Yield stress suspensions: from microstructure to macroscopic behavior – SUSPASEUIL

Dense suspensions such as fresh concrete consist of very polydisperse particles suspended in a viscous fluid. These materials are yield stress fluids: they flow only when the applied stress is large enough. When scale separation is possible between the large and small particles, these materials can be seen as suspensions of grains (noncolloidal particles) in a yield stress fluid (the colloidal paste). A good knowledge of the rheological properties of these materials, and of the link between these properties and their composition, is needed to ensure a good control of their casting properties (for industrial materials) and for the prediction of geophysical flows (e.g. debris flows). To date, the existing studies have focused only on very specific materials in link with their field of application. Therefore, a global understanding of their behavior still has to be achieved. To this end, one has to characterize their microstructure, their macroscopic properties, and the link between these two properties. Indeed, the particle distribution and their orientation in the suspension depend on the suspension flow history, which depends on the macroscopic properties of the suspension; these last properties depend in turn on the suspension microstructure. Moreover, at high particle volume fraction, contact networks may appear, which may lead to jamming.
The final aim of our project is to build a constitutive law of yield stress suspensions that includes a characterization of its microstructure and of this microstructure evolution. As no standard rheometric technique exists to characterize the macroscopic and structural anisotropies, and as it is challenging to take the nonlinear properties of the interstitial fluid into account, we propose to develop original experimental, numerical and theoretical tools, the coupling of which will allow us to achieve our goal.
We will study the macroscopic behavior of yield stress suspensions thanks to a new rheological test that will allow us to apply three-dimensional loadings. We will characterize the anisotropy of the macroscopic behavior for a given microstructure (obtained through a controlled strain history). On the other hand, we will develop a rheometrical device inserted in an X-ray microtomograph, which will allow us to characterize directly the suspension microstructure as a function of the strain history. A numerical tool devoted to simulate the flows of yield stress suspensions will also be developed; it will be used to predict the microstructure and its evolution as a function of the applied loading. Its strength will be its optimization and validation by comparison with the experiments succinctly described above. Finally, the link between the microstructure and the macroscopic behavior will be built through a change of scale method; it will be first validated experimentally, then numerically for complex flows.
Finally, these coupled studies should allow us to build the link between the microstructure and the macroscopic behavior, and between the strain history and the microstructure evolution, providing in the end a complete model of macroscopic behavior of yield stress suspensions validated experimentally.
This field of investigation being basically new, we will limit our study to the fundamental case of monodisperse spheres suspended into model yield stress fluids; we will nevertheless propose a validation of our approach on cementitious materials. The original tools that we will develop during this project and their ability to allow for the modeling of the behavior of model materials will naturally lead to study more complex materials.