High-Fidelity, Efficient Python Solver for Lid-Driven Cavity Flow: Contour Validation Against ANSYS Fluent

A short and trustworthy comparison between an in-house Python-based solver and the commercial CFD code ANSYS Fluent is provided for the benchmark case of the ۲D lid-driven cavity at Reynolds number ۱۰۰. The Python solver adopts the vorticity–streamfunction formalism and vectorized linear-algebra routines to solve steady-state velocity distributions on uniform grids of ۶۴ ۶۴ and ۱۲۸ ۱۲۸. Fluent computations are given as the reference solution. Velocity distributions from both methods are interpolated on the same fine grid for an overall visual comparison side by side, whereas localized error maps exhibit the absolute difference. The velocity centerline distributions (u and v) are compared quantitatively by means of root-mean-square (RMSE) and maximum absolute difference indicators. The results exhibit nearly perfect agreement of the principal vortex structure and boundary layer evolution, the majority of the error contained within the strong-gradient regions. The RMSE is contained below ۰.۱% of the lid velocity on both meshes, whereas the maximum deviations are contained within admissible limits despite the mesh refinement. The Python code is also significantly more efficient from the computational point of view, since the code runs around ۱۰× times faster on the ۶۴ ۶۴ mesh compared to Fluent, while around ۸× times faster on the ۱۲۸ ۱۲۸ mesh. The overall result is that an accurately implemented homemade Python code is capable of delivering simultaneously high fidelity and efficiency for standard laminar flow benchmarks.

کلیدواژه ها:

Lid-Driven Cavity ، Python CFD Solver ، Vorticity–Streamfunction ، Computational Efficiency