Constitutive model for shear-thickening suspensions: Predictions for steady shear with superposed transverse oscillations

TL;DR

This paper uses a tensorial constitutive model to predict how superposed transverse oscillations can unjam shear-thickening suspensions, successfully reproducing experimental viscosity drops and comparing well with simulations.

Contribution

The study applies a tensorial constitutive model to predict unjamming effects of transverse oscillations on shear-thickening suspensions, extending previous models to dynamic oscillatory conditions.

Findings

Model reproduces experimentally observed viscosity drops.

Qualitative agreement with microstructure and stress evolution.

Discrepancies highlight areas for model improvement.

Abstract

We recently developed a tensorial constitutive model for dense, shear-thickening particle suspensions that combines rate-independent microstructural evolution with a stress-dependent jamming threshold. This gives a good qualitative account for reversing flows, although it quantitatively over-estimates structural anisotropy [J.~J.~J.~Gillissen {\em et al.}, Phys. Rev. Lett. {\bf 123} (21), 214504 (2019)]. Here we use the model to predict the unjamming effect of superposed transverse oscillations on a steady shear flow in the thickened regime [N.~Y.~C.~Lin {\em et al.}, Proc.~Nat.~Acad.~Sci.~USA {\bf 113}, 10774 (2016)]. The model successfully reproduces the oscillation-mediated viscosity drop observed experimentally. We compare the time-dependent components of the stress and microstructure tensors to discrete-element simulations. Although the model correctly captures the main qualitative behaviour, it generally over-predicts the microstructural anisotropy in steady shear, and it under-predicts the number of particle contacts in oscillating shear. It also does not fully capture the correct variation in phase angle between the transverse component of the microstructure and the shear rate oscillations, as the amplitude of the latter is increased. These discrepancies suggest avenues for future improvements to the model.