Title:Effects of non-parallelism on standard and magnetorheological measurements

Abstract:Aiming towards the magnetorheological characterisation of whole human blood, we evaluated the suitability of our experimental setup for steady shear measurements with low-viscosity fluids. Previous measurements with a rotational rheometer equipped with a magnetorheological cell returned low and inconsistent apparent-viscosity values. In this work, a parametric study was conducted, experimentally and numerically, to evaluate the possible error sources and define an experimentally reliable window. Steady shear measurements were carried out with Newtonian fluids using two geometries: parallel-plates at different gap heights, and cone-plate. A clear decrease in measured viscosity with parallel-plate gap reduction was found, along with a slight overestimation of the cone-plate. Numerical results corroborated the experimental observations, pointing towards a small inclination of the bottom plate (between 0.1° and 0.3°). The numerical study has also shown that geometry non-parallelism can lead to significant flow alterations, particularly for cone-plate geometries where contraction/expansion flow can be generated, critically deviating from the canonical Couette flow and provoking non-negligible errors. Additional experimental and numerical work was conducted to evaluate the effects of the non-parallelism on magnetorheological measurements. The geometrical asymmetry results in severe alterations of the microstructural dynamics, possibly leading to an underestimation of the magnetorheological response. This work has shown that geometry non-parallelism can have critical repercussions on simple shear flows, with fundamentally different implications from a general gap-error, and can be identified through comparative measurements with Newtonian fluids in different geometries (parallel-plates and cone-plate)