EUVIMEDObjevte EUVIMED
Crossref

A symbolic-regression-based turbulence modeling method with structural consistency constraints

Xu Liu, Xiaoguang Wang, Zhilong Chen

Physics of Fluids · 2026-07-01 · 38 (7)

DOI: 10.1063/5.0343339 ↗

At a glance

Journal article in Physics of Fluids, published online on 2026-07-14. Written by Xu Liu and 2 co-authors. 32 references. Funded by: National Natural Science Foundation of China, Sichuan Science and Technology Program.

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

A novel explicit turbulence modeling framework is developed for high-Reynolds-number airfoil stall flows dominated by adverse pressure gradients (APG). The proposed framework integrates weighted proper orthogonal decomposition (wPOD) with a cross-condition structural consistency constraint. High-fidelity turbulence model correction fields are first obtained under multiple flow conditions through a field inversion procedure. An APG-wPOD is then applied to extract dominant low-dimensional structures shared across different conditions, effectively suppressing data noise in the inversion results. Within the resulting reduced modal space, an adaptive sampling strategy driven by the energy distribution of the dominant modes is employed to identify representative samples associated with stall onset and flow separation dynamics. Based on these samples, an explicit analytic correction model is constructed using symbolic regression. In addition, a cross-condition structural consistency constraint is imposed to ensure that the derived model preserves a stable and physically meaningful functional form across different flow conditions. Numerical results obtained for multiple operating conditions and different airfoil geometries demonstrate that the proposed method significantly improves prediction accuracy for high-Reynolds-number airfoil stall flows, while maintaining the predictive capability of the baseline turbulence model for attached-flow regimes. The framework also exhibits favorable numerical stability and strong cross-condition generalization performance.

Funding

  • National Natural Science Foundation of China
  • Sichuan Science and Technology Program

Cite

APA

Liu, X., Wang, X., & Chen, Z. (2026). A symbolic-regression-based turbulence modeling method with structural consistency constraints. Physics of Fluids, 38(7). https://doi.org/10.1063/5.0343339

Vancouver

Liu X, Wang X, Chen Z. A symbolic-regression-based turbulence modeling method with structural consistency constraints. Physics of Fluids. 2026;38(7). doi:10.1063/5.0343339

BibTeX
@article{Liu2026,
  author = {Liu, Xu and Wang, Xiaoguang and Chen, Zhilong},
  title = {A symbolic-regression-based turbulence modeling method with structural consistency constraints},
  journal = {Physics of Fluids},
  year = {2026},
  volume = {38},
  number = {7},
  doi = {10.1063/5.0343339}
}
Start for free →