Université Gustave Eiffel
  • Champs-sur-Marne, France
Recent publications
In this paper, stochastic planar stick–slip motions are investigated using a slider-on-belt model where the coefficient of friction (COF) of the contact interface is modelled as a random field. New three-variable stick–slip transition criteria are proposed to improve the accuracy and robustness of the algorithm. Stochastic analyses are performed concerning the Peak-to-Valley value of the displacement and friction forces and the time duration of the stick state. It is found that the correlation length of the COF random field is dominantly responsible for the stochastic behaviours of the system. In contrast, the belt velocity and the mean value of the COF have a significant influence on the time duration of the stick state compared with the corresponding deterministic slider-on-belt model.
Friction dampers are classically used in turbomachinery for bladed discs to control the levels of vibrations at resonance and limit the risk of fatigue failure. It consists of small metal components located under the platforms of the blades, which dissipate the vibratory energy through friction when a relative displacement between the blades and the damper appears. It is well known that the shape of such component has a strong influence on the damping properties and should be designed with a particular attention. With the arrival of additive manufacturing, new dedicated shapes for these dampers can be considered, determined with specific numerical methods as topological optimisation (TO). However, the presence of the contact nonlinearity challenges the use of traditional TO methods to minimise the vibration levels at resonance. In this work, the topology of the damper is parametrised with the moving morphable components (MMC) framework and optimised based on meta-modelling techniques: here kriging coupled with the efficient global optimisation (EGO) algorithm. The level of vibration at resonance is computed based on the harmonic balance method augmented with a constraint to aim directly for the resonant solution. It corresponds to the objective function to be minimised. Additionally, a mechanical constraint based on static stress analysis is also considered to propose reliable damper designs. Results demonstrate the efficiency of the method and show that damper geometries that meet the engineers’ requirements can be identified.
The removal of the non-steroidal anti-inflammatory drug (NSAID) Naproxen (NAX) in water by hydroxyl radicals (•OH) was performed by electrochemical advanced oxidation processes either with Pt or BDD anodes and a 3D carbon felt cathode. The degradation of NAX by (•OH vs. electrolysis time) was well fitted to a pseudo-first-order reaction rate kinetic. The detected reaction intermediates (aromatic compounds and carboxylic acids) were experimentally monitored during the process via LC, while density functional theory (DFT) was applied to uncover undetected intermediates, some for the first time in literature. The formation of toxic intermediates with higher toxicity than NAX were identified, such as IMS4b (6-Methoxy-1-[1-(6-methoxynaphthalen-2-yl) ethyl] naphthalen-2-ol), catechol, and glycolic acid. Based on these data, a detailed oxidation pathway of NAX by •OH was proposed. The evolution of solution toxicity indicated that formed toxic intermediates were subsequently removed during the TOC removal process. Finally, almost complete mineralization of NAX was achieved in simulated urine or wastewater, by the electro-Fenton treatment with an optimized dose of iron as catalyst, showing the EAOPs’ potential to efficiently remove NAX even from challenging matrices. In extension, the strategies developed can be applied to the treatment of other NSAIDs.
Most real optimization problems are defined over a mixed search space where the variables are both discrete and continuous. In engineering applications, the objective function is typically calculated with a numerically costly black-box simulation. General mixed and costly optimization problems are therefore of a great practical interest, yet their resolution remains in a large part an open scientific question. In this article, costly mixed problems are approached through Gaussian processes where the discrete variables are relaxed into continuous latent variables. The continuous space is more easily harvested by classical Bayesian optimization techniques than a mixed space would. Discrete variables are recovered either subsequently to the continuous optimization, or simultaneously with an additional continuous-discrete compatibility constraint that is handled with augmented Lagrangians. Several possible implementations of such Bayesian mixed optimizers are compared. In particular, the reformulation of the problem with continuous latent variables is put in competition with searches working directly in the mixed space. Among the algorithms involving latent variables and an augmented Lagrangian, a particular attention is devoted to the Lagrange multipliers for which a local and a global estimation techniques are studied. The comparisons are based on the repeated optimization of three analytical functions and a beam design problem.
A silica hollow microbottle resonator (HMBR) combined with a pair of curved silicon micro-mirrors on the outside wall of the microbottle is proposed and numerically investigated using the Finite-Difference Time-Domain (FDTD) algorithm. The microbottle has only 32 μm length, 26 μm width and 1.5 μm wall thickness. The curved micro-mirrors overcome the light diffraction loss through light focusing, while the microbottle, in which gas analytes are introduced, provides additional light confinement and hence improves the performance of the sensor. The obtained Q-factor is about 4590 at 1543.21 nm and the free spectral range (FSR) is more than 31 nm. An internal sensitivity of 1567 nm per refractive index unit (RIU) is achieved in the near-infrared (NIR), which is the highest ever reported for an refractive index (RI) gas sensor based on HMBR. With the introduction of an air gap layer between the silica HMBR and the silicon micro-mirrors, both the Q-factor and sensitivity have been improved to 6729 and 1730 nmRIU− 1 respectively. We believe that the proposed architecture will be used in future sensing applications.
Traditional agriculture has been the pillar of development on the planet for centuries. But with exponential population growth and increasing demand, farmers will need water to irrigate the land to meet this demand. Because of the scarcity of this resource, farmers need a solution that changes the way they operate. With the advent of new technologies, the notion of Agriculture 4.0 has become a reality to keep up with and meet the demand. With the addition of artificial intelligence and IoT through the collection and processing of agricultural data, decisions have become more and more precise to facilitate decision-making. This paper proposes an intelligent and flexible irrigation approach with low consumption and cost that can be deployed in different contexts. This approach is based on machine learning algorithms for smart agriculture. For this, we used a set of sensors (soil humidity, temperature, and rain) in an environment that ensures better plant growth for months, from which we collected data based on an acquisition map using the Node-RED platform and MongoDB. We used many different models based on the collected data: KNN, Logistic Regression, Neural Networks, SVM, and Naïve Bayes. The results showed that K-Nearest Neighbors is better with a recognition rate of 98.3% and a root mean square error (RMSE) of 0.12, compared to other models (LR, NN, SVM, NB). and towards the end, we provided a web application that brings together the various data emitted by the sensors as well as the prediction of our models to allow better visualization and supervision of our environment.
In this paper, we contribute a new B-spline based interval field decomposition method as a non-probabilistic approach that takes into account local effects in interval field modelling. With B-spline basis functions, the interval field formulation is highly intuitive and easy to construct. The computational efficiency outperforms the traditional local interval field decomposition method. The explicit expression of the proposed method facilitates the use of optimisation methods in determining field bounds where deterministic local values are available. The proposed method can incorporate the use of truncated hierarchical B-spline basis functions and multi-patch stitching method that facilitate modelling of inhomogeneous interval fields, which effectively address the spatial variability of the parameters describing the interval field. A numerical case of a simply supported beam with non-deterministic material parameters subjected to external loads is performed to illustrate the applicability of the proposed method.
The aim of this paper is twofold. On one hand, we strive to give a simpler proof of the optimality of greedy controls when the cost of interventions is control-affine and the dynamics follow a state-constrained controlled SIR model. This is achieved using the Hamilton–Jacobi characterization of the value function, via the verification argument and explicit trajectory-based computations. Aside from providing an alternative to the Pontryagin complex arguments in Avram et al. (Appl Math Comput 418:126816, 2022) (see also Avram et al. in Appl Math Comput 423:127012, 2022), this method allows one to consider more general classes of costs; in particular state-dependent ones. On the other hand, the paper is completed by linear programming methods allowing one to deal with possibly discontinuous costs. In particular, we propose a brief exposition of classes of linearized dynamic programming principles based on our previous work and ensuing dual linear programming algorithms. We emphasize the particularities of our state space and possible generations of forward scenarios using the description of reachable sets.
Recycled Concrete Aggregates (RCA) are mainly composed of two different materials: natural aggregate and attached mortar. Hence, RCA may have a much significant heterogeneity compared to natural aggregates. This heterogeneity limits their reuse in concrete due to the uncertainty in their expected behaviour. To clarify the effect of the attached paste volume on their properties, a single source of RCA was separated with a water jig according to the density of the particles. This original approach allows to avoid any disturbance of the data analysis with uncontrolled variation of paste nature or natural aggregate nature, as it is generally the case in literature collecting RCA from different sources. The used RCA had a narrow granular fraction (10–14 mm). After sorting tests, the sorted RCA was characterized by the measure of their density, water absorption, particle size distribution, particle shape and the attached mortar content. The results show that very large disparities exist in the density and the water absorption even in a narrow granular fraction, which can be explained by the dispersion of the attached mortar content in the RCA. These results allowed to established experimental relationships between the relative density and the others RCA properties. The developed relationships can been used to evaluate the attached mortar content for further laboratory study for example.
This paper studies the asymptotic behavior of the constant step Stochastic Gradient Descent for the minimization of an unknown function, defined as the expectation of a non convex, non smooth, locally Lipschitz random function. As the gradient may not exist, it is replaced by a certain operator: a reasonable choice is to use an element of the Clarke subdifferential of the random function; another choice is the output of the celebrated backpropagation algorithm, which is popular amongst practioners, and whose properties have recently been studied by Bolte and Pauwels. Since the expectation of the chosen operator is not in general an element of the Clarke subdifferential of the mean function, it has been assumed in the literature that an oracle of the Clarke subdifferential of the mean function is available. As a first result, it is shown in this paper that such an oracle is not needed for almost all initialization points of the algorithm. Next, in the small step size regime, it is shown that the interpolated trajectory of the algorithm converges in probability (in the compact convergence sense) towards the set of solutions of a particular differential inclusion: the subgradient flow. Finally, viewing the iterates as a Markov chain whose transition kernel is indexed by the step size, it is shown that the invariant distribution of the kernel converge weakly to the set of invariant distribution of this differential inclusion as the step size tends to zero. These results show that when the step size is small, with large probability, the iterates eventually lie in a neighborhood of the critical points of the mean function.
The installation of shoulders on rural roads to create more forgiving roads encourages drivers to cut corners on right-hand bends, but the underlying mechanisms are poorly understood. Since eye movements and steering control are closely coupled, this study investigated how the presence of a shoulder influences drivers' gaze strategies. To this end, eighteen drivers negotiated right-hand bends with and without a shoulder on a simulated rural road. In the presence of a shoulder, participants modified their visual sampling of the road by directing their gaze further inside the bend. At the same time, their lane position was deviated inward throughout the bend and the vehicle spent more time out of the lane. These results suggest that the shoulder influences the visual processes involved in trajectory planning. Recommendations are made to encourage drivers to keep their eyes and vehicle in the driving lane when a shoulder is present.
Light earth is a natural insulating material composed of earth and vegetal fibres. It can be used to insulate existing and new buildings to reduce energy and resources consumption, and excavated earth generation by the construction sector. A pedological database is crossed with suitability thresholds in order to evaluate spatially the availability of earth resources. Then, the soil suitability is mapped, and suitable soil amounts metrics are estimated for Brittany territory. A sensitivity analysis is performed to understand the potential variability of the results. Study estimates that 48% of Brittany's soil horizons are suitable for light earth building. Every year, 1.3 Mt of suitable soil are excavated in Brittany. Using only these excavated earths, all existing and new buildings in Brittany could be insulated with light-earth in less than 8 years. This study shows that suitable earth availability is not a limiting factor to develop light earth insulation in Brittany.
The browning phenomenon is a pathology affecting Mn-bearing medieval stained-glass-windows in potash-lime-silicate glass system. In order to unravel the potential implication of microorganisms in the appearance of this pathology, three model glasses respectively containing no MnO, 1 wt% and 2 wt% MnO were altered at circumneutral pH, with and without organic exudates (oxalic acid (OA) 1000 μM and siderophore desferrioxamine B (DFOB) 50–1000 μM) likely to be produced by bacteria and fungi. In the absence of exudates, the dissolution rates are inversely dependent on the Mn content of the glasses (0.8, 0.5 and 0.4 g m⁻²d⁻¹ respectively for no MnO, 1 wt% and 2 wt% MnO glasses). In contact with exudates, an opposite trend is observed. The prevalent mechanisms are interpreted as a strong ligand-promoted dissolution for DFOB (dissolution rate increase up to 270%) and a dominant proton-promoted dissolution for OA (dissolution rate increase up to 60%). When DFOB and OA are added together, the effect of DFOB on the dissolution rate is prevailing, while OA effect is tangible on the stoichiometry of the dissolution of the alteration. These results suggest that an indirect biological activity could be involved in the mobilization of Mn from a Mn-bearing glass, thus playing a role in the appearance of the browning phenomenon.
Thermal bowing is one of the major concerns when analysing the structural behaviour of masonry walls in fire conditions, along with the loss of resistance due to material degradation. This contribution proposes a new homogenization method aimed at determining the thermal deformed shape of a masonry wall exposed to fire. First, the homogenized thermo-elastic characteristics of a natural stone masonry wall are determined taking into account the material properties of stone and mortar as functions of temperature increase, as well as the geometrical characteristics of their assembly. As a result, membrane and bending stiffness coefficients, as well as thermal-induced efforts, of the wall modelled as a homogenized plate are obtained. Such homogenized thermo-mechanical characteristics make it possible to determine the deformed shape of the wall after a certain time of fire exposure, through the formulation and the resolution of a finite element plate problem, making use of the previous homogenized stiffness characteristics. Numerical simulations show that the presence of joints has little influence on the deformed shape of the masonry wall compared to a homogeneous stone-only wall. They appear to be in rather good agreement with previous full-scale experimental tests carried out on limestone masonry walls, therefore providing a first validation of the proposed method, even though further improvements could be envisaged as regards a more reliable determination of the temperature gradient across the wall thickness.
The durability, which refers to the ability of earthen structures to ensure their functionality over time while maintaining their required mechanical performance, is a key issue in evaluating the effectiveness of lime treatment. In this study, the effect of wetting-drying cycles on the compressibility and the microstructure was investigated with lime-treated/untreated samples considering the wetting fluid and the maximum aggregate size (S0.4, Dmax = 0.4 mm; S5, Dmax = 5 mm) effects. Results showed that the wetting-drying cycles caused an increase of void ratio and changed the bi-modal porosity to tri-modal characteristics for the untreated samples, while it led to a reversible variation of void ratio and an unchanged bi-modal pore size distribution characteristics for lime-treated samples, indicating the soil improvement by lime treatment. Thereby, the wetting-drying cycles made the compression curve of untreated samples change from convex to linear shape, while their effect was visible but not significant for the lime-treated soil. Higher decrease of macro-pore void ratio with loading was obtained on the lime-treated samples under wetting-drying cycles compared to the as-compacted samples, indicating a slight softening of soil structure by the wetting-drying cycles. Regarding the effect of wetting fluid nature, synthetic seawater resulted in a higher compressibility than deionised water. This was attributed to the presence of higher quantity of macro-pores in the samples wetted by synthetic seawater induced by the shrinkage of clay fraction. The aggregate size had slight effect on the compressibility of as-compacted samples due to their similar production of cementitious compounds and matric suction. Nevertheless, after wetting-drying cycles, the lime-treated samples S5 had lower decrease of macro-pore void ratio than the lime-treated samples S0.4, due to the significant reduction of macro-pore population with wetting-drying cycles.
We study a Schrödinger equation modeling the dynamics of an electron in a crystal in the asymptotic regime of small wave-length comparable to the characteristic scale of the crystal. Using Floquet Bloch decomposition, we obtain a description of the limit of time averaged energy densities. We make a rather general assumption assuming that the initial data are uniformly bounded in a high order Sobolev spaces and that the crossings between Bloch modes are at worst conical. We show that despite the singularity they create, conical crossing do not trap the energy and do not prevent dispersion. We also investigate the interactions between modes that can occurred when there are some degenerate crossings between Bloch bands.
Cigarette butts (CB) rank at the top of littered waste materials and can cause a serious impact on the environment. CB are mainly composed of cellulose acetate (CA) fibers, a polymer that has poor biodegradability. Following the growing concern to reduce pollution, this study presents an innovative way to recycle industrial treated CA fibers by incorporating them in cementitious mortars as partial replacement of sand. CA fibers are found to be porous, with a total porosity of 97%, to have a bulk density of 65 ± 2 kg/m³ when compacted and to have a high water absorption of about 853 wt% that is higher than other bio-sourced materials. Optimal formulations based on workability tests are 0.2 wt% of sand replacement by CA without the use of a superplasticizer (SP), and 1.3 wt% with 3 wt% of SP. The compressive and bending strengths, and the total shrinkage of the reference mortar are close to those the of the 0.2 wt% CA contained mortars, as they have close porosity. The latter increases for 1.3 wt% CA (and SP) contained mortar leading to a decrease in mechanical strength and an in increase in shrinkage. Thermogravimetric analyses (TGA) reveal that the quantity of water released by the CA fibers enhances the cement hydration.
We consider two-phase composites whose microstructures are two-dimensional and generated by the periodic replication of a convex polygonal cell containing a single inclusion embedded in a matrix. Adopting the framework of nonlinear conductivity, we address the problem of finding the inclusion shape that optimizes the effective energy. A conceptually simple but numerically effective approach is presented, in which the inclusion shape is parameterized by the Fourier coefficients of a scalar periodic function f that defines its polar representation. Truncating the Fourier expansion to a finite order turns the shape optimization problem into a finite-dimensional constrained optimization problem that can be solved using a numerical algorithm of choice. Explicit expressions of the function to optimize and its gradient are provided and can easily be evaluated from a finite-element model. The proposed approach is applied to perfectly conducting inclusions in a power law matrix. Results for the three types of regular tessellations (square, hexagonal and triangular) are presented and compared with the Vidgergauz [23, 24] microstructures giving the extremal inclusions in the linear case. The proposed method gives very simple representations of the extremal inclusions, which could useful for manufacturing the microstructures considered. The obtained nonlinear effective conductivities are compared with known Hashin–Shtrikman-type nonlinear bounds, which contributes to shed some light on the optimality of those bounds.
Institution pages aggregate content on ResearchGate related to an institution. The members listed on this page have self-identified as being affiliated with this institution. Publications listed on this page were identified by our algorithms as relating to this institution. This page was not created or approved by the institution. If you represent an institution and have questions about these pages or wish to report inaccurate content, you can contact us here.
2,109 members
Mahdi Zargayouna
  • Génie des Réseaux de Transports Terrestres et Informatique Avancée (GRETTIA)
Nathalie Fabry
  • Dispositifs d'Information et de Communication à l'Ère Numérique (Dicen)
Alexandra Fort
  • Laboratoire Ergonomie et Sciences Cognitives pour les Transports (LESCOT)
Champs-sur-Marne, France