Portail HAL IMT Mines Albi
Not a member yet
5440 research outputs found
Sort by
Thermomechanical performance of continuous carbon fibre composite materials produced by a modified 3D printer
No full textInternational audienceFirst of all, this article aimed to evidence the role of a modified printer developed for continuous carbon fibre reinforced PolyAmide (cCF/PA6-I) together with the use of a fully open slicing step on the printing quality and the longitudinal/transverse tensile and in-plane shear properties. A comprehensive assessment of the microstructure and properties with a similar material (cCF/PA6-I), but produced with a commercial printer (i.e., Markforged® MarkTwo) has been achieved. Our customised printer and the open slicer used have made possible to better control the print conditions (i.e., layer height and distance between filaments), to reduce the porosity from more than 10% to about 2% and improve the mechanical properties.Moreover, the understanding of the behaviour of these 3D printed composites with wide-ranging external temperatures is mandatory for future use in a severe environment and/or development of new thermally active 4D printed composites.The 3D printed cCF/PA6-I composites have been then thermomechanically characterised along different printing directions (0, 90 and ± 45°) from −55 to +100 °C. Unlike the longitudinal properties that hardly change with temperature, the transverse and in-plane shear stiffness and strength of these 3D printed composites were particularly sensitive to temperature variations, with decreases of 25–30% and 30–55%, respectively. This was due to the high sensitivity of the polymer matrix, the fibre/matrix and interfilament interfaces when the composites were loaded along those directions, because damages induced by internal thermal stresses. Fractography has also been carried out to reveal damage mechanisms
Pyrolysis of wood and PVC mixtures: thermal behaviour and kinetic modelling
No full textInternational audienceWood waste containing halogenated compounds such as polyvinyl chloride (PVC) is in abundant supply, although the pyrolysis of such waste feedstock for energy production may cause corrosion and environmental problems due to the release of HCl gas. Hence, there is a need to understand the pyrolysis behaviour of chlorine-contaminated wood in order to develop methods that minimise the impact of chloride species on pyrolysis equipment and product quality. In literature, few studies exist on the kinetic analysis of wood and PVC co-pyrolysis. The existing models assume a single-step reaction with an n-order reaction mechanism for the entire process, which may lead to large errors in the kinetic parameters estimated. Therefore, in this paper, we develop and validate a multi-step kinetic model that predicts the pyrolysis behaviour and reaction mechanism of poplar wood (PW) pellet with different contents of PVC (0, 1, 5, 10, 100 wt%). Using data from thermogravimetric analysis of the pellets at heating rates of 5, 10 and 20 °C/min, we determined the apparent kinetic parameters by combining Fraser-Suzuki deconvolution, isoconversional methods and master plot procedures. Our model fitted the experimental data well with a deviation of less than 4.5%. Our results show that the addition of 1 wt% PVC to PW decreases the activation energy of hemicellulose and cellulose pyrolysis in PW from 136.3 to 101.6 kJ/mol and from 216.7 to 108.2 kJ/mol, respectively. This demonstrates the importance of acid hydrolysis reactions between the cellulosic fibres of PW and HCl released from PVC dehydrochlorination. Furthermore, we found that a nucleation and growth mechanism best represents the rate-limiting interactions between PVC and PW, which we linked to the formation of metal chloride crystals from acid-base reactions between HCl and PW minerals. Our kinetic model is an improvement of current models for the co-pyrolysis of wood and PVC, and can be readily used in a reactor-scale model of a pyrolyser or gasifier due to its relative simplicity
Préparation, caractérisation et utilisation de catalyseurs biosourcés à base de nickel et de fer pour la production d'hydrogène
No full textInternational audienceThe Paris Agreement (2015) engaged the 195 signatory nations to reduce greenhouse gas emissions. The valorisation of biomass and biowaste into hydrogen and biofuels may play a significant role in reaching this objective, due to the carbon neutrality and high availability of these bioresources. The Water Gas Shift reaction (WGS), which allows enhancing the conversion of syngas into hydrogen, is involved in many thermochemical conversion processes such as gasification, Fisher-Tropsch Synthesis, and biogas reforming. It typically occurs between 180 and 400°C in the presence of noble or transition metal-based catalysts, including iron (Fe) and nickel (Ni) [1] . Commercial catalysts present a high environmental impact and a high dependence on critical or precious metals [2] . To overcome this problem, catalysts may be produced from bioresources with a high metal content, which requires a deep knowledge on the complex and heterogeneous structure of bioresources and on metal catalytic activity. The objective of this work is to prepare, characterize and utilize biosourced catalysts from bioresources rich in catalytic elements for enhancing hydrogen production through WGS reaction. In a first approach, raw fern and willow were selected due to their ability to cumulate heavy metals from soil by phytoremediation. For catalysts preparation, a part of raw biomass was pyrolyzed under N2 from 25°C to 800°C at 2°C/min, followed by an isothermal step at 800°C for an hour. Impregnation was carried out by Wetness impregnation (WI) method for biomass samples and Incipient Wetness impregnation (IWI) for biochar samples [2] , in both cases with an objective of 3 wt%biochar of Ni or Fe. Catalysts were then characterized in terms of organic element content (CHNS analysis), inorganic element content (ICP-AES) and its dispersion on the carbonaceous matrix (SEM, TEM), thermal stability (TGA-DSC), surface area (BET, N2), textural properties and surface chemical groups (TPD, TPR, XRD). Catalysts were then tested in WGS reaction. A fixed bed reactor (8 mm diameter, 25 cm long) was filled with 1.00±0.01 g of each biosourced catalyst and an inert bed of α-Al2O3 [1] . Permanent gases were analysed using an in-line µ-GC/TCD device. A first screening of the catalysts was carried out in Reverse WGS conditions (RWGS) to facilitate reactant introduction in gas phase, the reactor working at 400°C, 3 bar and a ratio of reactants equal to 1:1 (CO2:H2). In a second step, the best catalysts were tested in WGS reaction conditions (220°C, 3 bar, 1:5.5 CO:H2O). Biosourced catalyst structure was compared before and after the chemical reaction. The produced biosourced catalysts showed their thermal stability in TGA up to 500°C. Concerning the impregnation method, WI required 2 to 4 times less time and up to 3 times less metal nitrate than IWI catalysts and therefore biomass impregnation was facilitated. This may be explained by the aromatic nature of biochar leading to formation of Van der Waals bonds between the metals introduced by impregnation and the carbonaceous structure [2] . In the case of biomass, metals make covalent bonds with biomass structure during impregnation, which results in a strong retention of metals by the carbonaceous structure after pyrolysis. Fern biochar IWI impregnated with Ni showed promising activity in RWGS with an initial CO2 conversion of 20% (~35% at equilibrium) and an average selectivity for CO compared to CH4 of 88%. Willow biochar WI impregnated with Fe catalysts lower CO2 conversion, but a higher selectivity to CO (>98%). Non impregnated fern and willow biochars showed lower CO2 conversion with no CH4 production. Future work will test the selected biosourced catalysts in WGS conditions, to identify the key parameters of the biosourced catalysts leading to the maximum syngas conversion rate, as well as H2 productivity and selectivity, as a function of WGS operating conditions
Prévision des arrivées quotidiennes d'appels dans un centre d'appels entrants
No full textNational audienceWhen it comes to managing inbound call centers throughout the days of the week, we are met with a challenge concerning the incertitude revolved around the volume of incoming calls. By training either statistical or neural network forecasting models, we are able to anticipate the number of incoming calls within a certain degree of error. Of crucial importance is determining what type of model to train and how to configure this model’s hyperparameters in an optimized manner. By benchmarking different optimized forecasting models, we were able to generate daily call volume forecasts for an inbound call center
Développement d'un panneau sandwich multifonctionnel à base de mousse expansive pour application en intérieur cabine avion
No full textInternational audienceThe objective of the DEFLECT project (Clean Sky 2, 2018-2020) is to replace the sandwich panels currently used in aircraft electrical cabinets. These pieces are mainly made of aluminium or composite sandwich panels with fiberglass and phenolic resin, with a Nomex® honeycomb core. This second solution, lighter compared to aluminium, however, presents a challenge related to the integration of secondary functions without resorting to sometimes complex and costly machining steps. Therefore, it was decided to develop a new concept of sandwich panel based on expandable foam and composite fiberglass skins. The materials were selected based on their specific fire resistance. The project thus allowed the development of a demonstrator, which underwent a complete characterisation campaign. The study was divided into two parts: initially, the manufacturing process was optimized through several physicochemical tests; then the panel was characterized according to several specifications defined in agreement with the client. The developed panel validated the concept while reducing the overall volume, mass and cost of the electrical cabinet.L'objectif du projet DEFLECT (Clean Sky 2, 2018-2020) est le remplacement des panneaux sandwich, actuellement utilisés dans les meubles électriques d'avion. Ces meubles sont principalement en aluminium ou en panneaux sandwich en composite en fibres de verre et résine phénolique avec une âme en nida Nomex®. Cette deuxième solution, plus légère par rapport à l'aluminium présente cependant une difficulté liée à l'intégration des fonctions secondaires sans passer par des étapes d'usinages, parfois complexes et couteuses. Il a donc été décidé de développer un nouveau concept de panneau sandwich à base de mousse expansive et des peaux en composite en fibre de verre. Les matériaux ont été sélectionnés par rapport à leur résistance au feu spécifique. Le projet a ainsi permis le développement d'un démonstrateur, qui a fait l'objet d'une campagne de caractérisation complète. L'étude a été divisée en deux parties : dans un premier temps le procédé de fabrication a été optimisé à l'aide de plusieurs essais physico-chimiques ; puis le panneau a été caractérisé selon plusieurs spécifications définies en accord avec le donneur d'ordre. Le panneau développé a permis de valider le concept tout en réduisant le volume, la masse et le coût globale du meuble électrique
Combined Effect of Shaking Orbit and Vial Orientation on the Agitation-Induced Aggregation of Proteins
No full textInternational audienceOrbital shaking in a glass vial is a commonly used forced degradation test to evaluate protein propensity for agitation-induced aggregation. Vial shaking in horizontal orientation has been widely recommended to maximize the air-liquid interface area while ensuring solution contact with the stopper. We evaluated the impact of shaking orbit diameter and frequency, and glass vial orientation (horizontal versus vertical) on the aggregation of three proteins prepared in surfactant-free formulation buffers. As soon as an orbit-specific frequency threshold was reached, an increase in turbidity was observed for the three proteins in vertical orientation only when using a 3 mm agitation orbit, and in horizontal orientation only when using a 30 mm agitation orbit. Orthogonal analyses confirmed turbidity was linked to protein aggregation. The most turbid samples had a visually more homogeneous appearance in vertical than in horizontal orientation, in line with the predicted dispersion of air and liquid phases obtained from computational fluid dynamics agitation simulations. Both shaking orbits were used to assess the performance of nonionic surfactants. We show that the propensity of a protein to aggregate in a vial agitated in horizontal or vertical orientation depends on the shaking orbit, and confirm that Brij® 58 and FM1000 prevent proteins from agitation-induced aggregation at lower concentrations than polysorbate 80
An Ant Colony System for the Skilled, Multi-depot VRP with Due Dates and Time Windows
No full textInternational audienceThis article introduces a real-worl maintenance scheduling problem that can be defined as a Skilled Multi-Depot Vehicle Routing Problem with Due Dates and Time Windows, or Skill-MDVRPDDTW, and addresses two methods to solve it. One is a greedy heuristic inspired from the real-world planning processes used in a water service management context, and the other is a version of the Ant Colony System algorithm, widley used in the literature for the Vehicle Routing Problem and its variants and adapted to fit the features of the real-world maintenance problem. Both the problem and the algorithms are positionned in the literature and mathematically formulated, then experiments and results are discussed and compared through a set of various indicators
Conception d'un modèle basé sur les données pour l'évaluation de la planification de la continuité d'activité à l'aide d'indicateurs de robustesse
No full textThe importance of business continuity planning (BCP) has been underscored by the increase in risky events such as natural disasters and pandemics. Despite the critical nature of these plans, evaluating their effectiveness has remained a challenge. This thesis project proposes a data-driven approach to evaluate BCPs through the use of robustness indicators. To address the research question of how to design a data collection process for BCP evaluation, this project follows a three-step approach. Firstly, a comprehensive literature review establishes a state-of-the-art understanding of business continuity management. Secondly, a proposed data collection model considers different types of data, including physical and societal factors, and outlines approaches for their collection. The proposed model incorporates a data-driven approach to ensure accuracy and reliability in the evaluation process. This model is designed to address specific use cases within the scope of BCP, including a pediatric ward and an urban transportation company. Thirdly, a Proof of Concept (POC) analysis is conducted for these use cases to demonstrate the relevance and impact of the collected data on their respective business continuity strategies. The POC involves the design of robustness indicators for the evaluation of BCPs, and the analysis will provide valuable insights about the entire proposed model. The proposed approach aims to measure the ability of BCPs to maintain critical business functions and minimize disruptions. Introducing robustness indicators as part of the evaluation process represents a contribution to the field of business continuity management. The proposed data collection model considers various data types to provide a comprehensive evaluation of BCPs. The POC analysis demonstrates the practical application of the proposed model and the value of robustness indicators in assessing BCP effectiveness. The results of this study were examined through the following cases: (i) a data collection model and its approaches, (ii) a BC assessment grid, (iii) a BC ontology and a knowledge graph, and (iv) dynamic dashboards for presenting patients' and personnel's data along with their sociological profiles. This thesis project aims at improving the BCP by providing a robustness evaluation model. The proposed model has the potential to improve the assessment of business continuity planning and improve organizations' ability to withstand disruptive events. The research findings can be reproduced within organizations seeking to enhance their business continuity planning and risk management strategies.L'importance de la planification de la continuité d'activité a été, consécutivement à l'augmentation d'événements à risque tels que les catastrophes naturelles et les pandémies, reconnue comme essentielle. Bien que cette planification soit discutable, l'évaluation de son efficacité demeure un défi. Cette thèse est consacrée aux données qui servent à évaluer différents Plans de Continuité d'Activité (PCA) à partir de l'utilisation d'indicateurs de robustesse. La question à laquelle tente de répondre cette thèse porte plus spécifiquement sur la conception d'un processus de collecte de données pour l'évaluation de PCAs. Afin de répondre à notre questionnement nous emprunterons un cheminement constitué de trois étapes. La première se présente sous la forme d'une revue systématique de la littérature qui permet de proposer un état de l'art de la gestion de la continuité d'activité. La seconde aborde un modèle de recueil de données qui associe différents types de données, aussi bien des facteurs physiques que sociétaux, et décrit les approches qui permettent de les collecter. Le modèle proposé intègre une approche prioritairement centrée sur les données dans l'objectif de renforcer l'exactitude et la fiabilité du processus d'évaluation. La troisième étape est consacrée à une preuve de concept (POC) effectuée pour démontrer la pertinence et l'impact des données collectées à propos de la continuité d'activité. Le POC implique de concevoir des indicateurs de robustesse pour l'évaluation des PCAs, alors que l'analyse fournit des informations précieuses à propos de l'intégralité du modèle proposé. L'approche retenue vise à mesurer la capacité des PCAs à maintenir les fonctions critiques d'une organisation et à minimiser les perturbations dont elle souffre. L'introduction d'indicateurs de robustesse dans le cadre du processus d'évaluation représente une contribution au domaine de la gestion de la continuité d'activité. Le modèle de collecte de données proposé prend en compte différents types d'indicateurs afin de fournir une évaluation complète des PCAs. L'analyse du POC permet de simuler la faisabilité du modèle proposé et la pertinence des indicateurs de robustesse dans l'évaluation de l'efficacité des PCAs. Les résultats de cette étude ont été traités à travers : (i) un modèle de collecte de données et ses approches, (ii) une grille d'évaluation de la continuité d'activité, (iii) une ontologie de la continuité d'activité et un graphe de connaissances, et (iv) des tableaux de bord dynamiques pour présenter les données des patients et du personnel ainsi que leurs profils sociologiques. Ce travail ambitionne de contribuer à l'amélioration des PCAs en fournissant un modèle d'évaluation de la robustesse. Les résultats de la recherche souhaitent être utiles pour les organisations cherchant à améliorer leur planification de la continuité d'activité et leurs stratégies de gestion des risques
Towards a novel Data Mining System for Medication Error Management
No full textCopyright IEEEInternational audienceMedication errors associated with the Medication Use Process present significant risks and require effective management. However, current practices do not address these risks. There is a lack of awareness that medication errors are unintended risks, and a lack of dedicated systems to manage them. To overcome these limitations, a digital system exploiting massive medical data (big data) is proposed. By integrating various data sources, this system aims to provide adaptable medication errors’ management and continuous improvement. This article presents an overview of the system requirements and highlights the potential of Data Mining in healthcare. Implementing this system could revolutionize medication error management and improve patient safety
Supply Chain Network Design for a New Circular Business: a Case Study in Electric Conversion of ICE Vehicles
No full textInternational audienceThis paper explores the decision support for supply chain network design (SCND) in the context of a new circular business. The number and locations of each supply chain actor, as well as product flows are determined during the SCND process. Different supply chain configurations are evaluated based on economic and environmental aspects using agent-based modeling and discrete event simulation techniques. A case study on retrofit service for Internal Combustion Engine (ICE) vehicles is proposed. The paper provides managerial insights that will assist circular business providers in making informed decisions regarding SCND