Thesis--Massachusetts Institute of Technology. / Includes bibliographical references (p. 264-270).

Thesis (M. S.)--Chemical Engineering, Georgia Institute of Technology, 2006. / Dr. Athanasios Nenes, Committee Chair ; Dr. Amyn Teja, Committee Member ; Dr. Rodney Weber, Committee Member.

Thesis (Ph. D.)--Mechanical Engineering, Georgia Institute of Technology, 2007. / Mark Papania, MD, Committee Member ; Mark Allen, Committee Member ; Yves Berthelot, Committee Member ; Ari Glezer, Committee Member ; F. Levent Degertekin, Committee Chair ; Andrei G. Fedorov, Committee Chair.

In particulate flows, the flow inertia impacts the motion and size distribution of the particles and this in turn, has a strong implication on global behavior of the emulsions such as their rheological properties. As such, the central goal of most of the investigations on dispersed multiphase flow, so far, has been to understand the phase distribution of particles and to correlate the global behavior of the system with this parameter. For pressure-driven particulate flows in a channel, it is known that the velocity gradient in the channel leads to a lateral force whose magnitude and direction depends on the viscosity and density ratios of the fluids and the drop deformation. This lateral (lift) force is the primary reason behind the various observed modes of phase distribution of the particles. Unfortunately, most of the studies conducted so far have been concerned with the solid particles and for flows at low to moderate Reynolds numbers. Little is known about the dynamics of deformable drops at high Reynolds numbers. The goal of this study is to bridge the gap by direct numerical simulations. A front tracking/finite difference technique is used to solve the Navier-Stokes equations in the fluids inside and outside of the drops. Initially, the drops are randomly distributed in the computational domain their evolutions are followed for a sufficiently long time so that the system reaches a quasi-steady state. The statistics about the flow then will be extracted. The flow inertia is increased incrementally by increasing the pressure gradient.

In the recent years, there has been a growing interest in droplet-based (digital) microfluidic systems due to their ability to handle multiple discrete samples in a self-contained configuration compared to continuous flow systems. Various methods for droplet manipulation are currently available based on hydrodynamic, electrostatic, chemical, photonic and thermal interactions. High speed, controlled response and minimal thermal loading with least contamination are required in practical applications, especially in chemistry and biology. Although, thermal actuation of droplets has been recognized as an attractive choice due to a wide range of thermomechanical properties that can be exploited, the previous studies yielded limited success in addressing issues such as droplet evaporation, contamination, pinning, hysteresis and irreversibility that are associated with using solid substrate platforms In order to overcome shortcomings of traditional approaches, a novel thermally-actuated droplet manipulation platform based on using an inert liquid film was proposed and its working mechanisms were studied. Droplets at the air-liquid interface of immiscible liquids usually form partially submerged lens shapes (e.g. water on oil). In the thermally-induced motion of droplets on the free surface of immiscible liquid films, lens-shaped droplets move from warm toward cooler regions. In addition to this structure, we showed that droplets released from critical heights above the target liquid can sustain the impact iv and at the end maintain a spherical ball-shape configuration above the surface, despite undergoing large deformation. It was discovered in this study that such spherical droplets migrate in the opposite direction to lens droplets when subject to a thermal gradient; i.e. direction of increasing temperatures. The existence of this metastable spherical state above the free surface and its transition into more stable lens configuration was investigated using optical diagnostic tools and theoretical analysis. Opposite direction of motion observed for droplets at the free surface of immiscible liquids was explained based on droplet shape at the interface and the dynamics of thin liquid films subject to lateral thermal gradients: mainly 1) deformation of the free surface and 2) development of an outward moving flow (hot to cold) at the free surface due to surface tension gradients caused by thermal gradients. A lens droplet moves due to the free surface flow caused by Marangoni convection which is from hot to cold. On the other hand, the spherical droplet moves towards the maximum depression on the free surface, occurring at the hottest region as a result of the balance between gravity and drag forces from the opposing free surface flow. The proposed theoretical models predict experimental observations of droplet motion due to thermal gradients satisfactorily. Opposite responses of thermally-induced motion of lens and spherical droplets on a thin liquid layer, were characterized experimentally and compared to theory by v studying droplet motion in an exponentially-decaying temperature field maintained across the length of a shallow liquid layer. The effect of droplet size and magnitude of thermal gradient (slope) on drop velocity were investigated. The down-scaling effect is prominent, which shows that the proposed concept of droplet manipulation could be used favorably in miniaturized platforms. Based on the theoretical development and measurements obtained from meso-scale experiments, a silicon-based droplet transportation platform with embedded metal film micro heaters was developed. A thin layer of a chemically-inert and thermally stable liquid was chosen as the carrier liquid. Heaters were interfaced with control electronics and driven through a computer graphical user interface. By creating appropriate spatio-temporal thermal gradient maps, transport of droplets on predetermined pathways was demonstrated with a high level of controllability and speed.

No description available.

In this research it is attempted to determine whether the pressure drop for a polymer melt flowing transversely through a square array of cylinders can be predicted using purely viscous models or whether a viscoelastic constitutive equation is required. To predict these pressure drops, finite element calculations were performed using the generalized Newtonian fluid (GNF) model with the Bird-Carreau viscosity function, as well as two viscoelastic constitutive equations, the Phan-Thien and Tanner (PTT) model and the Rivlin-Sawyers (RS) model with the Pap ana sta siou, Scriven, and Macosko (PSM) damping :function. The constitutive equations were fit to the steady shear viscosity of a LLDPE melt and a LDPE melt. The PTT and RS models were also fit to uniaxial extensional stress growth data for each melt. The predictions of the pressure drop by means of the finite element calculations and a capillary model based on Darcys Law were compared to pressure drops from experiments performed with the two polymer melts. The agreement between experimental data and theoretical predictions was best for the calculations using the PTT model. The calculations using the PTT model as the constitutive equation indicate that time dependent fluid properties and extensional rheology must be correctly predicted by the constitutive equation used if accurate pressure drops of viscoelastic fluids flowing through banks of cylinders are to be calculated. This research is also concerned with the comparison of the results of numerical simulation of confined flow past a cylinder to birefringence data for two polymer melts. The Phan-Thien and Tanner (PTT) constitutive equation and the Rivlin-Sawyers constitutive equation with the Papanastasiou, Scriven, and Macosko (PSM) damping function were each fit to the shear viscosity and extensional viscosity data of both linear low-density polyethylene (LLDPE) and low-density polyethylene (LDPE) melts to determine the values of the model parameters. Finite element calculations were carried out using the 4x4SUPG and 4x4SU methods for the PTT model and the method developed by Dupont et al. for the RS model. Isochromatic birefringence patterns calculated from the predicted stress field and the stress-optic law were compared to birefringence data. Agreement was found between the birefringence data and the numerical predictions, except in the immediate vicinity of the cylinder surface. Large extensional stresses were observed and predicted along the centerline downstream of the cylinder for LDPE. This behavior was not observed or predicted for LLDPE. Stress fields obtained from birefringence measurements for LDPE flowing past three cylinders in a channel indicate an effect of deformation history on the flow behavior of LDPE. It is shown that the PTT model does not correctly predict the rheological behavior of LDPE as a function of shear history because the time scale of structural recovery is much longer than the relaxation time associated with viscoelasticity. / Ph. D.

Self-cleaning, anti-corrosion, anti-icing, dropwise-condensation, and drag-reduction are some applications in which superhydrophobic surfaces are implemented. To date, all the studies associated with superhydrophobic surfaces have been dedicated to understanding the liquid interaction with surfaces that are macroscopically smooth. The current study investigates the solid-liquid interaction of such surfaces which are fully decorated with macroscopic ridges (ribbed surfaces). In particular, the drop motion and impact on our newly designed non-wettable ribbed surface have been investigated in this work. Our experimental investigations have shown that liquid drops move faster on the ribbed surfaces due to lower friction induced by such a surface pattern. Moreover, an impacting droplet shows shorter contact time on ribbed surfaces. This concludes that ribbed surface pattern can be an efficient alternative design for the related applications. Besides the experimental studies, the theoretical analyses done in this work have led to, firstly a scaling model to predict descent velocity of a rolling viscous drops on an inclined non-wettable surface more accurately. Secondly, for curved superhydrophobic surfaces a scaling model which correlates the contact time of the impacting drop to its impact velocity has been developed. At the end, the knowledge obtained from this work has led to a special surface design which exhibits a contact time shorter than the inertial-capillary time scale, an unprecedented phenomenon.

No description available.

Runtime analysis of applications can help to gain insight into control flow of applications as well as detect performance issues. This work presents efficient means for integrating runtime monitoring facilities into the DROPS operating system and uses these to analyse performance and behavior of L4-based applications such as L4Linux.