Fluent Guided Meshing Workflows

Mesh Generation

• The Rapid Octree mesher is now improved in the following ways:

  • When creating a volumetric refinement region for the Rapid Octree mesher, the ability to define it's extents and orientation using a body of influence (BOI) is now supported as a full feature. A BOI is an arbitrary closed surface that you have previously imported or created, and may represent your intended refinement region better than the other options (that is, a box or frustum). (Refinement Regions)

  • When using the boundary projection treatment method with the Rapid Octree mesher, the ability to select a discrete mesh optimization (DMO) scheme for boundary mesh optimization is now supported as a full feature. The DMO scheme can result in significantly less defeaturing compared to the scheme available in previous releases (which is still available, and is now referred to as the conjugate gradient scheme). (Boundary Mesh Optimization)

  • When using the boundary projection treatment method with the Rapid Octree mesher, the ability to specify surface sizing definitions that have proximity refinement settings is now supported as a full feature. Such proximity refinement settings may be useful when you have gaps and smaller openings, as it can avoid the unwanted closing of flow paths due to defeaturing. (Custom Boundary Sizes)

  • When specifying the volume to be meshed by the Rapid Octree mesher, you can now use generic region definitions, each of which allow you to specify a material point and the associated face zones that are contained in or bound the volume. Such definitions can be helpful when you have a geometry where face zones overlap or should be discarded (which can introduce ambiguity in the boundary face zone assignment) or for a case where only a smaller sub-volume of a complex model is to be meshed. (Specifying the Volume)

  • When defining the geometry for the Rapid Octree mesher, you can now specify a reference size, which is used to expand the currently defined bounding box and thereby adjusts what cell sizes are possible in the mesh. This can be a more convenient way to precisely adjust the cell size (compared to manually changing the bounding box extents), and is especially helpful when the geometry that you input has exterior boundaries (for an interior flow problem, for example). (Defining the Reference Size)

Licenses

• A new capability level named CFD HPC Ultimate is available for the Fluent CPU or GPU solver. This license includes the capabilities of the CFD Enterprise license (as described in Program Capabilities) and also allows the solvers to run on any number of CPU cores or GPUs without using additional HPC licenses. It can be selected in Fluent Launcher (as described in Selecting the Licensing Level) or with the command line option -license=hpc_ultimate.

  The CFD HPC Ultimate license was first introduced in service pack 2025 R1.01, and is now improved in the following ways:

  • Aero, Icing, and Polyflow workspaces are now supported with the CFD HPC Ultimate license.

  • Concurrent parametric workflows in Fluent are now supported with the CFD HPC Ultimate license.

User Experience

• Access to Ansys Engineering Copilot (Ansys Engineering Copilot User's Guide) is added to the Learning and Support ribbon tab.

• You can now submit batch jobs using the Fluent Launcher. Refer to Setting General Options in Fluent Launcher in the Fluent User's Guide for additional information.

Files

• You can save 3D data (pictures) in USD format for use in NVIDIA Omniverse. Refer to Using the Save Picture Dialog Box in the Fluent User's Guide for additional information.

• You can write case settings corresponding to the Results node in the Outline View tree to .txt and HTML format. Refer to Exporting, Copying, and Importing Case Settings in TSV, Text, HTML Format in the Fluent User's Guide for additional information.

• The third-party software VKI is upgraded to Ceetron-SAM 2.1.0 to address security concerns for file import/export.

• CGNS is updated to version 4.5.

• Transient data export to EnSight DVS format now automatically includes all defined report definitions. Residuals data is also available.

• The user interface is now improved for reading the lightweight data of a case or mesh file, that is, reading only the case settings of the file without any of the data associated with the mesh. When setting up a case that has a large cell count, this allows you to review and/or modify the settings without having to wait for the reading of the mesh, on a machine that does not have the memory needed for the mesh (such as a personal computer or laptop).

  You can also now display surface meshes in the graphics window when you have read lightweight data. This can be helpful when setting up boundary conditions, monitors, animations, and so on.

  For further details on reading lightweight data, see Reading the Lightweight Data of Mesh or Case Files.

Solver-Meshing

• You can now easily disconnect one or more cell zones from all adjacent cell zones to which they are connected through shared faces and/or nodes, so that the cell zones can move relative to each other (for example, as part of a sliding mesh or dynamic mesh simulation). (Disconnecting Cell Zones)

• When using a dynamic mesh zone in a simulation that models ablation or erosion / accretion, the handling of sharp corners in the mesh is now improved, which may improve performance, accuracy and robustness. For details about ablation or erosion / accretion, see The Ablation Condition at Wall Boundaries and Particle Erosion Coupled with Dynamic Meshes, respectively.

• For overset meshes, you can now manually identify component meshes as collar meshes, so that they are sure to be protected from being cut by any other boundary during the hole cutting step. This can be useful in instances when the automatic treatment of overlaps is insufficient and when other hole cutting control methods are not as convenient to set up. (Defining Collar Meshes)

• When creating non-conformal mesh interfaces, the text command for enabling adjustable tolerances (define/mesh-interfaces/auto-options/set-one-to-one-pairing-tolerance) is improved. You now have the option of specifying an absolute tolerance, so that you can directly define the distance that is used in the pairing of boundary zones for mesh interfaces (rather than having to specify a relative length factor that is then multiplied by the largest mesh length scale of the interface boundaries). This provides you with greater control, and may allow you to eliminate unnecessary mesh interfaces or produce missing mesh interfaces. (Using a Non-Conformal Mesh in Ansys Fluent)

Cell Zones and Boundary Conditions

• The perforated wall model is now supported with the density-based solver.

• The following improvements are now available to assist with the proper setup of adjacent cell zones with different materials:

  • When changing the material of a cell zone, a warning is now printed in the console if an interior boundary zone is detected between that cell zone and an adjacent cell zone with a different material but the same type (fluid or solid).

  • During a mesh check, the check now fails if an interior boundary zone is detected between cell zones of different materials, and (if the mesh/check-verbosity text command is set to 3) a warning is now printed in the console. Note that the check will not fail if the case uses the battery model, heat exchanger model, or the solid oxide fuel cells (SOFC) model, as these models allow interior zones between different materials.

  • It is now easy to slit all interior boundary zones that are between not only solid but also fluid cell zones of different materials to create coupled wall / wall-shadow pairs, through the mesh/modify-zones/slit-interior-between-diff-materials text command.

  • In previous releases, when you fuse two uncoupled face zones (for example, an uncoupled wall / wall-shadow pair), this always produces a single face zone of type interior. Now the Fluent software checks during the fusing operation if the adjacent cell zones are of different materials or material types, and if they are, automatically slits the resulting interior zone into a coupled wall / wall-shadow pair, so that it is no longer necessary to slit it manually.

Materials

• The non-Newtonian Carreau viscosity model now uses a more general formulation of the Yasuda-Carreau model, which includes a shape parameter that controls the transition of the viscosity profile between zero-shear and infinite-shear viscosities. (Power Law for Non-Newtonian Viscosity)

Moving Reference Frames

• For cases that have one or more moving reference frame zones, the ability to use a time-step-size-independent continuity discretization is now fully supported. Such discretization reduces the degree to which the results are dependent on the time step size for transient simulations that use frame motion zones with the pressure-based solver. (Solution Strategies for a Single Moving Reference Frame)

Turbulence

• Optimized LES Numerics are now available for the GPU solver, which typically allows for a reduction in the number of iterations used per time step. (Optimized LES Numerics )

• For the GEKO model with the GPU solver, the Wall Distance Free formulation is now supported. For details, see Generalized k-ω  (GEKO) Model.

• The Viscous Heating option label in the Viscous Model dialog box is now changed to Viscous Work, in order to better reflect the effect of this option.

Turbo Setup Workflow

• Fluent's Turbo Workflow has been renamed to the Turbo Setup Workflow. For more information, see Using the Turbomachinery Setup Guided Workflow.

• Fluent's Turbo Workflow has been enhanced such that, if a completed setup is subsequently reopened, you can choose to hide or disable the Workflow tab in the user interface. Additionally, if there are no associated workflow files available, then the Workflow tab is now hidden by default. For more information, see Using the Turbomachinery Setup Guided Workflow.

Turbomachinery

• When Turbo Models is enabled, the Dynamic Mesh Zones dialog box now allows you to select one or more cell zones to create multiple dynamic mesh zones of the same type (for example, blades) simultaneously. (Creating and Applying Dynamic Mesh Zones)

Reacting Flows

• The dual-potential cell zone condition is now available for modeling electrochemical reaction in a porous medium. (Electrochemical Reaction in Porous Medium)

Discrete Phase Model

• For contour plots of DPM particle sampling on a planar surface, the <surfname>_signif_conc_estimate and <surfname>_signif_conc_time_integral expressions have been added to evaluate the local discrete phase concentration. (Contour Plots of DPM Particle Sampling Results on a Planar Surface)

Multiphase Models

• An enhancement to the source term linearization for the Wet Steam model is available to improve robustness and convergence speed in some cases. (Setting Up the Wet Steam Model)

• The quadrature-based moment method (QBMM) has been released as a full feature. The method can be helpful for cases where a higher order of accuracy for solving population balance equations is preferred. This was a beta feature in previous release. (The Quadrature-Based Moments Method (QBMM))

• The Lee evaporation/condensation model now supports a user-defined method for specifying the From Phase Frequency and To Phase Frequency model parameters using a DEFINE_MASS_TR_PROPERTY user-defined function (UDF).

• For the mixture multiphase simulations that involve the semi-mechanistic boiling model, the following enhancements have been implemented (Using the Semi-Mechanistic Boiling Model):

  • You can now specify Vapor Heat Transfer Coefficient for the single phase heat flux. For clarity, the previous Heat Transfer Coefficient parameter has been renamed to Liquid Heat Transfer Coefficient.

  • When the semi-mechanistic boiling model is used, the Boiling Model Expert Options are now always visible in the Evaporation-Condensation Model dialog box. In previous releases, they were displayed by using the text user interface (TUI).

  • The Critical Liquid Volume Fraction is now limited by a maximum value of 0.65.

Eulerian Wall Film Model

• Film mass and heat fluxes through source terms are now include in the evaluation of the film mass flow rate and film heat transfer rate in the film flux reports. Film heat source term input now must be in the form of (here, is the film mass, is the film liquid specific heat, is the film temperature, and is the reference temperature of 298.15K).

Electric Potential Field and Electrochemistry Model

• The swelling model is now available for the resolved lithium-ion battery model. (Lithium-ion Battery Swelling Model in the Fluent Theory Guide)

• For the electrolysis and H2 pump model, the following features have been added:

  • A new device, PEM fuel cell, is now available. This device can be used for modeling PEM fuel cell involving multiphase flows. (Electrolysis and H2 Pump Model)

  • A new device, high temperature PEM electrolysis, is now available. This device is suitable for modeling single-phase electrolyzers that operate at high temperatures. (Electrolysis and H2 Pump Model)

  • The unresolved 1D modeling approach has been added. This approach is an extension of the existing unresolved 0D approach and accounts for water transport across the unresolved membrane electrode assemblies (MEA) zones. (Unresolved 1D Modeling Approach)

  The swelling model is now available for the resolved lithium-ion battery model. (Lithium-ion Battery Swelling Model in the Fluent Theory Guide)

Battery Model

• A new, more general thermal abuse model, called N-equation kinetics model, has been added. This model incorporates a user-specified number of user-defined thermal abuse reactions. (General N-Equation Model)

Solver Initialization

• A text command is now available to flush the profile memory for a boundary. In some cases with solution-dependent profiles on boundary zones, re-initialization may not remove all data from the boundary and a complete flush may be needed. (Flushing Profile Memory on Boundary Zones)

• When using the pressure-based solver with a turbulence model, you can now enable an alternative treatment option, such that the effects of turbulence wall functions are removed at walls where you have specified zero shear. This treatment may enhance convergence and solution behavior. (Specified Shear)

Fluent Native GPU Solver

• It is now possible to enable many-to-many CPU/GPU remapping for the native GPU solver using Fluent Launcher. (CPU/GPU Remapping)

• The exporting of solution data calculated by the native GPU solver as EnSight DVS files for postprocessing in EnSight is now supported as a full feature. As part of the export, you can specify the file name, cell zones and/or surfaces, quantities, and so on, similar to when exporting from the CPU solver. Exporting after a calculation (either steady or transient) or during a transient calculation is supported. (Exporting Solution Data as EnSight DVS Files)

• Cylindrical profiles are now supported for cell zones and boundary conditions.

• The 3D fan cell zone condition is now supported.

• The following boundary conditions are now supported:

  • Inlet Vent

  • Exhaust Fan

• GPU Solver performance improvements have been made for the following models:

  • S2S model with non-changing geometries/boundary conditions

  • Sliding mesh

  • Flamelet Generated Manifold (FGM)

• Modeling electric potential is now supported.

• You can now track the current and peak memory of GPUs.

• Viscous work is now supported for applicable turbulence models.

• The ability to enable asynchronous outputting for the native GPU solver is now supported as a full feature. This can improve the solver performance during the calculation (with the speed increasing by up to a factor of 2) when postprocessing monitors, rendering animations, autosaving case / data files, and/or automatically exporting data. (Asynchronous Outputting).

• The use of expressions for report definitions and unsteady statistics with the GPU solver is now supported as a full feature for fully supported models. For a list of supported fields and other details, see Using Expressions with Report Definitions and Unsteady Statistics.

• The Discrete Phase Model (DPM) features supported with the GPU solver have been upgraded from beta to release. For details, refer to the following sections:

  • Discrete Phase Model

  • ????

  • Limitations for the DPM

  • Solver Settings for the DPM

  • Postprocessing for the DPM

Density-based Solver

• The compatibility between the two-temperature model and CHEMKIN mechanism import has been enhanced. This enhancement allows you to import CHEMKIN format kinetic mechanism, thermodynamic database, and transport property database into the two-temperature model. Additionally, the issue of Blottner curve fit coefficients being incorrectly set to zero during the CHEMKIN import has been addressed. 

Graphics, Reporting, and Postprocessing

• Annotations are now associated with graphics objects (mesh, contours, vectors, pathlines, particle tracks, LIC plots) once they are added to the display. This means on re-display, any associated annotations are displayed along with the graphics object. Refer to Adding Text to the Graphics Window in the Fluent User's Guide for additional information.

• The ranges for X and Y axes on 2D plots, such as XY Plots and Histograms, can now be controlled individually, allowing you to manually limit one range while having the other computed automatically.

• You now have the option to manually specify the increments of major and minor gridlines in 2D plots. Refer to ???? for additional information.

• When printing the memory usage to the console in Windows (as described in Memory Information), the resident memory will now be printed along with the virtual memory by node and by host. Note that for Windows, the virtual memory corresponds to private bytes and the resident memory corresponds to the working set; for further details, see the Microsoft online documentation.

• A preference for surface caching is added to improve performance of mesh redisplay and incremental display, even across mesh objects. This is beneficial for cases of all sizes. Refer to Graphics Performance in the Fluent User's Guide for additional information.

• The preference for improving the display and interactive performance of mesh surfaces is extended to the global mesh display. The preference is Optimize Input Data in the Graphics branch of Preferences (File > Preferences...).

• A new Viscous Dissipation field variable (in the Temperature... category) is now available when the energy equation is enabled and the Viscous Work option is enabled in the Viscous Model dialog box.

Parallel Processing

• The -platform=<x> command line option (which is available only on x86 Linux, and can enhance performance when running on processors that support the AVX2 instruction set) is improved:

  • The <x>=intel setting is now more reliable. With this setting, AVX2 binaries are used for the host and node processes, along with the Intel MKL sparse LDU smoother.

  • The <x>=amd setting is now fully supported. With this setting, AVX2 binaries are used for the host and node processes.

• When partitioning the mesh for a simulation with multiple processes or GPUs and the Laminar, EDC, or PDF transport model, you can now enable a Stiff Chemistry weighting to more evenly distribute the computational load corresponding to the reaction integration. (Partitioning)

Adjoint Solver Module

• The changes listed below have been made to the design tool settings.

  In the Design Change tab:

  • Morphing method selection has been changed to a drop-down list.

  • The constraint method is now specified by either enabling or disabling the Enhanced Constraint Satisfaction option, for the enhanced constraint method and standard constraint method, respectively.

  • The default morphing method has been changed from Polynomials with the standard constraint method to Polynomials with the enhanced constraint method.

  • Freeform Scaling Scheme settings are now hidden unless the Polynomials morphing method with the standard constraint method is specified.

  • The Applied Conditions list has been renamed to Applied Constraints.

  In the Region Conditions tab:

  • The Invariant option has been renamed to Shape Preserving.

• The design tool settings visibility option has been added to the design tool which allows for either basic settings or advanced settings to be exposed.

  When basic is selected for the design tool settings visibility, the settings listed below will be hidden.

  Design Change settings:

  • Morphing method

  • Fix surfaces...

  Objectives settings:

  • Increase Value and Decrease Value

  Region Conditions settings:

  • Continuity Order and Definition options for Region Boundary Continuity

  Numerics settings:

  • All Numerics settings are hidden

• When Advanced is selected for design tool settings visibility, the following Numerics settings have been changed:

  • Prescribed motion settings are now hidden when using the polynomials morphing method with the standard constraint method (can still be specified using the TUI).

  • In the Freeform Motion group box, Constraint Relaxation and Parameter Relaxation settings are now hidden (can still be specified using the TUI).

  • The Tolerances settings group box has been renamed to Residual Tolerances.

Parametric Studies

• Parametric reports exported as PowerPoint slides have various enhancements, such as an embedded Excel spreadsheet representation of the parametric table, as well as the ability to select specific filtered XY plots and slider-based images and image comparisons and add them to your exported PowerPoint presentations. For more information, see Enhancing Exported PowerPoint Parametric Reports.

• You are able to access visualization tools specific for your optiSLang-generated design point study. For more information, see Visualizing Your Parametric Study Data Using optiSLang.

• You can specify various ways to customize the report-related settings that will be applied while design points are updated in the parametric study. See Automatically Setting Design Point Report Options for more information.

Cloud Computing

• Signing into Ansys Cloud from within Fluent now opens a separate authentication panel within Fluent itself by default. You can choose to authenticate using a separate web browser page using the appropriate preference. See Accessing Ansys Cloud for more information about Ansys Cloud within Fluent.

System Coupling

• Rigid Body FSI is now available via System Coupling and Fluent. (Supported Capabilities and Limitations)

Expressions

• You can rename named expressions, including those referenced by other expressions and those in use in boundary conditions and elsewhere. The new name of a renamed expression will automatically be updated at the dependent locations, ensuring they do not become invalid. Refer to Named Expressions in the Fluent User's Guide for additional information.

Web Interface

• Various graphics-based enhancements have been introduced, such a more interactive triad in the View Arc, as well as interactive triads when working with multiple plane surfaces (see Creating and Displaying Multiple Plane Surfaces for more information).

• You can use the Graphics panel to adjust the colors and associations related to displaying objects in the graphics window. See Working With Graphics Objects and Your Simulation Results for more information.

• You can play animations, view individual frames in the animation, as well as pause and resume the animation at any frame. See Performing Calculation Activities for more information.

• Several improvements have been made to panels and selection lists in order to enhance usability. See Property Panels for more information.

• For cell zone and boundary conditions, within their respective context menus and/or panels, you can choose to Display the surface or to Create Surface if one does not already exist.

• When displaying the results of your simulation, you can edit various aspects of the colormap by double-clicking it in the graphics window. See Color Maps for more information.

• You can choose a specific color for any selected object(s) in the graphics window by using the Color by > Uniform color option from the context menu. Use the Esc key to release any selection(s) in the graphics window. Subsequently, you can randomize the colors of unselected objects displayed in the graphics window using Color by > Randomize colors option from the context menu.

• You can set the transparency of selected objects in the graphics window using the Set transparency option. This displays a slider control (and a numerical input control for values greater than or equal to 0 and less than or equal to 100) where you can adjust the visibility of the selected object. Use the Reset transparency option to disable your previous visibility/transparency settings.

• The View options in the View Arc now contains a toggle between the perspective projection view and the orthographic projection view (which itself reveals the ability to optionally display a ruler in the graphics window).

Beta Features

• There are also some exciting new enhancements available as beta features that you may be interested in trying out. Detailed documentation is in the Fluent 2025 R2 Beta Features Manual.

Fluent Icing

Fluent Icing allows you to easily conduct in-flight icing simulations within a dedicated Fluent Application Client environment. The following additional functionality has been added in this release.

• Ice Displacement and Remeshing Updates

  Multishot with remeshing workflows no longer require wall boundaries to be separated along sharp edges to capture these features properly during the wrapping process. This is a major usability upgrade, allowing users to continue the icing process without modifying their original grid boundary topologies.

  The remeshing script has been relocated to the cortex/resources/templates/prjapp/flicing/ directory and is no longer sourced from the FENSAP-ICE installation path.

  Ice smoothing is now enforced at wall/symmetry plane intersections in 3D remeshing simulations to improve the robustness of the surface wrapping process and to reduce the occurrence of prism and tetrahedral meshing failures in these regions.

  The STL Ice Thickness Limit option restricts the iced STL file to iced facets only, based on a user-defined ice thickness threshold. This provides several benefits:

  • STL file size is reduced, saving storage space.

  • Excessive and spotty mesh refinement caused by proximity detection between the STL and the original surface is avoided, resulting in smaller iced meshes during multishot icing.

  • Remeshing time is noticeably reduced by reading a smaller STL file, minimizing proximity sizing checks, and wrapping with a more optimal sizing function.

  • More optimal iced grids with reduced mesh sizes allow larger problems with more detailed ice shapes to be solved using the same computing resources.

  • STLs isolated to iced areas are easier to manipulate for downstream 3D printing processes, where ice shapes can be manufactured and attached to actual aircraft surfaces.

• User Interface, User Experience Upgrades

  The default run output file prefix is now set to out.*, instead of using the case file name. You can still use the case file or a custom prefix by modifying the Output Filename option.

  Numbered file outputs now use six-digit numbering instead of two-digit, improving solution loading and animation workflows, especially in ice protection CHT simulations.

• Non-Conformal Interface Support

  One-to-one matching mesh interfaces are now supported. These interfaces must have good overlap and be manually paired using the Matching option in the Mesh Interfaces panel.

• CFX Turbo Import (Beta)

  A new file import mode allows conversion of multi-domain CFX .res files into .cas and .dat files for use in engine icing simulations. Each domain is separated into individual simulations in the project panel. You can select a subset of domains for import and rearrange their order as needed.

• Mixing Plane Radial Profiles for DPM and Vapor (Beta)

  A mixing plane model has been developed to transfer particles and vapor between turbo rows by redistributing particle sizes into new bins based on their updated sizes and radial positions. The resulting data is stored as a radial profile in a file, which can be reused across runs and grids. The model captures changes due to wall-particle interactions such as splashing (droplets), shattering (ice crystals), and centrifugal effects. This feature enables multi-row engine icing simulations with evolving particle size distributions—functionality not supported in FENSAP-ICE-TURBO.

• Ice Shedding with Parallel Stress Analysis and Crack Propagation (Beta)

  A new multi-CPU stress solver, Iterative – Parallel, has been added to the rotor ice shedding model to accelerate calculations, particularly for propeller blades using unstructured grids. The previous single-CPU solver using the LU method has been renamed Direct – Serial.

Fluent Aero

Fluent Aero allows you to easily explore the aerodynamic performance of aircraft under a wide range of flight regimes, from subsonic to hypersonic conditions, all within a dedicated Fluent Application Client environment. A streamlined workflow guides you through the creation of a matrix of flight conditions or design points where single and multiple flight parameters, such as angle of attack, Mach number, altitude, etc., can vary. Most common models, solvers, and convergence settings of Fluent are tuned using the latest best practices for external aerodynamic problems and are available in Fluent Aero’s user interface. In this manner, simulations can be conducted in a quick and user-friendly environment. The full capabilities of the Fluent Solution workspace remain accessible when its session is displayed through Workspaces → Solution within the Fluent Aero ribbon. Tutorials are available to provide examples on how to conduct exploratory simulations using single and multiple flight parameters.

In this release, a broad range of features have been added and are described below.

• AET: Aerodynamic Extraction Tool

  The Aerodynamic Extraction Tool (AET) has been updated to automatically extract blade sections from a 3D blade using Ansys Discovery™ instead of Ansys SpaceClaim®. Since Discovery and SpaceClaim are used as background processes of the AET, the user experience and overall workflow remains unchanged.

• Ionized data for HFSS

  Fluent Aero now contains an ionized data converter that allows the export of point-cloud data containing permittivity and conductivity plasma properties. This converter can only be used with Fluent Aero species solutions that were obtained at high-speed flow conditions.

  This data can be read by HFSS to analyze how radio frequencies are affected. It replaces an existing Python script, offering enhanced functionalities such as automatic plasma property creation for multiple solutions, optimal cloud point generation to reduce computational costs, and an interactive user interface for visualizing the plasma sheath and volume.

• Mesh Adaption

  The Mesh Adaption feature using PUMA with the Combined Hessian Indicator has been moved to full release to enhance the precision and accuracy of all conditions defined inside Fluent Aero’s design point table while automating the post-processing of all mesh adaption final and cycle solutions.

  Management of cycle solutions generated by mesh adaption simulations have been significantly improved from its previous beta version. For instance:

  • It is now possible to use the Interrupt and Resume commands to pause a cyclical solution to visualize its solution and make appropriate changes to its convergence settings.

  • After completion of a simulation, it is possible to use the Restart and Continue to Update commands in a mesh adaption simulation to either improve the convergence of a given mesh adaption cycle, to increase the number of mesh adaption cycles to conduct, or to carry out more iterations to further converge a final adapted solution.

    Furthermore, you can also now access the Save Results command to store the results of a particular cycle. You may then choose to Run Next DP or Run Next Cycle (Beta) to proceed directly to the next design point or adaptation cycle.

  Post-processing of cycle solutions has been enhanced. Graphics and Scene objects can now be independently selected for image creation and saving from their respective Properties panels. A separate Results object has been introduced to control the image type and output settings. This enables saving images for the current solution (Current Solution), all design points (All Design Points), all cycles of the current design point (Current DP - All Cycles), or all cycles of all design points (All Design Points - All Cycles).

  Detailed documentation on the various ways to interact with a cycle solution is available in the User Guide.

  (Beta) When beta features are enabled, it is now possible to select an intermediate cycle to define as the final cycle of a particular design point calculation.

• Aerodynamic Coefficients: Parameter Search (Beta)

  Similar to Mesh Adaption, improvements were made to the Parameter Search feature to facilitate interacting with design point cycle solutions with Parameter Search enabled. This includes improvements to the Interrupt, Resume, Restart, Continue to Update, Save Results and Run Next DP features.

Polyflow

The Polyflow workspace is a means to explore manufacturing applications of highly viscous materials such as polymer extrusion, blowmolding, fiber spinning, etc. within the Fluent environment.

• Support of CFD HPC Ultimate license has been added.

• Disagglomeration model allows to study the evolution of the size of solid particles in mixing flows.

• Adaptive meshing for large variation of field (beta feature) allows to refine the mesh according to the variation of a given field in finite elements.

• Surface tension for 2D and 3D free surfaces (beta feature) allows to take surface tension into account in free surfaces.

• Calibrator model (beta feature) allows to study the performance of extrusion calibrators for kinematic and thermal aspects.

• Improvement of the Polyflow Workbench Add-in (beta feature).

Meshing Workflows

• Cutcell meshing, including prism generation for cutcell meshes is no longer supported.

• In previous releases, Spaceclaim (SCDM) was used as the default import route for importing .scdoc and .scdocx files. In this release, .scdoc and .scdocx files are now imported using the Discovery (DSCO) import route.

• A dynamic min area setting has been added based on the mesh scale to minimize issues if the mesh is scaled when going to the solver.

Mesh Generation

• When using the Rapid Octree mesher, the ability to set a global feature angle refinement level is removed, in order to promote clarity during setup. The ability to define such mesh refinement is now only available by creating size functions with defined angular refinement settings for one or more face zones. (Custom Boundary Sizes)

Files

• The requisite library for importing Abaqus ODB files is removed for Ansys Release 2025 R2 due to security concerns. Contact Ansys support if your workflow requires importing Abaqus ODB files.

Solver-Meshing

• The merging of zones is now performed faster. For further details on zone merging, see Merging Zones.

Cell Zones and Boundary Conditions

• When modeling porous media using the physical velocity porous formulation, the inputs for the velocity values in the velocity inlet boundary conditions are now no longer erroneously divided by the porosity. (Porous Media Conditions)

Turbulence

• The formulation of the SBES shielding function in the GPU solver has been consolidated with the Fluent-CPU solver. A term used in the original SBES shielding function to detect regions away from walls, which are not sufficiently refined for LES resolution, has been deactivated to match the Fluent-CPU behavior.

  The following command can be used to reactivate the term and recover the Fluent-GPU behavior prior to the 2025 R2 release:

  (rpsetvar 'sbes-bf-include-sub-function1? #t)

Reacting Flows

• For non-premixed and partially premixed combustion cases, a more accurate method has been implemented that calculates the heat release rate from the PDF table. In previous releases, the heat release rate was calculated using the cell enthalpy.

Discrete Phase Model

• When generating contour plots of DPM particle sampling, the accuracy of the calculation of the discrete phase flux through the sampling plane has been improved. If particles cross the plane in both directions, only the net flux through the individual square element of the plane is calculated. Note the following:

  • Result values may now be negative where they were positive before.

  • If only negative values are present, you can ignore the sign and use the absolute value of each data point.

  • If both positive and negative values are present, the net mass flux has different directions at different locations.

• The erosion/accretion model has been changed for cases with the Accumulate Erosion/Accretion Rates option is enabled in the Sample Trajectories / Compute Erosion dialog box. When reading a data file generated prior to 25R2 into your Fluent session, all accumulated erosion/accretion rates are discarded. After recomputing the erosion/accretion rates and saving the data file, the rates will be stored in the file and can be loaded into future sessions.

Multiphase Models

• In previous versions of Ansys Fluent, the default value of the coefficient in the Troshko-Hassan model formulation was 0.75, which was divided by 0.75 inside the code. The default value of has changed to 1, and the division has been removed. The solution results should not change compared to previous versions.

• For the VOF and mixture multiphase models, the alternative energy treatment for mass transfer has been improved resulting in more accurate solution.

• The depth function used in the TMA spectrum for random waves has been corrected to be consistent with DNV Recommended Practice - DNV-RP-C205 - Environmental Conditions and Environmental Loads, 2019. In previous releases of Ansys Fluent, the depth function was based on the 2010 edition of the same publication. (TMA Spectrum)

Density-Based Solver

• For High Speed Numerics with the density-based solver, there are now improved relaxation and iterative parameters for transient calculations employing the implicit CFL-based solution method. These modifications increase the robustness of the inner-iterations for high-speed flows and allow improved convergence when using large implicit time-steps.

Fluent Native GPU Solver

• To ensure that you have access to the full functionality of the GPU solver on one or multiple NVIDIA GPUs, you must now install CUDA version 12.8. The supported driver version is 570 or newer, and the minimum supported hardware is the Pascal generation. (Supported GPUs and Drivers)

• For transient cases with CHT solids and the solid timestepping option enabled, the treatment for solid timestepping for the CPU-based Fluent solver has been changed in this release. However, the GPU solver still uses the previous treatment.

Parallel Processing

• The resident memory required for some cases may increase approximately 25 MB per process due to third-party updates.

• The InfiniBand interconnect is now selected by default and when auto-select is used with Open MPI for both x86 and Arm architectures. (Starting Parallel Ansys Fluent on a Linux System Using Command Line Options)

Adjoint Solver Module

• Changes to the design tool may cause journals that were created in a previous release to fail. Failure to read journals may be caused by the design tool changes listed below.

  • Journal attempting to load a case from mesh.

  • Journal attempting to load a case that was created prior to Release 23R1, which does not have the adjoint module enabled.

  • Journals may fail due to the following changes to the default morphing method, from Polynomials with the standard constraint method to Polynomials with the enhanced constraint method:

    • Journal attempts to specify polynomials-standard numerics settings.

    • Journal attempts to set freeform scaling method and factor settings.

    • Journal attempts to define increase/decrease objective type while using the polynomials-standard morphing method, such that the target change is inappropriate unless it is scaled back significantly. The polynomials-enhanced method will simply achieve the target change without scaling it back, resulting in bad defomations.

    • Journal attempting to load a case from a mesh file. The resulting deformations will not be the same between standard and enhanced constraint methods.

    If a journal read failure was caused by the change to the default morphing and constraint method, adding the following line to the journal file may resolve the issue:

    adjoint design-tool design-change select-constraint-method standard

Polyflow

• • The number of processors used by the workspace solver has been increased from 2 to 4 processors.

  • Mesh objects created by the workspace no longer use generic naming conventions (mesh-1, mesh-2, etc.) by default. The new default is to use the topological names found in the mesh file. The behavior can be changed using Preferences. See Setting Preferences for Polyflow in the Fluent Workspaces User's Guide for details.