The prevailing concept that positive-edge to negative-face attraction accounts for the rheological behavior of montmorillonite suspensions at low electrolyte concentration was investigated. In one experiment, Mg2+ released from Na-montmorillonite was measured at several NaCl concentrations; in a second experiment, the viscosity, η, and the extrapolated shear stress, θ, were measured at several clay concentrations, pHs, and NaCl concentrations; and in a third experiment, the absorbance, A, was measured at two wavelengths (450 and 760 nm) at different clay and electrolyte concentrations. The released Mg2+ decreased with increasing NaCl concentration until it became zero at a NaCl concentration between 0.01 and 0.02 M, depending on pH. Thereafter, it increased with increasing NaCl concentration. Both θ and η were highly correlated with the amount of released Mg2+. Also, A remained constant until the NaCl concentration corresponded to that at the minimum of θ. Thereafter, it increased and became linearly related to θ. These results suggest: (1) positive-edge to negative-face interaction cannot solely account for the rheological properties of montmorillonite at low electrolyte concentration, and (2) the release of octahedral Mg2+ from montmorillonite affects θ, because it reduces the negative charge on the particles and, thereby, the repulsive force between them.

Earthquake-induced rock shear through a bentonite-filled deposition hole in a repository for spent nuclear fuel is an important scenario for the safety analysis because it may cause substantial damage to the canister hosting the spent fuel. Appropriate tools to investigate the effects on the buffer and the canister are required.

The study described here explored the laboratory tests conducted to develop a material model of the bentonite buffer to be used in the simulations, the material models that these tests have provided and finite element (FE) simulations of three scale tests of a rock shear for comparison between modelled and measured results. The results were used for validation of the material models and the calculation technique that was used for modelling different rock-shear cases.

The laboratory study consisted of swelling-pressure tests and tests to determine shear strength and stress-strain properties. The material model is elastic-plastic with a nonlinear stress-strain relation which depends on the density of the bentonite buffer and is a function of the strain rate. The three scale tests were modelled using the Abaqus finite element code. Good agreement between modelled and measured results was observed, in spite of the complexity of the models and the difficulties associated with measuring stresses and strains under the very fast shear.

The modelling results thus validate the modelling of the SR-Site. The modelling technique, the element mesh and the material models used in these analyses are well fitted and useful for this type of modelling.

The solubility of milk protein concentrate (MPC) powders was influenced by the method used for preparing the concentrate, drying conditions, and the type of dryer used. Increasing total solids of the ultrafiltered concentrates (23% total solids, TS) by diafiltration to 25% TS or evaporation to 31% TS decreased the solubility of MPC powders (80–83% protein, w/w dry basis), with ultrafiltration followed by evaporation to higher total solids having the greater detrimental effect on solubility. High shear treatment (homogenisation at 350/100 bar, microfluidisation at 800 bar or ultrasonication at 24 kHz, 600 watts) of ultrafiltered and diafiltered milk protein concentrates prior to spray drying increased the nitrogen solubility of MPC powders (82% protein, w/w dry basis). Of the treatments applied, microfluidisation was the most effective for increasing nitrogen solubility of MPC powders after manufacture and during storage. Manufacture of MPC powders (91% protein, w/w dry basis) prepared on two different pilot-scale dryers (single stage or two stage) from milk protein concentrates (20% TS) resulted in powders with different nitrogen solubility and an altered response to the effects of microfluidisation. Microfluidisation (400, 800 and 1200 bar) of the concentrate prior to drying resulted in increased long term solubility of MPC powders that were prepared on a single stage dryer but not those produced on a two stage spray dryer. This work demonstrates that microfluidisation can be used as a physical intervention for improving MPC powder solubility. Interactions between the method of preparation and treatment of concentrate prior to drying, the drying conditions and dryer type all influence MPC solubility characteristics.

The economic challenge for PSA regarding the maintenance of their mould foundry park isto predict and improve its lifespan in preventing any risk of initiation and propagationof cracks in thermal fatigue. A method for assessing the number of cycles of castingtooling thermal fatigue used by PSA was developed using the criterion of Fatemi-Socie [A.Fatemi, D. Socie, Fatigue Fract. Engng.Mater. Struct. 11 (1988) 149-165, D.F.Socie, multiaxial fatigue damage models, ASME 109 (1987)293-298], associated with a shear failure mode, which predicts the thermal fatigue damageby assessing stress and strain tensors on different facets around any material point.

An analytical model for determining the shear strength of concrete pile caps under the failure mode of diagonal-compression originally based on the softened strut-and-tie model is proposed. The failure probabilities of reinforced concrete pile caps are investigated by Monte Carlo method. The results indicate that the proposed model can accurately predict the shear strength of the pile caps. The distribution of the failure probabilities for pile caps designed to ACI 318-02 Appendix A and the proposed design method are more uniform than that designed to the ACI 318-99. The ACI 318-99 is very conservative and cannot provide a consistent safety for pile caps design. It is suggested that the procedures in the ACI 318-02 Appendix A should be moved to the main body of ACI 318-02 and the proposed design method should be incorporated into the current reinforced concrete pile cap design methods.

Dans les pays européens à sismicité modérée comme la France, la plupart des bâtiments en béton armé a été construite avant l'application des normes parasismiques modernes. En raison de l'absence de prise en compte du séisme lors de leur dimensionnement ou de la modification de l'aléa sismique, de nombreuses structures en béton armé ne respectent pas les normes actuelles sans être pour autant moins sûres. Pour des raisons de sûreté, les structures industrielles et nucléaires font l'objet de réévaluations sismiques régulières. L'application des méthodes de dimensionnement donne des résultats irréalistes et surtout peuvent occulter les points faibles de la structure. Cet article présente les modèles non-linéaires utilisés lors de réévaluations sismiques. Ces modélisations ont été appliquées à la modélisation de structures portique et de nœuds d'ossature testées sur table vibrante à Saclay et sur le mur de réaction du CCR d'Ispra.

The shear failure probabilities of reinforced concrete beam-column joints have been investigated by Monte Carlo method. The theoretical shear strength of joints is based on the softened strut-and-tie model proposed by Hwang and Lee (2002). The random variables included in this study are the strengths of concrete, the ultimate compression strain of concrete, the strengths of reinforcement, the dimensions of cross-section, and the model error of theoretical shear strength of joints. The shear failure probabilities of joints with SD 280 flexural reinforcement in the beams designed using the ACI Code are all higher than 0.04. The joints designed according to the softened strut-and-tie model are safer than those designed according to the ACI Code. The shear failure probabilities of exterior joints are higher than those of interior joints. The shear failure probabilities of joints with SD 280 flexural reinforcement in the beams are higher than those of joints with SD 420 flexural reinforcement.

Des particules soumises à un champ de force peuvent entraîner d'autresparticules non soumises à ce champ par interactions hydrodynamiques. On étudie unprocessus continu dans lequel les deux types de particules sont transportées parun écoulement de cisaillement perpendiculaire au champ de force. Cet entraînement est modélisé par une étude statistique de la suspension en écoulement de Stokes. Le champ de forces etl'écoulement modifient la probabilité de paire de particules de types différents. On montre que si le cisaillement est trop important, il apparaît desdoublets permanents, ce qui augmente l'entraînement. Sinon, cette probabilitéest anisotrope, mais est cependant proche de celle calculée pour un champ de forceseul, probabilité qui est isotrope. On présente alors un modèle d'entraînementbasé sur les vitesses moyennes des deux espèces de particules. Ce problèmes'applique en particulier à la technique de séparation dite par cellule SPLITTen chimie analytique. Dans cette technique, des particules marquées soumises àun champ sont séparées mais entraînent les particules non marquées. Le modèleest appliqué à l'optimisation du dimensionnement de la cellule SPLITT.