Recent Submissions
Visual analysis of sequential data
(2025) Munz-Körner, Tanja; Weiskopf, Daniel (Prof. Dr.)
Sequential and temporal data is omnipresent in various areas of our lives. It is characterized by a sequence of data points in a fixed order, possibly with a temporal component. With an increasing amount of data being generated and collected, and different types of data originating from various domains, appropriate methods are needed to examine, interpret, understand, and draw conclusions from complex processes. Depending on the use case, the amount of data, and the target group, different analysis methods have to be chosen or developed. While visualization alone can already provide interesting insights into the data, interactive visual analysis helps users extract additional information by letting them focus on specific parts of the data and exploring it from different perspectives. Techniques such as brushing and linking and multiple coordinated views (multiple visualizations for the same data that are linked) help realize such an examination.
In this thesis, several approaches for visually analyzing sequential data are presented. The focus lies particularly on two key application areas: eye tracking and the interpretability of machine learning (ML) methods. Additionally, the use of dimensionality reduction methods during preprocessing for visualization is an important concept of this work. In all these areas, sequential or temporal components play important roles. They can be the subject of exploration, used as input data to trigger complex processes, represent internal mechanisms within methods, or be the output of a process. Users may want to examine or compare them to understand the data better. In the area of eye tracking analysis, this thesis presents a visual analysis approach that addresses the influence of various filter settings (parameter choices) on the data being visualized and interpreted. Additionally, a method is presented that combines temporal data from different sources to enable a better comparison of this data. Preprocessing steps play a crucial role in both methods to allow meaningful visualizations of the data and subsequent examination of the data. Next, various ML approaches are considered. The interpretability of ML techniques is currently a very important and challenging topic. Especially ML models in the area of natural language processing (NLP) deal with sequential components as input data, and also, the internal operations follow sequential processing steps. This thesis demonstrates that, in the field of NLP, internal information from neural machine translation (NMT), visual question answering (VQA), and text classification tasks can be made available to users for an enhanced understanding of internal mechanisms and to improve prediction results. Toward the end of this thesis, dimensionality reduction techniques are applied as a preparation step for visualizing sequential data. First, dimensionality reduction is used in an interactive system to examine text classification in the context of ML. However, interpreting 2D visualizations of dimensionally reduced sequential data requires careful consideration due to the possibility of data loss, misleading projections, and potential misinterpretation of the visualization itself. Therefore, in this work, visualization approaches are presented that address this challenge to provide methods to prevent misinterpretation. Overall, all presented interactive visualization approaches of this thesis use sequential data as input, and the visual analysis techniques help users during data exploration, interpretation, for debugging purposes, or to improve prediction results generated with ML models.
A parametric design integrated sampling and general training approach for optimal control oriented surrogate models of light-related quantities
(2025) Weber, Simon Oskar; Subramaniam, Sarith; Leistner, Philip
This study presents a general method for determining optimal control oriented surrogate models of light-related quantities. It is termed the incidence operator method comprising of sampling, model training and model export.
In contrast to matrix-based methods, machine learning constitutes a fundamental component of this new method. This entails the ability to represent complex dependencies of light-related quantities on adjustable material and geometric properties, as well as the possibility to export models using a standardised format (e.g. FMI, ONNX). Furthermore, components were developed to streamline the sampling of training data from parametric designs. The associated higher resolution reveals spatial discontinuities, for which a novel modelling approach and integration methodology have been developed.
The incidence operator method was validated with the enhanced two-phase method using two scenes. Based on annual simulations, the simple scene demonstrates high agreement (nMAE<2.1%), while the complex fenestration scene exhibits spatial discontinuities resulting in an increased deviation (nMAE<11.9%).
Biocomposite natural fibre pultruded profiles and application possibilities in architecture : case-studies in lightweight structures
(Stuttgart: Institut für Tragkonstruktionen und Konstruktives Entwerfen, Universität Stuttgart, 2025) Spyridonos, Evgenia; Dahy, Hanaa (Assoc. Prof. Dr.-Ing. Arch)
The selection of materials in the construction industry plays a pivotal role in advancing sustainability goals. Material selection through natural resources is constrained, and therefore, attention has shifted towards the development of novel materials. Fibre-reinforced polymer composites can be reliable substitutes for conventional building materials, particularly because of their high strength-to-weight ratio, which can also reduce material waste. While synthetic fibres, such as glass and carbon, are commonly employed in construction, there is a growing trend toward more sustainable material resources, with increased usage of natural fibres. Natural fibres are derived from various sources, with plant fibres being the most popular due to their high strength, low density, and accessibility as cost-effective agricultural by-products.
This study discusses the creation and evaluation of the LeichtPRO-Profiles, pultruded biocomposites intended to be integrated into structural systems as load-bearing elements. Pultrusion, a technology for manufacturing linear fibre-reinforced composites, is a well-established, reliable method. This study explores the optimisation of pultrusion technology through a multidisciplinary co-development approach, examining alternative fibres like flax and hemp and presenting an optimised matrix formulation tailored to specific applications.
The study elaborates on the composition and performance of these natural fibre pultruded profiles, showcasing their mechanical capabilities through rigorous experimentation and testing. A primary objective is the application of the profiles in active-bending structures, emphasising the significance of understanding the material's bending behaviour. The most significant case study is a 10-metre-span active-bending structure, the first large-scale structural demonstrator implementing the new material. Nevertheless, additional mechanical and small-scale physical tests confirmed that their structural performance suits a range of other applications. This is demonstrated through several prototypes encompassing applications such as reciprocal, tensegrity, and deployable structures.
The extensive case studies presented in this work showcase the applicability of this product to a wide range of applications spanning various scales and thematic contexts. The properties of the developed pultruded profiles demonstrate their suitability for multiple applications, paving the way for their market availability and development of similar biocomposite products.
Experimental and numerical investigations of transient conjugate heat transfer processes
(2025) Hartmann, Christopher; Weigand, Bernhard (Prof. Dr.-Ing. habil.)
Effective cooling of components exposed to high thermal loads is a key challenge in aircraft engine development. Analyzing thermal loads during flight missions is critical, as they fluctuate with varying operating conditions. Accurate assessment requires considering coupled heat transfer processes and transient effects.
The calculation of slow, transient phenomena was optimized by enhancing a coupling environment between a finite element and a finite volume solver. A wide range of boundary conditions and geometries were experimentally investigated. An existing ITLR test rig was adapted, and four geometries were examined.
The rig enables independent, reproducible control of inlet velocity and temperature, allowing the study of various test cycles. Wall temperatures were measured with high resolution using infrared thermography, and wall heat fluxes were calculated. Numerical simulations complemented the experiments.
The data support validation of the coupling environment and showed good agreement with simulations. A variable, adaptive, experiment-specific coupling step size reduced computation time while preserving accuracy.
A method was developed to enhance prediction accuracy and account for local dissipation, targeting heat transfer coefficients and friction factors in complex flows. Experimental data were analyzed using heat transfer correlations. A close relationship between two local parameters was observed, which enabled the development of a simplified correlation. The resulting model included coefficients that could be linked to established laws for turbulent boundary layer flows. One parameter correlates with local pressure gradients, near wall streamlines and friction factor distributions, while the other yields a Reynolds analogy factor that was used to estimate wall shear stresses. The model agreed well with simulations and proved universally applicable.
Quantum B-algebras
(2013) Rump, Wolfgang
The concept of quantale was created in 1984 to develop a framework for non-commutative spaces and quantum mechanics with a view toward non-commutative logic. The logic of quantales and its algebraic semantics manifests itself in a class of partially ordered algebras with a pair of implicational operations recently introduced as quantum B-algebras. Implicational algebras like pseudo-effect algebras, generalized BL- or MV-algebras, partially ordered groups, pseudo-BCK algebras, residuated posets, cone algebras, etc., are quantum B-algebras, and every quantum B-algebra can be recovered from its spectrum which is a quantale. By a two-fold application of the functor “spectrum”, it is shown that quantum B-algebras have a completion which is again a quantale. Every quantale Q is a quantum B-algebra, and its spectrum is a bigger quantale which repairs the deficiency of the inverse residuals of Q. The connected components of a quantum B-algebra are shown to be a group, a fact that applies to normal quantum B-algebras arising in algebraic number theory, as well as to pseudo-BCI algebras and quantum BL-algebras. The logic of quantum B-algebras is shown to be complete.
Crystal structure of (3aS,4R,5S,6R,6aS)-4,5,6-trihydroxy-5,6-O-isopropylidene-3,3a,4,5,6,6a-hexahydro-1H-cyclopent[c]isoxazole, C9H15NO4
(2014) Gültekin, Zeynep; Frey, Wolfgang; Jäger, Volker
C9H15NO4, orthorhombic, P21212 (No. 18), a = 8.8078(6) Å, b = 21.209(1) Å, c = 5.3420(4) Å, V = 997.9 Å3, Z = 4, Rgt(F) = 0.045, wRref(F2) = 0.123, T = 293 K.
Crystal structure of (3S,4S)-4,5-O-cyclohexylidene-4,5-dihydroxy-3- methylamino-pentanoic acid hydrochloride, (C12H22NO4)Cl
(2014) Frey, Wolfgang; Henneböhle, Marco; Jäger, Volker
C12H22ClNO4, monoclinic, P1211 (no. 4), a = 6.2327(3) Å, b = 7.9360(3) Å, c = 14.1536(5) Å, β = 99.709(4)°, V = 690.1 Å3, Z = 2, Rgt(F) = 0.042, wRref(F2) = 0.115, T = 293 K.
Crystal structure of (3aS,4R,5S,6R,6aS)-1-benzyl-4,5,6-trihydroxy-5,6-O-isopropylidene-3,3a,4,5,6,6a-hexahydro-1H-cyclopent[c]isoxazole, C16H21NO4
(2014) Gültekin, Zeynep; Frey, Wolfgang; Jäger, Volker
C16H21NO4, monoclinic, C121 (No. 5), a = 39.820(2) Å, b = 5.3038(3)Å, c = 15.4373(9)Å, β = 110.222(4)°, V = 3059.4 Å3, Z = 8, Rgt(F) = 0.068, wRobs(F2) = 0.184, T = 293 K.
Crystal structure of tetracerium(III) trisulfide heptaoxodisilicate(IV), Ce4S3[Si2O7]
(2014) Hartenbach, Ingo; Schleid, Thomas
Ce4O7S3Si2, tetragonal, I41/amd (No. 141), a = 12.0543(8) Å, c = 14.2351(9) Å, V = 2068.4 Å3, Z = 8, Rgt(F) = 0.027, wRref(F2) = 0.060, T = 298 K.
Eine offene Steuerungsarchitektur für die Anwendung von Methoden des maschinellen Lernens in der Lasermaterialbearbeitung
(Stuttgart : Fraunhofer Verlag, 2025) Reiff, Colin; Verl, Alexander (Univ.-Prof. Dr.-Ing. Dr. h.c. mult.)
Laserprozesse sind ein vielseitiges Verfahren zur Materialbearbeitung. Allein durch die gezielte Einstellung der Prozessparameter lassen sich unterschiedlichste Fertigungsverfahren realisieren. Die Identifikation geeigneter Prozessparameter stellt jedoch aufgrund der Vielzahl von Parametern eine wesentliche Herausforderung bei der Beherrschung von Laserprozessen dar. Methoden des maschinellen Lernens bieten das Potenzial, diesem Umstand zu begegnen, können aber aufgrund technischer Restriktionen, wie der mangelnden Offenheit und Durchgängigkeit bestehender Steuerungsarchitekturen, nur bedingt eingesetzt werden.
Im Rahmen dieser Arbeit wird eine offene und durchgängige Steuerungsarchitektur entworfen, die es erlaubt, Methoden des maschinellen Lernens zur Ermittlung und Anwendung geeigneter Prozessparameter zur Laufzeit des Steuerungssystems einzusetzen. Die Steuerungsarchitektur wird anhand einer lernenden Prozesssteuerung für die Laserbeschriftung demonstriert und validiert. Mit Hilfe eines künstlichen neuronalen Netzes, das automatisiert trainiert, in die Steuerung integriert und zur Laufzeit ausgeführt wird, ist es möglich, farbige Bilder auf Edelstahl zu lasern.