MEAM Seminar Series Summer 2012

This research focuses on the dynamics of soft elastic particles in viscous flows. The particle is assumed to be an incompressible, neo-Hookean elastic solid. To describe the elastic deformation, an evolution equation for the extra stress tensor is derived. A monolithic finite element solver which uses the Arbitrary Lagrangian-Eulerian moving mesh technique is implemented to solve the velocity, pressure and stress in both fluid and solid phase simultaneously. When the velocity gradient of the external flow is constant (e.g., pure strain, simple shear), it is demonstrated that an ellipsoidal particle will undergo a homogeneous deformation with a uniform field of stress and pressure. An analytical theory has been developed to describe the finite-strain, time-dependent response of the neo-Hookean elastic particles by using a polarization technique originally developed for the classical Eshelby's problem in the linear elasticity.  A set of coupled, nonlinear, first-order ODEs is obtained for the evolution of the uniform stress fields in the particle, as well as for the shape and orientation of the particle. In the dilute limit, the macroscopic rheological properties of a suspension of elastic particles are thus derived conveniently. The theory has been successfully applied to the soft particle in motion in both simple shear and extensional flows. Other objects of this study, including the deformation of elastic particles undergoing electrophoresis and possible experimental measurements, are also reported.

A solid solution can spontaneously separate into phases, e.g. spinodal decomposition, that self assemble into patterns. This process can be significantly affected by material anisotropy, including heterogeneous elastic properties, non-dilational transformation strains, and orientation-dependent interfacial energy. A Cahn-Hilliard type phase field model is developed that incorporates chemical, interfacial and elastic energies and couples naturally to externally-imposed mechanical fields. Aggregation in the bulk and in thin films under patterned external tractions is investigated through simulations. The kinetics of aggregation and morphology of precipitates are shown to depend strongly upon material anisotropy  and through coupling with external load. A major contribution in this work is the demonstration that the observation in simulations can be interpreted theoretically using Eshelby-type estimated of self and interaction engeries between precipitates and external loads.

Trussed structures are prevalent in everyday life. We use truss construction for the interior structural support of large buildings (Eiffel Tower or power transmission towers) as wella as for temporary scaffolding to support workers and materials along the exterior. These structures are usually assembled on-site by human workers. However, there are many instances when construction of these structures is better suited for robots. These situations include environments with extreme temperature, high toxicity (e.g. nuclear incidents or structurally unsafe areas (e.g. collapsed buildings). In these instances, more automated approaches are needed to build these structures.

This thesis will develop general methods for automated robotic construction. In order to test these approaches, a robotic construction infrastructure was constructed at the GRASP Lab of the University of Pennsylvania. This system is comprised of several quadrotor helicopters with simple grippers, simple modular construction elements, and a motion capture system. Using the limitations and constraints of this infrastructure, several algorithms for constructing truss-like cubic structures using these modular elements are developed. First, an algorithm, which can construct a class of cubic structures without holes, is described. This algorithm is extended to several new algorithms capable of hole closure, which effectively expands the class of cubic structures to those that contain holes. Finally, these approaches are generalized for a larger class of lattice types including tetrahedral lattices.

Wear is one of the main factors that hinders the performance of probes for atomic force microscopy (AFM), including for the widely used amplitude modulation (AM-AFM) mode. To conduct consistent and quantitative AM-AFM wear experiments, we have developed a protocol that involved controlling the tip-sample interaction regime, determining the tip-sample interaction geometry, calculating the peak repulsive force and normal stress over the course the wear test, and quantifying the wear volume using high-resolution transmission electron microscopy imaging. The peak repulsive tip-sample interaction force is estimated from a closed-form equation that uses the Derjaguin-Müller-Toporov interaction model accompanied by an effective tip radius measurement procedure, which combines transmission electron microscopy with blind tip reconstruction. The contact radius, deformation, and stress are estimated by numerically solving a general contact mechanics model recently developed for the adhesive contact of arbitrary axisymmetric punch shapes by Zhou et al., J. of the Mech. and Phys. of Solids, 2011.   The applicability of this protocol is demonstrated experimentally by using diamond-like carbon coated and silicon nitride coated silicon probes scanning ultrananocrystalline diamond samples in repulsive-mode AM-AFM. We discuss the important role that the assumed tip shape geometry plays in calculating both the interaction forces and the contact stresses. Contact stress is significantly affected by the tip geometry while the peak repulsive force is mainly determined by experimental parameters, specifically, the free oscillation amplitude and amplitude ratio. There is no sign of fracture or plastic deformation in the case of diamond-like carbon; wear could be characterized as a gradual atom-by-atom process. In contrast, silicon nitride wears through removal of the cluster of atoms and plastic deformation.

Touching a real physical object with a hand-held tool causes the tool to experience high-frequency (tactile) accelerations that reflect the mechanical characteristics of the contact. These haptic signals provide salient cues about changes in tool-surface contact state and enable effortless identification of material and surface properties. While humans make extensive use of these cues, robots almost universally cannot sense them. Fortunately, the recent availability of low-cost MEMS-based accelerometers makes it practical to instrument robotic systems so that they too can take advantage of these cues.

This talk presents a suite of methods we have developed for enabling robots to be more aware of their physical interactions in operating conditions ranging from direct teleoperation to full autonomy. We will focus on the design and control of a haptic system capable of accurately recreating tactile acceleration signals experienced by a teleoperated robot in real time. This system has been implemented on the Intuitive Surgical da Vinci Surgical System, an FDA-approved telerobotic system that natively provides no haptic feedback. Building on prior work, we use MEMS-based accelerometers to provide real-time measurement of the high frequency accelerations experienced by the robot as a result of environmental contact. We use a dedicated linear voice coil actuator to generate high fidelity recreations of the tactile acceleration signals for the user to feel at the operator interface. This approach involves signal processing methods used to enhance the measured accelerations and modeling to carefully control the acceleration output of the voice coil actuator. The provided feedback feels natural and promises to reduce the operator's cognitive load and increase their situational awareness.

 

Slender rods are ubiquitous in nature and widely used in engineering construction. Although most of these slender structures are mechanically supported, under large compressive load, they will still buckle and cause catastrophic structural failure. It is so far not well understood how Coulomb friction caused by the supporting medium affects the stability of rods. In the first part of the seminar, we will discuss the distinct mechanisms when compressed rods in frictional and frictionless contact lose stability: a frictionless rod buckles as its stiffness becomes negative; a rod in frictional contact, in contrast, can bear substantial amount of perturbation without buckling after its stiffness turns negative. Buckling is initiated as perturbation tolerance decreases below the level set by the environment, at a much higher critical load. In the second part of the seminar, we will discuss the 1D to 2D to 3D configuration transitions of the constrained rods as they buckle. We will see that this transition can be tuned by and is highly sensitively to the supporting matrix stiffness. This property may be useful for future photonic devices.

This research is under the supervision of Dr. Katia Bertoldi at Harvard University and Dr. Pedro M. Reis at MIT. The study for the first part of the talk is funded by Schlumberger.

Nanoscale fluid phenomena are important in an increasing number of applications involving flows in micrometer- and nanometer-scale channels such as lab-on-a-chip and MEMS devices. At these scales, flow features can deviate from traditional continuum behavior due to the large surface effects in nanofluidics for example increased viscosity in nanochannels, solid-like fluid ordering near solid interfaces, and nearly frictionless flow in nanotubes. Atomistic methods, such as Molecular Dynamics (MD) simulations, are capable of fully resolving the flow fields, fluid proprieties, and interfacial structure in such confined geometries but macroscopic simulations however, are still out of reach due to the prohibitively large computational expense of modeling discrete particles. Hybrid atomistic-continuum (HAC) models offer a solution. HAC models limit the use of MD to only a small region where atomistic-level resolution is necessary and use continuum methods away from the region. These HAC models are a reasonable way to meet the twin goals of accuracy and efficiency. A HAC model for nano-scale fluid flow is presented in this talk. The domain is decomposed into an atomistic region where MD is used and a continuum domain where the Navier-Stokes equations are solved. The two domains are coupled through an overlap region where the solutions in the two domains are consistent. The continuity of mass, momentum, and energy is ensured through constraints on each region. The accuracy of the HAC model is demonstrated through the simulation of sudden start Couette flow and unsteady heat transfer in a nano-channel. Potential applications as well as planned future extensions to multiple phase flow will be discussed.

Nanomechanical logic based on nanoelectromechanical systems (NEMS) switches promises a significant reduction in total switching energy over conventional solid-state transistors. This energy savings is achieved using electrodes that are brought in and out of contact to form an electrically-controlled mechanical switch. Unlike the leaky dielectric conduction channels employed in fully-electronic transistors, no current flows when the NEMS switch is off (open). However, the reliability of the electrical contact interface in NEMS switches is a crucial barrier to their commercialization. The adhesiveness and reactivity of conventional contact materials (i.e. metals) results in permanent adhesion, or the buildup of insulating tribofilms, at the contact. Careful selection and testing of next-generation contact materials, and expanding the scientific understanding of electromechanical contact degradation, is imperative to achieve lifetimes above a quadrillion cycles, a requirement for the commercialization of NEMS logic.

Studying alternative electrical contact materials typically requires costly and time-consuming fabrication of a new device for each material under investigation. In the present work, a new rapid prototyping method was developed to interrogate the electrical robustness and adhesiveness of any contact material without the need for new device fabrication. The methodology is based on dynamic atomic force microscopy (AFM) to perform quantitative studies of single asperity contacts. Candidate contact materials can be effectively probed in conditions representative of NEMS switches (e.g., nanometer dimensions, nanonewton applied forces, relevant applied voltages across the contacts, and billions of contact cycles), allowing physical processes dominating the performance of nanoscale switches to be elucidated. This methodology can be performed in most commercial AFM systems

The new methodology was used to investigate the reliability of platinum-platinum electrical contacts for two billion cycles of contact. Increases in contact resistance up to four orders of magnitude were observed, similar to microscale/multi-asperity investigations, validating that this approach is representative of device-level behavior. Further insights into the physical and chemical phenomena leading to the degradation of platinum electrical contacts will be provided. Alternate contact materials with potential for superior performance will then be discussed.

Life immersed in a fluid is nothing unusual for an organism. They cope and take advantage of water or wind currents to move, feed, and reproduce. These fluidic environments range from simple, Newtonian fluids that flow like water to complex fluids that possess shear-rate dependent viscosity and viscoelasticity. Understanding the main mechanisms that govern the locomotion/motility of microorganisms in simple and complex fluids is of much practical interest with potential impact on drug delivery, robotics, medicine, and surveillance. The main question to be addressed in this talk is: how does the fluidic environment affect the motility behavior of micro-organisms?

In this talk, I will investigate the swimming behavior of the nematode Caenorhabditis (C.) elegans in both Newtonian and complex fluids. The nematode Caenorhabditis elegans is a small (1 mm long) roundworm that generates traveling waves to propel itself. In addition, C. elegans are an attractive organism since its genome has been completely sequenced and their complete cell lineage established. Using optical microscopy and a high-speed camera, we develop a set of methods to simultaneously track the nematode and trace the flow fields in order to characterize the motility behavior of C. elegans and the induced flow fields. In Newtonian fluids, we found that: (i) flow fields in Newtonian fluids exhibit an exponential decay trend, which agrees with the theoretical results in the low-Reynolds-number regime; (ii) calculations of propulsive forces from both resistive force theory and slender body theory agree well with the experimental data. As the nematode swims in viscoelastic fluids, we find that the presence of elasticity in the fluid can decrease the swimming speed and efficiency of C. elegans by up to 35% compared to Newtonian fluids of the same shear viscosity. As elasticity in the fluid is increased, the swimming speed also decreases. The flow fields generated by the swimmer may hold the answer to this phenomenon.

This work is supported by NSF-CAREER (CBET)-0954084 and Army Research Office.

Fluids containing polymers are of much interest spanning the petroleum, semiconductor, pharmaceutical, and chemical processing industries. They are also frequently encountered in everyday life in foods, paints, and cosmetics. Fluids containing polymer molecules do not flow like water. Even when flowing slowly, these fluids can exhibit hydrodynamic instabilities and a new type of turbulence - the so-called purely elastic turbulence even at low Reynolds numbers (Re) where linear viscous forces dominate non-linear inertial forces. These phenomena, driven by the extra elastic stresses in the flow due to the presence of polymer molecules in the fluid were experimentally observed in flows around objects (cylinders), Couette cells, and curved microchannels. A common feature of the above-mentioned geometries is the presence of curved streamlines, which are necessary for infinitesimal perturbations to be enhanced by the normal stress imbalances in viscoelastic flows. Thus, it is a common assumption that in the absence of curvature and inertia, the flow of viscoelastic fluids is linearly stable.

Here we present experimental results that suggest the existence of a nonlinear instability in flows with parallel streamlines at low Re. We perform experiments in a long, straight microchannel that is 100 μm deep, 100 μm wide and 3.3 cm long. The channel is divided into two main regions: a short (~ 0.3 cm) region where a linear array of cylinders (0 ≤ n ≤ 15) is positioned in order to introduce perturbations to the flow, and a long (~ 3.0 cm) parallel flow region in which the fate of an initial disturbance is monitored; a channel devoid of cylinders is also used for control. The flow is investigated using both dye advection and particle tracking velocimetry. Results show that the initial disturbance is sustained far downstream in the parallel shear geometry above certain Wissenberg number (Wi), and increase non-linearly with Wi even at vanishing small Re. Above a critical Weissenberg number (Wi > 5.4) and a critical number of obstacles (n ≥ 2), a sharply increase of velocity fluctuations together with a hysteresis loop indicate presence of a subcritical elastic instability. This scenario is akin to the transition from laminar to turbulent flow of Newtonian fluids in pipe and channel flows, except that the instability is caused by the nonlinear elastic stresses and not inertia.