SU2

SU2 is an open-source computational fluid dynamics (CFD) platform designed for aerodynamic analysis, multiphysics simulation, and design optimization. Originally developed by the Stanford University Aerospace Design Lab, SU2 has grown into a widely used tool in aerospace research and high-performance engineering environments.

The software allows engineers and researchers to simulate fluid flow behavior around complex geometries such as aircraft wings, turbine blades, and automotive bodies. SU2 supports both compressible and incompressible flow simulations, making it suitable for a wide range of aerodynamic and fluid mechanics problems.

One of SU2’s defining strengths is its design optimization capability. Engineers can automatically adjust geometry parameters and evaluate multiple design iterations to improve aerodynamic performance, reduce drag, or optimize thermal behavior.

Because it is open source, SU2 can be extended and customized for research workflows. It is commonly used in aerospace engineering programs, CFD research labs, and high-performance computing environments.

Key Features

• Discrete adjoint solver computing sensitivities of aerodynamic objectives with respect to thousands of design variables simultaneously at near single-solver computational cost
• SU2_CFD covering Euler, Navier-Stokes, and RANS flow with Spalart-Allmaras and SST turbulence models, explicit and implicit time integration, and MPI parallelism
• Gradient-based aerodynamic shape optimization through automated design loops coupling CFD, adjoint computation, mesh deformation, and shape parameterization
• Non-ideal compressible CFD for dense vapor and supercritical fluid modeling in ORC turbine and supercritical CO₂ applications
• Turbomachinery solver extensions with mixing-plane formulation, non-reflective boundary conditions, and multistage axial and radial turbomachinery support
• Incompressible flow solver with conjugate heat transfer for thermal management and cooled component analysis
• SU2_DEF mesh deformation using Free Form Deformation and Hicks-Henne parameterization for 3D and 2D shape optimization
• SU2_MSH adaptive mesh refinement guided by adjoint solutions for efficient high-fidelity computation
• Hybrid MPI-OpenMP parallel discrete adjoint computation scaling across large distributed-memory HPC clusters
• Python wrapper and API for programmatic integration with machine learning frameworks, uncertainty quantification tools, and external optimization drivers

Best For

Aerospace engineers, CFD researchers, turbomachinery designers, and computational engineering teams at universities, national laboratories, and advanced industrial R&D groups who need gradient-based aerodynamic shape optimization, adjoint-driven sensitivity analysis, and high-fidelity multiphysics simulation on unstructured meshes.

Particularly suited to teams with the numerical methods background and Linux HPC infrastructure to deploy and operate a research-grade open-source solver without vendor support.

Who It’s Not For

Engineering teams who need a commercially supported, GUI-driven CFD tool with integrated pre-processing, meshing, and post-processing in a validated industrial workflow. ANSYS Fluent, Simcenter STAR-CCM+, and SimScale serve those organizations more directly.

SU2 rewards deep investment in CFD numerics and HPC infrastructure. Teams without familiarity with adjoint methods, unstructured mesh generation, and Linux environments will find the barrier to productive use significantly higher than commercial alternatives.

Platform

Linux, macOS, and Windows. Linux is the primary platform and the environment in which HPC cluster deployment operates most reliably.

Windows support available through native build and Cygwin. Python 3.x required for the optimization driver and Python wrapper. MPI library required for parallel execution.

SU2GUI browser-based graphical interface available for users who prefer visual configuration over config-file-based setup.

Paraview and Tecplot compatible output for post-processing visualization.

Pricing

Completely free and open source under the GNU Lesser General Public License v2.1. Commercial use permitted without royalty or notification.

The SU2 Foundation accepts donations and organizational sponsorships to sustain development.

Community support through the CFD-Online SU2 forum, GitHub issue tracking, and a Slack workspace for active contributors.

Commercial support contracts are not offered by the SU2 Foundation. Organizations requiring guaranteed response SLAs need to engage third-party CFD consulting firms familiar with the codebase.

Pros

• Discrete adjoint solver enables gradient-based aerodynamic shape optimization at a computational cost and accessibility that no commercial open-source tool matches
• LGPL license permits integration into commercial and proprietary engineering workflows without license fee or restriction
• Hybrid MPI-OpenMP parallel adjoint scaling supports large-scale shape optimization on modern distributed HPC cluster architectures
• Python wrapper and API enable tight integration with machine learning optimization drivers, uncertainty quantification frameworks, and external design tools
• Active research community continuously expanding turbomachinery, ORC fluid, hypersonics, and FSI capabilities through academic and industrial contributions
• SU2GUI browser interface lowers the configuration barrier for engineers who need CFD analysis without deep config-file workflow familiarity

Cons

• No commercial support with community forums and GitHub as the primary help channels and no guaranteed response time or vendor escalation path
• Steep learning curve requiring familiarity with CFD numerics, adjoint methods, unstructured mesh generation, and Linux HPC environments before productive use
• No integrated meshing or pre-processing; SU2 ingests meshes from external tools such as GMSH, CGNS, and ANSYS Meshing rather than generating them internally
• Post-processing requires ParaView, Tecplot, or equivalent external visualization tools beyond the SU2GUI
• Windows and macOS builds are functional but less tested than Linux, and HPC cluster deployment is effectively Linux-only in practice

Rating

⭐ 4.3 / 5

Editorial Take

SU2 is a powerful research-focused CFD platform for aerodynamic analysis and optimization. Its open-source nature and strong design optimization tools make it particularly valuable for aerospace research and advanced engineering simulations.

Alternatives

OpenFOAM, ANSYS Fluent, Simcenter STAR-CCM+, DAFoam, MACH-Aero, ADflow, SimScale

Used In

• Aerodynamic shape optimization for aircraft wings and fuselages

• High-lift system design and optimization

• Multistage turbomachinery aerodynamic design

• Wind turbine blade shape optimization

• Supersonic and hypersonic vehicle aerodynamics

• ORC turbine and supercritical fluid flow analysis

• Academic CFD and adjoint methods research

• NASA and national laboratory aerosciences programs

Founded

2012 (Stanford University Aerospace Design Lab)