Computational fluid dynamics simulation

Finite element analysis (FEA) has been successfully used for product engineering for decades. Along with that, high-fidelity modeling approaches and more pragmatic ones were continuously developed, which let you obtain sufficiently accurate results faster.

Today, engineers can and must choose the level of accuracy that best fits their needs to answer engineering questions with minimum computational effort. The level of accuracy ranges from high-fidelity modeling techniques that enable the prediction of real behavior within a few percent or even less to quick methods that enable quick trend predictions.

Today, certification and verification processes for FEA simulation tools are well established. They will remain a critical ingredient to the progress of FEA, its reliability and trust in digital twins and its establishment in novel areas. While predictive simulation will continuously reduce the need for expensive measurements and prototyping, it will continue to require rigorous FE methods and best practices validation through experiments. 

Meshfree CFD methodologies offer an appealing alternative approach to mesh-based methods for selected applications. When rapidly getting results is a priority over the highest accuracy, smoothed particle hydrodynamics (SPH) is an efficient tool. However, both methods have their place, and depending on the requirements for time to solution vs. required accuracy, it may be beneficial to choose a mesh-based or meshless approach.

Most of the flows around us and relevant to product developments are turbulent in nature. Over decades, science and industry have established close relations to incorporate turbulence descriptions into the Navier-Stokes equation. For example, meshing the most suited turbulence model for a given application and CFD project heavily depends on accuracy versus simulation speed requirements.

Generally, turbulence modeling can be classified into three main categories: statistical modeling, also known as Reynolds Average Navier-Stokes (RANS), scale-resolving simulation (SRS), like large-eddy simulation (LES) or detached-eddy simulations (DES) and ultimately, direct numerical simulation (DNS), which does not make any modeling assumptions on turbulence.

Learning CFD requires time, dedication, thorough study and practice. It is critical to understand the underlying fundamental physics of fluid dynamics and the Navier-Stokes equation, grasp numerical methods and their limitations and practice the hands-on usage of the actual computational fluid dynamics software tool. Thanks to automation and continuous improvement of User Interfaces in modern computational fluid dynamics software, the barriers to high-fidelity CFD will further decrease on all levels shifting, the scope to exploring results and making simulation-based decisions. It is also critical to understand fundamental fluid dynamics to judge the results and make meaningful engineering decisions based on CFD results.

The choice of hardware for a CFD project really depends on your project, budget, and current priorities. Some recommendations: x86 CPUs have run simulations for ages now. Every solver was initially developed and verified for this platform. Look for CFD hardware with maximum cache – servers, workstations, and laptops. Graphic Processing Units (GPUs) support many solvers nowadays, and software will adapt to this even further. This solution is very energy efficient. Pay close attention that your required solvers are supported and meet the memory requirements of your use case. This lets you get the most out of multi-GPU workstations and GPU clusters. ARM processors support everything but the graphical user interface. This is an approach for cost-efficient computing, especially on cloud services. Provided your computational fluid dynamics simulation tool supports ARM technology. Generally, cloud-based CFD simulation is a straightforward solution. No investment in expensive computer hardware, no idle cost and scalable on demand.

Computational fluid dynamics simulation software is used in a wide range of engineering applications whenever there is a need to understand or predict fluid flow and heat transfer and the resulting effect on the design of a product or system. In industrial product design, computational fluid dynamics simulation has progressed to simulating the multiphysics behavior in complex geometries, enabling companies to fully understand and optimize their product design virtually before building a prototype.

Industries where computational fluid dynamics simulation is widely used include:

• Aerospace
• Automotive
• Chemical
• Consumer products
• Marine (ship design, propulsion systems and engine design)
• Electronics
• Energy (nuclear, oil & gas and power generation)
• Building services
• Life sciences
• Turbomachinery
• Sports
• Other general applications involving fluid flow and heat transfer