Publications

Absolute accuracy of industrial robot is required for most of industrial applications. However, positioning errors of several millimeters are induced by many factors. Hybrid calibration, combining analytical model and learning-based regression, can compensate for most of the positioning error, including payload effects. However, when the payload changes, hybrid calibration has to be performed again. In this paper, hybrid calibration is applied on an industrial robot in two different sub-workspaces, with two different payloads. The results of this method have been compared to other calibration approaches, and highlight that hybrid calibration provides a higher final accuracy. Moreover, two data-efficient and pragmatic approaches are proposed, to address the issue of changing payload. Both methods are based on hybrid calibration. The first one uses previously-acquired knowledge to drastically reduce the number of measurements necessary to update a trained learning model with another payload. The second one uses a model trained separately for two different payloads and interpolates the outputs to compensate for new payloads without any additional measurement. The datasets used are available at: https://doi. org/10.57745/DWUC0H. The methods have been experimentally validated using a compensation algorithm and compared to other approaches, and show that the positioning error can be reduced by 95%.

This study models the evolution of lithium demand in France’s electric vehicle sector through 2100, using system dynamics to account for the interplay of extraction, recycling, and production projects. Following the European Union's 2023 decision to ban internal combustion engine vehicles by 2035, France has focused on securing critical raw materials, including lithium, to support its electrification goals. The system dynamics model incorporates the growth of the French electric vehicles fleet, lithium extraction from a newly established mine, and recycling rates mandated by European regulations. The results compare the scenario with the simulation and show when the electric vehicle industry will not be self-sufficient in lithium.

Research on nanocellular foams is motivated in part by the promise of physical properties, in particular mechanical properties, that can go beyond the classical mechanical framework. However, due to the difficulty in obtaining foams of a given density but different cell sizes, determining the effect of cell size on the mechanical properties of polymer foams remains a challenge. To overcome this difficulty, studies on the mechanical behaviour of mesocellular, microcellular and nanocellular polymer foams have been compiled in this review article. After describing the different cellular structures between meso-, micro- and nanocellular foams, the mechanical properties are examined as a function of relative density and cell size. It is shown that for small strains and at low strain rates, nanocellular foams exhibit mechanical behaviour predicted by the Gibson and Ashby model. Relative density remains the first important factor to be taken into account when studying the Young’s modulus and buckling stress of nanocellular foams. The focus then shifts to fracture properties, as microcellular foams have already been shown to be far superior to more conventional foams. As studies are still scarce and different methodologies have been used, no general conclusions can be drawn. However, the fracture and impact properties could be greatly improved by this change in scale. The local confinement of molecular chains in polymeric nanocellular foams or the relaxation of the triaxial stress state in front of the crack tip could explain these observations.

The capitalisation on and transfer of technological, engineering and scientific knowledge associated with empirical know-how is an important issue for the sustainability and development of manufacturing. Indeed, certain sectors of industry are facing the increasing ageing of the labour force, recruitment difficulties and high staff turnover, leading to a loss of knowledge and know-how. In a context of numerical and digital transition and the migration of processes to industry 4.0, one of major challenges manufacturers face today is their capacity to build intelligent platforms for acquiring, storing and transferring their know-how and knowledge. It is crucial to create new media and tools for staff training and development capable of capturing knowledge and reusing it to create a project history through expertise and data collection. This paper presents the methodology and guidelines for implementing electronic knowledge books (eK-Books), along with their uses. The eK-Book is a semantic web-based hypertext medium (channel) allowing stakeholders to capitalise on, structure and transfer knowledge by using concept maps, process maps, influence graphs, downloadable documents, web pages and hypermedia knowledge sheets. They are intended for engineers, expert or novice technicians, manufacturers, sector coordinators and plant managers, as well as trainers and learners. They are usable and manageable in all types of environments and with different levels of accessibility. This paper highlights (1) the transfer knowledge capacity of eK-Books and (2) their usability in two agri-food sectors namely (1) the cheese sector with protected designation of origin (PDO) and protected geographical indication (PGI), and (2) the butchery and cold meat sectors.

The derivation of the Navier–Stokes equation in continuum mechanics leads to a number of consequences which are discussed in depth. In spite of its very high representativity of real flows, this equation presents some artefacts due to the whole notion of the continuous medium. An alternative to the Navier–Stokes equation is proposed, based on the conservation of energy per unit mass instead of momentum. The classical inertial frame of reference is replaced by a set of local frames of reference where interactions are treated as cause and effect. Invoking the principle of equivalence between energy and mass, the latter is eliminated from the quantities used in this new formalism. All quantities, variables and physical properties are thus expressed in units of mass. The law of motion is established in the form of the conservation of acceleration, an energy per unit of mass and length. The acceleration is thus written in the form of a Helmholtz–Hodge decomposition, in two terms, the first curl-free and the second divergence-free as a function of two potentials, scalar and vector. Maxwell’s idea of federating the laws of electrodynamics and magnetism to establish electromagnetism is taken up here to establish the new law of motion as a nonlinear wave equation. This approach makes it possible to demonstrate that this law is relativistic from the start. The form of the equation of motion in two Lagrangians gives access to symmetries related to the conservation of certain quantities according to Noether’s theorem.

AbstractMicrophysiological systems (MPSs) reconstitute tissue interfaces and organ functions, presenting a promising alternative to animal models in drug development. However, traditional materials like polydimethylsiloxane (PDMS) often interfere by absorbing hydrophobic molecules, affecting drug testing accuracy. Additive manufacturing, including 3D bioprinting, offers viable solutions. GlioFlow3D, a novel microfluidic platform combining extrusion bioprinting and stereolithography (SLA) is introduced. GlioFlow3D integrates primary human cells and glioblastoma (GBM) lines in hydrogel‐based microchannels mimicking vasculature, within an SLA resin framework using cost‐effective materials. The study introduces a robust protocol to mitigate SLA resin cytotoxicity. Compared to PDMS, GlioFlow3D demonstrated lower small molecule absorption, which is relevant for accurate testing of small molecules like Temozolomide (TMZ). Computational modeling is used to optimize a pumpless setup simulating interstitial fluid flow dynamics in tissues. Co‐culturing GBM with brain endothelial cells in GlioFlow3D showed enhanced CD133 expression and TMZ resistance near vascular interfaces, highlighting spatial drug resistance mechanisms. This PDMS‐free platform promises advanced drug testing, improving preclinical research and personalized therapy by elucidating complex GBM drug resistance mechanisms influenced by the tissue microenvironment.

Molten salt–based processes and hydrofluxes are highly sensitive to mixture composition and require knowledge of the combined melting point for successful materials syntheses. In particular processes using hydroxide–based fluxes (pure salt melts) and hydrofluxes (salt melts containing 15–50% HO) have been shown to be interesting environments to synthesize inorganic materials in high oxidation states. The development of tools to predict these properties is desirable to inform the implementation of processes using these mixtures. In this work, we use an artificial neural network model to estimate the melting points of fluxes and hydrofluxes comprising of quaternary mixtures of NaOH, KOH, LiOH, and H2O. A database of 1644 data points collected from 47 different sources was used in the training of the model. Melting points were predicted from the molar fractions of each component (4 independent variables)...

For the crystallization of an API in supercritical CO2, a two – step nucleation mechanism involves the apparition of metastable liquid droplets in the vapour phase composed of the API dissolved in the CO2, before crystallization. To find out the pressure and temperature conditions such a two – step mechanism could be observed, we studied the stability / metastability / instability for {(S)-Naproxen + CO2} and {(RS)-Ibuprofen + CO2} vapour binary mixtures. Thermodynamic computations proposed in the paper, have shown that a mixture of API and CO2 at elevated pressures can be unstable and/or metastable with respect to a liquid-vapour equilibrium and at the same time with respect to a solid-vapour equilibrium. Depending on the degree of supersaturation, such a mixture can potentially first decompose via spinodal decomposition into coexisting liquid and vapour phases, which turn due to nucleation and growth theory to a solid-fluid equilibrium.

With increasing global ecological and environmental pollution problems, synthetic fiber reinforcements cannot meet recent environmental goals and sustainable development requirements. Under these circumstances, the manufacturing of continuous natural fiber reinforcements filled with biodegradable and recyclable polymer composites began to receive more attention. As one of the fastest-growing processes in the last decade, Fused Filament Fabrication (FFF) shows great potential in printing continuous natural fiber-reinforced polymer composites (CNFRCs). However, several challenges, such as low mechanical properties and difficulty in printing, need to be overcome. This chapter attempts to analyze the current research status of using CNFRCs as a feeder for 3D-printing and provides a research outlook on the existing technology landscape, influencing material and printing parameters as well as reinforcement geometry and stacking orientation.

Purpose: In the context of the Anthropocene and the widespread use of ICT, the growth of digital e-waste should be tackled. In this paper, we examine the role of the of end-of-life (EoL) in relation to the other life cycle stages in life cycle assessments (LCA) of digital equipment. We investigate how LCAs of digital equipment model the EoL, what the results are for the environmental impacts of EoL treatments, and if the LCAs published after the ETSI standard and the ITU-T 1410 recommendation follow EoL modeling criteria. Method: We did a systematic literature review with three main research criteria. The LCA must (1) concern digital equipment or compare digital devices, (2) consider several impact categories, and (3) be from cradle-to-grave. As the number of scientific papers found was relatively small, we included LCAs of manufacturers, master thesis, and technical reports when they were appropriated. We found twenty-six references from academic and industrial sources corresponding to our research criteria. Results and discussions: Our review is structured according to the ISO 14040-14044 standards. EoL modeling is mainly in the scenario description, which is the first step of the LCA. For EoL modeling, we examine the allocation procedure, scenarios, distribution of scenarios, data source, transport modeling, EoL treatment rates, purity and quality of recycled materials, treatment of specific and hazardous components, and informal flow modeling. We identify the missing elements in the studies and compare them to ETSI and ITU-T 1410. We find it unclear whether ETSI and ITU-T 1410 are being followed in LCAs, as there are no noticeable changes in the EoL modeling. EoL often generates low environmental impacts compared to the other life cycle stages, except for some impact categories. Choices made in EoL modeling, such as using substitution with avoided impact approach or unspecified allocation, can result in low-impact estimates. We underline the need for more transparency in EoL allocation modeling to enable comparisons and interpretation of results from clear reporting. Conclusion: There is no clear consensus on how to model EoL in an LCA. As the EoL of a digital device is highly uncertain, modeling needs to be more consistent and detailed. LCA practitioners should go beyond these guidelines. We identify several missing elements in current LCAs. We provide recommendations for future LCAs, including providing more detail on the substitution approach, considering informal flows, using primary data, and implementing a hybrid methodology.