EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

A Real‐Time Physics‐Driven Digital Twin for S‐CO 2 Centrifugal Compressors: A One‐Dimensional Modelica‐Based Framework With Open Hybrid Calibration

He Ren, Liping Chen, Xiaojun Ye

Digital Twins and Applications · 2026-01 · 3 (1)

DOI: 10.1049/dgt2.70050 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Digital Twins and Applications, online erschienen am 21.08.2026. Verfasst von He Ren und 2 weiteren Autor:innen. Die Arbeit nennt 19 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Supercritical carbon dioxide (S‐CO 2 ) Brayton cycles are promising for next‐generation clean energy systems, yet their centrifugal compressors operate near the critical point, where drastic thermophysical property variations impose fundamental challenges on real‐time digital twin (DT) deployment. High‐fidelity three‐dimensional (3D) flow simulations, although physically detailed, require hours to days per simulation and cannot satisfy the sub‐second response demands of online monitoring and predictive maintenance. Lumped‐parameter performance maps, in contrast, are essentially static interpolations that lack extrapolation fidelity for unseen transient conditions. To bridge this gap, this paper presents a physics‐driven, reusable unit model for S‐CO 2 centrifugal compressors using the equation‐based object‐oriented modelling language Modelica. The compressor is decoupled along the flow path into four modular sub‐models—inlet duct, impeller, vaneless diffuser and volute—each spatially discretised by a staggered‐grid finite volume method (FVM) to preserve near‐critical physical fidelity. A semi‐implicit fractional‐step solver that exploits the block‐tridiagonal sparse structure inherent to one‐dimensional (1D) chain topologies is employed to achieve a computational‐to‐physical time ratio below 0.6. Furthermore, an open hybrid modelling architecture with standardised physical–virtual data interfaces is proposed, enabling future online residual feedback and adaptive parameter calibration through reserved hybrid ports. Steady‐state validation against experimental data covering 19 operating points shows that, after offline calibration of correction coefficients, prediction errors for outlet total temperature, outlet total pressure, total pressure ratio and isentropic efficiency in open‐loop mode are bounded within 1.64%, 4.14%, 4.09% and 3.74%, respectively. Transient open‐loop prediction demonstrates consistent dynamic tracking of outlet total temperature and pressure‐ratio transients, confirming the intrinsic extrapolation capability of the calibrated physics kernel. The proposed framework provides a plug‐and‐play, reusable building block for system‐level S‐CO 2 Brayton cycle twins, supporting future online state estimation, anomaly detection and predictive maintenance upon activation of the reserved hybrid ports.

Zitieren

APA

Ren, H., Chen, L., & Ye, X. (2026). A Real‐Time Physics‐Driven Digital Twin for S‐CO 2 Centrifugal Compressors: A One‐Dimensional Modelica‐Based Framework With Open Hybrid Calibration. Digital Twins and Applications, 3(1). https://doi.org/10.1049/dgt2.70050

Vancouver

Ren H, Chen L, Ye X. A Real‐Time Physics‐Driven Digital Twin for S‐CO 2 Centrifugal Compressors: A One‐Dimensional Modelica‐Based Framework With Open Hybrid Calibration. Digital Twins and Applications. 2026;3(1). doi:10.1049/dgt2.70050

BibTeX
@article{Ren2026,
  author = {Ren, He and Chen, Liping and Ye, Xiaojun},
  title = {A Real‐Time Physics‐Driven Digital Twin for S‐CO 2 Centrifugal Compressors: A One‐Dimensional Modelica‐Based Framework With Open Hybrid Calibration},
  journal = {Digital Twins and Applications},
  year = {2026},
  volume = {3},
  number = {1},
  doi = {10.1049/dgt2.70050}
}
Kostenlos starten →