Inertial Dense Suspensions (IDeS); Flows in between Laminar and Turbulent

Inertial Dense Suspensions are very common in nature and industry, with examples including blood flow, recycling, waste slurry transport, additive manufacturing, and energy storage solutions. While they are of great practical importance, these suspension flows are poorly understood. A direct consequence is that we cannot predict even the most basic properties of these flows with sufficient accuracy. For instance, the current best models over- or underestimate the measured pressure drop in a suspension pipe flow by 30-40% for various flow rates. This means that designing or controlling these flows is difficult, as there is no reliable relation between the flow rate and pressure difference (i.e. pumping power) for a given suspension.Our inability to predict these flows stems in part from the fact that they defy the conventional dichotomy of ‘laminar’ versus ‘turbulent’ flow; they inhabit a terra incognita. Most theoretical studies so far have been limited to confined suspensions in the viscous regime. In the applications mentioned, this idealized approach is not valid: Due to the finite size and velocity of the suspended particles, several inertial phenomena occur, such as particle-induced fluctuations, a gradual transition to turbulence, and shear-induced migration. These are at best only qualitatively understood, yet they dramatically alter the flow behavior.We can only now start investigating these flows systematically. The opacity of suspensions severely limited our ability to obtain experimental data using conventional flow measurement techniques. In recent years, I pushed the capabilities of non-optical techniques (e.g. ultrasound, MRI), which have opened up this important field. Using unique experimental facilities supplemented with resolved simulations, I will obtain the much-needed data and study the fundamental physics involved in this problem. I will use this to construct a comprehensive conceptual model and then reduce it to formulate predictive rules.