This research has aimed to investigate the possibility of applying a neural network algorithm into the structural design process of bascule bridge leaves, by creating a workflow in Grasshopper. The demand for this tool, originates from the fact that the current design process is experienced as linear and slow, and does not fit the dynamic design environment within Amsterdam.

The foundation for the generative design workflow, is a parametric model for an orthotropic bridge deck made out of steel. The model is made out of main beams, which are modeled as tubular profiles, cross beams and rib elements, which are modeled as line elements, and a bridge deck plate on top, to which all elements are welded. The parametric model is controlled by a total of 24 parameters which describe its dimensions, profiles and spacing of elements. Based on the user input for three parameters, namely the bridge deck length, bridge deck width and distance to the rotational axis, the neural network algorithm does a suggestion for the remaining parameters, based on information in the bridge database. The bridge database consists of 35 bascule bridges. A design for a steel bridge leaf is then generated. In the next step, the generated structure is forwarded into SCIA Engineer via a.xml file, for structural analysis. Within SCIA Engineer, the structural performance is assessed based on occurring Von Mises stresses under defined load combinations.

In the validation stage of the research, three steps were undertaken. In the first validation step, the predicting behavior of the neural network was optimized based on the mean squared error. The predicting behavior of the workflow was assessed using 5-fold cross validation. It appeared that the neural network had a prediction accuracy of 25.91%. Therefore, in the second validation step, the neural network’s complexity was reduced. In the improved model, the neural network predicts fourteen parameters. The total predicting accuracy of the improved model is equal to 61.07%.In the last validation step, five random cases were generated, of which their SCIA model output was compared to simplified models and the predicting behavior of the neural network was assessed. In two out of five cases, the neural network immediately suggests a good structure, while in two others, only one parameter had to be altered to create a viable structure. From validation of the SCIA models, it appeared that the moment, shear force and stress distribution for both the main beam and crossbeam showed consistent behavior through all five models. For the main beams, a simple MatrixFrame model was constructed for comparison. There were significant differences in the magnitude of occurring moments and shear forces between the MatrixFrame model and SCIA model. The magnitudes of moments, shear forces and stresses in the cross beams were compared to a hand calculation of a clamped beam. It was concluded that the SCIA models generated in the validation stage of the research, behaved correctly.

In the results section, the output of the workflow was compared to two reference cases, the Berlagebrug and Elizabeth Admiraalbrug. For the Berlagebrug, the SCIA model had such a significant error that no useful results could be extracted from this case study. For the Elizabeth Admiraalbrug, the workflow generated a structure which was approximately twice as large in steel mass. This has resulted in lower unity checks compared to the original design. A qualitative comparison between the current design process and when the created workflow would be implemented, was also done.

Based on the quantitative results obtained in the research, it was concluded that the neural network algorithm in the application developed in its current form, will not significantly improve the structural design process, due to a lack of consistency in generated results.

Wind turbines are decommissioned when they reach their endoflife. The reason for reaching endoflife varies per wind turbine, but they all have in common that the wind turbine blades can not be recycled and thus end up as waste. Since a growing amount of wind turbines will reach their endoflife the coming decade, the waste problem will increase. To solve or limit this problem, the lifetime of a blade can be extended, or a sustainable endoflife treatment can be established. This is currently not available yet. This research focuses on repurposing wind turbine blades. The objective is not to find one application, but rather to propose a method that enables large scale processing of wind turbine blades, for applications in the construction industry. Since to date it is uncommon to repurpose wind turbine blades, Two processes are proposed to give guidance on how to repurpose wind turbine blades. The first process is meant to repurpose as many blades as possible, by sawing the blade in plates and beams that can be applied to construction projects. The second process gives guidance to a project team that wants to incorporate wind turbine blades in their construction project, which can raise awareness for the waste problem. Both processess are supported by a grasshopper script. This script enables the user to either find an efficient mapping of elements that can be extracted from the blade, or gives design freedom depending on the aim of the repurpose project. The first process is further worked out. A geometry model of an actual wind turbine blade is analyzed and categorized in panels with a homogeneous material layup. On this panels, a nesting pattern of plates and beams, as commonly used in the construction industry, is projected which can later be sawn into relative flat plates and beams. With this, almost half of the blade panels can be turned into useful construction elements. The cross section of a blade consists of various functions, and thus various material structures. Three categories are distinguished: Spar caps, shell panels and the shear web. For each of these categories, the material layup and properties are analysed, which has led to an advice on possible applications. Spar caps are strong, stiff and have a homogeneous cross section. They can be repurposed as structural beams. Shell panels are light and stiff and can be repurposed as partition walls, formwork or flooring. The shear web is light, stiff, and entails good insulation properties. This can be repurposed as building cladding. Although this research does not contain sample tests, the material properties of the blade are provided and are of significant value, despite their 27 years of service.

The concept of sustainability takes into account a problem of urbanisation which causes a phenomenon of urban sprawl. In order to fight it, the strategy of urban densification is introduced and building upward extensions is one of the methods. However, investigating the feasibility of placing an extension on top requires a significant amount of time and effort from structural engineers. Therefore, this research aims to explore computational tools that would simplify the initial exploration process and make the option of extension more appealing. Theoretical background part of the thesis analyses conditions and limitations for building reuse, upward extension and use of timber. The main impediments that could prevent realisation of such project are found. The practical part of the research aims to explore parametric modelling and machine learning (ML) in order to produce a proof of concept ML tool which predicts how much an existing building can be extended based on its parameters. In order to have a sufficient amount of data, parametric modelling is used. For the first step, information is collected about seven realised upward extension projects in the Netherlands. This data defines the object for the parametric model which is built using Rhino plug-in Grasshopper. One part of the model analyses the existing structures with the goal to establish structural reserve while another part examines extensions. Six parameters defining existing buildings are chosen for iteration, by changing these parameters, a database of 8 652 fictional projects is generated. For the ML, the data is analysed and used for supervised learning where regression task is performed. Multiple linear regression (MLR) and nonlinear regression (NLR) algorithms are applied to make predictions which are validated using 5-fold cross validation (CV) and evaluated by calculating errors. Finally, the research presents a possibility and an example of using ML for extension exploration. The use of ML model during the initial stage of upward extension projects seems very appealing due to the quick feedback with limited information provided. A useful database of realised projects with a selection of structural parameters together with the incorporation of non-structural ones and a customised ML model should provide accurate predictions. However, the current problem of lack of data about realised extension projects needs to be solved.

Although the concept of a submerged floating tunnel (SFT) originates from the early 1900s, the cross-section is always assumed to be circular or rectangular. In this research, it is investigated whether this assumption is valid. In order to find the optimal cross-section, the optimization process is split into two targets. The first targets aims for an optimization of material use considering the large hydrostatic water pressure. The second target guarantees a tensile force in the tethers that support the SFT, while keeping the buoyancy weight ratio (BWR) as close to 1.0 as possible. This is beneficial for both the tunnel tube and the tether system. The optimal cross-section considering the optimization of material use is an 'egg-shape' or an oval which deviates slightly from a circle. This is modelled by the form-finding process as executed in the Grasshopper software. This is due to the relative pressure difference between SFT-top and SFT-bottom. The ovalization decreases for increasing depth. The second target results in an ellipsoidal shape, with a flat bottom and convexly shaped top (in case the SFT is anchored to the seabed). This shape reduces the drag and turbulence on the SFT, while a lift force is generated contributing to the tensile force in the tethers. An analogy with an aeroplane wing is made. The magnitude of the lift and drag force is evaluated by solving for the Panel Method in Python. A significant reduction in drag and turbulence becomes visible. Moreover, some significant lift forces are generated by the encountering flow. The wave load must be compensated by the BWR to assure a tensile force in the tethers. All together, two types of solutions are presented. The ideal solution would be an inner (concrete) tube which absorbs the water pressure with a steel exoskeleton to reduce the drag and generate a lift force. The exoskeleton has a flat bottom and a convexly shaped top (when anchored to the seabed) A compromise between the targets is also possible and presented within the research. This compromise also has a flatter bottom and a more convexly shaped top, but still has some appearances of a circle.

This graduation thesis comprises a feasibility study and design of a floating classroom which can offer students a safe learning environment in typhoon and flood-prone areas by designing a classroom which is typhoon resilient with a floating foundation. This research is a follow-up study to the Finch Floating Homes project. One month of field research has provided more insight into the local problems in the schools, climate, materials and wishes from the school. These insights in combination with a literature study on school design, typhoon resiliency and indoor comfort are the basic principles for the design. Additional parameters are set up by analyzing the Finch Floating Home and the Makoko floating school and the design principles of the Finch’s buildings. In the design framework, there is a design analysis in which the choices for the design are explained followed by a structural analysis in which the design is tested for its resistance to high wind speeds that can occur with a typhoon and for its buoyancy by using the calculation software Technosoft and Axisvm. Also, passive design strategies which are implemented in the design are explained to make the classroom indoor comfortable. The result is a design of a floating classroom that is able to offer the students a safe and comfortable learning environment which is ready to be built.

Summary The objective of this project was to illustrate the value of nature-based circularity by illustrating its potential in creating composed and engineered circular products at large, and that of Mycelium-based sandwich panels as perfect embodiment of such product in specific. To effectively explore the feasibility of these panels as a natural and circular alternative, an industry-oriented product development approach was used. Therefore, this research focussed on parameters thought to be crucial in generating interest from industry players. Within the limitation of the research these were the structural behaviour and scaled production of the product. In an extensive analysis of this novel material both general characteristics as exact properties of Mycelium-Based Composites (MBCs) were explored. This project succeeded in getting a ‘feel’ for it and has applied that deeper understanding in tackling the further steps in the research. A potentially promising industry niche was found in the distribution centre construction sector and its dynamics were studied. It was found that this niche sector is extremely sustainability-savvy, accounting for a total share of 16% of all BREEAM certification issued in 2018. While at the same time being responsible for 16% of the nation rigid foam insulation demand. This large quantity simply applied fossil-fuel derived foam was identified as ideal low bearing, ‘positive impact’, fruit. Through interviews with industry experts and an industry analysis several quantitative and qualitative boundary conditions was collected that formed a foundation for the remaining research. It was decided to perform a number of 4-point bend test, exploring the fungal sandwich panel’s potential to serve as roof plate. Both lose MBC as sandwich panel samples were tested. Based on the data provide by the experiments could be concluded both the properties of the material and the product are currently insufficient for the envisioned application. However, the found result were of an order of magnitude interring enough for further research. The MBC is approximately two to three times weaker than traditional foams. Although, the sandwich samples showed ten times better behaviour its full potential was far from realised. Weak adhesion of the individual components and the following premature delamination were identified as the main cause for this underperformance.Summary | 7 Further development is needed on a material, product and application level to achieve applicability, suggestions of potential fruitful development directions have been made. The order of magnitude of the results is however of such comparable quantity that, especially less demanding, employments of the product can on a short-term be accomplished. It is therefore interesting to see that this research has shown how the production process of mycelium-based sandwich panels could look like. By converting button mushroom facilities, a batch-based production system could be put into place with a scale adequate for the demands from the distribution centre construction industry. Twelve of the envisioned facilities could produce all insulation material for the complete sector. In this explorative project it has been shown that this nature-based technology holds potential. Sandwich panels can be made with a certain quality and quantity. But although current product performance is not sufficient for the studied application other widespread uses can be foreseen. So, although further research is certainly needed in the development of composed mycelium-based circular products it can be concluded that based on qualitative and quantitative benchmarks these products can and likely will serve as a sustainable and circular alternative to traditional, oil-based, rigid insulation products.

The energy transition is a hot topic, businesses also need to think about manners to transition to sustainable energy. Within this graduation the- sis, a stakeholder is involved named the ‘Nederlandse Spoorwegen’, in short NS. The stakeholder has the ambition to make their station’s en- ergy neutral. To achieve this, electricity generating assets needs to be implemented on every asset they own. One of these assets is the P+R parking plots near stations. An easy way of implementing solar into the parking plots is by integrating the solar panels into a carport structure. Current solar carport designs which have been constructed so far, are purely focusing on the aspect of generating the maximum amount of electricity and neglecting the aspect of design. The NS has the ambition to make the carport design sustainable in appearance and material use. Besides the sustainable appearance of the design, the design should be applicable in every P+R parking plots. This results in a modular sustain- able carport design that is orientation independent. For the design, a solar cell technology was needed. Three generations of solar cells were found in literature, but the third generation was not further researched since this generation isn’t commercially available yet. The remaining two generations of solar cell technologies (1st and 2nd generation) were researched on the following topics; performance, de- sign, and sustainability. The information found in this literature research directly fed the multi-criteria analysis (MCA) method called the analyt- ical hierarchy process. With this MCA method, polycrystalline silicon solar cell technology was selected to be the best suitable for the design. Another downside to the current solar carport designs is that these de- signs don’t fully exploit the structural capabilities of the solar panels. In the final design, a connection is designed and analyzed to exploit the structural capabilities of the solar panels, whilst still keeping the trans- parent nature of the solar panels intact. The final design features recycled rail tracks in the structure of the car- port. Besides being made from high-grade steel and the shape of the rail tracks suits their integrating into the structure, the direct link to the stakeholder was also recognized as a benefit for the final selection of this material. To further enhance the sustainable appearance/function of the carport, a green wall is implemented to absorb the rainwater. Thus, im- proving the water absorption in the asphalt dominated landscape of the P+R parking plots. In particular, the design of the sustainable solar carport for the NS is analyzed on solar radiation performance and structural performance (on carport scale and on connection scale). To gain an understanding if the new connection is beneficial on a carbon footprint scale, the newly designed connection is compared to a standard aluminum transom and mullion system.

Nowadays, precast technology is mainly used for the production of concrete shell elements. Computer Numerically Controlled (CNC) milling techniques is the main production method of single or double-curved concrete elements, creating astonishing shapes with high accuracy. However, these techniques are also accompanied with high costs and large material waste. Skilled carpentry or curved steel formwork are other techniques used for the production of concrete shells. An alternative to the aforementioned techniques is the “Flexible Mould” method being developed at Delft University of Technology. This method is based on mass customization of the produced panels, due to the adaptability of the formwork's shape according to the specified geometry. The concept of the “Flexible Mould” will be used as a starting point for the current thesis.

Concrete shells are mainly used for decorative reasons such as claddings and facade elements. Their structural capacity has still to be improved so that they can be applied as load-bearing elements and avoid sub-optimal or even “bad” designs. Several researches are being conducted in order to analyze and improve the load bearing capacity of concrete shells produced with the adjustable formwork technique. Applying conventional steel reinforcement to take up possible occurring tensile stresses in concrete shells produced with the Flexible Mould, has been proved to be possible only with limited diameters due to construction requirements.

In the first part of the thesis, a research was conducted on possible improvements in the production of concrete shells with the flexible mould, so that they could be considered feasible to be applied in real life structures. Having conducted a literature review on production parameters related to the “Flexible Mould” concept, Computer Aided Design (CAD) was used as the main tool to create the digital geometry of double-curved concrete elements. Dimensional limitations imposed from the existing test set-up of the adjustable formwork at TU Delft laboratories, as well as the inherent geometry of shell elements, had to be taken into account. As a next step towards the implementation of double-curved concrete elements in real life structures, non-load-bearing laboratory tests took place. These tests were performed for illustrating reasons with sand to be the main material used instead of concrete. During these tests, a steel wire mesh diamond shaped was applied as the elastic layer of the flexible mould, being a different material in comparison with the previous studies on the flexible mould. Grid distance between the actuators was the main parameter investigated during this part of the research.

In the second part of the thesis, a study was conducted on possible improvements in the flexural tensile capacity of concrete shells via prestressing steel reinforcement. CAD design and Finite Element Modeling (FEM) were the main tools used. Due to shell's complex geometry, the research was conducted first for simple geometrical shapes under linear static analyses. Two and three dimensional elements were investigated being modeled as plane stress elements and solids respectively. Linear or single-curved prestressed concrete elements were analyzed with regard to their deflection and stress field. The number of concrete elements as well as the influence of an intermediate element in-between them simulating a joint, was an additional investigated parameter.

Subsequently, a four-point bending that will take place at TU Delft, was simulated with Finite Element Modeling (FEM), investigating its structural response under both linear and nonlinear static analyses. The created (FEM) comprises of two centrally prestressed concrete elements connected with a mortar joint in between them. No bond interaction between the post-tensioning steel reinforcement and the concrete is incorporated in the FE model. The type and size of the applied reinforcement is investigated in this model. The main goal of this part was to analyze the structure's response for loads exceeding the cracking load of the concrete, taking into account a brittle concrete failure at joint's location.

In the final part of the thesis, a post-tensioned arch structure segmented in a series of concrete elements and joints, was first designed with CAD software and then analyzed under a linear static analysis via FEM. Although more complex, a three-dimensional (3D) modeling was conducted, aiming at describing the three dimensional state of stress and deflection field of the arch. Based on a draft CAD model, several geometrical parameters had to be adjusted in order to meet construction requirements of the final concrete shell structure. Modeling and meshing requirements imposed from the Finite Element Analysis (FEA) software were also taken into consideration for the final shape of the arch.

The current thesis is a part of a research program that is being performed at TU Delft, aiming at developing the Flexible Mould with regard to construction and structural application of concrete shells in real life structures.

Shell structures appear quite often in our daily life. From aircraft fuselages to storage silos and boat hulls, these structures are employed for their structural efficiency and the aesthetics they provide. They are characterised by a small thickness in comparison to their other dimensions and by curvature in their unstressed state. Despite their structural efficiency transferring loads in their plane, thin concrete shells are not employed quite often. The traditional construction method is a labour intensive process which includes the fabrication of a formwork system, usually made out of plywood, on which concrete is cast. The necessary reinforcement in the form of bars although it is quickly placed, it has the drawback of requiring frequent joints and overlapping. Its bending and assembling is also considered difficult for curved shapes like shells. As it can be seen, great costs emerge from this process and this was a trigger factor for the formation of the research question. A different approach for the construction of shells employs prefabricated elements which are connected on site and create the complete thickness of the concrete shell. Although high construction speed and high quality are achieved with this method, significant reduction of costs is still not feasible, as the formwork is very expensive and re-use of it, is only occasionally possible. However, an adjustable mould can produce elements of different dimensions, shape and curvature, decreasing substantially the formwork costs which comprise large part of the total shell construction costs. The flexible mould method and the principle of deformation after casting that were introduced by Prof. R.H. Schipper find application in this project. This project aims to answer whether it is possible to construct a concrete shell using the flexible mould prefabricated elements, without putting aat risk the stability of the structure. A comparison between the structural feasibility of a monolithic shell and the one of a segmented prefabricated shell is the method that will be utilised to answer the research question.

This thesis concerns research into the parametric modelling and optimization of steel-concrete bridge typologies typically found in the Netherlands, in order to further the automation of preliminary bridge designs. The research question posed is: "How can a parametric bridge design model be composed to fit into the
overall road design process in order to get a comparative overview of structurally sound bridge designs of different typologies?". Within the literature review some background information is provided about the design of steel-concrete bridges, parametric design, finite element modelling and evolutionary solvers for optimization. With this information seven different bridges are designed according to boundary conditions and requirements posed in a fictitious case. These bridge designs cover the beam-, truss-, arch- and cable stayed
bridge typologies. The designs are then parametrically modelled and converted to a finite element model in a single parametric design environment, after which they are structurally analysed and optimized for minimum costs. The result is that, for the fictitious case considered, the Warren truss without verticals is the
solution with the lowest costs. This bridge design is able to carry the loads, according to load model 1 from the Euro Codes,with a lower amount of structural steel compared to the girder beam bridges, and the absence of expensive cables in the structure makes it less expensive than the arch and cable stayed bridges. Due to the scope defined at the start of the research, there are some limitations. The costs calculated for the bridges are purely based on the bridge structure itself, foundations and approach structures are not considered and this has to be kept in mind when comparing the bridges. Finally, the research into how to implement this design and optimization tool into the overall design process has not been performed, however, the model is composed in such a way that it will be compatible with minor changes to the start and end of the process.