Session M18: Memory Formation in Matter: From Reading the Past to Designing the Future

Global Memory From Local Hysteresis and Disorder in a Jammed Solid

Cyclically shearing a soft 2D jammed material can train it with a memory of strain amplitude that can be retrieved later. Using experiments with particles at an oil-water interface, we show that this memory arises from the hysteresis of particles' rearrangements within the material. A stable population of rearrangements is created by training, and disorder ensures that they collectively discriminate among different inputs. These results point to a generic way of encoding memories in hysteresis and disorder, and explain why jammed and dilute suspensions must have different kinds of memory. The behavior in jammed systems is reminiscent of the return-point memory found in ferromagnets and many other systems, but in light of jammed systems' unquenched disorder, it is unexpected.   

Memory formation in cyclically deformed glasses

Cyclically shear deformed glasses reach, after a transient regime, periodic orbits involving transitions between basins of energy minima that are visited repeatedly during repeated cycles of deformation, if the amplitude of deformation is not too large. These periodic orbits encode information of their deformation history, and have been shown to retain memory of training at multiple strain amplitudes. Such memory effects, in the presence of plastic rearrangements involved in transitions between basins, is surprising, but appears to be generic. Investigations of the nature of such periodic orbits reveals rich structure, whose analysis permits investigation of the presence or otherwise of features such as return point memory, and suggests protocols by which the encoded information may most efficiently be recovered.   

Memories in a jar

Self-assembly has emerged as a powerful technique for synthesizing structures on the nano and micro-scale. The basis of this development is the use of biopolymers, like DNA, to design specific interactions between multiple species of components, allowing the spontaneous assembly of complex structures. Inspired by biological systems, where the same set of components can assemble many different complexes, we can design mixtures of shared components that have memory of many distinct structures and are capable of assembling each at will. Moreover, these structures can transition one to another when an appropriate switch is triggered. In this talk I will discuss examples of memories modeled as structures made of DNA coated colloidal particles.   

Memory Effects in Amorphous Mechanical Systems

Amorphous mechanical systems such as thin sheets crumpled into a soft ball, elastic foams and granular systems can exhibit a long-lasting Kovacs-like memory effect: a slow non-monotonic volume or stress relaxation, the shape of which contains information on mechanical perturbations applied to the system long before the measurement was taken. The talk will discuss previous and recent observations, and our attempts to uncover the mechanism underlying memory retention and related phenomena in these systems.   

Directed aging: Training elastic responses in disordered systems

Disordered materials exist in a rugged energy landscape and become trapped in metastable states that are not a global energy minimum. These far-from-equilibrium systems relax slowly as they search for lower-energy configurations. This aging often leads to material degradation and is thus thought of as undesirable. Here I show how this aging can be directed to produce a final state that has advantageous properties and unique functions not typically found naturally [1]. Aging a system subject to external constraints leaves a memory of how it was aged and creates stresses that direct its evolution [2]. Aging obeys a natural “greedy algorithm”: the material simply follows the path of most rapid and accessible relaxation. During aging, the material modifies stressed regions differently from those under less stress. Our goal is to find out what range of behaviors can be achieved by directed-aging protocols. We do this through experiments and with simulations that use simple models of the aging process in disordered networks. Our experimental networks are laser cut from EVA foam and then aged under external stress. (We can apply heat not only to accelerate the aging process but also to help the material retain a memory of how it was aged after the system has been re-cooled.) Using different applied stresses, we can achieve different functions; a memory of the stress protocol can be read out by measuring the mechanical response of the aged systems. Directed aging provides a pathway for modifying and tuning a material’s elastic properties in the non-linear as well as the linear regime without having to control the material at the microscopic level.

[1] N. Pashine, D. Hexner, A.J. Liu, S. R. Nagel “Directed aging, memory and Nature's greed,” ArXiv:1903.05776.
[2] D. Hexner, N. Pashine, A.J. Liu, S. R. Nagel “Effect of aging on the non-linear elasticity and memory formation in materials” ArXiv:1909.00481.