In road environments, real-time knowledge of local weather conditions is an essential prerequisite for addressing the twin challenges of enhancing road safety and avoiding congestions. Currently, the main means of quantifying weather conditions along a road network requires the installation of meteorological stations. Such stations are costly and must be maintained; however, large numbers of cameras are already installed on the roadside. A new artificial intelligence method that uses road traffic cameras and a convolution neural network to detect weather conditions has, therefore, been proposed. It addresses a clearly defined set of constraints relating to the ability to operate in real-time and to classify the full spectrum of meteorological conditions and order them according to their intensity. The method can differentiate between five weather conditions such as normal (no precipitation), heavy rain, light rain, heavy fog and light fog. The deep-learning method’s training and testing phases were conducted using a new database called the Cerema-AWH (Adverse Weather Highway) database. After several optimisation steps, the proposed method obtained an accuracy of 0.99 for classification.

The amount and characteristics of plastic waste in the environment are closely monitored with the regular expansion of their collection to smaller debris sizes and new compartments, such as atmosphere, soil and even food. These data, subject to constant re-evaluation, are necessary to assess the threat that plastic particles pose. In this article, we take stock of the most recent measurements of their size distribution on the ocean surface. We confront the observed size distribution to the commonly invoked degradation modes of plastics and propose a scenario that accounts for the quite unusual features of the size distribution of floating debris. We conclude that delamination alone could be a major mechanism explaining the large number of small microplastics (<150 μm) in the ocean.

Rubber-based nanocomposites prepared by solid-phase mixing with precipitated silica nanoparticles are typically strongly aggregated systems with different levels of spatial organization, as highlighted by our group in the past [1]. The strategy that we developed these past years was to investigate such systems based on the study of simplified industrial samples with ingredients limited to a strict minimum. The analysis of small-angle X-ray scattering data can then be performed on the scale of a micrometric simulation box. Tens of thousands of “model” nanoparticles are embedded in the matrix, and their dispersion strongly affects both the mechanical properties of the material, and the scattered intensity. A statistical method based on a reverse Monte Carlo solution of this many-parameter scattering problem will be presented, showing that some key features like percolation can be described [2]. Another key feature of rubber nanocomposites refers to the influence of the filler surfaces on the polymer structure and dynamics, and some recent progress will be discussed [3]. In particular, we have studied blends of short and long chains, where one chain type is deuterated, by small-angle neutron scattering. Different degrees of spatial segregation could be identified recently, including a peculiar, “fish-shaped” interfacial gradient characterized also by reverse Monte Carlo simulations, this time of the interface. [1] Guilhem P. Baeza, Anne-Caroline Genix, C. Degrandcourt, Laurent Petitjean, Jérémie Gummel, Marc Couty, Julian Oberdisse, Macromolecules 2013, 46, 317 (cover article) [2] Musino D, Genix A-C, Chauveau E, Bizien T, Oberdisse J, Nanoscale 2020, 12:3907 [3] A.C. Genix, V. Bocharova, B. Carroll, P. Dieudonné-George, M. Sztucki, R. Schweins, A. P. Sokolov, and Julian Oberdisse, ACS Applied Materials and Interfaces, 2021, 13, 36262

We use the impact of drops on a small solid target as a tool to investigate the behavior of viscoelastic fluids under extreme deformation rates. We study two classes of transient networks: semidilute solutions of supramolecular polymers and suspensions of spherical oil droplets reversibly linked by polymers. The two types of samples display very similar linear viscoelastic properties, which can be described with a Maxwell fluid model, but contrasting nonlinear properties due to different network structure. Upon impact, weakly viscoelastic samples exhibit a behavior qualitatively similar to that of Newtonian fluids: A smooth and regular sheet forms, expands, and then retracts. By contrast, for highly viscoelastic fluids, the thickness of the sheet is found to be very irregular, leading to instabilities and eventually formation of holes. We find that material rheological properties rule the onset of instabilities. We first provide a simple image analysis of the expanding sheets to determine the onset of instabilities. We then demonstrate that a Deborah number related to the shortest relaxation time associated to the sample structure following a high shear is the relevant parameter that controls the heterogeneities in the thickness of the sheet, eventually leading to the formation of holes. When the sheet tears-up, data suggest by contrast that the opening dynamics depends also on the expansion rate of the sheet.

Mucus is a biological gel protecting several tissues. Its key properties result from a crucial balance between solid-like and fluid-like behavior, ensured by the nonpermanent nature of the bonds between its macromolecular constituents. Our understanding of the micrometer-scale response of mucus to an applied stress is still rudimentary, although in living organisms, stresses acting on mucus are ubiquitous, from bacterial penetration to coughing and peristalsis. We show that under a modest applied stress, in the mechanical linear regime, the microscopic dynamics of pig gastric mucus transiently accelerate by up to 2 orders of magnitude. A simple model rationalizes this previously unrecognized fluidization mechanism stemming from elastic recoil following bond breaking and generalizes our findings to networks with reversible bonds.Mucus is a biological gel covering the surface of several tissues and ensuring key biological functions, including as a protective barrier against dehydration, pathogen penetration, or gastric acids. Mucus biological functioning requires a finely tuned balance between solid-like and fluid-like mechanical response, ensured by reversible bonds between mucins, the glycoproteins that form the gel. In living organisms, mucus is subject to various kinds of mechanical stresses, e.g., due to osmosis, bacterial penetration, coughing, and gastric peristalsis. However, our knowledge of the effects of stress on mucus is still rudimentary and mostly limited to macroscopic rheological measurements, with no insight into the relevant microscopic mechanisms. Here, we run mechanical tests simultaneously to measurements of the microscopic dynamics of pig gastric mucus. Strikingly, we find that a modest shear stress, within the macroscopic rheological linear regime, dramatically enhances mucus reorganization at the microscopic level, as signaled by a transient acceleration of the microscopic dynamics, by up to 2 orders of magnitude. We rationalize these findings by proposing a simple, yet general, model for the dynamics of physical gels under strain and validate its assumptions through numerical simulations of spring networks. These results shed light on the rearrangement dynamics of mucus at the microscopic scale, with potential implications in phenomena ranging from mucus clearance to bacterial and drug penetration.ASCII and Excel files for all the datasets shown in the figures of the main text and SI Appendix have been deposited in Zenodo (DOI: 10.5281/zenodo.5533877) (74).

La compréhension des phénomènes de déformation de gouttes après impact constitue un enjeu considérable pour des secteurs industriels tels que l'impression par jet d'encre, l'application de pesticides, le revêtement par pulvérisation, etc. Poussés par ces motivations, nous étudions les déformations subies par des gouttes de fluides Newtoniens, rhéo-fluidifiants, et de gels transitoires après impact, et de billes élastiques molles dans une autre configuration expérimentale.Dans un premier temps, les gouttes sont impactées sur une surface couverte d'une fine couche d'azote liquide qui en s'évaporant fournit les conditions idéales pour étudier la déformation de nappes liquides grâce à un effet Leidenfrost froid. Dans ces conditions, deux régimes sont identifiés : un régime capillaire pour lequel le diamètre maximal de la nappe ne dépend pas de la viscosité du fluide et un régime visqueux pour lequel celui-ci décroit lorsque la viscosité augmente. Cette décroissance est due à la dissipation visqueuse lors de l'extension de la nappe et est contrôlée par la viscosité extensionnelle biaxiale. Pour des solutions polymériques, nous montrons que cette viscosité prend en compte le caractère rhéo-fluidifiant du fluide. Dans ces conditions, nous avons impacté des fluides plus complexes qui se comportent comme des fluides de Maxwell. Un modèle d’oscillateur harmonique amorti non linéaire prédit le diamètre maximal de la nappe, d_max, ainsi que le temps, t_max, nécessaire pour atteindre d_max. Un bon accord quantitatif entre expérience et théorie est obtenu pour des fluides Newtoniens jusqu'à des viscosités élevées et des fluides de Maxwell dont le temps de relaxation caractéristique est plus grand, plus petit ou comparable au temps caractéristique de l'expérience, t_max.Dans une deuxième partie, les gouttes sont impactées sur de petites surfaces solides, appelées cibles, complexifiant le champ d'écoulement auquel les gouttes sont soumises. En effet, la portion de la nappe se trouvant en contact avec la cible est soumise à un cisaillement qui engendre une dissipation visqueuse de cisaillement et la portion s'étendant hors de la cible est uniquement sujette à la dissipation extensionnelle biaxiale. Nous prédisons l'expansion maximale de la nappe en fonction de la viscosité pertinente identifiée en évaluant quantitativement les dissipations dues au cisaillement et à l'extension biaxiale. Nous montrons, finalement, une corrélation entre le diamètre maximal de la nappe et le diamètre de la cible. A nouveau, l'importance de considérer le caractère rhéo-fluidifiant des solutions de polymères, pour rationaliser les résultats expérimentaux, est démontrée.Nous nous intéressons ensuite à l’émergence et à la dynamique du bourrelet entourant une nappe visqueuse en expansion. Nous présentons des mesures directes du champ d'épaisseur de la nappe qui permettent de mesurer les évolutions temporelles de la forme et du volume du bourrelet. Un modèle analytique est développé pour prédire la vitesse et le taux de remplissage du bourrelet. Nous identifions et comparons les importances relatives des contributions de la tension de surface, des forces d'inertie et de la viscosité à la vitesse de remplissage. Nous montrons que la contribution des forces d'inertie ne peut être négligée. Un très bon accord est trouvé entre valeurs théoriques et expérimentales.Finalement, nous utilisons un tensiomètre à bille tournante pour étudier la déformation biaxiale de billes élastiques molles. Ce dispositif permet de déformer une bille élastique immergée dans un fluide plus dense par la mise en rotation autour de son axe d'un capillaire qui les contient. Avec l'hypothèse d'une déformation homogène de la bille, nous parvenons à découpler théoriquement la mesure du module élastique de celle de la tension de surface. Ces mesures sont confrontées avec succès aux mesures expérimentales.