HAL Université de Savoie
Not a member yet
59053 research outputs found
Sort by
In situ observation of phase transitions in La2NiO4+δ bulk and thin film samples via Raman spectroscopy
No full textInternational audienceLa2NiO 4+δ has attracted increasing interest in recent years, both as oxygen electrode in solid oxide fuel cells and electrolysers due to its high electrochemical activity at intermediate-to-low temperatures, and as key component of memristive devices for neuromorphic computing, owing to its variable oxygen stoichiometry. The integration of La 2 NiO 4+δ into devices operating at different temperatures and oxygen partial pressures requires knowledge of the effects of hyper-stoichiometry (δ) on its crystalline structure. La 2 NiO 4+δ is known to accommodate oxygen at interstitial sites allowing for large δ values, up to ~ 0.16. In addition, the O-doping -temperature phase diagram is known to be complex, exhibiting several phase transitions with increasing δ. Herein, we use Raman spectroscopy to monitor the effects of O-doping in the phase diagram and the various structures it contains. Throughout this work, we studied this material in its usual ceramic form, as well as in the form of thin films. Results are discussed in terms of phase transitions, chemical expansion, and some of the possible consequences of the low mean grain size inherent to such thin films
Ecoute sismique de la faille de San Jacinto (Californie) avec un réseau dense de longue durée
No full textImproving seismic records near fault zones where large earthquakes nucleate is key to understanding rupture processes such as the link between aseismic slow slip and microseismicity, foreshock origins, and asperity roles in mainshock nucleation. The Parkfield SAFOD project pioneered seismic data collection at depth near the San Andreas Fault, but maintaining borehole instrumentation long-term has proven challenging. We investigate how a dense surface seismic array near an active fault can be maintained over several years and provide novel seismic observations.As part of the ERC-funded FaultScan project, we deployed a long-term (2.5 years), dense (300 stations) seismic array at Piñon Flat Observatory (PFO), near the San Jacinto Fault, one of Southern California's most active faults with potential for high-magnitude earthquakes (M > 7). Our goal was detecting weak signals like microseismicity and hidden tremors, and measuring localized seismic velocity variations. The first chapter describes the experiment and assesses data characteristics and quality. We evaluate how slant-stacking at P- and S-wave velocities toward known seismic events improves signal-to-noise ratio by up to factor 11 compared to a permanent PFO station. We also study how wind noise affects data quality and show that average background noise at 1-10 Hz frequencies is a good wind speed proxy.The second chapter focuses on detecting microseismic events and tremor-type signals. Slant-stacking at P- and S-wave velocities in the San Jacinto Fault azimuth allows extracting 10 times more events than standard catalogs. Analyzing temporal distribution of these additional events, we identified robust seismic activity decrease along the fault lasting several months following Hurricane Hilary, which struck Southern California in late August 2023. While this quiescence origin remains unclear, we favor the hypothesis that increased shallow crustal loading from heavy rainfall (few kPa) perturbed the fault stress field, causing temporary seismicity reduction. We adapted our slant-stacking procedure to extract long-lasting (minutes to hours) seismic tremors by applying one-bit temporal normalization and median filtering to suppress high-amplitude signals. We detected multiple coherent energy bursts, mostly anthropogenic with known origins (freight trains) and identified a previously unobserved persistent, narrow-band, monochromatic signal detectable across regional distances exceeding 100 km, likely from large industrial turbines. During 2.5-year data acquisition, we detected tremor signals from the San Jacinto fault direction with spectral content similar to tectonic tremors (2–8 Hz). Future work will localize these signals and track their temporal evolution, potentially offering new insight into slow-slip episodes.Finally, this thesis explores body wave reconstruction through noise correlation functions, using trains 20 km from the array as seismic sources to infer temporal velocity variations at depth along the San Jacinto Fault. Through full waveform modeling and sensitivity kernel analysis, I examined effects of noise source position variability and near-surface velocity changes on recovered correlations. Modeling shows train position uncertainties can induce travel-time biases of several milliseconds, while P-wave correlations exhibit reduced sensitivity to near-surface perturbations compared to localized fault zone velocity changes.Results demonstrate that long-term, dense seismic arrays are technically feasible and offer valuable opportunities to detect and characterize diverse seismic phenomena. Combining advanced array processing, improved detection techniques, and ability to extract weak and long-duration signals opens new avenues for understanding fault behavior and environmental influences on seismicity.Améliorer les enregistrements sismiques près des zones de faille où naissent les grands séismes est essentiel pour comprendre les processus de rupture, comme le lien entre glissement lent asismique et microsismicité, l'origine des répliques précurseurs, et le rôle des aspérités. Le projet Parkfield SAFOD a été pionnier dans la collecte de données sismiques en profondeur près de la faille de San Andreas, mais maintenir ce type d'instrumentation long terme s'est révélé difficile. Nous étudions comment un réseau sismique dense en surface peut être maintenu sur plusieurs années.Dans le cadre du projet FaultScan financé par l'ERC, nous avons déployé un réseau sismique dense (300 stations) long terme (2,5 ans) à l'Observatoire de Piñon Flat, près de la faille de San Jacinto, l'une des failles les plus actives du sud de la Californie avec un potentiel de séismes de forte magnitude (M > 7). L'objectif était de détecter des signaux faibles comme la microsismicité, des tremors cachés dans le bruit, et mesurer les variations de vitesse sismique. Le premier chapitre décrit l'expérience et évalue les caractéristiques des données. Le stacking aux vitesses des ondes P et S améliore le rapport signal sur bruit jusqu'à un facteur 11 par rapport à une station permanente. Le bruit de fond moyen entre 1 et 10 Hz est un bon indicateur de la vitesse du vent.Le deuxième chapitre se concentre sur la détection d'événements microsismiques et de signaux de type tremor. Stacker dans l'azimut de la faille de San Jacinto permet d'extraire 10 fois plus d'événements que le catalogue standard. En analysant la distribution temporelle de ces événements, nous avons identifié une diminution robuste de l'activité sismique le long de la faille durant plusieurs mois après l'ouragan Hilary (fin août 2023). Nous favorisons l'hypothèse qu'une charge crustale superficielle accrue due aux fortes pluies a pu perturber le champ de contrainte, entraînant une réduction temporaire de la sismicité. Nous avons adapté notre procédure pour extraire des tremors sismiques longue durée en appliquant une normalisation temporelle et un filtrage médian pour supprimer les signaux de forte amplitude. Nous avons détecté plusieurs pics d'énergie cohérents, principalement anthropiques (trains) et identifié un signal persistant, à bande étroite, détectable sur des distances régionales dépassant 100 km, probablement issu de grandes turbines industrielles. Nous avons détecté quelques signaux tremor venant de la direction de la faille avec un contenu spectral similaire aux tremors tectoniques (2–8 Hz).Cette thèse explore aussi la reconstruction d'ondes de volume à travers les fonctions de corrélation de bruit, utilisant des trains à 20 km du réseau comme sources sismiques pour déduire les variations temporelles de vitesse en profondeur. Grâce à la modélisation de forme d'onde complète et l'analyse de noyaux de sensibilité, j'ai examiné les effets de la variabilité de position de la source et des changements de vitesse près de la surface. La modélisation montre que les incertitudes dans la position du train peuvent induire des biais de temps de trajet de plusieurs millisecondes, tandis que les corrélations d'ondes P montrent une sensibilité réduite aux perturbations de surface.Les résultats démontrent que les réseaux sismiques denses long terme sont techniquement faisables et offrent des opportunités pour détecter et caractériser divers phénomènes sismiques. La combinaison de traitement avancé, de techniques de détection améliorées, et de la capacité à extraire des signaux faibles ouvre de nouvelles voies pour comprendre le comportement des failles et les influences environnementales sur la sismicité
High‐Frequency Harmonic Tremor Associated With a Water‐Filled Crevasse at Tête Rousse Glacier, Mont‐Blanc Massif
No full textInternational audienceAbstract Tête Rousse glacier is a small polythermal glacier in the Mont‐Blanc massif (French Alps) that released a large outburst flood in 1892 when a water‐filled intraglacial cavity suddenly drained. A new water‐filled cavity was detected again in the central part of the glacier by a geophysical campaign in 2007. It has been pumped three times since to avoid another catastrophic flood. The volume of water in this central reservoir has decreased recently but recent geophysical surveys suggested that significant volumes of water could be stored in the upper part of the glacier. Here, we describe a seismic tremor signal detected in May 2022 probably generated by a water‐filled reservoir within the glacier. The tremor started on 15 May a few days after temperature first increased above 0°C. Tremor amplitude was stronger in the evening and is correlated with water level measured in a crevasse about 230 m downglacier. The time delay between temperature and tremor or water level is consistent with the time needed for water to infiltrate within the snow and into the glacier. We used different methods to locate this signal both from amplitude decay and from P and S waves arrival times. Both methods provide a similar location near the northern boundary of the glacier. Ground penetrating radar surveys performed in May 2024 have since detected a water‐filled reservoir near this location. These results validate our interpretation of this seismic tremor being produced by changes in water‐level in this reservoir
PEM fuel cell durability under accelerated stress test by voltage cycling: Effect of high operating temperature
No full textInternational audienc
Branching fraction measurement of the decay
No full textInternational audienceA measurement of the branching fraction for the decay is presented using collision data collected by the LHCb experiment at a centre-of-mass energy of 13 TeV. The analysis is based on data recorded between 2016 and 2018, corresponding to an integrated luminosity of . The result is obtained using decays as a normalisation channel. The measured branching fraction is , where the uncertainties are statistical, systematic, and due to the limited knowledge of the normalisation mode branching fraction, respectively. This result improves the precision of the branching fraction measurement by a factor of two compared to the previous best measurement and sets a more stringent bound on lepton flavour universality in quark transitions. It is consistent with previous measurements, and the extracted lepton flavour universality test observable, , agrees with the Standard Model prediction
Evaluation of the adhesion between a catalyst-coated membrane and various types of GDLs for PEMFC
No full textInternational audienceThe proton exchange membrane fuel cell (PEMFC) is a promising technology for decarbonizing some anthropic activities. Many studies haveshown how the choice of components for the membrane electrode assembly (MEA) and the assembly process can influence theperformance and lifetime of the PEMFCs [1, 2]. Although it is of major interest for optimizing the performance and durability ofPEMFCs, few studies have quantified the level of adhesion between the catalyst layer (CL) and the membrane or the gas diffusion layer(GDL). [1 - 5].A previous study showed that applying pressure during the MEA assembly improves adhesion between layers by increasing the interface’sfriction coefficients, but only up to a certain limit [6]. Further investigations are needed to better understandthis phenomenon. In this context, specific experimental tests were developed to determine the mechanical properties of five-layer MEA interfaces. The objective of this study is to evaluate the level of adhesion between a membranecoated or not with a catalyst layer and two types of GDLs (SGL 22BB and Freudenberg H14Cx653). Figure1 shows that the static friction coefficient μs between a GDL and the membrane can vary by more than 50%,and even 100%, depending on a catalyst coated membrane (CCM) and on the type of GDL used. Moreover,the effect of the CL on the friction coefficient is only observed with the GDL 22BB, which exhibited asignificant increase of μs, whereas the μs values are similar for the GDL H14Cx653, regardless of the CL’spresence on the membrane. The evolution of the surface topography with applied mechanical pressure wasalso characterized to better understand the fundamental phenomena taking place at the interfaces. These datawill be useful for analyzing the influence of different types of CL deposition (on the membrane (CCM) or onthe GDL (GDE) as well as the assembly conditions of MEAs on the durability of PEMFCs
Using a seismic network for automatic detection, localization and characterization of mass movements in the Mont-Blanc massif.
No full textInternational audienceIn the Mont-Blanc massif (western European Alps), seismic stations record numerous signals originating from surface mass movements, such as rockslides, rockfalls, and serac avalanches. The large number of recorded signals makes the automation of the processing workflow essential for practical application. These seismic waveforms differ significantly from those generated by earthquakes, making standard algorithms unsuitable for their analysis. The signals typically exhibit an emergent onset, making it challenging to precisely determine their start time. Moreover, the arrival times of P and S waves, routinely used for earthquake localization, cannot be easily identified. The seismic records also vary in length, reflecting the differing durations of the associated phenomena.To analyze such data using a seismic network, we adapted selected algorithms to address these challenges. For detection, we chose the STA/LTA algorithm, and for localization, we used amplitude decay algorithm and BackTrackBB software, which exploits wave field coherence. To test these algorithms, we created a reference dataset consisting of large, well-documented mass movements. The dataset was developed using regional mass movement databases, webcam image analysis, direct observations made by a network of observers, and seismic data from the Sismalp network. This reference dataset enabled us to fine-tune the algorithms and automate the processing of waveforms related to mass movements
