Title: Local rigid clusters encode mechanical memories in shear-jammed dense suspensions

Speaker: Prof. Sayantan Majumdar, Raman Research Institute, Bangalore

Abstract: Dense granular suspensions of frictional particles transform into a transient solid-like state under applied stress/strain perturbation, a phenomenon known as shear-induced jamming. Once the perturbation is removed, the system relaxes back to the liquid-like state usually within a few seconds. Here we uncover a novel and unexpected memory formation in these systems, that helps the system to enhance the shear-thickening response under suitable training protocols. In our experiments, we report that if we repeatedly perturb the system in the same direction, the system gradually transforms to a much stronger solid-like jammed state. However, if the perturbation direction is alternatively reversed, the jammed state becomes weaker or, not reached at all. Remarkably, the same effect persists even when the time gap between the successive perturbations is much larger than the bulk stress relaxation time-scale of the system. This implies that the liquid-like state after the complete stress relaxation encodes a structural memory. Using in situ boundary imaging, we confirm that such direction-dependent enhancement/weakening of mechanical response also reflects in the shear-induced dilation behaviour of the sample. Numerical simulations suggest that such an effect originates from the gradual structural evolution of the system under repeated unidirectional perturbations. Such perturbations increase the number and size of the locally rigid-clusters and thereby enhance the strength of the solid-like jammed state. We observe that such local rigidity also plays a crucial role in stress-anisotropy and shear-reversal response of the system. Our results may open up the possibility of designing smart materials where the mechanical properties can be reversibly tuned by encoding structural memories in the system.