Guilai Han

This paper numerically investigates the effects of thermochemical nonequilibrium on the shock interactions and aerodynamic heating over the V-shaped leading edge. The numerical methodology is validated against experimental data from the JF22 hypervelocity wind tunnel. A comparative study is performed using the calorically perfect gas (CPG) model, as well as the one-temperature and two-temperature five-species chemical reaction models of air. The results demonstrate that thermochemical nonequilibrium significantly modifies the post-shock thermodynamic state through internal energy excitation and chemical reactions, thereby altering the shock-layer structure and detached-shock standoff distance. The different relaxation processes predicted by the one-temperature and two-temperature models lead to distinct shock interaction patterns, with the one-temperature model inducing a transition from Mach reflection to regular reflection from the same family. Compared with the CPG model, thermochemical nonequilibrium models predict enhanced wall heat flux due to the coupled effects of modified shock-layer states, conductive heat transfer, and species enthalpy transport. In particular, under the fully catalytic wall condition, atomic recombination releases chemical formation enthalpy, contributing to the species enthalpy diffusion term and the enhanced wall heat flux predicted by the thermochemical nonequilibrium models. These findings demonstrate that accurate thermochemical nonequilibrium modelling is essential for the reliable design and performance assessment of high-enthalpy scramjet inlets.

Accurate aerodynamic prediction under limited high-fidelity data remains a central challenge in multi-fidelity modeling. This work proposes a geometry-aware fusion framework, termed ConvexHull-GP, for aerodynamic surrogate construction from heterogeneous data sources. The method combines low- and high-fidelity Gaussian process models with a convex-hull distance criterion that explicitly distinguishes interpolation and extrapolation regions in the high-fidelity sample domain: inside the hull, predictions are provided by the high-fidelity GP, whereas outside the hull, a boundary-anchored discrepancy correction is applied to the low-fidelity prediction to improve extrapolation robustness while preserving global trends. The framework is evaluated on the Laboratory Technology Vehicle benchmark over Mach number and angle-of-attack conditions, and is systematically compared with representative multi-fidelity baselines including MF-IDW, GPy-CoKriging, and GPy-InvDistMean for both normal-force and pitching-moment prediction under multiple high-fidelity sampling scenarios. To further assess transferability, additional studies on non-LTV aerodynamic configurations, including typical wing--body and airfoil-type datasets reported in the literature, are incorporated to examine performance across different geometries, flow regimes, and data-source combinations. Results show that ConvexHull-GP provides consistently stronger overall accuracy and extrapolation stability, particularly in data-scarce settings, indicating that geometry-informed regional fusion is a reliable and practical strategy for multi-fidelity aerodynamic modeling with favorable generalization and accuracy-efficiency balance.

Many engineering problems are high dimensional, multiscale, and strongly nonlinear, which makes purely theory-driven modeling difficult. Although deep learning can be accurate, it typically requires large datasets and often yields grey-/black-box models with limited physical interpretability. This work proposes functional learning (FL), a physics-guided genetic-programming (GP) framework for aerodynamic prediction that embeds physical constraints throughout the GP pipeline, including governing relations and similarity parameters, dimensional consistency, continuity and differentiability, asymptotic limits, and intrinsic symmetries. FL performs dimensionless normalization and symmetry-based data augmentation, seeds the initial population with established theoretical and empirical formulas, and maintains an external gene pool of reusable function blocks decomposed from physical relations and correlations. During evolution, physically meaningful dimensionless combinations are used for feature engineering and reproduction, and candidates with non-physical singularities or asymptotic violations are filtered during evaluation. Numerical experiments and practical applications show that FL accelerates convergence and yields concise, physically plausible analytical expressions with strong generalization under limited training data.

Hypersonic aircraft face severe lightning strike threats due to their poorly conductive thermal protection layers and inadequate lightning protection. These threats can lead to structural failure and mission loss. To investigate gasdynamic responses and potential mechanical damage mechanisms during lightning events, this study establishes a numerical framework for lightning channel simulation based on the Navier–Stokes equations. The model comprehensively integrates Joule heating, viscous radiation, short-wavelength thermal radiation, and electromagnetic effects under local thermodynamic equilibrium. Validation against experimental data confirms accurate prediction of lightning channel temperature, pressure, radius, and shock wave development. Analysis of near-wall processes reveals distinct thermal stratification within lightning channels. Viscous wall effects significantly alter near-wall morphology and thermodynamics but have a negligible influence on shock wave evolution, which closely matches the one-dimensional case. The wall sustains complex stress states featuring high-frequency oscillatory shear stress and high-magnitude normal pressure as primary mechanisms for material tearing and fracture. Therefore, spectral experiments should account for stratified structures; protection designs should extend beyond attachment zones due to impact risks on adjacent structures; high vibration frequencies call for attention to the integrity of high-stiffness components and the prevention of fastener loosening; high pressure significantly elevates the risk of fracture and tearing failure in brittle materials.

A nine-species, twelve reaction steps surrogate-kerosene model was employed to investigate the flow field characteristics of a kerosene-fueled oblique detonation engine. Validation against both numerical and experimental results confirms the model's validity; on this basis, the flow field characteristics of combustor and influencing factors are investigated. Research results indicate that: (1) Typical wave structure within the combustor is characterized by the presence of oblique shock wave, oblique detonation wave, separation shock wave, reattachment shock wave, reflected shock wave, and separation zone. (2) Shock and expansion waves affect the species distribution, with the expansion process facilitating the transformation of intermediate products into final products. (3) As flight altitude increases, the separation shock wave and oblique detonation wave angles decrease, enlarging the separation zone scale and displacing the reattachment shock wave origin downstream, and the reflected shock wave vanishes entirely. (4) As the equivalence ratio is increased from 0.8 to 1.2, both the separation shock wave and oblique detonation wave angles enlarge and the reattachment shock wave migrates upstream. With the equivalence ratio further increased, a counter-trend emerges. (5) Wave structure oscillations at a frequence identical to that of the incoming flow, while the reflected shock wave are repeatedly generated and annihilated as the separation zone contracts and expands. Oscillatory waveforms of pressure and density within the separation zone deviate markedly from the incoming flow. (6) Spatial non-uniformities in fuel mixing exert a virtually negligible influence on the wave structure.