Rheological and mechanical properties of mono-disperse hard-sphere silica suspensions (Ludox® HS-... more Rheological and mechanical properties of mono-disperse hard-sphere silica suspensions (Ludox® HS-40) are investigated as a function of particle volume fraction (φp ranging from 0.22 to 0.51), using shear rate tests, oscillatory methods, indentation and ultrasonic technique [1]. Four regimes are identified with increasing particle content; they are related to the existence and the behavior of the electrical double layer (EDL) around each particle. Semidiluted suspensions (0.22 ≤ φp ≤ 0.30), which do not flocculate due to the strong interparticle repulsions, show Newtonian behavior (I). Their viscosity is higher than predicted by classical models, due to the EDL, which effectively increases the particle radius. At higher particle contents, the suspensions exhibit two viscoelastic (VE) shear-thinning regimes separated by a critical particle volume fraction. It discriminates semi-concentrated fluids (II, 0.31 ≤ φp ≤ 0.35), in which particles form aggregates of quickly increasing size, from concentrated (III 0.35 ≤ φp ≤ 0.47) strain sensitive samples with paste-like aspect and high elastic modulus. Regime II can be explained by a sudden partial collapse of the EDL as it is progressively squeezed with increasing φp. At higher particle concentration (IV, φp = 0.51), the suspensions undergo a glass transition and behave as an elastic-brittle solid with high Young’s modulus. The particles are trapped within the cages constituted by their neighbors, and the structure is at arrest.
<div> <div> <div> <p... more <div> <div> <div> <p>When a fluid can experience a "jammed" state, it will flow only when the local deviatoric stress becomes greater than a critical stress, the so-called  "yield-stress". Jamming can be caused by entangled dislocations in a mineral, or by the existence of a hard skeleton in a two-phase fluid. According to recent numerical modeling, a Bridgmanite lower mantle would predominantly deform by pure dislocation climb; and due to dislocations interactions, it would flow only for local deviatoric stress greater than a critical yield stress which depends on dislocation density. In a first set of fluid mechanics experiments in such a visco-plastic fluid, we showed that hot plumes would develop with a much thicker morphology than in newtonian fluids. Scaling laws derived from the experiments tightly relate the buoyancy and diameter of the hot plumes to the value of the yield-stress, as well as to the mantle microstructure (such as dislocation density and vacancy concentration). Yield stress values between 1 and 10 MPa, implying dislocation densities between 10<sup>8</sup> and 10<sup>10</sup> m<sup>−2</sup>, would be sufficient to explain the thick plumes morphology observed in seismic tomography images; while low vacancy concentrations could explain the 1000 km-depth horizon also seen in tomography. </p> <p>In a second set of experiments, we show that the existence of a yield stress in a Bridgmanite lower mantle will also act as a filter regarding slab penetration in the lower mantle. Given slab buoyancy, a typical slab, 100 km-thick, could not overcome the lower mantle yield stress. So most of single slabs would be expected  to stagnate in the transition zone. However a pile of folded slab with a typical thickness around 400 km would have sufficient buoyancy and would penetrate into the lower mantle. This could explain the seismic tomographic observations regarding slabs in the transition zone and in the lower mantle, without the need to invoke a compositional stratification there.</p> </div> </div> </div>
<div> <div> <div> <d... more <div> <div> <div> <div> <p>The origin of the Eocene-Oligocene European Cenozoic Rift System (ECRIS) is debated in terms of driving forces, far-field or near field, Alpine slab-pull or active plume. An analysis of residual (non-isostatic) topography over Africa and Europe reveals domains elongated parallel to the absolute motion of plates in a hot-spot reference frame. The East African Rift (EAR) and the ECRIS sit on top of such positive anomalies. A recent whole mantle tomographic model (French et al., 2013; French & Romanowicz, 2015; Davaille & Romanowicz, 2020) shows in addition that the low shear-wave velocity zones of the lower and upper mantle are organized with a bundle of vertical plumes and horizontal fingers pointing in the same direction parallel to the absolute motion of Africa and Eurasia, thus parallel to the main rifts. The case of the EAR and its magmatic extension toward the north across the Arabian Plate is particularly clear with several levels of such fingers. The northward migration of the first volcanism from Ethiopia to Armenia between the Eocene and the Late Miocene suggests that the asthenosphere moves faster than the plates and thus drives plate motion (Faccenna et al., 2013). We propose a simple model where plates are driven by basal drag, following an upwelling from the low-velocity anomalies below Africa and toward subduction zones. The EAR develops as lithospheric weak zones on top of the positive anomalies of residual topography due to the underlying low velocity anomalies elongated parallel to the absolute motion. This indicates an interplay between large-scale convection, a small-scale fingering instability, and lithospheric deformation. The development of the Eocene-Oligocene short-lived ECRIS and its interference with Mediterranean slab dynamics are then discussed in the framework of this simple model.</p> </div> </div> </div> </div>
Ever since the theory of Plate Tectonics in the 1960’s, the dream of the geomodeler has been to g... more Ever since the theory of Plate Tectonics in the 1960’s, the dream of the geomodeler has been to generate plate tectonics self-consistently from thermal convection in the laboratory. By selfconsistenly, I mean that the configuration of the plate boundaries is in no way specified a priori, so that the plates develop and are wholly consumed without intervention from the modeler. The reciepe is simple : put a well-chosen fluid in a fishtank heated from below and cooled from above, wait and see. But the « well-chosen » is the difficult part... and the interesting one.
The planets and rocky satellites of our solar system show very different evolutions and present-d... more The planets and rocky satellites of our solar system show very different evolutions and present-day dynamics. These result from the convective regimes prevailing in their mantle. However, the conditions necessary for convection to generate Plate Tectonics on Earth, episodic resurfacing on Venus, heat pipes on Io, or no resurfacing on Mars, remain strongly debated. The difficulty comes from the complexity of rocks rheology : viscous at high temperature and on long time-scale, brittle at low temperature and short time-scale. This « soft matter » behaviour can be recovered in the laboratory using aqueous colloidal dispersions, whose rheology varies from viscous to elasto-visco-plastic to brittle when their temperature, and/or their water or ionic content, vary. We therefore have investigated the physics of thermal and solutal convection in those systems. They show a diversity of convective regimes, including the ones encountered in rocky mantles; and the conditions under which heat pip...
Mid-ocean ridges exhibit significant changes in their structural, morphological, and volcanic cha... more Mid-ocean ridges exhibit significant changes in their structural, morphological, and volcanic characteristics with changes in lithospheric thickness and/or spreading velocity. However, to separate the respective roles of those two partly correlated effects is difficult with only field data. We therefore designed a series of laboratory experiments using colloidal silica dispersions as an Earth analogue. Saline water solutions placed in contact with these fluids, cause formation of a skin through salt diffusion, whose rheology evolves from purely viscous to elastic and brittle with increasing salinity. Applying a fixed spreading rate to this preformed, brittle plate results in cracks, faults and axial ridge structures. Lithospheric (skin) thickness at a given extension rate can be varied by changing the surface water layer salinity. Moreover, the mechanical properties of the skin can also be independently controlled by changing the type of colloid. We focus here on cases where the spr...
Introduction: Planets long-term cooling, as well as surface phenomena such as plate tectonics, vo... more Introduction: Planets long-term cooling, as well as surface phenomena such as plate tectonics, volcanoes and earthquakes, are mainly controlled by the existence and patterns of convective motions in their solidstate mantle. The latter strongly depend on rocks’rheology. Two main regimes are usually considered: « Stagnant Lid » where convective motions develop below a lithosphere which remains stagnant, and « Plate Tectonics » with the continuous rejuvenation of the planet surface by subduction of its cold gravitationnally unstable lithosphere. However the necessary conditions for the latter regime are still debated. We run new laboratory experiments with visco-elastoplastic colloidal dispersions to explore the problem, and a diversity of regimes was observed besides the two end members already mentioned. Moreover, an experiment always passed through a succession of different regimes, and this should also be the case for a planet.
Using laboratory experiments and theoretical modeling, we recently showed that plumes could induc... more Using laboratory experiments and theoretical modeling, we recently showed that plumes could induce roll-back subduction around large coronae. When a hot plume rises under a brittle and visco-elasto-plastic skin/lithosphere, the latter undergoes a flexural deformation which puts it under tension. Radial cracks and rifting of the skin then develop, sometimes using pre-existing weaknesses. Plume material upwells through the cracks (because it is more buoyant) and spreads as a axisymmetric gravity current above the broken denser skin. The latter bends and sinks under the combined force of its own weight and that of the plume gravity current. However, due to the brittle character of the upper part of the experimental lithosphere, it cannot deform viscously to accomodate the sinking motions. Instead, the plate continues to tear, as a sheet of paper would do upon intrusion. Several slabs are therefore produced, associated with trenches localized along partial circles on the plume, and stro...
The earliest compositional differentiation of the terrestrial planets, the formation of their out... more The earliest compositional differentiation of the terrestrial planets, the formation of their outgassed atmospheres, and the existence of condensed water oceans over their solid mantles, are conditioned by magma ocean (MO) formation and solidification. Recent studies have suggested that depending on the planet initial water content and its orbital distance to the sun, two types of conditions, with (I) or without (II) water ocean, can prevail at the end of the MO phase. We use a coupled interior-atmosphere model of MO thermal evolution to go further and study systematically the influence of the planet initial CO2 content on the resulting surface conditions (temperature, volatiles, condensed water) at the end of the MO phase. The position of the boundary between the two regimes is shown to depend also on the initial CO2 content.
Recent tomographic imaging of the mantle below major hotspots shows slow seismic velocities exten... more Recent tomographic imaging of the mantle below major hotspots shows slow seismic velocities extending down to the core-mantle boundary, confirming the existence of mantle plumes. However, these plumes are much thicker than what was previously thought based on convection in a newtonian or a power-law mantle. In those two rheological frameworks, hot « fat » plumes can only be produced either if hotter plumes are also more viscous than ambiant mantle, or if the plume conduit is thickened by a recirculation of thermo-compositional heterogeneities within the conduit itself.
La Terre est un systeme multiechelles extremement complexe. Sa dynamique est loin d'etre comp... more La Terre est un systeme multiechelles extremement complexe. Sa dynamique est loin d'etre comprise, d'une part parce que les phenomenes entrant en jeu ne sont pas tous identifies, d'autre part parce que les observations disponibles, bien que tres diverses, ne sont souvent que parcellaires. Predire la dynamique des systemes naturels constitue donc un defi pour le (geo)physicien, car la dynamique terrestre est bien souvent un probleme sous-contraint. C'est pourquoi le progres en geophysique necessite obligatoirement deux approches complementaires : l'etude des phenomenes physiques fondamentaux, qui permet de construire un cadre interpretatif rigoureux pour les observations, et le respect des observations de tout genre. D'ou un va-et-vient constant entre disciplines, et entre observations, mesures et modeles physiques. Ce court article en donne quelques exemples.
Dans les manteaux des planètes rocheuses et dans les dispersions colloïdales, une variation de te... more Dans les manteaux des planètes rocheuses et dans les dispersions colloïdales, une variation de température ou de composition peut induire des variations de densité, la gélification de leur microstructure et une modification drastique de leur viscosité. Mouvements de convection et formation d’une peau faiblement déformable en surface peuvent donc coexister. L’étude de la convection dans les dispersions colloïdales peut fournir des informations précieuses sur la physique des planètes. Ainsi, nous avons montré que les panaches thermiques sont importants pour casser la peau de surface et induire la subduction de celle-ci, un phénomène qui pourrait être actif sur Vénus aujourd’hui.
Rheological and mechanical properties of mono-disperse hard-sphere silica suspensions (Ludox® HS-... more Rheological and mechanical properties of mono-disperse hard-sphere silica suspensions (Ludox® HS-40) are investigated as a function of particle volume fraction (φp ranging from 0.22 to 0.51), using shear rate tests, oscillatory methods, indentation and ultrasonic technique [1]. Four regimes are identified with increasing particle content; they are related to the existence and the behavior of the electrical double layer (EDL) around each particle. Semidiluted suspensions (0.22 ≤ φp ≤ 0.30), which do not flocculate due to the strong interparticle repulsions, show Newtonian behavior (I). Their viscosity is higher than predicted by classical models, due to the EDL, which effectively increases the particle radius. At higher particle contents, the suspensions exhibit two viscoelastic (VE) shear-thinning regimes separated by a critical particle volume fraction. It discriminates semi-concentrated fluids (II, 0.31 ≤ φp ≤ 0.35), in which particles form aggregates of quickly increasing size, from concentrated (III 0.35 ≤ φp ≤ 0.47) strain sensitive samples with paste-like aspect and high elastic modulus. Regime II can be explained by a sudden partial collapse of the EDL as it is progressively squeezed with increasing φp. At higher particle concentration (IV, φp = 0.51), the suspensions undergo a glass transition and behave as an elastic-brittle solid with high Young’s modulus. The particles are trapped within the cages constituted by their neighbors, and the structure is at arrest.
<div> <div> <div> <p... more <div> <div> <div> <p>When a fluid can experience a "jammed" state, it will flow only when the local deviatoric stress becomes greater than a critical stress, the so-called  "yield-stress". Jamming can be caused by entangled dislocations in a mineral, or by the existence of a hard skeleton in a two-phase fluid. According to recent numerical modeling, a Bridgmanite lower mantle would predominantly deform by pure dislocation climb; and due to dislocations interactions, it would flow only for local deviatoric stress greater than a critical yield stress which depends on dislocation density. In a first set of fluid mechanics experiments in such a visco-plastic fluid, we showed that hot plumes would develop with a much thicker morphology than in newtonian fluids. Scaling laws derived from the experiments tightly relate the buoyancy and diameter of the hot plumes to the value of the yield-stress, as well as to the mantle microstructure (such as dislocation density and vacancy concentration). Yield stress values between 1 and 10 MPa, implying dislocation densities between 10<sup>8</sup> and 10<sup>10</sup> m<sup>−2</sup>, would be sufficient to explain the thick plumes morphology observed in seismic tomography images; while low vacancy concentrations could explain the 1000 km-depth horizon also seen in tomography. </p> <p>In a second set of experiments, we show that the existence of a yield stress in a Bridgmanite lower mantle will also act as a filter regarding slab penetration in the lower mantle. Given slab buoyancy, a typical slab, 100 km-thick, could not overcome the lower mantle yield stress. So most of single slabs would be expected  to stagnate in the transition zone. However a pile of folded slab with a typical thickness around 400 km would have sufficient buoyancy and would penetrate into the lower mantle. This could explain the seismic tomographic observations regarding slabs in the transition zone and in the lower mantle, without the need to invoke a compositional stratification there.</p> </div> </div> </div>
<div> <div> <div> <d... more <div> <div> <div> <div> <p>The origin of the Eocene-Oligocene European Cenozoic Rift System (ECRIS) is debated in terms of driving forces, far-field or near field, Alpine slab-pull or active plume. An analysis of residual (non-isostatic) topography over Africa and Europe reveals domains elongated parallel to the absolute motion of plates in a hot-spot reference frame. The East African Rift (EAR) and the ECRIS sit on top of such positive anomalies. A recent whole mantle tomographic model (French et al., 2013; French & Romanowicz, 2015; Davaille & Romanowicz, 2020) shows in addition that the low shear-wave velocity zones of the lower and upper mantle are organized with a bundle of vertical plumes and horizontal fingers pointing in the same direction parallel to the absolute motion of Africa and Eurasia, thus parallel to the main rifts. The case of the EAR and its magmatic extension toward the north across the Arabian Plate is particularly clear with several levels of such fingers. The northward migration of the first volcanism from Ethiopia to Armenia between the Eocene and the Late Miocene suggests that the asthenosphere moves faster than the plates and thus drives plate motion (Faccenna et al., 2013). We propose a simple model where plates are driven by basal drag, following an upwelling from the low-velocity anomalies below Africa and toward subduction zones. The EAR develops as lithospheric weak zones on top of the positive anomalies of residual topography due to the underlying low velocity anomalies elongated parallel to the absolute motion. This indicates an interplay between large-scale convection, a small-scale fingering instability, and lithospheric deformation. The development of the Eocene-Oligocene short-lived ECRIS and its interference with Mediterranean slab dynamics are then discussed in the framework of this simple model.</p> </div> </div> </div> </div>
Ever since the theory of Plate Tectonics in the 1960’s, the dream of the geomodeler has been to g... more Ever since the theory of Plate Tectonics in the 1960’s, the dream of the geomodeler has been to generate plate tectonics self-consistently from thermal convection in the laboratory. By selfconsistenly, I mean that the configuration of the plate boundaries is in no way specified a priori, so that the plates develop and are wholly consumed without intervention from the modeler. The reciepe is simple : put a well-chosen fluid in a fishtank heated from below and cooled from above, wait and see. But the « well-chosen » is the difficult part... and the interesting one.
The planets and rocky satellites of our solar system show very different evolutions and present-d... more The planets and rocky satellites of our solar system show very different evolutions and present-day dynamics. These result from the convective regimes prevailing in their mantle. However, the conditions necessary for convection to generate Plate Tectonics on Earth, episodic resurfacing on Venus, heat pipes on Io, or no resurfacing on Mars, remain strongly debated. The difficulty comes from the complexity of rocks rheology : viscous at high temperature and on long time-scale, brittle at low temperature and short time-scale. This « soft matter » behaviour can be recovered in the laboratory using aqueous colloidal dispersions, whose rheology varies from viscous to elasto-visco-plastic to brittle when their temperature, and/or their water or ionic content, vary. We therefore have investigated the physics of thermal and solutal convection in those systems. They show a diversity of convective regimes, including the ones encountered in rocky mantles; and the conditions under which heat pip...
Mid-ocean ridges exhibit significant changes in their structural, morphological, and volcanic cha... more Mid-ocean ridges exhibit significant changes in their structural, morphological, and volcanic characteristics with changes in lithospheric thickness and/or spreading velocity. However, to separate the respective roles of those two partly correlated effects is difficult with only field data. We therefore designed a series of laboratory experiments using colloidal silica dispersions as an Earth analogue. Saline water solutions placed in contact with these fluids, cause formation of a skin through salt diffusion, whose rheology evolves from purely viscous to elastic and brittle with increasing salinity. Applying a fixed spreading rate to this preformed, brittle plate results in cracks, faults and axial ridge structures. Lithospheric (skin) thickness at a given extension rate can be varied by changing the surface water layer salinity. Moreover, the mechanical properties of the skin can also be independently controlled by changing the type of colloid. We focus here on cases where the spr...
Introduction: Planets long-term cooling, as well as surface phenomena such as plate tectonics, vo... more Introduction: Planets long-term cooling, as well as surface phenomena such as plate tectonics, volcanoes and earthquakes, are mainly controlled by the existence and patterns of convective motions in their solidstate mantle. The latter strongly depend on rocks’rheology. Two main regimes are usually considered: « Stagnant Lid » where convective motions develop below a lithosphere which remains stagnant, and « Plate Tectonics » with the continuous rejuvenation of the planet surface by subduction of its cold gravitationnally unstable lithosphere. However the necessary conditions for the latter regime are still debated. We run new laboratory experiments with visco-elastoplastic colloidal dispersions to explore the problem, and a diversity of regimes was observed besides the two end members already mentioned. Moreover, an experiment always passed through a succession of different regimes, and this should also be the case for a planet.
Using laboratory experiments and theoretical modeling, we recently showed that plumes could induc... more Using laboratory experiments and theoretical modeling, we recently showed that plumes could induce roll-back subduction around large coronae. When a hot plume rises under a brittle and visco-elasto-plastic skin/lithosphere, the latter undergoes a flexural deformation which puts it under tension. Radial cracks and rifting of the skin then develop, sometimes using pre-existing weaknesses. Plume material upwells through the cracks (because it is more buoyant) and spreads as a axisymmetric gravity current above the broken denser skin. The latter bends and sinks under the combined force of its own weight and that of the plume gravity current. However, due to the brittle character of the upper part of the experimental lithosphere, it cannot deform viscously to accomodate the sinking motions. Instead, the plate continues to tear, as a sheet of paper would do upon intrusion. Several slabs are therefore produced, associated with trenches localized along partial circles on the plume, and stro...
The earliest compositional differentiation of the terrestrial planets, the formation of their out... more The earliest compositional differentiation of the terrestrial planets, the formation of their outgassed atmospheres, and the existence of condensed water oceans over their solid mantles, are conditioned by magma ocean (MO) formation and solidification. Recent studies have suggested that depending on the planet initial water content and its orbital distance to the sun, two types of conditions, with (I) or without (II) water ocean, can prevail at the end of the MO phase. We use a coupled interior-atmosphere model of MO thermal evolution to go further and study systematically the influence of the planet initial CO2 content on the resulting surface conditions (temperature, volatiles, condensed water) at the end of the MO phase. The position of the boundary between the two regimes is shown to depend also on the initial CO2 content.
Recent tomographic imaging of the mantle below major hotspots shows slow seismic velocities exten... more Recent tomographic imaging of the mantle below major hotspots shows slow seismic velocities extending down to the core-mantle boundary, confirming the existence of mantle plumes. However, these plumes are much thicker than what was previously thought based on convection in a newtonian or a power-law mantle. In those two rheological frameworks, hot « fat » plumes can only be produced either if hotter plumes are also more viscous than ambiant mantle, or if the plume conduit is thickened by a recirculation of thermo-compositional heterogeneities within the conduit itself.
La Terre est un systeme multiechelles extremement complexe. Sa dynamique est loin d'etre comp... more La Terre est un systeme multiechelles extremement complexe. Sa dynamique est loin d'etre comprise, d'une part parce que les phenomenes entrant en jeu ne sont pas tous identifies, d'autre part parce que les observations disponibles, bien que tres diverses, ne sont souvent que parcellaires. Predire la dynamique des systemes naturels constitue donc un defi pour le (geo)physicien, car la dynamique terrestre est bien souvent un probleme sous-contraint. C'est pourquoi le progres en geophysique necessite obligatoirement deux approches complementaires : l'etude des phenomenes physiques fondamentaux, qui permet de construire un cadre interpretatif rigoureux pour les observations, et le respect des observations de tout genre. D'ou un va-et-vient constant entre disciplines, et entre observations, mesures et modeles physiques. Ce court article en donne quelques exemples.
Dans les manteaux des planètes rocheuses et dans les dispersions colloïdales, une variation de te... more Dans les manteaux des planètes rocheuses et dans les dispersions colloïdales, une variation de température ou de composition peut induire des variations de densité, la gélification de leur microstructure et une modification drastique de leur viscosité. Mouvements de convection et formation d’une peau faiblement déformable en surface peuvent donc coexister. L’étude de la convection dans les dispersions colloïdales peut fournir des informations précieuses sur la physique des planètes. Ainsi, nous avons montré que les panaches thermiques sont importants pour casser la peau de surface et induire la subduction de celle-ci, un phénomène qui pourrait être actif sur Vénus aujourd’hui.
Uploads