The Art of Computational Fluid Dynamics (CFD)
Most analysts will agree that CFD is often regarded as an art because of its highly complex and non-linear nature. We’ve all gone through the growing pains of trying to get a new class of sequences to converge, or even past the first iteration, all-the-while, ensuring accuracy of CFD simulations.
Reasons for this are numerous; sometimes it’s due to the underlying physics, sometimes solver design and limitations get in the way. In the latest Ansys release, the team decided to abstain from big ‘flashy new features’, instead, concentrating on core improvements.
Consequently, the latest Ansys release has put forth a massive effort in significantly increasing speed, usability and robustness. This encompasses all Ansys fluids simulation offerings: Fluent and CFX.
At the end of the day, analysts can get their job done more efficiently and reliably. This is the core objective for both SimuTech Group and Ansys.
New Ansys CFD Enhancements | Organizing Detailed Settings by Project
One impressive step forward in Ansys Fluent’s usability is exemplified in the addition of a beta feature which automatically adjusts detailed settings based on the type of problem being solved. This minimizes the amount of tweaking that an analyst has to do when tackling a new type of problem.
Also, lists can be now shown in a hierarchy based on name prefixes. So if you have a large number of surfaces all starting with the same prefix, they will be clumped together and indented for easy visualization and multi-selection.
Clicking the prefix in the list selects all surfaces with that prefix, while CTRL + clicking expands the list so you can see which surfaces had that prefix. This behavior can be toggled on and off.
In CFX, an engineer was often forced to code CEL (CFX Expression Language) or Fortran routines to monitor and post-process advanced quantities. As one example of improvements in this area, CFX has introduced means of monitoring of aerodynamic damping during simulations involving blade motion.
Furthermore, a new robust outlet boundary condition was added to handle turbo machinery applications, in which the entire operating curve of the device can be simulated, from stall to choke with less user intervention.
These features compliment an already comprehensive suite of turbomachinery and general CFD tools, meaning that the engineer spends less time customizing code, and more time focused on the engineering problem at hand.
The Role of AI | Ensuring Accuracy of CFD Simulations
Peripheral utilities such as documentation or help files also play large role in successful learning and simulation implementation. On this front, ANSYS has significantly revamped the help files, making the search tools a lot more powerful and organized.
It now allows filtering by physics, application, type of manual (tutorial, user guide etc), and by book. It also intelligently interprets queries and suggests alternatives if search is made with mistakes.
Parallel enhancements have been made to the already impressive scalability of both Fluent and CFX. Scaling on larger numbers of processors is more linear, even when the partitions are small and parallel communication is high.
GPU Augmentations | CFD Hardware Optimizations
One notable leap forward is the introduction of GPU computing in Fluent. The Fluent AMG solver now takes advantage of single or multiple GPU’s in the pressure-based solver. This means that workstations and clusters can be augmented with GPUs to maximize the benefits of hardware and software investments.
Improved CFD Usability | Fluids Applications
These features amongst many other improvements in the details of the workflow have significantly impacted the usability of the Fluids applications. At the same time, ensuring accuracy of CFD Simulations.
Additionally, coupled with optimizations in high performance computation. Fluent and CFX have never been quicker in getting you to your engineering goals.