arXiv subjects

Marco Laudato

Publications and source records attributed to Marco Laudato.

12 recordsLinked to original sources

arXiv2025

Neural-Network Closures for Complex-Shaped Particles in the Force-Coupling Method

A data-driven surrogate framework to accelerate particle-resolved modelling of quasi-dilute suspensions of rigid, non-spherical particles in Stokes flow is introduced. A regularized-Stokeslet boundary element method (BEM) is implemented to compute hydrodynamic responses in canonical linear flows, focusing on the particle stresslet and angular velocity for spheroids, and additionally the chiral thrust for helicoidal particles. For spheroids, the BEM solver is validated against available analytical benchmarks (Faxen-type relations for the stresslet and Jeffery's theory for rotation), and parameter choices for surface discretization and regularization are selected through systematic convergence studies. For helicoidal particles, where no analytical solutions exist, accuracy is quantified via Richardson-style self-convergence, complemented by tests of linearity, frame objectivity, and chirality-dependent symmetries. The resulting datasets are used to train a neural-operator surrogate that maps local flow descriptors and particle configuration to the corresponding stresslet, rotation, and thrust at negligible evaluation cost. Across independent test sets spanning random orientations and flow types, the surrogate achieves median relative errors below 1% for the deviatoric stresslet (95th percentile below 3%) and comparable accuracy for angular velocity and thrust. The combination of validated BEM generation and fast inference provides a practical route to coupling complex particle shapes into mesoscale solvers such as the force-coupling method, enabling large-ensemble studies of microstructure and suspension rheology.

physics.flu-dyn↗

arXiv2025

A Neural-Operator Surrogate for Platelet Deformation Across Capillary Numbers

Reliable multiscale models of thrombosis require platelet-scale fidelity at organ-scale cost, a gap that scientific machine learning has the potential to narrow. We train a DeepONet surrogate on platelet dynamics generated with LAMMPS for platelets spanning ten elastic moduli and capillary numbers (0.07 - 0.77). The network takes in input the wall shear stress, bond stiffness, time, and initial particle coordinates and returns the full three-dimensional deformation of the membrane. Mean-squared-error minimization with Adam and adaptive learning-rate decay yields a median displacement error below 1%, a 90th percentile below 3%, and a worst case below 4% over the entire calibrated range while accelerating computation by four to five orders of magnitude. Leave-extremes-out retraining shows graceful extrapolation: the held-out stiffest and most compliant platelets retain sub-3% median error and an 8% maximum. Error peaks coincide with transient membrane self-contact, suggesting improvements via graph neural trunks and physics-informed torque regularization. These results classify the surrogate as high-fidelity and position it for seamless coupling with continuum CFD, enabling platelet-resolved hemodynamic simulations in patient-specific geometries and opening new avenues for predictive thrombosis modeling.

physics.flu-dyn↗

arXiv2025

Neural Operator Modeling of Platelet Geometry and Stress in Shear Flow

Thrombosis involves processes spanning large-scale fluid flow to sub-cellular events such as platelet activation. Traditional CFD approaches often treat blood as a continuum, which can limit their ability to capture these microscale phenomena. In this paper, we introduce a neural operator-based surrogate model to bridge this gap. Our approach employs DeepONet, trained on high-fidelity particle dynamics simulations performed in LAMMPS under a single shear flow condition. The model predicts both platelet membrane deformation and accumulated stress over time, achieving a mode error of ~0.3% under larger spatial filtering radii. At finer scales, the error increases, suggesting a correlation between the DeepONet architecture's capacity and the spatial resolution it can accurately learn. These findings highlight the importance of refining the trunk network to capture localized discontinuities in stress data. Potential strategies include using deeper trunk nets or alternative architectures optimized for graph-structured meshes, further improving accuracy for high-frequency features. Overall, the results demonstrate the promise of neural operator-based surrogates for multi-scale platelet modeling. By reducing computational overhead while preserving accuracy, our framework can serve as a critical component in future simulations of thrombosis and other micro-macro fluid-structure problems.

physics.flu-dyn↗

arXiv2025

Sparse wavefield reconstruction and denoising with boostlets

Boostlets are spatiotemporal functions that decompose nondispersive wavefields into a collection of localized waveforms parametrized by dilations, hyperbolic rotations, and translations. We study the sparsity properties of boostlets and find that the resulting decompositions are significantly sparser than those of other state-of-the-art representation systems, such as wavelets and shearlets. This translates into improved denoising performance when hard-thresholding the boostlet coefficients. The results suggest that boostlets offer a natural framework for sparsely decomposing wavefields in unified space-time.

eess.AS↗

arXiv2024

High-Fidelity Description of Platelet Deformation Using a Neural Operator

The goal of this work is to investigate the capability of a neural operator (DeepONet) to accurately capture the complex deformation of a platelet's membrane under shear flow. The surrogate model approximated by the neural operator predicts the deformed membrane configuration based on its initial configuration and the shear stress exerted by the blood flow. The training dataset is derived from particle dynamics simulations implemented in LAMMPS. The neural operator captures the dynamics of the membrane particles with a mode error distribution of approximately 0.5\%. The proposed implementation serves as a scalable approach to integrate sub-platelet dynamics into multi-scale computational models of thrombosis.

physics.comp-ph↗

arXiv2024

A boostlet transform for wave-based acoustic signal processing in space-time

Sparse representation systems that encode signal architecture have had a profound impact on sampling and compression paradigms. Remarkable examples are multi-scale directional systems, which, similar to our vision system, encode the underlying architecture of natural images with sparse features. Inspired by this philosophy, we introduce a representation system for wave-based acoustic signal processing in 2D space--time, referred to as the \emph{boostlet transform}, which encodes sparse features of natural acoustic fields using the Poincar\'e group and isotropic dilations. Boostlets are spatiotemporal functions parametrized with dilations, Lorentz boosts, and translations in space--time. Physically speaking, boostlets are supported away from the acoustic radiation cone, i.e., having broadband frequency with phase velocities other than the speed of sound, resulting in a peculiar scaling function. We formulate a discrete boostlet frame using Meyer wavelets and bump functions and examine its sparsity properties. An analysis with experimentally measured fields indicates that discrete boostlet coefficients decay significantly faster and attain superior reconstruction performance than wavelets, curvelets, shearlets, and wave atoms. The results demonstrate that boostlets provide a natural, compact representation system for acoustic waves in space-time.

physics.flu-dyn↗

arXiv2024

Sound generation mechanisms in a collapsible tub

Collapsible tubes can be employed to study the sound generation mechanism in the human respiratory system. The goals of this work are (a) to determine the airflow characteristics connected to three different collapse states of a physiological tube and (b) to find a relation between the sound power radiated by the tube and its collapse state. The methodology is based on the implementation of computational fluid dynamics simulation on experimentally validated geometries. The flow is characterized by a radical change of behavior before and after the contact of the lumen. The maximum of the sound power radiated corresponds to the post-buckling configuration. The idea of an acoustic tube law is proposed. The presented results are relevant to the study of self-excited oscillations and wheezing sounds in the lungs.

physics.flu-dyn↗

arXiv2020

Covariant Momentum Map Thermodynamics for Parametrized Field Theories

A general-covariant statistical framework capable of describing classical fluctuations of the gravitational field is a thorny open problem in theoretical physics, yet ultimately necessary to understand the nature of the gravitational interaction and a key to quantum gravity. Inspired by Souriau's symplectic generalization of the Maxwell-Boltzmann-Gibbs equilibrium in Lie group thermodynamics, we investigate a spacetime-covariant formulation of statistical mechanics for parametrized first-order field theories, as a simplified model sharing essential general covariant features with canonical general relativity. Starting from a covariant multi-symplectic phase space formulation, we define a general-covariant notion of Gibbs state in terms of the covariant momentum map associated with the lifted action of the diffeomorphisms group on the extended phase space. We show how such a covariant notion of equilibrium encodes the whole information about symmetry, gauge and dynamics carried by the theory, associated to a canonical spacetime foliation, where the covariant choice of a reference frame reflects in a Lie algebra-valued notion of local temperature. We investigate how physical equilibrium, hence time evolution, emerges from such a state and the role of the gauge symmetry in the thermodynamic description.

math-ph↗

arXiv2018

A Pedagogical Intrinsic Approach to Relative Entropies as Potential Functions of Quantum Metrics: the $q$-$z$ Family

The so-called $q$-z-\textit{Rényi Relative Entropies} provide a huge two-parameter family of relative entropies which includes almost all well-known examples of quantum relative entropies for suitable values of the parameters. In this paper we consider a log-regularized version of this family and use it as a family of potential functions to generate covariant $(0,2)$ symmetric tensors on the space of invertible quantum states in finite dimensions. The geometric formalism developed here allows us to obtain the explicit expressions of such tensor fields in terms of a basis of globally defined differential forms on a suitable unfolding space without the need to introduce a specific set of coordinates. To make the reader acquainted with the intrinsic formalism introduced, we first perform the computation for the qubit case, and then, we extend the computation of the metric-like tensors to a generic $n$-level system. By suitably varying the parameters $q$ and $z$, we are able to recover well-known examples of quantum metric tensors that, in our treatment, appear written in terms of globally defined geometrical objects that do not depend on the coordinates system used. In particular, we obtain a coordinate-free expression for the von Neumann-Umegaki metric, for the Bures metric and for the Wigner-Yanase metric in the arbitrary $n$-level case.

quant-ph↗

arXiv2017

Tomographic Reconstruction of Quantum Metrics

In the framework of quantum information geometry we investigate the relationship between monotone metric tensors uniquely defined on the space of quantum tomograms, once the tomographic scheme chosen, and monotone quantum metrics on the space of quantum states, classified by operator monotone functions, according to Petz classification theorem. We show that different metrics can be related through a change of the tomographic map and prove that there exists a bijective relation between monotone quantum metrics associated with different operator monotone functions. Such bijective relation is uniquely defined in terms of solutions of a first order second degree differential equation for the parameters of the involved tomographic maps. We first exhibit an example of a non-linear tomographic map which connects a monotone metric with a new one which is not monotone. Then we provide a second example where two monotone metrics are uniquely related through their tomographic parameters.

math-ph↗

arXiv2017

Dynamical Vector Fields on the Manifold of Quantum States

In this paper we shall consider the stratified manifold of quantum states and the vector fields which act on it. In particular, we show that the infinitesimal generator of the GKLS evolution is composed of a generator of unitary transformations plus a gradient vector field along with a Kraus vector field transversal to the strata defined by the involutive distribution generated by the former ones.

quant-ph↗

arXiv2017

Differential Calculus on Manifolds with a Boundary. Applications

This paper contains a set of lecture notes on manifolds with boundary and corners, with particular attention to the space of quantum states. A geometrically inspired way of dealing with these kind of manifolds is presented,and explicit examples are given in order to clearly illustrate the main ideas.

math-ph↗