Olivier Castelnau

The phase stability, crystal structure, and melting of Nb have been examined under high pressure shock compression to 365 GPa using ultrafast x-ray diffraction measurements on two x-ray free electron laser facilities. On compression, Nb remains stable in the bcc phase up to 220 GPa, with coexistence of bcc-Nb and liquid from 249 to 298 GPa, and complete melting at 301 GPa, with melt identified by diffuse liquid diffraction. Melting initiates at higher pressure than expected based on theoretical predictions, and the data consistently excludes the presence of other suggested phases of Nb at these pressures, including and hcp, thereby resolving the long-standing structural discrepancy. Singh-type strength analysis provides experimental evidence for a change in sign of the elastic anisotropy parameter at 121 GPa, consistent with the previously reported heat-induced hardening to heat-induced softening transition in Nb.

The development of high-throughput X-ray diffraction (XRD) techniques presents significant challenges to traditional data analysis frameworks. The present work demonstrates that deep convolutional neural networks (CNNs) can reliably quantify key microstructural parameters of polycrystalline materials, specifically crystallite size and heterogeneous strain (microstrain) extracted from powder XRD patterns, within milliseconds. A CNN was trained using synthetic data generated with a dedicated code based on the GSAS-II software. We employ the example of monoclinic ZrO2 to critically examine the theoretical performances of the CNN. The accuracy of the predictions was systematically investigated within a crystallite size range of 5 – 1000 nm and a microstrain range of 0.05-2%. In this context, the role of the resolution of the diffractometer and the XRD peak profile shape is discussed in details. For instance, on a low-resolution laboratory diffractometer it is possible to determine the crystallite size within a 5–100 nm range with 99.7% accuracy in the absence of microstrain, and 97.3% accuracy when microstrains are present. Higher accuracies are obtained with a high-resolution diffractometer and over an extended range of size and microstrain. The potential of the CNN is demonstrated by the analysis of the crystallite size and microstrains in MgAl2O4 formed by a solid-state reaction between MgO and Al₂O₃ using in situ XRD at 1200 °C performed at the ESRF. The CNN can determine the crystallite size and microstrain in 0.004 s per diagram with values that deviate by ~7% and ~8% from a conventional Rietveld refinement, respectively, which is contained within the statistical uncertainties of these parameters. We further demonstrate that the same training dataset can be used for other regression problems, for instance phase fraction quantification, with no additional computational cost and only minor modifications to the network architecture. This work paves the way for real-time data analysis at synchrotron facilities.

Marine-terminating glaciers play a significant role he loss of ice mass from Greenland. Iceberg calvings are thought to account for about half of Greenland ice mass loss. These events produce seismic waves (glacial earthquakes) recorded by global seismic networks and contain information such as the iceberg volume and the forces acting during the event. Previous work showed that seismic inversion, coupled to numerical modeling, can be used to decipher the glacial earthquake signal [Sergeant 2019]. However, the lack of consideration for the hydrodynamic forces applied onto the glacier leads to large uncertainties in the iceberg volume estimations. Therefore, based on previous work, a Computational Fluid Dynamics (CFD) model is employed to model the complex fluid/structure interaction between a capsizing iceberg and the ocean. The simulations match results from laboratory experiments with great accuracy (rotation kinematics, effect of calving type, hydrodynamic pressure, etc.) [De Pinho Dias 2025].In this presentation, we will show how the forces applied onto a simple model of Helheim glacier during an iceberg calving depend on geometrical parameters (iceberg height, aspect ratio, water depth, iceberg drop height). The pressure force exerted onto the glacier front has a significant magnitude of more than 50 % of the iceberg/glacier contact force which acts in the opposite direction.The forces are then converted into seismic signals and show a very good match with the signal recorded at 8 stations in Greenland.In addition, we will show the path of particle tracers advected by calving-induced water currents. Sergeant, A. et al. (2019) ‘Monitoring Greenland ice sheet buoyancy-driven calving discharge using glacial earthquakes’, Annals of Glaciology, 60(79), pp. 75–95. doi:10.1017/aog.2019.7. De Pinho Dias, N., Leroyer, A., Mangeney, A. and Castelnau, O., 2025. Fluid-structure modeling of iceberg capsize. Ocean Engineering, 336, p.121765. doi:10.1016/j.oceaneng.2025.121765

Marine-terminating glaciers in shallow, cold Greenland fjords are not in direct contact with warm and deeper Atlantic Water. In these fjords, stratification near glacier termini is primarily driven by freshwater inputs. Surface runoff enhances near-surface stratification, while subglacial discharge drives localized mixing at depth. Meanwhile, the role of calving events as drivers of fjord-scale mixing remains poorly constrained. Here, we combine hydrographic measurements and oxygen isotope data with remote sensing and three-dimensional computational fluid dynamics simulations to resolve the immediate fjord response to a major calving event. We show that a large, buoyancy-driven calving event triggers rapid vertical mixing that redistributes water masses throughout the upper and intermediate water column, entraining Polar Water into the surface layer and inducing episodic contributions from deeper fjord waters. This mixing results in a salinity increase in the upper ∼ 25 m, counteracting the expected summer freshening from glacier meltwater. Oxygen isotope (δ18O) profiles and simulations further indicate episodic inputs from deeper water masses. Together, these results demonstrate that calving events can exert significant, and previously poorly constrained, controls on mixing in cold, shallow glacier–ocean systems.

Iceberg calving is a complex process often followed by the capsize of the newborn iceberg because of the torque created by the buoyancy and gravity forces. In the case of kilometer-scale icebergs, calving/capsize events can trigger seismic waves (glacial earthquakes) recorded hundreds of kilometers away by global seismic networks. These recordings contain information on the seismic source such as the calved-iceberg volume as well as the contact force applied by the iceberg on the glacier, the glacier dynamical response but also water waves and flow following the capsize.To obtain an accurate estimation of the iceberg volume, it is necessary to couple seismic inversion of glacial earthquakes with numerical modeling of the capsize [Sergeant 2019]. Therefore, based on our previous work, we use a Computational Fluid Dynamics (CFD) model to simulate the fluid-structure interaction between the ocean and an iceberg capsizing against a glacier terminus. The model reproduces with great accuracy lab experiments (rotation kinematics, effect of calving type, hydrodynamic pressure, etc).In this talk, we will focus on field-scale simulations. We will show that the forces applied on the glacier terminus due to the hydrodynamic pressure and to the iceberg-glacier contact appear to have similar amplitudes, are of opposite signs, and depend on the iceberg geometry. We also give a first estimation of the glacier deformation under the action of these forces. Our CFD simulations make it possible to compute the water velocity field in which values exceed 10 m/s close to the 200-meters-high-iceberg capsize.Sergeant, A. et al. (2019) ‘Monitoring Greenland ice sheet buoyancy-driven calving discharge using glacial earthquakes’, Annals of Glaciology, 60(79), pp. 75–95. doi:10.1017/aog.2019.7.

The transport of heat from the interior of the Earth drives convection in the mantle, which involves the deformation of solid rocks over billions of years. Significant advancements have been made over recent years to study lower mantle assemblages under relevant pressure and temperature conditions, which have confirmed the usual view that ferropericlase is weaker than bridgmanite. However, natural strain rates are 8 to 10 orders of magnitude lower than those observed in the laboratory, and remain inaccessible to us. Once the physical mechanisms of the deformation of rocks and their constituent minerals have been identified, it is possible to overcome this limitation thanks to multiscale numerical modeling, which allows for the determination of rheological properties for inaccessible strain rates. This presentation will demonstrate how this theoretical approach can be used to describe the elementary deformation mechanisms of bridgmanite and periclase. These descriptions are compared with available experimental results in order to validate the theoretical approach. In a subsequent phase, the impact of very slow strain rates on the activation of the aforementioned mechanisms is evaluated. Our findings indicate that significant alterations in deformation mechanisms can occur in response to changes in strain rate.

One of the major questions in climate science is to improve the accuracy of sea-level rise prediction, for which mass loss of the polar ice caps has a significant contribution. In this work, the focus is on buoyancy-dominated capsize of large icebergs. The capsizes generate specific seismic signals, which in turn can be analysed and used as a unique tool to study the long term evolution of such large icebergs capsize and the glacier response.To better quantify ice mass loss due to iceberg calving at marine terminating glaciers, coupling iceberg calving simulation and inversion of the seismic waves generated by these events and recorded at teleseismic distances is necessary. To achieve our task, a complex fluid/structure model of the iceberg capsize is required to obtain accurate forces history acting on the glacier terminus. The simulated forces can then be compared to the force inverted from the seismic signal. Therefore, based on our recent work, we implement a Computation Fluid Dynamics (CFD) approach to reach a high fidelity modelling of the iceberg capsize. First work using the experimental data of an iceberg capsize showed the need and ability of CFD computations to precisely reproduce the iceberg kinematics for different cases. We will present more advanced CFD configurations, including the contact between the capsizing iceberg and a rigid glacier front. Computation results are compared and validated against lab scale experiments, where we outline that some 3D effects cannot be neglected. We will also present full scale capsize simulations, in which the mixing of ocean layers occurs. In particular, we will quantify the transport of particles within the ocean to illustrate the potential change of nutriments distribution or of pressure experienced by local fauna due to iceberg calving.

Rocks from the Earth mantle and polar ices have in common a nonlinear rheology and low crystal symmetries leading often to a limited number of independent slip systems for the glide or climb of dislocations. Both deform at elevated homologous temperatures, mostly under creep. Very large plastic deformation occurs during large scale geophysical flows, leading to pronounced crystallographic texture and an associated anisotropic rheology. Polar ice is a pure material, whereas several mineral phases are present simultaneously the mantle. The mantle deforms at extremely slow strain-rates, 10 orders of magnitude smaller than standard laboratory strain-rates, and thus the estimation of the mantle behaviour requires a drastic extrapolation from lab data. A consequence of the features outlined above is that deformation of mantle rocks or polar ices leads to a strong heterogeneity of the stress and strain-rate fields inside the polycrystalline aggregates, at the intragranular (micron) scale. This field heterogeneity has strong implication in terms of texture evolution, recrystallization, but also on the effective flow stress. Another consequence is that simple or ad-hoc micromechanical models are often inaccurate when the goal is to estimate the in situ nonlinear and anisotropic rheology, and the microstructure evolution at large strain, as the activation of slip systems is highly sensitive to stress fluctuations. In this presentation, we will review existing mean-field models for polycrystalline aggregates, show their capabilities / limitations with respect to reference full-field solutions, and show the benefit of the fully-optimized second order self-consistent scheme recently proposed by Song and Ponte Castañeda [2018]. Examples for ice and few mantle minerals will be given for illustrative purpose. D. Song and P. Ponte Castañeda, Fully optimized second-order homogenization estimates for the macroscopic response and texture evolution of low-symmetry viscoplastic polycrystals, Int. J. plasticity 110 (2018), 272–293