Refine
Year of publication
Document Type
- Preprint (1037)
- Doctoral Thesis (928)
- Article (497)
- Report (399)
- Master's Thesis (30)
- Conference Proceeding (28)
- Diploma Thesis (24)
- Periodical Part (21)
- Working Paper (15)
- Lecture (11)
Language
- English (3026) (remove)
Keywords
- AG-RESY (47)
- PARO (25)
- Visualisierung (16)
- SKALP (15)
- Wavelet (13)
- finite element method (12)
- Case-Based Reasoning (11)
- Inverses Problem (11)
- Optimization (11)
- RODEO (11)
Faculty / Organisational entity
- Kaiserslautern - Fachbereich Mathematik (1045)
- Kaiserslautern - Fachbereich Informatik (739)
- Kaiserslautern - Fachbereich Physik (288)
- Kaiserslautern - Fachbereich Maschinenbau und Verfahrenstechnik (251)
- Fraunhofer (ITWM) (205)
- Kaiserslautern - Fachbereich Elektrotechnik und Informationstechnik (113)
- Kaiserslautern - Fachbereich Chemie (97)
- Kaiserslautern - Fachbereich Biologie (90)
- Kaiserslautern - Fachbereich Sozialwissenschaften (71)
- Kaiserslautern - Fachbereich Wirtschaftswissenschaften (32)
We present an optimal control approach for the isothermal film casting process with free surfaces described by averaged Navier-Stokes equations. We control the thickness of the film at the take-up point using the shape of the nozzle. The control goal consists in finding an even thickness profile. To achieve this goal, we minimize an appropriate cost functional. The resulting minimization problem is solved numerically by a steepest descent method. The gradient of the cost functional is approximated using the adjoint variables of the problem with fixed film width. Numerical simulations show the applicability of the proposed method.
The desire to model in ever increasing detail geometrical and physical features has lead to a steady increase in the number of points used in field solvers. While many solvers have been ported to parallel machines, grid generators have left behind. Sequential generation of meshes of large size is extremely problematic both in terms of time and memory requirements. Therefore, the need for developing parallel mesh generation technique is well justified. In this work a novel algorithm is presented for automatic parallel generation of tetrahedral computational meshes based on geometrical domain decomposition. It has a potential to remove this bottleneck. Different domain decomposition approaches and criteria have been investigated. Questions regarding time and memory consumption, efficiency of computations and quality of generated surface and volume meshes have been considered. As a result of the work parTgen (partitioner and parallel tetrahedral mesh generator) software package based on the developed algorithm has been created. Several real-life examples of relatively complex structures involving large meshes (of order 10^7-10^8 elements) are given. It has been shown that high mesh quality is achieved. Memory and time consumption are reduced significantly, and parallel algorithm is efficient.
The main concern of this contribution is the computational modeling of biomechanically relevant phenomena. To minimize resource requirements, living biomaterials commonly adapt to changing demands. One way to do so is the optimization of mass. For the modeling of biomaterials with changing mass, we distinguish between two different approaches: the coupling of mass changes and deformations at the constitutive level and at the kinematic level. Mass change at the constitutive level is typically realized by weighting the free energy function with respect to the density field, as experimentally motivated by Carter and Hayes [1977] and computationally realized by Harrigan and Hamilton [1992]. Such an ansatz enables the simulation of changes in density while the overall volume remains unaffected. In this contribution we call this effect remodeling. Although in principle applicable for small and large strains, this approach is typically adopted for hard tissues, e.g. bone, which usually undergo small strain deformations. Remodeling in anisotropic materials is realized by choosing an appropriate anisotropic free energy function. <br> Within the kinematic coupling, a changing mass is characterized through a multiplicative decomposition of the deformation gradient into a growth part and an elastic part, as first introduced in the context of plasticity by Lee [1969]. In this formulation, which we will refer to as growth in the following, mass changes are attributed to changes in volume while the material density remains constant. This approach has classically been applied to model soft tissues undergoing large strains, e.g. the arterial wall. The first contribution including this ansatz is the work by Rodriguez, Hoger and McCulloch [1994]. To model anisotropic growth, an appropriate anisotropic growth deformation tensor has to be formulated. In this contribution we restrict ourselves to transversely isotropic growth, i.e., growth characterized by one preferred direction. On that account, we define a transversely isotropic growth deformation tensor determined by two variables, namely the stretch ratios parallel and perpendicular to the characteristic direction. <br> Another method of material optimization is the adaption of the inner structure f a material to its loading conditions. In anisotropic materials this can be realized by a suitable orientation of the material directions. For example, the trabeculae in the human femur head are oriented such that they can carry the daily loads with an optimum mass. Such a behavior can also be observed in soft tissues. For instance, the fibers of muscles and the collagen fibers in the arterial wall are oriented along the loading directions to carry a maximum of mechanical load. If the overall loading conditions change, for instance during a balloon angioplasty or a stent implantation, the material orientation readapts, which we call reorientation. The anisotropy type in biomaterials is often characterized by fiber reinforcement. A particular subclass of tissues, which includes muscles, tendons and ligaments, is featured by one family of fibers. More complex microstructures, such as arterial walls, show two fiber families, which do not necessarily have to be perpendicular. Within this contribution we confine ourselves to the first case, i.e., transversely isotropic materials indicated by one characteristic direction. The reorientation of the fiber direction in biomaterials is commonly smooth and continuous. For transverse isotropy it can be described by a rotation of the characteristic direction. Analogous to the theory of shells, we additionally exclude drilling rotations, see also Menzel [2006]. However, the driving force for these reorientation processes is still under discussion. Mathematical considerations promote strain driven reorientations. As discussed, for instance, in Vianello [1996], the free energy reaches a critical state for coaxial stresses and strains. For transverse isotropy, it can be shown that this can be achieved if the characteristic direction is aligned with a principal strain direction. From a biological point of view, depending on the kind of material (i.e. bone, muscle tissue, cartilage tissue, etc.), both strains and stresses can be suggested as stimuli for reorientation. Thus, whithin this contribution both approaches are investigated. <br> In contrast to previous works, in which remodeling, growth and reorientation are discussed separately, the present work provides a framework comprising all of the three mentioned effects at once. This admits a direct comparison how and on which level the individual phenomenon is introduced into the material model, and which influence it has on the material behavior. For a uniform description of the phenomenological quantities an internal variable approach is chosen. Moreover, we particularly focus on the algorithmic implementation of the three effects, each on its own, into a finite element framework. The nonlinear equations on the local and the global level are solved by means of the Newton-Raphson scheme. Accordingly, the local update of the internal variables and the global update of the deformation field are consistently linearized yielding the corresponding tangent moduli. For an efficient implementation into a finite element code, unitized update algorithms are given. The fundamental characteristics of the effects are illustrated by means of some representative numerical simulations. Due to the unified framework, combinations of the individual effects are straightforward.
This paper presents a wavelet analysis of temporal and spatial variations of the Earth's gravitational potential based on tensor product wavelets. The time--space wavelet concept is realized by combining Legendre wavelets for the time domain and spherical wavelets for the space domain. In consequence, a multiresolution analysis for both, temporal and spatial resolution, is formulated within a unified concept. The method is then numerically realized by using first synthetically generated data and, finally, several real data sets.
Ion energy spectra of a laser-produced Ta plasma have been investigated as a function of the flight distance from the focus. The laser (Nd:YAG, 20 ns, 210 mJ) is incident obliquely (45°) and focused to an intensity of about 10^11 W cm-2. The changes in the ion distributions have been analysed for the Ta+ to Ta6+ ions in an expansion range 64 - 220 cm. With increasing distance from the target, a weak but monotonic decrease is observed for the total number of ions, which is essentially due to the decrease in the number of the more highly charged species. For the Ta+ and Ta2+ ions the net changes approximately cancel. A more sophisticated picture of the recombination dynamics is obtained, however, if the changes within individual groups of ions expanding with different velocities are compared. Here, in the same spectrum, both increasing and decreasing ion numbers can be observed. This can be interpreted as direct evidence of recombination and its dependence on temperature, density and charge.
The particle flux produced by an obliquely incident Nd Q-switched pulse (20 ns) on a Ta target has been analysed with regard to its angular distribution resolved for both its neutral and ion components. The laser intensity has been varied in the range between about 10^10 - 10^11 W cm-2, which is appropriate for many low-irradiance applications. It is observed that, at all emission angles and for the whole range of laser intensities, the number of neutral species clearly dominates the composition of the particles. At 1.3 x 10^10 W cm-2 the total number of emitted particles is 4 x 10^14, scaling as E_L^¾ with the laser energy. While for relatively low laser energies the angular distribution shows the usual smooth cos-behaviour, an additional strong directive emission cone, superimposed upon the cos-distribution, develops if the laser energy is enhanced. Both the strength and the width strongly depend on the laser intensity. While at lower intensities a fit by a cos^n function with n ~ 10 seems appropriate, n increases to 26 at an intensity of 10^11 W cm-2 . It can be assumed that secondary energy transfer processes that are not yet fully understood are responsible for this anomalous emission.
The visualization of numerical fluid flow datasets is essential to the engineering processes that motivate their computational simulation. To address the need for visual representations that convey meaningful relations and enable a deep understanding of flow structures, the discipline of Flow Visualization has produced many methods and schemes that are tailored to a variety of visualization tasks. The ever increasing complexity of modern flow simulations, however, puts an enormous demand on these methods. The study of vortex breakdown, for example, which is a highly transient and inherently three-dimensional flow pattern with substantial impact wherever it appears, has driven current techniques to their limits. In this thesis, we propose several novel visualization methods that significantly advance the state of the art in the visualization of complex flow structures. First, we propose a novel scheme for the construction of stream surfaces from the trajectories of particles embedded in a flow. These surfaces are extremely useful since they naturally exploit coherence between neighboring trajectories and are highly illustrative in nature. We overcome the limitations of existing stream surface algorithms that yield poor results in complex flows, and show how the resulting surfaces can be used a building blocks for advanced flow visualization techniques. Moreover, we present a visualization method that is based on moving section planes that travel through a dataset and sample the flow. By considering the changes to the flow topology on the plane as it moves, we obtain a method of visualizing topological structures in three-dimensional flows that are not accessible by conventional topological methods. On the same algorithmic basis, we construct an algorithm for the tracking of critical points in such flows, thereby enabling the treatment of time-dependent datasets. Last, we address some problems with the recently introduced Lagrangian techniques. While conceptually elegant and generally applicable, they suffer from an enormous computational cost that we significantly use by developing an adaptive approximation algorithm. This allows the application of such methods on very large and complex numerical simulations. Throughout this thesis, we will be concerned with flow visualization aspect of general practical significance but we will particularly emphasize the remarkably challenging visualization of the vortex breakdown phenomenon.
Analog sensor electronics requires special care during design in order to increase the quality and precision of the signal, and the life time of the product. Nevertheless, it can experience static deviations due to the manufacturing tolerances, and dynamic deviations due to operating in non-ideal environment. Therefore, the advanced applications such as MEMS technology employs calibration loop to deal with the deviations, but unfortunately, it is considered only in the digital domain, which cannot cope with all the analog deviations such as saturation of the analog signal, etc. On the other hand, rapid-prototyping is essential to decrease the development time, and the cost of the products for small quantities. Recently, evolvable hardware has been developed with the motivation to cope with the mentioned sensor electronic problems. However the industrial specifications and requirements are not considered in the hardware learning loop. Indeed, it minimizes the error between the required output and the real output generated due to given test signal. The aim of this thesis is to synthesize the generic organic-computing sensor electronics and return hardware with predictable behavior for embedded system applications that gains the industrial acceptance; therefore, the hardware topology is constrained to the standard hardware topologies, the hardware standard specifications are included in the optimization, and hierarchical optimization are abstracted from the synthesis tools to evolve first the building blocks, then evolve the abstract level that employs these optimized blocks. On the other hand, measuring some of the industrial specifications needs expensive equipments and some others are time consuming which is not fortunate for embedded system applications. Therefore, the novel approach "mixtrinsic multi-objective optimization" is proposed that simulates/estimates the set of the specifications that is hard to be measured due to the cost or time requirements, while it measures intrinsically the set of the specifications that has high sensitivity to deviations. These approaches succeed to optimize the hardware to meet the industrial specifications with low cost measurement setup which is essential for embedded system applications.
The conversion efficiency of laser energy into kinetic ion energy in a laser-produced plasma has been investigated for two quite different targets: graphite and tantalum. The laser energy (intensity) varied from several mJ to 200 mJ (1O^9 to 7 x 10^10 W cm-2) which is appropriate to many applications of a laser produced ion source. The conversion efficiency as a function of the laser energy was directly determined by differential measurements of the charge, kinetic energy and angular emission distribution of the plasma ions in absolute units. Whilst for the Ta target a nearly constant efficiency of about 30% was observed, the graphite result shows an unexpectedly strong enhancement of the transfer efficiency of up to 80% in the laser intensity range around 1.5 x l0^10 W cm-2. It is assumed that the results are related to the difference in the surface roughness of the targets.
GPU Stereo Vision
(2007)
To analyze scenery obstacles in robotics applications depth information is very valuable. Stereo vision is a powerful way to extract dense range information out of two camera images. In order to unload the CPU the intensive computation can be moved to GPU, taking advantage of the parallel processing capabilities of todays consumer level graphics hardware. This work shows how an efficient implementation on the GPU can be realized utilizing the NVIDIA Cuda framework.