In this article the correlation between the average molar mass and the melt flow rate (MFR) is achieved. Based on the example of Polyoxymethylene (POM) it is shown that a high average molar mass is associated with a low MFR (high viscosity). On the basis of this dependency, the mechanical properties of static and dynamic tensile strength, elastic modulus, hardness and notched impact strength are investigated. It was found that the characteristic values of static tensile strength, elastic modulus and hard-ness increase with increasing MFR (decreasing viscosity). On the other hand the dynamic long-term properties and notched impact strengths decrease with increasing MFR. / Im Beitrag wird der Zusammenhang zwischen der mittleren molaren Masse und des Schmelzfließindex (MFR) hergestellt. Dabei wird am Beispiel von Polyoxymethylen (POM) ersichtlich, dass eine hohe mittlere molare Masse mit einem geringen MFR (hochviskos) einhergeht. Basierend auf dieser Abhängigkeit werden die mechanischen Eigenschaften statische und dynamische Zugfestigkeit, E-Modul, Härte sowie Kerbschlagzähigkeit untersucht. Dabei konnte festgestellt werden, dass die Kenngrößen statische Zugfestigkeit, E-Modul und Härte mit steigendem MFR (abnehmende Viskosität) zunehmen. Die dynamischen Langzeiteigenschaften und Kerbschlagzähigkeiten sinken hingegen mit zunehmendem MFR.

Nb3Sn Rutherford cables are used in CERN’s superconducting 11 T dipole and MQXF quadrupole magnets, which are proposed for the instantaneous luminosity (rate of particle collisions) upgrade of the Large Hadron Collider (LHC) by a factor of five to a High Luminosity-Large Hadron Collider (HL-LHC). Nb3Sn-based conductors are the key technology for the envisioned Future Circular Collider (FCC) with an operating magnetic dipole field of 16 T. The baseline superconductor of the LHC dipole magnets is Nb–Ti, whereas an operation above 10 T is not possible due to the current carrying performance limitations of this superconductor at higher magnetic fields. Therefore, a superconducting material such as Nb3Sn has to be used with proven performance capabilities of 10 T and above. The conductor choice towards Nb3Sn-based cables affects the magnet manufacturing process, as it requires a heat treatment up to 650°C, an epoxy resin impregnation and introduces mechanical diffculties as the superconducting filaments are brittle and strain sensitive. A mechanical over loading of the filaments lead to irreversible conductor damage. The designs of 11 and 16 T magnets are supposed to push the conductor towards its mechanical and electrical performance limitations. The magnetic field induced forces on the current carrying conductor are balanced by mechanical pre-loading of the magnet. Thereby the highest controlled mechanical pre-load for the 11 T dipole magnet is set at ambient temperature. The mechanical stress limits of Nb3Sn-based cables have been investigated at cryogenic temperatures. The material strength and stiffness of the cable insulation system, formed by glass-fibre-reinforced resin, is increased at low temperatures. The ultimate stress values, determined at cryogenic temperature, are therefore not conservative. The ultimate stress limitation of the insulated conductor is assumed to be lower at ambient temperature. The cable limitations at ambient temperature need to be known for the ongoing magnet manufacturing process and also for future design approaches. Furthermore, the compressive stress–strain behaviour of a coil conductor block at ambient temperature is the key material characteristic, in order to recalculate the stress level in the coil during the assembly process. Existing approaches using an indirect strain measurement method provide uncertainties in the low-strain regime, which is the essential strain range for a material compound consisting of major fractions composed of heat-annealed copper and epoxy resin. Compressive stress–strain data of an impregnated conductor block are required, based on a direct strain measurement system, as available data have been collected on samples based on a different strand type and insulation system. The elaborated direct strain measurements can be correlated to strain gauge data, measured directly on a coil. The stress distribution in a Nb3Sn Rutherford cable need to be understood and validated to understand strain-induced degradation effects in the insulated conductor. This knowledge can also help to optimise the stress distribution envisioned magnet designs. The stress–strain state in the copper and Nb3Sn phase of a loaded conductor block has to be determined experimentally. This dissertation describes a test protocol and first elaborated results on the investigated stress limitations of a Nb3Sn Rutherford cable under homogeneous load applied in transversal direction. The compressive stress–strain behaviour of impregnated Nb3Sn Rutherford cable stacks was investigated experimentally. This includes a detailed report on the sample manufacturing process, measurements performed and validation of results through a comparison with the elaborated data of cable stacks extracted from a coil. The presented results from neutron diffraction measurements of loaded cable stacks allow the determination of the stress–strain level of the copper and Nb3Sn phase in the impregnated conductor. The relevant measured results have been recalculated with numerical calculations based on the Finite Element Method (FEM).:1. Introduction 1 1.1. The LHC and the HL-LHC project 1.2. The FCC study 1.3. Superconducting materials for accelerator magnets 1.4. Multi-filamentary wires and Rutherford cables 1.5. Coil manufacturing process 1.6. Magnet coil assembly 1.7. Objectives of this thesis 2. Theory: fundamental principles 17 2.1. Analytical calculation: sector coil dipole 2.2. Mechanical behaviour of composite materials 2.3. Failure criteria and strength hypotheses for materials 2.4. Compressive tests 2.5. Fundamental principles of Neutron scattering 2.5.1. Test apparatus and measurement method 2.5.2. Lattice plane and Miller indices 2.5.3. Bragg diffraction and interference 2.5.4. Diffraction-based strain calculation 2.5.5. Diffraction-based stress calculation 2.6. Fundamental principles of FEM 3. Homogeneous transversal compression of Nb3Sn Rutherford cables 3.1. Superconducting cable test stations 3.2. The FRESCA test facility and specific sample holder 3.3. The sample description 3.4. Experimental procedure 3.5. Review of existing contact pressure measurement system 3.6. Compressive test station 3.7. Validation of the pressure-sensitive films 3.8. Press punch 3.9. Improvement of the contact stress distribution 3.9.1. First test: cable pressed between the bare tools 3.9.2. Second test: tool shimmed with a soft Sn96Ag4 3.9.3. Third test: tool shimmed with a soft Sn60Pb40 3.9.4. Fourth test: tool shimmed with a soft indium 3.9.5. Fifth test: tool shimmed with a polyimide film 3.10. Test results 3.11. Conclusion 4. Material characterisation by a compression test 4.1. Test set-ups for compressive tests and validation 4.2. Sample preparation 4.3. Compressive stress–strain measurement 4.4. Ten-stack sample stiffness estimation-based composite theories 4.5. Dye penetration test on loaded and unloaded samples 4.6. Conclusion 5. Neutron diffraction measurements 80 5.1. Test set-up for neutron diffraction measurement 5.2. The samples 5.3. Experiment: lattice stress–strain measurements 5.4. Conclusion 6. Simulation and modelling of Nb3Sn cables 6.1. The models 6.2. The 2D simulation results 6.3. The 3D simulation results 6.4. Conclusion 7. Comprehensive summary 7.1. Summary 7.2. Critical review 7.3. Next steps Appendix 113 A. Calculation of the magnetic field components in a sector coil without iron B. Approaches for the determination of diffraction elastic constants C. Manufacturing drawings D. FEM calculation results of the 2D model E. FEM calculation results of the 3D model F. Source Codes Bibliography

The synthesis of in-situ bulk metallic glass composites (BMGCs) with crystals that undergo a martensitic transformation under loading is possibly the most effective method to improve the plasticity of metallic glasses at room temperature. These martensitic or shape memory BMGCs are typically fabricated via solidification of glass-forming melts, which requires the meticulous selection of the chemical composition and the proper choice of the processing parameters (particularly the cooling rate) in order to ensure that the glassy matrix coexists with the desired amount of austenitic phase having suitable morphology and characteristics. Unfortunately, a relatively limited number of alloy systems, where austenite and glassy matrix coexist over a wide range of compositions, is available. Additionally, the necessity for rapid heat extraction and the corresponding high cooling rates essential for glass formation by melt solidification set an inherent limit to the achievable dimensions of BMGs and BMGCs specimens. The aim of this thesis is to study the effectiveness of powder metallurgy as an alternative to solidification for the synthesis of shape memory BMGCs. Ni50.6Ti49.4 and Zr48Cu36Al8Ag8 metallic glass powders were selected as the constituents of the composites because they have been extensively investigated and represent well the characteristic behavior of metallic glass and shape memory phases. BMGCs with different volume fractions of NiTi phase were fabricated using pressure-assisted sintering via hot pressing and their microstructure, mechanical properties and deformation mechanism were investigated. Particular focus was placed upon identifying the individual contributions of the martensitic transformation and shear band formation to plasticity as well as their mutual interaction at different length scales using a multidisciplinary approach involving experiments and simulations. BMG composites were synthesized by hot pressing of powder mixtures consisting of Zr48Cu36Al8Ag8 metallic glass and different amounts of Ni50.6Ti49.4 particles (10, 20, 40 and 60 vol.%) using the optimized consolidation parameters (temperature-time-pressure) determined for the monolithic BMG. All composites are characterized by a relatively uniform particle distribution and good interface bonding without any sign of reaction between the metallic glass and NiTi. The NiTi particles are progressively less isolated with increasing volume fraction of NiTi up to 40 % and, for the BMGC with 60 vol.% NiTi, the glassy particles are no longer connected and the NiTi phase becomes the continuous matrix. This is not a trivial achievement as the change of matrix while maintaining the structure of the constituent phases would not be easily obtained by solidification of melts with such different compositions. The size of the samples (10 mm diameter and 9 - 11 mm height) is larger than the characteristic BMGCs synthesized by casting and can, in principle, be scaled up to larger dimensions, demonstrating the effectiveness of this approach for overcoming the size limitation inherent to glass formation via solidification. In contrast to the monolithic BMG, which does not show any sign of plasticity, the BMGCs exhibit macroscopic plastic deformation that progressively increases with increasing NiTi content along with distinct strain-hardening. The BMG composites have similar fracture strength, which is comparable with the monolithic BMG, and exhibit a distinct double yield behavior, similar to shape memory BMGCs fabricated by casting. The deformed BMGCs exhibit a high density of shear bands, again in agreement with what observed for similar BMGCs fabricated by casting. These findings not only demonstrate that BMGCs with tunable microstructures and thus with optimized deformability can be synthesized by pressure-assisted sintering but, thanks to the phase stability of the components across such a wide range of compositions, also offer an excellent platform to examine fundamental aspects in the field of martensitic BMGCs. The confining stress exerted by the surrounding glassy matrix was quantified at the macroscale via a hybrid Voigt-Reuss mixture, which considers intermediate weighted combinations of stiff and compliant behaviors. In this way, the macroscopic stress required to initiate the martensitic transformation from B2 to B19´ can be described with rather good accuracy. The confining effect was further investigated by in-situ high-energy X-ray diffraction to have access to the strain tensor of the B2 phase as a function of loading. The results indicate that the confining stress along the direction perpendicular to the loading axis is particularly strong because the expansion of the B2 phase is constrained by the elastic matrix. A mechanism responsible for shear band formation in shape memory BMGCs is proposed. The stress field generated by the martensitic transformation in the contiguous glass would activate the adjacent shear transformation zone (STZ, the elementary units of plasticity in BMGs). The stress field induced by the activated STZ in the surrounding material then triggers the activation of the following STZs along the path of a potential shear band, in an autocatalytic process resembling the domino effect. The shear band formed in this way propagates through the glassy phase and, when impinging a B2 particle, the associated stress field would locally trigger the martensitic transformation, starting again the process. Molecular dynamics (MD) simulations of a martensitic BMGC show that the structural perturbation generated by the martensitic transformation is indeed transmitted to the adjacent glassy matrix and, in turn, to the developing shear band, in agreement with the proposed mechanism. The individual contribution of the glassy phase to the residual strain after each loading-unloading cycle was quantified assuming that the NiTi phase behaves in the same manner across the different specimens. The glass contribution was then correlated to the shear band density to obtain the plastic strain resulting from shear banding for a given amount of NiTi phase, a quantity that could be effectively used in the design of plastically-deformable BMGCs with shape memory particles. The martensitic transformation in the composites becomes progressively more irreversible with increasing strain. A large contribution to the martensite stabilization may come from the residual stress induced by the shear bands, in accordance with the finite element method (FEM) simulations, showing that residual stresses in the composites suppress the reverse transformation after unloading. These finding corroborates the hypothesis that the residual elastic stress field generated by the shear bands may be fundamental for stabilizing the martensitic phase by restraining the atoms at the glass-crystal interface from rearranging back to form austenite. This process can be reversed by proper heat treatment. The findings presented in this thesis offer the opportunity to synthesize shape memory BMG composites with enhanced plasticity and strain-hardening capability along with larger dimensions than those typically achieved by solidification. The powder metallurgy approach provides the necessary versatility in materials design and resulting properties of the composites via the control over the fundamental microstructural features, such as volume fraction, size, morphology and distribution of the second phase. Additionally, materials processing in the solid state gives a virtually infinite choice among the possible composite components, a degree of freedom not usually given when processing via solidification.:Abstract iii Kurzfassung vii Motivation and objectives xi 1 Theoretical background and state-of-the-art 1 1.1 Bulk metallic glasses (BMGs) 1 1.1.1 Formation of metallic glasses 2 1.1.2 Mechanical properties of BMGs 5 1.1.3 Shear bands in metallic glasses 8 1.2 Bulk metallic glass matrix composites 19 1.2.1 Fabrication of BMG composites 20 1.2.2 In-situ BMG composites 27 1.2.3 Ex-situ BMG composites 43 2 Experiments and simulations 57 2.1 Sample preparation 57 2.1.1 Starting materials 57 2.1.2 Powder mixing 59 2.1.3 Powder consolidation 60 2.2 Materials characterization 61 2.2.1 Composition analysis 61 2.2.2 Laboratory X-ray diffraction 61 2.2.3 High-energy X-ray diffraction and strain analysis 62 2.2.4 Viscosity measurements 63 2.2.5 Differential scanning calorimetry 64 2.2.6 Density measurements 64 2.2.7 X-ray computed tomography 65 2.2.8 Optical microscopy and scanning electron microscopy 65 2.2.9 Transmission electron microscopy 66 2.2.10 Elastic constants measurements 66 2.2.11 Mechanical tests 67 2.3 Molecular dynamic simulations 67 2.4 Finite element simulations 68 3 Pressure-assisted sintering of single-phase Zr48Cu36Al8Ag8 metallic glass and Ni50.6Ti49.4 powders 73 3.1 Synthesis and properties of single-phase Zr48Cu36Al8Ag8 bulk metallic glass 73 3.2 Synthesis and properties of single-phase Ni50.6Ti49.4 shape memory alloy 80 4 Pressure-assisted sintering of BMG composites with shape memory crystals: Microstructure and mechanical properties 87 4.1 Microstructure of BMG composites 87 4.2 Effect of NiTi volume fraction on mechanical properties 90 4.3 Effect of confinement of the glassy phase on the martensitic transformation 95 5 Deformation mechanism of shape memory BMG composites 101 5.1 Martensitic transformation and shear band formation 101 5.2 Mechanism of shear band formation in shape memory BMG composites 107 6 Reversibility of the martensitic transformation in shape memory BMG composites 113 6.1 Martensite stabilization in NiTi alloy and BMG composites 113 6.2 Simulation of the martensite stabilization effect in BMG composites 119 6.3 Effect of heat treatment on the martensitic reverse transformation 121 7 Summary and outlook 125 References 131 Acknowledgements 155 Publications 157 Erklärung 159

Background: To ensure adequate function after orthopedic tumor reconstruction, it is important to reattach the remaining soft tissue to the implant. This study aimed at obtaining mechanical properties of textile muscle-implant and muscle-bone connections in a preliminary test. Methods: Two groups of soft-tissue attachment were mechanically tested and compared: Native bone-muscle samples obtained from human femora and muscles attached to a prosthetic implant by means of Trevira® attachment tubes. Additionally, muscle samples were tested with muscle fibers aligned parallel and perpendicular to the tension load. A uniaxial load was exerted upon all samples. Results: Failure loads of 26.7 ± 8.8 N were observed for the native bone-muscle group and of 18.1 ± 9.9 N for the Trevira® group. Elongations of 94.8 ± 36.2 % were observed for the native bone-muscle group and 79.3 ± 51.8 % for the Trevira® group. The location of failure was mainly observed in the central area of the muscle fibers. Muscle fibers with parallel fiber orientation (47.6 ± 11.5 N) yielded higher tensile strength than those with perpendicular fiber orientation (14.8 ± 4.1 N). Conclusions: Our experiments showed that higher forces were transmitted in the origin and insertion areas than in areas of flat soft tissue reconstruction using attachment tubes. The data indicate that the tested material allows reattaching muscles, but without reinforcing the insertion site. Therefore, attachment tubes with region-dependent and potentially anisotropic material behavior might be advantageous to optimize muscle-bone load transmission after surgery, which may allow lower complication rates and shorter physical recovery.

In the last half of the XX century and the first years of the XXI century a large amount of methods for the determination of mechanical and tribological properties of materials on the micro- and nano-scale were developed. However, some problems and disadvantages are kept up to now. The combined application of normal and lateral forces allows to extend the possibilities of conventional contact mechanical approaches for investigations of mechanical and tribological behaviour of bulk and coated materials. Due to the unique construction of the Lateral Force Unit (LFU) the lateral force can be applied to the sample during normal indentation by the commercial nanoindenter UMIS 2000. Thus, the presented thesis reports the detailed study of the LFU capabilities for the determination of mechanical properties and tribological behavior on the micro-scale. At first it was found that the most appropriate standard position for the correct combined application of normal and lateral forces is the LFU inclination by 3.3° with respect to the UMIS stage. This standard position allows to minimize the influence of different factors on the measuring process. It was shown that the shape of normal displacement-time curves is the most convenient after the thermal drift correction for the simplification of the determination of such parameters as the maximal normal displacement and the residual normal deformation obtained by lateral force application. It was found that the crack formation can be detected as the observation of sudden change of lateral displacement in lateral force-displacement curve together with normal displacement in normal displacement-time curve. These investigations were performed for single-crystal sapphire. For the first time the crack in single-crystal sapphire was detected by the contact mechanical method in situ. The critical tensile stress for the crack formation in single-crystal sapphire was determined as 9.68+-0.22 GPa. It was established that the onset of plastic deformation can be detected by the observation of shape change of lateral force-displacement curve together with the appearance of residual normal deformation in normal displacement-time curve. These investigations were done for bulk BK7 glass and silicon dioxide film with thickness of 951 nm on silicon substrate. The yield strength for the silicon dioxide film was evaluated as 6.83+-0.02 GPa. It was found that the static friction of materials couples can be evaluated by the analysis of lateral force-displacement curves with the error of 5-10 %. The static friction coefficients for fused silica, BK7 glass, single-crystal sapphire as well as SiO2, DLC and CrN0.08 coatings were determined against diamond, tungsten carbide and sapphire spherical indenter with different radii. The effect of normal load on static friction for fused silica and BK7 glass against 10.5 µm diamond spherical indenter was also studied. It was found that the onset of plastic deformation leads to a significant change of static friction. / In der zweiten Hälfte des 20. Jahrhunderts und während den ersten Jahren des 21. Jahrhunderts wurden zahlreiche Methoden zur Untersuchung mechanischer und tribologischer Materialeigenschaften auf der Mikro- und Nanometerskala entwickelt. Trotz der Fortschritte auf diesem Gebiet blieben vielfältige Fragestellungen unbeantwortet oder waren mit den vorhandenen experimentellen Untersuchungsmethoden nicht zugänglich. Mit der kombinierten Belastung aus Lateral- und Normalkräften wurden die etablierten Messverfahren um einen viel versprechenden Ansatz zur Charakterisierung mechanischer sowie tribologischer Eigenschaften erweitert, der sowohl für Massiv- als auch Schichtmaterialien anwendbar ist. Die einzigartige Konstruktion einer Lateralkrafteinheit bietet als separates Bauteil die Möglichkeit während eines Standardeindringversuches mittels des kommerziellen Nanoindenters UMIS 2000 bei normaler Last, eine laterale Belastung zu überlagern. Die vorliegende Arbeit zeigt eine detaillierte Studie der Einsatzmöglichkeiten der Lateralkrafteinheit hinsichtlich der Charakterisierung mechanischer Eigenschaften und tribologischen Materialverhaltens auf der Mikrometerskala. Zunächst wurde herausgefunden, dass eine Verkippung der Lateralkrafteinheit von 3,3° gegenüber dem UMIS-Rahmen notwendig ist, um eine hochgenaue und definierte Belastung aus lateraler und normaler Kraft auf die Probe auszuüben. Mit dieser durchgeführten Korrektur der Ausrichtung gelang es weitere auf den Messprozess einwirkende Effekte zu minimieren. Nach der Korrektur der thermischen Drift scheinen die gemessenen Normalverschiebungs-Zeit-Kurven für die Bestimmung von mechanischen Parametern wie maximaler Verschiebung oder bleibender Eindrucktiefe bei lateraler Belastung geeignet zu sein. Als ein weiteres Ergebnis gelang es, durch die kombinierte Belastung der Kraftkomponenten Bruchversagen nachzuweisen. Das Materialversagen wurde durch eine abrupte Änderung der lateralen Verschiebung im Last-Verschiebungs-Diagramm angezeigt. Mit dieser Methode wurde erstmalig in-situ das Bruchversagen am Beispiel des einkristallinen Saphirs detektiert. Die kritische Zugspannung, die zur Bruchbildung bei Saphir führte, war 9,68+-0,22 GPa. Die Analyse der Kurvenform der Kraft-Verschiebungs-Kurven für die Lateralbelastung im Zusammenhang mit dem Auftreten von bleibender Deformation in den zugehörigen Verschiebungs-Zeit-Kurven der normalen Belastung liefert den Beginn der plastischen Deformation. Massive BK7-Glasproben sowie SiO2-Schichten wurden untersucht. Für die Fließspannung der SiO2-Schicht wurde ein Wert von 6,83+-0,02 GPa ermittelt. Der Haftreibungskoeffizient für verschiedene Materialpaarungen wurde aus den Last-Verschiebungs-Kurven mit einer Genauigkeit von 5-10 % berechnet. Zu den untersuchten Materialien gehörten Quarz, einkristallines Saphir, BK7-Glas sowie SiO2-, DLC- und CrN0.08-Schichten, die mit Diamant, Wolframkarbid und Saphir-Indentern gepaart wurden. Zusätzlich wurde der Einfluss der Normallast auf den Haftreibungskoeffizienten für Quarz und BK7-Glas gegen Diamant studiert. Es zeigte sich, dass der Beginn der plastischen Deformation zu signifikanten Änderungen der Haftreibung führt.

Mechanische Eigenschaften, wie z.B. der Elastizitätsmodul oder die Fließspannung, sind wichtige Materialgrößen, um ein Material zu charakterisieren. Dies kann beispielsweise dazu dienen, ein Bauelement eines MEMS unter Berücksichtigung seiner Funktion zu optimieren. Daher ist es nötig, eine Messmethode zur Verfügung zu haben, die diese Größen auch in kleinen Dimensionen korrekt bestimmen kann, insbesondere auch in dünnen Schichten. Deshalb wurde ein eigenes Konzept basierend auf der Kombination von elastischer Modellierung und Nanoindentationsexperimenten in unserer Arbeitsgruppe entwickelt. Dieses Konzept beruht auf der Theorie der sphärischen Indentation in geschichtete Materialien (Image Load Method). In einem nächsten Schritt wurde dieser theoretische Ansatz erweitert, indem das Modell eines effektiven Indentors mittels des Erweiterten Hertzschen Ansatzes in das ursprüngliche Modell implentiert wurde. Zur gleichen Zeit wurden neue experimentelle Möglichkeiten entwickelt, die auf der Applikation einer definierten Lateralkraft in einem Indentationsexperiment beruhen. Bei der Auswertung dieser neuen experimentellen Methoden stellte sich heraus, dass die auf dem theoretischen Modell basierenden Fittingprozeduren einen subjektiven Faktor aufweisen, sodass je nach Nutzer der Auswertesoftware unterschiedliche Ergebnisse erhalten werden. Der Einfluss intrinsischer Spannungen auf Indentationsexperimente wurde ebenfalls bisher noch nicht systematisch untersucht. Daher ist es die Aufgabe dieser Arbeit, ebendiese offenen Fragen zu beantworten und die Methode der Nanoindentation weiter zu optimieren, um dieser Messmethode neue Anwendungsgebiete zu eröffnen. Die Untersuchungen zum Einfluss der intrinsischen Spannung auf die experimentell erhaltenen mechanischen Eigenschaften einer dünnen Schicht beinhalten ein Modellexperiment mit einer Formgedächtnislegierung (NiTinol), in welcher mittels einer eigens konstruierten Biegevorrichtung definierte biaxiale Spannungszustände eingestellt werden können. Dabei konnte gezeigt werden, dass die Berechnung des Von- Mises-Spannungsfeldes mit dem Wert der intrinsischen Spannung korrigiert werden kann, so dass das erhaltene Maximum der Von-Mises-Spannung dem tatsächlichen Wert der Fließspannung des Materials entspricht. In der vorliegenden Arbeit werden des Weiteren detaillierte Untersuchungen der Entlastungskurven von Referenzmaterialien (BK7-Glas) und geschichteten Materialien (CrN Schicht auf Si) durchgeführt, die auf Berkovich- Indentationsmessungen beruhen. Dabei wurde insbesondere die Auswerteroutine basierend auf dem Konzept des effektiven Indentors dahingehend weiterentwickelt, dass der bisherige subjektive Einfluss erheblich reduziert werden konnte. Diese generell anwendbare Auswerteroutine (d0-Fit) zeichnet sich vor allem durch ein hohes Maß an Nachvollziehbarkeit und Reproduzierbarkeit aus. Mit der gleichzeitigen Anwendung einer Normal- und einer Lateralkraft in einem Indentationsexperiment mit einem spitzen Indentor (Berkovich) ist es möglich, weitere Informationen über die mechanischen Eigenschaften der untersuchten Probe zu gewinnen. Dabei wurde eine kritische Lateralkraft gefunden, die der kritischen Normalkraft einer partiellen Be- und Entlastung mit sphärischen Indentoren analog ist. Hierbei konnte die Möglichkeiten sowie Grenzen demonstriert werden, die das Modell des effektiven Indentors mit dem erweiterten Hertzschen Ansatzes bei der Auswertung der erhalten Messkurven bereitstellt. Diese Untersuchungen mit den bereits erwähnten Referenzmaterialien haben den Charakter eines empirischen Ansatzes. / Mechanical parameters, such as Young’s modulus or yield strength, are important material properties to characterize a material. These parameters can be used to optimize a construction unit of a MEMS with respect to its function for example. Therefore a measurement technique is needed that allows the determination of such mechanical properties even at very small length scales and especially in thin films. To assess the mechanical properties at small length scales and/or layered structures an experimental approach based on nanoindentation measurements and corresponding elastic modeling was developed within our working group. This approach uses the elastic theory of spherical indentation in layered structures based on a potential theory (Image Load Method). In a next step this theoretical approach was extended with implementation of the concept of the effectively shape indenter employing an extended Hertzian approach. At the same time new experimental techniques were developed opening the possibility to apply well defined lateral loads to nanoindentation experiments accompanied with precise measurement of the lateral loads and displacements. The theoretical model bases on fitting procedures of the experimentally obtained curves. During the evaluation of this new experimental nanoindentation approach a subjective factor within the fitting procedures was found, so that depending on the user different results can be derived. Furthermore the influence of intrinsic stresses on the nanoindentation data was not investigated systematically so far. The task of this work is therefore the answering of these open questions and to optimize the method of nanoindentation to open new application possibilities for this new nanoindentation approach. The investigations of the influence of the intrinsic stress on experimentally obtained mechanical properties of a thin film bases on a model experiment with a shape memory alloy (NiTinol). With the help of special designed bending device biaxial stress state can be induced in this material. It was shown that the calculation of the von Mises stress field can be corrected with the value of the intrinsic stress so that the obtained maximum of the von Mises stress corresponds to the yield strength of the material. Moreover it was shown that the onset of phase transformation from austenite to martensite under indentation loads corresponds to the von Mises stress criterion. In the present work detailed analysis of the unloading curves obtained with Berkovich nanoindentation on reference materials (BK7 borosilicate glass) and layered materials (CrN on Si substrate) was performed. The evaluation procedure was refined with respect to the subjective factor. The found procedure (d0-fit) is applicable in a general way and is characterized by a high degree of traceability and reproducibility. Using mixed loading conditions with a normal and a lateral load application at the same time with a sharp indenter (Berkovich) further information on the mechanical characteristics of a material can be derived. A critical lateral force (CLF) was found which is analogous to the critical normal force in loading-partial unloading indentation with spherical indenters. During this investigation the possibilities as well as the limitations of the theoretical model based on the effectively shaped indenter together with the extended Hertzian approach for the analysis of experimentally obtain unloading curves was shown. It should be noted that these investigations with reference materials have empirical character.

Der Bericht fasst die Ergebnisse eines Forschungsprojektes zusammen, in dem Möglichkeiten zur Optimierung der maschinellen Unkrautregulierung untersucht wurden, um eine Reduzierung des Handarbeitsaufwands zu erreichen. Mit der erprobten Gerätekombination ist eine Senkung der Handarbeit gegenüber dem bisher in der Praxis üblichen Verfahren um 50 bis 60 Prozent möglich. In der optimalen Variante zur Unkrautregulierung wurde bei einem Ertrag von 98 dt/ha (2013) bzw. 108 dt/ha (2014) und einem kalkulierten Durchschnittserlös von 3,00 €/kg die Gewinnzone erreicht. Die Ergebnisse richten sich an Praktiker und Berater aus dem ökologischen und konventionellen Gartenbau.

Kohlenstoffnanoröhren (Carbon Nanotubes, CNTs) gelten seit einigen Jahren als vielversprechendes neuartiges Material für verschiedenste Anwendungen in der Technik unterschiedlicher Fachgebiete. Von besonderem Interesse, z.B. in Leichtbaustrukturen, sind die postulierten exzellenten mechanischen Eigenschaften der einzelnen CNTs hinsichtlich Steifigkeit und Festigkeit. Diese auf der Nanoskala identifizierten Eigenschaften sollen auch in makroskopischen Bauteilen zu besonders guten mechanischen Eigenschaften führen. Demonstriert werden kann dies zum Beispiel an einer neuartigen Faser, die aus einer Vielzahl individueller Kohlenstoffnanoröhren gesponnen wurde. An dieser Faser durchgeführte Tests zeigen jedoch, dass die Eigenschaften nicht in der gewünschten Höhe von der Nanoskale auf die Makroskale übertragen werden. Um diesen Effekt erklären und evtl. beheben zu können, sowie für das Design von Strukturen aus Nanoröhren ('Superstrukturen') und einige weitere Anwendungen, sind Simulationsmodelle nötig, die die grundlegenden mechanischen (elastischen) Eigenschaften beschreiben können und zudem mit einer sehr großen Anzahl beteiligter CNTs und damit Atome umgehen können. Betrachtet man dies zusätzlich unter dem Aspekt, dass, beispielsweise zu Designzwecken, jeweils Rechnungen zu mehreren Varianten notwendig sind, ist verständlich, dass für jeden Durchlauf nur eine begrenzte Menge an Rechenzeit aufgebracht werden soll. Daher wird in der vorliegenden Arbeit ein mechanisches Modell der Kohlenstoffnanoröhren entwickelt, das die geforderte Aufgabe um ein Vielfaches schneller als quantenmechanische Methoden oder auch klassische Molekulardynamik behandeln kann. Basis hierfür ist ein molekularmechanischer Ansatz, der ein Ersatzmodell der betrachteten Kohlenstoffnanoröhre aus Balkenelementen erzeugt. Die zur Definition des Balkenfachwerks nötigen Balkeneigenschaften werden hierbei aus einem zugrundeliegenden chemischen Kraftfeld abgeleitet, das die kovalenten Bindungen zwischen den Atomen der Nanoröhre beschreibt. Der Ansatz ist damit in die Klasse der 'molecular structural mechanics' (MSM) Ansätze einzuordnen. Ausgangspunkt der vorliegenden Arbeit ist zunächst ein etabliertes MSM-Modell, dessen Schwächen in der vorliegenden Arbeit analysiert werden. Dabei wird festgestellt, dass der bisher verwendete MSM-Ansatz nicht energetisch konsistent zum zugrundeliegenden chemischen Kraftfeld ist. Dieser Umstand wird zunächst durch die Entwicklung eines modifizierten MSM-Modells behoben. Anschließend wird gezeigt, dass dieses Modell energetisch konsistent zum eingesetzten Kraftfeld ist. Um weitere Fortschritte mit dem gewählten molekularmechanischen Ansatz zu erzielen, wird dann ein verallgemeinertes MSM-Modell auf Basis eines fortschrittlichen chemischen Kraftfeldes entwickelt, das weitere Nachteile des ursprünglichen Ansatzes behebt und universeller einsetzbar ist. Das Modell wird dann zur Bestimmung der elastischen Konstanten von Armchair und Zig-zag CNTs eingesetzt und die erhaltenen Ergebnisse diskutiert.:1. Grundlagen 2. Modellbildung und Simulation einwandiger Kohlenstoffnanoröhren 3. Ergebnisse und Diskussion zum Zweck der Modellentwicklung 4. Ergebnisse und Diskussion der elastischen Parameter einwandiger CNTs 5. Zusammenfassung und Ausblick / For several years now, Carbon Nanotubes (CNTs) are seen as a promising new material for manifold applications in new technologies from different fields. The predicted excellent mechanical properties such as high strength and stiffness are of particual interest e.g. in lightweight structures. The nanoscopic propertiers are prone to lead to good mechanical properties also in macrosopic parts. This can be demonstrated for instance on the basis of a novel type of carbon fiber which is spun out of a multitude of individual carbon nanotubes. However, tests of the fibre show that the outstanding properties on the nanoscale are not fully transfered to the macroscale. In order to explain this effect as well as for designing structures made out of nanotubes (so called super structures) and other applications, models for simulations are needed. These models should be capable of reproducing the basic (elastic) mechnical properties of the nanotubes as well as to be capcable of dealing with a large number of participating nanotubes and hence atoms. Considering the additional aspect that multiple calculations of similar systems, e.g. for design purposes, are required, it is easy to understand, that for each calculation only a limited amount of computational effort is affordable. Hence, in the present work a mechanical model for the carbon nanotubes is developed which can fulfil the requested task in a much shorter time than quantummechanical or moleculardynamic calculations. The model is based on a molecular mechanics approach which creates a substitute model for the carbon nanotube based on beam elements. The parameters mandatory to define the beam elements in the beam framework are obtained on the basis of a chemical force field forming the foundation of the approach. The chemical force fields describes the properties of the covalent bonds in the carbon nanotube. As a result, the proposed model can be classified to be part of the molecular structural mechanics (MSM) approaches. Starting point of the present work is a well known MSM-model which is at first analyzed in order to identify its drawbacks. During this investigation it is found, that the model used so far is not consistent in terms of energy to its underlying chemical force field. This problem is fixed by the development of a modified MSM-approach. It is shown that this modified approach is now consistent to the underlying chemical force field in terms of energy. In order to further improve the method, a generalized, advanced MSM-framework is developed on the basis of a sophisticated chemical force field. This advanced framework resolves further drawbacks of the models and enables a more general application of the model. The obtained model is then used to calculate and discuss the elastic constants of Armchair and Zig-zag Carbon Nanotubes.:1. Grundlagen 2. Modellbildung und Simulation einwandiger Kohlenstoffnanoröhren 3. Ergebnisse und Diskussion zum Zweck der Modellentwicklung 4. Ergebnisse und Diskussion der elastischen Parameter einwandiger CNTs 5. Zusammenfassung und Ausblick

The large elastic limit, the strength close to the theoretical limit, the excellent magnetic properties and good corrosion resistance of bulk metallic glasses (BMGs) make them promising for several applications such as micro-geared motor parts, pressure sensors, Coriolis flow meters, power inductors and coating materials. The main limitation of these materials is their reduced macroscopic ductility at room temperature, resulting from an inhomogeneous deformation concentrated in narrows shear bands. The poor ductility can be overcome by the incorporation of a ductile second phase in the glassy matrix to form composites, which exhibit a better balance between strength and ductility. Different types of BMG composites have been developed to date but considerable plastic strain during tensile or bending tests has been only obtained for composites with in-situ formation of the second phase during solidification. Among these in-situ formed composites, significant tensile ductility has been only observed for two types of alloys so far: TiZrBe-based and CuZr-based BMG composites. The former precipitate dendrites of the cubic β-(Ti,Zr) phase in the glass matrix, whereas the latter combine spherical precipitates of the cubic B2-CuZr shape memory phase within the glass. The CuZr-based BMG composites have certain advantages over the TiZrBe-based composites such as the absence of Be, which is a toxic element, and exhibit a strong work-hardening behaviour linked to the presence of the shape memory phase. This concept of “shape memory” BMG composites has been only applied to CuZr-based alloys so far. It is worth investigating if such a concept can be also used to enhance the plasticity of other BMGs. Additionally, the correlation between microstructure, phase formation and mechanical properties of these composites is still not fully understood, especially the role of the precipitates regarding shear band multiplication as well as the stress distribution in the glassy matrix, which should be significantly influenced by the precipitates. The aim of the present work is to develop a new family of shape memory bulk metallic glass composites in order to extend the concept initially developed for CuZr-based alloys. Their thermal and mechanical properties shall be correlated with the microstructure and phase formation in order to gain a deeper understanding of the fundamental deformation mechanisms and thermal behaviour. A candidate to form new shape memory BMG composites is the pseudo-binary TiCu-TiNi system because bulk glassy samples with a critical casting thickness of around 1 mm have been obtained in the compositional region where the cubic shape memory phase, B2-TiNi, precipitates. This phase undergoes a martensitic transformation to the orthorhombic B19-TiNi during cooling at around 325 K. The B2- and B19-TiNi exhibit an extensive deformation at room temperature up to 30% during tensile loading. Compositions in the Ti-Cu, Ti-Cu-Ni, Ti-Cu-Ni-Zr, Ti-Cu-Ni-Zr-(Si) and Ti-Cu-Ni-Co systems were selected based on literature data and on a recently proposed λ+Δh1/2 criterion, which considers the effect of atomic size mismatch between the elements and their electronic interaction. Samples were then produced by melt spinning (ribbons) and Cu-mould suction casting (rods and plates). The investigation started in the Ti-Cu system. A low glass-forming ability (GFA) was observed with formation of amorphous phase only in micrometer-thick ribbons and the results showed that the best glass former is located around Ti50Cu50. Considering that the GFA of the binary alloys can be further improved with additions of Ni, new Ti-Cu-Ni shape memory BMG composites were then developed in which the orthorhombic Ti(Ni,Cu) martensite precipitates in the glassy matrix. These alloys exhibit a high yield strength combined with large fracture strain and the precipitates show a reversible martensitic transformation from B19 to B2-type structure at a critical temperature around 320 K (during heating). The amorphous matrix stabilizes the high-temperature phase (B2 phase), which causes different transformation temperatures depending on whether the precipitates are partially or completely embedded in the glassy matrix. The deformation starts in the softer, crystalline phase, which generates a heterogeneous stress distribution in the glassy matrix and causes the formation of multiple shear bands. The precipitates also have the important function to block the fast movement of shear bands and hence retard fracture. However, the size of such composites is limited to 1 mm diameter rods because of their low GFA, which can be further improved by adding CuZr. New Ti-Cu-Ni-Zr composites with diameter ranging from 2 to 3 mm were developed, which consist mainly of spherical precipitates of the cubic B2-(Ti,Zr)(Cu,Ni) and the glassy phase. The interrelation between composite strength and volume fraction of B2 phase was analysed in detail, which follows the rule of mixture for values lower than 30 vol.% or the load-bearing model for higher values. The fracture strain is also affected by the volume fraction of the respective phases with a maximum observed around 30 vol.% of B2 phase, which agrees with the prediction given by the three-body element model. It was observed that the cubic B2 phase undergoes a martensitic transformation during deformation, resulting in a strong work hardening and a high fracture stress of these alloys. The GFA of the Ti-Cu -based alloys can be further increased by minor additions of Si. A maximum GFA is observed for additions of 1 and 0.5 at.% Si to binary Ti-Cu or quaternary Ti-Cu-Ni-Zr alloys, respectively. This optimum GFA results from the formation of a lower amount of highly stable Ti5Si3 precipitates, which act as nuclei for other crystalline phases, and the increased stability of the liquid and the supercooled liquid. The addition of Co has the opposite effect. It drastically decreases the GFA of Ti-Cu-Ni alloys and both the martensitic transformation temperature and their mechanical behaviour seem to correlate with the number and concentration of valence electrons of the B2 phase. The transformation temperature decreases by increasing the concentration of valence electrons. An excellent combination of high yield strength and large fracture strain occurs for Ti-Cu-Ni-Zr and Ti-Cu-Ni-Zr-Si alloys with a relatively low amount of CuZr, with a fracture strain in compression almost two times larger than the one usually observed for CuZr-based composites. For instance, the Ti45Cu39Ni11Zr5 alloy exhibit a yield strength of 1490±50 MPa combined with 23.7±0.5% of plastic strain. However, a reduced ductility was found for the CuZr-richer Ti-Cu-Ni-Zr compositions, which results from the precipitation of the brittle Cu2TiZr phase in the glassy matrix. The present study extends the concept of “shape memory BMG matrix composites” originally developed for CuZr-based alloys and delivers important insights into the correlation between phase formation and mechanical properties of this new family of high-strength TiCu-based alloys, which upon further optimization might be promising candidates for high-performance applications such as flow meters, sensors and micro- and mm-sized gears. / Auf Grund der hohen Elastizitätsgrenze, Festigkeiten, die nahe an der theoretischen Grenze liegen, sehr guten magnetischen Eigenschaften, sowie einer guten Korrosionsbeständigkeit erscheint der Einsatz massiver metallischer Gläser (BMG) vielversprechend in zahlreichen Gebieten, wie z.B. in Mikro-Getriebemotorteilen, Coriolis-Massendurchflussmessern, Drucksensoren, Speicherdrosseln und als Beschichtungsmaterialien. Der Einsatz dieser Materialien wird jedoch hauptsächlich durch ihre begrenzte makroskopische Duktilität bei Raumtemperatur eingeschränkt. Diese resultiert aus einer inhomogenen Verformung, die in schmalen Scherbändern konzentriert ist. Die unzureichende Duktilität kann durch das Einbringen einer zweiten, duktilen Phase in die Glas-Matrix verbessert werden, so dass Komposite gebildet werden. Diese Komposite weisen in der Regel immer noch hohe Festigkeiten auf, lassen sich aber gleichzeitig deutlich besser plastisch verformen. Es wurden bereits verschiedene Arten von massiven metallischen Glas-Matrix-Kompositen entwickelt. Jedoch konnte die plastische Verformbarkeit in Zug- oder Biegeversuchen nur in den Materialien erhöht werden, in denen sich die zweite Phase bei der Erstarrung ausscheidet. Unter diesen in-situ Kompositen konnte eine signifikante Duktilität lediglich für zwei Legierungstypen beobachtet werden: massive metallische Gläser auf TiZrBe- und auf CuZr-Basis. Die Ausscheidungen der kubischen β-(Ti,Zr) Phase wachsen dendritenartig in die Glas-Matrix, wohingegen sich in letzterem Legierungstypen sphärische Ausscheidungen der Formgedächtnislegierung, B2-CuZr, im Glas bilden. CuZr-Basislegierungen haben dabei den großen Vorteil, dass sie kein Be enthalten, welches toxisch ist. Außerdem weisen diese Komposite auch dank der Formgedächtnisphase eine starke Kaltverfestigung auf. Das Konzept, massive metallische Formgedächtnis-Glas-Matrix-Komposite herzustellen, um die mechanischen Eigenschaften zu optimieren, wurde bisher nur auf CuZr-Basislegierungen angewandt. Es soll mittels dieser Arbeit nun erforscht werden, ob dieses Konzept auf andere massive metallische Gläser übertragbar ist. Des Weiteren ist der Zusammenhang zwischen Gefüge, Phasenbildung und mechanischen Eigenschaften der Komposite noch nicht vollständig verstanden, insbesondere die Rolle der Ausscheidungen in Bezug auf die Scherbandbildung und die Spannungsverteilung in der Glas-Matrix. Das Ziel der vorliegenden Arbeit ist die Entwicklung einer neuen Klasse massiver, metallischer Formgedächtnis-Glas-Matrix Komposite um das Konzept, welches ursprünglich für CuZr-Basislegierungen entwickelt wurde, zu erweitern. Die thermischen und mechanischen Eigenschaften sollen mit dem Gefüge und der Phasenbildung in Beziehung gesetzt werden, um so die fundamentalen Verformungsmechanismen und ihre Ursachen besser zu verstehen. Der Ausgangspunkt bei der Herstellung neuer massiver metallischer Formgedächtnis-Glas-Matrix Komposite ist das pseudobinäre TiCu-TiNi-System. In diesem System konnten massive Glasproben mit einem kritischen Gießdurchmesser von circa 1 mm hergestellt werden und zwar in dem Zusammensezungsbereich, in dem die kubische Formgedächtnisphase, B2-TiNi, gebildet wird. Während der Abkühlung findet in diesen Kompositen bei etwa 325 K eine martensitische Umwandlung der B2-Phase zur orthorhombischen B19-TiNi Phase statt. B2- und B19-TiNi weisen eine gute Verformbarkeit von bis zu 30% bei Raumtemperatur unter Zugbelastung auf. Die hier erzeugten Ti-Cu, Ti-Cu-Ni, Ti-Cu-Ni-Zr, Ti-Cu-Ni-Zr-(Si) und Ti-Cu-Ni-Co-Legierungen basieren auf Literaturangaben und Vorhersagen bezüglich der Glasbildungsfähigkeit in diesen Systemen mittels λ+Δh1/2-Kriterium, welches die Auswirkungen der Atomgrößenunterschiede der Elemente und deren elektronische Wechselwirkung einbezieht. Die Proben wurden im Schmelzspinnverfahren (Bänder) und mittels Saugguss in einer Cu-Kokille (Stäbe und Bleche) hergestellt. Die Weiter- und Neuentwicklung von Legierungen, beginnt mit dem Ti-Cu-System. Die Glasbildungsfähigkeit in diesem binären System ist nur gering, so dass lediglich mikrometerdicke amorphe Bänder hergestellt werden können. Die Ergebnisse zeigen, dass der beste Glasbildner eine Zusammensetzung von etwa Ti50Cu50 hat. Die Glasbildungsfähigkeit von binären Legierungen kann durch die Zugabe von Ni weiter verbessert werden. Dies führte innerhalb dieser Arbeit zur Entwicklung neuer Ti-Cu-Ni Formgedächtnis-Glas-Matrix Komposite, in welchen die orthorhombische Martensitphase in der Glas-Matrix ausgeschieden wird. Diese ternären Legierungen zeigen eine hohe Zugfestigkeit in Kombination mit einer hohen Bruchdehnung. Beim Überschreiten einer Temperatur von etwa 320 K vollziehen die Ausscheidungen eine reversible martensitische Umwandlung vom B19- zum B2-Strukturtyp. Durch die amorphe Matrix wird die Hochtemperaturphase (B2 Phase) stabilisiert. Dies verursacht unterschiedliche Umwandlungstemperaturen im Kompositmaterial, die davon abhängig sind, ob die Ausscheidungen nur teilweise oder vollständig in der Matrix eingebettet sind. Die Verformung beginnt in der weichen kristallinen Phase, welche eine heterogene Spannungsverteilung in der Glas-Matrix erzeugt und eine hohe Dichte an Scherbändern in der Matrix verursacht. Die Ausscheidungen haben zudem die Funktion, die Ausbreitung der Scherbänder zu blockieren und das Versagen des Materials zu verzögern. Die Größe der Komposite ist jedoch auf Grund der geringen Glasbildungsfähigkeit auf einen Stabdurchmesser von ca. 1 mm begrenzt. Dies kann mit dem Zulegieren von CuZr verbessert werden. Es wurden hier auf diese Weise neue Ti-Cu-Ni-Zr Komposite entwickelt, deren Durchmesser zwischen 2 und 3 mm liegt. Diese bestehen hauptsächlich aus sphärischen Ausscheidungen der kubischen B2-(Ti,Zr)(Cu,Ni)- und der Glasphase. Die wechselseitige Beziehung zwischen der Streckgrenze und dem Volumenanteil der B2-Phase wurde im Detail untersucht. Für kristalline Volumenanteile kleiner als 30 Vol.-% folgt die Streckgrenze der Mischungsregel und für größere Volumenanteile dem „lasttragenden Modell“ (load bearing model). Die Bruchdehnung wird ebenfalls vom Volumenanteil der Phasen beeinflusst und zeigt ein Maximum bei etwa 30 Vol.-% an B2-Phase. Dies stimmt mit der Vorhersage des „Drei-Element-Modells“ überein. Es wurde festgestellt dass die kubische B2-Phase während der Verformung eine martensitische Umwandlung durchführt, was die starke Kaltverfestigung und die hohen Bruchspannungen dieser Legierungen zur Folge hat. Die Glasbildungsfähigkeit von TiCu-Basislegierungen kann im Gegenzug weiterhin durch geringe Si-Zusätze gesteigert werden. Hierbei tritt jeweils ein Maximum bei Zusätzen von 1 und 0,5 at-% Si zu binären Ti-Cu- oder zu quarternären Ti-Cu-Ni-Zr-Legierung auf. Das Optimum der Glasbildungsfähigkeit ist das Ergebnis sowohl eines geringeren Anteils hochschmelzender Ti5Si3-Ausscheidungen, die als Keimbildner für andere kristalline Phasen dienen, als auch der erhöhten Stabilität der Schmelze sowie der unterkühlten Schmelze. Der Zusatz von Co wiederum hat einen gegenteiligen Effekt. Er vermindert die Glasbildungsfähigkeit von Ti-Cu-Ni-Legierungen drastisch. Zudem scheinen sowohl die martensitische Umwandlungstemperatur als auch das mechanische Verhalten mit der Zahl und Konzentration der Valenzelektronen der B2-Phase zu korrelieren. Die Umwandlungstemperatur sinkt mit steigender Valenzelektronenkonzentration. Eine ausgezeichnete Kombination von hoher Streckgrenze und Bruchdehnung tritt für die Legierungen Ti-Cu-Ni-Zr und Ti-Cu-Ni-Zr-Si mit einem relativ geringen CuZr-Anteil auf. Die Bruchdehnung unter Druck ist fast zweimal höher als es für CuZr-Basis-Komposite gewöhnlich beobachtet worden ist. Die Legierung Ti45Cu39Ni11Zr5 zeigt beispielsweise eine Streckgrenze von 1490±50 MPa in Kombination mit einer plastischen Dehnung von 23,7±0,5%. Für die CuZr-reicheren Ti-Cu-Ni-Zr Zusammensetzungen wurde jedoch eine geringere Duktilität festgestellt, was das Resultat spröder Cu2TiZr-Ausscheidungen in der Glas-Matrix ist. Die vorliegende Arbeit erweitert folglich das Konzept der „Formgedächtnis-Glas-Matrix Komposite“, welches bisher auf CuZr-basierte Legierungen beschränkt war und liefert wichtige Einblicke in die Beziehung zwischen Phasenbildung und mechanischen Eigenschaften der neuen Klasse hochfester TiCu-Basislegierungen, welche nach weiterer Optimierung vielversprechend sein könnten für Hochleistungsanwendungen wie Durchflussmesser, Sensoren und mikrometer- und mm-große Antriebe.

Selective laser melting (SLM) is a powder-based additive manufacturing technique consisting of the exact reproduction of a three dimensional computer model (generally a computer-aided design CAD file or a computer tomography CT scan) through an additive layer-by-layer strategy. Because of the high degree of freedom offered by the additive manufacturing, parts having almost any possible geometry can be produced by SLM. More specifically, with this process it is possible to build parts with extremely complex shapes and geometries that would otherwise be difficult or impossible to produce using conventional subtractive manufacturing processes. Another major advantage of SLM compared to conventional techniques is the fast cooling rate during the process. This permits the production of bulk materials with very fine microstructures and improved mechanical properties or even bulk metallic glasses. In addition, this technology gives the opportunity to produce ready-to-use parts with minimized need for post-processing (only surface polishing might be required). Recently, significant research activity has been focused on SLM processing of different metallic materials, including steels, Ti-, Ni- and Al-based alloys. However, most of the research is devoted to the parameters optimization or to feasibility studies on the production of complex structures with no detailed investigations of the structure-property correlation. Accordingly, this thesis focuses on the production and structure-property correlation of Al-12Si samples produced by SLM from gas atomized powders. The microstructure of the as-prepared SLM samples consists of supersaturated primary Al with an extremely fine cellular structure along with the residual free Si situated at the cellular boundaries. This microstructure leads to a remarkable mechanical behavior: the yield and tensile strengths of the SLM samples are respectively four and two times higher than their cast counterparts. However, the ductility is significantly reduced compared with the cast samples. The effect of annealing at different temperatures on the microstructure and resulting mechanical properties of the SLM parts has been systematically studied by analyzing the size, morphology and distribution of the phases. In addition, the mechanical properties of the SLM samples have been modeled using micro- structural features, such as the crystallite and matrix ligament sizes. The results demonstrate that the mechanical behavior of the Al-12Si SLM samples can be tuned within a wide range of strength and ductility through the use of the proper annealing treatment. The Al-Si alloys are generally used as pistons or cylinder liners in automotive applications. This requires good wear resistance and sufficient strength at the operating temperature, which ranges between 373 – 473 K. Accordingly, the tensile properties of the SLM samples were also tested at these temperatures. Changing the hatch style during SLM processing vary the texture in the material. Hence, samples with different hatch styles were produced and the effect of texture on their mechanical behavior was evaluated. The results show that the hatch style strongly influences both the mechanical properties and the texture of the samples; however no direct correlation was observed between texture and mechanical properties. The wear properties of the Al-12Si material was evaluated using pin-on-disc and fretting wear experiments. These experiments show that the as-prepared SLM samples exhibit better wear resistance than their cast counterparts and the SLM heat-treated samples. Finally, the corrosion investigations reveal that the SLM samples have similar corrosion behavior as the cast specimens under acidic conditions. A major drawback for the wide application of SLM as an industrial processing route is the limited size of the products. This is a direct consequence of the limited dimensions of the available building chambers, which allow for the production of samples with volumes of about 0.02 m3. A possible way to overcome this problem would be the use of the welding processes to join the small SLM objects to form parts with no dimensional limitations. In order to verify this possibility, friction welding was employed to join Al-12Si SLM parts. The results indicate that friction welding not only successfully permits the join materials manufactured by SLM, but also helps to significantly improve their ductility. This work clearly demonstrates that SLM can be successfully used for the production of Al-12Si parts with an overall superior performance of the mechanical and physical properties with respect to the conventional cast samples. Moreover, the mechanical properties of the SLM samples can be widely tuned in-situ by employing suitable hatch styles or ex-situ by the proper heat treatment. This might help the development of SLM for the production of innovative high-performance Al-based materials and structures with controlled properties for automotive and aerospace applications.