Brain of Materials ist eine Plattform für innovative und smarte Materialien, die Designern, Entwicklern und Ingenieuren diverser Branchen dabei hilft, ihre Produktentstehungsprozesse zu optimieren und zu beschleunigen.

Due to coming emission legislations, the temperature is expected to increase in heavy-duty diesel engines, specifically in the hot-end of the exhaust system affecting components, such as exhaust- and turbo manifolds. Since the current material in the turbo manifold, a ductile cast iron named SiMo51, is operating close to its limits there is a need for material development in order to maintain a high durability of these components. When designing for increased life, many material properties need to be considered, for example, creep-, corrosion- and fatigue resistance. Among these, the present work focuses on the latter two up to 800°C improving the current material by additions of Cr, for corrosion resistance, and Ni, for mechanical properties. The results show improved high-temperature corrosion resistance in air from 0.5 and 1wt% Cr additions resulting in improved barrier layer at the oxide/metal interface. However, during oxidation in exhaust-gases, which is a much more demanding environment compared to air, such improvement could not be observed. Addition of 1wt% Ni was found to increase the fatigue life up to 250°C, resulting from solution strengthening of the ferritic matrix. However, Ni was also found to increase the oxidation rates, as no continuous SiO2-barrier layers were formed in the presence of Ni. Since none of the tested alloys showed improved material properties in exhaust gases at high temperature, it is suggested that the way of improving performance of exhaust manifolds is to move towards austenitic ductile cast irons or cast stainless steels. One alloy showing good high-temperature oxidation properties in exhaust atmospheres is an austenitic cast stainless steel named HK30. This alloy formed adherent oxide scales during oxidation at 900°C in gas mixtures of 5%O2-10%H2O-85%N2 and 5%CO2-10%H2O-85%N2 and in air. In the two latter atmospheres, compact scales of (Cr, Mn)-spinel and Cr2O3 were formed whereas in the atmosphere containing 5%O2 and 10%H2O, the scales were more porous due to increased Fe-oxide formation. Despite the formation of a protective, i.e. compact and adherent, oxide scale on HK30, exposure to exhaust-gas condensate showed a detrimental effect in form of oxide spallation and metal release. Thus, proving the importance of taking exhaust-gas condensation, which may occur during cold-start or upon cooling of the engine, into account when selecting a new material for exhaust manifolds. / <p>QC 20130508</p>

This report evaluates the possibility of using sequential learning in a material development setting to help predict material properties and speed up the development of new materials. To do this a Random forest model was built incorporating carefully calibrated prediction uncertainty estimates. The idea behind the model is to use the few data points available in this field and leverage that data to build a better representation of the input-output space as each experiment is performed. Having both predictions and uncertainties to evaluate, several different strategies were developed to investigate performance. Promising results regarding feasibility and potential cost-cutting were found using these strategies. It was found that within a specific performance region of the output space, the mean difference in alloying component price between the cheapest and most expensive material could be as high as 100 %. Also, the model performed fast extrapolation to previously unknown output regions, meaning new, differently performing materials could be found even with very poor initial data. / I denna rapport utvärderas möjligheten att använda sekventiell maskininlärning inom materialutveckling för att kunna prediktera materials egenskaper och därigenom förkorta materialutvecklingsprocessen. För att göra detta byggdes en Random forest regressionsmodell som även innehöll en uppskattning av prediktionsosäkerheten. Tanken bakom modellen är att använda de relativt få datapunkter som generellt brukar vara tillgängliga inom materialvetenskap, och med hjälp av dessa bygga en bättre representation av input-output-rummet genom varje experiment som genomförs. Med både förutsägelser och osäkerheter att utvärdera utvecklades flera olika strategier för att undersöka prestanda för de olika kandidatmaterialen. Genom att använda dessa strategier kunde lovande resultat vad gäller genomförbarhet och potentiell kostnadsbesparing hittas. Det visade sig att, för specifika prestandakrav, den genomsnittliga skillnaden i pris mellan den billigaste och den dyraste materialkemin kan vara så hög som 100 %. Vad gäller övriga resultat klarade modellen av att snabbt extrapolera initial data till tidigare okända regioner av output-rummet. Detta innebär att nya material med ny typ av prestanda kunde hittas även med mycket missanpassad initial träningsdata.

This report evaluates the possibility of using sequential learning in a material development setting to help predict material properties and speed up the development of new materials. To do this a Random forest model was built incorporating carefully calibrated prediction uncertainty estimates. The idea behind the model is to use the few data points available in this field and leverage that data to build a better representation of the input-output space as each experiment is performed. Having both predictions and uncertainties to evaluate, several different strategies were developed to investigate performance. Promising results regarding feasibility and potential cost-cutting were found using these strategies. It was found that within a specific performance region of the output space, the mean difference in alloying component price between the cheapest and most expensive material could be as high as 100 %. Also, the model performed fast extrapolation to previously unknown output regions, meaning new, differently performing materials could be found even with very poor initial data. / I denna rapport utvärderas möjligheten att använda sekventiell maskininlärning inom materialutveckling för att kunna prediktera materials egenskaper och därigenom förkorta materialutvecklingsprocessen. För att göra detta byggdes en Random forest regressionsmodell som även innehöll en uppskattning av prediktionsosäkerheten. Tanken bakom modellen är att använda de relativt få datapunkter som generellt brukar vara tillgängliga inom materialvetenskap, och med hjälp av dessa bygga en bättre representation av input-output-rummet genom varje experiment som genomförs. Med både förutsägelser och osäkerheter att utvärdera utvecklades flera olika strategier för att undersöka prestanda för de olika kandidatmaterialen. Genom att använda dessa strategier kunde lovande resultat vad gäller genomförbarhet och potentiell kostnadsbesparing hittas. Det visade sig att, för specifika prestandakrav, den genomsnittliga skillnaden i pris mellan den billigaste och den dyraste materialkemin kan vara så hög som 100 %. Vad gäller övriga resultat klarade modellen av att snabbt extrapolera initial data till tidigare okända regioner av output-rummet. Detta innebär att nya material med ny typ av prestanda kunde hittas även med mycket missanpassad initial träningsdata.

Egg consumption in Sweden is very high, which means a lot of egg shells are thrown out. For example the average swede consumes about 350 eggs every year. In this thesis I have investigated how eggshells can be applied in new areas as a resource, and I want to inspire the reader and encourage them to rethink what is called waste. Is it really waste or a possible resource? The work began with literature search, interviews and a survey to understand people's views on eggshells. I have concluded that not a lot of people know what eggshells are made of, and what they are used for. In a literature search I came across many scientific articles on eggshells. According to King’ori (2011) eggshells are used as fertilizers, used in medicine, cosmetic surgery, and dental care, calcium supplements, in the food industry and in crafts. When eggshell is thrown as food waste it creates problems in recycling stations, the eggshell is like sediment and lies at the bottom of tanks and pipes. The recycling staff have to perform additional work when cleaning tanks and pipes as eggshells are like sediment and lay at the bottom. The staff have to manually shoot out eggshells with many other useful materials that could have been digested and become biogas. This is an exploratory thesis where I investigate the material eggshell through a material-driven design process to find new sustainable uses. I follow the material-driven design process method by Karana et al.( 2015) plus my own added steps during the process. Material driven design process is a process where designers use a material to find the attributes of the material. These are then presented in new contexts or already existing contexts, this means that the material on the existing product can be replaced. This differs from traditional design, where the material is chosen based on wanted attributes. This report will give you an understanding of the material-driven design process-method and how I have carried out a material-driven design process, to develop different material properties of eggshells. The report ends with a concluding discussion and presents new eggshell-based materials that I have produced in the form of a demonstrator. It is a book which has two different materials, a hard material and a paper material which are combined and put together with a steel thread. I will also share my knowledge about eggshells and what potential it has in the future.

Electron beam powder bed fusion (PBF-EB) is an additivemanufacturing (AM) method based on layer-by-layer melting of apowder bed. The technology is industrialized in certain applicationsbut still considered as immature and is not as widely used as laserbeam-based systems (PBF-LB). PBF-EB can offer several benefits overPBF-LB such as process cleanliness, thermal efficiency, fast beam speed,higher power and energy transfer, low residual stresses in built partsand a good signal environment for process monitoring. This can beadded on top of the general benefits of AM such as geometricalfreedom, manufacturing efficiency, easy design revisions, short leadtimes and so on. This suggests that PBF-EB holds potential as atechnology for the sustainable production of materials andcomponents. This thesis investigates how PBF-EB can be furtherdeveloped to create new and unique materials features. This isachieved by introducing innovative methods for material processingand by further developing the PBF-EB process itself. The thesisintroduces a charge-free heating method for PBF-EB and the resultssuggest an enhanced processability of difficult-to-process materialsand powders. A method for building multi-materials in PBF-EB isintroduced and demonstrated by the manufacturing of direct andlamellar transitions between different alloys. Methods for processmonitoring and powder bed resistivity evaluation are proposed andxiidemonstrated. It is concluded that the results presented in this thesisenabled new PBF-EB processing modes, increased the knowledge ofthe process, and introduced a new material group by demonstratingthat ceramics can be processed at high temperatures (~1600C). / <p>Vid tidpunkten för framläggningen av avhandlingen var följande delarbeten opublicerade: delarbete 2 och delarbete 4 (inskickat).</p><p>At the time of the defence the following papers were unpublished: paper 2 and paper 4 (submitted).</p>

Increasing demand for structural health monitoring in space highlights the need to make the creation of these systems more accessible. This study investigates the potential of additive manufacturing to achieve this goal by characterizing a self-sensing material made of a commercially available 3D-printed continuous carbon fibre filament. The results demonstrate the feasibility of converting the filament into a strain sensor with improved sensitivity compared to conventional foil strain gauges. Mechanical and electromechanical properties of the self-sensing material were characterized, including an ultimate tensile strength of 45.09 ± 3.45 MPa, a failure strain of 38.93 ± 3.41%, and a base resistance of 759.11Ω. The tensile gauge factor was calculated to be 467.06 ± 375.90 within the strain range of 0% to 3.8% with a linearity (R2) of 0.93. For the first time, a systematic literature review compares mechanical and electromechanical properties to enable material selection for mechanical design incorporating self-sensing material. The study highlights that the spread of material properties in a group of materials indicates how well-developed a material is for self-sensing purposes. This study advances our understanding of the feasibility of using additive manufacturing to create self-sensing materials for structural health monitoring systems and opens up new avenues for further research.

<p dir="ltr">Vocabulary is a crucial element in second language learning. However, researchers in vocabulary acquisition express concerns about students’ successful acquisition of vocabulary (e.g., no significant gain after one semester of instruction (Clark & Ishida, 2005)) and the limited classroom instruction dedicated to vocabulary. In an effort to address these issues, the present study developed an online vocabulary learning system intended for use as homework, incorporating relevant theories, hypotheses, and empirical findings from existing literature and investigated its effectiveness employing a mixed-methods design.</p><p dir="ltr">For the quantitative component, students’ vocabulary gains were measured across three aspects of vocabulary knowledge (breadth/size, depth, and speed of access). A three-week experiment was conducted with students enrolled in the third-semester Japanese language course at a US Midwest institution. The final dataset included 54 students’ data. The experimental group (<i>n</i> = 28) utilized the new system to learn target words, while the control group (<i>n</i> = 26) used the current system employed in the course. The current system is also operated online and includes two types of exercises (i.e., listen-and-repeat and flashcards). ANCOVAs were employed to identify any significant differences between the groups, controlling for their pretest scores. Additionally, regression analyses were conducted to explore the relationship between the time the experimental group students spent learning new words using the new system and their outcomes, while also controlling for their pretest scores.</p><p dir="ltr">For the qualitative component, eight students from the same participant pool as the quantitative component participated in one-hour focus group discussions, conducted separately for the experimental and control groups.</p><p dir="ltr">The quantitative analysis revealed no significant differences between the groups; however, it was found that the time spent by the experimental group learning new words using the system significantly predicted two aspects of vocabulary knowledge. The qualitative data offered insights into potential explanations for the lack of significant differences between the groups, including the influence of students’ motivation on the experiment and the perceived difficulty level of the vocabulary exercises implemented in the new system. Based on the results of the present study, numerous suggestions are made for future development projects of similar systems and research.</p>