Institut Mines-Télécom
Recent publications
Fluvial sediments need to be periodically dredged from waterways. At the same time, the building sector has to find high-volume alternatives to the extraction of non-renewable mineral resources. In this context, recent research has shown the benefit of reusing calcined sediments as supplementary cementitious materials. On a different note, this study examines the hydration and performance of blends of calcined fluvial sediment with low amounts (10 to 30 wt%) of hydrated lime with the aim of keeping the carbon footprint of the binder the lowest as possible. The effect of different curing conditions (moist curing at 20 °C or 50 °C and dry curing at 65%RH) is studied on compressive strength, reaction kinetics (fixed lime by TGA and heat of hydration), reaction products (XRD), microstructure (SEM) and porosity (MIP). The outcomes of the study revealed that heat curing for 3 days and 10 wt% of lime results in satisfactory strength development for the design of bio-based concretes (compressive strength of 9 MPa). Furthermore, the highest compressive strength of 17.5 MPa was achieved at 110 days at 20 °C for the mix with 20 wt% of lime. The amount of unreacted lime was found to be the highest for the mix including 30 wt% of lime for which the degree of reaction of lime at 90 days was around 0.7 for both curing temperatures. Under dry curing, a large part of the initially available lime was carbonated and some reaction products underwent carbonation leading to very low strength by 110 days.
Turbulence is known to enhance gas-liquid mass transfer and mixing of high Schmidt number dissolved gases in water by deforming the concentration boundary layer that develops at the interface. Fundamental mechanisms of surface renewal and injection have been progressively evidenced throughout the last decades, via fundamental experiments of low mean shear turbulence interacting with flat interfaces in water. However, and despite the obvious influence of non-Newtonian behaviours on gas liquid mass transfer in industrial and environmental applications, not such study exists (to the best of the author’s knowledge) on whether and how these mechanisms apply in shear-thinning dilute polymer solutions (DPS). Following a previous work on near surface hydrodynamics, turbulent mass transfer and mixing is studied in a weakly shear-thinning fluid, and compared to the Newtonian, water case. Stereoscopic Particle Image Velocimetry (SPIV) and Inhibited Planar Laser Induced fluorescence are used simultaneously to measure the local liquid phase velocity and dissolved gas concentration fields respectively. Coupled measurements are used to estimate the turbulent mass fluxes, which are interpreted using a conditional quadrant analysis. Results show that in DPS as well as in water, surface renewal is the most frequent mechanism, but injection events are the most efficient in terms of mass transfer. Even at a low concentration, the polymer significantly modifies the signature of those mass transfer events, by enhancing scalar stretching and injection mass transfer outside of the viscous sub-layer, altering classical gradient models.
The most developing issue in analysing complex networks and graph mining is link prediction, which can be studied for both content and structural-based analysis in a social network. Link prediction deals with the prediction of missing links by determining whether a link can be created between two nodes in a future snapshot of a given directed weighted graph. Existing link prediction methods are only studied for unsigned graphs and work on principles of the common neighbourhood. However, the link prediction problem can also be studied for signed graphs where signed links can give an interesting insight into user associations. Obstruction of studies in this domain is caused by imbalance of class, i.e., positive links are frequent than negative ones, and forbearance of hidden communities. A signed network is a combination of dense and hidden communities. A hidden community structure is overlooked by majority of existing applications, taking dense community structure, i.e., one whole graph as input for developing a link prediction model. Hence, complete network information is required by majority of existing approaches, which seems unrealistic in modern social network analytics. In this article, we exploit hidden network communities to address link prediction problem in the signed network, focusing on negative links. A number of observation were made regarding negative links and a principle ensemble framework, i.e., E NeLp, is proposed, having two phases, i.e, network embedding and classifier prediction. Using a probabilistic embedding framework, network representation of hidden signed communities is learned, which were then passed to a learning classifier to predict negative links, keeping intact the ensemble framework. Despite the limited availability of signed network datasets, an extensive experimental study was performed to evaluate E NeLp pertinency, robustness, and scalability. The performance result shows that E NeLp can be a promising consideration for addressing link prediction tasks in signed networks and gives encouraging results.
The biomass bottom ash (BBAS) from Roubaix and Lens (France) with different physical–chemical characteristics was used as reactive filters for phosphorus (P) removal. The principal objective of our study was to evaluate the P removal potential with the two BBAS, and batch tests were performed to determine the influence of several parameters including the origin and type of BBAS, particle size, chemical composition, initial P concentration, contact time and water quality (real wastewater, synthetic wastewater) on P removal performance. Then, a series of laboratory scale column tests were performed with real wastewater doped to 30 mg P/l during the 8 months of experiments. To determine the mechanisms of P removal by the two BBAS, a monitoring of pH, calcium concentration and P concentration was performed using a fraction collector. In addition, a series of physical–chemical analyses such as XRD, SEM–EDS and sequential extraction were performed before and after 8 months of filter operation. The maximum phosphorus removal capacity (PRC) of the real wastewater solutions (100 mg P/l) in batch ranged from 1.56 to 1.96 mg P/g after 7 days of contact for BBAL and BBAR, respectively. The maximum P removal capacities in the columns varied from 4.15 to 9.72 g P/kg material after 8 months of experimentation for BBAL and BBAR, respectively. The results of XRD, SEM–EDS, sequential extraction confirm that the main mechanism of P removal was through precipitation of Ca-P (as hydroxyapatite). The P removal performance of the real wastewater was mainly related to the particle size, specific surface area and CaO dissolution of each BBAS. However, the results show that wastewater treatment plants represent an important source of Ca2+ ions to further promote the precipitation of hydroxyapatite on the surface, resulting in increased P removal efficiency.
Video super-resolution aims to reconstruct a high-resolution video from a low-resolution video corresponding to a magnification scale. Video super-resolution, as a fundamental computer vision task, is widely used in various fields. Particularly, in the field of endoscopic, high-resolution endoscopic videos help doctors to observe more details of lesions and improve the accuracy and speed of diagnosis. A novel deformable Transformer network is proposed to solve the super-resolution problem of endoscopic video data. To address the problem that the Transformer’s self-attention module cannot effectively capture local information, the self-attention module is improved by using convolution operations to increase the local feature capture capability of the self-attention module. In order to compensate for the deficiency of Transformer for continuous inter-frame alignment, a new bidirectional deformable convolutional network is designed as the feed-forward module of Transformer to achieve frame-to-frame feature alignment and feature propagation using deformable convolution. A high-resolution dataset for endoscopic video super-resolution is produced using endoscopic surgery videos. Our proposed deformable Transformer network is demonstrated to have the best performance with the competitive number of parameters in endoscopic imaging so far by comparing the performance of other existing video super-resolution methods in the endoscopic dataset through sufficient experiments. Our proposed deformable Transformer network improves the PSNR metric by 0.97 dB over the state-of-the-art method in the RGB channel, while reducing the number of network parameters by 0.39 million.
Many rail services around the world continue to use diesel as the primary fuel source and enclosed railway stations have been identified as a possible hotspot for exposure to harmful diesel exhaust exposures. Little is known about the occupational exposure to air pollution for railway station workers due to their mobility around the station and variations in station design. A detailed understanding of the concentration of black carbon (BC), a diesel exhaust tracer, inside railway stations and the factors driving occupational exposures is required to minimize occupational exposure. Real-time personal exposure to BC was measured during 60 work-shifts encompassing different roles at three large enclosed railway stations of different design in London, Birmingham and Edinburgh (UK). Sampling was conducted by the train station workers over a period of 27 days between January 2017 to October 2018. Worker shift-mean BC exposures ranged 0.6–20.8 μg m⁻³ but 1-min peak exposures reached 773 μg m⁻³, with train dispatchers experiencing the highest BC exposures. Station design, job role, and frequency of diesel trains were the main drivers of occupational BC exposure. Elevated exposures for some station workers indicate that mitigation measures to reduce their exposure should be implemented to lower the risk of occupational health impacts. These could include improving ventilation and reducing engine emissions.
To overcome the curse of dimensionality of particle filter (PF), the block PF (BPF) partitions the state space into several blocks of smaller dimension so that the correction and resampling steps can be performed independently on each block. Despite its potential performance, this approach has a practical limitation. BPF requires an additional input compared to classical PF: the partition from which blocks are defined. A poor choice will not offer the expected performance gain. In this paper, we formulate the partitioning problem as a clustering problem and propose a data-driven partitioning method based on constrained spectral clustering (CSC) to automatically provide an appropriate partition. We design a generalized BPF that contains two new steps: (i) estimation of the state vector correlation matrix from particles, (ii) CSC using this estimate to determine blocks. The proposed method succeeds in providing an online partition into blocks restricted in size and grouping the most correlated state variables. This partition allows the BPF to escape the curse of dimensionality by reducing the variance of the filtering distribution estimate while limiting the level of bias. Since our approach relies on particles that are already necessary to generate as part of BPF, the computation overhead is limited.
Whether a harbinger of a new era or an anomaly, the year 2020 confronted students and teachers alike with the necessity of reassessing and reformulating teaching and learning possibilities and practicalities. In this very subjective text, we examine some of our own experiences of higher education under lockdown and physical distancing conditions. In an apparent paradox, the changed conditions simultaneously added more stress and uncertainty to the students’ learning process while also providing the learners with more confidence to question the established norms. Against the background of ongoing systemic collapse, we explore our own and our students’ stories and poems chronicling learning in a time of crisis and constraint. Drawing on critiques of modern consumer capitalism underpinning management education, we use the experience of a ruptured semester to propose a reinterpretation of management learning as rooted in the paradoxes of desire and longing: for success, career, but also for enlightenment, revelation, social change and togetherness. We ask the reader to embrace the poetic and libidinal aspects of desire and longing as central to the transformative potential of the learning encounter and propose to reconstitute the basis for education as rooted in desire and longing: for contact, for learning, for revelation.
Mineral foam concretes are classified as lightweight concretes, which are characterized by the presence of air voids in the materials due to the use of a suitable foaming agent. Their physical and mechanical properties are strongly governed by microstructural properties, which depend on several factors such as foam quantity, water/binder ratio (W/B), and the mixing process. In this study, mineral foams were prepared with a mixed foaming method. Gypsum is used as a binder, flax shives as vegetative aggregates, and Betomouss as the foaming agent (sodium lauryl ether sulfate as the main tensioactive molecule). The impact of mixing conditions (stirring rate, stirring time, and W/B ratio) on the performances of foams is investigated. The results show that varying these mixing conditions will induce high variation of foam performances at the fresh and hardened states. Furthermore, increasing the stirring rate and stirring time generates an increase of the air-void system, finally causing materials to decrease foam density. Consequently, this entails an improvement in thermal conductivity and a reduction of mechanical properties. In addition, increasing the W/B ratio leads to an increase in spreadability, density, and mechanical properties. In contrast, foams with a low W/B ratio have the best thermal properties.
Aerosol pollutant particles indoors significantly affect public health. The conventional wisdom is that natural ventilation will alleviate the dispersion of airborne or aerosol particles. However, we show that the problem is far more complex and that natural ventilation should be applied under specific conditions to be effective. We performed several simulations of a simplified (and easily reproducible) room with a window opening, and aerosol particles stratified layers. Opening a window can scatter particles present in stratified layers indoors and, potentially, contribute to the degradation of indoor air quality for a significant period of time. Moreover, we show that thermal instabilities arising from the temperature gradients due to temperature differences between the indoor and outdoor environment spread the particles randomly indoors, adversely affecting air quality and architectural design. Recommendations for more efficient natural ventilation minimizing aerosol pollutant particles dispersed indoors are provided.
Multifunctional heat exchangers (HEX) are very important devices which are used in many industrial applications (aerospace, aeronautics, automobile, process and chemical engineering, etc). Improving their designs for an optimal overall performance is still a wide window for both research and development. In the present contribution, a new HEX design is proposed as two concentric elliptically-deformed tubes of complex geometry. Employing advanced computational fluid dynamics (CFD), we show how a combined Poiseuille-Taylor-Couette (PTC) analogous flow between two rotating concentric elliptically-deformed annular tubes, can efficiently enhance heat and mass transfer at low to medium Reynolds numbers (Re<=488). Furthermore, we illustrate how chaotic advection can be controlled by regulating the local flow inertia and rotational forces imposed by a combined flow between: i- Poiseuille flow, and ii- Taylor-Couette analogous flow that is generated by interchanging the clockwise/anticlockwise rotation of the inner and outer walls. Results are presented and discussed using Lagrangian tracking and Poincaré sections techniques that describe the different underlying physical phenomena of mixing and heat transfer.
This work presents the design of a combined control and state estimation approach to simultaneously maintain optimal water levels and maximize hydroelectricity generation in inland waterways using gates and ON/OFF pumps. The latter objective can be achieved by installing turbines within canal locks, which harness the energy generated during lock filling and draining operations. Hence, the two objectives are antagonistic in nature, as energy generation maximization results from optimizing the number of lock operations, which in turn causes unbalanced upstream and downstream water levels. To overcome this problem, a two-layer control architecture is proposed. The upper layer receives external information regarding the current tidal period, and determines control actions that maintain optimal navigation conditions and maximize energy production using model predictive control (MPC) and moving horizon estimation (MHE). This information is provided to the lower layer, in which a scheduling problem is solved to determine the activation instants of the pumps that minimize the error with respect to the optimal pumping references. The strategy is applied to a realistic case study, using a section of the inland waterways in northern France, which allows to showcase its efficacy.
One major research topic is to characterize the mechanical behaviour of actual reinforced pavement structures from laboratory experimentation and take it into account for the design. This investigation aims to verify the effect of fiberglass geogrid presence on interface linear viscoelastic (LVE) behaviour separately and as a system along with the bituminous mixture layers. To conduct the research, two different fiberglass geogrids, with ultimate tensile strength (UTS) of 100 and 50 kN/m, and tack coat made of straight-run bitumen and modified by polymer were combined for the fabrication of three reinforced configurations. In addition, two unreinforced configurations were also fabricated. The first was a single layer slab and the second was a double-layered slab composed of two bituminous mixtures (same type) bonded layers by a tack coat. Complex modulus tests were carried out in specimens cored in two different directions, vertically (V) and horizontally (H) cored. The experimental data were fitted using the 2 Springs, 2 Parabolic Elements and 1 Dashpot (2S2P1D) model. The test results showed that all interfaces’ complex modulus obtained for V specimens were LVE. Moreover, complex viscous properties of the interfaces were obtained from the used binder. The interface containing polymer modification presented the highest stiffness.
Unsaturated alcohols are volatile organic compounds (VOCs) that characterize the emissions of plants. Changes in climate together with related increases of biotic and abiotic stresses are expected to increase these emissions in the future. Ozonolysis is one of the oxidation pathways that control the fate of unsaturated alcohols in the atmosphere. The rate coefficients of the gas-phase O3 reaction with seven C5-C8 unsaturated alcohols were determined at 296 K using both absolute and relative kinetic methods. The following rate coefficients (cm3 molecule-1 s-1) were obtained using the absolute method: (1.1 ± 0.2) × 10-16 for cis-2-penten-1-ol, (1.2 ± 0.2) × 10-16 for trans-2-hexen-1-ol, (6.4 ± 1.0) × 10-17 for trans-3-hexen-1-ol, (5.8 ± 0.9) × 10-17 for cis-3-hexen-1-ol, (2.0 ± 0.3) × 10-17 for 1-octen-3-ol, and (8.4 ± 1.3) × 10-17 for trans-2-octen-1-ol. The following rate coefficients (cm3 molecule-1 s-1) were obtained using the relative method: (1.27 ± 0.11) × 10-16 for trans-2-hexen-1-ol, (5.01 ± 0.30) × 10-17 for trans-3-hexen-1-ol, (4.13 ± 0.34) × 10-17 for cis-3-hexen-1-ol, and (1.40 ± 0.12) × 10-16 for trans-4-hexen-1-ol. Alkenols display high reactivities with ozone with lifetimes in the hour range. Rate coefficients show a strong and complex dependence on the structure of the alkenol, particularly the relative position of the OH group toward the C═C double bond. The results are discussed and compared to both the available literature data and four structure-activity relationship (SAR) methods.
Observations of key gaseous trace pollutants, namely NO, NOy, CO, SO2 and O3, performed at several curb, residential, industrial, background and free-troposphere sites were analyzed to assess the temporal and spatial variability of pollution in Cyprus. Notably, the analysis utilized one of the longest datasets of 17 years of measurements (2003–2019) in the East Mediterranean and the Middle East (EMME). This region is considered a regional hotspot of ozone and aerosol pollution. A trend analysis revealed that at several stations, a statistically significant decrease in primary pollutant concentration is recorded, most likely due to pollution control strategies. In contrast, at four stations, a statistically significant increase in ozone levels, ranging between 0.36 ppbv y⁻¹ and 0.82 ppbv y⁻¹, has been observed, attributed to the above strategies targeting the reduction of nitrogen oxides (NOx) but not that of Volatile Organic Compounds (VOCs). The NO and NOy, and CO levels at the Agia Marina regional background station were two orders of magnitude and four times lower, respectively, than the ones of the urban centers. The latter denotes that local emissions are not negligible and control a large fraction of the observed interannual and diurnal variability. Speciation analysis showed that traffic and other local emissions are the sources of urban NO and NOy. At the same time, 46 % of SO2 and 40 % of CO, on average, originate from long-range regional transport. Lastly, a one-year analysis of tropospheric NO2 vertical columns from the TROPOMI satellite instrument revealed a west-east low-to-high gradient over the island, with all major hotspots, including cities and powerplants, being visible from space. With the help of an unsupervised machine learning approach, it was found that these specific hotspots contribute overall around 10 % to the total NO2 tropospheric columns.
This research exposes a new methodology to follow the evolution of polymer crystallinity during material extrusion thanks to the use of thermocouples and fast scanning calorimetry (FSC). Micrometric thermocouples allowed recording the thermal history at different locations on a polypropylene (PP) part, which was then used to reproduce the thermal process conditions in FSC. This method allows to precisely study the evolution of PP crystallinity as function of time, which is an important feature to predict the printability of polymers, since crystallization is known to restrain fusion between beads. The results highlighted the dependence of the periodic thermal stress undergone by PP on several characteristics (e.g. layer surface area, location in the part, filling strategy). Therefore, the crystallization behavior was found to be different depending on the location in the part. Crystallization was delayed at the center of the part compared to the edges because the nozzle sweep frequency was higher at the center, resulting in lower cooling kinetics.
The hardware primitives known as Physically Unclonable Functions (PUFs) generate unique signatures based on uncontrollable variations which occur during the manufacturing process of silicon chips. These signatures are in turn used for securing Integrated Circuits either as a secret key for cryptographic modules, or as a medium for authenticating devices. Naturally being a security primitive, PUFs are the target for attacks as such it is important to mitigate such vulnerabilities. This paper in particular investigates PUFs’ vulnerability to power-based modeling attacks. Here, we expand upon our previous simulation based Cross-PUF attacks by targeting PUFs realized in real-silicon; namely, we consider PUFs deployed in Field-Programmable Gate Array (FPGA) fabrics. In Cross-PUF attacks, a model of a reference PUF is used to attack another PUF realized from the same HSPICE simulated design or the same bitstream in FPGA. We also investigate the impact of such attacks on multi-bit parallel PUFs. The HSPICE simulation results are compared vis-a-vis with the FPGA implementation outcome of these attacks confirming the effectiveness of such simulations. Finally we show that a combination of Dual Rail logic and Random Initialization logic, named DRILL, can be effectively used to thwart such power-based modeling attacks.
Radicals and their precursors play a central role in the chemical transformations occurring in indoor air and on indoor surfaces. Such species include OH, HO2, peroxy radicals, nitrous acid, reactive chlorine species, NO3, N2O5, Criegee intermediates, and glyoxal and methylglyoxal. Recent advances on instrumental analysis and modeling studies have demonstrated the need for a wider range of measurements of radical species and their precursors in indoor air. This work reviews measurement techniques and provides considerations for indoor measurements of several radicals and their precursors. Techniques to determine the actinic flux are also presented owing to the relevance of photolytically-initiated processes indoors. This review is also intended to provide pointers for those wanting to learn more about measurements of radicals indoors.
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.
1,300 members
Luigi Iannone
  • Département d'Informatique et Réseaux
Jean-Marc Le Caillec
  • Département Image et Traitement de l'Information
Zribi Amin
  • Département Signal et Communications
Alexander Pelov
  • Département Réseaux, Sécurité et Multimédia (RSM)
Antoine Beugnard
  • Département Informatique (Bretagne)
37-39 rue Dareau, 75014, Paris, France
Head of institution
France’s #1 group of Engineering and Management graduate schools