Computers & Fluids最新文献

{"title":"Towards scalable surrogate models based on neural fields for large scale aerodynamic simulations","authors":"Giovanni Catalani ,&nbsp;Jean Fesquet ,&nbsp;Xavier Bertrand ,&nbsp;Frédéric Tost ,&nbsp;Michael Bauerheim ,&nbsp;Joseph Morlier","doi":"10.1016/j.compfluid.2025.106929","DOIUrl":"10.1016/j.compfluid.2025.106929","url":null,"abstract":"<div><div>This paper introduces a novel surrogate modeling framework for aerodynamic applications based on Neural Fields. The proposed approach, MARIO (Modulated Aerodynamic Resolution Invariant Operator), addresses non parametric geometric variability through an efficient shape encoding mechanism and exploits the discretization-invariant nature of Neural Fields. It enables training on significantly downsampled meshes, while maintaining consistent accuracy during full-resolution inference. These properties allow for efficient modeling of diverse flow conditions, while reducing computational cost and memory requirements compared to traditional CFD solvers and existing surrogate methods. The framework is validated on two complementary datasets that reflect industrial constraints. First, the AirfRANS dataset consists of a two-dimensional airfoil benchmark with non-parametric shape variations. Performance evaluation of MARIO on this case demonstrates an order of magnitude improvement in prediction accuracy over existing methods across velocity, pressure, and turbulent viscosity fields, while accurately capturing boundary layer phenomena and aerodynamic coefficients. Second, the NASA Common Research Model features three-dimensional pressure distributions on a full aircraft surface mesh, with parametric control surface deflections. This configuration confirms MARIO’s accuracy and scalability. Benchmarking against state-of-the-art methods demonstrates that Neural Field surrogates can provide rapid and accurate aerodynamic predictions under the computational and data limitations characteristic of industrial applications.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"306 ","pages":"Article 106929"},"PeriodicalIF":3.0,"publicationDate":"2026-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145683872","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"On the spurious modes associated with the pressure centred low Mach number fix for compressible flows","authors":"Jonathan Jung ,&nbsp;Ibtissem Lannabi ,&nbsp;Vincent Perrier","doi":"10.1016/j.compfluid.2025.106945","DOIUrl":"10.1016/j.compfluid.2025.106945","url":null,"abstract":"<div><div>Density based finite volume schemes usually used for the computation of compressible flows are known to be, in general, not accurate in the low Mach number limit. A very popular low Mach number fix that has been proposed in different forms consists in centring the gradient of the pressure. In this article, we first perform some numerical experiments for proving that this low Mach number fix may lead to non-convergence on the momentum, especially if unstructured meshes are used. We also show that on the long time limit of the wave system, the same kind of problem may occur.</div><div>On triangular meshes, we perform a full analysis of the spurious modes of the wave system, and make evidence that the numerical scheme exhibits some oscillatory spurious modes. On simple triangular mesh configurations, these spurious modes are explicitly built. Then on general triangular meshes, we explain how to compute a basis of the spurious modes by relying on the ones built on simple configurations. A major outcome of this article is that the dimension of the spurious modes space is very large, approximately equal to the number of nodes of the mesh.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"306 ","pages":"Article 106945"},"PeriodicalIF":3.0,"publicationDate":"2026-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Stabilization of a hypersonic boundary layer via the oxidized surface of carbon/silicon carbide composite","authors":"Liangzhong Fan ,&nbsp;Qingyong Zhu ,&nbsp;Zhihui Li ,&nbsp;Dongheng Lai","doi":"10.1016/j.compfluid.2025.106948","DOIUrl":"10.1016/j.compfluid.2025.106948","url":null,"abstract":"<div><div>The evolution of second-mode waves in hypersonic boundary layers may be affected by the surface morphology of thermal protection systems. In this study, passive oxidation experiment is conducted on carbon/silicon carbide (C/SiC) thermal protection materials, yielding oxide-layer thicknesses of approximately 0.6–0.9 mm. Furthermore, the relationship between oxide layer thickness and composite material composition is analyzed. A strategy is proposed to mitigate second mode waves by tailoring the compositional distribution of C/SiC composites to form a wavy oxidized surface. Numerical simulations are performed to investigate the effect of the wavy surfaces with wave heights of 0.6 mm and 1.8 mm on the evolution of near-wall disturbances at Mach 6. When the wavy surface extends downstream of the synchronization point, it significantly suppresses second mode wave amplitude. However, when positioned only upstream of the synchronization point, a wave height of 0.6 mm instead amplifies disturbance waves. Increasing the wave height to the average boundary layer thickness (1.8 mm) shifts the peak disturbance location upstream but does not reduce its amplitude. Additionally, an idealized oxidized surface with a linearly graded composite composition is examined. Compared to an arc-shaped wavy surface, the idealized variable-curvature wavy surface provides more sustained suppression of the disturbance wave amplitude. Further analysis suggests that the modulation of disturbance waves by pressure oscillations induced by the wavy surface may be a key mechanism for reducing the second mode amplitude.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"306 ","pages":"Article 106948"},"PeriodicalIF":3.0,"publicationDate":"2026-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145787886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A pressure-based hybrid framework for sub- and supersonic compressible two-phase flow with non-equilibrium phase change","authors":"Lubing Xu ,&nbsp;Yanfei Li ,&nbsp;Xiao Ma ,&nbsp;Hengjie Guo ,&nbsp;Shijin Shuai ,&nbsp;Alexander Alekseevich Shevelev ,&nbsp;Matvey Kraposhin","doi":"10.1016/j.compfluid.2025.106912","DOIUrl":"10.1016/j.compfluid.2025.106912","url":null,"abstract":"<div><div>Pressure-based hybrid Kurganov-Noelle-Petrova (KNP) -PIMPLE algorithms have been widely used in all-Mach compressible flows to ensure the solution stability and maintaining physical consistency, but it suffers from high numerical diffusion at contact discontinuities and resulting in non-physical numerical oscillations at the two-phase regions, limiting the physical understanding of non-equilibrium phase change (flash boiling). In this work, a new hybrid solver named <em>twoPhaseMixingHLLCFoam</em> was proposed to avoid the drawbacks with three improvements: 1) Harten-Lax-van Leer-Contact (HLLC) scheme was used to replace the KNP scheme as a part of the hybrid framework to reduce the numerical diffusion; 2) The interpolation method of sound speed for Riemann solvers was modified to correctly reproduce the sound speed in two-phase buffer regions; 3) A new blending function based on local sound speed for identification of compressible and incompressible regimes was introduced.</div><div>The solver has been verified and validated against the analytical and experimental results. The improvement in numerical diffusion and the capability to resolve the pressure waves and shock waves propagation in gas-gas, gas-liquid and liquid-liquid flows were verified by a series of typical 1D cases. Then, the solver’s capability in resolving complex flow structures using 2D cases was demonstrated, especially in terms of the effectiveness of the transition between compressible and incompressible fluxes for all-Mach flows containing sub-, tran- and supersonic regimes simultaneously. Finally, a numerical study of acetone flash boiling jet has been conducted to demonstrate the solver’s capability of resolving under-expanded compressible two-phase jets with phase change. Scaling tests have shown that parallel performance of the new and original hybrid solvers is almost the same. Due to the adoption of a generalized mixture model, the new solver can be extended to multicomponent and even multiphase scenarios.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106912"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145615892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Three-dimensional super-resolution reconstruction of turbulent flow using 3D-ESRGAN with random sampling strategy","authors":"Linqi Yu,&nbsp;Yanyun Chen,&nbsp;Mustafa Z. Yousif,&nbsp;Hee-Chang Lim","doi":"10.1016/j.compfluid.2025.106890","DOIUrl":"10.1016/j.compfluid.2025.106890","url":null,"abstract":"<div><div>This study introduces a deep learning framework that uses an enhanced three-dimensional super-resolution generative adversarial network (3D-ESRGAN) to reconstruct high-resolution turbulent flow fields from low-resolution data. To minimize the reliance on complete datasets during training, a random sampling strategy is used. In each training epoch, approximately 1 % of spatial points are randomly chosen from both the predicted and ground-truth fields for loss computation. This method mimics sparse sensor measurements in real-world experiments, allowing the model to learn accurate mappings based on limited spatial observations. Furthermore, physics-guided loss functions are incorporated to ensure consistency with the underlying physical laws, thereby improving the reliability of the reconstructed flow fields. Subsequently, the framework is tested on two typical flow scenarios: (1) flow over a finite wall-mounted square cylinder at <span><math><mrow><msub><mrow><mi>Re</mi></mrow><mi>d</mi></msub><mo>=</mo><mn>500</mn></mrow></math></span>, and (2) fully developed turbulent channel flow at <span><math><mrow><msub><mrow><mi>Re</mi></mrow><mi>τ</mi></msub><mo>=</mo><mn>180</mn></mrow></math></span> and 500. Both scenarios are generated using direct numerical simulation (DNS). Finally, the results are presented and analyzed both qualitatively and quantitatively. The qualitative findings show that the reconstructed fields effectively restore the key vortex structures and turbulence features absent in the coarse data. Quantitative comparisons with the ground-truth data confirm high accuracy in terms of velocity profiles, Reynolds stresses, probability density functions (PDFs), and energy spectra. Additionally, the relative error regarding the streamwise velocity magnitude is calculated, and all cases show a low error rate of approximately 5 %. In summary, the findings confirm the efficacy of combining GAN-based super-resolution with a random sampling strategy for accurate and data-efficient 3D turbulence reconstruction. The results suggest that the proposed framework can be successfully applied in real-world scenarios where only sparse measurements are accessible.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106890"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145518589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generalization-aware structured mesh smoothing via graph neural networks","authors":"Jingzhu Pei ,&nbsp;Xinhai Chen ,&nbsp;Yang Liu ,&nbsp;Zhichao Wang ,&nbsp;Qingyang Zhang ,&nbsp;Qinglin Wang ,&nbsp;Xiang Gao ,&nbsp;Jie Liu","doi":"10.1016/j.compfluid.2025.106895","DOIUrl":"10.1016/j.compfluid.2025.106895","url":null,"abstract":"<div><div>Structured meshes provide computationally efficient discretization for scientific simulations through geometrically ordered mesh units. Because of their strict orthogonality constraints, it is still challenging to generate high-quality structured meshes. Although traditional optimization techniques employ iterative node relocation to enhance element quality, they suffer from accuracy-efficiency trade-offs. Existing intelligent methods based on supervised learning attempt to circumvent these limitations through data-driven optimization. However, these methods are inherently limited by their dependence on domain-specific training data and their inability to generalize across diverse complex geometries. To overcome these problems, we propose Structured-Mesh Graph Neural Network-based smoothing(SMGNN-Smoothing), a generalization-aware unsupervised method for structured mesh optimization. Our method integrates three key innovations: (1) a graph neural network that aggregates neighborhood features to predict optimized node configurations, (2) an adaptive normalization technique enabling consistent processing of multi-resolution meshes, (3) a well-designed loss function controls the whole training process, StructureLoss. SMGNN-Smoothing realizes excellent optimization performance across multiple quality metrics. It outperforms existing supervised learning methods and shows strong generalization capability. Compared with optimization-based smoothing, it achieves an order-of-magnitude improvement in computational efficiency.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106895"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145570079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hybrid High-order formulations with turbulence modelling capabilities for incompressible flow problems","authors":"Lorenzo Botti ,&nbsp;Daniele A. Di Pietro ,&nbsp;Francesco Carlo Massa","doi":"10.1016/j.compfluid.2025.106915","DOIUrl":"10.1016/j.compfluid.2025.106915","url":null,"abstract":"<div><div>We propose a Hybrid High-Order (HHO) formulation of the incompressible Navier–Stokes equations, that is well suited to be employed for the simulation of turbulent flows. The spatial discretization relies on hybrid velocity and pressure spaces and the temporal discretization is based on Explicit Singly Diagonal Implicit Runge-Kutta (ESDIRK) methods. The formulation possesses some attractive features that can be fruitfully exploited when high-fidelity computations are required, namely: pressure-robustness, conservation of volume enforced cell-by-cell up to machine precision, robustness in the inviscid limit, implicit high-order accurate time stepping with local time step adaptation, reduced memory footprint thanks to static condensation of both velocity and pressure, possibility to exploit inherited <em>p</em>-multilevel solution strategies to improve performance of iterative solvers. After demonstrating the relevant properties of the scheme in practice, performing challenging 2D and 3D test cases, we consider the simulation of the Taylor–Green Vortex flow problem at Reynolds 1 600.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106915"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145615827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Modeling advection-dominated flows with space-local reduced-order models","authors":"T. Van Gastelen ,&nbsp;W. Edeling ,&nbsp;B. Sanderse","doi":"10.1016/j.compfluid.2025.106911","DOIUrl":"10.1016/j.compfluid.2025.106911","url":null,"abstract":"<div><div>Reduced-order models (ROMs) are often used to accelerate the simulation of large physical systems. However, traditional ROM techniques, such as proper orthogonal decomposition (POD)-based methods, often struggle with advection-dominated flows due to the slow decay of singular values. This results in high computational costs and potential instabilities.</div><div>This paper proposes a novel approach using space-local POD to address the challenges arising from the slow singular value decay. Instead of global basis functions, our method employs local basis functions that are applied across the domain, analogous to the finite element method, but with a data-driven basis. By dividing the domain into subdomains and applying the space-local POD, we obtain a sparse representation that generalizes better outside the training regime. This allows the use of a larger number of basis functions compared to standard POD, without prohibitive computational costs. To ensure smoothness across subdomain boundaries, we introduce overlapping subdomains inspired by the partition of unity method.</div><div>Our approach is validated through simulations of the 1D and 2D advection equation. We demonstrate that using our space-local approach, we obtain a ROM that generalizes better to flow conditions not included in the training data. In addition, we show that the constructed ROM inherits the energy conservation and non-linear stability properties from the full-order model. Finally, we find that using a space-local ROM allows for larger time steps.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106911"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145615891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Direct Poisson solver combining domain decomposition and influence matrix methods and its application to DNS of oscillating grid turbulence","authors":"Toru Yamada ,&nbsp;Ryuga Sumi ,&nbsp;Yohei Morinishi","doi":"10.1016/j.compfluid.2025.106891","DOIUrl":"10.1016/j.compfluid.2025.106891","url":null,"abstract":"<div><div>This study presents the application of a direct Poisson solver combining domain decomposition and influence matrix methods to the direct numerical simulation (DNS) of oscillating grid turbulence (OGT). Solving the pressure Poisson equation is one of the major challenges in computational fluid dynamics. Traditional direct methods are accurate but difficult to be applied to large-scale problems, while iterative methods can suffer from slow convergence. The hybrid approach enables the use of the direct methods for the problem with complex computational geometry. The results show that this direct solver significantly reduces computation time compared to the iterative BiCGSTAB method. The DNS results are validated by comparison of experimental data, demonstrating good agreement in the vertical distribution of velocity fluctuation intensity. Therefore, the efficiency of this direct method for simulating the OGT turbulent flows is confirmed.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106891"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145479252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and assessment of an improved Baldwin-Lomax turbulence model for complex hypersonic flows in two and three dimensions","authors":"Jian Shen ,&nbsp;Xianliang Chen ,&nbsp;Lin Fu","doi":"10.1016/j.compfluid.2025.106917","DOIUrl":"10.1016/j.compfluid.2025.106917","url":null,"abstract":"<div><div>Accurate prediction of hypersonic turbulent boundary layers is critical for the design of hypersonic vehicles. Traditional turbulence models were originally developed for incompressible flows and are commonly extended to compressible conditions by employing density-weighted averages. In a recent study, Chen et al. (J. Fluid Mech., 2024, vol 987, A7) proposed an improved Baldwin-Lomax (BL) turbulence model by incorporating velocity transformations and the temperature-velocity relation. Their modifications yielded notable improvements for high-speed zero-pressure-gradient flat-plate flows. Building upon this foundation, the present study introduces further enhancements to Chen et al.’s BL model to improve its accuracy and robustness for complex hypersonic configurations. The improved BL turbulence model is implemented into a standard computational fluid dynamics (CFD) solver and validated against direct numerical simulation results and experimental data across two- and three-dimensional hypersonic cases involving pressure gradients, cold walls, and shock/boundary-layer interactions. The results show that the improved BL turbulence model generally achieves superior accuracy in attached flow regions compared to the baseline BL, Spalart-Allmaras and k–<em>ω</em> SST turbulence models. These findings highlight the model’s potential for practical use in hypersonic flow simulations, offering a valuable tool for aerospace engineering applications.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"305 ","pages":"Article 106917"},"PeriodicalIF":3.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145570082","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}