HAL Portal UBS (Université Bretagne Sud)
Not a member yet
14201 research outputs found
Sort by
: Portrait of an Amateur by Himself and Others
No full textInternational audienceBased on his publications and a number of archives, this article analyzes the way É. Rivière defines and presents himself as a researcher. This self-representation is contrasted with the viewpoints of his contemporaries to highlight a series of contrasts and tensions. É. Rivière embodies a polygraph who is increasingly at odds with the imperatives of specialization within prehistory, so much so that the credibility of his work is undermined. Well-connected to the major Parisian institutions, he is poorly integrated into the local networks of amateurs working in the field. This positioning and lack of legitimacy contributed to the many polemics that punctuated the prehistorian's career.the prehistorian’s career.Cet article analyse, à partir de ses publications et de quelques archives, la manière dont É. Rivière se définit et se met en scène en tant que chercheur. Cette autoreprésentation est confrontée aux points de vue de ses contemporains afin de mettre en lumière une série de contrastes et de tensions. É. Rivière incarne une figure de polygraphe en décalage croissant avec les impératifs de spécialisation qui s'imposent au sein de la préhistoire, si bien que la crédibilité de ses travaux en est fragilisée. Bien connecté aux grandes institutions parisiennes, il est mal inséré dans les réseaux locaux d'amateurs qui oeuvrent sur le terrain. Ce positionnement et ce déficit de légitimité favorisent les nombreuses polémiques qui ponctuent la carrière du préhistorien
Bridging Arbitrary and Tree Metrics via Differentiable Gromov Hyperbolicity
No full textInternational audienceTrees and the associated shortest-path tree metrics provide a powerful framework for representing hierarchical and combinatorial structures in data. Given an arbitrary metric space, its deviation from a tree metric can be quantified by Gromov's δhyperbolicity. Nonetheless, designing algorithms that bridge an arbitrary metric to its closest tree metric is still a vivid subject of interest, as most common approaches are either heuristical and lack guarantees, or perform moderately well. In this work, we introduce a novel differentiable optimization framework, coined DELTAZERO, that solves this problem. Our method leverages a smooth surrogate for Gromov's δ-hyperbolicity which enables a gradient-based optimization, with a tractable complexity. The corresponding optimization procedure is derived from a problem with better worst case guarantees than existing bounds, and is justified statistically. Experiments on synthetic and real-world datasets demonstrate that our method consistently achieves state-of-the-art distortion
Estimation of compressive strength in thick continuous fibre composites: Quantifying the role of initial fibre misalignment
No full textInternational audienceThe present study investigates the compressive strength of continuous fibre composites. These materials comprise almost ten different fibre/matrix pairs with four epoxy-based thermoset resin systems, three different carbon fibres and three manufacturers. The experimental values obtained from flexural testing on thick laminates (with 10 samples for each pair) are compared to an analytical compressive failure criterion. This latter model incorporates ease of use in engineering design and mechanistic contributions, including material (non-linear shear behaviour of the ply), manufacturing (initial fibre misalignment) and structural effect (loading, neighbourhood of ply). A comparison between the experimental and estimated values, using the standard deviation of the frequency distribution of fibre misalignment, demonstrates a satisfactory or excellent outcome. The comparison for a pair produced less favourable results, with greater discrepancies observed</div
Patrimoine en mouvement : le corps archive du danseur
No full textInternational audienceHow can we preserve living heritage, particularly when it involves bodies in motion? In dance, this question intersects with the challenges of building a contemporary repertoire and archiving. Living memory is preserved through the kinetic memory of performers. While bodies archive dance, the art form and its underlying choreographic intentions leave other imprints: writings, scores, recordings, costumes, and more. Les Carnets Bagouet are pioneering in their approach to transmission, preserving the memory of the late choreographer’s work. Traces of all kinds, bearing witness to the context and process of creation or the dissemination of the work, become the focus of archival research. By examining the case of the Centre chorégraphique national de Montpellier (National Choreographic Centre of Montpellier) and the choreographers Dominique Bagouet and Mathilde Monnier, who successively directed it, this article aims to articulate the concept of the “body-archive” within a broader framework of archival reflections, including “the two bodies of the archive”. This approach enables us to consider the archiving of a work not only retrospectively, in the context of reenactments, reappropriations, and reinterpretations, but also during the creation process itself, taking into account the unique challenges of choreographic art and the potential future uses of the archived material.Comment faire patrimoine du vivant, qui plus est quand il s’agit de corps en mouvement ? En danse, cette question croise celles de la constitution d’un répertoire contemporain et de mise en archives. La conservation de la mémoire vivante s’opère via la mémoire kinésique des interprètes. Les corps archivent la danse, mais la danse et l’intention chorégraphique qui y préside laissent d’autres empreintes : écrits, partitions, captations, costumes etc. Les Carnets Bagouet font figure de pionniers dans l’entreprise de transmission. Ce sont les passeurs de mémoire de l’œuvre du chorégraphe disparu. Les traces de toute nature témoignant du contexte et du processus de création ou de la diffusion de l’œuvre deviennent lieu de questionnements archivistiques. En développant l’exemple du Centre chorégraphique national de Montpellier et des chorégraphes Dominique Bagouet et Mathilde Monnier qui l’ont dirigé, cette proposition entend articuler le concept du corps archive dans un ensemble plus vaste de réflexions archivistiques ainsi que sur « les deux corps de l’archive » permettant de mettre en perspective l’archivage d’une œuvre non pas seulement a posteriori quand il s’agit d’entreprise de reenactment, de réappropriations, de relectures, mais également au moment de la création en tenant compte des enjeux de l’art chorégraphique et des possibles desseins envisagés
Commémorer à défaut d’inhumer. Patrimonialisations transfrontalières des défunts de la guerre d’Espagne dans la région de Saragosse
No full textInternational audienc
Airborne wind energy: the KEEP project *
No full textA ground-generation airborne wind energy system is considered. The device consists of a kite attached to an oscillating arm in the horizontal plane that generates electricity. The kite's trajectory is constrained to follow a prescribed horizontal lemniscate pattern. An implicit modeling approach is adopted where the control force required to maintain the eight-shaped trajectory is computed. This system is reduced to a two-dimensional second-order ordinary differential equation. Its limit cycles are studied numerically, with specific equilibria serving as starting points to obtain physically admissible periodic motions. These cycles depend on several design parameters including line length, arm inertia, and braking coefficient, that are tuned to maximize the average power generated by the arm's oscillations. The system's behavior is further analyzed across a typical range of wind speeds, demonstrating robustness to wind variations while maintaining positive line tension throughout the cycle
Climate change metrics: bridging IPCC AR6 updates and dynamic life cycle assessments
No full textInternational audienceClimate change metrics result from analytical simplification of complex and diverse climate models that are generally not deeply investigated by Life Cycle Assessment communities. We investigated the Sixth Assessment Report of the Intergovernmental Panel on Climate Change to properly gather updated metric equations, climate parameters and associated uncertainties. Metrics are mainly designed with a single gas pulse emission at t0 whereas multi-gas and multi-time pulse emissions are mostly encountered in LCA modelling. Therefore, common static and relative metrics might not suit dynamic climate change assessments (dCCA) that differentiate pulse timing and gas contributions over time. This study focuses on absolute and dynamic metrics – cumulative radiative forcing (AGWP or ΔF) and global temperature change (AGTP or ΔT) – applied to well-mixed greenhouse gases. Cumulative radiative forcing assessment at 20, 100, 500 years appears sufficient. Global temperature change metrics have some advantages that offset their higher uncertainties. (1) Degree Celsius unit better suits peak warming targets. (2) Positive and negative peaks, as well as long-term temperature change, partly alleviate the time horizon decision issue while assessing product systems. (3) Graphical representations are comparable to simultaneously depict short- and long-lived climate forcers. In future assessment reports, IPCC is invited to recall climate equations and updated parameters values in a pedagogical way and to adopt peak and long-term temperature change metrics. dCCA recommendations are to plot ΔF and ΔT temporal profiles of product systems up to 600 years and use suggested metrics. This should enable going towards climate neutrality with more clarity, transparency and understanding
Machine Learning Predictions for the Comparative Mechanical Analysis of Composite Laminates with Various Fibers
No full textInternational audienceThis article addresses the complex behavior of composite laminates under varied layer orientations during tensile tests, focusing on carbon fiber and epoxy matrix composites. Data characterizing the mechanical load behavior are obtained using twelve composite laminates with different layer orientations and the DIGIMAT-VA software (version 2023.3). First, these data were used to elaborate a complex comparative analysis of composite laminates from the perspective of materials science. Composite laminates belong to three classes: unidirectional, off-axis oriented, and symmetrically balanced laminates, each having a specific behavior. From the perspective of designing a new material, a prediction model that is faster than the finite element analysis is needed to apply this comparative analysis’s conclusions. As a novelty, this paper introduces several machine learning prediction models for composite laminates with 16 layers arranged in different orientations. The Regression Neural Network model performs best, effectively replacing expensive tensile test simulations and ensuring good statistics (RMSE = 34.385, R2 = 1, MAE = 19.829). The simulation time decreases from 34.5 s (in the case of finite element) to 0.6 s. The prediction model returns the stress–strain characteristic of the elastic zone given the new layer orientations. These models were implemented in the MATLAB system 2024, and their running proved good models’ generalization power and accuracy. Even specimens with randomly oriented layers were successfully tested
Goodness-of-fit testing of the distribution of posterior classification probabilities for validating model-based clustering
No full textWe present the first method for assessing the relevance of a model-based clustering result in both parametric and non-parametric frameworks. The method directly aligns with the clustering objective by assessing how well the conditional probabilities of cluster memberships, as defined by the mixture model, fit the data. By focusing on these conditional probabilities, the procedure applies to any type and dimension of data and any mixture model. The testing procedure requires only a consistent estimator of the parameters and the associated conditional probabilities of classification for each observation. Its implementation is straightforward, as no additional estimator is needed. Under the null hypothesis, the method relies on the fact that any functional transformation of the posterior probabilities of classification has the same expectation under both the model being tested and the true model. This goodness-of-fit procedure is based on a empirical likelihood method with an increasing number of moment conditions to asymptotically detect any alternative. Data are split into blocks to account for the use of a parameter estimator, and the empirical log-likelihood ratio is computed for each block. By analyzing the deviation of the maximum empirical log-likelihood ratios, the exact asymptotic significance level of the goodnessof-fit procedure is obtained.</div
Macroevolutionary Divergence along Allometric Lines of Least Resistance in Frog Hindlimb Traits and Its Effect on Locomotor Evolution
No full textInternational audienceUnderstanding whether and why microevolutionary patterns of trait covariation match macroevolutionary divergence is essential for linking evolution at different timescales. However, recent work has focused on developmental constraints for alignment between intraspecific variation and divergence, neglecting a potential role of natural selection on function to connect these scales. Here, we compare the support for the selection and constraint hypotheses to explain both phenotypic trait covariation and species divergence. To test these hypotheses, we collected data on hindlimb and jumping performance traits within and across species of two frog genera. We compared patterns of within-species phenotypic variation (the P matrix) with divergence and selective covariance matrices, from which we could extract the major axes of the realized adaptive landscape (AL), the directions in which adaptive peaks shifted the most over evolutionary time. We also tested whether the major axes of the AL were related to selection on jumping performance. We found high alignment between patterns of variation across scales. Most divergence occurred in allometric size, defined as the first eigenvector of the P matrix. However, jumping performance gradients were unaligned with the major axes of the AL and the P matrix. Across species, however, evolution of maximum acceleration showed a strong negative relationship with changes in allometric size. We infer that the jumping peak evolved under fluctuating selection, and species have tracked the peak along the direction of most within-species variation, allometric size. We conclude that long-term hindlimb divergence was constrained by developmental interactions among traits associated with growth and not net directional selection. Nonetheless, divergence on size indirectly influenced jumping evolution.</div