Question about underlying methodology of ROM Builder in Ansys Fluent / Workbench

[stefvandriessche04]

Hello everyone,

I am currently working with Ansys Fluent together with the ROM Builder in Ansys Workbench to construct a Reduced Order Model (ROM) for CFD simulations. While I understand the practical workflow (snapshot generation → training → ROM evaluation), it is not fully clear to me which reduced-order modelling methodology is used internally by Ansys ROM Builder.

More specifically, I am trying to understand the theoretical framework behind the implementation:

• Is the ROM approach intrusive or non-intrusive?
• Does it rely on POD-Galerkin projection, response surfaces, interpolation-based surrogate modelling, or another technique?

If anyone can clarify the modelling approach used internally (or point me to relevant documentation), that would be very helpful.

Thanks in advance for your help! 🙂