Search

We present a stochastic method for reconstructing missing spatial and velocity data along the trajectories of small objects passively advected by turbulent flows with a wide range of temporal or spatial scales, such as small balloons in the atmosphere or drifters in the ocean. Our approach makes use of conditional generative diffusion models, a recently proposed data-driven machine learning technique. We solve the problem for two paradigmatic open problems, the case of 3D tracers in homogeneous and isotropic turbulence, and 2D trajectories from the NOAA-funded Global Drifter Program. We show that for both cases, our method is able to reconstruct velocity signals retaining non-trivial scale-by-scale properties that are highly non-Gaussian and intermittent. A key feature of our method is its flexibility in dealing with the location and shape of data gaps, as well as its ability to naturally exploit correlations between different components, leading to superior accuracy, with respect to Gaussian process regressions, for both pointwise reconstruction and statistical expressivity. Our method shows promising applications also to a wide range of other Lagrangian problems, including multi-particle dispersion in turbulence, dynamics of charged particles in astrophysics and plasma physics, and pedestrian dynamics.

Heavy and light particles are commonly found in many natural phenomena and industrial processes, such as suspensions of bubbles, dust, and droplets in incompressible turbulent flows. Based on a recent machine learning approach using a diffusion model that successfully generated single tracer trajectories in three-dimensional turbulence and passed most statistical benchmarks across time scales, we extend this model to include heavy and light particles. Given the particle type - tracer, light, or heavy - the model can generate synthetic, realistic trajectories with correct fat-tail distributions for acceleration, anomalous power laws, and scale dependent local slope properties. This work paves the way for future exploration of the use of diffusion models to produce high-quality synthetic datasets for different flow configurations, potentially allowing interpolation between different setups and adaptation to new conditions.

We present TURB-Hel, a database formed by two datasets of incompressible homogeneous and isotropic turbulence, maintained in a statistically stationary state by fully helical forcing. The aim is to provide a dataset that clearly exhibits the phenomenon of the helicity cascade from the large to the small scales generated by a large-scale forcing that breaks the mirror symmetry. This database offers the possibility to realize a wide variety of analyses of fully developed turbulence from the sub-grid scale filtering up to the validation of an a posteriori LES. TURB-Hel is available for download using the SMART-Turb portal http://smart-turb.roma2.infn.it.

We address the problem of data augmentation in a rotating turbulence set-up, a paradigmatic challenge in geophysical applications. The goal is to reconstruct information in two-dimensional (2D) cuts of the three-dimensional flow fields, imagining to have spatial gaps present within each 2D observed slice. We evaluate the effectiveness of different data-driven tools, based on diffusion models (DMs), a state-of-the-art generative machine learning protocol, and generative adversarial networks (GANs), previously considered as the best-performing method both in terms of point-wise reconstruction and the statistical properties of the inferred velocity fields. We focus on two different DMs recently proposed in the specialized literature: (i) RePaint, based on a heuristic strategy to guide an unconditional DM for flow generation by using partial measurements data and (ii) Palette, a conditional DM trained for the reconstruction task with paired measured and missing data. Systematic comparison shows that (i) DMs outperform the GAN in terms of the mean squared error and/or the statistical accuracy; (ii) Palette DM emerges as the most promising tool in terms of both point-wise and statistical metrics. An important property of DMs is their capacity for probabilistic reconstructions, providing a range of predictions based on the same measurements, enabling for uncertainty quantification and risk assessment.

In this work we extend a previously derived $n$- $^9$Be optical potential up to 500 MeV and apply it to the system $n$- $^{12}$C, finding excellent results for the energy dependence of the total cross sections. Results obtained with a standard optical model calculation are compared to those from the eikonal formalism in order to asses the accuracy of the latter as a function of the nucleon incident energy. For comparison, single folded (s.f.) nucleon-target potentials are also obtained using $^{12}$C densities from different models. These potentials are sensitive to the density used and none of them reproduce the characteristics of the phenomenological potential nor the cross section results. We then calculate nucleus-nucleus ($NN$) potentials and total reaction cross sections for some "normal" and exotic projectile nuclei on $ ^ {12} $C within the eikonal formalism. We find that single folded (S.F.) projectile-target imaginary potentials and double folded (D.F.) potentials can produce similar energy dependence of the reaction cross sections but the S.F. results agree better with experimental data provided the radius parameter of the phenomenological $n$-target potential is allowed to be energy dependent. We conclude that the results previously obtained for a $^9$Be target are quite general, at least for light systems, and that a S.F. $NN$ potential built on a phenomenological nN potential can constitute an interesting and useful alternative to D.F. potentials., Comment: 11 pages, 8 figures

Lagrangian turbulence lies at the core of numerous applied and fundamental problems related to the physics of dispersion and mixing in engineering, bio-fluids, atmosphere, oceans, and astrophysics. Despite exceptional theoretical, numerical, and experimental efforts conducted over the past thirty years, no existing models are capable of faithfully reproducing statistical and topological properties exhibited by particle trajectories in turbulence. We propose a machine learning approach, based on a state-of-the-art diffusion model, to generate single-particle trajectories in three-dimensional turbulence at high Reynolds numbers, thereby bypassing the need for direct numerical simulations or experiments to obtain reliable Lagrangian data. Our model demonstrates the ability to reproduce most statistical benchmarks across time scales, including the fat-tail distribution for velocity increments, the anomalous power law, and the increased intermittency around the dissipative scale. Slight deviations are observed below the dissipative scale, particularly in the acceleration and flatness statistics. Surprisingly, the model exhibits strong generalizability for extreme events, producing events of higher intensity and rarity that still match the realistic statistics. This paves the way for producing synthetic high-quality datasets for pre-training various downstream applications of Lagrangian turbulence., Comment: We include Supplementary Information as a separate PDF file available in the Tex Source download area

Physiological solvent flows surround biological structures triggering therein collective motions. Notable examples are virus/host-cell interactions and solvent-mediated allosteric regulation. The present work describes a multiscale approach joining the Lattice Boltzmann fluid dynamics (for solvent flows) with the all-atom atomistic molecular dynamics (for proteins) to model functional interactions between flows and molecules. We present, as an applicative scenario, the study of the SARS-CoV-2 virus spike glycoprotein protein interacting with the surrounding solvent, modeled as a mesoscopic fluid. The equilibrium properties of the wild-type spike and of the Alpha variant in implicit solvent are described by suitable observables. The mesoscopic solvent description is critically compared to the all-atom solvent model, to quantify the advantages and limitations of the mesoscopic fluid description., Comment: 15 pages, 6 figures

We present thread-safe, highly-optimized lattice Boltzmann implementations, specifically aimed at exploiting the high memory bandwidth of GPU-based architectures. At variance with standard approaches to LB coding, the proposed strategy, based on the reconstruction of the post-collision distribution via Hermite projection, enforces data locality and avoids the onset of memory dependencies, which may arise during the propagation step, with no need to resort to more complex streaming strategies. The thread-safe lattice Boltzmann achieves peak performances, both in two and three dimensions and it allows to sensibly reduce the allocated memory ( tens of GigaBytes for order billions lattice nodes simulations) by retaining the algorithmic simplicity of standard LB computing. Our findings open attractive prospects for high-performance simulations of complex flows on GPU-based architectures.

We present TURB-Lagr, a new open database of 3d turbulent Lagrangian trajectories, obtained by Direct Numerical Simulations (DNS) of the original Navier-Stokes equations in the presence of a homogeneous and isotropic forcing. The aim is to provide the community interested in data-assimilation and/or Lagrangian-properties of turbulence a new testing-ground made of roughly 300K different Lagrangian trajectories. TURB-Lagr data are influenced by the strong non-Gaussian fluctuations characteristic of turbulence and by the rough and non differentiable fields. In addition, coming from fully resolved numerical simulations of the original partial differential equations, they offer the possibility to apply a wide range of approaches, from equation-free to physics-based models. TURB-Lagr data are reachable at http://smart-turb.roma2.infn.it

Inference problems for two-dimensional snapshots of rotating turbulent flows are studied. We perform a systematic quantitative benchmark of point-wise and statistical reconstruction capabilities of the linear Extended Proper Orthogonal Decomposition (EPOD) method, a non-linear Convolutional Neural Network (CNN) and a Generative Adversarial Network (GAN). We attack the important task of inferring one velocity component out of the measurement of a second one, and two cases are studied: (I) both components lay in the plane orthogonal to the rotation axis and (II) one of the two is parallel to the rotation axis. We show that EPOD method works well only for the former case where both components are strongly correlated, while CNN and GAN always outperform EPOD both concerning point-wise and statistical reconstructions. For case (II), when the input and output data are weakly correlated, all methods fail to reconstruct faithfully the point-wise information. In this case, only GAN is able to reconstruct the field in a statistical sense. The analysis is performed using both standard validation tools based on $L_2$ spatial distance between the prediction and the ground truth and more sophisticated multi-scale analysis using wavelet decomposition. Statistical validation is based on standard Jensen-Shannon divergence between the probability density functions, spectral properties and multi-scale flatness.

David Brink was one of the leading theoretical nuclear physicists of his generation. He made major contributions to the study of all aspects of nuclear physics embracing nuclear structure, nuclear scattering, and nuclear instability. His wide ranging interests and interactions with theorists and experimentalists alike helped him in providing both theoretical analysis and interpretations and suggesting experiments. He had the gift of visualising complex problems in simple terms and provided clear analysis of the underlying processes. He was an expert on the use of semi-classical methods which provided an intuitively clear picture of complex phenomena. His research work and books are characterised by scientific clarity, transparency, and depth. David possessed outstanding skills in mathematical computation, and he was an expert on special functions, group theory, and the Feynman path integral method. David had many research students and collaborated with a large number of scientists from across the world, for whom he was a source of scientific and human inspiration and admiration. His most fundamental belief was that research was a means of trying to discover and understand the beauties of Nature and explain them in simple terms to others. His absolute belief in the value of truth and his unselfish and generous attitude in sharing knowledge makes him an outstanding figure in contemporary Nuclear Physics., Comment: A slightly shortened version of this article was published in the Biographical Memoirs of Fellows of the Royal Society,Volume 73

A GPU-accelerated version of the lattice Boltzmann method for efficient simulation of soft materials is introduced. Unlike standard approaches, this method reconstructs the distribution functions from available hydrodynamic variables (density, momentum, and pressure tensor) without storing the full set of discrete populations. This scheme shows satisfactory numerical stability, significantly lower memory requirements, and data access cost. A series of benchmark tests of relevance to soft matter, such as collisions of fluid droplets, is discussed to validate the method. The results can be of particular interest for high-performance simulations of soft matter systems on future exascale computers., Comment: 12 pages, 7 figures

Data reconstruction of rotating turbulent snapshots is investigated utilizing data-driven tools. This problem is crucial for numerous geophysical applications and fundamental aspects, given the concurrent effects of direct and inverse energy cascades, which lead to non-Gaussian statistics at both large and small scales. Data assimilation also serves as a tool to rank physical features within turbulence, by evaluating the performance of reconstruction in terms of the quality and quantity of the information used. Additionally, benchmarking various reconstruction techniques is essential to assess the trade-off between quantitative supremacy, implementation complexity, and explicability. In this study, we use linear and non-linear tools based on the Proper Orthogonal Decomposition (POD) and Generative Adversarial Network (GAN) for reconstructing rotating turbulence snapshots with spatial damages (inpainting). We focus on accurately reproducing both statistical properties and instantaneous velocity fields. Different gap sizes and gap geometries are investigated in order to assess the importance of coherency and multi-scale properties of the missing information. Surprisingly enough, concerning point-wise reconstruction, the non-linear GAN does not outperform one of the linear POD techniques. On the other hand, supremacy of the GAN approach is shown when the statistical multi-scale properties are compared. Similarly, extreme events in the gap region are better predicted when using GAN. The balance between point-wise error and statistical properties is controlled by the adversarial ratio, which determines the relative importance of the generator and the discriminator in the GAN training. Robustness against the measurement noise is also discussed.

Deep neural networks are rapidly emerging as data analysis tools, often outperforming the conventional techniques used in complex microfluidic systems. One fundamental analysis frequently desired in microfluidic experiments is counting and tracking the droplets. Specifically, droplet tracking in dense emulsions is challenging as droplets move in tightly packed configurations. Sometimes the individual droplets in these dense clusters are hard to resolve, even for a human observer. Here, two deep learning-based cutting-edge algorithms for object detection (YOLO) and object tracking (DeepSORT) are combined into a single image analysis tool, DropTrack, to track droplets in microfluidic experiments. DropTrack analyzes input videos, extracts droplets' trajectories, and infers other observables of interest, such as droplet numbers. Training an object detector network for droplet recognition with manually annotated images is a labor-intensive task and a persistent bottleneck. This work partly resolves this problem by training object detector networks (YOLOv5) with hybrid datasets containing real and synthetic images. We present an analysis of a double emulsion experiment as a case study to measure DropTrack's performance. For our test case, the YOLO networks trained with 60% synthetic images show similar performance in droplet counting as with the one trained using 100% real images, meanwhile saving the image annotation work by 60%. DropTrack's performance is measured in terms of mean average precision (mAP), mean square error in counting the droplets, and inference speed. The fastest configuration of DropTrack runs inference at about 30 frames per second, well within the standards for real-time image analysis., Comment: 24 pages, 7 figures, and 2 video files

We report new dynamical modes in confined soft granular flows, such as stochastic jetting and dripping, with no counterpart in continuum viscous fluids. The new modes emerge as a result of the propagation of the chaotic behaviour of individual grains -- here, monodisperse emulsion droplets to the level of the entire system as the emulsion is focused into a narrow orifice by an external viscous flow. We observe avalanching dynamics and the formation of remarkably stable jets -- singlefile granular chains -- which occasionally break, resulting in a non-Gaussian distribution of cluster sizes. We find that the sequences of droplet rearrangements that lead to the formation of such chains resemble unfolding of cancer cell clusters in narrow capillaries, overall demonstrating that microfluidic emulsion systems could serve to model various aspects of soft granular flows, including also tissue dynamics at the meso-scale., Comment: 9 pages, 6 figures

We investigate the dynamics of a phase-separating binary fluid, containing colloidal dumbbells anchored to the fluid-fluid interface. Extensive Lattice Boltzmann-Immersed Boundary method simulations reveal that the presence of soft dumbbells can significantly affect the curvature dynamics of the interface between phase-separating fluids, even though the coarsening dynamics is left nearly unchanged. In addition, our results show that the curvature dynamics exhibits distinct non-local effects, which might be exploited for the design of new soft mesoscale materials. We point out that the inspection of the statistical dynamics of the curvature can disclose new insights into local inhomogeneities of the binary fluid configuration, as a function of the volume fraction and aspect ratio of the dumbbells., Comment: 15 pages, 14 figures

Multiple emulsions are a class of soft fluid in which small drops are immersed within a larger one and stabilized over long periods of time by a surfactant. We recently showed that, if a monodisperse multiple emulsion is subject to a pressure-driven flow, a wide variety of nonequilibrium steady states emerges at late times, whose dynamics relies on a complex interplay between hydrodynamic interactions and multibody collisions among internal drops. In this work, we use lattice Boltzmann simulations to study the dynamics of polydisperse double emulsions driven by a Poiseuille flow within a microfluidic channel. Our results show that their behavior is critically affected by multiple factors, such as initial position, polydispersity index, and area fraction occupied within the emulsion. While at low area fraction inner drops may exhibit either a periodic rotational motion (at low polydispersity) or arrange into nonmotile configurations (at high polydispersity) located far from each other, at larger values of area fraction they remain in tight contact and move unidirectionally. This decisively conditions their close-range dynamics, quantitatively assessed through a time-efficiency-like factor. Simulations also unveil the key role played by the capsule, whose shape changes can favor the formation of a selected number of nonequilibrium states in which both motile and nonmotile configurations are found., Comment: Post peer-review version. 11 pages, 6 figures. Movies upon request

The problem of classifying turbulent environments from partial observation is key for some theoretical and applied fields, from engineering to earth observation and astrophysics, e.g. to precondition searching of optimal control policies in different turbulent backgrounds, to predict the probability of rare events and/or to infer physical parameters labelling different turbulent set-ups. To achieve such goal one can use different tools depending on the system's knowledge and on the quality and quantity of the accessible data. In this context, we assume to work in a model-free setup completely blind to all dynamical laws, but with a large quantity of (good quality) data for training. As a prototype of complex flows with different attractors, and different multi-scale statistical properties we selected 10 turbulent 'ensembles' by changing the rotation frequency of the frame of reference of the 3d domain and we suppose to have access to a set of partial observations limited to the instantaneous kinetic energy distribution in a 2d plane, as it is often the case in geophysics and astrophysics. We compare results obtained by a Machine Learning (ML) approach consisting of a state-of-the-art Deep Convolutional Neural Network (DCNN) against Bayesian inference which exploits the information on velocity and enstrophy moments. First, we discuss the supremacy of the ML approach, presenting also results at changing the number of training data and of the hyper-parameters. Second, we present an ablation study on the input data aimed to perform a ranking on the importance of the flow features used by the DCNN, helping to identify the main physical contents used by the classifier. Finally, we discuss the main limitations of such data-driven methods and potential interesting applications.

In this brief review we discuss the basic theoretical concepts used in the experimental studies of the most common cases of direct reactions such as (a) elastic scattering, (b) inelastic scattering, (c) Coulomb excitation, (d) transfer reactions and (e) breakup reactions., Comment: Contribution to the "Handbook of Nuclear Physics", Springer, 2022, edited by I. Tanihata, H. Toki and T. Kajino

We present LBcuda, a GPU accelerated version of LBsoft, our open-source MPI-based software for the simulation of multi-component colloidal flows. We describe the design principles, the optimization and the resulting performance as compared to the CPU version, using both an average cost GPU and high-end NVidia GPU cards (V100 and the latest A100). The results show a substantial acceleration for the fluid solver reaching up to 200 GLUPS (Giga Lattice Updates Per Second) on a cluster made of 512 A100 NVIDIA cards simulating a grid of eight billion lattice points. These results open attractive prospects for the computational design of new materials based on colloidal particles., Comment: 27 pages, 5 figures

Two-dimensional (2D) transition-metal monochalcogenides have been recently predicted to be potential photo(electro)catalysts for water splitting and photoelectrochemical (PEC) reactions. Differently from the most established InSe, GaSe, GeSe, and many other monochalcogenides, bulk GaS has a large band gap of ca. 2.5 eV, which increases up to more than 3.0 eV with decreasing its thickness due to quantum confinement effects. Therefore, 2D GaS fills the void between 2D small-band-gap semiconductors and insulators, resulting of interest for the realization of van der Waals type-I heterojunctions in photocatalysis, as well as the development of UV light-emitting diodes, quantum wells, and other optoelectronic devices. Based on theoretical calculations of the electronic structure of GaS as a function of layer number reported in the literature, we experimentally demonstrate, for the first time, the PEC properties of liquid-phase exfoliated GaS nanoflakes. Our results indicate that solution-processed 2D GaS-based PEC-type photodetectors outperform the corresponding solid-state photodetectors. In fact, the 2D morphology of the GaS flakes intrinsically minimizes the distance between the photogenerated charges and the surface area at which the redox reactions occur, limiting electron-hole recombination losses. The latter are instead deleterious for standard solid-state configurations. Consequently, PEC-type 2D GaS photodetectors display a relevant UV-selective photoresponse. In particular, they attain responsivities of 1.8 mA W-1 in 1 M H2SO4 (at 0.8 V vs reversible hydrogen electrode -RHE-), 4.6 mA W-1 in 1 M Na2SO4 (at 0.9 V vs RHE), and 6.8 mA W-1 in 1 M KOH (at 1.1 V vs RHE) under 275 nm illumination wavelength with an intensity of 1.3 mW cm-2.

The development of high-power density vanadium redox flow batteries (VRFBs) with high energy efficiencies (EEs) is crucial for the widespread dissemination of this energy storage technology. In this work, we report the production of novel hierarchical carbonaceous nanomaterials for VRFB electrodes with high catalytic activity toward the vanadium redox reactions (VO2+/VO2+ and V2+/V3+). The electrode materials are produced through a rapid (minute timescale) low-pressure combined gas plasma treatment of graphite felts (GFs) in an inductively coupled radio frequency reactor. By systematically studying the effects of either pure gases (O2 and N2) or their combination at different gas plasma pressures, the electrodes are optimized to reduce their kinetic polarization for the VRFB redox reactions. To further enhance the catalytic surface area of the electrodes, single-/few-layer graphene, produced by highly scalable wet-jet milling exfoliation of graphite, is incorporated into the GFs through an infiltration method in the presence of a polymeric binder. Depending on the thickness of the proton-exchange membrane (Nafion 115 or Nafion XL), our optimized VRFB configurations can efficiently operate within a wide range of charge/discharge current densities, exhibiting energy efficiencies up to 93.9%, 90.8%, 88.3%, 85.6%, 77.6%, and 69.5% at 25, 50, 75, 100, 200, and 300 mA cm-2, respectively. Our technology is cost-competitive when compared to commercial ones (additional electrode costs < 100 euro m-2) and shows EEs rivalling the record-high values reported for efficient systems to date. Our work remarks on the importance to study modified plasma conditions or plasma methods alternative to those reported previously (e.g., atmospheric plasmas) to improve further the electrode performances of the current VRFB systems.

In the ever-increasing energy demand scenario, the development of novel photovoltaic (PV) technologies is considered to be one of the key solutions to fulfil the energy request. In this context, graphene and related two-dimensional (2D) materials (GRMs), including nonlayered 2D materials and 2D perovskites, as well as their hybrid systems, are emerging as promising candidates to drive innovation in PV technologies. The mechanical, thermal, and optoelectronic properties of GRMs can be exploited in different active components of solar cells to design next-generation devices. These components include front (transparent) and back conductive electrodes, charge transporting layers, and interconnecting-recombination layers, as well as photoactive layers. The production and processing of GRMs in the liquid phase, coupled with the ability to on-demand tune their optoelectronic properties exploiting wet-chemical functionalization, enable their effective integration in advanced PV devices through scalable, reliable, and inexpensive printing-coating processes. Herein, we review the progresses in the use of solution-processed 2D materials in organic solar cells, dye-sensitized solar cells, perovskite solar cells, quantum dot solar cells, and organic-inorganic hybrid solar cells, as well as in tandem systems. We first provide a brief introduction on the properties of 2D materials and their production methods by solution-processing routes. Then, we discuss the functionality of 2D materials for electrodes, photoactive layer components-additives, charge transporting layers, and interconnecting layers through figures of merit, which allow the performance of solar cells to be determined and compared with the state-of-the-art values. We finally outline the roadmap for the further exploitation of solution-processed 2D materials to boost the performance of PV devices., Comment: Chemical Society Reviews, 2021, Advance Article

Harvesting all sources of available clean energy is an essential strategy to contribute to healing current dependence on non-sustainable energy sources. Recently, triboelectric nanogenerators (TENGs) have gained visibility as new mechanical energy harvester offering a valid alternative to batteries, being particularly suitable for portable devices. Here, the increased capacitance of a few-layer graphene-based electrode is obtained by incorporating nitrogen-doped graphene (N_graphene), enabling a 3_fold enhancement in TENGs power output. The dependence of TENGs performance on the electronic properties of different N_graphene types, varying in the doping concentration and in the relative content of N-pyridinic and N-graphitic sites is investigated. These sites have different electron affinities, and synergistically contribute to the variation of the capacitive and resistive properties of N-graphene and consequently, TENG performance. It is demonstrated that the power enhancement of the TENG occurs when the N_graphene, an n-semiconductor, is interfaced between the positive triboelectric material and the electrode, while a deterioration of the electrical performance is observed when it is placed at the interface with the negative triboelectric material. This behavior is explained in terms of the dependence of N_graphene quantum capacitance on the electrode chemical potential which shifts according to the opposite polarization induced at the two electrodes upon triboelectrification.

The state-of-the-art deep learning-based object recognition YOLO algorithm and object tracking DeepSORT algorithm are combined to analyze digital images from fluid dynamic simulations of multi-core emulsions and soft flowing crystals and to track moving droplets within these complex flows. The YOLO network was trained to recognize the droplets with synthetically prepared data, thereby bypassing the labor-intensive data acquisition process. In both applications, the trained YOLO + DeepSORT procedure performs with high accuracy on the real data from the fluid simulations, with low error levels in the inferred trajectories of the droplets and independently computed ground truth. Moreover, using commonly used desktop GPUs, the developed application is capable of analyzing data at speeds that exceed the typical image acquisition rates of digital cameras (30 fps), opening the interesting prospect of realizing a low-cost and practical tool to study systems with many moving objects, mostly but not exclusively, biological ones. Besides its practical applications, the procedure presented here marks the first step towards the automatic extraction of effective equations of motion of many-body soft flowing systems., Comment: 31 pages, 15 figures