Eric E. Keaveny

Optimization of chiral structures for microscale propulsion.

Recent advances in micro- and nanoscale fabrication techniques allow for the construction of rigid, helically shaped microswimmers that can be actuated using applied magnetic fields. These swimmers represent the first steps toward the development of microrobots for targeted drug delivery and minimally invasive surgical procedures. To assess the performance of these devices and improve on their design, we perform shape optimization computations to determine swimmer geometries that maximize speed in the direction of a given applied magnetic torque. We directly assess aspects of swimmer shapes that have been developed in previous experimental studies, including helical propellers with elongated cross sections and attached payloads. From these optimizations, we identify key improvements to existing designs that result in swimming speeds that are 70-470% of their original values.

A comparative study between dissipative particle dynamics and molecular dynamics for simple- and complex-geometry flows

The purpose of this study is to compare the results from molecular-dynamics and dissipative particle dynamics (DPD) simulations of Lennard-Jones (LJ) fluid and determine the quantitative effects of DPD coarse graining on flow parameters. We illustrate how to select the conservative force coefficient, the cut-off radius, and the DPD time scale in order to simulate a LJ fluid. To show the effects of coarse graining and establish accuracy in the DPD simulations, we conduct equilibrium simulations, Couette flow simulations, Poiseuille flow simulations, and simulations of flow around a periodic array of square cylinders. For the last flow problem, additional comparisons are performed against continuum simulations based on the spectral/hp element method.

Modeling the magnetic interactions between paramagnetic beads in magnetorheological fluids

In this study, we develop and compare new and existing methods for computing the magnetic interactions between paramagnetic particles in magnetorheological (MR) fluids. The commonly employed point-dipole methods are outlined and the inter-particle magnetic forces, given by these representations, are compared with exact values. An alternative finite-dipole model, where the magnetization of a particle is represented as a distribution of current density, is described and the associated computational effort is shown to scale as O(N). As the dipole moments and forces given by this model depend on the length scale of the current distribution, a sensitivity analysis is performed to reveal a proper choice of this length scale. While the dipole models give a good estimation of the far-field interactions, as two particles come into contact, higher order multipoles are needed to properly resolve their interaction. We present the exact two-body calculation and describe a procedure to include the higher multipoles arising in a pairwise interaction into a dipole model. This inclusion procedure can be integrated with any dipole or higher-multipole calculation. Results from relevant three-body problems are compared to exact solutions to provide information as to how well the inclusion procedure performs in simulations of self-assembly and estimating the yield strength of structures.

Experiments and theory of undulatory locomotion in a simple structured medium

Undulatory locomotion of micro-organisms through geometrically complex, fluidic environments is ubiquitous in nature and requires the organism to negotiate both hydrodynamic effects and geometrical constraints. To understand locomotion through such media, we experimentally investigate swimming of the nematode Caenorhabditis elegans through fluid-filled arrays of micro-pillars and conduct numerical simulations based on a mechanical model of the worm that incorporates hydrodynamic and contact interactions with the lattice. We show that the nematodes path, speed and gait are significantly altered by the presence of the obstacles and depend strongly on lattice spacing. These changes and their dependence on lattice spacing are captured, both qualitatively and quantitatively, by our purely mechanical model. Using the model, we demonstrate that purely mechanical interactions between the swimmer and obstacles can produce complex trajectories, gait changes and velocity fluctuations, yielding some of the life-like dynamics exhibited by the real nematode. Our results show that mechanics, rather than biological sensing and behaviour, can explain some of the observed changes in the worms locomotory dynamics.

Force-coupling method for flows with ellipsoidal particles

The force-coupling method, previously developed for spherical particles suspended in a liquid flow, is extended to ellipsoidal particles. In the limit of Stokes flow, there is an exact correspondence with known analytical results for isolated particles. More generally, the method is shown to provide good approximate results for the particle motion and the flow field both in viscous Stokes flow and at finite Reynolds number. This is demonstrated through comparison between fully resolved direct numerical simulations and results from the numerical implementation of the force-coupling method with a spectral/hp element scheme. The motion of settling ellipsoidal particles and neutrally buoyant particles in a Poiseuille flow are discussed.

Fluctuating force-coupling method for simulations of colloidal suspensions

Abstract The resolution of Brownian motion in simulations of micro-particle suspensions can be crucial to reproducing the correct dynamics of individual particles, as well as providing an accurate characterisation of suspension properties. Including these effects in simulations, however, can be computationally intensive due to the configuration dependent random displacements that would need to be determined at every time step. In this paper, we introduce the fluctuating force-coupling method (FCM) to overcome this difficulty, providing a fast approach to simulate colloidal suspensions at large-scale. We show explicitly that by forcing the surrounding fluid with a fluctuating stress and employing the FCM framework to obtain the motion of the particles, one obtains the random particle velocities and angular velocities that satisfy the fluctuation–dissipation theorem. This result holds even when higher-order multipoles, such as stresslets, are included in the FCM approximation. Through several numerical experiments, we confirm our analytical results and demonstrate the effectiveness of fluctuating FCM, showing also how Brownian drift can be resolved by employing the appropriate time integration scheme and conjugate gradient method.

Applying a second-kind boundary integral equation for surface tractions in Stokes flow

A second-kind integral equation for the tractions on a rigid body moving in a Stokesian fluid is established using the Lorentz reciprocal theorem and an integral equation for a double-layer density. A second-order collocation method based on the trapezoidal rule is applied to the integral equation after appropriate singularity reduction. For translating prolate spheroids with various aspect ratios, the scheme is used to explore the effects of the choice of completion flow on the error in the numerical solution, as well as the condition number of the discretized integral operator. The approach is applied to obtain the velocity and viscous dissipation of rotating helices of circular cross-section. These results are compared with both local and non-local slender-body theories. Motivated by the design of artificial micro-swimmers, similar computations are performed on previously unstudied helices of non-circular cross-section to determine the dependence of the velocity and propulsive efficiency on the cross-section aspect ratio and orientation. Overall, we find that this formulation provides a stable numerical approach with which to solve the flow problem while simultaneously obtaining the surface tractions and that the appropriate choice of completion flow provides both increased accuracy and efficiency. Additionally, this approach naturally avails itself to known fast summation techniques and higher-order quadrature schemes.

Large-scale simulation of steady and time-dependent active suspensions with the force-coupling method

We present a new development of the force-coupling method (FCM) to address the accurate simulation of a large number of interacting micro-swimmers. Our approach is based on the squirmer model, which we adapt to the FCM framework, resulting in a method that is suitable for simulating semi-dilute squirmer suspensions. Other effects, such as steric interactions, are considered with our model. We test our method by comparing the velocity field around a single squirmer and the pairwise interactions between two squirmers with exact solutions to the Stokes equations and results given by other numerical methods. We also illustrate our methods ability to describe spheroidal swimmer shapes and biologically-relevant time-dependent swimming gaits. We detail the numerical algorithm used to compute the hydrodynamic coupling between a large collection (104-105) of micro-swimmers. Using this methodology, we investigate the emergence of polar order in a suspension of squirmers and show that for large domains, both the steady-state polar order parameter and the growth rate of instability are independent of system size. These results demonstrate the effectiveness of our approach to achieve near continuum-level results, allowing for better comparison with experimental measurements while complementing and informing continuum models.

Simulating Brownian suspensions with fluctuating hydrodynamics.

Fluctuating hydrodynamics has been successfully combined with several computational methods to rapidly compute the correlated random velocities of Brownian particles. In the overdamped limit where both particle and fluid inertia are ignored, one must also account for a Brownian drift term in order to successfully update the particle positions. In this paper, we present an efficient computational method for the dynamic simulation of Brownian suspensions with fluctuating hydrodynamics that handles both computations and provides a similar approximation as Stokesian Dynamics for dilute and semidilute suspensions. This advancement relies on combining the fluctuating force-coupling method (FCM) with a new midpoint time-integration scheme we refer to as the drifter-corrector (DC). The DC resolves the drift term for fluctuating hydrodynamics-based methods at a minimal computational cost when constraints are imposed on the fluid flow to obtain the stresslet corrections to the particle hydrodynamic interactions. With the DC, this constraint needs only to be imposed once per time step, reducing the simulation cost to nearly that of a completely deterministic simulation. By performing a series of simulations, we show that the DC with fluctuating FCM is an effective and versatile approach as it reproduces both the equilibrium distribution and the evolution of particulate suspensions in periodic as well as bounded domains. In addition, we demonstrate that fluctuating FCM coupled with the DC provides an efficient and accurate method for large-scale dynamic simulation of colloidal dispersions and the study of processes such as colloidal gelation.

An in-depth numerical study of the two-dimensional Kuramoto–Sivashinsky equation

The Kuramoto–Sivashinsky equation in one spatial dimension (1D KSE) is one of the most well-known and well-studied partial differential equations. It exhibits spatio-temporal chaos that emerges through various bifurcations as the domain length increases. There have been several notable analytical studies aimed at understanding how this property extends to the case of two spatial dimensions. In this study, we perform an extensive numerical study of the Kuramoto–Sivashinsky equation (2D KSE) to complement this analytical work. We explore in detail the statistics of chaotic solutions and classify the solutions that arise for domain sizes where the trivial solution is unstable and the long-time dynamics are completely two-dimensional. While we find that many of the features of the 1D KSE, including how the energy scales with system size, carry over to the 2D case, we also note several differences including the various paths to chaos that are not through period doubling.