Part I deals with the simulation of fluid dynamics using the lattice-Boltzmann method. Microfluidic devices often feature two-dimensional, repetitive arrays. Flows through such devices are pressure-driven and confined by solid walls. We have defined new adaptive generalised periodic boundary conditions to represent the effects of outer solid walls, and are thus able to exploit the periodicity of the array by simulating the flow through one unit cell in lieu of the entire device. The so-calculated fully developed flow describes the flow through the entire array accurately, but with computational requirements that are reduced according to the dimensions of the array. / Part II discusses the problem of separating macromolecules like proteins or DNA coils. The reliable separation of such molecules is a crucial task in molecular biology. The use of Brownian ratchets as mechanisms for the separation of such particles has been proposed and discussed during the last decade. Pressure-driven flows have so far been dismissed as possible driving forces for Brownian ratchets, as they do not generate ratchet asymmetry. We propose a microfluidic design that uses pressure-driven flows to create asymmetry and hence allows particle separation. The dependence of the asymmetry on various factors of the microfluidic geometry is discussed. We further exemplify the feasibility of our approach using Brownian dynamics simulations of particles of different sizes in such a device. The results show that ratchet-based particle separation using flows as the driving force is possible. Simulation results and ratchet theory predictions are in excellent agreement. / Part III deals with the co-evolution of networks and dynamic models. A group of agents occupies the nodes of a network, which defines the relationship between these agents. The evolution of the agents is defined by the rules of the dynamic model and depends on the relationship between agents, i.e., the state of the network. In return, the evolution of the network depends on the state of the dynamic model. The concept is introduced through the adaptive SIS model. We show that the previously used criterion determining the critical infected fraction, i.e., the number of infected agents required to sustain the epidemic, is inappropriate for this model. We introduce a different criterion and show that the critical infected fraction so determined is in good agreement with results obtained by numerical simulations. / This thesis comprises three parts, reporting research results in Fluid Dynamics (Part I), Particle Separation (Part II) and Co-evolving Networks (Part III). / We further discuss the concept of co-evolving dynamics using the Snowdrift Game as a model paradigm. Co-evolution occurs through agents cutting dissatisfied links and rewiring to other agents at random. The effect of co-evolution on the emergence of cooperation is discussed using a mean-field theory and numerical simulations. A transition between a connected and a disconnected, highly cooperative state of the system is observed, and explained using the mean-field model. Quantitative deviations regarding the level of cooperation in the disconnected regime can be fully resolved through an improved mean-field theory that includes the effect of random fluctuations into its model. / Graeser, Oliver = 自主個體為本系統與在流體力學、分子分離、共同演化網絡上的應用 / 顧皓森. / Adviser: Hui Pak-Ming. / Source: Dissertation Abstracts International, Volume: 73-01, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2010. / Includes bibliographical references (leaves 204-216). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese. / Graeser, Oliver = Zi zhu ge ti wei ben xi tong yu zai liu ti li xue, fen zi fen li, gong tong yan hua wang luo shang de ying yong / Gu Haosen.

Thesis (MSc)--Stellenbosch University, 2004. / ENGLISH ABSTRACT: In the ongomg search for better and faster ways to characterize complex polymer systems, it is often necessary to couple different analytical techniques in order to obtain information on more than one distributed property. In this study, the coupling of chromatography and spectroscopy to atomic force microscopy (AFM) was attempted for the first time, and thus the term "LC-FTIR-AFM" was coined. This new hyphenated technique combines the separation power of liquid chromatography (LC) and the ability of infrared spectroscopy (IR) to identify almost any organic compound, with the AFM's ability to be used for thermal analysis of individual phases in polymer mixtures. The first two steps of this new technique comprise (i) the separation of compounds in a mixture via gradient polymer elution chromatography (GPEC) and (ii) the identification of each compound by means of LC-FTIR analysis. In the final step, LC-FTIR analysis is coupled to AFM through the use of the LC-FTIR interface. A number of polymer mixtures were analysed by means of the novel technique that was developed, in order to establish its validity and value as a characterization technique of the future. The influence of fllm thickness and molar mass on the thermal parameters of individual components in mixtures, measured by this technique, were also investigated. This technique adds a new dimension to conventional thermal analysis methods, since it allows the thermal transitions of individual polymer phases in multiphase polymers to be resolved directly after separation and identification. / AFRIKAANSE OPSOMMING: In die voortdurende soektog na beter en vinniger maniere om komplekse polimeersisteme te karakteriseer, is dit soms nodig om verskillende analitiese tegnieke met mekaar te koppel ten einde inligting aangaande twee of meer verspreide eienskappe te bekom. Gedurende hierdie studie is daar gepoog om chromatografie en spektroskopie met atoominteraksie-mikroskopie (atomic force microscopy, AFM) te koppel. Gevolglik het die term "LC-FTIR-AFM" ontstaan. Hierdie nuwe koppelingstegniek kombineer die kragtige skeidingspotensiaal van vloeistofchromatografie en die vermoë van infrarooispektroskopie om byna enige organiese verbinding positief te identifiseer, met die atoominteraksie-mikroskoop se potensiaal om as 'n termiese analise metode vir individuele fases in polimeermengsels te dien. Die eerste twee stappe van die tegniek behels (i) die skeiding van verbindings in 'n mengsel deur middel van gradient-hoë-druk-vloeistofchromatografie en (ii) die identifisering van afsonderlike verbindings deur vloeistofchromatografie gekoppel aan infrarooispektroskopie. Gedurende die finale stap word vloeistofchromatografie en infrarooispektroskopie aan die atoominteraksie-mikroskoop gekoppel deur gebruik te maak van die LC-FTIR koppelingsapparaat. 'n Aantal polimeermengsels is geanaliseer deur die nuwe tegniek hierbo beskryf, ten einde die geldigheid en waarde daarvan as 'n analitiese metode vir die toekoms vas te stel. Die invloed van film diktes en molekulêre massa op die termiese oorgange van individuele komponente in mengsels, soos gemeet deur hierdie metode, is ook ondersoek. Hierdie tegniek voeg 'n nuwe dimensie tot konvensionele termiese analise metodes deurdat dit die bepaling van termiese oorgange van individuele polimeerfases III multifase polimere, direk na afloop van skeiding en identifikasie moontlik maak.