images

When you run the Podman container, you must specify the Ansys license file.

To launch the container in solution mode, use:

sudo podman run -it --name ansys-inc -e ANSYSLMD_LICENSE_FILE=<license file or server> ansys_inc 3ddp -gu

To launch the container in meshing mode, use:

sudo podman run -it --name ansys-inc -e ANSYSLMD_LICENSE_FILE=<license file or server> ansys_inc 3ddp -gu -meshing

Run Podman container using PyFluent#

  1. Install PyFluent.

  2. Use the following Python code to run the container:

import os
import ansys.fluent.core as pyfluent
os.environ["ANSYSLMD_LICENSE_FILE"] = "<license file or server>"
custom_config = {'fluent_image': 'ansys_inc:latest', 'mount_source': f"{os.getcwd()}", 'auto_remove': False}
solver_session = pyfluent.launch_fluent(container_dict=custom_config, use_podman_compose=True)