Abstract: We demonstrate that deep learning techniques can be used to predict motility induced phase separation (MIPS) in suspensions of active Brownian particles (ABPs) by creating a notion of phase at the particle level. Using a fully connected network in conjunction with a graph neural network we use individual particle features to predict to which phase a particle belongs. From this, we are able to compute the fraction of dilute particles to determine if the system is in the homogeneous dilute, dense, or coexistence region. Our predictions are compared against the MIPS binodal computed from simulation. The strong agreement between the two suggests that machine learning provides an effective way to determine the phase behavior of ABPs and could prove useful for determining more complex phase diagrams.

Publication: arXiv
ID: CaltechAUTHORS:20210106-130209799

]]>

Abstract: By introducing the notion of a dynamic overlap concentration scale, we identify universal and previously unreported features of the mechanical properties of active colloids. These features are codified by recognizing that the characteristic length scale of an active particle's trajectory, the run-length, introduces a new concentration scale ϕ∗. Large-scale simulations of repulsive active Brownian particles (ABPs) confirm that this new run-length dependent concentration, which is the trajectory-space analogue of the overlap concentration in polymer solutions, delineates distinct concentration regimes in which interparticle collisions alter particle trajectories. Using ϕ∗ and concentration scales associated with colloidal jamming, the mechanical equation-of-state for ABPs can be collapsed onto a set of principal curves that contain a number of previously overlooked features. The inclusion of these features qualitatively alters previous predictions of the behavior for active colloids as we demonstrate by computing the spinodal for a suspension of purely-repulsive ABPs. Our findings suggest that dynamic overlap concentration scales should be of great utility in unraveling the behavior of active and driven systems.

Publication: arXiv
ID: CaltechAUTHORS:20201005-103534186

]]>

Abstract: We thank IBS for their comments which question our interpretation of the universal viscosity divergence near the flow-arrest transition in constant stress and pressure rheology of hard-sphere colloidal suspensions [2]. IBS introduced two Péclet numbers: Pe_0 = ẏ_ɑ^2=d0 and Pe = ẏ_ɑ^2/d(φ), with ẏ the strain rate, ɑ the particle size, d0 the isolated single-particle diffusivity and d_0 the long- time at-rest self-diffusivity, and considered three regimes: (i) Pe_0 < Pe ≪ 1, (ii) Pe_0 ≪ 1 ≪ Pe, and (iii) 1 ≪ Pe_0 < Pe.

ID: CaltechAUTHORS:20160815-101200112

]]>

Abstract: We present a general theory for determining the force (and torque) exerted on a boundary (or body) in active matter. The theory extends the description of passive Brownian colloids to self-propelled active particles and applies for all ratios of the thermal energy kBT to the swimmer's activity k_sT_s = ζU^2_0T_R/6, where ζ is the Stokes drag coefficient, U_0 is the swim speed and T_R is the reorientation time of the active particles. The theory has a natural microscopic length scale over which concentration and orientation distributions are confined near boundaries, but the microscopic length does not appear in the force. The swim pressure emerges naturally and dominates the behavior when the boundary size is large compared to the swimmer's run length ℓ = U_0T_R. The theory is used to predict the motion of bodies of all sizes immersed in active matter.

ID: CaltechAUTHORS:20160120-145959501

]]>

Abstract: We introduce fractal liquids by generalizing classical liquids of integer dimensions d=1,2,3 to a fractal dimension df. The particles composing the liquid are fractal objects and their configuration space is also fractal, with the same non-integer dimension. Realizations of our generic model system include microphase separated binary liquids in porous media, and highly branched liquid droplets confined to a fractal polymer backbone in a gel. Here we study the thermodynamics and pair correlations of fractal liquids by computer simulation and semi-analytical statistical mechanics. Our results are based on a model where fractal hard spheres move on a near-critical percolating lattice cluster. The predictions of the fractal Percus-Yevick liquid integral equation compare well with our simulation results.

ID: CaltechAUTHORS:20150713-082840560

]]>