Self-organization and memory formation in two-dimensional jammed deformable matter under cyclic compression

Rahul Nayak, Satyavani Vemparala, Pinaki Chaudhuri · 2026

Read Self-organization and memory formation in two-dimensional jammed deformable matter under cyclic compression by Rahul Nayak, Satyavani Vemparala, Pinaki Chaudhuri (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-22398.txt.

Introduction

Nayak, Vemparala and Chaudhuri study soft ring assemblies squeezed cold past the jamming point. Further squeezing buckles the rings, softening the material on large scales. Under repeated squeezing, uniform rings anneal toward a reversible path through an ordered state. Mixed-size rings instead settle into stable, hysteretic limit cycles. Those cycles store a robust memory of training history, kept even under later overdriving. Large-scale hysteresis traces to lopsided microscale deformations with the contact network mostly intact. Particle softness thus governs organisation and memory in jammed soft matter.

Catalog record

CREATORRahul Nayak, Satyavani Vemparala, Pinaki Chaudhuri
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44913
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.22398

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Self-organization and memory formation in two-dimensional jammed deformable matter under cyclic compression Rahul Nayak, Satyavani Vemparala, Pinaki Chaudhuri Abstract We study the athermal mechanical response of deformable ring assemblies to quasistatic compression. Beyond jamming, further densification induces buckling of rings, resulting in macroscopic mechanical softening. Under cyclic compression, monodisperse systems anneal toward a nearly reversible path passing through an ordered state, whereas polydisperse systems converge to stable, hysteretic limit cycles. These limit cycles encode a robust memory of the training history that is retained even under subsequent overdriving. We show that macroscopic hysteresis in the disordered packings originates from directionally asymmetric non-affine deformations at the microscale while keeping contact network largely intact. Our findings demonstrate how particle deformability governs collective self-organization and memory formation in jammed soft matter. I Introduction The jamming transition describes the emergence of rigidity in disordered collections of particles, such as grains, foams, and emulsions 22 . While the scaling laws and structural properties of jammed systems are well-established for particles modeled as point objects interacting via hard-core or soft repulsive potentials 29 ; 23 ; 41 ; 10 , many physical and biological constituents, ranging from cells and microgels to polymeric rings, possess internal degrees of freedom that allow for significant shape deformations 34 ; 42 ; 24 . Unlike rigid objects, these deformable entities adapt to local packing constraints according to their internal elasticity, altering the mechanical response of the assembly in ways that point-particle models cannot capture. The bulk behavior of these systems is therefore determined by both inter-particle contacts and the elastic energy stored in shape deformations, thereby linking particle-scale mechanics to the macroscopic response 6 ; 40 . The consequences of this coupling extend well beyond structural packing, viz. assemblies of deformable particles serve as model systems for studying glassy dynamics in highly compliant constituents 32 ; 15 ; 14 ; 13 ; 28 ; 7 , for modeling biological tissues 30 ; 21 , and for understanding the rheology of squishy matter 33 ; 2 ; 12 ; 8 . In regimes where particle deformations become large, the standard assumptions of granular mechanics, e.g. small strains, Hertzian contacts, and fixed particle shape, break down, requiring frameworks that explicitly couple particle-level shape change to the overall mechanical response 9 . Crucially, in these deformable particle assemblies, individual particle identities and interstitial voids persist even at high densities, which requires modeling approaches distinct from confluent frameworks used to study dense cellular packings, where cell boundaries are shared and voids are entirely absent 5 ; 17 ; 36 . This non-confluent distinction has consequences for localized stress transmission and contact network stability under driving. How the interplay between shape deformability and the persistence of interstitial voids influences the capacity of jammed assemblies to self-organize and encode memory under cyclic mechanical driving is our focus. Disordered solids encode memories of their mechanical history, a phenomenon demonstrated across a range of systems, from sheared non-Brownian suspensions and bubble rafts to crumpled elastic sheets 35 ; 31 ; 25 ; 19 ; 37 ; 1 . Atomistic and elastoplastic modeling have established that single and multiple memories can be encoded via cyclic driving, which acts as a mechanical annealing process that progressively drives the system toward stable reversible limit cycles 11 ; 20 . The encoded memory can, however, be fragile – it is erased when the driving amplitude exceeds the training amplitude, and is ultimately bounded by the irreversibility transition, the critical amplitude beyond which the system enters a regime of plastic diffusion 26 . Microscopically, memories are organized via localized plastic rearrangements around soft spots that, through training, become reliably activated and can be modeled as hysteretic two-state units; when interactions between these units become frustrated, they induce multi-periodic cycles and enable the emergence of complex memory behaviors 39 ; 27 . The vast majority of these studies, however, concern systems where changes in particle shape play no role – whether deformability introduces qualitatively new memory mechanisms remains an open question that this work addresses directly. To explore these aspects, we model two-dimensional athermal assemblies of deformable soft matter as ring polymers and investigate their response to quasistatic compression and cyclic driving. Beyond jamming, individual rings buckle under compression, producing a compliant “squishy jammed” state marked by a pronounced softening of the bulk modulus. Under cyclic driving, these assemblies self-organize into stable limit cycles whose nature depends critically on structural disorder – monodisperse systems undergo athermal annealing toward near-crystalline, nearly reversible paths, recovering memory through ordering, while polydisperse systems converge to persistent hysteresis loops encoding robust structural memory within a disordered state, stable against extreme overdriving and nested subcycles. This resilience originates in the topological stability of the self-organized contact network, preserved even under large post-buckling deformations. The macroscopic hysteresis arises from directionally asymmetric non-affine deformations that redistribute internal contact forces within a largely frozen network, a sub-topological irreversibility mechanism absent in rigid-particle systems. These findings reveal that the ability of rings to absorb compressive stress through shape change protects the self-organized topology under driving, giving rise to a class of highly compliant jammed assemblies capable of robustly encoding and reproducing trained mechanical states. The remainder of this paper is organized as follows. In Sec.II, we describe the ring polymer model and the three systems studied, as well as the quasistatic compression protocol. Results are presented in Sec.III, followed by a discussion in Sec.IV. II Model and methods We consider a two-dimensional assembly of N=1000 deformable particles modeled as ring polymers 38 ; 13 . For the interaction between monomers, we used a combination of Weeks-Chandler-Andersen (WCA) potential U_{LJ}(r)=\begin{cases}4\epsilon[(\frac{\sigma_{m}}{r})^{12}-(\frac{\sigma_{m}}{r})^{6}]+\epsilon&\text{if }r\leq 2^{\frac{1}{6}}\sigma_{m}\\ 0&\text{if }r>2^{\frac{1}{6}}\sigma_{m}\end{cases} (1) between all monomers, and finite extensible nonlinear elastic (FENE) potential U_{FENE}(r)=-\epsilon k_{F}R_{F}^{2}\ln[1-(\frac{r}{R_{F}\sigma_{m}})^{2}]\quad\text{if }r<R_{F}\sigma_{m} (2) between bonded monomers, where \sigma_{m} is the diameter of each monomer (and the unit of length), \epsilon is the unit of energy, k_{F}=15 is the spring constant, and R_{F}=1.5 is the maximum extension of the bond 38 . To model the bending flexibility of the ring backbone, we implement an angular potential acting on the angle \phi between consecutive bond vectors: V(\phi)=K_{\phi}(1-\cos(\phi-\pi)) (3) We consider three model systems. (i) A monodisperse system (labelled M0 ) in which all rings have the same number of monomers ( n_{m}=100 ) and the same angular stiffness K_{\theta}=100 . (ii) A stiffness-polydisperse system (labelled K30 ) in which all rings have n_{m}=100 monomers but differ in angular stiffness, with K_{\theta} sampled from a uniform distribution of mean \overline{K_{\theta}}=100 . (iii) A size-polydisperse system (labelled P30 ) in which all rings have the same stiffness K_{\theta}=100 but differ in the number of monomers, with n_{m} sampled from a uniform distribution of mean \overline{n_{m}}=100 . In both polydisperse systems, polydispersity is defined as the ratio of the standard deviation of the distribution to its mean, expressed as a percentage; following our prior study 28 , we use 30\% polydispersity in both cases. Throughout this work, all energy, density, and pressure values are reported in reduced Lennard-Jones units. Initial athermal configurations are generated by sampling equilibrium supercooled states at T=1.0 28 . These configurations are then instantaneously quenched to their respective inherent structures using conjugate gradient minimization. This sampling is performed at low

Read the complete text