Scientific publications of the Saarland University
Not a member yet
14321 research outputs found
Sort by
Trauma and physical pain: an urban-rural comparison of sexually abused girls in Burundi
No full textABSTRACT
Background: Childhood sexual abuse (CSA) is a significant risk factor for physical and
psychological difficulties, especially in (post-)war- and conflict affected regions. Survivors
often suffer from both post-traumatic stress disorder (PTSD) and physical pain. Avoidance
may contribute to symptom chronification. Yet, in rural areas daily physical obligations may
render physical activities unavoidable for survivors. Living conditions post-CSA may play a
decisive role in symptom development and chronification.
Objective: The aim of this study was to compare PTSD symptom severity and physical pain
intensities post-CSA in rural and urban children/adolescents in Burundi. Additionally, the
relationship between PTSD symptom severity and physical pain intensity was analysed.
Method: The sample comprised 138 sexually abused female children/adolescents (M = 15.9,
SD = 1.7), of whom 65 lived in urban and 73 lived in rural areas. Local psychologists assessed
PTSD symptom severity and pain intensities using standardized questionnaires.
Results: A Fisher’s Exact Test revealed significantly lower pain intensities in rural compared to
urban children/adolescents. No significant group differences in PTSD symptom severity were
found using a T-Test. PTSD symptom severity correlated positively with pain intensity in the
total sample (τOverall = .16) and in the urban subgroup (τUrban = .16).
Conclusion: The results indicate a relationship between physical pain intensity and living
conditions, potentially through physical activity levels post-CSA. The findings also support a
link between PTSD and physical pain, highlighting the importance of avoidance for the
development and maintainance of both in young female CSA survivors.Antecedentes: El abuso sexual infantil (ASI) constituye un factor de riesgo significativo para el
desarrollo de dificultades físicas y psicológicas, especialmente en regiones afectadas por la
guerra y conflictos (post guerra). Las sobrevivientes suelen presentar trastorno de estrés
postraumático (TEPT) y dolor físico. La evitación podría contribuir a la cronificación de los
síntomas. Sin embargo, en zonas rurales las obligaciones físicas cotidianas pueden hacer
que las actividades físicas sean inevitables para las sobrevivientes. Las condiciones de vida
posteriores al ASI podrían desempeñar un rol decisivo en la aparición y cronificación de los
síntomas.
Objetivo: Comparar la gravedad de los síntomas de TEPT y la intensidad del dolor físico
posterior al ASI en niñas y adolescentes de zonas rurales y urbanas en Burundi. Además, se
analizó la relación entre la gravedad del TEPT y la intensidad del dolor físico.
Método: La muestra incluyó a 138 niñas/adolescentes víctimas de abuso sexual (M = 15.9, DE =
1.7), 65 residentes en áreas urbanas y 73 en áreas rurales. Psicólogos locales evaluaron la
gravedad del TEPT y las intensidades de dolor mediante cuestionarios estandarizados.
Resultados: La prueba exacta de Fisher mostró intensidades de dolor significativamente
menores en las participantes rurales en comparación con las urbanas. No se encontraron
diferencias significativas entre los grupos respecto de la gravedad del TEPT mediante
prueba t. En la muestra total se observó una correlación positiva entre la gravedad del TEPT
y la intensidad del dolor (τGeneral = .16), así como en el subgrupo urbano (τUrbano = .16).
Conclusión: Los resultados sugieren una relación entre la intensidad del dolor físico y las
condiciones de vida, posiblemente mediada por los niveles de actividad física posteriores al
ASI. Los hallazgos también respaldan la asociación entre TEPT y dolor físico, destacando la
importancia de su evitación y mantención en el desarrollo de ambos, tanto en niñas y
adolescentes víctimas de ASI
Anomaly detection in longitudinal clinical profile
No full textThis thesis investigates anomaly detection in longitudinal clinical data, with a particular focus on anti-doping applications where athlete monitoring requires identifying subtle, temporally embedded deviations in biological profiles. Unlike single-sample assessments, longitudinal data allows the analysis of intra-individual dynamics over time, supporting the detection of abnormal patterns that may otherwise remain hidden. A major challenge in anti-doping is the use of sophisticated manipulation strategies by some athletes to evade positive doping tests. An example is sample swapping, in which athletes’ biological samples may be deliberately substituted with those of another individual or with previously stored “clean” samples. Such practices undermine the reliability of conventional testing methods, which typically assume each sample to be authentic and unaltered. In contrast, longitudinal anomaly detection allows for the identification of inconsistencies within an athlete’s biological trajectory, thereby offering a means of uncovering irregularities suggestive of potential swapping events. Detecting such anomalies is difficult due to different challenges related to longitudinal data or the domain of anti-doping itself. To address these challenges, this thesis is categorized into three parts. The first part of the thesis introduces methods for longitudinal anomaly detection that address the key challenges of irregular sampling intervals, heterogeneous profile lengths, limited numbers of samples per athlete, and the scarcity of ground-truth labels. Two complementary architectures are proposed. The Self Attention-based Convolutional Neural Network (SACNN) addresses these issues by constructing structured subsequences from irregular profiles and applying attention-weighted convolutional layers to learn structural-temporal dependencies, thereby capturing subtle contextual anomalies such as sample swapping. In parallel, the Subsampling-based Convolutional Neural Network (SCNN) handles the challenges through a subsampling and aggregation strategy, where triplet-based segments are used to capture differential consistency, allowing reliable anomaly detection even in profiles with as few as two samples. Both models reduce reliance on explicit anomaly labels by learning individualized baselines. They are trained under high-specificity constraints, with evaluations performed on real-world anti-doping datasets including DNA-verified anomalies. The second part of the thesis incorporates metabolic pathway structures into model learning, ensuring that outputs are not only accurate but also biologically plausible. Two complementary approaches are proposed. Structural–Temporal Tokenization for Large Language Models (STT-LLM) introduces a novel tokenization strategy that encodes both the metabolic structure and temporal behaviour of clinical parameters from longitudinal profiles, enabling resource-efficient language models to process clinical data while preserving biological context. In parallel, Graph-based Modelling for Metabolism Pathways (GRAMP) embeds the steroid metabolic network into a graph attention architecture, allowing the model to detect pathway-consistent anomalies through information propagation across metabolically linked biomarkers. The third part of the thesis focuses on interpretable and domain-informed reasoning for decision support. Two complementary explanation tools are proposed. Metabolism Pathway-driven Prompting (MPP) uses structured graphs of the steroid metabolism pathway to guide language models in generating textual explanations for flagged anomalies, linking detected deviations to plausible biological mechanisms. Digital Athlete Passport (DAP) offers a visual analytics approach, projecting high-dimensional longitudinal clinical profiles into lower-dimensional spaces to visualize deviations and trajectory shifts, supported by PCA-based interpretation and centroid tracking. All models are integrated into CASPIAN, a software framework that allows domain experts to flexibly combine detection, structure-aware modelling, and interpretability methods. Together, these contributions provide a comprehensive approach to anomaly detection in longitudinal clinical profiles, allowing biologically grounded and explainable monitoring in high-stakes domains such as anti-doping and beyond.Diese Arbeit untersucht die Erkennung von Anomalien in longitudinalen klinischen Daten, mit besonderem Schwerpunkt auf Anti-Doping-Anwendungen, bei denen die Überwachung von Sportlern die Identifizierung subtiler, zeitlich eingebetteter Abweichungen in biologischen Profilen erfordert. Im Gegensatz zu Einzelprobenbewertungen ermöglichen Längsschnittdaten die Analyse der intraindividuellen Dynamik im Zeitverlauf und unterstützen so die Erkennung abnormaler Muster, die sonst möglicherweise verborgen bleiben würden. Eine große Herausforderung im Anti-Doping-Bereich ist der Einsatz ausgeklügelter Manipulationsstrategien durch einige Sportler, um positive Dopingtests zu umgehen. Ein Beispiel hierfür ist der Probenaustausch, bei dem die biologischen Proben von Athleten absichtlich durch die Proben einer anderen Person oder durch zuvor gelagerte „saubere” Proben ersetzt werden. Solche Praktiken untergraben die Zuverlässigkeit herkömmlicher Testmethoden, bei denen in der Regel davon ausgegangen wird, dass jede Probe authentisch und unverfälscht ist. Im Gegensatz dazu ermöglicht die longitudinale Anomalieerkennung die Identifizierung von Unstimmigkeiten innerhalb der biologischen Entwicklung eines Athleten und bietet somit eine Möglichkeit, Unregelmäßigkeiten aufzudecken, die auf einen möglichen Probenaustausch hindeuten. Die Erkennung solcher Anomalien ist aufgrund verschiedener Herausforderungen im Zusammenhang mit Längsschnittdaten oder dem Bereich der Dopingbekämpfung selbst schwierig. Um diesen Herausforderungen zu begegnen, ist diese Arbeit in drei Teile gegliedert. Der erste Teil der Arbeit stellt Methoden zur Längsschnitt-Anomalieerkennung vor, die sich mit den zentralen Herausforderungen unregelmäßiger Probenahmeintervalle, heterogener Profillängen, begrenzter Probenanzahlen pro Athlet und der Knappheit von Ground-Truth- Labels befassen. Es werden zwei sich ergänzende Architekturen vorgeschlagen. Das Self Attention-based Convolutional Neural Network (SACNN) geht diese Probleme an, indem es aus unregelmäßigen Profilen strukturierte Teilsequenzen konstruiert und aufmerksamkeitsgewichtete Faltungsschichten anwendet, um strukturelle und zeitliche Abhängigkeiten zu lernen und so subtile kontextuelle Anomalien wie Probenvertauschungen zu erfassen. Parallel dazu bewältigt das Subsampling-based Convolutional Neural Network (SCNN) die Herausforderungen durch eine Subsampling- und Aggregationsstrategie, bei der tripletbasierte Segmente verwendet werden, um unterschiedliche Konsistenzen zu erfassen, was eine zuverlässige Anomalieerkennung selbst bei Profilen mit nur zwei Proben ermöglicht. Beide Modelle reduzieren die Abhängigkeit von expliziten Anomalie-Labels, indem sie individualisierte Baselines lernen. Sie werden unter hochspezifischen Einschränkungen trainiert, wobei die Bewertungen anhand realer Anti-Doping-Datensätze einschließlich DNA-verifizierter Anomalien durchgeführt werden. Der zweite Teil der Arbeit bezieht Stoffwechselwegstrukturen in das Modelllernen ein und stellt so sicher, dass die Modellausgaben nicht nur genau, sondern auch biologisch plausibel sind. Es werden zwei sich ergänzende Ansätze vorgeschlagen. Structural-Temporal Tokenization for Large Language Models (STT-LLM) führt eine neuartige Tokenisierungsstrategie ein, die das metabolische strukturelle und zeitliche Verhalten klinischer Parameter aus Längsschnittprofilen codiert, sodass ressourceneffiziente Sprachmodelle klinische Daten unter Beibehaltung des biologischen Kontexts verarbeiten können. Parallel dazu bettet Graphbased Modelling for Metabolism Pathways (GRAMP) das Steroid-Stoffwechselnetzwerk in eine Graph-Attention-Architektur ein, wodurch das Modell durch Informationsverbreitung über metabolisch verknüpfte Biomarker wegkonsistente Anomalien erkennen kann. Der dritte Teil der Arbeit konzentriert sich auf interpretierbare und domäneninformierte Argumentation zur Entscheidungsunterstützung. Es werden zwei sich ergänzende Erklärungsinstrumente vorgeschlagen. Metabolism Pathway-driven Prompting (MPP) verwendet strukturierte Graphen des Steroidstoffwechselwegs, um Sprachmodelle bei der Generierung von textuellen Erklärungen für markierte Anomalien anzuleiten und erkannte Abweichungen mit plausiblen biologischen Mechanismen zu verknüpfen. Digital Athlete Passport (DAP) bietet einen visuellen Analyseansatz, bei dem hochdimensionale longitudinale klinische Profile in niedrigdimensionale Räume projiziert werden, um Abweichungen und Trajektorienverschiebungen zu visualisieren, unterstützt durch PCA-basierte Interpretation und Zentroid-Tracking. Alle Modelle sind in CASPIAN integriert, einem Software-Framework, das es Fachleuten ermöglicht, Erkennungs-, strukturbewusste Modellierungs- und Interpretierbarkeitsmethoden flexibel zu kombinieren. Zusammen bieten diese Beiträge einen umfassenden Ansatz zur Anomalieerkennung in longitudinalen klinischen Profilen und ermöglichen eine biologisch fundierte und erklärbare Überwachung in Bereichen mit hohem Risiko, wie z. B. Anti-Doping und darüber hinaus
How socioeconomic status affects a child's education – Investigating objective and subjective factors involved in shaping educational success in Germany
No full textDifferences in educational trajectories between social backgrounds can only be partially explained by differences
in cognitive abilities and are therefore considered educational inequalities. In this study, multiple constructs
involved in the prediction of educational success were investigated in a joint approach to specify their unique
contributions and to identify mechanisms associated with how socioeconomic status (SES) influences education.
Multiple regression analyses were conducted on N = 2273 children (aged 10 to 12). The effect of SES on
educational success was found to function via two mechanisms: First, the effect of school grades and home
environment on the assignment to secondary school was moderated by SES showing stronger influence at higher
SES levels. In contrast, being conscientious exerted a stronger influence for low SES children. Second, high SES
children were more likely to display characteristics that positively affected their academic performance (e.g.,
higher self-perceived ability, educational aspiration, cognitive abilities). Overall, the disadvantage of children
with low SES can be explained by the central findings that (1) school grades played a lesser role for low SES
children in their recommendation for further educational paths after primary school, and (2) high SES children
showed higher self-perceived abilities and higher educational aspirations unrelated to their cognitive abilities
which was associated with higher educational success. Why these mechanisms occur and where they originate
should be further investigated considering additional factors
Pro-apoptotic shift in human aniridia-derived limbal stromal cells under prolonged high-glucose stress, In vitro
No full textMetabolic stress can profoundly influence cell survival pathways in limbal stromal cells (LSCs). This study
investigated the effects of prolonged supraphysiological glucose exposure on apoptotic responses in both healthy
LSCs and LSCs derived from patients with congenital aniridia (AN-LSCs).
Primary human LSCs (n = 12) and AN-LSCs (n = 8) were cultured under high-glucose conditions (70 mM) for
either 48 or 72 h. Apoptotic cell populations were quantified using Annexin V/PI flow cytometry. Gene
expression of key apoptosis-related markers—CASP3/7/8/9/10, BCL2, BAX, BID, CDKN1A, CDKN1B, XIAP,
BIRC5, and TNFα—was assessed by qPCR, and corresponding protein levels were determined via flow cytometry
and ELISA.
Both cell types showed a significant increase in apoptosis after 72 h of high-glucose exposure compared to 48 h
(LSCs: p = 0.0021; AN-LSCs: p = 0.0017). At the mRNA level, CASP3 was upregulated in both cell types (p =
0.0210; p = 0.0396), while XIAP was downregulated (p = 0.0312; p = 0.0141). In LSCs, CASP10 and CDKN1A
(p21) mRNA levels increased (p = 0.0369; p = 0.0495), whereas BAX expression decreased (p = 0.0097). In AN LSCs, CASP7 level increased (p = 0.0032), BIRC5 (Survivin) expression decreased (p = 0.0113). Protein analysis
confirmed increased levels of Caspase-3, Caspase-7, and p21 in both groups after 72 h (p ≤ 0.0305), accompanied
by a significant decrease in XIAP (p = 0.0078; p = 0.0348). Additionally, Survivin protein was markedly reduced
in AN-LSCs after prolonged treatment (p = 0.0182).
Prolonged high-glucose exposure induces a shift in LSCs from an initial protective state to enhanced apoptotic
signaling. This transition, marked by increased pro-apoptotic and reduced anti-apoptotic markers, suggests that
sustained metabolic stress can override early survival mechanisms in both LSCs and AN-LSCs. These insights may
help guide the development of targeted therapies for aniridia-associated keratopathy
Curcumin and Its Derivatives in Hepatology: Therapeutic Potential and Advances in Nanoparticle Formulations
Get PDFCurcumin, a plant-derived polyphenol, shows promise in hepatology for treating
both malignant and non-malignant liver diseases and a subset of extrahepatic cancers.
Curcumin has hepatoprotective, anti-inflammatory, antifibrotic, and antiproliferative properties, as is evident in preclinical and clinical studies. This highlights its potential as an
adjunct to established cancer therapies, especially in the context of hepatocellular carcinoma
and secondary liver malignancies. Curcumin also demonstrates potential in metabolic
dysfunction-associated steatotic liver disease (MASLD), owing to its antifibrotic and lipidlowering effects. However, its clinical use is limited, relating to its poor bioavailability and
rapid metabolism. Nanotechnology, including liposomal and polymeric carriers, alongside
synthetic curcumin derivatives, offers strategies to enhance the bioavailability and pharmacokinetic properties. We propose to revisit the use of curcumin in nanoparticle preparations
in chronic liver disease and summarize current evidence in this review article
π-Lewis Base Activation of Carbonyls and Hexafluorobenzene
Get PDFWe report hitherto elusive side-on η2
-bonded palladium(0) carbonyl (anthraquinone, benzaldehyde) and arene
(benzene, hexafluorobenzene) palladium(0) complexes and present the catalytic hydrodefluorination of hexafluorobenzene by cyclohexene. The comparison with respective cyclohexene, pyridine and tetrahydrofuran complexes reveals that
the experimental ligand binding strengths follow the order THF<C6H6<C6F6<cyclohexene<pyridine<benzaldehyde<anthraquinone. To understand this surprising order, the complexes’ electronic structures were elucidated by
nuclear magnetic resonance (NMR), single crystal X-Ray diffraction (sc-XRD), ultraviolet/visible (UV/Vis) electronic
absorption, infrared (IR) vibrational, Pd L3-edge X-ray absorption (XAS), and X-ray photoelectron (XP) spectroscopic
techniques, complemented by Density Functional Theory (DFT) calculations including energy decomposition (EDANOCV) and effective oxidation state (EOS) analyses. For benzene, pyridine and cyclohexene, bonding follows the
donor/acceptor picture of the Dewar–Chatt–Duncanson model. In stark contrast, hexafluorobenzene, benzaldehyde and
anthraquinone bind via essentially the π-channel only and thus as π-analogues of Z-acceptor ligands. This contribution
elucidates the control of functional-group selectivity in palladium(0) catalysis and delineates a novel strategy to activate
electron-deficient π-systems
Modifying the antibacterial performance of Cu surfaces by topographic patterning in the micro- and nanometer scale
Get PDFAntimicrobial surfaces are a promising approach to reduce the spread of pathogenic microorganisms in various
critical environments. To achieve high antimicrobial functionality, it is essential to consider the material-specific
bactericidal mode of action in conjunction with bacterial surface interactions. This study investigates the effect of
altered contact conditions on the antimicrobial efficiency of Cu surfaces against Escherichia coli and Staphylococcus aureus. The fabrication of line-like periodic surface patterns in the scale range of single bacterial cells was
achieved utilizing ultrashort pulsed direct laser interference patterning. These patterns create both favorable and
unfavorable topographies for bacterial adhesion. The variation in bacteria/surface interaction is monitored in
terms of strain-specific bactericidal efficiency and the role of corrosive forces driving quantitative Cu ion release.
The investigation revealed that bacterial deactivation on Cu surfaces can be either enhanced or decreased by
intentional topography modifications, independent of Cu ion emission, with strain-specific deviations in effective
pattern scales observed. The results of this study indicate the potential of targeted topographic surface functionalization to optimize antimicrobial surface designs, enabling strain-specific decontamination strategies
Scalable Manufacturing Method for Model Protein-Loaded PLGA Nanoparticles: Biocompatibility, Trafficking and Release Properties
Get PDFBackground and Objectives: Drug delivery systems (DDSs) offer efficient treatment solutions to challenging diseases such as central nervous system (CNS) diseases
by bypassing biological barriers such as the blood–brain barrier (BBB). Among DDSs,
polymeric nanoparticles (NPs), particularly poly(lactic-co-glycolic acid) (PLGA) NPs, hold
an outstanding position due to their biocompatible and biodegradable qualities. Despite
their potential, the translation of PLGA NPs from laboratory-scale production to clinical
applications remains a significant challenge. This study aims to address these limitations
by developing scalable PLGA NPs and evaluating their potential biological applications.
Methods: We prepared blank and model-protein-loaded (albumin–FITC and wheat germ
agglutinin-488 (WGA-488)) fluorescent PLGA NPs using the traditional double-emulsion
method combined with the micro-spray-reactor system, a novel approach that enables
fine particle production enabling scale-up applications. We tested the biocompatibility
of the NPs in living RPMI 2650 and neuroblastoma cell lines, as well as their trafficking
and uptake. Release kinetics of the encapsulated proteins were investigated through confocal microscopy and in vitro release studies, providing insights into the stability and
functionality of the released proteins. Results: The formulation demonstrated sustained
and prolonged protein release profiles. Importantly, cellular uptake studies revealed that
the NPs were not internalized. Furthermore, encapsulated WGA-488 protein retained
its functional activity after release, validating the integrity of the encapsulation and release processes. Conclusions: The proof-of-concept study on NP manufacturing and an
innovative drug trafficking and release approach can bring new perspectives on scalable
preparations of PLGA NPs and their biological applications
Studies concerning the electronic structure and the electron transfer in electron poor intermetallics
No full textjThis doctoral thesis contains results in the field of new ternary intermetallic aluminum compounds with the general composition MxAlyTz with M = Ca-Ba, Y, La-Nd, Sm-Lu and T being a transition metal of group nine and ten (Co-Pt). The goal was to synthesize new compounds and investigate their chemical bonding situation, charge transfer and influence of spectroscopic measurements. The first part shows results of the binary compounds AEAl2 and AEAl4 with AE = Ca, Sr, Ba and the influence of the AE element on the chemical shift in 27Al MAS NMR, supported by quantum chemical calculations. The second part contains new ternary aluminum compounds with a divalent cationic element M, which include Ca, Sr and Eu. Besides structure and 27Al MAS NMR investigations, quantum calculations were performed giving insight into the chemical bonding situation as well as the charge transfer occurring in the compounds. In addition to the analytic methods mentioned above, electrical resistance measurements of the new compound SrAl8Rh2 and magnetic susceptibility data of the solid solution Eu2Al15Pt6–xTx with T = Pd, Ir and Au were recorded, showing the electronic behavior and the influence of the substitution on the magnetic behavior respectively. The third part contains compounds with formally the trivalent cationic species. The series from the ternary compounds MAl5Pt3 and M2Al16Pt9 could be extended by Y, La-Nd, Sm, Gd-Er or La-Nd, Sm and Gd, respectively. Magnetic measurements were conducted for the MAl5Pt3 series with M = Y, Ce-Nd, Gd-Ho showing Pauli-paramagnetism for YAl5Pt3 whereas the compounds with Nd, Gd, Tb, Dy and Ho exhibit an antiferromagnetic ordering at low temperatures. Additionally, quantum chemical calculations, XPS and 27Al MAS NMR measurements of YAl5Pt3 have been carried out. In the fourth chapter, preliminary results of new ternary phases, namely CaAl5Ni2, Eu5Al0.70(1)Pd2.30(1) and an extension of the RE4Al13Pt9 series with RE = La-Nd are summarized.Diese Dissertationsschrift beinhaltet Ergebnisse aus dem Gebiet neuer intermetallischer Aluminiumverbindungen mit der allgemeinen Zusammensetzung MxAlyTz, wobei M = Ca-Ba, Y, La-Nd, Sm-Lu und T ein Element der Gruppen neun oder zehn (Co-Pt) enthalten. Neben der Synthese neuer Verbindungen wird außerdem über Untersuchungen der chemischen Bindung und des Ladungsübertrags sowie der Einfluss auf spektroskopische Analysen berichtet. Im ersten Teil dieser Arbeit werden neue Ergebnisse im binären System AEAl2 und AEAl4 über den Einfluss der Erdalkalimetalle Ca, Sr und Ba auf die chemische Verschiebung des 27Al MAS NMR Signals gezeigt, die Untersuchungen wurden dabei mit Hilfe von quantenchemischen Rechnungen unterstützt. Der zweite Teil der Arbeit berichtet über neue ternäre Aluminiumverbindungen, indem das Metall M als divalentes Kation in der Struktur vorzufinden ist, diese inkludieren Ca, Sr and Eu. Dabei wurden neben der Strukturanalyse und 27Al MAS NMR Untersuchungen, quantenchemische Rechnungen durchgeführt, um die Bindungssituation und den Ladungsübertrag zu untersuchen. Zusätzlich wurden elektrische Widerstandsmessungen bei SrAl8Rh2 und Untersuchungen der magnetischen Eigenschaften an Eu2Al15Pt6–xTx mit T = Pd, Ir und Au aufgenommen, welche den Einfluss der Substitution auf das magnetische Verhalten aufweisen. Der dritte Teil umfasst die Verbindungen in dem das Metall M den trivalenten Zustand annimmt. Dies sind die Reihen basierend auf divalenten Verbindungen MAl5Pt3 und M2Al16Pt9, welche mit Y, La-Nd, Sm, Gd-Er beziehungsweise La-Nd, Sm und Gd erweitert werden konnten. Von den Verbindungen der MAl5Pt3 Serie mit M = Y, Ce-Nd, Gd-Ho wurden die magnetischen Eigenschaften vermessen, wobei YAl5Pt3 Pauli-paramagnetisch ist, die Phasen mit Nd, Gd, Tb, Dy und Ho hingegen einen antiferromagnetischen Übergang bei niedrigen Temperaturen aufweisen. Zusätzlich wurden quantenchemische Rechnungen, XPS und 27Al MAS NMR Messungen an YAl5Pt3 durchgeführt. Im vierten Kapitel der Thesis werden erste Ergebnisse neu gefundener Phasen vorgestellt, diese enthalten Erkenntnisse zu CaAl5Ni2, Eu5Al0.70(1)Pd2.30(1) und eine Erweiterung der RE4Al13Pt9 Serie mit den RE = La-Nd.Deutsche Forschungsgemeinschaft (DFG), Projektbezeichnung: Studien zur Bindungssituation und zum Elektronenübertrag in elektronenarmen intermetallischen Verbindunge
Prozess- und Betriebsmittelentwicklung für die Absicherung der Inbetriebnahme umfelderfassender Sensoren automatisierter Fahrzeuge in der Produktionslinie
Get PDFVor dem Hintergrund der Zunahme hochautomatisierter und autonomer Fahrfunktionen stei-gen die Anforderungen an Fahrzeughersteller, die Sicherheit und Funktionsfähigkeit der Fahr-zeuge nachzuweisen. Hierzu existieren bereits verschiedene Ansätze in der Produktentwick-lung und zur Unterstützung der Prototypenphase. Eine Lösung für die Funktionsabsicherung in der Fahrzeugproduktion existiert zum aktuellen Zeitpunkt noch nicht. Zur Adressierung dieser Herausforderung stellt die vorliegende Arbeit einen Prozess zur Funktionsabsicherung hochautomatisierter und autonomer Fahrzeuge in der Montagelinie vor. Ziel dieses Prozesses ist die präzise, robuste und nachvollziehbare Ermittlung der extrinsi-schen Kalibrierung der umfelderfassenden Sensoren. Um einer weiteren Parallelisierung des End-of-Lines entgegenzuwirken, zielt die entwickelte Lösung auf eine Integration in die Fließ-montagelinie ab. Für die Prozessentwicklung wird zunächst eine Analyse des Produktes und des Stands der Technik im Bereich der Fahrzeuginbetriebnahme und Funktionsabsicherung durchgeführt. Darauf aufbauend werden die Anforderungen an den Zielprozess definiert. Nach der Prozessentwicklung im Sinne eines Grob- und eines Lösungskonzeptes wird der Prozess in einem Technologiedemonstrator implementiert und validiert. Die Arbeit schließt mit einem Fazit und einem Ausblick auf den weiteren Forschungsbedarf.Due to the increase of highly automated and autonomous driving functions, the demands on vehicle manufacturers to ensure the safety and functionality of vehicles are increasing. Vari-ous approaches concerning the product development and the prototype phase exist, but there is no solution for the validation of the vehicles functions in the production line. To address this challenge, this thesis presents a process for the function validation of highly automated and autonomous vehicles on the assembly line. The aim of this development is the precise, robust and reproducible determination of the extrinsic calibration of the environmental sensors. In order to counteract further parallelization of the end-of-line, the developed concept focuses an integration into the flow assembly line. For the process development, an analysis of the product and the state of the art in the field of vehicle commissioning is first carried out. On this basis, requirements for the development of a target process are defined. After the process development in the terms of a rough and a solution concept, the process is implemented in a technology demonstrator and this implementation is validated. The thesis concludes with a summary and an outlook on the need for further research