### Refine

#### Year of publication

- 2006 (24) (remove)

#### Document Type

- Doctoral Thesis (24) (remove)

#### Language

- English (24) (remove)

#### Keywords

- IMRT (2)
- Kontinuumsmechanik (2)
- Optimization (2)
- computational mechanics (2)
- "Slender-Body"-Theorie (1)
- Ananasgewächse (1)
- Asymptotik (1)
- Atom optics (1)
- Atomoptik (1)
- Beam orientation (1)

#### Faculty / Organisational entity

This thesis introduces so-called cone scalarising functions. They are by construction compatible with a partial order for the outcome space given by a cone. The quality of the parametrisations of the efficient set given by the cone scalarising functions are then investigated. Here, the focus lies on the (weak) efficiency of the generated solutions, the reachability of effiecient points and continuity of the solution set. Based on cone scalarising functions Pareto Navigation a novel, interactive, multiobjective optimisation method is proposed. It changes the ordering cone to realise bounds on partial tradeoffs. Besides, its use of an equality constraint for the changing component of the reference point is a new feature. The efficiency of its solutions, the reachability of efficient solutions and continuity is then analysed. Potential problems are demonstrated using a critical example. Furthermore, the use of Pareto Navigation in a two-phase approach and for nonconvex problems is discussed. Finally, its application for intensity-modulated radiotherapy planning is described. Thereby, its realisation in a graphical user interface is shown.

The topic of this thesis is the coupling of an atomistic and a coarse scale region in molecular dynamics simulations with the focus on the reflection of waves at the interface between the two scales and the velocity of waves in the coarse scale region for a non-equilibrium process. First, two models from the literature for such a coupling, the concurrent coupling of length scales and the bridging scales method are investigated for a one dimensional system with harmonic interaction. It turns out that the concurrent coupling of length scales method leads to the reflection of fine scale waves at the interface, while the bridging scales method gives an approximated system that is not energy conserving. The velocity of waves in the coarse scale region is in both models not correct. To circumvent this problems, we present a coupling based on the displacement splitting of the bridging scales method together with choosing appropriate variables in orthogonal subspaces. This coupling allows the derivation of evolution equations of fine and coarse scale degrees of freedom together with a reflectionless boundary condition at the interface directly from the Lagrangian of the system. This leads to an energy conserving approximated system with a clear separation between modeling errors an errors due to the numerical solution. Possible approximations in the Lagrangian and the numerical computation of the memory integral and other numerical errors are discussed. We further present a method to choose the interpolation from coarse to atomistic scale in such a way, that the fine scale degrees of freedom in the coarse scale region can be neglected. The interpolation weights are computed by comparing the dispersion relations of the coarse scale equations and the fully atomistic system. With this new interpolation weights, the number of degrees of freedom can be drastically reduced without creating an error in the velocity of the waves in the coarse scale region. We give an alternative derivation of the new coupling with the Mori-Zwanzig projection operator formalism, and explain how the method can be extended to non-zero temperature simulations. For the comparison of the results of the approximated with the fully atomistic system, we use a local stress tensor and the energy in the atomistic region. Examples for the numerical solution of the approximated system for harmonic potentials are given in one and two dimensions.

The fast development of the financial markets in the last decade has lead to the creation of a variety of innovative interest rate related products that require advanced numerical pricing methods. Examples in this respect are products with a complicated strong path-dependence such as a Target Redemption Note, a Ratchet Cap, a Ladder Swap and others. On the other side, the usage of the standard in the literature one-factor Hull and White (1990) type of short rate models allows only for a perfect correlation between all continuously compounded spot rates or Libor rates and thus are not suited for pricing innovative products depending on several Libor rates such as for example a "steepener" option. One possible solution to this problem deliver the two-factor short rate models and in this thesis we consider a two-factor Hull and White (1990) type of a short rate process derived from the Heath, Jarrow, Morton (1992) framework by limiting the volatility structure of the forward rate process to a deterministic one. In this thesis, we often choose to use a variety of modified (binomial, trinomial and quadrinomial) tree constructions as a main numerical pricing tool due to their flexibility and fast convergence and (when there is no closed-form solution) compare their results with fine grid Monte Carlo simulations. For the purpose of pricing the already mentioned innovative short-rate related products, in this thesis we offer and examine two different lattice construction methods for the two-factor Hull-White type of a short rate process which are able to deal easily both with modeling of the mean-reversion of the underlying process and with the strong path-dependence of the priced options. Additionally, we prove that the so-called rotated lattice construction method overcomes the typical for the existing two-factor tree constructions problem with obtaining negative "risk-neutral probabilities". With a variety of numerical examples, we show that this leads to a stability in the results especially in cases of high volatility parameters and negative correlation between the base factors (which is typically the case in reality). Further, noticing that Chan et al (1992) and Ritchken and Sankarasubramanian (1995) showed that option prices are sensitive to the level of the short rate volatility, we examine the pricing of European and American options where the short rate process has a volatility structure of a Cheyette (1994) type. In this relation, we examine the application of the two offered lattice construction methods and compare their results with the Monte Carlo simulation ones for a variety of examples. Additionally, for the pricing of American options with the Monte Carlo method we expand and implement the simulation algorithm of Longstaff and Schwartz (2000). With a variety of numerical examples we compare again the stability and the convergence of the different lattice construction methods. Dealing with the problems of pricing strongly path-dependent options, we come across the cumulative Parisian barrier option pricing problem. We notice that in their classical form, the cumulative Parisian barrier options have been priced both analytically (in a quasi closed form) and with a tree approximation (based on the Forward Shooting Grid algorithm, see e.g. Hull and White (1993), Kwok and Lau (2001) and others). However, we offer an additional tree construction method which can be seen as a direct binomial tree integration that uses the analytically calculated conditional survival probabilities. The advantage of the offered method is on one side that the conditional survival probabilities are easier to calculate than the closed-form solution itself and on the other side that this tree construction is very flexible in the sense that it allows easy incorporation of additional features such as e.g a forward starting one. The obtained results are better than the Forward Shooting Grid tree ones and are very close to the analytical quasi closed form solution. Finally, we pay our attention to pricing another type of innovative interest rate alike products - namely the Longevity bond - whose coupon payments depend on the survival function of a given cohort. Due to the lack of a market for mortality, for the pricing of the Longevity bonds we develop (following Korn, Natcheva and Zipperer (2006)) a framework that contains principles from both Insurance and Financial mathematic. Further on, we calibrate the existing models for the stochastic mortality dynamics to historical German data and additionally offer new stochastic extensions of the classical (deterministic) models of mortality such as the Gompertz and the Makeham one. Finally, we compare and analyze the results of the application of all considered models to the pricing of a Longevity bond on the longevity of the German males.

In this thesis, we have dealt with two modeling approaches of the credit risk, namely the structural (firm value) and the reduced form. In the former one, the firm value is modeled by a stochastic process and the first hitting time of this stochastic process to a given boundary defines the default time of the firm. In the existing literature, the stochastic process, triggering the firm value, has been generally chosen as a diffusion process. Therefore, on one hand it is possible to obtain closed form solutions for the pricing problems of credit derivatives and on the other hand the optimal capital structure of a firm can be analysed by obtaining closed form solutions of firm's corporate securities such as; equity value, debt value and total firm value, see Leland(1994). We have extended this approach by modeling the firm value as a jump-diffusion process. The choice of the jump-diffusion process was a crucial step to obtain closed form solutions for corporate securities. As a result, we have chosen a jump-diffusion process with double exponentially distributed jump heights, which enabled us to analyse the effects of jump on the optimal capital structure of a firm. In the second part of the thesis, by following the reduced form models, we have assumed that the default is triggered by the first jump of a Cox process. Further, by following Schönbucher(2005), we have modeled the forward default intensity of a firm as a geometric Brownian motion and derived pricing formulas for credit default swap options in a more general setup than the ones in Schönbucher(2005).

Matter-wave Optics of Dark-state Polaritons: Applications to Interferometry and Quantum Information
(2006)

The present work "Materwave Optics with Dark-state Polaritons: Applications to Interferometry and Quantum Information" deals in a broad sense with the subject of dark-states and in particular with the so-called dark-state polaritons introduced by M. Fleischhauer and M. D. Lukin. The dark-state polaritons can be regarded as a combined excitation of electromagnetic fields and spin/matter-waves. Within the framework of this thesis the special optical properties of the combined excitation are studied. On one hand a new procedure to spatially manipulate and to increase the excitation density of stored photons is described and on the other hand the properties are used to construct a new type of Sagnac Hybrid interferometer. The thesis is devided into four parts. In the introduction all notions necessary to understand the work are described, e.g.: electromagnetically induced transparency (EIT), dark-state polaritons and the Sagnac effect. The second chapter considers the method developed by A. Andre and M. D. Lukin to create stationary light pulses in specially dressed EIT-media. In a first step a set of field equations is derived and simplified by introducing a new set of normal modes. The absorption of one of the normal modes leads to the phenomenon of pulse-matching for the other mode and thereby to a diffusive spreading of its field envelope. All these considerations are based on a homogeneous field setup of the EIT preparation laser. If this restriction is dismissed one finds that a drift motion is superimposed to the diffusive spreading. By choosing a special laser configuration the drift motion can be tailored such that an effective force is created that counteracts the spreading. Moreover, the force can not only be strong enough to compensate the diffusive spreading but also to exceed this dynamics and hence to compress the field envelope of the excitation. The compression can be discribed using a Fokker-Planck equation of the Ornstein-Uhlenbeck type. The investigations show that the compression leads to an excitation of higher-order modes which decay very fast. In the last section of the chapter this exciation will be discussed in more detail and conditions will be given how the excitation of higher-order modes can be avoided or even suppressed. All results given in the chapter are supported by numerical simulatons. In the third chapter the matterwave optical properties of the dark-state polaritons will be studied. They will be used to construct a light-matterwave hybrid Sagnac interferometer. First the principle setup of such an interferometer will be sketched and the relevant equations of motion of light-matter interaction in a rotating frame will be derived. These form the basis of the following considerations of the dark-state polariton dynamics with and without the influence of external trapping potentials on the matterwave part of the polariton. It will be shown that a sensitivity enhancement compared to a passive laser gyroscope can be anticipated if the gaseous medium is initially in a superfluid quantum state in a ring-trap configuration. To achieve this enhancement a simultaneous coherence and momentum transfer is furthermore necessary. In the last part of the chapter the quantum sensitivity limit of the hybrid interferometer is derived using the one-particle density matrix equations incorporating the motion of the particles. To this end the Maxwell-Bloch equations are considered perturbatively in the rotation rate of the noninertial frame of reference and the susceptibility of the considered 3-level \(\Lambda\)-type system is derived in arbitrary order of the probe-field. This is done to determine the optimum operation point. With its help the anticipated quantum sensitivity of the light-matterwave hybrid Sagnac interferometer is calculated at the shot-noise limit and the results are compared to state-of-the-art laser and matterwave Sagnac interferometers. The last chapter of the thesis originates from a joint theoretical and experimental project with the AG Bergmann. This chapter does no longer consider the dark-state polaritons of the last two chapters but deals with the more general concept of dark states and in particular with the transient velocity selective dark states as introduced by E. Arimondo et al. In the experiment we could for the first time measure these states. The chapter starts with an introduction into the concept of velocity selective dark states as they occur in a \(\Lambda\)-configuration. Then we introduce the transient velocity selective dark-states as they occur in an particular extension of the \(\Lambda\)-system. For later use in the simulations the relevant equations of motion are derived in detail. The simulations are based on the solution of the generalized optical Bloch equations. Finally the experimental setup and procedure are explained and the theoretical and experimental results are compared.

In this thesis diverse problems concerning inflation-linked products are dealt with. To start with, two models for inflation are presented, including a geometric Brownian motion for consumer price index itself and an extended Vasicek model for inflation rate. For both suggested models the pricing formulas of inflation-linked products are derived using the risk-neutral valuation techniques. As a result Black and Scholes type closed form solutions for a call option on inflation index for a Brownian motion model and inflation evolution for an extended Vasicek model as well as for an inflation-linked bond are calculated. These results have been already presented in Korn and Kruse (2004) [17]. In addition to these inflation-linked products, for the both inflation models the pricing formulas of a European put option on inflation, an inflation cap and floor, an inflation swap and an inflation swaption are derived. Consequently, basing on the derived pricing formulas and assuming the geometric Brownian motion process for an inflation index, different continuous-time portfolio problems as well as hedging problems are studied using the martingale techniques as well as stochastic optimal control methods. These utility optimization problems are continuous-time portfolio problems in different financial market setups and in addition with a positive lower bound constraint on the final wealth of the investor. When one summarizes all the optimization problems studied in this work, one will have the complete picture of the inflation-linked market and both counterparts of market-participants, sellers as well as buyers of inflation-linked financial products. One of the interesting results worth mentioning here is naturally the fact that a regular risk-averse investor would like to sell and not buy inflation-linked products due to the high price of inflation-linked bonds for example and an underperformance of inflation-linked bonds compared to the conventional risk-free bonds. The relevance of this observation is proved by investigating a simple optimization problem for the extended Vasicek process, where as a result we still have an underperforming inflation-linked bond compared to the conventional bond. This situation does not change, when one switches to an optimization of expected utility from the purchasing power, because in its nature it is only a change of measure, where we have a different deflator. The negativity of the optimal portfolio process for a normal investor is in itself an interesting aspect, but it does not affect the optimality of handling inflation-linked products compared to the situation not including these products into investment portfolio. In the following, hedging problems are considered as a modeling of the other half of inflation market that is inflation-linked products buyers. Natural buyers of these inflation-linked products are obviously institutions that have payment obligations in the future that are inflation connected. That is why we consider problems of hedging inflation-indexed payment obligations with different financial assets. The role of inflation-linked products in the hedging portfolio is shown to be very important by analyzing two alternative optimal hedging strategies, where in the first one an investor is allowed to trade as inflation-linked bond and in the second one he is not allowed to include an inflation-linked bond into his hedging portfolio. Technically this is done by restricting our original financial market, which is made of a conventional bond, inflation index and a stock correlated with inflation index, to the one, where an inflation index is excluded. As a whole, this thesis presents a wide view on inflation-linked products: inflation modeling, pricing aspects of inflation-linked products, various continuous-time portfolio problems with inflation-linked products as well as hedging of inflation-related payment obligations.

In this study, 27 marine bacteria were screened for production of bioactive metabolites. Two strains from the surface of the soft coral Sinularia polydactyla, collected from the Red Sea, and three strains from different habitats in the North Sea were selected as a promising candidates for isolation of antimicrobial substances. A total of 50 compounds were isolated from the selected bacterial strains. From these metabolites 25 substances were known from natural sources, 10 substances were known as synthetic chemical and herein are reported as new natural products, and 13 metabolites are new. Two substances are still under elucidation. All new compounds were chemically and biologically characterized. Pseudoalteromonas sp. T268 produced simple phenol and oxindole derivatives. Production of homogentisic acid and WZ 268S-6 from this bacteria was affected by the salinity stress. WZ 268S-6 shows antimicrobial and cytotoxic activities. Its target is still unclear. Isolation of isatin from this strain points out for the possibility of using this substance as a chemotaxonomical marker for Alteromonas-like bacteria. A large number of nitro-substituted aromatic compounds were isolated from both Salegentibacter sp. T436 and Vibrio sp. WMBA1-4. They may be derived from metabolism of phenylalanine or tyrosine. From Salegentibacter sp. T436, 24 compounds were isolated, of which four compounds are new and six compounds were known as synthetic chemicals. WZ 436S-16 (dinitro-β-styrene) is the most potent antimicrobial and cytotoxic compound. It inhibits the oxygen uptake by N. coryli and causes apoptosis in the human promyelocytic leukaemia (HL-60 cells). From Vibrio sp. WMBA1-4, 13 new alkaloids were isolated, of which four were known as synthetic products and herein are reported as new substances from natural sources. The majority of these compounds show antimicrobial and cytotoxic activities. The cytotoxic activity of WMB4S-11 against the mouse lymphocytic leukaemia (L1210 cells) is due to the inhibition in the protein biosynthesis, while the remaining cytotoxic alkaloids have no effect on the synthesis of macromolecules in this cell line. The antibacterial activity of WMB4S-2, -11, -12, -13 and the antifungal activity of WMB4S-9 are not due to the inhibition in the macromolecules biosynthesis or in the oxygen uptake by the microorganisms. The biological activity of these nitro-aromatic compounds from Salegentibacter sp. T436 and Vibrio sp. WMBA1-4 is influenced by the presence of a nitro group and its position in respect to the hydroxyl group, number of the nitro groups, and the type of substitutions on the side chain. In diaryl-maleimide derivatives, types and position of substitution on the aryl rings, on the maleimide moity, and the hydrophobicity of the aryl ring itself lead to variations in the extent of the bioactivity of these derivatives. This is the first time that vibrindole (WMB4S-14) and turbomycin B or its noncationic form (WMB4S-15), isolated from Vibrio sp., are reported as cytotoxic compounds. WMB4S-15 inhibits the biosynthesis of macromolecules in L1210 cells. The structural similarity between some of the metabolites in this study and previously reported compounds from sponges, ascidians, and bryozoan indicates that the microbial origin of these compounds must be considered.

This thesis discusses methods for the classification of finite projective planes via exhaustive search. In the main part the author classifies all projective planes of order 16 admitting a large quasiregular group of collineations. This is done by a complete search using the computer algebra system GAP. Computational methods for the construction of relative difference sets are discussed. These methods are implemented in a GAP-package, which is available separately. As another result --found in cooperation with U. Dempwolff-- the projective planes defined by planar monomials are classified. Furthermore the full automorphism group of the non-translation planes defined by planar monomials are classified.

This thesis deals with modeling aspects of generalized Newtonian and of non-Newtonian fluids, as well as with development and validation of algorithms used in simulation of such fluids. The main contribution in the modeling part are the introduction and analysis of a new model for the generalized Newtonian fluids, where constitutive equation is of an algebraic form. Distinction between shear and extensional viscosities leads to anisotropic viscosity model. It can be considered as a natural extension of the well known (isotropic viscosity) Carreau model, which deals only with shear viscosity properties of the fluid. The proposed model takes additionally into account extensional viscosity properties. Numerical results show that the anisotropic viscosity model gives much better agreement with experimental observations than the isotropic one. Another contribution of the thesis consists of the development and analysis of robust and reliable algorithms for simulation of generalized Newtonian fluids. For such fluids the momentum equations are strongly coupled through mixed derivatives appearing in the viscous term (unlike the case of Newtonian fluids). It is shown in this thesis, that a careful treatment of those derivatives is essential in deriving robust algorithms. A modification of a standard SIMPLE-like algorithm is given, where all the viscous terms from the momentum equations are discretized in an implicit manner. Moreover, it is shown that a block diagonal preconditioner to the viscous operator is good enough to be used in simulations. Furthermore, different solution techniques, namely projection type methods (consists of solving momentum equations and pressure correction equation) and fully coupled methods (momentum and continuity equations are solved together), are compared. It is shown, that explicit discretization of the mixed derivatives lead to stability problems. Further, analytical estimates of eigenvalue distribution for three different preconditioners, applied to the transformed system arising after discretization and linearization of the momentum and continuity equations, are provided. We propose to apply a block Gauss-Seidel preconditioner to the transformed system. The analysis shows, that this preconditioner is able to cluster eigenvalues around unity independent of the transformation step. It is not the case for other preconditioners applied to the transformed system as discussed in the thesis. The block Gauss-Seidel preconditioner has also shown the best behavior (among all preconditioners discussed in the thesis) in numerical experiments. Further contribution consists of comparison and validation of numerical algorithms applied in simulations of non-Newtonian fluids modeled by time integral constitutive equations. Numerical results from simulations of dilute polymer solutions, described by the integral Oldroyd B model, have shown very good quantitative agreement with the results obtained by differential Oldroyd B counterpart in 4:1 planar contraction domain at low Weissenberg numbers. In this case, the Weissenberg number is changed by changing the relaxation time. However, contrary to the differential Oldroyd B model, the integral one allows to perform stable simulations also in the range of high Weissenberg numbers. Moreover, very good agreement with experimental observations has been achieved. Simulations of concentrated polymer solutions (polystyrene and polybutadiene solutions), modeled by the integral Doi Edwards model, supplemented by chain length fluctuations, have shown very good qualitative agreement with the results obtained by its differential approximation in 4:1:4 constriction domain. Again, much higher Weissenberg numbers can be achieved when the integral model is used. Moreover, very good quantitative results with experimental data of polystyrene solution for the first normal stress difference and shear viscosity defined here as the quotient of a shear stress and a shear rate. Finally, comparison of the two methods used for approximating the time integral constitutive equation, namely Deformation Field Method (DFM) and Backward Lagrangian Particle Method (BLPM), is performed. In BLPM the particle paths are recalculated at every time step of the simulations, what has never been tried before. The results have shown, that in the considered geometries both methods give similar results.

For the last decade, optimization of beam orientations in intensity-modulated radiation therapy (IMRT) has been shown to be successful in improving the treatment plan. Unfortunately, the quality of a set of beam orientations depends heavily on its corresponding beam intensity profiles. Usually, a stochastic selector is used for optimizing beam orientation, and then a single objective inverse treatment planning algorithm is used for the optimization of beam intensity profiles. The overall time needed to solve the inverse planning for every random selection of beam orientations becomes excessive. Recently, considerable improvement has been made in optimizing beam intensity profiles by using multiple objective inverse treatment planning. Such an approach results in a variety of beam intensity profiles for every selection of beam orientations, making the dependence between beam orientations and its intensity profiles less important. This thesis takes advantage of this property to accelerate the optimization process through an approximation of the intensity profiles that are used for multiple selections of beam orientations, saving a considerable amount of calculation time. A dynamic algorithm (DA) and evolutionary algorithm (EA), for beam orientations in IMRT planning will be presented. The DA mimics, automatically, the methods of beam's eye view and observer's view which are recognized in conventional conformal radiation therapy. The EA is based on a dose-volume histogram evaluation function introduced as an attempt to minimize the deviation between the mathematical and clinical optima. To illustrate the efficiency of the algorithms they have been applied to different clinical examples. In comparison to the standard equally spaced beams plans, improvements are reported for both algorithms in all the clinical examples even when, for some cases, fewer beams are used. A smaller number of beams is always desirable without compromising the quality of the treatment plan. It results in a shorter treatment delivery time, which reduces potential errors in terms of patient movements and decreases discomfort.

The new international capital standard for credit institutions (“Basel II”) allows banks to use internal rating systems in order to determine the risk weights that are relevant for the calculation of capital charge. Therefore, it is necessary to develop a system that enfolds the main practices and methods existing in the context of credit rating. The aim of this thesis is to give a suggestion of setting up a credit rating system, where the main techniques used in practice are analyzed, presenting some alternatives and considering the problems that can arise from a statistical point of view. Finally, we will set up some guidelines on how to accomplish the challenge of credit scoring. The judgement of the quality of a credit with respect to the probability of default is called credit rating. A method based on a multi-dimensional criterion seems to be natural, due to the numerous effects that can influence this rating. However, owing to governmental rules, the tendency is that typically one-dimensional criteria will be required in the future as a measure for the credit worthiness or for the quality of a credit. The problem as described above can be resolved via transformation of a multi-dimensional data set into a one-dimensional one while keeping some monotonicity properties and also keeping the loss of information (due to the loss of dimensionality) at a minimum level.

Traffic flow on road networks has been a continuous source of challenging mathematical problems. Mathematical modelling can provide an understanding of dynamics of traffic flow and hence helpful in organizing the flow through the network. In this dissertation macroscopic models for the traffic flow in road networks are presented. The primary interest is the extension of the existing macroscopic road network models based on partial differential equations (PDE model). In order to overcome the difficulty of high computational costs of PDE model an ODE model has been introduced. In addition, steady state traffic flow model named as RSA model on road networks has been dicsussed. To obtain the optimal flow through the network cost functionals and corresponding optimal control problems are defined. The solution of these optimization problems provides an information of shortest path through the network subject to road conditions. The resulting constrained optimization problem is solved approximately by solving unconstrained problem invovling exact penalty functions and the penalty parameter. A good estimate of the threshold of the penalty parameter is defined. A well defined algorithm for solving a nonlinear, nonconvex equality and bound constrained optimization problem is introduced. The numerical results on the convergence history of the algorithm support the theoretical results. In addition to this, bottleneck situations in the traffic flow have been treated using a domain decomposition method (DDM). In particular this method could be used to solve the scalar conservation laws with the discontinuous flux functions corresponding to other physical problems too. This method is effective even when the flux function presents more than one discontinuity within the same spatial domain. It is found in the numerical results that the DDM is superior to other schemes and demonstrates good shock resolution.

Wetting of a solid surface with liquids is an important parameter in the chemical engineering process such as distillation, absorption and desorption. The degree of wetting in packed columns mainly contributes in the generating of the effective interfacial area and then enhancing of the heat and mass transfer process. In this work the wetting of solid surfaces was studied in real experimental work and virtually through three dimensional CFD simulations using the multiphase flow VOF model implemented in the commercial software FLUENT. That can be used to simulate the stratified flows [1]. The liquid rivulet flow which is a special case of the film flow and mostly found in packed columns has been discussed. Wetting of a solid flat and wavy metal plate with rivulet liquid flow was simulated and experimentally validated. The local rivulet thickness was measured using an optically assisted mechanical sensor using a needle which is moved perpendicular to the plate surface with a step motor and in the other two directions using two micrometers. The measured and simulated rivulet profiles were compared to some selected theoretical models founded in the literature such as Duffy & Muffatt [2], Towell & Rothfeld [3] and Al-Khalil et al. [4]. The velocity field in a cross section of a rivulet flow and the non-dimensional maximum and mean velocity values for the vertical flat plate was also compared with models from Al-Khalil et al. [4] and Allen & Biggin [5]. Few CFD simulations for the wavy plate case were compared to the experimental findings, and the Towel model for a flat plate [3]. In the second stage of this work 3-D CFD simulations and experimental study has been performed for wetting of a structured packing element and packing sheet consisting of three elements from the type Rombopak 4M, which is a product of the company Kuhni, Switzerland. The hydrodynamics parameters of a packed column, e. i. the degree of wetting, the interfacial area and liquid hold-up have been depicted from the CFD simulations for different liquid systems and liquid loads. Flow patterns on the degree of wetting have been compared to that of the experiments, where the experimental values for the degree of wetting were estimated from the snap shooting of the flow on the packing sheet in a test rig. A new model to describe the hydrodynamics of packed columns equipped with Rombopak 4M was derived with help of the CFD–simulation results. The model predicts the degree of wetting, the specific or interfacial area and liquid hold-up at different flow conditions. This model was compared to Billet & Schultes [6], the SRP model Rocha et al. [7-9], to Shi & Mersmann [10] and others. Since the pressure drop is one of the most important parameter in packed columns especially for vacuum operating columns, few CFD simulations were performed to estimate the dry pressure drop in a structured and flat packing element and were compared to the experimental results. It was found a good agreement from one side, between the experimental and the CFD simulation results, and from the other side between the simulations and theoretical models for the rivulet flow on an inclined plate. The flow patterns and liquid spreading behaviour on the packing element agrees well with the experimental results. The VOF (Volume of Fluid) was found very sensitive to different liquid properties and can be used in optimization of the packing geometries and revealing critical details of wetting and film flow. An extension of this work to perform CFD simulations for the flow inside a block of the packing to get a detailed picture about the interaction between the liquid and packing surfaces is recommended as further perspective.

Uncoupling protein1 (UCP1) in brown adipose tissue was discovered earlier as the main uncoupling source of respiration. We describe the basic facts and a modest contribution of our group to the area of research on mitochondrial uncoupling proteins. After defining the terms uncoupling, leak, proton-mediated uncoupling, we discuss the assumption that due to its low abundance, uncoupling protein 2 (UCP2) can provide only mild uncoupling, i.e. can decrease the proton motive force by several mV only. A fatty acid cycling mechanism is described as a plausible explanation for the protonophoretic function of all uncoupling proteins together with our experiments supporting it. A speculation for the phylogenesis of all uncoupling proteins can be deduced by estimated UCP2 content in several tissues, and details of its activation are explained on the basis of our experiments. In the present study a solubilization and refolding method for UCP2 from inclusion bodies was developed and characterized. As it was known and also demonstrated from previous experiments on UCP1 that fatty acids are substrates, we used the same procedure to study the function of UCP2. Utilizing spin-labelled fatty acids (SLFA) for our experiments we demonstrated the binding of fatty acids to UCP2, and the competition of other natural fatty acids like oleic acid, palmitic acid, arachidonic acid and eicosatrienoic acid to the preformed complex emphasizes the presence of a fatty acid binding site for mitochondrial UCP2. The findings were observed by EPR spectroscopy where the highly immobilized spectra with presence of spin-labelled fatty acid eventually end up as free spin label spectra with a particular concentration of the natural fatty acid added to the UCP2 bound with spin-labelled fatty acid. This fits in significantly with the earlier findings of UCP1 and also leads to assumption of functional explanation about the physiological relevance between the uncoupling proteins functions. The present study, in which representative and sensitive parameters for EPR spectroscopy were established, at the same time describes the concentration effects of fatty acids upon the protein bound with spin-labelled fatty acids which are much of importance in comparison to physiological levels, being in the micromolar range (µM) as compared with milli molar (mM) as for UCP1 previously. In appropriate examples, different fatty acids are used and compared with competitors like alkylsulfonates also emphasizing the function of the protein. And the studies with the effect of nucleotides inhibition demonstrate that there exists a putative binding site for fatty acids. Much significance lies in demonstration with the spin-labelled-ATP studies where competition of ATP to the protein bound to spin-labelled ATP explains about the inhibition effect of nucleotides on the UCP2. So the present study applies different methods for the functional characterization of UCP2. The studies of natural fatty acids and alkylsulfonates with UCP2 bound to spin-labelled fatty acid, and study of nucleotide inhibition on UCP2 are closely related and give the much awaited answer to the question of functional similarities between UCP1 and UCP2. This supports the discussion of many groups which predict the functional similarity between these two proteins based upon sequence homology. Also many attempts have been reported in literature to explain the physiological functional relevance where by this present study can also be added to as we now suppose from the present conclusions of our experiments.

Tropical geometry is a rather new field of algebraic geometry. The main idea is to replace algebraic varieties by certain piece-wise linear objects in R^n, which can be studied with the aid of combinatorics. There is hope that many algebraically difficult operations become easier in the tropical setting, as the structure of the objects seems to be simpler. In particular, tropical geometry shows promise for application in enumerative geometry. Enumerative geometry deals with the counting of geometric objects that are determined by certain incidence conditions. Until around 1990, not many enumerative questions had been answered and there was not much prospect of solving more. But then Kontsevich introduced the moduli space of stable maps which turned out to be a very useful concept for the study of enumerative geometry. A well-known problem of enumerative geometry is to determine the numbers N_cplx(d,g) of complex genus g plane curves of degree d passing through 3d+g-1 points in general position. Mikhalkin has defined the analogous number N_trop(d,g) for tropical curves and shown that these two numbers coincide (Mikhalkin's Correspondence Theorem). Tropical geometry supplies many new ideas and concepts that could be helpful to answer enumerative problems. However, as a rather new field, tropical geometry has to be studied more thoroughly. This thesis is concerned with the ``translation'' of well-known facts of enumerative geometry to tropical geometry. More precisely, the main results of this thesis are: - a tropical proof of the invariance of N_trop(d,g) of the position of the 3d+g-1 points, - a tropical proof for Kontsevich's recursive formula to compute N_trop(d,0) and - a tropical proof of Caporaso's and Harris' algorithm to compute N_trop(d,g). All results were derived in joint work with my advisor Andreas Gathmann. (Note that tropical research is not restricted to the translation of classically well-known facts, there are actually new results shown by means of tropical geometry that have not been known before. For example, Mikhalkin gave a tropical algorithm to compute the Welschinger invariant for real curves. This shows that tropical geometry can indeed be a tool for a better understanding of classical geometry.)

The study provides insights into the dynamic processes of vascular epiphyte vegetation in two host tree species of lowland forest in Panama. Further, a novel approach is presented to examine the possible role of host tree identity in the structuring of vascular epiphyte communities: For three locally common host tree species (Socratea exorrhiza, Marila laxiflora, Perebea xanthochyma) we created null models of the expected epiphyte assemblages assuming that epiphyte colonization reflected random distribution of epiphytes in the forest. In all three tree species, abundances of the majority of epiphyte species (69 – 81 %) were indistinguishable from random, while the remaining species were about equally over- or underrepresented compared to their occurrence in the entire forest plot. Permutations based on the number of colonized trees (reflecting observed spatial patchiness) yielded similar results. Finally, a Canonical Correspondence Analysis also confirmed host-specific differences in epiphyte assemblages. In spite of pronounced preferences of some epiphytes for particular host trees, no epiphyte species was restricted to a single host. We conclude that the epiphytes on a given tree species are not simply a random sample of the local species pool, but there are no indications of host specificity either. To determine the qualitative and quantitative long-term changes in the vascular epiphyte assemblage of the host tree Socratea exorrhiza, in the lowland forest of the San Lorenzo Crane Plot, we followed the fate of the vascular epiphyte assemblage on 99 individuals of this palm species, in three censuses over the course of five years. The composition of the epiphyte assemblage changed little during the course of the study. While the similarity of epiphyte vegetation decreased on single palm individuals through time, the similarity analyzed over all palms increased. Even well-established epiphyte individuals experienced high mortality with only 46 % of the originally mapped individuals surviving the following five years. We found a positive correlation between host tree size and epiphyte richness and detected higher colonization rates of epiphytes per surface area on larger trees. Epiphyte assemblages on single Socratea exorrhiza trees were highly dynamic while the overall composition of the epiphyte vegetation on the host tree species in the study plot was rather stable. We suggest that higher recruitment rates due to localized seed dispersal by already established epiphytes on larger palms promote the colonization of epiphytes on larger palms. Given the known growth rates and mortality rates of the host tree species, the maximum time available for colonization and reproduction of epiphytes on a given Socratea exorrhiza tree is estimated to be about 60 years. Changes in the epiphyte vegetation of c. 1000 individuals of the host tree species Annona glabra at Barro Colorado Island over the course of eight year were documented by means of repeated censuses. Considerable increase in the abundance of the dominating epiphyte species and ongoing colonization of the host tree species suggests that the epiphyte vegetation has not reached a steady state in the maximal 80 years since the establishment of the host tree. Epiphyte species composition as a whole was rather stable. We disentangled the relationship between epiphyte colonization and tree size/available time for colonization with the finding that tree size explained only a low proportion of colonization while other factors like connectivity to dispersal source and time explain may explain a larger part. Epiphyte populations are patchily distributed and examined species exhibit properties of a metapopulation with asynchronous local population growth, high local population turnover, a positive relationship between regional occurrence and patch population size, and negatively correlated relationship between extinction and patch occupancy. The documented metapopulation processes highlight the importance of not colonized suitable habitat for the conservation of epiphytes.

The validity of formulas w.r.t. a specification over first-order logic with a semantics based on all models is semi-decidable. Therefore, we may implement a proof procedure which finds a proof for every valid formula fully automatically. But this semantics often lacks intuition: Some pathological models such as the trivial model may produce unexpected results w.r.t. validity. Instead, we may consider just a class of special models, for instance, the class of all data models. Proofs are then performed using induction. But, inductive validity is not semi-decidable -- even for first-order logic. This theoretical drawback manifests itself in practical limitations: There are theorems that cannot be proved by induction directly but only generalizations can be proved. For their definition, we may have to extend the specification. Therefore, we cannot expect to prove interesting theorems fully automatically. Instead, we have to support user-interaction in a suitable way. In this thesis, we aim at developing techniques that enhance automatic proof control of (inductive) theorem provers and that enable user-interaction in a suitable way. We integrate our new proof techniques into the inductive theorem prover QuodLibet and validate them with various case studies. Essentially, we introduce the following three proof techniques: -We integrate a decision procedure for linear arithmetic into QuodLibet in a close way by defining new inference rules that perform the elementary steps of the decision procedure. This allows us to implement well-known heuristics for automatic proof control. Furthermore, we are able to provide special purpose tactics that support the manual speculation of lemmas if a proof attempt gets stuck. The integration improves the ability of the theorem prover to prove theorems automatically as well as its efficiency. Our approach is competitive with other approaches regarding efficiency; it provides advantages regarding the speculation of lemmas. -The automatic proof control searches for a proof by applying inference rules. The search space is not only infinite, but grows dramatically with the depth of the search. In contrast to this, checking and analyzing performed proofs is very efficient. As the search space also has a high redundancy, it is reasonable to reuse subproofs found during proof search. We define new notions for the contribution of proof steps to a proof. These notions enable the derivation of pruned proofs and the identification of superfluous subformulas in theorems. A proof may be reused in two ways: upward propagation prunes a proof by eliminating superfluous proof steps; sideward reuse closes an open proof obligation by replaying an already found proof. -For interactive theorem provers, it is essential not only to prove automatically as many lemmas as possible but also to restrict proof search in such a way that the proof process stops within a reasonable amount of time. We introduce different markings in the goals to be proved and the lemmas to be applied to restrict proof search in a flexible way: With a forbidden marking, we can simulate well-known approaches for applying conditional lemmas. A mandatory marking provides a new heuristics which is inspired by local contribution of proof steps. With obligatory and generous markings, we can fine-tune the degree of efficiency and extent of proof search manually. With an elaborate case study, we show the benefits of the different techniques, in particular the synergetic effect of their combination.

The primary object of this work is the development of a robust, accurate and efficient time integrator for the dynamics of flexible multibody systems. Particularly a unified framework for the computational dynamics of multibody systems consisting of mass points, rigid bodies and flexible beams forming open kinematic chains or closed loop systems is developed. In addition, it aims at the presentation of (i) a focused survey of the Lagrangian and Hamiltonian formalism for dynamics, (ii) five different methods to enforce constraints with their respective relations, and (iii) three alternative ways for the temporal discretisation of the evolution equations. The relations between the different methods for the constraint enforcement in conjunction with one specific energy-momentum conserving temporal discretisation method are proved and their numerical performances are compared by means of theoretical considerations as well as with the help of numerical examples.

With the burgeoning computing power available, multiscale modelling and simulation has these days become increasingly capable of capturing the details of physical processes on different scales. The mechanical behavior of solids is oftentimes the result of interaction between multiple spatial and temporal scales at different levels and hence it is a typical phenomena of interest exhibiting multiscale characteristic. At the most basic level, properties of solids can be attributed to atomic interactions and crystal structure that can be described on nano scale. Mechanical properties at the macro scale are modeled using continuum mechanics for which we mention stresses and strains. Continuum models, however they offer an efficient way of studying material properties they are not accurate enough and lack microstructural information behind the microscopic mechanics that cause the material to behave in a way it does. Atomistic models are concerned with phenomenon at the level of lattice thereby allowing investigation of detailed crystalline and defect structures, and yet the length scales of interest are inevitably far beyond the reach of full atomistic computation and is rohibitively expensive. This makes it necessary the need for multiscale models. The bottom line and a possible avenue to this end is, coupling different length scales, the continuum and the atomistics in accordance with standard procedures. This is done by recourse to the Cauchy-Born rule and in so doing, we aim at a model that is efficient and reasonably accurate in mimicking physical behaviors observed in nature or laboratory. In this work, we focus on concurrent coupling based on energetic formulations that links the continuum to atomistics. At the atomic scale, we describe deformation of the solid by the displaced positions of atoms that make up the solid and at the continuum level deformation of the solid is described by the displacement field that minimize the total energy. In the coupled model, continuum-atomistic, a continuum formulation is retained as the overall framework of the problem and the atomistic feature is introduced by way of constitutive description, with the Cauchy-Born rule establishing the point of contact. The entire formulation is made in the framework of nonlinear elasticity and all the simulations are carried out within the confines of quasistatic settings. The model gives direct account to measurable features of microstructures developed by crystals through sequential lamination.

Nowadays piezoelectric and ferroelectric materials are becoming more and more an interesting part of smart materials in scientific and engineering applications. Precision machining in manufacturing, micropositioning in metrology, common rail systems with piezo fuel injection control in automobile industry, and ferroelectric random access memories (FRAM) in microelectromechanical systems (MEMS) besides commercial piezo actuators and sensors can be very good examples for the application of piezoceramic and ferroelectric materials. In spite of having good characteristics, piezoelectric and ferroelectric materials have significant nonlinearities, which limit the applications in high performance usage. Domain switching (ferroelastic or ferroelectric) is the main reason for the nonlinearity of ferroelectric materials. External excessive electromechanical loads (mechanical stress and electric field) are driving forces for domain switching. In literature, various important experiments related to the non-linear properties of piezoelectric and ferroelectric materials are reported. Simulations of nonlinear properties of piezoelectric and ferroelectric materials based on physical insights of the material have been performed during the last two decades by using micromechanical and phenomenological models. The most significant experiments and models are deeply discussed in the literature survey. In this thesis the nonlinear behaviour of tetragonal perovskite type piezoceramic materials is simulated theoretically using two and three dimensional micromechanical models which are based on physical insights of the material. In the simulations a bulk piezoceramic material which has numerous grains is considered. Each grain has random orientation in properties of polarization and strain. Randomness of orientations is given by Euler angles equally distributed between \(0\) and \(2\pi\). Each element in the micromechanical model has been assumed to have the same properties of the real piezoelectric grain. In the first part of the simulations, quasi-static characteristics of piezoelectric materials are investigated by applying cyclic, rate independent, bipolar, uni-axial and external electrical loading with an amplitude of 2 kV/mm gradually starting from zero value in virgin state. Moreover, the simulations are undertaken for these materials which are subjected to quasi-static, uni-polar, uni-axial mechanical stress, namely compressive stress. The calculations are performed at each element based on linear constitutive equations, nonlinear domain switching and a probability theory for domain switching. In order to fit the simulations to the experimental data, some parameters such as spontaneous polarization, spontaneous strain, piezoelectric and dielectric constants are chosen from literature. The domain switching of each grain is determined by an electromechanical energy criterion. Depending on the actual energy related to a critical energy a certain probability is introduced for domain switching of the polarization direction. Same energy levels are assumed in the electromechanical energy relation for different types of domain switching like 90º and 180º for perovskite type tetragonal or 70.5º and 109.5º for rhombohedral microstructures. It is assumed that intergranular effects between grains can be modelled by such probability functions phenomenologically. The macroscopic response of the material to the applied electromechanical loading is calculated by using Euler transformations and averaging the individual grains. Properties of piezoelectric materials under fixed mechanical stresses are also investigated by applying constant compressive stress in addition to cyclic electrical loading in the simulations. Compressive stress is applied and kept constant before cyclic bipolar electrical loading is implemented. In the following chapters, a three-dimensional micromechanical model is extended for the simulation of the rate dependent properties of certain perovskite type tetragonal piezoelectric materials. The frequency dependent micromechanical model is now not only based on linear constitutive and nonlinear domain switching but also linear kinetics theories. The material is loaded both electrically and mechanically in separate manner with an alternating electrical voltage and mechanical stress values of various moderate frequencies, which are in the order of 0.01 Hz to 1 Hz. Electromechanical energy equation in combination with a probability function is again used to determine the onset of the domain switching inside the grains. The propagation of the domain wall during the domain switching process in grains is modelled by means of linear kinetics relations after a new domain nucleates. Electric displacement versus electric field hysteresis loops, mechanical strain versus mechanical stress and electric displacement versus mechanical stress for different frequencies and amplitudes of the alternating electric fields and compressive stresses are simulated and presented. A simple micromechanical model without using probabilistic approach is compared with the one that takes it into account. Both models give important insights into the rate dependency of piezoelectric materials, which was observed in some experiments reported in the literature. Intergranular effects are other significant factors for nonlinearities of polycrystalline ferroelectric materials. Even piezoelectric actuators and sensors show nonlinearities when they are operated with electrical loading, which is much lower than the coercive electric field level. Intergranular effects are the main cause of such small hysteresis loops. In the corresponding chapter, two basic field effects which are electrical and mechanical are taken into account for the consideration of intergranular effects micromechanically in the simulations of the two dimensional model. Therefore, a new electromechanical energy equation for the threshold of domain switching is introduced to explain nonlinearities stemming from both domain switching and intergranular effects. The material parameters like coercive electric field and critical spontaneous polarization or strain quantities are not implemented in the electromechanical energy relation. But, this relation contains new parameters which consider both mechanical and electrical field characteristics of neighbouring elements. By using this new model, mechanical strain versus electric field butterfly curves under small electrical loading conditions are also simulated. Hence, a rate dependent concept is applied in butterfly curves by means of linear kinetics model. As a result, the simulations have better matching with corresponding experiments in literature. In the next step, the model can be extended in three dimensional case and the parameters of electromechanical energy relation can be improved in order to get better simulations of nonlinear properties of polycrystalline piezoelectric materials.