Integrated Computational Materials, Process and Product Engineering – IC-MPPE 2022Hybrid conference

Machine learning (ML) has become a standard tool for accelerating innovation in the materials industry. A particularly well-suited ML application is materials design, wherein a user specifies a number of target properties and constraints (e.g., cost, manufacturability, etc.), and then uses ML to identify materials candidates most likely to meet those requirements. However, a number of key challenges must be overcome in order to realize the benefits of ML in materials design. In this talk, I will describe Citrine’s platform approach to these challenges, and some materials-specific ML method development results.

The digital transformation of materials science and engineering is driven by societal needs to develop novel materials which are compatible with a circular economy, save resource while they are produced and have extended lifetimes until they are re- or upcycled. Materials platforms help to share the tremendous effort within the community. The platforms are built to offer prototypical implementations of domain specific services (e.g. digital workflows and development environments, simulations, AI) and allow to publish structured reference datasets as best practices together with the underlying materials ontologies.
Starting with ‘Why should we go through all the effort?’, short examples (UrWerk, iBain, BW MaterialDigital, metallography) shall be shown of how data can be used either across institutions along the materials life cycle. Implications, of what could be done with a global materials data space shall be discussed very briefly. Next, we will discuss ‘How can a common data space could be implemented?’. Finally, the question ‘What needs to be done?’ will be discussed based on two example platforms, the BMBF ‘Plattform MaterialDigital’ (2019-2025) and the NFDI-MatWerk (2022-2026-maybe till 2031). Here, the mission and implementation methodology will be described as well as possible products to answer the needs in Materials Science and Engineering and the related stakeholders. The talk will be closed by a short glimpse of how the transformation can take place on an institutional scale.

The development of new and improved materials and the use of existing materials in new applications across different industries are a significant innovation driver and a key factor for the success and sustainability of industry and European society in general.
Today, many more large and small companies rely on numerical simulations to effectively and efficiently design and engineer new products and to optimise processes, thus minimising the need for expensive and time‐consuming prototyping and testing. Furthermore, the potential of materials modelling as a driver for radical increase in speed of product design and radical decrease in production costs and in‐service performance is recognised by manufacturing companies across Europe (SME’s and large corporations alike). In a tough, highly volatile and competitive market environment, innovation and time‐to‐market is critical, especially for companies that need to put differentiated products on the market every year. Materials modelling‐led product innovation can be a key differentiator for success in such competitive markets.
The EMMC association (www.emmc.eu) and its wide ranging community have been leading in building a highly interdisciplinary ecosystem; it merges the many strands of modelling and digitalisation advances to accelerate materials development and manufacturing. Key developments include advancing open and integrated modelling and innovation platforms, connected to powerful knowledge systems based on interoperability and ontologies to support the Industry 5.0 transition.
EMMC is contributing to a consolidated strategic roadmap, supporting the agile and inclusive European governance of advanced materials and the drawing up of a novel European strategic next-generation advanced materials agenda for a sustainable society. In particular, EMMC calls for a Materials and Process Acceleration Platform: an ecosystem uniting digital and materials capacities and competences in materials modelling, characterisation, robotics, data documentation, ontologies and ML/AI, which are orchestrated to accelerate the design, development and application of solution-oriented advanced materials and related processes/manufacturing.

To cope with the ever-growing needs for high-performance materials, a strong ongoing trend exists worldwide towards installation of material development and processing platforms termed materials acceleration platforms (MAPs). A common element of these initiatives is the transition from stand-alone simulation and characterization tools towards integrated environments combining databases, physical modeling, inverse design, machine learning and experimental testing in a common framework accessible to all contributors and stakeholders. Such platforms are expected to meet the future challenges related to materials design and to significantly accelerate development processes.
In this talk, we present the MAP “MCacceL” that is currently being implemented at Materials Center Leoben, where we showcase two use cases, i.e. the development of bainitic steels with improved strength and ductility and perovskites with increased energy storage capability. The graphical user interface of the platform will offer tools for automated and manual recording and structuring of experimental measurements, simulations, and literature data according to FAIR data principles. On this data basis, workflows for targeted materials optimization are developed using physics-based material models and machine learning approaches that can provide process-structure-property relationships. With interactive analysis and data exploration tools, the discovery and optimization of high-performance materials will be significantly shortened.

Every industry is driven by cost-efficiency. Although securing quality and safety marks the top priority in the aerospace industry the pressure for a cost-efficient supply chain grown steadily, especially for high cost components like titanium forgings. One significant impact on the costs of aerospace components is the material-efficiency, represented by the buy-to-fly ratio. Every saved kilogram of material from ingot and billet manufacturing to the manufacturing of forgings and the final component has a significant effect on the overall costs. With the introduction of additive manufacturing (AM) as a new, fundamentally different manufacturing method, also the microstructure design of already well-established materials could be challenged. As titanium alloys, specifically Ti-6Al-4V, is widely used for structural parts it is especially interesting to focus on optimizing its buy-to-fly ratio using AM. For bigger parts (> 0,5m) wire based technologies like Wire+Arc AM (WAAM) are the most promising to improve the buy-to-fly ratio form seven after conventional forging to potentially two without further topology optimization. As Ti-6Al-4V always shows a beta-microstructure after WAAM the question arises, if an optimized forging route towards beta-microstructure would lead to a similar cost benefit. Indeed, if a beta-microstructure is the goal from the start many processing steps can be skipped and the design of dies can be improved towards optimal near net shape forgings achieving similar cost savings like WAAM. Comparing those two processes geometrical (tolerances due to shape and distortion) and metallurgical properties (porosity, microstructure and mechanical properties) have to match customer specifications. This work compares the design and manufacturing challenges between WAAM and beta-forging and shows their differences in respect to their properties.

Low carbon lath martensitic steels such as PH15-5 are often used in high-performance applications e.g. structural components of aircrafts. Since the development and improvement of closed die forging routes for such parts is often time-consuming and expensive, especially for microstructurally complex materials such as PH15-5, a spatially resolved prediction of their mechanical properties as a function of microstructure is essential. Current frameworks are able to model the evolution of the austenite grain size throughout thermomechanical processing. However, prediction of the martensitic substructure size, which is a decisive factor for the mechanical properties during application, requires further research. Thus, a more detailed investigation of the correlation between austenitic grain size and martensitic block size is necessary. The martensite block size depends on multiple factors such as the prior austenite strength and grain size, as well as the temperature at which the transformation occurs. With each martensitic transformation sequence, stresses are introduced into the microstructure. It is assumed that they strengthen the remaining austenite and retard subsequent transformation sequences to lower temperatures, which affects the martensitic block size. The main aim of the present research is the improvement of martensitic block size predictions as a function of the austenite grain size. This will enable reliable modelling of mechanical properties based on fundamental parameters e.g. stress affected strength of austenite. In order to investigate the influence of the local but also mean stresses induced by the martensitic transformation on the remaining austenite in-situ synchrotron x-ray diffraction and in-situ high resolution electron backscatter diffraction measurements were conducted, respectively.

Forged Aluminum alloys have outstanding properties in light weight construction for highly loaded parts. Hence, these alloys are widely used as forgings for automotive chassis applications in millions of cars. In distinction to other manufacturers Krenhof GmbH operates in a niche market for so called “high performance cars” and “hypercars” as well as for racing applications. In this market the yearly production lies between some hundred parts up to ca. 15,000 cars a year. In comparison to typical mass production where some 100,000 cars a year are produced these special markets demand special challenges over the entire production chain in terms of materials and processes. For this purpose, often special materials and tailor-made production processes are used. The presentation gives an overview of the concepts which applied in terms of materials and production processes to fulfil the requirements of these special markets. Besides the technical perspective also economical aspects must be taken into account which often differ significantly from the framework conditions in mass production.

Modern bainitic steel is one very promising steel concept contributing to actual, international sustainability development goals reducing carbon emissions during production but also during application, as they combine low material costs (i.e. low alloying element content) and comparatively efficient production processes with enhanced mechanical properties such as strength and ductility in the final product. Their outstanding mechanical properties are based on a composite-like microstructure consisting of a soft phase, i.e. retained austenite and a hard phase, i.e. bainitic ferrite with potentially cementite precipitates. Whereas strength is mainly controlled by the bainitic phase, ductility is mainly affected by retained austenite’s phase fraction and mechanical stability against transformation to martensite during mechanical loading. Both parameters are a function of carbon content in retained austenite. Therefore, to predict and subsequently control the competing mechanisms of cementite precipitation in bainitic ferrite and carbon enrichment in austenite is decisive for the optimization of bainite’s ductility. To this end, thermo-kinetic simulations are carried out using MatCalc 1D cell simulations in order to investigate the influence of the chemical composition and local microstructure size (i.e. thickness of bainitic ferrite and retained austenite) on the maximum carbon concentration of retained austenite. Additionally, electron backscatter diffraction and transmission electron microscopy measurements are conducted for the determination of the thickness distribution of bainitic ferrite and retained austenite. Atom probe tomography measurements of the retained austenite carbon concentration validated the MatCalc 1D cell simulations. Finally, design criteria for chemical composition and microstructure size distributions of bainitic ferrite and retained austenite are derived for an optimized stability of retained austenite and ductility of bainitic steels.

On the example of Al 2024, it is shown how experimental analysis and computational materials simulation can interact to identify physical mechanisms that operate in the alloy in the course of component production. In particular, the rotary swaging (Rundkneten) process is taken as an example where more or less strong plastic deformation during shaping of the component is followed up by an annealing step. In the latter, the microstructure recrystallizes with a clear correlation between the amount of deformation that is introduced during rotary swaging and the resulting grain size. The experimental characterization is mainly performed with electron back scatter diffraction (EBSD), while the simulations are carried out with the Finite Element software ANSYS, which is used to determine the local distribution of strains as well as the software package MatCalc, which has been co-developed in several succeeding projects at the Materials Center Leoben, to simulate the evolution of microstructure.

The metals producing industry is continuously developing new processing capabilities and high-performance alloys with improved properties to meet societal demands in the transportation, energy, construction and health sectors. These developments can benefit from the use of advanced computational materials science tools to simulate microstructure evolution during processing of alloys including recrystallization, grain growth and phase transformation. Increasingly multi-scale modelling approaches are employed to gain further insight into the underlying microstructure mechanisms. In particular the interaction of alloying elements with migrating interfaces critically determines microstructure evolution rates. Here, atomistic scale simulations, e.g. using density functional theory (DFT), provide trend information that can assist alloy design. Further, meso-scale models, e.g. phase field models (PFMs), enable to simulate actual microstructures rather than average microstructure features such as fraction transformed and mean grain size. The status of these modelling approaches will be critically reviewed. A case study for recrystallization in gold will be presented to analyze the strengths and limitations of DFT informed microstructure modelling. Further, the status of similar multi-scale simulations will be discussed for phase transformation in steels.

Computational design of alloy materials relies on predictive modeling of alloy properties. While conventional ab initio methods offer such a possibility they are not efficient for disordered systems such as metallic solid solutions. The focus of our method development within the GreenALM project is an alternative approach based on the Green's function formalism. This approach brings two advantages: i) computationally very efficient modeling of disordered alloys within the coherent potential approximation (CPA); ii) highly scalable modeling using the supercell representation of alloys, which is a more demanding but more accurate technique than CPA. We demonstrate examples of the use of this computational methodology for studying properties of important alloy materials.

The world’s transition to renewable energy sources with reduced or zero greenhouse gas emissions gives hydrogen a special role of a sustainable energy source, as it can be produced and converted without CO2 emissions. However, there is a long-standing problem related to hydrogen-material interaction, also known as the hydrogen embrittlement. Most of the known-to-date structural materials suffer from it and new material solutions are requires ensuring safety and efficiency of the materials used for the new type of applications.
In this work, we would like to present the results of the successful application of a model-based approach to design of hydrogen resistant Ni-base alloy and steel development. The approach includes theoretical methods ranging from the atomic scale up to the component level, verified by the state-of-the-art experimental methods. Both theoretical and experimental methods applied in this work have been used for cross validation of the obtained results and applied in a complementary manner to investigate the problem of hydrogen-material interaction and possible acting mechanisms of hydrogen embrittlement across the scales.

Titanium alloys are used in many applications in aerospace jet engines from compressor blades through to critical rotating discs, each application having its own unique mechanical integrity requirements. Feedback from full scale component testing and service operation can be used to inform alloy composition and manufacturing routes for the future. Utilisation of a multiscale approach is vital for titanium alloys, where mechanical behaviour is determined by chemistry, microstructure and microtextural condition, all of which require characterisation, assessment and understanding for safety critical component application. This talk will consider some examples of threats in service and current understanding of key factors in alloy performance.

Compositionally Complex Alloys (CCAs) are a part of the High Entropy Alloy (HEA) concept, which currently draws much attention from both fundamental and applied research perspectives. For advanced high temperature applications, these alloys, strengthened by intermetallic particles, show high potential for replacing some types of state-of-the-art Ni-based superalloys in the future. Such alloys can be widely used in hot regions of high performance aircraft turbines. Within this application area, the need for novel materials with high yield strength, exceeding the strength levels of existing alloys for application at temperatures above 649°C, is strong. Materials that can withstand extreme conditions at even higher temperatures are needed to provide a better energy efficiency of the turbine engines by reduced fuel consumption.
In this work, we assess the potential of novel wrought CCAs, designed on the base of the state of the art alloy design concepts and predictive computational approaches. The composition of the new alloys was selected with the help of density functional theory (DFT) calculations and a newly developed DFT-based software toolkit (Material Design Toolkit) supporting high throughput calculations of multicomponent alloys with various degrees of compositional and magnetic disorder. The predictive power of the MDT has been assessed in a series of experiments and used in design of new CCA compositions.
Newly produced alloys have been characterized using high resolution microscopy investigations. The mechanical properties have been assessed using high temperature compression tests and comparing them to existing analogous commercial alloys. The results show that the designed alloys can reach very promising compressive yield strength at the application temperatures of 649°C and 800°C, surpassing conventionally used Ni-based alloys. Additionally, a potential of improved formability in a temperature range between 1050 and 1150°C can be anticipated from the results of performed deformation tests making them even more promising as wrought alloys.

There are a number of applications where materials are exposed to high temperatures and mechanical loads combined with specific environmental conditions. Applications are e.g. high temperature tooling which require materials combining oxidation resistance, high temperature strength and hardness as well as resistance against creep. The use of advanced materials in re-entry applications can be considered as another extreme use case. This presentation will provide an overview of material concepts which are developed for these demanding applications. Results from new material grades based on Ultra High Temperature Ceramics (UHTCs) as well as Refractory based High Entropy Alloys (HEAs) will be shown.
Powder metallurgical processing has been used to prepare full dense HEAs and UHTCs. Subsequent heat treatments were carried out in order to homogenize the microstructure. Microstructural analyses have been carried out in order to characterize the materials after processing. Oxidation testing was performed at temperatures up to 1.200°C. Additionally mechanical testing was performed and microstructural analysis was carried out to analyse the materials at different conditions.

Oxide-dispersion strengthened materials exhibit improved high temperature mechanical properties such as strength and creep resistance. However, their production via mechanical alloying is not only time-consuming and thus expensive but also in some cases difficult to implement. Therefore, in order to improve the efficiency of the mechanical alloying process, cryomilling was investigated for the production of oxide-dispersion strengthened alloys. For this purpose, milling was performed at room temperature and at cryogenic temperatures in a novel attritor ball mill using prealloyed fcc FeCrMnNiCo powders together with 1 wt.% yttria. Detailed investigations of the as-milled powders include X-ray diffraction and high-resolution scanning electron microscope as well as atom probe tomography, transmission electron microscopy and positron annihilation spectroscopy have been performed. The as-milled powders indicate an increased milling efficiency of cryomilling towards shorter milling times necessary to refine and dissolve a substantial amount of yttria but also yielded a more homogeneous yttria distribution for cryomilled powders at the same milling time. Nano-scaled and atomistic examinations suggest a vacancy assisted dissolution of yttria into nanoclusters. Furthermore, the replacement of yttria with metallic yttrium was examined in order to bind and reduce excess oxygen incorporated during the process. These experiments using a Fe-10Al-4Cr-4Y2O3 alloy were carried out under vacuum using a conventional ball mill and were mechanically tested in the consolidated state exceeding the creep resistance of the high-end Ni-based superalloy as CMSX single crystals at temperatures beyond 1100°C.

Materials Science Department, CENIMAT|I3N and CEMOP/UNINOVA, Faculty of Sciences and Technology, NOVA University Lisbon, 2829-516 Caparica, Portugal.

A massive increase of the internet of things (IoT) nodes is happening nowadays, since almost every device has embedded electronics. However, this will induce a high quantity of electronic waste (e-waste) accumulation. To overcome this serious societal problem, it is necessary to rethink the production process of some electronic devices and to reconsider their environmental footprint. By adopting more abundant metal oxides and processes, electronics waste is reduced, leading to the reduction of the carbon footprint. Printed and paper electronics fulfil these requirements and have been considered as key to surpass the high production costs, material’s waste and still allow large area production.

In this work we will present the most important landmarks achieved by these stimulating scientific areas as well as some insights to emerging applications such as green electronics.

Featuring "ultralow power consumption" challenges all the components of a smart system: starting from a power saving sensing device (the key enabling component of the application), but also continuing with efficient signal conditioning, acquisition and processing electronics, and ending with low energy protocols in the components in charge of data comunications. In this talk we will present our recent progresess on the optimization as a whole of a wireless gas sensor system with a really constrained power budget: less than 50 microwatts. Our gas sensor technology is based on the Micro Light Plate concept, a power efficient approach towards light activated semiconductor gas sensors. Turning the conductance signal generated by the sensor and driving its LED-based conditioning elements called for application-specific integrated circuits (ASICs) that involved a number of design strategies to keep the power budget under control. Finally, out of the many comunication protocols available, several options were choosen, tested and implemented; making compromises between datarate, range and energy needs. To meet these global requirements, simulations and models of different kinds were a must, and we will show how they guided us along the system integration path.

Advances in the miniaturization of electronics and manufacturing methods have empowered wearables with extremely fast processing capabilities while reducing their size at the same time. It has led to the widespread adoption of wearables in everyday consumer devices, various industrial applications, and personalized healthcare & remote diagnostics. With the increased computing power of wearable electronics along with the need to maintain a connection with the internet, the energy demand for wearables has significantly increased. The physical size of the energy storage elements cannot be accordingly increased to cope with the higher power requirements as it is limited by the intended application. Wearable devices, especially in health care and industrial maintenance applications, also desire and benefit from structural flexibility. An optimal solution to power the modern wearables is to i) discover novel printable materials to form higher energy density storage devices, ii) use advanced printing methods to increase energy power density or enable structural flexibility of current energy storage devices, and/or iii) develop methods and materials to form highly efficient energy harvesting devices. Here, we present an insight into the state-of-the-art printing methods (3D printing, inkjet printing, screen printing, electrohydrodynamic printing,) used to print novel materials for the creation of energy harvesters (triboelectric nanogenerators, solar cells, RF antenna-based, thermoelectric energy generators, and piezoelectric energy harvesters) and energy storage devices (electrochemical batteries, biochemical batteries, and supercapacitors).

Devices for the Internet-of-Things are often placed in remote locations or are embedded in vehicles or machines and thus need to be energy-autonomous. Highly promising energy sources are perovskite-based solar cells, that have overwhelmed the field of photovoltaics with skyrocketing efficiencies, while simultaneously allowing light weight and cost efficient fabrication and hence a low energetic (and financial) uptake. For integration it is of the essence to store superfluous energy for timespans of darkness. Here providing high charging voltages is an important component for energy density in batteries and even more in capacitors. Unfortunately voltages of high efficiency single solar cell architectures are limited by the optimal band gap for light absorption, which is why tandem concepts are highly desirable, where the voltages of sub-cells add up and are able to advance into formerly unreachable areas.
Here, we demonstrate perovskite/organic tandem cells with an open circuit voltage of 2.2V and efficiency of 24%. This sets a new milestone for these type of tandem devices with a realistic prospect of reaching >31%.
This progress is based on a threefold approach:
- In surprising contrast to these almost paradigmatic concerns, we
evidenced an outstanding operation stability of organic single
junctions that are based on the PM6:Y6 active system, if operated at
tandem-like conditions.
- We managed to overcome interfacial losses, that are the predominant
reasons limiting the output voltage of wide-gap perovskite cells. At
the same time, our findings evidence that the proper choice of charge
extraction layers can mitigate some detrimental degradation
mechanisms inside the cell.
- We introduce a novel recombination interconnect for the two
sub-cells, that is based on an ultra-thin (1.5nm) metal-like indium
oxide layer. This interconnect offers unprecedented low optical and
electrical losses, which unlocks the exploitation of the full
potential of the two sub-cells without any discount.

Rotary-percussive drilling in steel reinforced concrete subjects cemented tungsten carbide hammer drill bits to intensive wear from many active wear mechanisms as well as high mechanical and thermo-mechanical loading which may lead to premature failure due to overload or fatigue fracture. To ensure the required component lifetime and reliability in this challenging application, a large number of varied but ultimately complementary testing methods are necessary. Whether it is measuring the composition of the raw materials at parts per million accuracy or testing finished parts in drilling tests running to millions of impact load cycles and Megajoules of energy, great effort is used to characterise the materials, components and assemblies that make up a hammer drill bit. This presentation will highlight the micro-, meso- and macrosopic physical, mathematical and simulation-based testing methods employed to better understand the cause and effect chain of processing <-> microstructure <-> properties <-> performance for cemented tungsten carbide hammer drill bits. You might just be surprised how much science and engineering is behind this seemingly simple part that you use to make holes with!

Hardmetals are used as tool materials in various areas such as metal machining or metal forming, where they are exposed to high temperatures and high loads. Hardmetals have good high-temperature properties, but these properties are influenced by the microstructure and operating condition. Therefore, knowledge of limits for their failure-proof application are important for the design of hardmetal tools. The aim of the current work was the experimental determination of these limits and the connection of them with microstructural changes in WC-Co hardmetals at 700 °C and 800 °C. Six WC-Co hardmetal grades were tested under uniaxial cyclic compression at a stress ratio of R = $\sigma_{min}/\sigma_{max}$ = -$\infty$ in a vacuum. The investigated grades differ in their WC grain size (0.4 μm to 2.0 μm) and Co-content (6 wt.% to 12 wt.%). Limit stresses are determined by examining the influence of increasing stress ranges and their effect on the (residual) strain evolution at zero applied load as a function of microstructure. At low stress ranges, the residual strain was observed to stabilize after a characteristic number of load cycles. Advancing strain ratcheting occurred above a critical stress range (limit stress), above which strain stabilization does not occur any more. By scanning electron microscopy, the formation of microdefects, such as cavities and nanopores, at phase triple points, WC/WC grain boundaries, and WC/Co interfaces were observed to be promoted by advancing strain ratcheting. Due to the increase in microdefect density in the microstructure with increasing stress range, statements about early failure are possible based on the compressive stress-strain data and the course of the residual strain with increasing number of cycles. The limit stresses determined are therefore regarded as limits for the failure-proof use of tools and components.

Due to its high hardness and wear resistance as well as its good chemical stability, chemical vapor deposited TiB2 serves as protective coating in machining of Ti and Al alloys. To prevent B diffusion into the substrate, typically a TiN diffusion barrier layer is deposited beneath the TiB2 layer. However, the sharp transition from the face-centered cubic (fcc) TiN with its tensile residual stress to the hexagonal (h) TiB2, which typically exhibits compressive residual stress, results in poor adhesion. In order to strengthen the interface, a graded Ti(B,N) interlayer was implemented. In a first step, an interlayer providing a smooth transition from fcc-TiN, via TiBN containing 5, 15, 30 and 45 at.% B to h-TiB2 was synthesized. The graded coating and corresponding single layers were investigated in detail using cross-sectional and advanced high resolution methods. With increasing B content, a decreasing grain size and a rising h-TiB2-based phase fraction was observed, which strongly affects the residual stress, provoking a change from ~1 GPa tensile to ~1 GPa compressive across the coating thickness. The hardness almost linearly increases with rising B content from ~18 GPa for fcc-TiN to ~41 GPa for h-TiB2 and also the fracture stress and toughness increase from ~7 to ~13 GPa and ~4.6 to ~5.5 MPam1/2, respectively, with the only exception of the TiBN coating containing 45 at.% B, which exhibits a significant drop in hardness and fracture properties. The deteriorated mechanical properties can be related to a transition from a fcc-Ti(B,N) dominated structure at lower B contents to a highly defective h-Ti(B,N)2 dominated structure at 45 at.% B. Based on these findings, in a next step a further improved graded interlayer with B contents adjusted to below 15 at.% was synthesized, which exhibits excellent adhesion and superior performance in cutting application compared to the original TiN/TiB2.

Variation in lifetime of used coated cemented carbide tools in cutting applications are evident, even when using the same machine, same cooling conditions as well as the same steel material.
The common steel classification only defines the chemical composition of the steel grades, but the frequency and size of non-metallic inclusions is not clearly determined.
This study examines this factor for the apparent wear damage mechanisms and end of lifetime reached in a longitudinal turning application of a hot work steel X37CrMoV5-1.
The investigated turning inserts were based on WC-7wt.% Co coated with a TiN-TiCN-Al2O3 based CVD film.
To focus on the influence of the non-metallic inclusions, the hot work steel X37CrMoV5-1 from four different steel suppliers were tested with varying cutting parameters.
The chemical composition of the inclusions and their frequency were investigated using a scanning electron microscope. The obvious damage mechanisms on the turning inserts were documented with a digital microscope.
The chemical composition, a high number and the size of the inclusions significantly influence the end of lifetime of the turning inserts.

Accurate prediction of the wear evolution and damage of rails and wheels is essential for effective maintenance processes to be achieved. Knowing with more certainty when grinding/turning or replacement is needed can help in planning the processes better which in turn will help reduce costs and keep availability of track and vehicles at an optimum.
There are very sophisticated multi-body dynamics models and numerical models available for understanding vehicle performance and wheel/rail interface conditions. In order to develop tools for damage prediction, however, these must be combined with effective tribological wear, damage and friction models. This an area where many improvements can be made. Many models are semi-empirical, which is fine, but this means that they only work for those cases where the input data is available. There is a great need for physically based models to overcome this.
This presentation will give an overview of some recent work on wheel/rail tribology to help understand prevalent wear and friction mechanisms and some tools developed to improve their prediction aimed at improving the situation.

The last few years showed the infrastructure managers that it is not only necessary to create and maintain a superstructure able to carry the loads from the trains but a system also being resilient to environmental impacts like floods, fire, landslides or else. At the same time a fast-rising importance of green travel possibilities and the shift from road towards rail traffic is happening, not only due to the climate benchmarks of the EU. The goal of Austria – a rise of 30% of trains upon an only slowly growing infrastructure – shows that the timespan for maintenance is going to be reduced drastically. Knowing that our superstructure has a mean lifetime of 35 years also indicates that in secondary tracks it is possible to find much older superstructure. It is now time to design a maintenance reduced superstructure, less vulnerable towards environmental impacts and optimize the maintenance planning.
This presentation will give a short overview of the changes within superstructure and maintenance in the last few years and also give some insights towards ongoing research.

Currently railway turnouts have a significant higher demand for maintenance compared to standard / tangent track. The root cause behind this increased maintencance demand is based on the effect of higher forces in the wheel/ rail contact due to transition zones, e.g. in the crossing panel or in the switch panel.
In the European project Shift2Rail, voestalpine first analysed together with its industrial partners ÖBB Infrastruktur, Kirchdorfer Concrete Solutions and Getzner Werkstoffe the periodic maintenance tasks and the costs behind. The result of this analysis showed that highest maintenance costs occur for the exchange of crossings or half set of switches but also for ballast tamping and ballast cleaning to provide a good track geometry.
In the following, with the overall aim to reduce in the future the maintenance demand in turnouts, an upated turnout design was developed and out of this development finally two demonstrator turnouts were manufactured and installed in 2020 in Vienna. These demonstrator turnouts have numerous innovative features included:

1. High wear and RCF resistant rail grade 400 UHC and special switch rail machining TOZ+ to extend the lifetime of the half set of switches.
2. Elastic support below the crossing to reduce peak forces during the wheel transition process to extend the lifetime of the crossing
3. Introduction of an elastic fastening system for optimized stiffness distribution through the whole turnout and a new sleeper design including sleeper coupling systems and under sleeper pads to optimise the ballast load and furthermore the settlement of the turnout.

Finally, to prove the effectivness of the optimised design a large measurement campaign including permanent measurement systems was setup.

Modelling of squats
Sandra Baltic, Timna Gschwandl, Werner Daves

Squat-type rail cracks are characterised by a localised depression of the rail-wheel running band. A squat-type crack typically starts on a rail head surface and grows to depths of a few millimetres. Such cracks can grow within month or few years to a size, that rails have to be replaced. As this damage type is a worldwide problem it costs the railways billions of Euro per year.The problem of squats aroused mainly within the last 20 years and no solution to the problem is actually visible.
Possible initiation points are head checks, pittings but also scratches caused by grinding of rails.
Cross sectioning of rail samples revealed the crack initiation in planes perpendicular to the rail traffic direction. It is an aim to show the influence of factors that might trigger the squat initiation and formation.
The tendency of one defect to grow rather than another is analysed by calculating crack driving force value and directions.The results show that cracks are able to grow under particular wheel loading in the directions that are in accordance with observed squat growth. These results suggest lateral forces as the main suspected determinant for the squat initiation.
Residual stresses in rails have to be regarded, too. Stresses near the surface and especially near the plastified zone below the running band will strongly influence the crack path and growth rate of cracks. Such residual stresses are measured and are input for the simulation of squats.
Multiaxial cyclic experiments using tubular specimens are undertaken, to characterize the stress, plastic strain response of different rail materials and are used as input for squat simulations.
The lecture describes the actual efforts to accurately simulate squats from their initiation to a state the rail has to be maintained or replaced.

Since the large automotive companies go ahead with the electrification of the power train, also the demand for more powerful electrochemical energy storage systems is growing. However not only the automotive industry requires more powerful accumulators. In the area of mobile consumer electronics, the proceedings of the last years are compensated by the rising energy demand of the devices. In addition, the fast charging requirements for some consumer electronics products are approaching the same level on the same level compared to the automotive industry.

The specific energy of a lithium ion battery is given by the product of specific charge - more precisely the lithium storage ability per mass and the voltage difference of negative and positive electrode. As far as the negative electrode is concerned lithium and lithium compounds already represent the most reductive species of the electrochemical series. However carbonaceous materials that are used in state of the art lithium ion batteries are only able to store up to 1 lithium ion per 6 carbon atoms in the case of fully crystalline graphite, which results in a quite low charge density of 759 Ah∙l-1. The application of metals or semi metalloids, which are able to store lithium under formation of an intermetallic phase represents a very promising alternative, whereas especially silicon has turned out as a very promising material that would offer - at least theoretically – more than 3 times higher charge densities. It is well known that the high charge density comes along with several challenges like structural stability, instability of the SEI layer as well as special requirements for the cell design. Within this talk advantages and disadvantages of different silicon based material concepts in combination with their requirements for processing and cell design will be discussed.

To this date lithium dendrites remain one of the key challenges in solid-state Li-batteries (SSLB).1,2 The formation and growth of these dendrites cause an inevitable failure at charge rates far below the threshold set by industry (>3mA/cm2) and are believed to be driven by stress accumulation stemming from the deposited lithium itself. Upon reaching a critical pressure, the solid-state electrolyte (SSE) starts to crack, which propagates until the cell short-circuits.3 Herein, we propose that this fracturing is not of mechanical origin but can rather be related to an increased Li activity in the SSE near the protrusion, caused by the diffusion-controlled Li deposition. This destabilizes the local structure of the SSE and weakens it within a certain time frame. If current is applied for shorter durations than is required for the Li activity to reach critical values, as is with high frequency pulses, the structural destabilization of the SSE can be delayed, and higher critical current densities (CCD) can be reached. Applying 1 MHz current pulses, a sixfold increase of the CCD, compared to DC operation and a maximum value of around 6.3 mA/cm² was reached in this work. Since an adaption of the applied current waveform can be combined with other reported measures like interlayers, alloys or structured interfaces, the fast-charging goal set for electric vehicles, appears to become feasible.4,5,6 Hence, the application of pulsed currents represents a crucial step toward realization of SSLBs for electric vehicles and other emerging applications.

The development of the key components of battery-electric and fuel cell vehicles, namely the battery pack and fuel cell system including the related cooling system and hydrogen/air media supply for the latter, is closely linked to different challenges with regards to performance, lifetime and safety. In the case of battery-electric vehicles, the engineers must specifically take care of the proper selection of the battery cell type, the related packaging and thermal conditioning under different real-world driving and charging scenarios. For fuel cell electric vehicles, the sizing of the fuel cell stack and the related hydrogen/air supply system as well as the proper cooling system dimensioning are decisive for the overall system efficiency and achievable lifetime.
In the above context, scalable multi-physics simulation is adopted to support the development engineers in the different phases of the development process, from concept layout, to detailed component and sub-system development, to virtual integration and calibration. Depending on the specific development tasks, different simulation methods and tools are applied. These comprise behavioral models, usually neglecting geometrical details and reflecting components and sub-system behavior in a 0D/1D manner, and geometry resolving methods providing a detailed 3D space- and time-resolved insight into the various mechanical, fluid-flow, thermal, electrical and electrochemical processes.
Following a brief introduction of the methods and tools adopted, their application to support typical engineering tasks related to the development and optimization of batteries and fuel cell systems is presented. The overview of typical battery related development tasks comprises performance optimization, hazard prevention and lifetime assessment under real-world scenarios. The selected fuel cell related applications range from detailed anode and cathode flow-field design and membrane-electrode assembly component selection, to stack media supply and thermal management analysis, to system layout and components sizing. Finally, an outlook on future research and development needs is given.

Energy systems based on solid oxide fuel cells (SOFCs) and solid oxide electrolysis cells (SOECs) are among the most promising future technologies for sustainable clean energy supply and storage. SOFCs convert the chemical energy of a fuel into electrical energy with high efficiency and low emissions. A major advantage of SOFC technology is its fuel flexibility. SOFCs are capable of converting various fuels (natural gas, biogas, syngas, bio-ethanol, etc., in addition to pure hydrogen) directly into electrical energy. SOECs provide an efficient and sustainable means of storing electrical energy from volatile renewable sources. By reversing the SOFC principle, water is electrochemically split into the components oxygen and hydrogen using electricity, e.g. generated from wind energy, and thus electrical energy is stored in the form of chemical energy in the hydrogen. The switch from SOFC to SOEC mode (and vice versa) can be carried out at short notice, making it an extremely flexible system. In contrast to low-temperature electrolysis cells, SOECs are capable of electrolyzing carbon dioxide/water mixtures in addition to pure water (co-electrolysis) and converting them into synthesis gas (CO, H2) for the production of hydrocarbons, methanol, ammonia, etc. ("power-to-gas" concept) in a methanation reactor. SOECs thus contribute to a reduction of greenhouse gases.
In our research, we investigate fundamental relationships between material properties, morphology and electrochemistry of solid oxide cells with a focus on the air electrode and the air electrode-electrolyte interface. Together with our partners, competences in the field of material development, electrode and cell preparation, and electrochemical characterization, are bundled with detailed microstructural and chemical analyses. The goal is to develop a knowledge-based design approach for the next generation of solid oxide cells for sustainable and efficient energy storage and conversion.

The SBB infrastructure is one of the most densely used networks in the world, with heavy, slow freight and fast, frequent passenger traffic running on the same tracks.
The maintenance of the assets is a major challenge, because service interruptions for maintenance have to be integrated into the timetable for a very long time (> 1 year).
Based on rail treatment, the presentation shows which rolling contact fatigue defects (RCF) occur mainly at SBB-Track and how important it is to know the right time to intervene for early, still repairable defect elimination. Today we have a lot of calculation methods of material research for explain and predict RCF. So far so good, but how can these models to be scaled and applied to an entire railway network?

SBB has been working on applied research for years, many data prepared, and calculation methods have been created in the last decade. The currently very simple prediction system of RCF and the approach of replacing parts with more detailed material research based models of the MCL are to be explained on the SBB rail maintenance as example.
It is the central part, that the railways must move forward with the creation of data and calculation basics as SBB has done. But researcher community must move as well! Highly complex and therefore truthful models with detailed input parameters that an infrastructure operator cannot provide will never find a way into practice.
It will be shown in detail in preparation, what SBB has made and how it is planned to connect on the research work of MCL. An example are the changes in generations of locomotive technology, which are a co-trigger of RCF.
This example should be used to show how model changes or parameter replacements alone can influence positive the usability of the models.

With current development in sensor technologies, data processing, numerical simulations, and life prediction methodologies, there are increasing possibilities for monitoring and predicting track health. This has led to an increased interest in so-called “digital twins”, i.e. digital models of the actual railway track. Ideally, a “digital twin” should be updated as the track characteristics evolve, and also be able to predict future development. The current presentation will set out from the concept of a “digital twin” and focus on some challenges related to the concept. This includes how the current track status can be quantified and which parameters that should be employed in a track health quantification. Such a quantification requires operational parameters. However, measurable parameters are often not parameters that can be directly included in predictive models for track health. This calls for a translation that is sometimes far from obvious. An example of this is the translation between (fatigue) damage, which has to be translated to measurable physical deterioration (e.g., in the form of fatigue crack length). Finally, a full track health assessment will include a multitude of trac components and pertinent deterioration phenomena. To limit the scope, the presentation will highlight some of these and exemplify how they can be (and are) addressed.
The work is partly funded by the European Union’s Horizon 2020 research and innovation programme in the Shift2Rail projects In2Track3 under grant agreement No.101012456.

A railway network consists of myriad important components that finally form a smoothly running railway infrastructure when optimally coordinated upon each other. Switches are of utmost importance in this context, as they allow trains to change between tracks during ongoing operation and thus enable trouble-free services. Such functionality is naturally associated with high costs, as the moving parts lead to discontinuities in the points and the associated higher dynamic forces cause faster wear of the components itself. Consequently, not only the maintenance costs increase, but also the overall life cycle costs.
In order to counteract this trend, innovative methods and concepts are necessary so as to avoid even more pressure on the infrastructure operators' already strained budgets. This presentation shows essential problems in the turnout and presents the results of the recent years of intensive research and field-tests carried out in a close collaboration with several infrastructure managers. Future-oriented and practicable solutions will be unveiled that pave the way towards prescriptive maintenance and a sustainable railway infrastructure.

Railway turnouts feature discontinuities in wheel-rail contact geometry and track stiffness. These discontinuities results in large dynamic contact forces causing degradation of the rail surfaces and track irregularities over time.
Currently, researchers focus mostly on one damage mechanism at the time, however, the interplay and interconnection between those mechanisms have not yet been addressed.
The aim of the current work is therefore to develop and demonstrate a novel methodology based on the interconnected submodules to make such holistic predictions for accumulated track damage in turnouts and account for the interplay and interaction between damage patterns such as the development of track irregularities due to ballast and subsoil settlement, and rail profile changes due to plastic deformation and wear.
We found that at the beginning of the turnout service life the running surface of the rails in these areas change significantly due to plastic deformation, which in turn increases the dynamic impact forces from passing vehicles.
It is shown that the developed WSM methodology can predict accumulated track damage accounting for the coupling between different damage modes such as track settlement, rail profiles plastic deformation and wear. It can therefore be used for holistic assessment of turnouts e.g., in turnout design optimisation and material selection studies.

Various different damage mechanisms driven by local contact loads, contribute to material degradation during long-term railway operations. To assess the material specific damage behavior in crossings a numerical framework is proposed to predict the damage evolution based on three contact parameters. The framework transfers in-service load conditions to the local damage models using three parameters: pmax (maximum contact pressure), c (creepage) and 2a (contact patch length). By finally using wear and RCF models from literature, the proposed method offers a good prediction tool for the material specific assessment of damage in crossings or rails.

The presented work shows a hybrid model based and data driven condition monitoring approach for railway crossings. A virtual sensor concept is used to estimate the location and the length of the transition area. The transition area is a well-known crossing state indicator, which is evaluated by service technicians during visual inspections to rate the crossing condition. This feature can be additionally used to rate wheel profiles and thereby enable a whole system monitoring of rolling stock and track with a single monitoring concept. The monitoring concept is installed at two demonstrators and long-term data has been recorded over the last 2 years. The collected data set can be subsequently used to develop prediction models and thereby facilitate a condition-based maintenance strategy for the operators.

Power transistors based on wide band gap semiconductors have become commercially available in recent years. In the case of power conversion for e-mobility, silicon carbide (SiC) power transistors are a serious competitor to established silicon technology due to their low on-state resistance at high breakdown voltage. However, raw SiC wafer cost is a key cost driver and currently limits its widespread adoption. Gallium nitride (GaN), on the other hand, can be grown epitaxially on affordable foreign substrates such as silicon or sapphire and has been successfully commercialized for lower voltage applications using a lateral transistor architecture.
This talk will introduce the concept of vertical GaN membrane power transistors, a new class of power transistors which is currently developed by the European ECSEL project YESvGaN. It combines the cost advantages of GaN heteroepitaxy on affordable substrates with the current handling capabilities of a vertical transistor architecture which is needed for, e.g., the traction inverter of an EV. Besides an introduction to the concept itself and the current state of development, the talk will highlight the challenges from epitaxial growth, over chip processing to the interconnection technology which have to be mastered for vertical GaN membrane transistors to become the next generation of power semiconductors.

In this talk, the evolution of hybrid printed electronics for flexibles and wearables applications will be reviewed. Hybrid integration leverages the strengths of flexible and printed electronics to manufacture compliant and large-area electronics and silicon-based technologies for computation and signal communication. Here we will address the state-of-the-art ultra-thin-chip (UTC) development to comply with the flexibility of the printed substrates. Successful and reliable integration of the UTC to low Cost printed circuit boards such as PET either in a stop & go fabrication line via die-bonders or in a roll-to-roll line via in-line assembly modules are the most potent game-changer of printed hybrid electronics. Moreover, green electronics based on paper which have been considered as one of the emerging technologies will be assessed and the advancements and challenges in hybrid integration of UTCs on E-papers will be discussed.

Lithium (Li)-ion based batteries are one of the most used systems in decentralized storage systems, e-mobility or mobile electronic devices etc. However, an essential problem of state-of-the-art cells concerns the fast fading of the capacity during electrochemical cycling. The use of silicon (Si) as an active material in the anode provide promising prospects. The understanding of the microstructure of silicon-based anodes and its change with electrochemical cycling in connection with the electrochemical properties is highly crucial to develop more advanced Li-ion batteries. However, a big problem for the microstructure characterization concerns the complex hierarchical structures of the anode material going from m- down to nm-scales. Here, multi-method approaches are essential to cover the different scales with respect to resolution, contrast, and the representative volume of interest to gain sufficient statistical information while maintaining the ability to extract the needed information at relevant scales and with appropriate contrast modalities. Another challenge concerns the accurate and efficient analysis of the “big” microstructure data. In this work we discuss a multiscale correlated workflow suitable to investigate the microstructure of Si-based anodes in correlation with chemical element information and to gain a comprehensive, multiscale, representative picture of the intricate microstructure dictating the ultimate electrode performance. To handle the big image data accurately and to retrieve statistically relevant microstructure information we implement a machine learning based architecture for the segmentation of the gained image data of the Si-based composite anode.

The Internet of Things (IoT) is based on a network of trillions of electronic components such as sensors and actuators, which must communicate with one another and perform intelligent action on the environment. Autonomous, wireless operation is the key enabling factor for most IoT applications, which poses significant challenges for the energy supply of such devices. In many cases, in fact, the usage of grid connection or replaceable batteries is neither possible nor sustainable. The greenest and most efficient way to realize energy autonomous IoT devices is thus to combine energy harvesting modules with energy storing elements.

Ceramic ultracapacitors are attractive as energy storage materials for the IoT, because they combine high power density with high energy density. This way, the stored energy is quickly available for fast operation (i.e. to supply wireless transmission/receival peaks) and can be supplied over hours – depending on the power budget of the IoT device. In this presentation, an overview will be given on ceramic ultracapacitors from the atomistic description of energy density to the peculiarities of industrial processes. In particular, it will be focused on the interplay of chemical composition, microstructure, and system architecture, suggesting possible ways how these devices could be further improved or implemented.

Continuous casting is the most dominating process in modern steelmaking, with more than 95 % of the annual world steel production. Currently, great effort is made to develop online-capable quality prediction systems based on fundamental metallurgical concepts with fast runtimes. The essential part of so-called “quality index” strategies is, of course, their industrial verification but also their further improvement using high-fidelity offline software tools. The present work introduces a 2D offline heat transfer model as a development platform for continuous slab casting. The calculated temperature fields are critically verified with temperature measurements performed during the casting process in the first part. Then, simulation results of two quality indicators are presented, describing (i) the phenomenon of non-uniform heat withdrawal during initial solidification in the casting mold caused by the volume contraction during the peritectic transition and (ii) the risk of internal hot tear segregation formation with particular respect to transversal half-way cracks. The consideration of the actual chemical composition of the steel grade leads to a reasonable agreement between the calculated peritectic indicator and local temperature fluctuations in the mold recorded by permanently installed thermocouples. Based on systematic quality analysis after casting, the hot tear criterion can be successfully applied to reduce the risk of the formation of transverse half-way cracks by adjusting the casting parameters. Finally, a slight modification of the peritectic indicator is implemented into the online capable system “DynaQI” running at the industry partner’s slab caster. Despite the rapid calculation procedure of DynaQI the results show even more improved performance in quality prediction and confirm the approach of offline/online coupled development of quality indices.

In this paper we present the results of the simulation of a dross particle growth from a zinc melt. It is shown that despite the isotropy of the interfacial kinetics used in the simulation, the hydrodynamics alone could introduce an anisotropy in dross particle growth, similar to the natural anisotropy. The natural interfacial kinetics in faceted crystal growth are responsible for the faster growth of some facets and therefore their disappearance. A front tracking method with a cellular automaton approach was applied to the growth of free Fe2Al5 particle in a Zn melt at constant temperature. The dross particle growth was found to be dependent of the magnitude velocity flow and on the orientation of the flow melt with respect to the dross particle orientation. The boundary diffusion layer is modified by the flow and therefore the mass transfer and consequently the growth. To obtain in simulation the disappearance of a facet, longer simulation times should be done, however a good qualitative comparison was observed with the experimentally observed dross particles.

The requirements on new materials used in mobile communications are driven by the demand for higher data transfer rates. Single crystalline piezoelectric materials such as LiTaO3 and LiNbO3 have thereby qualified as substrates for precise and efficient frequency filters and are consequently employed in the newest 5G mobile communication network standards. Crystal growth in specific orientations are usually pursued to ensure optimized functional properties. The question arises whether such orientations may withstand thermo-mechanical loading during qualification and/or service.
In this work, biaxial strength measurements along with in-situ SEM fracture toughness experiments were performed on miniaturized specimens to investigate the mechanical performance of LiTaO3 and LiNbO3 materials. In addition, nanoindentation experiments showed onset of plastic deformation for specific loading scenarios which may arise in future SAW designs where thin film and multilayer architectures are increasingly important. Corresponding knowledge for the micro-scale may be utilized in the fabrication of LiTaO3 and LiNbO3 materials of particular orientations with optimized structural and functional properties.

Thermophysical properties of materials and the optimization of the heat transfer are becoming more and more important for industrial applications of micro- and nanoelectronic devices. Thin layers in the micrometer to nanometer range are used to give specific functions to the devices. Since the thermophysical properties of thin layers differ from bulk material, this data is required for precise predictions of thermal management. One way to obtain the thermal properties of thin layers is the optical-based Time Domain Thermoreflectance (TDTR) method. To carry out TDTR measurements with a low level of uncertainty, the samples under study must meet requirements related to the surface roughness and a low level of optical scattering. The range of samples analysable by TDTR can be extended by applying the so-called bidirectional heat flow approach. This approach opens the possibility to assess thermal properties of materials with rough surfaces as well. The validity of the implemented model was shown by the characterisation of a test sample with well-known thermal properties fabricated for this purpose out of poly(methyl methacrylate) (PMMA) roughened with acetone:ethanol. The results obtained by TDTR measurements were compared to literature values, demonstrating the applicability of the bidirectional heat flow approach for this setup.

Mechanical properties of metals and alloys depend critically on the microstructure, for e.g., yield strength increases with a decrease in average grain size, also known as the Hall-Petch effect. During thermomechanical treatments, microstructure processes such as grain growth and recrystallization significantly modify the average grain size and grain size distributions. Both these processes involve the migration of high-angle grain boundaries that are sensitive to the grain boundary structure and alloying additions. Grain boundary migration rates in metals, quantified by intrinsic mobility and solute segregation determined from the segregation energy are both anisotropic properties of a grain boundary. Nevertheless, most experimental studies assume effective grain boundary mobility and segregation energy as adjustable parameters to describe grain growth rates. Here, the phase-field method is used to simulate grain growth with anisotropic grain boundary properties to determine the representative grain boundary mobility and segregation energy. A critical discussion will be presented to identify the limits where a representative grain boundary can be defined. Within these limits, a representative segregation energy for solute elements informed from DFT simulations will be used to identify the candidate solute elements that may promote grain refinement of austenite in steels at high temperatures.

High-temperature laser scanning confocal microscopy (HT-LSCM) represents an excellent in-situ technique to observe austenite grain growth at elevated temperatures. In manufacturing processes, the grain size represents an important material parameter, especially for surface defect sensibility during continuous casting. Controlling the austenite grain size in further production steps determines the workability, process route and overall energy consumption to achieve the required product demands. Elements like niobium, phosphorus or even manganese tend to segregate on grain boundaries and inhibit grain boundary movement. This effect is well known as solute drag and can be indirectly characterized using HT-LSCM, enabling to provide accurate data for physical modelling of grain growth under isothermal and non-isothermal conditions. The present work focuses on the influence of carbon and phosphorus on the grain boundary movement of γ-Fe and summarizes selected results in the temperature range of 1050 - 1350 °C depending on initial C and P contents.

Nowadays, most steel grades are well described by commercial databases enabling to perform reliable calculations of thermodynamic properties and phase diagrams over a wide composition range. However, to further improve thermodynamic databases for advanced steels, e.g., silicon alloyed electrical steels, medium Mn-steels, or higher alloyed tool steels, there is still a need to provide new phase equilibrium data obtained by modern laboratory techniques.
This work presents an overview of the experimental procedure at the Chair of Ferrous Metallurgy at Montanuniversitaet Leoben. Taking the example of the selected systems, the workflow will be demonstrated. Investigations started with melting high-purity alloys by induction melting and subsequent centrifugal spin casting. Phase transitions of the alloys were systematically determined up to the liquidus temperature using the well-established differential thermal analysis (DTA) and differential scanning calorimetry (DSC) method. Since manganese shows strong evaporation tendency during melting leading to a possible damage of the DSC measuring devices, a micro-DTA-protected setup with closed crucible by tantalum lids is applied. Using these methods, all high-temperature phase transformations of peritectic, eutectic and eutectoid steel systems can be measured under equilibrium conditions. The measured phase equilibrium data are typically used to develop advanced thermodynamic databases for steel in the CALPHAD framework.

The production of steel coils with scrap material using an electric arc furnace (EAF) results in a very low CO2 emission compared to traditional production in blast furnace followed by basic oxygen steelmaking, but introduces many foreign elements by scrap. The impact of these foreign elements on the mechanical properties, such as the plastic strain ratio (r-value), is in many cases not understood entirely and the role of nano-precipitates are not captured by the process analysis. Predicting the r-value that determines the deep drawing capability of steel coils is a prerequisite for producing high-quality flat steel by EAF route.
In this work we apply AI regression models for predicting the r-value of steel coils from chemical composition and process parameters. The data from steel production and tensile tests was provided by voestalpine Stahl GmbH and includes a full chemical analysis, as well as many parameters from all process steps and the resulting mechanical properties. As a prerequisite for training of AI models, the data needs to be understood, analyzed, checked, and unreasonable data be removed (data cleaning). Additionally, methods for data fusion are investigated. The result is a machine-readable dataset fit for various modelling tasks. The used AI models include Random Forest Regression, Support Vector Regression, Artificial Neural Networks and Extreme Gradient Boost. In this poster the necessary steps of this workflow are summarized and a critical analysis of the applied models are presented.

The Materials Center Leoben (MCL) develops and implements simulation tools and software to characterize, understand and design materials properties at different length scales, using various methods closely interlinked with experiments. MCL constantly pursues novel and vanguardist Software development approaches. To streamline the software development workflow, MCL started the implementation of continuous integration and continuous deployment (CI/CD) tool infrastructure. CI automatizes the integration of code changes from multiple developers into a shared repository by combining automated building and testing units. At the same time, CD provides automated deployment services to deliver the changes to end-users. Based on gitlab (gitlab.mcl.at), a CI service is already available at MCL, and the first projects are already using the infrastructure. This poster will provide a general overview of the MCL-CI/CD development status, covering the infrastructure used and practical aspects of implementation and usage. In addition, an outlook on the future services and user projects will be provided.

The Materials Center Leoben has a broad spectrum of innovative, useful, and user-friendly software solutions for material, process and product development, covering different length scales and methods. Some examples at the macroscopic scale include tools to predict crack propagation in wheelset axles (INARA) and visualize residual stresses in rails (VRS). Going down to the atomistic level, the MCL developed a software to predict grain boundary segregation (SEGROcalc) and an ab-initio based toolkit to study chemically and magnetically complex alloys (GreenALM). This poster will provide a general overview of some representative MCL-Software, which has successfully simplified complex scientific methodologies into workflows/tools accessible to less-expert users. In addition, an outlook on future software development activities will be provided.

SAC305 (Sn–3.5 wt.% Ag–0.5 wt.% Cu) is one of the most promising alloys to replace Sn–Pb solder alloys in microelectronics packaging. During reliability testing, ball grid array (BGA) test vehicles are subjected to accelerated thermomechanical cycling in order to induce crack initiation and crack propagation. The thermomechanical behaviour of SAC305 solder balls varies widely because of the mechanically and thermally highly anisotropic properties of the Sn-matrix. The anisotropy of Sn plays a significant role in the reliability of the BGA since solder balls usually solidify as single or few-grained crystals.
The developed method workflow allows (1) the non-destructive inspection of entire BGAs for existing cracks in solder balls and (2) provides statistically relevant 3D information about the severity of the damage in each solder ball by applying semi-automatic image analysis. Therefore, primary causes for crack initiation and crack propagation in solder balls during thermomechanical fatigue testing can be gained.

In recent years, 3D integration technology plays an important role with respect to improved functionality and miniaturization of microelectronic devices. Through Silicon Via (TSV) technology enables electrical connection through the different layers of the 3D stacks. Those metallized vertical vias are introduced by etching holes into the silicon. The holes are then either coated or filled with a conductive material – open or closed TSVs are achieved, respectively. A major challenge concerns the time- and cost- efficient inspection of the TSVs (i.e., 10,000 TSVs/wafer) with respect to their quality. Therefore, the presented research:

1. Gains information about TSVs using the concept of Scanning Acoustic
   Microscope (SAM) interferometry.
2. Localizes and Classifies automatically the TSVs with high efficiency and
   accuracy by developing and applying two interconnected Convolutional Neural
   Networks (CNN).
3. Helps to understand the propagation of acoustic waves and its interaction
   with the TSVs using Elastodynamic Finite Integration (EFIT) based simulation.

Keywords: residual stress state; finite element modeling; railway axle; cold rolling; corrosion fatigue
Wheelset axles are among the most safety-critical components in the railway area. Mechanical surface treatments like the cold rolling process introduce residual compressive stresses at the surface of such components. Furthermore, it increases the fatigue strength even under the influence of corrosive media and thus the service life 1.
Experimental results on both laboratory specimens and real axles show how cold rolled components have an increased service life compared to not rolled components, even under the influence of corrosive media. Using calculations of the stress intensity factor of the residual stress state and the state under load in combination with the use of the cyclic R-curve, the growth of cracks at the surface and at defects can be predicted [2-3].
One of the difficulties in such calculations is the correct prediction of the residual stress distribution after cold rolling and the subsequent stress redistribution when defects such as notches or cracks appear, as for example shown in Figure 1. Already established calculation methods provide information about the level of stresses on the surface and their penetration depth, but cannot capture the actual distribution of residual stresses. For this reason, a method is needed that provides additional insight into the strain and stress distribution. Finite Element Analyses are the calculation method of choice for finding the application-specific ideal settings of the cold rolling parameters.
To predict the behavior of such critical components even in corrosive environments an integrated computational material and process model is needed that uses a combination of already established computational methods, experiments and simulations.

To cope with the ever-growing needs for high-performance materials, a strong ongoing trend exists worldwide towards installation of material development and processing platforms termed materials acceleration platforms (MAPs). A common element of these initiatives is the transition from stand-alone simulation or characterization tools towards integrated environments combining databases, physical modeling, inverse design, machine learning and experimental testing in a common framework. A core prerequisite for these platforms is the integrated computational materials engineering for the linking of multiple scales and models to establish process-microstructure-property relationships. Within the IC-MPPE-COMET project, such a platform is developed and the applicability to material design problems demonstrated for two different use cases.

The evolution from Digital Models to a Digital Shadow and a full Digital Twin is presented for a multi-physical lab-scale problem representing industrial manufacturing processes using induction hardening to produce components with defined material properties. The transfer to the industrial scale is a mid- to long-term process built on close cooperation between research and industry and must be specifically adapted to the individual manufacturing conditions. MCL contributes to the concept development and realization of Digital Twins of manufacturing processes in the industry. On industrial level, an example is shown for the forging of aerospace components at voestalpine Böhler Aerospace GmbH & Co KG

The aim of this work is to calculate local hydrogen concentrations in high-strength steel fasteners. For quantification, both, the trap densities and binding energies must be known. We present a process and microstructure sensitive model to quantify trap densities and corresponding binding energies to study of the distribution of hydrogen and local concentrations within the component. With this method products with increased hydrogen resistivity can be invented, e.g. by applying the so-called ”beneficial trapping” concept where energetically favorable ”traps” extract hydrogen from the interstitial lattice (mobile hydrogen) and consequently hinder its adverse effects.

Photonic quantum technologies relying on the generation and detection of light at the single-photon level play a central role in emerging quantum computing, quantum communication and quantum sensing applications. State-of-the-art experimental implementations most commonly place stringent requirements on the performance of the employed devices, components and systems, hence necessitating optimized material growth processes, nanofabrication technology, and system implementations. Here, recent progress within the Quantum Nano Photonics group at KTH Royal Institute of Technology will be summarized, in particular related to nanoscale materials development in the areas of single-photon generation and single-photon detection. Examples include atomic defects and quantum dots acting as sources for single and entangled photons, as well as superconducting nanowire devices employed as high-performance detectors for single photons in the visible and near-infrared wavelength ranges. Current trends in the respective fields will be outlined and emerging applications in quantum optics and beyond will be discussed.

Since 2013 zero-dimensional, spherical quantum dots (QDs) are used by the display industry for high-end devices with improved color gamut compared to non-QD LED TVs. In
contrast, the next generation of optoelectronic devices will most likely be based on quantum materials with superior core/shell structures like “Dot-in-GiantShell” (GS-QD), “Dot-in-Rod” (DR) or “Dot-in-Sheet” (DS) nanoheterostructures because of their unique optical properties like narrow band widths, high stability, increased brightness, huge absorption
coefficient in the blue, polarized emission and on/off switchability in an electric field. For producing these next generation particles Fraunhofer CAN has established a platform technology based on continuous flow to allow automating of the synthesis. This ensures the reproducibility and throughput needed to investigate their use in new optoelectronic applications. We will present how this technique is set-up and explain how the different geometries of the above mentioned particles are achieved.
In case of the Giant Shell particles we will show results for particles with diameters between 10 and 20 nm including TEM analysis, especially the influence of the diameter on quantum yield and absorption characteristics.
In case of the elongated Quantum Rods we will show results for green and red emitting particles with different
aspect ratios including TEM analysis and quantum yields. We will also show measurements proving the degree
of polarization (DOP) on a single particle level and the DOP results for aligned QR film on 1 cm2 electrodes.
In case of the Quantum Sheets we will present TEM and quantum yield measurements and compare their
properties to the 0D and 1D particles mentioned above.
We will also outline how these properties influence their use in certain applications to outline that the perfect QD for every application does not exist. The QD chosen has to fit to the reqirement of the application it should be used for.

This work presents a multi-stimuli responsive sensor for artificial skin applications. The sensor is responsive to surrounding changes in force, humidity and temperature. The developed design consists of a hydrogel core, responsive to temperature and humidity changes; and a piezoelectric shell for force detection. Swelling of the hydrogel core, in response to humidity and temperature, mechanically strains the piezoelectric shell and generates detectable electric charge. The two materials are combined into core-shell nanorod structures, using state-of-the-art vapor-based deposition techniques. These deposition techniques provide control over material’s mechanical, optical and electrical properties in addition to film’s conformity and uniformity. Moreover, the core-shell nanorods are deposited into a nanostructured UV-curable resin, which allows the fabrication of the embedded core-shell nanorods. 1. Piezoelectric zinc oxide is synthesized using plasma-enhanced atomic layer deposition (PE-ALD).In PE-ALD, substrate temperature defines the deposited film’s crystalline properties. A combination between (100) and (002) crystallographic orientations gives control over zinc oxide’s piezoelectric properties. In this work, piezoelectric zinc oxide layer with combined (100) and (002) preferential orientation is deposited at low temperatures, which is advantageous for when flexible substrates. 2. humidity and temperature responsive hydrogel, Poly-N-vinylcaprolactam (pNVCL), is synthesized using initiated chemical vapor deposition (iCVD). The dry vapor-phase technique gives control over the lower critical solution temperature (LCST), amongst other material properties. 3. The multi-stimuli responsive core-shell nanorods are deposited into nanostructured UV-curable polyurethane acrylate (PUA) resin, with nanoholes having a diameter d = 500 nm, height L = 500 nm and pitch = 1000 nm, achieved using UV nanoimprint lithography (UV-NIL).

Chemical sensors based on metal oxides (MOx), are one of the most promising gas sensing devices due to their high sensitivity to numerous gases, fast response, miniaturization, and simple production. The detection principle of these sensors is a conductivity change of the MOx sensing material due to chemical reactions of gases with surface molecules. Cross sensitivities and interference to humidity, however, are still significant drawbacks of these sensors.
Functionalization of MOx sensing films with catalytic nanoparticles (NP) is a highly promising technology for optimizing sensor performance. The huge variety of potential MOx - NP-combinations requires efficient screening technologies to find proper hybrid material mixtures which enable controlled adjustment of the sensor response to specific target gases. This is of high importance for realization of a multi gas sensor device capable for clear discrimination of single gas components out of a gas mixture.
In this work we introduce our approach for efficient screening of hybrid MOx – NP material combinations. We have developed a specific Si-platform chip along with a gas measurement setup which enables simultaneous characterization of 16 chemical sensor structures in parallel. The Si-chips feature an array of Ti/Pt electrodes for contacting ultrathin MOx sensing films, which are deposited by spray pyrolysis and structured by photolithography to a size of 50 x 100 µm$^2$. On these platform chips we tested three different MOx (SnO$_2$, ZnO, and CuO) before and after functionalization with mono- and bimetallic NPs (such as Au, Pt, Pd, and NiPt) towards several test gases (CO, HCmix, toluene, CO$_2$). Measurements are performed in a background gas of synthetic air at different relative humidity levels (25 – 75%) and at different operating temperatures up to 350°C. We will present the sensing performance results of various MOx-NP combinations exhibiting an optimized response to specific target gases.

A carbon-containing magnesia refractory product (MgO−C) as it is applied in steel ladles of the secondary steel metallurgy was investigated for the Mode I fracture behaviour. Usually, it is determined with the wedge splitting test to quantify the thermal shock behaviour, the materials' brittleness or to gain input parameters for thermomechanical Finite Element (FE) simulations. For high temperature testing of carbon-containing refractories, a gas purging, for example with argon, is required to avoid carbon oxidation in the sample. Laser speckle extensometers measure the displacement in a contactless manner. Based on the results obtained from the tests, fracture mechanical parameters such as the specific fracture energy and the nominal notch tensile strength were calculated. An inverse FE simulation yielded tensile strength, the total specific fracture energy, and the strain-softening behaviour. As refractories often show significant creep rates at application temperatures, the creep behaviour was also considered in the inverse evaluations. Finally, the contribution of creep to the determined fracture energy was quantified.

In power semiconductor applications, thin films are of utmost importance in terms of reliability. In order to design such reliable thin films, residual stresses determination and microstructural characterization plays a crucial role. Here, various advanced techniques like FE-EBSD, FESEM-TKD (Transmission Kikuchi diffraction) and high resolution angular EBSD (HR-EBSD) are carried out to perform microstructural characterization qualitatively as well as quantitatively. Particularly HR-EBSD provides novel insights with respect to the strain and stress distribution on the thin film surface as well as on the cross section. In particular knowledge about the evolution of residual stresses in the grain and along the grain boundaries are gained. The obtained HR-EBSD results are correlated with residual stress profile measurements performed along the thickness of the thin films. For the latter the ion layer removal method with a machine learning prediction model is used. The presented workflow shall help to design improved thin films for future industrial applications.

Acknowledgement:
The authors gratefully acknowledge the financial support under the scope of the COMET program within the K2 Center “Integrated Computational Material, Process and Product Engineering (IC-MPPE)” (Project No: P2.12 and P1.12). This program is supported by the Austrian Federal Ministries for Climate Action, Environment, Energy, Mobility, Innovation and Technology (BMK) and for Digital and Economic Affairs (BMDW), represented by the Austrian Research Promotion Agency (FFG), and the federal states of Styria, Upper Austria and Tyrol.

The principles of conventional residual stress analysis using XRD are explained and their further development to enable the determination of the residual stress depth profile are outlined. The actual investigations are focused on residual stress depth profiling concerning WC-Co hardmetals using X-ray diffraction techniques. Additionally, results for hardmetal samples coated with AlCrTiN layers are also presented. The measurements provide the residual stress depth profile near the substrate-coating interface. Besides, corresponding results from ceramic Si3N4 samples are shown. The grazing incidence X-ray diffraction method (GIXD) is an important tool for the determination of residual stress gradients near the material surface. Due to the relatively low penetration depth of X-rays in materials as investigated here, such methods are restricted to a region of a few microns near the surface. The application of this technique and its innovative modifications to obtain a sin2ψ plot at strictly constant mean penetration depth is presented. The equi-penetration grazing incidence X-ray diffraction (EP-GIXD) variant originates from a sin2ψ approach based on the determination of the lattice plane distances of various selected diffraction peaks at a specific angle of incidence. For each measurement dataset the condition ∂τ/∂ψ = 0 is here fulfilled and corresponds strictly to an individual mean penetration depth τ independent from the tilt angle ψ. The residual stress investigation of substrates near the substrate-coating interface is complicated by absorption effects in the coating layer. Rocking curve measurements i.e. the determination of the diffraction intensity as a function of the sample orientation, enable a quantification of the absorption properties of the coating. Finally, a mathematical formalism covering the transformation from the residual stress depth profile viewed as a function of the mean penetration depth to a profile as a function of the real depth is discussed.

An appealing idea to material scientists is to characterize the flow behavior of materials with minimal experimental effort while guaranteeing highly reliable results. Nanoindentation is one candidate technique to achieve this objective. Although established as standard method to extract hardness and Young’s modulus, the technique is not yet fully exploited regarding the determination of localized flow curves, since understanding the correlations between mechanical properties obtained by spherical indentation experiments and uniaxial data is extremely challenging. To correctly account for tip imperfections, a calibration procedures originating from fundamental geometrical considerations is applied. This sets the foundation for strain-rate controlled experiments and allows an experimental evaluation of the constraint factor in consideration of the mechanical properties and induced strain, which enables the extraction of reliable flow curves. These protocols are applied at room and elevated temperatures in order to demonstrate temperature effects on the work hardening behavior of different metal microstructures.

Condition Monitoring (i.e. estimating a physical system's current fault state from accessible observations (sensor readings)) can sigificantly lower a systems' operating cost by facilitating longer service intervals and shorter service interventions. Model-Based Condition Monitoring (MBCM) uses inverse models to infer the current fault state. Constructing accurate, robust and compact inverse models is a daunting task, though. Hybrid semiparametric modelling of physical systems (eg combining knowledge and observation) is a viable option to deal with these conflicting design goals.
This talk will look back at findings from past research activities as well as forward to the new prospects of probabilistic hybrid models.
Specifically, I will detail MCL's MBCM workflow collecting insights across multiple projects and industries including software tools developed and further ones planned. I will contrast classical block-based hybrid models with more flexible approaches arising from the rich field of physics-informed machine learning. Finally I will motivate the future focus on probabilistic hybrid models.

Induction hardening involves multiple physical problems from the electromagnetic, thermal, mechanical and metallurgical domain. These highly nonlinear coupled problems are usually solved by applying finite element methods (FEM) thanks to the available physics-based knowledge in the form of partial differential equations (PDEs). However, the material properties are coupled to the PDEs and they need to be known beforehand for an accurate FEM.

To estimate the thermo-physical material data, we inform a learning machine about the underlying physics described by known PDEs, a procedure called hybrid modeling. A learning machine based on artificial neural networks (ANNs) is employed to estimate the temperature response of an inductively heated sample. ANNs are the tool of choice when it comes to finding nonlinear relations between process parameters and the material response. The ANN is trained using measured temperature data T(r,t) (in time and space) while respecting PDEs describing the coupled electromagnetic-thermal problems with initial conditions (ICs) and boundary conditions (BCs). The thermo-physical properties, in this work thermal conductivity and specific heat, are considered as unknowns, which are embedded in the PDEs. The unknown material data are temperature dependent and are estimated by ANNs in an optimization problem. The optimization problem involves minimizing an objective function, which contains cost terms including temperature data fit loss, PDE residual loss, initial and boundary condition fit losses. Once the ANNs informed PDEs are optimized the material data incorporated in the PDEs are easily extracted by the nonlinear functions, which are approximated by optimized ANNs. We test and verify the approach for different samples (magnetic and non-magnetic) whose material data are already known (partially or totally) with high accuracy.

The manufacturing industries have already started to incorporate data driven and hybrid techniques into their associated material processes where utilization of physical and data-driven technologies along with Machine Learning (ML) and smart data handling have recently gained a tremendous momentum. Optimizing process controls through digital twining and improving operations and maintenances along with minimization of overheads and deficiencies are among main goals of the process digitalization drive. Hybrid physical-data driven modelling with its associated ML schemes along with Reduced Order Modelling (ROM) techniques are among the new trends in design and monitoring of light weight material processes. In this research work, an effective use of hybrid ROM scheme for predictive modelling of material processes have been investigated and its agility of dealing with optimization of process parameters and part performance has briefly been reviewed. Further attempts have also been made to employ the hybrid physical-data driven modelling for the real-time optimization of the process parameters using fast predictive-corrective models. With the introduction of these hybrid models, the combination of experimental, numerical simulation and mined data sources have been considered to setup proper semantic databases for data processing and filtering. Furthermore, the combination of hybrid models and ROM techniques along with ML modules have been employed for selected cases of material processes as pilot studies.

This lecture will highlight various sensors and sensor applications for wired and wireless CbM solutions Analog Devices provides in vibration monitoring. It will inform about trends and some practical use-cases our offerings fit into.

The present work investigates the technical feasibility of a condition monitoring setup aiming at the detection of RCF cracks, so called head checks, in railway rails via a wayside (i.e. stationary) setup using surface acoustic waves (SAW) as detection principle. The experimental SAW setup consists of a pitch-catch setup using piezo transducers equipped with comb adaptors to excite and measure narrowband Rayleigh waves with a center frequency of 1 MHz. SAW experiments were performed on a rail subjected to cyclic loading in a 1:1 wheel-rail test rig yielding the specific rolling contact fatigue, i.e. head checks. Elastic finite integration (EFIT) simulations were performed to analyze the surface and bulk wave propagation in the rail and to predict the signals at specific receiver positions. SAW transmission and reflection scenarios at cracks were analyzed numerically via modelled variations of gauge corner crack configurations in crack number (0, 1, 2 and 3) and depth (0, 0.5mm and 1mm). Both the numerical and the experimental results show a clear correlation between the appearance and intensity of the head check damage and the Rayleigh wave attenuation in transition mode.

....

The European industry needs to keep competitive advantage by addressing the needs for Industry 4.0 and be prepared for 5.0.

Already finalized European project Delphi4LED proposed new methods and standards to generate a first of a kind multi-domain LED digital twins. Compared to traditional approach, the newly developed LED digital models contributes to reduce the design phase thanks to faster computation at equivalent accuracy.

AI-TWILIGHT will generate prescriptive digital twins for the lighting industry. The project is addressing multiple application area: automotive, horticulture, streetlighting, general lighting. The ambition is to cover a large part of the lifecycle of a LED-based system (including the electronic driver): from product design flow to infrastructure development and maintenance.

The key technical and exploitation objectives of the AI-TWILIGHT consortium are:

To create digital twins of LED light-sources and electronics (driver)
To create self-learning models using AI and analytics techniques
To facilitate the implementation of the digital twins in digitalized design flow (for SSL product design) and facilitate their applications upstream, up to digital twins of lighting systems of large infrastructures (e.g. for building design).
To implement the AI-TWILIGHT methods, models and tools within consortium partners to harvest its benefits

AI-TWILIGHT LED application design flows will be combined with the CPS/IoT approaches and reliability aspects, supported by AI techniques. A self-learning digital twin will be result of the project.

Remaining useful life (RuL) model for electronic systems, e.g. for power electronics, deals with the prediction of possible service life and failures under different types of operating conditions. It is almost impossible to precisely predict future events; therefore, it is necessary to account for the different sources of uncertainty that affect lifetime estimations. Beside the RuL modelling aspect and the need of an accurate data base, the computation of the remaining useful life is meaningful in the context of condition-based monitoring and needs to be approached as an uncertainty propagation problem. The first impact factor is the aleatory uncertainty from inherently-variable inputs of the physical system and its environment such as material properties, process parameters and service parameter. The second impact factor is the epistemic uncertainty from a lack of knowledge or incomplete information about the system and its environment in service. Both uncertainties must be considered to provide the most accurate assessment of the RuL model. The work discusses the different sources of uncertainty and their usability for RuL modelling of power electronics systems.

Energy efficiency and saving in microelectronic devices goes along with thermal management, as their failure rate increases exponentially and their efficiency decreases linearly with the operating temperature. Due to the continuous trend of miniaturization, thermal management additionally gains importance and leads to an increase in device packing density. This triggers the need for heat dissipation and the development of heat dissipation strategies, which requires knowledge of the temperature-dependent thermophysical and structural properties.
Here, a closer look was taken at nm-thin $\text{Nb}_2\text{O}_5$ films, predominantly used in devices for optical, electrochromic, and sensing applications. Due to their geometrical restrictions, the thermal conductivity of such nanometre thin films can differ from their bulk counterpart, and it is strongly affected by their processing. E.g. depending on the applied substrate temperature, the $\text{Nb}_2\text{O}_5$ film grows either in the amorphous or crystalline form. Within this work, the thermal transport properties and the crystallographic structure of the $\text{Nb}_2\text{O}_5$ films were interrelated. The thermal conductivity of both amorphous and crystalline $\text{Nb}_2\text{O}_5$ was determined by the Time Domain Thermoreflectance. The average and local structure are determined by in-situ high-temperature X-ray diffraction and in-situ high-temperature Raman spectroscopy, respectively. The crystalline $\text{Nb}_2\text{O}_5$ film showed a decrease in thermal conductivity with increasing temperature. In contrast, the amorphous $\text{Nb}_2\text{O}_5$ film had a constant thermal conductivity of 2.2 ± 0.2 W/mK at temperatures below 275°C. Above that temperature, an abrupt increase in thermal conductivity to 2.8 ± 0.2 W/mK was recorded. This substantial increase in thermal conductivity cannot be linked to any macroscopic phase change but rather to a local phase rearrangement near the crystallization temperature evidenced by the Raman spectra analysis. This study serves as a guide to engineer future $\text{Nb}_2\text{O}_5$ based thin film devices and for their reliability and efficiency optimization.

The development of drugs/vaccines is very expensive and comes with huge failure rates, financial risks. For companies to be successful portfolios need to balance financial risks, investments and potential revenues. Recent years have seen big pharmaceutical companies focusing on specific diseases or conditions, potentially as a mean to understand the financial risks and returns better. In this context the role of process development is to deliver robust processes in short timelines and at limited technical risk. The Quality by Design (QbD) paradigm is seen as a measure to keep technical risks in check, as it helps to systematically explore the process knowledge and improve the understanding. However, QbD guided bioprocess development is time effective only if knowledge is transferred from one project to the next (horizontal knowledge transfer), one scale to the other (vertical knowledge transfer). Today knowledge is transferred across scales and projects when assessing the technical risks. However, this form of knowledge transfer is limited and despite the wide spread of platform processes, to some degree the process needs to be developed de novo for every new drug/vaccine candidate. This leaves a huge potential to accelerate process development.

In this contribution, we show how advanced machine-learning and hybrid modeling approaches can be exploited to transfer knowledge between scales and projects. In particular, we present a novel embedding approach that allows for transversal data analysis across process runs with different cell-lines and products. We showcase the added value of the approach for an upstream bioprocess development case, concluding that process development could be more efficient and multiple times faster.

Condition monitoring of industrial milling tools is of vital importance in industrial production environments when trying to improve those processes that see ever-increasing demands in terms of productivity and process safety. Diagnosis of the damage state of end milling tools in real-time is crucial for the quality control of the produced workpieces. In the model-based condition monitoring approach presented here, the tool degradation state is monitored by tracking coefficients of a mechanistic model for the moments determined during the milling operation by an instrumented tool holder. The presented monitoring approach is based on two metrics derived from the goodness-of-fit provided by the milling mechanistic model for the observed sensor date. Cumulative sum control charts are constructed to monitor changes occurring to the tool condition during the milling tool’s operation. Compared to previously suggested methods based solely on the values of force model coefficients, our approach facilitates a preciser and less conservative identification of the point of transition from homogeneous wear to onset of breakout formation at the cutting edges. The proposed strategy is suitable for real-time monitoring of the milling tool’s damage condition based on a mechanistic model.

Searching, finding and annotating patterns in time series data has become even more important with the advent of machine learning systems, which rely on massive amounts of training data to analyse, model and predict system behaviour. We present TimeFuse, a set of algorithms integrated in a software suite which enables engineers to find patterns in time series data based on selected similar data or sketched signal shapes. TimeFuse facilitates rapid identification and annotation of similar patterns in time series data. We employ an information retrieval approach based on the SAX algorithm to identify signal ranges similar to a specified search pattern with a strong emphasis on recall over precision. A streamlined user interface enables users to rapidly select, group and annotate identified clusters of similar signals. With TimeFuse, users can quickly prepare training data for modelling purposes based on sketches of their understanding of expected signal shapes or based on known patterns in historical data.

The voestalpine Group’s High Performance Metals Division focuses on producing and processing high-performance materials and customer-specific services such as heat treatment, high-tech surface treatments, and additive manufacturing processes—increasingly using digital technologies. Thanks to its unique sales and service network, the division offers its customers material availability and processing as well as local points of contact at about 140 sites around the world. Sawing machines are therefore a widespread asset type in the High Performance Metals division and offer high potential to generate value by date-driven optimization of operations. voestalpine High Performance Metals DIGITAL SOLUTIONS GmbH (subsidiary since 2021) built an IIoT based solution to increase transparency for sawing operations called HPMconnect. It uses a unique architecture in combination with cloud services and an IIoT platform to manage the system and leverage the created information. An edge device collects data from several machines, aggregates it and stores it securely in the chosen database – cloud or on-premise solutions are possible.
Legacy assets are equipped with a retrofit sensor box, measuring current and vibration to determine the actual state. All components are industry-proven automation standard from the DIGITAL SOLUTIONS IIoT development partner and global player in the automation control market, enabling competitive pricing and reliable service.
HPMconnect is productive at several locations in Europe and at the first location in Asia. Runtime data builds the foundation for further use-cases e.g. advanced analytics, tool performance, energy management and predictive maintenance. Designed as a platform, container technology is used to extend its functionality and value over time.
A very active community around HPMconnect contains current and future users, lean managers, production managers etc. Most members have operations background and the necessary know-how to shape future development to an increased value-add of the whole solution.