Data-driven fluid model parameters

Generated from SU2 commit 6b54b78464 on 2026-10-23. Do not edit this page. Edit the comment above the option in Common/src/CConfig.cpp instead.

INTERPOLATION_METHOD

Interpolation method used to determine the thermodynamic state of the fluid.

More in config_template.cfg: Interpolation method for thermodynamic state calculation for data-driven fluid model or flamelet fluid model. (MLP for multi-layer perceptron, LUT for 2D or 3D, unstructured look-up table)

• Type: one of the values below
• Values: LUT, MLP
• Default: LUT
• In config_template.cfg: INTERPOLATION_METHOD= MLP
• Mentioned in: Changes in Version 8, Setup Guide: Non-Ideal Compressible Flows, Setup Guide: Species Transport and Combustion, Incompressible, Laminar Combustion Simulation, Non-ideal compressible flows with physics-informed neural networks
• Used in 2 test case files, for example: laminar_premixed_h2_flame_cfd.cfg (MLP)
• Code: CConfig.cpp line 1292

FILENAMES_INTERPOLATOR

Input file for the interpolation method.

More in config_template.cfg: Name of the input file containing the information required for the interpolator to function. Provide list of .mlp files (See https://github.com/EvertBunschoten/MLPCpp for more information.) when using the MLP option for INTERPOLATION_METHOD or a single .drg file for the LUT INTERPOLATION_METHOD option.

• Type: list of strings
• In config_template.cfg: FILENAMES_INTERPOLATOR= (MLP_1.mlp, MLP_2.mlp, MLP_3.mlp)
• Mentioned in: Setup Guide: Non-Ideal Compressible Flows, Setup Guide: Species Transport and Combustion, Incompressible, Laminar Combustion Simulation, Incompressible, Laminar counterflow diffusion flame, Non-ideal compressible flows with physics-informed neural networks
• Used in 10 test case files, for example: lam_prem_ch4_cfd.cfg ((fgm_ch4.drg)), lam_partial_prem_ch4_cfd.cfg ((LUT_methane_3D.drg)), laminar_premixed_h2_flame_cfd.cfg ((MLP_TD.mlp, MLP_PD.mlp, MLP_PPV.mlp))
• Code: CConfig.cpp line 1294

DATADRIVEN_NEWTON_RELAXATION

Relaxation factor for Newton solvers in data-driven fluid model.

More in config_template.cfg: Relaxation factor for the Newton solvers in the data-driven fluid model

• Type: real number
• Default: 1.0
• In config_template.cfg: DATADRIVEN_NEWTON_RELAXATION= 0.8
• Mentioned in: Setup Guide: Non-Ideal Compressible Flows
• Code: CConfig.cpp line 1296

USE_PINN

\biref USE_PINN Use physics-informed approach for the entropy-based fluid model.

More in config_template.cfg: Use physics-informed neural network for calculation of the fluid caloric properties.

• Type: YES / NO
• Default: NO
• In config_template.cfg: USE_PINN= YES
• Mentioned in: Changes in Version 8, Setup Guide: Non-Ideal Compressible Flows, Non-ideal compressible flows with physics-informed neural networks
• Used in 1 test case file, for example: datadriven_nozzle.cfg (YES)
• Code: CConfig.cpp line 1298

CONFINEMENT_PARAM

Input Confinement Parameter for Vorticity Confinement

More in config_template.cfg: Set confinement parameter (0.00 by default)

• Type: real number
• Default: 0.0
• In config_template.cfg: CONFINEMENT_PARAM = 0.00
• Used in 1 test case file, for example: turb_ONERAM6_vc.cfg (0.05)
• Code: CConfig.cpp line 1301