Proceedings of the 4th International Seminar on Non-Ideal Compressible Fluid Dynamics for Propulsion and Power

Martin T. White, Tala El Samad, Ioannis K. Karathanassis, A. I. Sayma, Matteo Pini, Alberto GuardoneERCOFTAC Series 2023 · DOI · OpenAlex

Overview

The 4th International Seminar on Non-Ideal Compressible Fluid Dynamics (NICFD) for Propulsion and Power, held in 2022, serves as a comprehensive compendium of the current state-of-the-art in fluid modeling. This volume brings together 23 peer-reviewed papers focusing on fluids that deviate from the standard "ideal gas" simplification, particularly those operating near critical points or in complex two-phase states.

The editorial board for the proceedings includes experts from the University of Sussex, City University of London, Delft University of Technology, and Politecnico di Milano.

Addressing the Non-Ideal Regime

Traditional fluid dynamics often relies on the assumption of ideal gas behavior. However, modern high-efficiency engineering—ranging from carbon-neutral power plants and refrigeration to advanced rocket engines—increasingly operates in "non-ideal" regimes. These conditions include flows at supercritical pressures, proximity to the saturation curve, or liquid-vapor mixtures.

In these regions, fluid behavior becomes counter-intuitive and mathematically complex. Standard Computational Fluid Dynamics (CFD) tools frequently struggle to accurately predict phenomena such as shock waves, turbulence, and heat transfer. This collection addresses the critical need for improved numerical models, robust experimental data, and a deeper theoretical understanding of NICFD to support the next generation of clean energy and propulsion technologies.

Methodological Innovations

The proceedings highlight a diverse range of methodologies, bridging fundamental physics with modern data science:

  • High-Fidelity Simulation: Researchers utilized Direct Numerical Simulation (DNS) and Large Eddy Simulation (LES) to investigate wall-bounded turbulence and boundary layers in organic fluids.
  • Machine Learning Integration: The use of Artificial Neural Networks (ANN) has emerged as a key tool for developing subgrid-scale turbulence models tailored for dense gas flows.
  • Numerical Solver Advancement: Significant work was presented on adapting the open-source SU2 solver and developing pressure-based models for two-phase flows under generic equations of state.
  • Experimental Validation: Data was gathered from specialized facilities, including the ASTER asymmetric shock tube, the LUTsCO2 test loop, and supersonic blade cascades for Organic Rankine Cycles (ORC).
  • Uncertainty Quantification: New frameworks were introduced for estimating model-form uncertainty in Reynolds-Averaged Navier-Stokes (RANS) closures when applied to non-ideal fluids.

Key Insights and Results

Exploiting Non-Ideal Effects

One of the most promising findings is the ability to exploit non-ideal effects, such as the reduction in the speed of sound near the critical point, to design more compact and efficient turbomachinery.

Turbulence and Two-Phase Challenges

Research indicates that high-pressure transcritical conditions significantly alter the structure of wall-bounded turbulence compared to ideal gases, necessitating specialized "real-fluid" corrections. Furthermore, modeling "flashing" flows—where liquid suddenly vaporizes—remains a challenge, with non-equilibrium phenomena proving critical for accurate nozzle design.

Validation Successes

Several papers demonstrated that adapted CFD solvers, such as SU2, can now predict isentropic non-ideal flows with high accuracy. Additionally, a new two-phase nozzle design tool based on the "two-fluid model" was successfully validated for wet-to-dry expansions of organic fluids.

Future Directions and Limitations

Despite significant progress, the field faces ongoing challenges:

  • Data Scarcity: There remains a shortage of high-quality experimental data for multi-component mixtures (e.g., $CO_2$ with impurities), which are vital for industrial carbon capture.
  • Computational Efficiency: While DNS and LES provide deep insights, they are computationally expensive. A primary goal for future work is transitioning these high-fidelity insights into faster, industrial-grade RANS models.
  • Complex Mixtures: Research is increasingly shifting toward real-world working fluids, which are often complex mixtures rather than pure substances, adding further thermodynamic layers such as shifting critical points.

Summary updated

BibTeX
@inproceedings{white2023proceedings,
  author = {Martin T. White and Tala El Samad and Ioannis K. Karathanassis and A. I. Sayma and Matteo Pini and Alberto Guardone},
  title = {{Proceedings of the 4th International Seminar on Non-Ideal Compressible Fluid Dynamics for Propulsion and Power}},
  booktitle = {ERCOFTAC Series},
  year = {2023},
  doi = {10.1007/978-3-031-30936-6},
}
Figure