 |
 |
 |
 |
 |
 |
 |
The ANSYS CFX Solver
At the heart of ANSYS CFX is the coupled algebraic multigrid solver. Simply put, it produces accurate solutions to the linear equations with fast and reliable convergence.
The ANSYS CFX multigrid method has been evolving for more than 20 years and has been used to solve literally millions of simulations. It is:
ANSYS CFX: Fast, Reliable, Robust, Accurate Numerics 
Grid Refinement – Relative Number of Grid Nodes
The Coupled Multi-Grid Solver
Solving the Navier-Stokes equations quickly and reliably is a vital aspect of CFD analysis. ANSYS CFX software uses a unique solution strategy, based on coupled multigrid solver technology, that surpasses existing CFD methods in speed and robustness.
The ANSYS CFX solver is an efficient and powerful CFD engine, with high parallel efficiency and innovative smoothing and coarsening algorithms. The solver has been vigorously validated and is continuously improved. The sover is able to handle complex flow regimes, such as multiphase or reacting flows.
For large meshes which are increasingly required in engineering simulations, this translates into significant and predictable reductions in run times and faster project turnaround.
When combined with CFX’s outstanding parallelization capability, the speed at which free-surface flows now can be solved is revolutionary.
Coupled Solution
Many CFD codes use a segregated, semi-implicit solution scheme, where each equation is solved in sequence. In this approach, direct inter-equation coupling is neglected, and only accounted for in a deferred manner, requiring additional iterations to achieve convergence. The de-coupled nature of such methods also results in degraded non-linear performance as the grid is refined.
ANSYS CFX radically improves performance by solving the full hydrodynamic system of equations simultaneously across all grid nodes. This technique provides a robust and reliable solver which requires far fewer iterations to converge.
The coupled solver delivers better performance on all types of problems, but is particularly powerful in flows where inter-equation coupling is significant. Examples of this include rotating flow with strong Coriolis terms, combusting flows and high-speed flow with strong pressure gradients.
Multigrid: Performance in Fine Grids
The second important aspect of the ANSYS CFX solver is its multigrid approach. While the coupled aspect of the solver deals with local effects, the multigrid solver effectively deals with the long distance or 'long wavelength' effects.
This approach automatically generates a cascade of successively coarser grids, which allows the solution information to propagate rapidly across the entire computational domain. The solutions on the coarser meshes are used to accelerate the original fine grid solution.
Also, iterations performed on the coarse mesh are proportionally less expensive than finer grid iterations, so it is clear that these accelerations are also economical.
It is critical that numerical errors reduce rapidly and reliably as the mesh is refined. Numerical errors occur in the key transport terms advection, diffusion and in sources. The element technology of CFX yields accurate local gradients for diffusion and sub-grid source resolution. All of the industry-accepted first and second order advection schemes are offered. CFX is the only major commercial CFD software that enables second order advection by default and is robust and accurate.
Memory Usage
The ANSYS CFX solver provides high memory efficiency. One million unstructured tetrahedral mesh element problems can be run in 400 MB RAM. The software intelligently uses the memory available in order to dynamically optimize the balance of resource usage against computational speed.
Advanced Numerics
Careful discretisation is necessary to provide robust and accurate answers on the range of situations encountered in industrial CFD. ANSYS CFX software's default “high-resolution” discretisation delivers both of these. This adaptive numeric scheme locally adjusts the discretisation to be as close to Second-Order as possible, while ensuring the physical boundedness of the solution. The result is demonstrated and reliable second-order solutions for the full range of physics, meshes and element types.
