Sorry, you need to enable JavaScript to visit this website.
Partager

Publications

Sont listées ci-dessous, par année, les publications figurant dans l'archive ouverte HAL.

2026

  • Asymptotic homogenization of microstructured interfaces
    • Lebbe Nicolas
    • Maurel Agnès
    • Pham Kim
    , 2027. This chapter extends the homogenization framework developed in Chapter 1 to the description of wave propagation in layers of metamaterials with deeply subwavelength thicknesses: i.e. metasurfaces. In this regime, the thickness of the structured medium is negligible compared with the wavelength, leading to an interface-type homogenization instead of a volumetric one. We derive effective transmission conditions and associated surface parameters that replace the heterogeneous layer with equivalent transmission conditions, capturing the macroscopic impact of the underlying microstructure while preserving the essential physical features of the original system.
  • Interface homogenization in Electromagnetics: Localized surface plasmon resonance
    • Lebbe Nicolas
    • Maurel Agnès
    • Pham Kim
    , 2027. In this chapter, we extend the homogenization results derived in Chapter 2 for penetrable inclusions to incorporate localized plasmonic resonances arising in electromagnetic media containing particles with negative and dispersive permittivity. In particular, we show that the dispersion relations of the effective material parameters associated with plasmonic metasurfaces can be characterized through the solution of a single plasmonic eigenvalue problem.
  • Direct cyclic approach for the numerical simulation of fretting wear with plasticity
    • Caradec Quentin
    • Breuzé Matthieu
    • Fayard Jean-Luc
    • Prabel Benoît
    • Maitournam Habibou
    , 2026. Whereas the development of significant near-surface plastic strain is considered as a key feature of fretting wear mechanisms, few numerical investigations of fretting wear phenomena considering a cyclic elastoplastic material behavior have been conducted. Indeed, the integration of a path-dependent constitutive model in wear simulation frameworks is not straightforward. Due to the large number of cycles involved in wear processes, numerical methods devoted to wear simulation usually use a cycle jump to accelerate the computations, which allows to reach otherwise unrealistically high cycle numbers. When the material behavior depends on path-dependent variables, their evolution over the cycle jump needs to be described, which poses an additional difficulty. In the present paper, two different strategies are developed and compared for this purpose. First, plasticity is treated using the same cycle jump method as for wear evolution. This method effectively allows to accelerate the simulation but is limited to small acceleration factors, which is not sufficient to reach the desired high numbers of cycles. The second technique is based on the direct determination of an asymptotic state for cyclic plasticity. This method has the advantage of decoupling the calculation of wear evolution from the treatment of cyclic plasticity, which allows to use advanced efficient methods for wear evolution such as an implicit cycle jump. In the application case considered in this paper, this method proves to be very efficient and provides valuable insights into the high-cycle evolution of cyclic plasticity, exhibiting an interaction with wear evolution.
  • Internal combination resonance in a Micro Electro-Mechanical three-axis gyroscope: Experimental results and reduced-order modelling via invariant manifolds
    • Colombo Alessio
    • Villa Marco
    • Morelli Federico
    • Fedeli Patrick
    • Touzé Cyril
    • Frangi Attilio
    Mechanical Systems and Signal Processing, Elsevier, 2026, 260, pp.114933. This work investigates internal combination resonance in a Micro-Electro-Mechanical System (MEMS) three-axis gyroscope, where energy couples among three vibrational modes via the relationship ω1 + ω2 ≈ ω3 . The phenomenon is first characterised experimentally through frequency response measurements and spectrograms, revealing paired resonance peaks that emerge only under dual-frequency excitation. A theoretical framework grounded in complex normal form theory is then developed, deriving analytical expressions for the frequency-response curves which are shown to replicate the experimental findings. A reduced-order model of the fully coupled electromechanical problem is constructed via Direct Parametrisation of Invariant Manifolds applied to the full finite element mesh (1.3 million degrees of freedom), demonstrating quantitative agreement with experiments and validating the predicted mechanisms. By combining invariant manifold parametrisation with complex normal form theory, the methodology bridges the gap between analytical understanding and practical finite element reduction, yielding explicit solutions for nonlinear phenomena whilst maintaining computational efficiency for large-scale systems. The results validate invariant manifold methods as a robust tool for MEMS design and control applications. (10.1016/j.ymssp.2026.114933)
    DOI : 10.1016/j.ymssp.2026.114933
  • Analytical and experimental study of turbulence-induced vibrations in a cantilever-mounted coaxial cylinder
    • Plado Costante Loucas
    • Lagrange Romain
    • Kocher Maud
    • Puscas Maria Adela
    Journal of Fluids and Structures, Elsevier, 2026, 147, pp.104678. This paper investigates turbulence-induced vibrations in a confined annular flow configuration representative of pressurized water reactor internals. The experimental setup consists of a rigid inner cylinder mounted on two flexible rectangular plates inside a cylindrical vessel. Building on our previous work, analytical expressions for the fluid-elastic forces in quiescent fluid are derived, accounting for added mass, damping, partial immersion, viscous, and confinement effects. Theoretical predictions of the natural frequencies for various filling heights are obtained and systematically validated against experimental measurements. A new nondimensional formulation of the governing equations further yields an analytical expression for the root mean square displacement under turbulent forcing. The root mean square prediction is based on an effective homogeneous turbulence model characterized by a prescribed pressure power spectral density and correlation lengths drawn from the literature. Although dedicated measurements indicate that the turbulence in the present configuration is not strictly homogeneous, introducing a calibrated cut-off reduced frequency in the pressure spectrum yields a consistent and quantitatively accurate description of the root mean square response as a function of reduced velocity. (10.1016/j.jfluidstructs.2026.104678)
    DOI : 10.1016/j.jfluidstructs.2026.104678
  • Numerical benchmark to assess sensitivity to spatial discretization in quasi-static fracture simulations : Application to phase-field models
    • Zembra Edgar
    • Loiseau Flavien
    • Lazarus Véronique
    • Henry Hervé
    , 2026. Numerous numerical methods based on Linear Elastic Fracture Mechanics (LEFM), such as phase-field fracture models, enable the solution of crack propagation problems. Despite established mathematical properties (e.g., Γ-convergence), their numerical discretization can introduce systematic biases that persist even with spatial mesh refinement, such as artificial anisotropy in fracture properties at the element scale. Although acknowledged, these biases remain underexplored in practice. To address this gap, this work introduces a numerical benchmark for quantitatively evaluating discretization-induced biases. The benchmark focuses on the propagation of an off-centered crack in a band subjected to a sliding tensile load, where the crack follows an exponentially decaying trajectory toward the band center. This path is verified against a sharp crack LEFM reference solution. This non-trivial yet well-characterized crack path offers a robust basis for the quantitative assessment and the comparison of different crack propagation solvers. To illustrate this benchmark, we assess two phase-field implementations: one based on the finite element method and another on the finite difference method. These simulations reveal that discretization indeed introduces local artificial anisotropy. For structured meshes, crack path curvature is not captured, even for element sizes considered small enough in classical practice. For unstructured meshes, while the path is well-captured at smaller sizes, oscillations in the predicted trajectory are observed. Used here to assess the discretization biases of two specific numerical methods, this benchmark test is intended to evaluate other numerical methods for simulating crack propagation.
  • A robust computational framework for the mixture-energy-consistent six-equation two-phase model with instantaneous mechanical relaxation terms
    • Orlando Giuseppe
    • Haegeman Ward
    • Pelanti Marica
    • Massot Marc
    Computers and Fluids, Elsevier, 2026, 317, pp.107193. We present a robust computational framework for the numerical solution of a hyperbolic 6-equation single-velocity two-phase system. The system's main interest is that, when combined with instantaneous mechanical relaxation, it recovers the solution of the 5-equation model of Kapila. Several numerical methods based on this strategy have been developed over the years. However, neither the 5- nor 6-equation model admits a complete set of jump conditions because they involve non-conservative products. Different discretizations of these terms in the 6-equation model exist. The precise impact of these discretizations on the numerical solutions of the 5-equation model, in particular for shocks, is still an open question to which this work provides new insights. We consider the phasic total energies as prognostic variables to naturally enforce discrete conservation of total energy and compare the accuracy and robustness of different discretizations for the hyperbolic operator. Namely, we discuss the construction of an HLLC approximate Riemann solver in relation to jump conditions. We then compare an HLLC wave-propagation scheme which includes the non-conservative terms, with Rusanov and HLLC solvers for the conservative part in combination with suitable approaches for the non-conservative terms. We show that some approaches for the discretization of non-conservative terms fit within the framework of path-conservative schemes for hyperbolic problems. We then analyze the use of various numerical strategies on several relevant test cases, showing both the impact of the theoretical shortcomings of the models as well as the importance of the choice of a robust framework for the global numerical strategy. (10.1016/j.compfluid.2026.107193)
    DOI : 10.1016/j.compfluid.2026.107193
  • Développement d’un modèle prédictif de l’usure pour les composants REP
    • Fayard Jean-Luc
    • Bentivegna Filippo
    • Breuze Matthieu
    • Prabel Benoit
    • Maitournam Habibou
    • Stolz Claude
    , 2026. Les composants internes des réacteurs à eau pressurisée sont soumis à des sollicitations mécaniques et thermohydrauliques répétées, conduisant à des contacts de type impact-glissement, glissement ou fretting. Ces sollicitations peuvent entraîner une usure progressive des matériaux, caractérisée par une perte de matière et une modification de la géométrie des surfaces en contact. La prédiction de cette évolution constitue un enjeu important pour l’analyse de la durée de vie des composants, la maintenance et la sûreté des installations. Les modèles classiques d’usure, tels que la loi d’Archard ou les lois fondées sur l’énergie dissipée par frottement à la surface de contact, permettent de corréler certains résultats expérimentaux. Cependant, les coefficients associés à ces approches dépendent fortement de la géométrie, du chargement, du couple de matériaux et de l’environnement. Cette dépendance limite leur portée prédictive lorsque les conditions d’application changent. Pour dépasser ces limitations, des approches plus physiques peuvent être envisagées, notamment le modèle fondé sur l’endommagement proposé par Caradec et al., dans lequel la perte de matière est associée à l’accumulation progressive d’un endommagement local. Ce travail s’inscrit dans le développement d’une méthodologie numérique de prédiction de l’usure par fretting, appliquée à une configuration cylindre/plan modélisée en déformations planes sous Cast3M. Le chargement considéré consiste en une mise en contact sous effort normal constant, suivie d’un déplacement tangentiel cyclique imposé. Avant les calculs d’usure, le cadre numérique est validé à partir de la solution analytique de Hertz, en comparant la pression de contact et les contraintes en profondeur. Cette étape permet de préciser les choix de maillage, de type d’éléments et de post-traitement des champs mécaniques. Une nouvelle stratégie de maillage est ensuite mise en place afin de conserver un raffinement suffisant dans la zone de contact et de glissement, tout en réduisant le coût numérique des calculs multi-cycles. Le raffinement local est piloté par le nombre d’éléments dans la demi-largeur de contact hertzienne. Des études de sensibilité sont également réalisées afin de fixer les paramètres numériques nécessaires aux simulations d’usure, notamment la discrétisation du cycle de fretting et le facteur d’accélération utilisé pour réduire le coût des calculs multi-cycles. La contribution présente la comparaison de deux approches de simulation de l’usure : l’approche énergétique et l’approche fondée sur l’endommagement. Ces deux approches sont appliquées à un essai de fretting simple issu des travaux de Marc et al., pour lequel la mesure expérimentale disponible est le volume usé final. Enfin, une troisième approche est introduite. Elle se fonde sur l’exploitation des champs de contraintes et de déformations sous la zone de contact. L’objectif est d’identifier une grandeur mécanique caractéristique de l’usure dans le volume de matière, puis de définir un critère permettant de déterminer les zones usées. La pertinence de cette approche est évaluée en comparant ses prédictions à celles des approches fondées sur l’énergie dissipée par frottement à la surface de contact et sur l’endommagement local.
  • On the validity limits of the parametrisation method for invariant manifolds: an assessment of practical criteria for vibrating systems
    • Stabile André
    • Grolet Aurélien
    • Vizzaccaro Alessandra
    • Touzé Cyril
    Nonlinear Dynamics, Springer Verlag, 2026, 114 (16), pp.1058. The parametrisation method for invariant manifolds is a powerful technique for deriving reducedorder models in the context of nonlinear vibrating systems with geometric nonlinearities, allowing accurate computations of nonlinear normal modes. Thanks to arbitrary order asymptotic expansions, converged results are within reach and directly applicable to finite element structures. However, since it relies on a local theory and asymptotic expansions, the results are only valid up to a given amplitude, which defines the convergence radius of the approximation. The aim of this contribution is to investigate the validity limits of the approach and review the existing error estimates, with the concrete objective of proposing a practical approach to estimate the validity range during the computation, thus producing safe bounds within which the reduced-order model can be used. Three different criteria are assessed. The first one uses the error in the invariance equation as the distance to the fixed point increases. The second one is adapted from an upper bound criterion derived for normal form transforms and based on the potential singularities of the homological operator. The third one uses validity limits of series expansion through Cauchy and d'Alembert rules, which can be tested either on the reduced dynamics coefficients or those of the nonlinear mappings. The criteria are tested on a number of different examples that are representative of the situations encountered when dealing with nonlinear vibrations. The Duffing equation serves as a first benchmark that allows considering conservative oscillations, forced systems at primary resonance, and superharmonic resonance. The investigations are then extended to a vibrating system with two degrees of freedom. Finally, the different criteria are assessed on a finite element beam structure, and guidelines are formulated to generalise their practical use and produce accurate and easy-to-use error bounds in the context of model order reduction for nonlinear vibrating structures. (10.1007/s11071-026-12956-0)
    DOI : 10.1007/s11071-026-12956-0
  • Magnetic and electromagnetic vibration absorbers
    • Lo Feudo Stefania
    • Li Haiqin
    • Cumunel Gwendal
    • Li Ang
    • Touzé Cyril
    , 2026, 254, pp.171-202. In this chapter, two different devices that use magnetic and electromagnetic forces for the purpose of vibration mitigation, are presented. First, a magnetic vibration absorber, solely relying on magnetic forces to create a nonlinear restoring force, is studied. Thanks to an accurate representation of the magnetic field and together with geometric design rules, the device is capable of displaying different kind of linear and nonlinear forces and can be thus easily adjusted to create either a nonlinear tuned vibration absorber (with positive linear stiffness), a nonlinear energy sink (NES with vanishing linear stiffness and thus only nonlinear restoring force), or a bistable NES with negative linear restoring force. The performance of the different absorber choices is analysed to mitigate the vibration of a three-storey structure. Second, an electromagnetic vibro-impact NES is investigated. The system consists of a linear oscillator equipped with multiple turns of coil, coupled to a magnet attachment that undergoes two-sided inelastic impacts. The device is analysed in terms of vibration mitigation, and it is shown that the addition of the electromagnetic forces enhances the robustness of the device as compared to its counterparts that use only mechanical forces. (10.1007/978-3-032-23215-1_7)
    DOI : 10.1007/978-3-032-23215-1_7
  • Model-order reduction for vibrating systems with the direct parametrisation method for invariant manifolds : Hopf bifurcations, validity limits and limit cycle continuation
    • de Figueiredo Stabile André
    , 2026. Providing accurate predictions of the behavior of nonlinear mechanical systems is an ever-relevant challenge, especially given today’s competitive industrial context. The need to fully utilize structural capabilities pushes the exploitation of nonlinear phenomena forward, which in turn requires increasingly accurate models. However, high-fidelity models can quickly become prohibitively large, in particular in the context of vibrating systems, making time-history analyses or frequency response curve computations too expensive. In this context, model-order reduction (MOR) techniques emerge as attractive strategies that allow for reducing the dimension of the problem at hand while maintaining a good precision level. While a plethora of MOR strategies exist, this thesis focuses on the use of nonlinear normal modes (NNMs), defined as invariant manifolds in phase space, to construct reduced-order models (ROMs) through the parametrisation method for invariant manifolds, leveraging their properties and restricting the system's dynamics to evolve within them, such that the full-order model (FOM) is replaced by a ROM of minimal size by projection. The method has been formulated in a direct version, the direct parametrisation of invariant manifolds (DPIM), that can be efficiently applied to large finite element models, and which has been tested in several classes of problem of practical interest, with the current focus being the inclusion of additional effects. The contributions of this thesis fit into this context and thus aim to address certain unexplored aspects of the method. In addition to being a MOR strategy, the method is also a perturbative technique, and can be used to yield analytical solutions to low-dimensional systems with high interpretative power. A first objective of the thesis is to explore such solutions, obtained with the help of a symbolic code developed and made publicly available, and compare them with those from standard perturbative approaches. We show that in this context, no ad hoc assumptions on the magnitude of nonlinear and dissipative terms have to be made by the parametrisation method, which is not the case for most perturbative approaches. Although high-order solutions generally provide better convergence, due to the asymptotic and thus inherently local nature of the parametrisation method, there is always a certain validity limit for its approximations. This fact is of high concrete interest, as FOM solutions are usually not at hand for comparison with the ROM. Another axis of interest in this thesis is then the prediction of such validity limits. In particular, we adapt and compare several approaches existing in the literature, and show how they can be used in practice to produce safe bounds within which the method can be employed. A third subject of interest is the extension of the method to parameter-dependent systems, and in specific those subjected to a Hopf bifurcation. They present a high complexity level, since the dynamics of the system qualitatively changes at the bifurcation, and physically they also correspond to situations of great practical interest, such as flutter instabilities of airplane wings. In this contribution, modifications to the DPIM algorithm are proposed, allowing to treat systems depending on a parameter and experiencing Hopf bifurcations. The last point addressed in this thesis is concerned with the application of the parametrisation method to limit cycles rather than fixed points. In specific, an analytical continuation procedure is proposed, allowing to extend the validity limits of previous parametrisation computations. We show how this can be used to extend bifurcation diagrams obtained with ROMs for Hopf bifurcations, but also to extend the validity limits of the method when considering backbone curves and frequency response curves.
  • Coupled CFD–DEM modelling of damage to a breakwater armour layer made of Antifer cubes exposed to regular waves
    • Barcet Matthieu
    • Benguigui William
    • Laviéville Jérome
    • Benoit Michel
    • Fede Pascal
    • Bonometti Thomas
    Ocean Engineering, Elsevier, 2026, 363, pp.126525. The study aims to demonstrate the capabilities of a CFD-DEM (Computational Fluid Dynamics -Discrete Element Method) to predict wave-structure interactions and armour units motions in a breakwater composed of Antifer cubes. To do so, the fluids (air and water) are simulated using a Eulerian-Eulerian CFD solver and the contacts between the blocks and the slope are solved using a DEM code. The codes were coupled and validated in a previous study, using idealised configurations that did not reflect actual coastal structures. This work extends the previous study to a semi-realistic configuration where the blocks are positioned along a inclined wall to represent the armour layer of a rubble-mound breakwater. Both experiments and simulations are performed, and the results are compared to evaluate the accuracy of the model. The comparisons are made on the number of displaced armour units and on the amplitude of their displacement.<p>Two regular configurations are studied: a one-layer configuration and a two-layers configuration which improves the stability and better represents realistic breakwaters. Finally, simulations are performed on a randomly generated armour layer which the goal of improving the realism of the breakwater.</p> (10.1016/j.oceaneng.2026.126525)
    DOI : 10.1016/j.oceaneng.2026.126525
  • Relationship Between the Ferrite/Austenite Ratio and Pitting Corrosion Resistance of 2507 Super Duplex Stainless Steel Produced by the LP‐DED Process
    • Ballesio Sébastien
    • Queval Jordan
    • Comas Chloé
    • Dhondt Matthieu
    • Doudard Cédric
    • Szmytka Fabien
    Materials and Corrosion / Werkstoffe und Korrosion, Wiley-VCH Verlag, 2026, 77 (10), pp.1692-1705. In this study, the pitting corrosion behavior of 2507 super duplex stainless steel (SDSS) produced by the laser powder‐directed energy deposition (LP‐DED) process was investigated in both as‐built and heat treated (HT) conditions. Compared with the wrought SDSS and HT LP‐DED SDSS, the as‐built LP‐DED SDSS exhibits an increase in the ferrite content of approximately 75%. For HT LP‐DED SDSS specimens, potentiodynamic polarization tests revealed that the pitting potential decreased with increasing heat treatment temperature. The Pitting Resistance Equivalent Number (PREN) was calculated for each phase in each material examined in this study to elucidate the relationship between microstructure and pitting corrosion resistance. (10.1002/maco.70182)
    DOI : 10.1002/maco.70182
  • Modal experimental continuation of an aircraft during ground vibration test
    • Chukwu Augustus
    • Stephan Cyrille
    • Touzé Cyril
    • Doaré Olivier
    , 2026. Phase-locked loop experimental continuation method is extended to multi-input multi-output testing, for the identification of the nonlinear dynamics of an aircraft during ground vibration test. More specifically, backbone curve tracking in moderately nonlinear regimes is investigated. Force appropriation and phase control over multiple excitation points are presented, and particular attention is devoted to introduce weighted mode purity indices, to robustly assess the nonlinear mode isolation. A full-scale aircraft is used to illustrate the methods, revealing the benefits that can be obtained for the identification of nonlinear characteristics. (10.2514/6.2026-4777)
    DOI : 10.2514/6.2026-4777
  • Path-following methods for quasi-static crack propagation: Application to phase-field fracture
    • Loiseau Flavien
    • Lazarus Véronique
    , 2026. Numerical simulations of quasi-static crack propagation in brittle materials often suffer from numerical instabilities, such as snapback events, due to structural softening. In variational phase-field fracture models, these instabilities manifest as abrupt crack jumps, thereby impeding physical validity, as the energy minimization is performed over a significant crack increment. Moreover, they often prevent incremental force boundary conditions, as they may lead to a loss of force balance as soon as the crack starts propagating. This study introduces path-following methods to mitigate such instabilities in quasi-static phase-field fracture simulations. We evaluate existing methods alongside a novel approach, Control by Maximum Strain Increment Outside the Crack (CMSIOC), which enforces stable crack growth by constraining strain increments in the uncracked region. All studied methods rely on a single scalar control equation that depends solely on displacement fields, enabling integration into classical staggered solvers without major changes. Performance is assessed via three benchmark problems of increasing complexity, with results compared against Linear Elastic Fracture Mechanics (LEFM) references based on Griffith's criterion and the G-max criterion. Our findings show that CMSIOC: 1. Accurately follows the equilibrium path, avoiding abrupt crack jumps while preserving physical validity; 2. Supports force-controlled boundary conditions without loss of force balance during propagation; 3. Ensures uniform incremental crack growth, properly distributing the computational efforts across load steps. These results highlight CMSIOC's robustness as a model- and problem-independent solution for stable phase-field fracture simulations.
  • Path-following methods for phase-field simulation of quasi-static crack propagation
    • Loiseau Flavien
    • Lazarus Véronique
    International Journal of Solids and Structures, Elsevier, 2026, 334, pp.113974. The variational approach to fracture, particularly through its regularization as a phase-field model, has become a widely used tool for simulating the quasi-static propagation of cracks in structures. However, classic incremental loading can induce unstable crack growth, violating the quasi-static assumption, and in some cases, leads to a loss of force balance, preventing self-consistency and the estimation of dissipated energy during snapback instabilities. To address this challenge, path-following methods are investigated. Their aim is to adjust the applied load so that it stays at the propagation threshold, thereby preserving the quasi-static assumption and ensuring equilibrium solutions. In this work, we apply and evaluate multiple path-following methods within the framework of variational phase-field fracture models, which are developed to regularize linear elastic variational sharp crack evolution problems. Our study pursues two objectives. First, we review several existing path-following methods, with a focus on partitioned strategies based on the displacement field, which decouple the path-following control equation from the rest of the system, facilitating easier integration with staggered solvers. In addition, we introduce a new path-following method whose particularity is to limit the maximum strain increment outside the cracked regions. Second, we use the Γ-convergence to the sharp crack model to evaluate these methods across three crack propagation problems of increasing complexity. The comparison demonstrates that the proposed path-following method offers a simple yet highly effective approach to capture the equilibrium path in phase-field fracture simulations. This method robustly maintains the quasi-static assumption, ensuring physically meaningful results. By enabling accurate estimation of the energy dissipated during snapback instabilities, it paves the way for the rational design of more resistant heterogeneous materials. (10.1016/j.ijsolstr.2026.113974)
    DOI : 10.1016/j.ijsolstr.2026.113974
  • Path-Following Methods for Quasi-Static Crack Propagation in Phase-Field Fracture Models
    • Loiseau Flavien
    • Lazarus Véronique
    , 2026. Quasi-static crack propagation simulations in brittle materials often suffer from numerical instabilities, such as snapback events, due to structural softening. In phase-field fracture models, these instabilities manifest as abrupt crack jumps, impeding the physical validity as the energy minimization is performed over a significant crack increment. Moreover, they often prevent the imposition of incremental force boundary conditions, as they may lead to a loss of force balance as soon as the crack starts to propagate. This study explores the application of path-following methods to variational phase-field fracture simulation. It aims to identify a method that is robust, model- and problem-independent, and easy to implement in a classic staggered solver. To this aim, we evaluate existing path-following techniques alongside a novel proposition: Control by Maximum Strain Increment Outside the Crack (CMSIOC), which ensures stable crack propagation by limiting strain increments in the uncracked region. All studied methods introduce a single scalar control equation dependent solely on displacement fields, enabling seamless integration into classic staggered solvers without requiring significant modifications. To assess the performance of these methods, we compare them across three numerical benchmark problems of increasing complexity. For reference, we employ a sharp crack model from Linear Elastic Fracture Mechanics (LEFM) based on Griffith's theory and the G-max criterion. Our findings demonstrate that CMSIOC reliably follows the equilibrium path and closely replicates LEFM benchmark solutions. Hence, it enables the capture of snapback instabilities without abrupt crack jumps, also allowing for force-controlled boundary conditions without loss of force balance. Another advantage is that it provides nearly uniform crack growth per increment, thereby balancing the computational cost across load steps. This presentation will: (1) recall the concept of equilibrium paths in fracture mechanics, (2) introduce the CMSIOC, detailing its theoretical foundation and implementation, (3) present the results of the benchmark problems, and (4) conclude on this work with practical recommendations for integrating path-following methods into phase-field fracture simulations.
  • Experimental and Numerical Investigation of Aeroacoustic Installation Effects on Small-Scale Multi-Rotor Systems in Hover
    • El Ouni Khalil
    • Simiao Pitta Vinicius
    • Cotté Benjamin
    , 2026. This paper investigates aeroacoustic installation effects in small-scale multi-rotor systems in hover through a combined experimental and numerical approach. Measurements were conducted in an anechoic chamber for three configurations: an isolated propeller, a propeller with its supporting arm, and a tandem propeller arrangement representative of a quadcopter layout. The propellers operate at very low Reynolds numbers, around 3 × 10 4 , where transitional flow and viscous effects are important. A mid-fidelity aerodynamic model based on a NonLinear Vortex Lattice Method is considered to predict the aerodynamic forces used as input to Hanson's tonal noise model. The aerodynamic simulation is shown to be very sensitive to airfoil polar data. In the tandem configuration, unsteady loading noise plays a significant role. The results show that a small variation of rotational speed of rotors induces a change of directivity for the unsteady loading noise and the total noise. (10.2514/6.2026-3349)
    DOI : 10.2514/6.2026-3349
  • Theoretical Study of Turbulence Ingestion Noise for Propellers in Nonuniform Flows
    • Lavanant Romain
    • Cotté Benjamin
    • Serre Gilles
    • Mercier Jean-François
    , 2026. This paper presents an analytical model for predicting the broadband noise radiated by a propeller ingesting a spatially nonuniform turbulent flow. The formulation, derived from Goldstein's acoustic analogy, expresses the far-field power spectral density through azimuthal Fourier decompositions of the incident heterogeneous turbulence spectrum, thereby retaining inter-strip and blade-to-blade correlations along the blade span. The spatially varying turbulence parameters (integral length scale and turbulence intensity) that serve as inputs to the spectrum models are derived analytically from Prandtl's mixing-length shear-production approach. Two turbulence spectrum formulations are considered: a heterogeneous isotropic von Kármán spectrum and a heterogeneous anisotropic Kerschen-Gliebe spectrum. A comparison with experimental data for a rotor ingesting a turbulent boundary layer highlights the respective roles of anisotropy and heterogeneity. On the one hand, an isotropic turbulence description fails to reproduce the measured spectral humps (haystacks) at blade passing frequency harmonics; anisotropy with a stretched streamwise length scale is required. On the other hand, the heterogeneous model captures the progressive attenuation of higher-order haystacks through destructive interference between contributions of varying eddy sizes across the propeller disk, an effect that homogeneous models systematically overpredict. (10.2514/6.2026-3517)
    DOI : 10.2514/6.2026-3517
  • Extension du domaine de validité des modèles réduits fondés sur la méthode de paramétrisation de variétés invariantes : continuation analytique automatisée de cycles limites
    • de Figueiredo Stabile André
    • Grolet Aurélien
    • Touzé Cyril
    , 2026. Extension du domaine de validité des modèles réduits fondés sur la méthode de paramétrisation de variétés invariantes : continuation analytique automatisée de cycles limites
  • Accounting for grain crushing and pore collapse for strain localization in rocks
    • Collins-Craft Nicholas Anton
    • Sulem Jean
    • Stefanou Ioannis
    • Einav Itai
    , 2026. We present a model for crushable granular rocks by embedding breakage mechanics in the Cosserat continuum. This model features a dependence on an enriched set of state variables (the elastic strains and curvatures, the density, the solid fraction, and the breakage state variable), demonstrates a dependence of the yield surface on the Lode angle, breakage and solid fraction, and evolution laws that tightly couple the competing processes. We then outline the notion of linear stability analysis and how we use this technique to obtain both the thickness and orientation of any shear bands that may form in the system. The model is then calibrated from data available in the literature on Fontainebleau sandstone and other similar granular rocks. We compare the model predictions with experimental measures of both the stress-strain response, and the width and angle of shear bands, and find good agreement with the results that have been previously reported.
  • Bubble rise dynamics in a quiescent liquid and impact on a cylinder
    • Beltran F.
    • Benguigui W.
    • Merigoux N.
    • Bonometti T.
    • Colin C.
    International Journal of Multiphase Flow, Elsevier, 2026, 199 (May), pp.105672. This study investigates the dynamics of a single bubble rising in a quiescent liquid and impacting a fixed cylinder using a resolved two-fluid approach. The resolved two-fluid approach is validated against experimental data and compared with a one-fluid approach through 2D axisymmetric and 3D simulations across a wide range of Reynolds and E &amp; ouml;tvos numbers, and density ratios. The two-fluid model accurately reproduces the bubble shape, terminal velocity, and impact dynamics, showing agreement with both experimental observations and the one-fluid approach. A detailed analysis on the impact force coefficient exerted by the bubble on the cylinder is conducted with the two-fluid approach by varying the bubble's Reynolds and E &amp; ouml;tvos numbers and the density, viscosity, and bubble-to-cylinder diameter ratios (Reb is an element of [1, 80],Eo is an element of [10, 116], rho l/rho g is an element of [25, 1000], mu l/mu g is an element of [10, 100], and db/Dc is an element of [0.5, 1.0]). This study reveals that, when varying one dimensionless number at a time, the bubble Reynolds number has the most significant influence on the impact force coefficient, followed by the bubble-to-cylinder diameter ratio and the E &amp; ouml;tv &amp; ouml;s number, while the effects of viscosity and density ratios are weaker. A correlation on the impact force coefficient (associated to the force applied by the bubble on the cylinder at impact) is proposed and may be useful for Euler-Lagrange point-particle methods. (10.1016/j.ijmultiphaseflow.2026.105672)
    DOI : 10.1016/j.ijmultiphaseflow.2026.105672
  • An augmented lagrangian XFEM formulation for stabilizing near-tip singularities in fluid-driven fractures modeling
    • Faivre Maxime
    • Martin Alexandre
    • Massin P.
    • Giot Richard
    • Golfier Fabrice
    Computer Methods in Applied Mechanics and Engineering, Elsevier, 2026, 451, pp.118649. This paper enhances a fully-coupled hydromechanical model developed in the XFEM by introducing a new augmented Lagrangian formulation to stabilize the hydrodynamical behavior at interfaces. While the original model stabilized the mechanical part, it inadequately represented pressure fields near high gradients, particularly at the tip. The new formulation projects fields into an appropriate function space, enabling the use of a standard lumping technique on the mass matrix from the discretized time derivative in Reynolds’ lubrication equation. It also incorporates a lumping approach tailored to parabolic PDEs to stabilize the bulk behavior. Since Reynolds’ equation becomes singular near the tip in impermeable porous media, a virtual opening is introduced to regularize the fluid flow. The improved model is validated with analytical solutions, demonstrating its effectiveness in eliminating spurious pressure oscillations near discontinuities. (10.1016/j.cma.2025.118649)
    DOI : 10.1016/j.cma.2025.118649
  • Steady-State Algorithm with Structural Periodicity: Application to Computation of Railways’ Ballast Plastic Strains
    • Badinier Thibault
    • Maiolino Siegfried
    • Maitournam Habibou
    Geotechnics, MDPI, 2026, 6 (1), pp.29. The geometry of ballasted railway tracks is crucial for ensuring railway safety and efficiency. This paper introduces the use of innovative steady-state algorithms designed to compute plastic strains in linear geotechnical structures like railway ballast layers, within Finite Element Methods (FEMs). Facing the specificities of moving loads, traditional step-by-step algorithms, while simple and adaptable, are computationally expensive and time-consuming. In contrast, the proposed steady-state algorithms leverage an Eulerian approach to describe the movement of loads significantly reducing computational time while maintaining accuracy. This paper proposes these algorithms as a methodological improvement and demonstrates the applicability and efficiency of the method for non-periodic structures, as well as for periodic structures, such as railway tracks with evenly spaced sleepers. This paper demonstrates the applicability and efficiency of theses algorithms through comparative studies with traditional methods on typical railway structures. The results show that the presented algorithm not only matches the accuracy of step-by-step methods but also drastically reduces computation time and data storage requirements. This advancement has practical applications for railway infrastructure managers, enabling more efficient and accurate predictions of track geometry evolution and preventing incidents through improved maintenance strategies. (10.3390/geotechnics6010029)
    DOI : 10.3390/geotechnics6010029
  • Un modèle filaire hybride poutre-coque pour la rupture de tuyauteries à haute énergie
    • Coron Amalio
    , 2026. Les entreprises du secteur de l'énergie exploitent des installations comportant de vastes réseaux de tuyauteries. La compréhension et la modélisation du comportement de ces systèmes, aussi bien en fonctionnement nominal qu'en situation accidentelle, sont cruciales. Cette thèse présente le développement d'un modèle filaire haute fidélité pour la modélisation de réseaux de tuyauteries industriels en situation accidentelle.Des modélisations filaires enrichies de tuyauterie ont été développées dans la littérature. Depuis les travaux de von Kármán en 1911, ces modèles ont suscité un intérêt particulier, notamment dans les contributions de Wood, Bathe et Almeida, Militello et Huespe, Weicker, Karamanos et Arbind. Contrairement aux modèles de poutre classique, le point commun des modèles enrichis réside dans la prise en compte des déformations des sections. Celles-ci sont décrites par une modélisation en coque cylindrique de la tuyauterie, dont la cinématique est développée en série de Fourier sur la section circulaire. La cinématique ainsi proposée est donc plus riche et plus complète qu'un modèle de poutre simple.Un tel modèle a été développé à EDF R&D lors de travaux antérieurs à cette thèse. Il combine un comportement de poutre et de coque pour décrire la déformation globale de la tuyauterie ainsi que la déformation des sections transverses. Les hypothèses d'Euler-Bernoulli sont utilisées pour le modèle poutre et les hypothèses de Kirchhoff-Love sont retenues pour la cinématique de coque. De plus, les déplacements de la coque sont développés en série de Fourier en fonction de la coordonnée angulaire de la section. Les coefficients de la série de Fourier constituent alors les inconnues cinématiques du modèle, en complément de ceux du modèle de poutre. Ce modèle de tube sain a été discrétisé par la méthode des éléments finis et implémenté dans un code de calcul en dynamique explicite.Dans cette thèse, l'élément hybride poutre-coque pour les tuyauteries saines est étendu pour l'étude des fissures progressives. La cinématique de la tuyauterie saine est donc complétée pour représenter une tuyauterie fendue. Dans cette perspective, des inconnues cinématiques supplémentaires sont incorporées au modèle pour permettre l'ouverture des sections transverses.Cette nouvelle cinématique est construite en étudiant un tube fendu sur toute sa longueur sous des chargements dans chaque direction de l'espace. Ce tube est modélisé comme un anneau fendu. La réponse mécanique de cet anneau fendu est exprimée en fonction de la coordonnée angulaire de la section de manière cohérente avec la formulation coque développée en série de Fourier. Les nouveaux termes sont ajoutés à ceux de la coque, avec les nouvelles composantes du déplacement associé. Les résultats de cette nouvelle cinématique sont ensuite analysés. Un nouvel élément fini, appelé tube fendu, a été développé sur la base de cette cinématique enrichie.Dans un premier temps, la capacité de la nouvelle cinématique à représenter le comportement d'une tuyauterie fissurée est évaluée. Les champs de déplacement d'une simulation éléments finis de coque sont projetés sur les composantes de la cinématique de tube fendu. Chaque champ et sa projection associée sont comparés.Les résultats de simulations numériques du modèle de tube fendu sont ensuite comparés avec des solutions de référence. Ces solutions sont soit analytiques, soit numériques obtenues par des simulations en éléments de coque. Les cas de comparaisons sont d'abord des cas uniformes de tuyau fendu sous plusieurs chargements (pression, force linéique). Le modèle est ensuite utilisé pour la représentation d'un tuyau qui comporte une partie saine et une partie fissurée reliées.