CFX-Pre and the CFX User Environment

Setting up a CFD analysis can be simple or complex, depending on the problem being solved. ANSYS CFX offers a pre-processing interface and tools that make sure the simple things are simple, while still providing for the most complex cases and advanced use of the software.

ANSYS CFX-Pre offers a modern, consistent and intuitive interface for the definition of complex CFD problems. As a starting point, it gives complete flexibility in importing and then manipulating meshes (translating, copying, scaling, etc.).

For standard single and multi-phase problems, CFX-Pre offers a quick setup wizard to guide users through the steps from model selection to boundary conditions and solver settings. Similarly for turbomachinery applications, the TurboPre wizard streamlines set up of multistage analyses, automating the definition of component interfaces and helping create standard boundary conditions for rotating machinery.

For general cases, a toolbar guides users through the key steps in problem set-up, using input panels that adapt automatically to the selected models. The evolving problem definition is displayed in a tree view to allow users to directly access any section of it at any time. Once problem definition is complete, users can feed it directly to the ANSYS CFX solver.

In addition, CFX-Pre offers lots of options for customization and automation: user-defined GUI extensions to allow advanced users to create customized input panels with application-specific terminology and explanatory sketches, a customizable model library with pre-defined set-ups for complex or frequently repeated cases, batch execution of macros and recorded session files and full access to the Perl programming language for maximum programmability.

CFX Expression Language (CEL) and CFX Command Language (CCL) form the foundation of the flexibility of ANSYS CFX and its user environment.

CEL is a powerful definition language to allow users to incorporate their own custom models quickly directly in the standard ANSYS CFX GUIs. For example, users can take advantage of CEL to add new physical models, create additional solution variables, define property relationships and set boundary conditions and profiles. CEL syntax is intuitive and easy to learn, and includes many pre-defined functions and operators to allow users to easily customize their simulations in any number of ways.

CCL is the intuitive text-based command-file that can be used to as an alternative to the GUI to access the solver, implement physics, define boundary conditions and set solver parameters. Parametric studies can be quickly defined by editing command files and changing the appropriate values. This also enables you to run ANSYS CFX software in batch mode or to integrate it in optimization and design systems.Solving the highly non-linear Navier-Stokes equations for fluid flow both reliably and robustly is no simple task. A combination of factors help ensure that users get fast, robust and reliable convergence as well as excellent parallel performance, efficient memory usage and accurate answers.

ANSYS CFX – Solver Technology

Since its origins, ANSYS CFX has focused on a solution strategy using coupled algebraic multi-grid techniques to bring users fast and reliable convergence that is completely scalable with mesh size, requires no user input or numerical adjustments, is insensitive to high aspect ratio mesh cells to allow boundary layers to be captured efficiently and accurately and uses second-order advection schemes by default, for maximum accuracy in all simulations. The solver delivers excellent 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.

ANSYS CFX was designed from the ground up for maximum parallel efficiency. This deep-rooted characteristic has become all the more essential since the advent of multi-core processors and cluster computing. With impressive CPU and memory scaling, ANSYS CFX is inherently prepared for the evolution of computing power.
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.

Careful discretization is necessary to provide robust and accurate answers on the range of situations encountered in industrial CFD. ANSYS CFX software’s default “high-resolution” discretization delivers on both counts. This adaptive numeric scheme locally adjusts the discretization 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 accurate solutions for the full range of physics, meshes and element types.
The vast majority of industrial flows are turbulent, so ANSYS CFX has always placed special emphasis on providing leading turbulence models to capture the effects of turbulence accurately and efficiently.

ANSYS CFX Scalability on a 10M node transonic airfoil benchmark example. Data collected on a cluster of AMD Opteron 2218 processors, showing the benefit of high speed interconnects

 

Turbulence Modeling

For statistical turbulence models, ANSYS CFX provides numerous common two-equation models and Reynolds-Stress models. However, particular focus is placed on the widely-tested Shear Stress Transport (SST) turbulence model, as it offers significant advantages for non-equilibrium turbulent boundary layer flows and heat transfer predictions. The SST model is as economical as the widely-used k-ε model, but offers much higher fidelity, especially for separated flows, providing excellent answers on a wide range of flows and near-wall mesh conditions. The SST model is complemented by numerous other turbulence modeling innovations in ANSYS CFX, including an automatic wall treatment for maximum accuracy in wall shear and heat transfer predictions, and a number of extensions to capture effects like streamline curvature.

ANSYS CFX has also led developments of a commercial implementation of a laminar-to-turbulent transition model. Using CFD to predict the location where laminar boundary layer become turbulent is critical to improving efficiency and/or longevity of equipment in turbomachinery, aerospace, marine and many other industries. The Menter-Langtry γ−θ laminar-turbulent transition model™ gives users a powerful tool to capture various types of transition mechanisms in their CFD simulation.

In addition, ANSYS CFX provides a number of scale-resolving turbulence models, like Large and Detached Eddy Simulation (LES and DES) models. A particular highlight here is the development of the novel Scale-Adaptive Simulation (SAS) model. This model provides a steady solution in stable flow regions while resolving turbulence in transient instabilities like massive separation zones, without an explicit grid or timestep dependency. The SAS model has shown excellent results on numerous validation cases, and provides an excellent option for applications in which resolution of turbulence is required.
Optimizing heat transfer is often critical in many types of industrial equipment, like turbine blades, engine blocks and combustors, as well as in the design of buildings and structures. In any such applications, an accurate prediction of convective heat transfer is essential. In many of them, the diffusion of heat in solids and/or heat transfer by radiation also plays an important role.

A heated jet in cross-flow, where predictions with the SAS turbulence model provide much improved predictions compared to a RANS model

 

Heat Transfer and Radiation

ANSYS CFX features the latest technology for combining fluid dynamics solutions using Conjugate Heat Transfer (CHT) for the calculation of thermal conduction through solid materials. Users can create the solid domain meshes for CHT regions independently, and then use Generalize Grid Interfaces (GGI) to attach any non-conformal meshes. Additional related features include the ability to account for heat conduction through thin baffles, thermal resistance at contact areas between solids and through coatings on solid surfaces, and advection in CHT solids due to their motion.

ANSYS CFX also incorporates a wealth of models to capture all types of radiative heat exchange in and between fluids and solids, whether these are fully and semi-transparent to radiation or opaque. The most flexible model is the Monte Carlo model, which simulates the physical interactions between photons and their environment by tracing a representative number of rays through the simulation domain. It can simulate any variation from optically thick to thin (or transparent) media, both in fluids and solids. To maximize efficiency, the radiation mesh can be automatically coarsened in regions where the changes in the radiation field are small.

ANSYS CFX also gives the user the choice of different spectral models to account for wavelength dependencies in their simulation, as well as allowing for scattering effects to be taken into account.
Numerous CFD applications involve not just a single fluid phase, but rather multiple phases. From its roots over two decades ago, ANSYS CFX has gained vast experience in the simulation of such multiphase flows. This experience is built into the current technology. A complete suite of both Euler-Euler and Euler-Lagrange models allow ANSYS CFX simulations to capture the interplay between multiple fluid phases, like gases and liquids, dispersed particles and droplets, and free surfaces. All of these models benefit strongly from the coupled solver technology to achieve robust and scalable multi-phase flow solutions.

Detailed analysis of the fluid flow through the turbine of an automotive turbocharger together with the temperature in the exhaust manifold metal using CHT technology in ANSYS CFX Courtesy of BorgWarner Turbo & Emission Systems

 

Multiphase Flow

The Eulerian multiphase model features a wealth of options to capture the exchange of mass, momentum and energy. This includes numerous drag and non-drag force models, as well as robust models for phase change due to cavitation, evaporation, condensation and boiling. Additionally, Multiple Size Group (MUSIG™) model allows the user to simulate the effect of turbulent break-up and coalescence of different bubble sizes.

The free surface flow option in ANSYS CFX allows the simulations of open channel flow, flow around ship hulls, tank filling and sloshing, Pelton turbines and many other situations. A special compressive discretization scheme is used to maintain a sharp interface at the free surface. Optionally, users can have two distinct velocity fields, to allow for separation to be simulated in conjunction with strong mixing or entrainment.

The Lagrangian particle transport model users to simulate disperse phases discretely distributed in a continuous phase, such as liquid sprays or airborne solid particles. The functionality is extended by a large number of additional models for phenomena such as primary and secondary spray break-up, particle-wall interaction, wall erosion due to particle impact, particle-particle collision and coal combustion.

For disperse phases that equilibrate quickly with their surroundings, such as small bubbles or particles rising in tundishes or settling under gravity in clarifiers, the algebraic slip model in ANSYS CFX is a very efficient option.

Nucleate boiling downstream of spacers in a fuel rod bundle assembly Courtesy of Dr. E. Krepper, FZ Dresden

 

Rotating Machinery

ANSYS CFX is a leading CFD code in this demanding industry, where the requirements in terms of accuracy, robustness and speed are among the highest. Over two decades of use for rotating machinery simulation has ensured that ANSYS CFX provides all the models and infrastructure for accurate, robust and efficient modeling of all types of pumps, fans, compressors and gas and hydraulic turbines. ANSYS provides all needs for turbomachinery analysis, starting with the geometry design and mesh generation tools, ANSYS BladeModeler and ANSYS TurboGrid, made expressly to meet the needs of turbomachinery design and analysis. In addition, the tight connections to ANSYS Structural Mechanics allow Fluid-Structure Interaction (FSI) to be captured whenever required.

Within ANSYS CFX, tailored pre- and post-processing tools complement a full suite of interface models to capture the interaction between rotating and stationary components.

The transient rotor-stator capability resolves the true transient interaction between components for maximum accuracy. It can be applied to individual pairs of blade passages or to the entire 360 degree machine. Setup and use is as simple as it is with the other frame change models and it is even possible to combine transient and steady-state frame change interfaces in the same computation. Complementing this is the inclusion of second-order time differencing which delivers greater transient accuracy.

The stage interface model is a simpler model, which provides faster solutions than the full transient rotor-stator model. It enables a steady state computation to be used by performing circumferential averaging of the variables at the interface.

Another way to model the interaction of rotating and stationary parts is with ANSYS CFX software’s frozen-rotor model, which is useful when the circumferential flow variation that each blade passage experiences is large during a full revolution. With this option, computations are again performed in a steady-state mode, based on the assumption of quasi-steady flow around the rotating component at every rotation angle. The additional rotational effects (coriolis and centrifugal terms) are included in the rotating regions, and the frame change across the sliding interface is accommodated automatically when linking the different regions of the solution.

 

Streamlines shown in cross section through energy recovery turbine of HPB

 

Chemical Reactions and Combustion

Whether simulating combustion design in gas turbines, automotive engines, or coal-fired furnaces, or assessing fire safety in and around buildings and other structures, ANSYS CFX provides a rich framework to model chemical reactions and combustion associated with fluid flow.

A rich library of pre-defined chemical reactions, that can be easily edited and extended by users, as well as the integration of ANSYS Reactive Integrated Flamelet (RIF) for detailed chemistry tables, provide a complete range of options for all situations. These are rounded out with models for auto- and spark ignition, pollutant formation (NOx, soot), residual exhaust gases, knock, wall quenching, flame extinction and more.

The Eddy-Dissipation Model (EDM) and Finite Rate Chemistry (FRC) models are provided in ANSYS CFX for relatively fast and slow reactions, respectively, in comparison to the mixing of reactants due to the turbulent fluid flow. Simulations are, however, not limited to either extreme: the two models can be combined, with the reaction rate being taken as the minimum of the two, both for single and multi-step reactions, from pre-defined or user-defined reactions.

In situations where fuel and oxidant are fed into a system separately, and the chemistry is assumed to be relatively fast,  the laminar flamelet model with presumed Probability Density Function (PDF) offers a practical and efficient means to depict the detailed of chemistry of hundreds of species, without having to solve hundreds of transport equations.

The Burning Velocity Model (BVM) is well suited for combustion in which oxidant and fuel are pre-mixed or partially pre-mixed, and the flames are steady, such as in gas turbines. It is also coupled with the laminar flamelet PDF model to model post-flame front mixing and reaction.

Like the BVM model, the Extended Coherent Flamelet Model (ECFM) is also suited for pre-mixed or partially pre-mixed fuel/oxidant combinations, and can also capture post-flame front mixing using laminar flamelet PDF model. An additional degree of freedom makes it, however, more suitable for unsteady flames and moving geometries, as found, for example, in internal combustion engines. It also includes an option to include the quenching effect walls can have on flames.

Instantaneous and time-averaged temperature predictions in the simulation of a combustion chamber using the scale-resolving SAS turbulence model. Courtesy of German Aerospace Center (DLR), Institute of Combustion Technology

 

Material Properties

The detailed behaviour of materials under the influence of flow conditions like pressure or temperature can have a critical effect on the accuracy of CFD predictions. ANSYS CFX provides a wide range of material modeling options to ensure nothing stands in the way of achieving the highest fidelity solutions possible.

ANSYS CFX comes with a rich database of material properties for a large range of liquids, gases, and solids. Both ideal and real fluid behaviour can be modeled, using well-established and advanced equations of state.  Numerous relations are also available for viscosity and conductivity variations, from Sutherland’s formula to models based on kinetic theory. For non-Newtonian fluids, an ample selection of viscosity models is provided to account for their shear-rate dependent behaviour.

Should a simulation involve a proprietary material, or any other material or material property not already included in the material database, ANSYS CFX users can take advantage of the flexibility of the CFX User Environment and the power of the CFX Expression Language (CEL). CEL allows easy definition of any number of new materials or dependencies of material properties on flow conditions like pressure, temperature, shear-strain rate and more. Users can enter any algebraic expressions for such custom models directly in the CFX-Pre GUI, avoiding the need to program separate external routines and using the simple CEL syntax. It is as straight-forward as writing them down on paper!

 

Prediction of wetness dispersion under non-equilibrium conditions for quantification of thermodynamic performance in a low-pressure steam turbine Courtesy of Siemens AG Ref: Gerber A.G., Sigg R., Völker L., Casey M.V., Sürken N., “Predictions of Non-Equilibrium Phase Transition in a Model Low-Pressure Steam Turbine” Journal of Engineering for Power and Energy, Vol 221, No A6, September 2007, pp 735-744

 

Modeling the Interaction Between Fluids and Solids

Beyond the powerful range of options for capturing FSI with ANSYS structural mechanics software, many options exist directly within ANSYS CFX to model the effect of solid motion on fluid flow.

The robust and flexible algorithm to deform a given fluid volume mesh in ANSYS CFX tolerates even very large boundary displacements. These displacements may be defined explicitly by the user, e.g., using CEL, or be the implicit result of an FSI simulation with ANSYS Structural Mechanics software, or from the rigid body solver within ANSYS CFX. In all cases, boundary displacements are diffused into the interior volume mesh, while ensuring that small or near-wall elements are deformed less. This maintains good boundary layer resolution and allows for larger mesh deformations with a single mesh topology.

In situations where the boundary motion is more extreme, and a single mesh topology is simply insufficient to model the entire displacement, ANSYS CFX provides options for automatic re-meshing, when required during a simulation. The automatic re-meshing allows users to connect to ANSYS ICEM CFD, exploiting its scripted, batch meshing capabilities to drive the automatic re-meshing from within ANSYS CFX. Alternatively, users can integrate any other scriptable meshing software.

The immersed solids method is an additional FSI option in ANSYS CFX that allows simulation of unlimited motion of solid objects through fluid regions, as it avoids any mesh deformation or re-meshing. As an immersed solid passes through a fluid, the region of overlap is determined and the fluid solution is adjusted to reflect the presence of the solid by applying appropriate source terms. The solid motion can be defined by the user, with complete flexibility, or it can be an implicit result from the rigid body solver within ANSYS CFX.

A fully integrated and implicit 6-degree-of-freedom rigid body solver (beta functionality in Release 12) in ANSYS CFX permits boundary, domain, or sub-domain motion to be an implicit result of the forces and moments acting on a rigid body of a given mass and defined moments of inertia. Applications include store separation from aircraft and ship motion under the influence of waves.

The flow through a screw pump simulated using the immersed solid technology to capture the motion of the rotors