DAFoam: Discrete Adjoint with OpenFOAM for High-fidelity Gradient-based Optimization

DAFoam develops an efficient discrete adjoint method to perform high-fidelity gradient-based design optimization with the MACH-Aero framework. DAFoam has the following features:

• It uses a popular open-source package OpenFOAM for multiphysics analysis.
• It implements an efficient discrete adjoint approach with competitive speed, scalability, accuracy, and compatibility.
• It allows rapid discrete adjoint development for any steady and unsteady OpenFOAM primal solvers with modifying only a few hundred lines of source codes.
• It supports design optimizations for a wide range of disciplines such as aerodynamics, heat transfer, solid mechanics, hydrodynamics, and radiation.

DAFoam is distributed using the GPL-v3 license, and its source code is avaiable from Github

Download the DAFoam image and follow the rest of steps in Get started to run your first DAFoam optimization!