Three parallel algorithms for classical molecular dynamics are presented. The first assigns each processor a fixed subset of atoms; the second assigns each a fixed subset of inter-atomic forces to compute; the third assigns each a fixed spatial region. The algorithms are suitable for molecular dynamics models which can be difficult to parallelize efficiently—those with short-range forces where the neighbors of each atom change rapidly. They can be implemented on any distributed-memory parallel machine which allows for message-passing of data between independently executing processors. The algorithms are tested on a standard Lennard-Jones benchmark problem for system sizes ranging from 500 to 100,000,000 atoms on several parallel supercomputers--the nCUBE 2, Intel iPSC/860 and Paragon, and Cray T3D. Comparing the results to the fastest reported vectorized Cray Y-MP and C90 algorithm shows that the current generation of parallel machines is competitive with conventional vector supercomputers even for small problems. For large problems, the spatial algorithm achieves parallel efficiencies of 90% and a 1840-node Intel Paragon performs up to 165 faster than a single Cray C9O processor. Trade-offs between the three algorithms and guidelines for adapting them to more complex molecular dynamics simulations are also discussed.

Fast parallel algorithms for short-range molecular dynamics

Microfabricated integrated circuits revolutionized computation by vastly reducing the space, labor, and time required for calculations. Microfluidic systems hold similar promise for the large-scale automation of chemistry and biology, suggesting the possibility of numerous experiments performed rapidly and in parallel, while consuming little reagent. While it is too early to tell whether such a vision will be realized, significant progress has been achieved, and various applications of significant scientific and practical interest have been developed. Here a review of the physics of small volumes (nanoliters) of fluids is presented, as parametrized by a series of dimensionless numbers expressing the relative importance of various physical phenomena. Specifically, this review explores the Reynolds number Re, addressing inertial effects; the Peclet number Pe, which concerns convective and diffusive transport; the capillary number Ca expressing the importance of interfacial tension; the Deborah, Weissenberg, and elasticity numbers De, Wi, and El, describing elastic effects due to deformable microstructural elements like polymers; the Grashof and Rayleigh numbers Gr and Ra, describing density-driven flows; and the Knudsen number, describing the importance of noncontinuum molecular effects. Furthermore, the long-range nature of viscous flows and the small device dimensions inherent in microfluidics mean that the influence of boundaries is typically significant. A variety of strategies have been developed to manipulate fluids by exploiting boundary effects; among these are electrokinetic effects, acoustic streaming, and fluid-structure interactions. The goal is to describe the physics behind the rich variety of fluid phenomena occurring on the nanoliter scale using simple scaling arguments, with the hopes of developing an intuitive sense for this occasionally counterintuitive world.

/pdf/microfluidics-fluid-physics-at-the-nanoliter-scale-4bpahtd8qa.pdf

Microfluidics: Fluid physics at the nanoliter scale

Electrowetting has become one of the most widely used tools for manipulating tiny amounts of liquids on surfaces. Applications range from 'lab-on-a-chip' devices to adjustable lenses and new kinds of electronic displays. In the present article, we review the recent progress in this rapidly growing field including both fundamental and applied aspects. We compare the various approaches used to derive the basic electrowetting equation, which has been shown to be very reliable as long as the applied voltage is not too high. We discuss in detail the origin of the electrostatic forces that induce both contact angle reduction and the motion of entire droplets. We examine the limitations of the electrowetting equation and present a variety of recent extensions to the theory that account for distortions of the liquid surface due to local electric fields, for the finite penetration depth of electric fields into the liquid, as well as for finite conductivity effects in the presence of AC voltage. The most prominent failure of the electrowetting equation, namely the saturation of the contact angle at high voltage, is discussed in a separate section. Recent work in this direction indicates that a variety of distinct physical effects?rather than a unique one?are responsible for the saturation phenomenon, depending on experimental details. In the presence of suitable electrode patterns or topographic structures on the substrate surface, variations of the contact angle can give rise not only to continuous changes of the droplet shape, but also to discontinuous morphological transitions between distinct liquid morphologies. The dynamics of electrowetting are discussed briefly. Finally, we give an overview of recent work aimed at commercial applications, in particular in the fields of adjustable lenses, display technology, fibre optics, and biotechnology-related microfluidic devices.

/pdf/electrowetting-from-basics-to-applications-14e6zoxse5.pdf

Electrowetting: from basics to applications

Intermolecular And Surface Forces

We present a method to solve initial and boundary value problems using artificial neural networks. A trial solution of the differential equation is written as a sum of two parts. The first part satisfies the initial/boundary conditions and contains no adjustable parameters. The second part is constructed so as not to affect the initial/boundary conditions. This part involves a feedforward neural network containing adjustable parameters (the weights). Hence by construction the initial/boundary conditions are satisfied and the network is trained to satisfy the differential equation. The applicability of this approach ranges from single ordinary differential equations (ODE), to systems of coupled ODE and also to partial differential equations (PDE). In this article, we illustrate the method by solving a variety of model problems and present comparisons with solutions obtained using the Galerkin finite element method for several cases of partial differential equations. With the advent of neuroprocessors and digital signal processors the method becomes particularly interesting due to the expected essential gains in the execution speed.

/pdf/artificial-neural-networks-for-solving-ordinary-and-partial-1pse3ki6jh.pdf

Artificial neural networks for solving ordinary and partial differential equations

Neural algorithm for solving differential equations

We report that a droplet dispensed from a micropipette almost always has a considerable electrical charge of a magnitude dependent on the constituents of the droplet, on atmospheric humidity and on the coating material of pipette tip. We show that this natural electrification of a droplet originates from the charge separation between a droplet and pipette tip surface by contact with water due to the ionization of surface chemical groups. Charge on a droplet can make it difficult to detach the droplet from the pipette tip, can decrease its surface tension, can affect the chemical characteristics of solutions due to interactions with charged molecules, and can influence the combination and localization of charged bio-molecules; in all cases, the charge may affect results of experiments in which any of these factors is important. Thus, these findings reveal experimental parameters that should be controlled in experiments that use micropipettes.

/pdf/spontaneous-electrical-charging-of-droplets-by-conventional-54kgmf84rm.pdf

Spontaneous electrical charging of droplets by conventional pipetting

Electrophoretic motion of a charged droplet in a dielectric fluid under an electric field has been investigated experimentally for use as a microdroplet actuation method. The effects of the droplet size, electric field strength, and electrolyte concentration and ion species on the charging of an aqueous droplet have been examined. The amount of electrical charging has been measured by two different methods: indirect measurement using the image analysis of droplet motion and direct measurement using the electrometer. Quantitative comparison of the droplet charge measured experimentally and the theoretical value of a perfectly conductive sphere shows that an aqueous droplet is less charged than the corresponding perfectly conductive sphere. The limiting effect on electrical charging is more significant for an electrolyte droplet, and the effect is positively correlated to the electrolyte concentration rather than the ion species. This implies that the low electrical conductivity of water is not a major cause of the limiting effect. The scaling law of the charging amount for a deionized water droplet nearly follows that of the perfect conductor, whereas for an electrolyte droplet, the scaling law exponent is slightly higher. Some advantages and potentials of the current droplet actuation method are also discussed in comparison with the conventional ones.

Electrophoresis of a charged droplet in a dielectric liquid for droplet actuation.

Liquid droplets possess many practically important applications and academically interesting issues. Accurate flow data are necessary to correlate the hydrodynamic characteristics with the physicochemical processes occurring inside a droplet. However, the refraction of light at the droplet surface makes it difficult to measure the flow field inside the droplet accurately. To resolve this problem, two correction methods based on the ray tracing technique are employed. One is the image mapping method and the other is the velocity mapping method. For this, a mapping function between the image plane and the object plane is derived. The two correction methods are applied to the flow inside evaporating droplets of different ethanol concentrations for measuring their velocity fields, using a PIV method. The results obtained with the two methods are nearly identical. The major differences between the original results and the corrected results are found in the locations of the vortex centres and the magnitude of velocity vectors. Between the two correction methods, the velocity mapping method is recommended, because it is more convenient and recovers a greater number of velocity vectors, compared with the image restoration method.

https://iopscience.iop.org/article/10.1088/0957-0233/15/6/009/pdf

In Seok Kang

Papers

Neural algorithm for solving differential equations

Spontaneous electrical charging of droplets by conventional pipetting

Electrophoresis of a charged droplet in a dielectric liquid for droplet actuation.

Quantitative visualization of flow inside an evaporating droplet using the ray tracing method

Electrical charging of a conducting water droplet in a dielectric fluid on the electrode surface