2025 R1 What's New - Ansys Fluent
2025 R1 What's New - Ansys Fluent
2025 R1
UI/UX Updates
• Desktop Interface
• Fluent Web Interface
CPU Solver Updates
• Ansys Cloud Burst Compute
• New and Enhanced Physics
• Meshing Updates
14%
22%
MIUPS
60
40.6
40 CPU
24.19
14.31 A100
20 7.02
0
128 256 1 GPU 2 GPUs 4 GPUs 8 GPUs
CPUs 4 CPUs 8
nodes nodes
6 ©2025 ANSYS, Inc.
Non-adiabatic Flamelet Generated Manifold (FGM) (β)
PRECCINSTA burner
• 3M cells
• One additional transport equation • Fuel, CH4, 1 atm
- Total (sensible + combustion) enthalpy
• Supported Wall Boundary Conditions
- Heat flux and adiabatic
- Specified temperature Static Temperature
- Coupled walls
• Support for DO radiation and DPM
coupling
• CHT is not yet supported Mean Progress Variable Progress Variable variance
Target applications:
o Gas turbines and combustion 10% higher computational cost compared to Adiabatic FGM
7 ©2025 ANSYS, Inc.
Acoustics with Ffowcs Williams & Hawkings
Target applications:
o Turbomachinery blades
o Airfoils
o External Aero
9 ©2025 ANSYS, Inc.
Turbulence: Curvature Correction
Hydro-cyclone
• Available for all omega-based
turbulence models, including
transition
- Improves turbulence calculation
for highly curved or swirling flows
Target applications:
o Cyclones
o Pipe flows with sharp bends or turns
• Wall treatment for temperature made consistent between GPU and CPU solvers for LES
- Can cause changes for GPU results when the first cell center is in the buffer layer
- Transition between sublayer and log regions are smoother with the new treatment
Target applications:
o Gas turbines
o External aerodynamics
o Can be applied to all LES calculations
- Clustering enhancements
• Split-angle option in TUI to avoid defeaturing
/solve/models/radiation/s2s-parameters/split-angle
Target applications:
o Electronics Cooling
o Underhood Thermal Management
Target applications:
o Water hammer, hydro-acoustics Area weighted average pressure at the outlet
o Solution stabilization for large pressure peaks
19 ©2025 ANSYS, Inc.
Volume of Fluid Method (β)
22.28
14.01
9.68
5.61
128 cores 256 cores 512 cores 1 GPU 2 GPUs 4 GPUs 8 GPUs
• Surface tension for VOF must be considered when Capillary or Weber numbers
become small
• Continuum surface force model with constant surface tension coefficient
• Optional boundary-zone-based wall adhesion
• Modified body-force-weighted
pressure discretization
- Most cases show improved stability for VOF
with the modified body-force-weighted Mixing tank
formulation compared to other methods Pseudo-transient with MBFW
• Particle Monitors
- Show Mass Flow/Change Rate option available for
DPM reports if you want to create a report for the
mass flow rate rather than the total mass
• Breakup Model
- SSD breakup model constants now accessible and
injection specific
• Reduce the case/data loading time when restart the GPU solver calculation and
enabling native GPU post-processing
- Command line option: -lite
- Unsupported objects like iso-surfaces are disabled
• Work for both first-time and restart runs
• Once saved with lite, cas/dat can only be read in this mode
• Usability enhancements
- Automatically reload Python UDF (in the next iteration block) when .py file is
modified
- Automatically wrap Real() around floating-point values
- Scheme commands for listing, loading, and unloading Python UDFs
• (gpuapp-add-udf "vel_profile.py" "t_profile.py")
• (gpuapp-remove-udf "t_profile.py")
• (gpuapp-list-udf)
- More macros available, e.g., area vector, cell volume, time step size, …
• For details see Fluent Beta Manual
• Access Fluent rpvar from Python UDF, and automatically reload UDF when its value is
modified
ref_density = UDF.rpgetvar('reference-density')
• Specify gravity vector, turbulence viscosity, and curvature correction coefficient in UDF
• Access field value in neighboring c0/c1 cells of a face
• General-purpose define_execute_at_end and define_adjust
V-velocity
def udf_fix_velocity(c_centroid, c_u, c_v, c_w):
x, y, z = c_centroid
if math.fabs(x) < 0.005 and math.fabs(y) < 0.005: Modify u/v/w to
return 0., 0., 0.
else: approx. a solid
return c_u, c_v, c_w
UDF.add_cell_zone_adjust(udf_fix_velocity, UDF.all_cell_zones(), (c_u))
• The gRPC server can be started from command line or TUI: /server/start-grpc-server
- There is no filter on the PyFluent side to hide unsupported or incompatible features
for the GPU solver. Test your scripts carefully and verify that all defined settings work
as intended.
14%
solver
3501
1859
1478
1041
557
8 16 24 32 64 96 128 256
MIUPS 557 1041 1478 1859 3501 4688 6092 7892
• Solid time-stepping
- Independent timestep for CHT solids
• Reduces calculation time when only final state is of interest
- Also appropriate for LES+CHT
Steady
𝑑𝑡𝑓 = 𝑑𝑡𝑠 = 10 𝑑𝑡𝑓 = 10, 𝑑𝑡𝑠 = 100
Fluid
Solid
Benefits Benefits
• Applications with thousand of zones (such as batteries, fuel cells, multiphase • More workflows possible with full remote sessions from mesh generation (β) to results
flows, radiation and species models and more) now have much faster evaluation.
responsiveness for shorter meshing and solve time. • Enhanced graphics improve productivity/ease within the interface.
• Transient post-processing can now be performed directly in the Fluent solver
for an improved, single interface workflow.
• Multi-select graphics objects in the tree and use context menu to open all of them in
new windows
• OpenGL2 Mesa
- Software driver that supports high-quality output
• Slower compared with OpenGL2 and DX11 because of missing
hardware acceleration
• Comparable in quality and feature set
• Human-readable settings can be exported from the context menu of the Setup node
- Useful to compare settings between cases with third-party tools
- In the file selection dialog, specify the name with .html to change the format
Setup
----- Text export
General
-------
Solver
------
HTML export
Type : pressure-based
Time : steady
• New fast zone merging on case read (β) 100 cell zones
100 interior face zones
- Automatically used when beta options are enabled
- Merging zones in single step
• Improves solver performance and Zone merging time 24.2 25.1 Speedup
8 cores
reduces memory
Cell zones 76 [s] 5.6 [s] 14x
Interior face zones 513 [s] 16 [s] 32x
Battery pack
1070 cell zones, 29M cells
8562 face zones
• Residuals of adjoint calculations can now be monitored with the web interface
- Access to adjoint settings remains beta in 2025 R1
Animate particle
tracks and pathlines
Ansys Cloud Burst Compute New & Enhanced Physics Meshing Updates
Newly available Cloud service for Fluent Users: Submit jobs • Electrochemistry Workflow Improvements: Dramatically • Water-Tight Meshing (WTM) / Fault Tolerant Meshing
from a local Fluent case to the Ansys Burst to Cloud service faster set up time for models with many zones, solver (FTM): Significant Task Performance speed ups (up to 30X)
for one-off Cloud resource usage. robustness for LI-battery models significantly reduces number for cases with many parts/zones
of time steps required, capture capacity fade effects in the
battery aging model • Skin Assemblies (Modular Wrapping): faster meshing with
• Aerospace Advancements: Fluent Aero mesh adaption (β) FTM for very large assemblies by wrapping sub-assemblies
allows you to update Mesh Adaption or Parametric Search before meshing the full model
Cycles for all design points
• Optimization Enhancements: Extended constraints for multi-
objective shape optimization, back-to-CAD workflow ()
updates the original CAD model using the optimized shape.
Benefits Benefits Benefits
Enhancements to Fluent CPU solver physics capabilities Meshing speed-ups improve the efficiency and length of pre-
Users who have limited in-house access to Cloud or on-premise
resources can now use the Ansys Burst to Cloud service for
drives more robust, quick and accurate CFD workflows processing workflows.
individual jobs requiring Cloud resources. across multiple applications.
• After submission, the input files are transferred to the server and the job is queued
Target applications:
o Selective catalytic reduction (SCR)
85 ©2025 ANSYS, Inc.
Lagrangian Wall Film (LWF)
Target applications:
o Selective catalytic reduction (SCR)
with deposit chemistry
Old New
Method Method
Case reading 30 min 40s
Data reading 10 min. 7s
Solution initialization 1 min 10s
6s5p module with bodies for interconnectors 6s5p module that needs virtual connection file
90 ©2025 ANSYS, Inc.
Thermal Abuse Model
0.16A/m2
cathode
anode
capillary pressure
UDS Anisotropic
ElectricCurrent
FilmHeatTransferRate
HeatTransferRate
MassFlow
• Improved UI
• Automation of initialization and pre-
processing
- Create floating disks automatically before solving
iteration
- Restart VBM calculations without extra setup steps
- Run VBM with changing meshes (Fluent Icing,
Mesh adaption, change disk position, etc.)
• Averaging of Aerodynamic Coefficients using Power Spectral Density (PSD) & Moving
Averages (β)
- Supported in Standard Fluent Aero, Parametric Search & AET Workflows
Averaging of Aerodynamic Coefficients
PSD + Moving Average PSD + Moving Average
Simple Average
• Equality Constraints
- Example wing:
• Decrease drag &…
• Maintain/Increase lift &…
• Maintain wing sectional area
• Inequality Constraints
- Example fuel venturi mixer:
• Minimize outlet fuel non-uniformity &…
• Maintain/reduce pressure-drop &…
• Maintain ratio of air-to-fuel mass flow rate
• If the first scheme of the blended strategy diverges, the second scheme might not be
able to bring the residuals down again
- A modified version of Residual Minimization is available as first scheme to avoid this
initial divergence
• Workflow to update the CAD in Ansys Discovery after morphing the mesh in Fluent
- Export the initial and deformed mesh as STL
- Load the original geometry and both STLs in Discovery
- Use a script to update the initial CAD with the Warp function:
• Warp.WarpBodyByMeshNew(GetRootPart().Bodies[0], GetRootPart().Meshes[0], GetRootPart().Meshes[3], False, 0.0001)
• Bodies[0]: CAD face; Meshes[0]: Undeformed STL; Meshes[3]: Deformed STL; 0.0001: Tolerance; all must be adjusted to the case!
• Limitations
- Remeshing is not supported, the topology of the mesh must stay the same
• The STL export of the Design Tool is improved to keep the ordering of the surface mesh nodes intact
across multiple partitions and large deformations
- It can be necessary to split the original geometry into faces based on Named
Selections to deform them one-by-one
107 ©2025 ANSYS, Inc.
Back-to-CAD workflow (β)
• Warp.WarpBodyByMeshNew( Body 0
GetRootPart().Bodies[0], Body 1
GetRootPart().Meshes[0], Body 2
GetRootPart().Meshes[3],
Mesh 0
False,
Mesh 1
0.0001)
Mesh 2
• Check body name with: Mesh 3
- print(GetRootPart().Bodies[0].GetName()) Mesh 4
Mesh 5
• Experiment with the tolerance by orders of magnitude to warp the CAD face
successfully
• Blade Row
- A reference radial position can now be set for computing the radial equilibrium at
pressure outlets
• Previously the hub was taken as the reference radial position
• Aeromechanics:
- Fast Mesh motion for periodic displacement
• Solves mesh motion once at the start of the simulation, allowing for faster flutter analysis
• 2D polyhedra support
- Hanging node conversion to polyhedral by default when reading older case file
• Can be disabled using /file/convert-hanging-nodes-during-read?
• 2D PUMA
- Default adaption method
• Can be changed to legacy hanging node adaption using /mesh/adapt/set/method
• Refinement levels from Fluent Meshing imported during I/O and mode transfer when
reading Rapid Octree meshes
- Allows for more intuitive PUMA adaption and better mesh quality in Fluent
PBNS
DBNS
Speed-up Factor 24R2 -> 25R1 Speed-up Factor 24R2 -> 25R1
400 14
350 12
300 10
250 8
200 6
150 4
100 2
50 0
Pattern with Pattern with Pattern with Full workflow
0 1000 regions 2500 regions 10000 regions for furnace,
Import Geometry Describe Geometry Table update of 20k 4800 regions,
execution zones 19k zones
• Generate Multizone mesh options for situations when the full model is meshed with
MultiZone
- Control if re-merge of separated faces
- Control if merge of body label bodies
• Performance improvements
• The list of available zones/labels is filtered automatically to show only items without
boundary layer definition
- Release for WTM, beta for FTM
- Simplifies setup for complex models
Filtered automatically
for next BL definition
30
25
20
15
10
0
File load Taks switch Tree select Drag-drop Copy-paste Create Transform Undo Transform Full workflow
• Can be deactivated:
New smooth mesh coarsening option available
/mesh/rapid-octree/advanced-
meshing-options/auto-align-
surface-normals? no
• Default Menu
TUI GUI
- Input objects
• Old: /mesh/rapid-octree/geometry
• New: /mesh/rapid-octree/input-object
- Default size
• Old: /mesh/rapid-octree/mesh-sizing/boundary-cell-size
• New: /mesh/rapid-octree/mesh-sizing/default-boundary-cell-size