EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Transient Multiphase Flow Nested-Closure Extended Physics-Informed Neural Network (TM-NC-XPINN): Boundary-Supervised Reconstruction of Gas–Liquid Pipe Flow

Jun Chen, Zhiyu Xu, Jueqi Li, Yukun Mao, Junbo Zhang, Jingwen Song, Daqiang Feng, Jianing Li, Haoyu Wang, Jianjun Zhu

Processes · 15.11.2026 · 14 (22) · 3673

DOI: 10.3390/pr14223673 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Processes, online erschienen am 15.11.2026. Verfasst von Jun Chen und 9 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Frontier Interdisciplinary Exploration Research Program of China University of Petroleum, Beijing, Oil & Gas Major Project.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We investigate a physics-constrained neural network for reconstructing transient gas–liquid pipe flow under prescribed initial and boundary conditions. The method combines phase mass and mixture momentum equations with a horizontal drift-flux closure, bounded trainable corrections, and differentiable interpolation of SRK property tables. It uses temporal domain decomposition, a physics-loss schedule based on the training step, and a liquid-phase integral mass constraint. Training uses only the initial spatial profile, inlet gas and liquid mass-flow histories, and outlet pressure. Interior reference fields at positive times are reserved for post-training evaluation and are excluded from the loss and checkpoint selection. The baseline slug-flow case has overall methane/water feed mole fractions of 0.04/0.96. Its relative RMSEs are 0.64% for pressure, 3.52% for liquid velocity, 0.83% for gas void fraction, and 1.41% for gas velocity. The gas void fraction MAE is 0.0026. Gas compressibility and interphase slip influence the spatiotemporal evolution of the gas-phase variables. Additional slug, bubbly, and stratified-related cases are considered to assess reconstruction across flow states. All six cases are trained independently for 20,000 optimization steps.

Förderung

  • National Natural Science Foundation of China
  • Frontier Interdisciplinary Exploration Research Program of China University of Petroleum, Beijing
  • Oil & Gas Major Project

Zitieren

APA

Chen, J., Xu, Z., Li, J., Mao, Y., Zhang, J., Song, J., Feng, D., Li, J., Wang, H., & Zhu, J. (2026). Transient Multiphase Flow Nested-Closure Extended Physics-Informed Neural Network (TM-NC-XPINN): Boundary-Supervised Reconstruction of Gas–Liquid Pipe Flow. Processes, 14(22), 3673. https://doi.org/10.3390/pr14223673

Vancouver

Chen J, Xu Z, Li J, Mao Y, Zhang J, Song J, et al. Transient Multiphase Flow Nested-Closure Extended Physics-Informed Neural Network (TM-NC-XPINN): Boundary-Supervised Reconstruction of Gas–Liquid Pipe Flow. Processes. 2026;14(22):3673. doi:10.3390/pr14223673

BibTeX
@article{Chen2026,
  author = {Chen, Jun and Xu, Zhiyu and Li, Jueqi and Mao, Yukun and Zhang, Junbo and Song, Jingwen and Feng, Daqiang and Li, Jianing and Wang, Haoyu and Zhu, Jianjun},
  title = {Transient Multiphase Flow Nested-Closure Extended Physics-Informed Neural Network (TM-NC-XPINN): Boundary-Supervised Reconstruction of Gas–Liquid Pipe Flow},
  journal = {Processes},
  year = {2026},
  volume = {14},
  number = {22},
  pages = {3673},
  doi = {10.3390/pr14223673}
}
Kostenlos starten →