YJYu Jiao

Yu Jiao

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yu Jiao has a total of 31 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    NeuralFVM: Neural-physics-based finite volume method for turbulent flows using the k-ω model

    by Tingkai Xue, Yu Jiao, Te Ba, Jingliang Wang, Juntao Yang, Simon See, Boyang Chen, Claire E. Heaney, Christopher C. Pain, Chang Wei Kang, et al.


  • NP4Buoyancy: An end-to-end differentiable neural-physics framework for buoyancy-driven flows

    by Yu Jiao, Han Zhang, Tingkai Xue, Boyang Chen, Juntao Yang, Te Ba, Simon See, Chang Wei Kang, Claire E. Heaney, Christopher C. Pain, et al.


  • Open Access

    Ultrasonic-Net: a physics-informed neural network for ultrasonic field prediction in cleaning systems

    by Yu Jiao, Alexander Bußmann, Steffen J. Schmidt, Marcus Giglmaier, Nikolaus A. Adams

  • Open Access

    NP4Buoyancy: An End-to-End Differentiable Neural-Physics Framework for Buoyancy-Driven Flows

    by Yu Jiao, Han Zhang, Tingkai Xue, Boyang Chen, Yang Juntao, Te Ba, Simon See, Kang Chang Wei, Claire E. Heaney, Christopher Pain, et al.

    We present an end-to-end differentiable neural-physics framework for buoyancy-driven flows, NP4BUOYANCY, which bridges high-fidelity computational fluid dynamics (CFD) and modern machine-learning (ML) ecosystems. Unlike data-driven surrogate models, NP4BUOYANCY is a solver-level implementation of the incompressible Navier–Stokes equations under the Boussinesq approximation, in which spatial discretization, multigrid solvers, diagnostics, and automatic differentiation are unified within a single computational graph. The framework is implemented entirely in PyTorch, enabling performance portability across heterogeneous Central Processing Unit/Graphics Processing Unit (CPU/GPU) architectures and seamless integration with ML components. At its core, NP4BUOYANCY reformulates classical geometric multigrid (GMG) methods as fixed-weight convolutional operators. All convolution kernels, including those for smoothing and inter-grid transfer, are derived analytically rather than learned from data, preserving physical interpretability while leveraging optimized GPU tensor primitives and enabling a differentiable pipeline. The framework is systematically validated on buoyancy-driven two-dimensional and threedimensional (2D/3D) Rayleigh–Bénard convection. Across a wide range of Rayleigh numbers (103 ⩽ Ra ⩽ 1011), the solver shows close agreement with benchmark velocity statistics, local and global Nusselt numbers, and heat-transfer scaling laws, while achieving substantial GPU acceleration relative to CPU execution. Beyond forward simulation, NP4BUOYANCY enables gradient-based inverse modelling through native automatic differentiation. By combining multi-fidelity optimization with gradient checkpointing, the framework supports end-to-end trajectory optimization on high-resolution grids within practical memory constraints. Numerical experiments demonstrate successful reconstruction of initial thermal plume structures from sparse observations, without offline training data or manual adjoint derivation. Overall, NP4BUOYANCY provides a validated, scalable, and fully differentiable platform for buoyancy-driven convection, enabling adjoint-free inverse analysis and physics-consistent, artificial intelligence (AI)-augmented CFD.

  • Open Access

    Effects of cylinder-wall distance on re-shock interaction with fuel cylinders at near-critical thermodynamic states

    by Yu Jiao, Steffen J. Schmidt, Nikolaus A. Adams

    In high-speed propulsion systems, for instance, scramjets and ramjets, shock waves impacting the fuel-environment gas interfaces at near-critical thermodynamic states is a phenomenon that is frequently encountered. The effects of shock waves on fuel cylinders with the inclusion of real-fluid features remains a relatively underexplored area of research. In this study, we investigate the influence of re-shock on the shocked fuel cylinder at near-critical thermodynamic states. The spacing between the fuel cylinder and the wall is varied from 1.5R to 9R. The simulation employs compressible multi-component equations and real-fluid thermodynamic relationships to model the evolution of the cylinder and the dynamics of the surrounding gas flow. A hybrid numerical scheme based on finite volume is used to capture shocks and interfaces. The approach has been validated against reference data and shows excellent agreement. The effects of varying cylinder-wall distances on the evolution of the surface of the re-shocked fuel cylinder are thoroughly analyzed. The analysis presented covers intricate shock and re-shock impingement, cylinder deformation, cylinder displacements, and vortex development. A range of features at different stages of evolution are analyzed quantitatively and qualitatively. These include wave pattern evolution, pressure redistribution, shift properties (center-of-mass parameters), cylinder geometry deformation characteristics (cylinder width), baroclinic vorticity distribution, circulation histories, and enstrophy progression. The analysis also quantifies changes over time in the characteristic mixing degree between the fuel and ambient gas. Two additional parameters, the area and the mass fraction-weighted area of the mixing region, are included. Current studies could provide valuable insights for experimental design and industrial applications involving gas-fuel mixing processes at near-critical thermodynamic states.


  • Open Access

    Effect of gas cavity size and eccentricity on shock interaction with a cylinder at near-critical conditions

    by Yu Jiao, Steffen J. Schmidt, Nikolaus A. Adams

    In this study, we investigate the impact of gas cavity size and eccentricity on the interaction of shockwaves with a cavity-embedded fuel-liquid cylinder under near-critical conditions. We analyze a range of scenarios involving both eccentric and concentric cavities, varying cavity radii (0-0.875R), eccentricity angles (0°–180°), and distances (0R-0.45R). Our methodology entails modeling the evolution of the fuel cylinder and surrounding gas flow using compressible multi-component equations, employing a finite-volume-based hybrid numerical framework capable of accurately capturing shocks and interfaces. Additionally, real-fluid thermodynamic relationships are employed, validated against reference data, showing excellent agreement. Mesh independence studies are provided. We analyze the shock impingement characteristics, deformation of the cylinder and cavity, and the formation of vortices. Various phenomena at different evolution stages are explored, including wave pattern evolution, jet formation, cavity breakup, baroclinic vorticity distribution, and circulation histories. Size and eccentricity of the cavity determine time intervals between wave contact with the cylinder and with the cavity, thereby influencing the evolution of wave patterns and interface deformation. We propose an analytical model for deposited circulation, obtained by appropriately combining the Yang, Kubota, and Zukoski (YKZ) and the Zhang and Zou (ZZ) models, which agrees well with numerical findings for cases involving smaller cavities. However, for larger cavities, as the cavity gradually reaches the cylinder surface, induced coupling effects invalidate the model. Furthermore, we introduce four predictive fits for the center-of-mass position of the shocked cylinder under near-critical conditions. These fits—the Time-Size Polynomial Prediction Fit, the Time-Eccentricity Polynomial Prediction Fit, the Time-Eccentricity Distance Polynomial Prediction Fit, and the Connecting Rod Prediction Fit—are tailored for cases involving cavities of varying sizes, eccentricity angles, and distances. Demonstrating good predictive performance, these fits offer valuable insights into the mixing behavior of liquid fuel sprays in a diverse range of near-critical environments and high-speed propulsion systems.


  • Simulating shock interaction with a cavity-embedded cylinder/droplet using a real-fluid hybrid scheme at near-critical conditions

    by Yu Jiao, Steffen J. Schmidt, Nikolaus A. Adams

    Examining the influence of shock waves on cylinders and droplets at near-critical conditions, especially when accounting for real fluid effects, represents a relatively unexplored frontier. This research gap becomes even more relevant when extending the investigation to three-dimensional scenarios. The underlying evolution mechanisms at these conditions remain elusive, with limited existing literature. In this study, we present a thorough exploration employing two-dimensional and three-dimensional numerical simulations of a droplet with an embedded gas cavity subjected to a normal shock wave at near-critical conditions. Our approach involves modeling the cylinder/droplet and the surrounding gas flow using the compressible multicomponent equations, incorporating real fluid thermodynamic relationships, and implementing a finite-volume-based hybrid numerical framework capable of capturing shocks and interfaces. To establish the reliability of our approach, we validate it against reference data, demonstrating excellent agreement. We also conduct mesh independence studies, both qualitatively and quantitatively. Our analysis is comprehensive, considering the intricacies of shock impingement, the morphological deformation of the cylinder/droplet and cavity, and the development of vortices. We discuss and analyze various phenomena, including the evolution of wave patterns, jet formation, sheet formation, hole appearance, the emergence of petal-shaped structures or lobes, ligament formation, shear-induced entrainment, and internal cavity (bubble) breakup. We compare the results obtained from the cylinder/droplet with a cavity to those from a planar shock wave impacting a pure cylinder/droplet. We provide a holistic view of the two-dimensional cylinder and three-dimensional droplet's evolution before and after the impact of a shock wave, accompanied by quantitative data regarding the positions of characteristic points along the column over time. Our analysis further scrutinizes the geometrical characteristics of the cylinder and the trends in the distribution of baroclinic vorticity at various stages. Our findings reveal that the presence of a gas cavity plays a pivotal role in shaping the shock wave, which, in turn, influences the generation and distribution of baroclinic vorticity. This leads to a transformation in the unstable evolution process of both the cylinder and the droplet. Importantly, shock waves impacting the evolving interfaces of the cylinder/droplet and the internal gas cavity generate baroclinic vorticity, which subsequently affects the transport and distribution of vorticity, thereby influencing the evolution of the cylinder/droplet interface. In the case of three-dimensional droplets, baroclinic vorticity induces complex, intricate three-dimensional structure transformations.


  • Open Access

    An efficient and robust all-Mach consistent numerical scheme for simulation of compressible multi-component fluids including surface tension, cavitation, turbulence modeling and interface sharpening on compact stencils

    by Yu Jiao, Steffen J. Schmidt, Nikolaus A. Adams

    We present an efficient, fully conservative numerical scheme valid in the entire range of highly to weakly compressible flows using a single-fluid four equation approach together with multi-component thermodynamic models. The approach can easily be included into existing finite volume methods on compact stencils and enables handling of compressibility of all involved phases including surface tension, cavitation and viscous effects. The mass fraction (indicator function) is sharpened in the two-phase interface region using the algebraic interface sharpening technique Tangent of Hyperbola for INterface Capturing (THINC). The cell face reconstruction procedure for mass fractions switches between an upwind-biased and a THINC-based scheme, along with proper slope limiters and a suitable compression coefficient, respectively. For additional sub-grid turbulence modeling, a fourth order central scheme is included into the switching process, along with a modified discontinuity sensor. The proposed All-Mach Riemann solver consistently merges the thermodynamic relationship of the components into the reconstructed thermodynamic variables (like density, internal energy), wherefore we call them All Mach THINC-based Thermodynamic-Dependent Update (All-Mach THINC-TDU) method. Both, liquid-gas and liquid-vapor interfaces can be sharpened. Surface tension effects are taken into account by using a Continuum Surface Force (CSF) model. An explicit, four stage low storage Runge Kutta method is used for time integration. The proposed methodology is validated against a series of reference cases, such as bubble oscillation,advection,deformation, shock-bubble interaction, a vapor bubble collapse and a multi-component shear flow. Finally, the approach is applied to simulate the three-dimensional primary break-up of a turbulent diesel jet.