Gas extraction in the north-eastern Groningen province of the Netherlands has resulted in an increase in induced earthquakes, causing significant damage to homes and affecting residents’ mental health. While reinforcement measures and compensations are being implemented, progress has been slow. The structural analysis for reinforcing these buildings follows various methodologies in accordance with Dutch standard NPR 9998, but these methods are time-consuming and repetitive. This study investigates the use of machine learning techniques to optimize and accelerate the structural analysis of unreinforced masonry (URM) residential buildings. The traditional seismic analysis of URM buildings is computationally expensive, requiring extensive time for model development and assessment. To address this, a classification model and a surrogate model were developed using Deep Neural Network. The data required for both models are obtained from the existing database of Royal HaskoningDHV. Both models provided results indicating whether the building requires reinforcement measures or not. The classification model employs binary classification, using specific building features to categorize structures into those requiring reinforcement
and those that do not. The model generates rapid, implementable outcomes, enabling efficient categorization of buildings. This swift assessment allows for prompt decision-making, particularly in identifying and prioritizing high-risk structures that need more detailed analysis. Surrogate models are efficient approximations of complex analysis, capable of predicting reliable structural responses. Here surrogate for SLaMA analysis is built. The SLaMA surrogate will serve as a proof of concept for developing surrogates of more complex and computationally demanding structural analyses. The results show that, for the classification model, despite sufficient data points, the correlation between input features and the target variable remains weak, primarily due to changes in NPR regulations over the years. This did not allow for a consistent assessment procedure; hence, the same building might be compliant with a recent NPR and not with the older NPRs. For the surrogate model, two different input combinations were experimented with, the analysis provides insights into the impact of input parameters, the need for dimensionality reduction, and the number of data points required for computationally expensive models. The input combinations helped to analyze the curse of dimensionality. The results showed that as more parameters were added, the model required significantly more data to maintain performance. Additionally, the relevance of the selected
reduced input features was observed, to see how well it captures the key aspects of the original analysis with less computational cost. It was observed that the selected features need to remain closely aligned with the original analysis to produce comparable results. In conclusion, while machine learning approaches offer significant potential for improving the efficiency of seismic analysis, further refinement and validation of the models are necessary to address their current limitations and enhance their applicability in real-world scenarios.

Aggregates make up around 70% of concrete by volume, leading to a high river sand and gravel extraction rate. This depletion of natural resources will soon become unsustainable as the demand will inevitably overcome the availability of natural aggregates. Moreover, about 200 million tons of basic oxygen furnace (BOF) steel slag (LD slag) are produced yearly worldwide and are predominantly disposed of in landfills. The need to find a sustainable alternative to substitute natural aggregates next to the great availability of steel slag opens the opportunity for its usage as aggregate in concrete. However, when used in concrete, LD slag undergoes an expansive reaction, resulting in cracks and concrete failure. This expansive behavior is widely attributed to the expansion of free lime. The instability has been widely investigated on the macro scale, but only a limited number of studies on the microscale investigated such a hypothesis.

The present study provides a thorough microscopic investigation of LD slag and its instability phenomenon. Firstly, X-ray diffraction (XRD) and Scanning Electron Microscopy (SEM) are used to characterize the steel slag’s chemical composition and phase distribution. In this study, a phase is defined as a portion of a system with similar chemical characteristics. The main phases of LD slag were dicalcium silicate (belite), calcium ferrite with potential alumina content, and wustite phase. During characterization, it was found 2 distinct phase distributions in LD slag particles, the first with small round and isolated belite phases and another with larger belite phases that are interconnected. This difference in phase distribution is attributed to the different cooling rates of the molten steel slag.

Then, possible reasons for the volumetric instability in LD slag were microscopically investigated, namely, the hydration of free lime, belite polymorphic transformation, hydration of magnesium oxide, and release of residual stress. Free lime and belite polymorphic transformation were not observed in the LD slag analyzed in this study and, therefore, were not considered as a mechanism of the volumetric instability. Moreover, Mg-rich phases have been observed in many isolated and interconnected LD slag particles, which are likely mechanisms of damage onset. Lastly, residual stress was qualitatively confirmed, and inhomogeneous hydration was also observed, which supports the hypothesis of redistribution of residual stress as a damage mechanism of concrete with LD slag as aggregate.

All findings and instability generation in concrete have been performed in concrete samples from accelerated curing conditions, such as high-temperature underwater curing. The test was validated by comparing the damage of the laboratory samples to a 15-year-old concrete left outdoors in the field, exposed to actual environmental conditions. It was observed that the damage features were similar, indicating that the accelerated curing conditions could well replicate the damage in real conditions and could predict the behavior of the concrete in practice.

Lastly, mitigation measures have been proposed, focusing on the concrete binder and particle treatment. It was found that the instability is delayed by the increasing development and compactness of the concrete matrix, likely by delaying water ingress. Moreover, by crushing LD slag particles, and thus -partially- releasing the residual stress present, the damage in concrete was reduced, delaying cracking initiation and concrete failure. However, it was found that despite delaying the damage onset, modifying the binder and crushing the aggregates could not ultimately prevent damage and instability.

In the Netherlands, there is a significant backlog of infrastructure maintenance and deferred projects, exacerbated by the increasing demand for infrastructure renovation funds due to bridges reaching the end of their life cycles. Monitoring and assessing bridges typically involve visual inspections, which are subjective, expensive, and time-consuming. Remote sensing techniques, particularly Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR), offer a quick and objective solution for monitoring and analysis. MT-InSAR shows promise for low-cost, network-wide continuous structural monitoring. However, there is no defined assessment method with specific damage indicators linked to bridge failures that can be applied to various types of bridges. This thesis presents a weighted multilevel assessment method using MT-InSAR techniques. In the first level, four damage indicators (DIs) are proposed and analyzed on a point spatial scale: velocity DI, relative velocity DI, six-month velocity DI, and deviation from mean time series DI. In the second level, three DIs are proposed and analyzed on a grid cell spatial scale: velocity DI, relative velocity DI, and mean cumulative displacement DI. In the final level, two DIs are proposed and analyzed along the length of the bridge: largest differential displacement DI and deflection ratio DI. This method is applied to two bridges, each facing distinct issues: settlement and fatigue. The main findings demonstrate that the proposed damage indicators can be effectively utilized in the assessment system to qualitative rate the two case studies. The assessments reveal slow downward movements and large localized upward movements, which characterize the distinct underlying issues. The capabilities of the assessment method show promise for simultaneously evaluating various types of bridge. Additionally, it was found that imposing a coherence (quality) constraint on the MT-InSAR data is not advisable, as it may filter out the most critical data. Furthermore, the data indicates a potential seasonal unwrapping error, significantly affecting the analysis. The proposed method confirms the potential capabilities of MT-InSAR techniques in bridge assessment and suggests that this approach could serve as a foundation for network-level bridge assessments in the future, contributing to a much-needed early warning system.

Masonry structures subjected to a lateral load, such as earthquakes and wind, can be prone to out-of-plane failure. It is known that geometrical properties and boundary conditions have a large influence on the out-of-plane bending capacity of walls. In this thesis, a numerical study is performed to evaluate the influence of the bond pattern on the two-way out-of-plane bending capacity of unreinforced brick masonry walls. A nonlinear finite element analysis with a 3D block-based model has been adopted to simulate the texture of the wall in detail. The study builds on the work of Chang et al., which was validated against the experimental benchmarks by Griffith et al. The numerical model can accurately estimate the initial stiffness and out-of-plane bending capacity, but the model is less accurate post-peak.
It is assumed that the material properties of the mortar in the bed joints are the same as the one of the head and collar joints. The masonry wall was first subjected to self-weight and a vertical pre-compression of 0.1 MPa. Afterwards, a monotonic out-of-plane pressure placed on the face of the masonry wall is applied. The numerical model takes both physical and geometrical nonlinearity into account.
In the first phase, different modelling assumptions were checked and a sensitivity study was performed. Prior to the study must the compressive behaviour of the brick be determined, as the bricks are elastic without a limit, thus nonlinear effects like crushing are not modelled. The resulting check determined that the minimum principal stress did not exceed the compressive strength of the brick.
A mesh sensitivity study has also been performed as a change in mesh size was needed. In the model of Chang, the face of the brick was divided over its length into two parts called a halfbrick (Hbrick). The bond patterns require a division of the face into 4 parts, a quarterbrick (Qbrick) or an octogonalbrick (Obrick) which maintains the aspect ratio of the Hbrick. Eventually, the Qbrick mesh is the best option as convergence issues arose for the Obrick.
Additionally, experiments have shown that cracks can propagate through the bricks, while this is not possible with a smeared cracking model. To simulate this effect a discrete cracking model within the brick was introduced. The possible locations of the cracks can easily be assumed in a masonry wall, to be in line with a mortar joint in the course above or below it. The change to the discrete cracking model did not influence the out-of-plane two-way bending capacity, while convergence issues arose, and it was subsequently not used.
Finally, in the experiment by Griffith et al., clamps are installed over the height of the end of the return walls. These clamps prevent any horizontal translations and rotations in the horizontal plane. These clamps are represented by tyings in the numerical model. In this study are these tyings removed and is the length of the return wall parametrised. A masonry wall without any return walls is fully one-way bending and to a length of 360 mm does the masonry wall show a transition between one-way to two-way bending. A return wall of at least 480 mm long is sufficiently long to achieve two-way bending.
A parametric study was performed considering five single wythe and three double wythe bond patterns: stretcher bond, quarter bat bond, lateral bond, half Flemish bond, stack bond, chain bond, English bond and Flemish bond. The double wythe masonry wall has been assembled by adding an outer shell to the single wythe wall and by decreasing the pre-compression to 0.0478 MPa to maintain the same vertical force as the single wythe bonds.
The highest out-of-plane bending capacity for the single wythe bond patterns is the half Flemish bond with 4.9 kPa with a spread of 5.4% and for the double wythe bond patterns it is the Flemish bond with 18.7 kPa with 7.2%. The spread between the highest and lowest out-of-plane bending capacity is small. All the walls, except the one in stacked bonded masonry, show a crack pattern in the form of a cross. The slopes of the diagonal cracks with respect to the horizontal bed joint line have also been studied and relates as follows: a shallow slope in the crack pattern coincides with a higher two-way out-of-plane bending capacity. While a steeper slope or the absence of a cross-shaped pattern coincides with a lower out-of-plane bending capacity. On the contrary, the initial stiffness of the masonry wall does not relate to the out-of-plane bending capacity.
To further confirm these findings, research is needed concerning different masonry types, considering relative properties between bricks and mortar joints, different properties between bed, head and collar joints, and spatial variability of properties.

Previous earthquake reports have highlighted the high vulnerability of masonry walls to failure when subjected to the out-of-plane (OOP) loading. Particularly, this OOP failure is a significant cause of collapse in unreinforced masonry (URM) buildings, especially those with slender walls and large openings. Moreover, in the past they were designed with minor regard to seismic design concepts. In Groningen, the Netherlands, where the majority of houses are constructed using unreinforced masonry, there have been frequent occurrences of low-intensity seismic activity, causing pre-damage to the walls. Typically, the walls get pre-damaged in the in-plane (IP) direction due to low-intensity shakes and settlements, since the box-like behaviour is established when the forces are low. Box-behaviour refers to the case where there is a strong connection between orthogonal walls and a stiff rigid floor diaphragm. As a consequence, the walls experience the desired IP shear failure during seismic loading. However, research regarding the OOP performance of pre-damaged URM walls is limited, and existing studies have shown that IP pre-damage decreases the ultimate OOP load-bearing capacity of the walls. To address this gap, this report aims to conduct a preliminary investigation on the OOP strength of pre-damaged URM walls, considering both walls without and with openings. Notably, there is no existing evidence in the literature regarding walls with openings on this specific topic. The primary research question can be stated as follows:
How does the in-plane pre-damage affect the out-of-plane load-bearing strength of URM walls?

This thesis employs a numerical modelling approach to address the research question. The software package DIANA 10.5 is used, adopting a simplified micro modelling method with shell elements to simulate the behaviour of masonry. Bricks are modelled with continuum shell elements, while the mortar joints are modelled using zero-thickness interface elements. Also, the potential vertical crack in the bricks is considered using the interface elements. The validation of the numerical modelling approach is performed in two steps. Firstly, the mechanical properties of the numerical model are calibrated using small-scale material tests. In the second step, the calibrated parameters are directly applied to the numerical monotonic analyses of full-scale walls, and the results are then compared to experimental test results. The calcium silicate (CS) brick masonry tests, from material to structural levels, conducted at the Delft University of Technology are selected as benchmarks for the calibration and validation of the numerical model. A good agreement is observed between the numerical and experimental results for the IP walls in terms of initial stiffness, peak shear force, and crack patterns. However, the results for the OOP walls are significantly overestimated, with an average overestimation of 35% for the peak force. This is attributed to the differences in boundary conditions (full or partial rotation restriction), type of tests (cyclic or monotonic) and loading conditions (displacement control or load control) between the numerical simulation and experimental tests. Nevertheless, the crack patterns are in good agreement with the experiments.

The validated model is used to investigate the OOP performance of pre-damaged walls, including both solid wall (without opening) and wall with an asymmetric opening. Two distinct approaches are considered to simulate the response of the pre-damaged walls. In the first approach, known as the reduced-parameters approach, a model is created with varying properties: reduced stiffness and strengths are assigned to locations where cracks were observed during the IP tests, while regular properties are maintained in other locations. Subsequently, the model is subjected to OOP analyses. In the second approach, known as the sequential loading approach, the wall is initially monotonically loaded in the IP direction that represents the pre-damage of the wall. Then, the OOP load is applied while maintaining the IP pre-deformation until the wall experiences failure. The study considers four different states of pre-damage, ranging from minor to extensive damage. For the solid wall, the damage levels are based on the observed damage during the IP tests up to the maximum drift of 0.2%. On the other hand, the damage levels for the wall with opening are derived from the damage observed during the IP monotonic pushover numerical simulation up to the maximum drift of 0.14%.

For the solid wall, it is found that both approaches give same results for low pre-applied IP damages, up to 0.06% drift. The reduction of the OOP peak strength is almost negligible until this drift level. As the damage increases, the reduction of the strength also sharply increases. At the maximum of 0.2% IP-drift, 40% reduction of the OOP peak strength of the wall is observed in the reduced-parameters approach, while in sequential loading approach, the wall immediately failed resulting in negligible OOP strength because of the severe IP damage. For the wall with opening, similar to the solid wall, no measurable influence of the OOP strength due to minor IP damages, up to 0.06% drift, could be observed in both approaches. The maximum reduction that could be observed is approximately 15% at maximum of 0.14% pre-applied IP drift. In the sequential loading approach, as the level of pre-damage is increased, the pre-peak stiffness of the force-displacement curve decreases, as expected for a pre-damaged wall. However, the reduced-parameters approach does not show this reduction, which could be attributed to the pre-damage applied only at specific locations. For both types of walls, there is no significant difference in the crack pattern of the undamaged and pre-damaged wall. The well-known envelope crack pattern is obtained in both cases.

To conclude, the out-of-plane load-bearing strength of URM walls is significantly affected by the presence of in-plane damages. The impact is minimal under minor in-plane damages but increases rapidly as the damage becomes more severe.

This thesis limits the numerical analyses to monotonic loading. However, for future research, it would be beneficial to extend the analyses to include cyclic or dynamic loads, as they provide a more realistic representation of seismic loading conditions. Moreover, it is recommended to incorporate various boundary conditions for the OOP loading. For instance, the research could be expanded to include a C-shaped wall configuration, where the lateral edges are supported by return walls, a scenario commonly encountered in practice.

This thesis investigates the usability and creation of a multi-criteria optimisation tool for a facade element consisting of two outer glass layers and a 3D-printed core. The chosen constraint is related to the structural requirements laid down by the NEN codes based on the ULS and SLS. The criteria are: the transparency criterion, which maximises the transparency of the panel; the insulation criterion, which minimises the U-value; and the daylight criterion, which maximises the indirect sunlight at the summer solstice and the total sunlight during the winter solstice.

Topology optimisation is the first considered optimisation algorithm that allows for complete freedom of the core design through discritisation of the core volume with finite elements. Nevertheless, topology optimisation is not considered usable in this tool as a result of the inability to add relevant criteria to topology optimisation integrated into software, the complexity of developing an optimisation tool with topology optimisation in Python or Matlab, and the intricate optimal core designs, which are not easily printable.

Next, a hexagon structure is used as the core design with some related variables, and it is created in Rhino 7. The daylight criterion is performed with the Ladybug library, and a mesh sensitivity study is conducted. The structural constraint uses Karamba3D, and Karamba3D’s accuracy is investigated with validation with Diana and stiffness experiments. The experiments revealed a large discrepancy between the stiffness observed in the experiments and the predicted stiffness in the Karamba3D and Diana models. The insulation and transparency criteria are added, and the genetic algorithm is used to determine the near-optimal solution. Finally, a case study based on the Echo building is used to compare the near-optimal sandwich panels with regular windows. Although the computational demand inhibits quick results, the sandwich panel performs well compared to the original window in relation to the three criteria.

The preservation of historical heritage buildings frequently requires the assessment of brick masonry structures. These structures exhibit complex behaviour due to the arrangement of units and the interaction between bricks and mortar. This study aims to compare two modelling techniques for simulating masonry's compressive response and understanding which technique performs better.

This study employs two block-based modelling techniques: a simplified brick-to-brick model and a detailed brick-to-brick model. The block-based modelling technique was selected to capture the intricate structural details and crack patterns of masonry specimens. Wallet specimens were evaluated to understand the compressive behaviour of masonry.

Two experimental benchmarks were chosen to validate the selected modelling approach. First, experimental results by Thamboo (2020), also modelled with a simplified brick-to-brick model by Zahra et al. (2021), were considered. However, to obtain numerical results comparable to the experimental ones, unrealistic input parameters had to be assumed (i.e., very low value of elastic modulus for the bricks); consequently, this benchmark was ultimately deemed unreliable. Hence, a second benchmark from Jafari (2021) was chosen which did not present such issues.

Considering the effect of confinement was crucial to obtain results in agreement with experiments. The detailed brick-to-brick model, explicitly modelling each brick-and-mortar joint with solid elements and accounting for mortar confinement, proved to be the only modelling approach able to simulate the second experimental benchmark. The simplified modelling technique, considering mortar joints as zero-thickness interfaces between solid elements representing the bricks, consistently underestimated the peak loads as mortar confinement was not included. This may explain why unrealistic input parameters were used by Zahra et al. (2021) to simulate the first benchmark. In general, the post-peak behaviour was difficult to simulate.

A sensitivity study was performed by varying the boundary conditions, Poisson’s ratio of mortar and integration scheme. The study on boundary conditions was performed on both the simplified and detailed models whereas the study on Poisson’s ratio and integration scheme was performed on the detailed model. The boundary conditions did not have a significant influence on the global behaviour of the specimen in the detailed or simplified models. Small variation in compressive strength and crack pattern were observed, but the effect on post-peak behaviour could not be evaluated due to numerical instability. The Poisson’s ratio of mortar was found to have a significant influence on the peak compressive strength of the masonry specimen. A higher Poisson's ratio resulted in a greater value in the peak compressive strength of masonry. A higher value of Poisson’s ratio increases the amount of confinement in mortar layers resulting in higher values of compressive strength of masonry. A high integration scheme (3×3×3) and a regular integration scheme (2×2×2) effectively captured crack propagation. On the contrary, a reduced integration scheme (1×1×1) could not do so effectively. The reason can be attributed to the higher number of integration points in the former cases which allows for better propagation of cracks.

These investigations provide precious insight into the choice of modelling techniques for simulating the compressive behaviour of brick masonry. The detailed modelling technique can be implemented to study the compressive behaviour of masonry specimens if mortar confinement is accounted for. However, further effort is necessary to reduce numerical instability to simulate the post-peak response. This can be adopted further to study the influence of, for example, bond pattern, specimen thickness (e.g., multi- vs single-wythe masonry) and specimen shape (e.g. wallets, prism, core) on the compressive behaviour of masonry.

Glass is a fascinating material that has also been used as a primary construction material in various remarkable buildings since the last century. Unfortunately, a lot of carbon dioxide is released during the production of glass because of the extreme heat required. The construction industry accounts for ten per cent of global CO2 emissions. If a focus in this sector is put on recycling, reducing and reusing, emissions can be drastically reduced because of less needed new building material. However, structural glass is currently hardly reused. A second challenge arises with existing demountable structures made of glass like the LocHal: these are not weatherproof, thus unsuitable as sheltered accommodation.

In response to the absence of adequate standardised structural glass building systems, this research project proposes a preliminary design for a modular, transportable art pavilion with an appropriate structural verification. The research question is therefore: ’How can glass be applied as load‐bearing material in temporary modular building units to realise easy‐to‐ (dis)assemble, transparent and transportable structures?’.

A case‐study is introduced to find an answer to the main research question. The fictive scenario is sketched to design a temporary art pavilion which stands for one to three months in a city centre in the Netherlands. After this period, the pavilion should be demounted and transported to the next destination. The information described here determines the boundary conditions for the design and calculations. The imaginary pavilion is 24 m in length, 10 m in width and 2.5 m in height. The inner walls in the pavilion are retained to create a natural walking path inside. On the short side of the pavilion, doors are inserted as an entrance and an exit. From the available literature, the transportability of the building elements and the requirement for thermal insulation appear to be important preconditions for the final design.

The design is based on four different types of prefabricated building elements: roof panels, wall panels, floor panels and base profiles. The roof panels are of 220 mm thick cross‐laminated timber (CLT), in length six or three metres and have a width of 2.5 m. The wall panels are of laminated glass, 2.5 m in height and come in two types. Insulated glass units (IGU’s) of 6 or 5 m long function as exterior walls. These consist of an outer sheet of 10 mm fully tempered glass, a 15 mm cavity of 90% argon gas and a triple laminated inner panel of 5.10.5 mm heat‐strengthened glass. The inner walls are of a composition of 5.10.5 mm heat‐strengthened glass. The glass is laminated with a SentryGlas® interlayer of 1.52 mm. CLT is also used for the floor panels, now with a thickness of 210 mm and lengths of 6 and 3 m. The base is defined by steel ’cap’ (RHSFB, lengths of 6 and 5 m) and ’hat’ (THQ, length of 4.25 m) profiles. Both profiles are 265 mm in height.

Roof connections, wall connections and base connections were designed and dimensioned, in total seven different types. The most innovative and structurally interesting joints are two wall connections. These connections consist of a so‐called ’coffee‐ cup‐hand’ system; titanium elements laminated 30 mm into the middle sheet of the wall panels. The wall panels are checked on maximum deflection and tensile stress, as well as the local tensile stress in the glass and in the SentryGlas® interlayer around the laminated titanium elements. All checks comply with the maximum allowable values described in the Eurocodes and in literature.

The conclusion is that the proposed design, with some enhancements to be made, satisfies the structural‐, building physics‐ and practical requirements as a transportable and a relative transparent building system for structural glass. With this building system, glass can be integrated as a load‐bearing material for designs of temporary structures.

Engineers who wish to apply this building system in practice are advised to first enhance the roof connections. For transportation means, the grid‐measurements should be decreased by 8% to fit the components in a regular container. For practice, it is also advised to deal with factors such as installations and drainage systems, which were not included in this study. Follow‐up research could focus on the adaptability of the building system when the building is used with a more permanent function. In addition, it is mechanically interesting to further investigate how the wall connections interact with each other in a 3D analysis and lab experiments.

The Circular Arch Bridge is a pedestrian bridge consisting out of four different recycled materials: cast glass, ceramic, circument and geopolymer concrete. Due to an increasing amount of vibration problems in modern pedestrian bridges they should not only be designed for static loads, but dynamic loads have to be considered as well. There are currently no guidelines in the Eurocodes for the dynamic assessment of a bridge like the Circular Arch Bridge. Therefore, this paper provides a research on how the dynamic assessment of the Circular Arch Bridge should be conducted.

The Circular Arch Bridge is considered a dry-stacked masonry arch bridge. A full dynamic assessment is conducted, with the use of finite element software, according to the guidelines for pedestrian bridges. The eigenvalues of the Circular Arch Bridge are not within the critical range for pedestrian induced dynamic loads. It is found that for everyday use the Circular Arch Bridge is within Comfort Class 2 from the Eurocode, thereby providing a medium comfort level. A special load case is added to the analysis representing a group of 60 people dancing on the bridge during experimental testing of the bridge. For this load case the Circular Arch Bridge is within Comfort Class 3 from the Eurocode, thereby providing a minimum level of comfort.

With a sensitivity analysis it is determined that the most sensitive input parameter of the finite element model is the modulus of elasticity of the interlayer. An investigation of the effect of large stiffness differences between the interlayer and the stones on the dynamic properties of the bridge is conducted. It is concluded that the large stiffness differences have a positive effect on the dynamic properties of the bridge. The effect of the dry-stacked assembly of the bridge on the dynamic properties is also investigated. This as a negative effect on the dynamic properties of the bridge.

Seismic events have shown the hazardous and expensive consequences resulting from seismic damage of non-structural elements. For this reason, the concern regarding the seismic performance of glazed curtain walls has acquired more relevance in the building industry; however, there is still a gap of information in the literature about this topic. To contribute with a broader panorama about the seismic performance of CW, this thesis project is focused on the seismic of the connections within a curtain wall.
This thesis intends to evaluate the performance of the connections that contribute to the seismic response of a glazed curtain wall with finite element models and experimental tests. To achieve this objective, several important points were taken into account. First, a literature study to understand what is a CW, and which are the components that may have influence in seismic design. Second, the two case-studies considered in this research are presented; in this section information about the specimens tested and types of tests performed is provided. Third, elaborate finite element models of the connections corresponding to each case study describing the criteria used. Finally, a comparison between the experimental and numerical results is elaborated to determine if the modelling approach is correct.
The experimental tests were carried out with the collaboration of Permasteelisa Group, and the numerical models were elaborated with DIANA FEA software. Five connections were analysed in this research: mullion to mullion, transom to starter sill, top aluminium bracket and hook, starter sill- bottom steel bracket, and silicone by a non-linear analysis with incremental displacements. At the end of this analysis it was observed a good agreement in the behaviour of the frame to frame connections with a minimal over estimation of the stiffness of certain frame to frame connections. Regarding the connections involving brackets, it was observed that the behaviour in the in-plane horizontal direction can be affected by the rotation of this elements because bolted connections shall be modelled as rotational springs. Finally about silicone it was observed that there exists a good match between the experimental results and the numerical models.
In conclusion, the modelling procedure was validated with experimental tests with relatively optimal results. Nonetheless, there are still limitations in the development of this topic that will be described in the recommendations but much more openness is expected in the future in terms of research on this topic.

Increasing housing demand in Europe and the need to be more sustainable are asking the construction industry to leave the traditional pathways and innovate. An emerging construction method with volumetric timber modules potentially offers the solution. However little is known about the robustness of this new innovation. Therefore, this thesis aims to advance the current research regarding collapse resistance in volumetric modular timber buildings. To achieve that goal, first, in this thesis, a literature study into robustness and modular buildings was performed. The study identified the importance of redundancy, proper connection design, ductility, and tying in modular structures for robustness. Secondly a case study was performed. The case study provided the structural design concept of a newly developed volumetric timber post and beam module by Lister Buildings, structurally designed by Pieters Bouwtechniek. The modules consist of glued laminated beams and columns with cross-laminated floor and ceiling slabs. With the modules, a hypothetical residential building of 5 storeys was created, which consecutively was decomposed into equivalent two-dimensional frame structures suited for 2D finite element analysis. To account for the mechanical behaviour of the connections in the models, nonlinear springs were used. To establish the springs, the component method was adopted. The spring behaviour was determined from the elastic, plastic and failure performance of the individual components of a connection in translational and rotational degrees of freedom. Then based on a notional element removal concept, multiple possible loss events of load bearing elements were analysed in the finite element software of Abaqus to examine the existence and development of alternative load paths. To do that, both nonlinear static and nonlinear dynamic analyses were performed. The results of this study indicate that the new design concept performs quite well on robustness. Several alternative load paths were identified, among which cantilever action, bridging action, and catenary action, depending on the removal event and scenario. In addition this thesis investigated the dynamic amplification factor (DAF) of the modular system. The results of this research imply that for alternative load path analysis of timber structures, a DAF of 2.0 should be applied in a (non)linear static analysis. More research will be necessary to examine whether a lower value than 2.0 can be used.

The gas production in the Groningen gas field of the Netherlands has caused a significant amount of shallowhuman induced earthquakes. Among various building typologies, Groningen is home to many Dutch historical churches constructed by unreinforced masonry which has shown to be highly vulnerable to these earthquakes. From the perspective of conservation and prevention of loss of our historical and cultural heritage, the structural assessment of these churches and their monitoring is of importance. The assessment of the seismic performance of historic churches is a challenging task because of their unique design, governed by macro element behaviour and nonlinear material behaviour. An accurate and reliable earthquake resistant assessment with appropriate numerical modelling of the structure and proper assumptions of all the uncertainties is crucial. The structural monitoring cannot be applied to all historical churches. By selecting representative case studies, indepth study regarding both numerical modelling and structural monitoring are possible. This information can then serve engineers and professionals to evaluate similar structures. The research gap lies in accurate modelling of historical churches to achieve reliable and faster results by linear dynamic modelling. The fundamental modes, eigenfrequencies and modal shapes, geometry and material properties, boundary conditions, connections between structural elements and loading variations are studied using 3D models with shell elements of the casestudy. Although, the models described cannot closely represent the real structure. Firstly, a regular thickness has been assigned to masonry walls and piers despite their irregularities. Secondly, the material properties have been estimated from similar structures in Groningen, and may not represent the actual material characteristics of this case study. Thirdly, the cavity ties has been disregarded because they have been assumed to be corroded (no actual information was available). Finally, the concrete slab at the entrance of the church is disconnected to the foundation, which was an approximation from the drawings but could not be verified. For the casestudy prior structural retrofitting (models 15); Numerical Model 3 with eliminated foundation and timber flooring and a rigid base at the ground level is expected to give the best approximation of dynamic response of the church. The fundamental frequency in X direction of the casestudy prior structural retrofitting can be approximated between 2.30Hz to 2.75Hz and the fundamental frequency in Y direction between 3.0Hz to 3.45Hz. The fundamental global modes in this model show maximum deformation in the timber roof of the mainstructure and tip of the belltower in globalnd Y directions. This indicates weakness of the casestudy prior structural retrofitting. The Old Church has undergone structural retrofitting in July 2018, to prevent further seismic damage on the structure. From the survey of the casestudy and the retrofitting measures, the recent findings (eg., cavity walls, installation of steel frame, steel and timber columns) are incorporated into the finite element model of the casestudy and various modelling variations are presented in models 6 to 10. From comparing the results of these models post structural modifications, Numerical Model8 (or Model9 which has almost similar results) with degraded material properties of masonry and timber diaphragms, a shallow masonry step foundation and a rigid base at the foundation level are predicted as the closest approximation of the dynamic properties of the church. The fundamental frequency of the casestudy post structural retrofitting can be predicted to lie between 1.5Hz2.6Hz in global X direction and between 1.1Hz1.65 Hz in global Y direction. It can be observed that the implemented measures make the church stiffer and the weakness of the structure post structural retrofitting is localised at masonry facade walls of the belltower, cavity wall between the belltower and mainstructure, tip of the belltower, lateral walls and timber roofing of the mainstructure. This research on simulating a numerical model that closely represents the dynamic properties of the Old Church in Garrelsweer for achieving reliable, computationally effective and faster results. This can be used as a basis to compare the results of ambient vibration testing and operational modal analysis, and a relevant guide to study how the numerical models can be defined for modelling similar structures.

Bridges are exposed to dynamic forces – such as pedestrians crossing a bridge or the wind force acting on a bridge – which cause vibration of the structure. When a structure is forced at one of its eigenfrequencies it deforms into a corresponding shape leading to major deformations, which, in the worst-case scenario, can lead to collapse of the structure, as happened to the Tacoma Narrows bridge in 1940. It is thus very important to know the dynamic characteristics of a structure. For glass structures no specific limit values are defined, this is why it is interesting to investigate the dynamic behaviour of a glass bridge This thesis aims to predict the structural response of TU Delft’s Glass Truss Bridge when subjected to static and dynamic loads. The studied bridge is a footbridge, connecting the TU Delft campus to the Green Village, and is unique in the aspect that glass diagonals have been used in its construction. Through an iterative procedure of modeling, virtual testing and updating the model of the bridge, the dynamic response of the bridge is predicted. The study shows that the response of the bridge is between the European Guidelines’ limits for (steel) footbridges. Maximum acceleration of the bridge is found at 500mm/s and the lowest eigenfrequency is out of range of that of joggers. The sensitivity analysis shows that the parameters having the most influence on the bridge’s response are the diameter of the glass diagonals and the soil density. Using the outcome of the sensitivity analysis a dynamic testing protocol is recommended that outlines the optimal test setup for modal testing of the bridge. Through the simulated dynamic tests ten of the twenty eigenfrequencies of the bridge are found. For each of these frequencies, the damping is approach and modeshapes are reconstructed based on the imaginary plots of the Frequency Response Functions. The latter show a good correlation with the modeshapes obtained from the model.

Steel trusses are the preferred structures when large distances are required to span due to its lightweight, the reduced deflection and its load bearing capacity. Although nowadays with the help of computational tools more efficient designs can be achieved, the geometrical shape of trusses has barely changed. This is mainly due to the ease of the required structural calculations and due to the available standard manufacturing and assembly procedures. Although a standard shape could be seen as an advantage, studies suggest that the relationship between the efficient use of material and design rationalization is not balanced, leading to design-overcapacity of the structure. The use of optimization procedures provides new opportunities to engineers. New design solutions and near-optimal structures can be achieved when implementing optimization processes. Optimization processes could generate economical alternatives in terms of weight reduction. Despite optimization techniques are becoming increasingly available, optimized structures are not commonly used in real practice due to the complex configuration an optimized structure might have, especially when using layout optimization. Due to the non-conventional node-members configuration, realizing these optimized trusses in real practice with conventional joining mechanisms, such as bolted and welded connections, will probably lead to complex designs and higher costs. Therefore, innovative joining technologies are required to be implemented. Snap-fit connections are widely used connecting mechanism for joining plastic components. Due to its simplicity and efficiency while connecting two or more elements together are suitable for different applications. Despite it is not commonly used in the building industry and it is frequently used on small scale applications, the principle based on snap-fit connections could be used to connect the structural members of optimized trusses. First, a literature study provides a clear understanding of truss structures, focusing mainly on the importance of their joints. Structural optimization is also discussed, paying special attention on a new layout-optimization approach, which is based on two steps: first, the less-weight possible layout is obtained (the benchmark) without taking into account any buildable considerations. The second step consists on rationalizing the benchmark contemplating practical and buildable restrictions. Finally, a representative case-study steel truss is presented. The second part of this research explores the structural design and the optimization procedure applied on the case-study truss. Here the layout optimization approach is applied using Peregrine (plugin for Grasshopper). After an optimized layout is generated, it is design structurally using Karamba3D. Finally, the case-study truss and the optimized truss are compared. To explore the opportunity of implementing optimized trusses in practice, the conceptual design of an innovative joint based on snap-fit connections is developed. A representative internal node from the optimized truss is chosen and based on a general concept, three design concepts are created. Extensive Finite Element (FE) analyses are carried out to study the structural behavior of the innovative joint during three main phases. All FE analysis were performed using the FE software Abaqus. To conclude, this research project is an exploratory feasibility study of the use of steel snap-fit joints as internal nodes for optimized steel trusses.

Gas extraction has caused increased seismic activity in the area of Groningen, bringing about considerable damage to houses. The houses are generally old masonry houses with flexible timber diaphragms. It is of importance to assess the response of these houses in order to apply proper retrofit measures. There has been substantial research on masonry, however there has been little exploration on timber diaphragms. For this reason, the TU Delft conducted cyclic tests on as-built and retrofitted timber diaphragms, as well as on the connections within the diaphragms. Timber diaphragms exhibit a nonlinear hysteresis under cyclic loading, which is caused by the nonlinear behaviour of the connections. This thesis focuses on the numerical modelling of the as-built and retrofitted diaphragm as tested at the TU Delft. Using ANSYS 18.2, the numerical models of the diaphragms are subjected to cyclic pushover analyses. As the connections within the diaphragms are the only source of nonlinearity, these are simulated first. For the nonlinear connections, it is essential that the numerical models can capture the pinching behaviour and strength degradation for cycles of the same amplitude. The numerical models of the connections are calibrated on experimental tests. With the models of the connections, the models for the as-built and the retrofitted diaphragm can be constructed. The results of the numerical models of the diaphragms are validated with experimental data. The models showed sufficient similarities with experimental results and are deemed suitable simulations of timber diaphragms.

In Groningen, seismic activity has increased due to the extraction of gas in the area. A large-scale research campaign has been launched with the aim to assess and safeguard structures in the region. However, an accurate assessment of these buildings turned out to be a challenge, due to the nonlinear behaviour of the masonry and the dynamic nature of a seismic load. A Nonlinear Time History (NLTH) analysis takes into account both these factors, but the computational demand of such a method is considerable. Another method that is widely used to analyse the seismic response of a structure is the Modal Response Spectrum (MRS) method. The computational demand of this method is considerably less compared to NLTH, but nonlinear material behaviour is only taken into account in an indirect manner via a behaviour factor, and the results are considered to be too conservative. A third method is the Nonlinear Pushover (NLPO) method. It takes nonlinear material behaviour into account and compared to NLTH, NLPO is computationally more efficient. Furthermore, an advantage is that it separates capacity from demand. Even though the NLPO method is commonly applied worldwide, its validity still needs to be proven for the Groningen case. Both objectives were studied by looking into a single case study, consisting of a low-rise URM apartment building. The behaviour of the structure is characterised by a weak and strong direction, in which the weak direction is characterised by a relatively low stiffness and lateral capacity compared to the strong direction. The seismic response of the structure is determined according to the MRS, NLPO and NLTH methods. Furthermore, the NLPO analyses are executed using two different computational discretisation methods, namely continuum FEM and macro EFM. DIANA is used as a FEM solver for the MRS, NLPO and NLTH analyses and 3MURI is used for the EFM model. Moreover, a modal and uniform lateral load pattern are taken into account for the NLPO analyses. The conclusions which are drawn from the case study can generally be applied to low-rise URM apartment buildings in Groningen. However, it must be noted that significant alterations in geometry and building materials might influence the results. Furthermore, modelling assumptions have been applied, and it is important to note that the possible influence of these assumptions, may partially limit the extent of the conclusions. Moreover, several limitations are inherent to the studied methods, and cannot be accounted for somehow. All analyses are performed by incrementally increasing the seismic load until one of the near collapse limit state criteria according to NPR 9998 is met. Furthermore, three target displacement methods are evaluated: the capacity spectrum method according to NPR 9998, the regular N2-method, included in the Eurocode 8, and an adaptation of the N2-method which is developed specifically for URM structures by Guerinni. The performance of the structure according to each of the methods is studied subsequently, by looking into the force-displacement behaviour, displacement profile and damage at failure, failure mechanisms and the maximum admissible seismic load. Two significant disadvantages of macro EFM were identified when comparing the results of the NLPO analyses using 3MURI and DIANA. First, the fact that out-of-plane behaviour is not taken into account in 3MURI could significantly influence the behaviour of a structure in terms of base shear capacity, which is especially true when structures are characterised by an extremely low total length of piers in the in-plane direction. Furthermore,DIANA allows for a more gradual softening behaviour, which helps the post-peak force redistribution. As a consequence, the maximum admissible seismic load according to DIANA could be higher. However, despite the two aforementioned disadvantages of the macro EFM method as implemented in 3MURI, all other relevant results of both methods are similar. The fact that the two identified disadvantages of 3MURI can only result in more conservative results, suggests that macro EFM, as implemented in 3MURI, is a suitable computational discretisation method for the seismic assessment according to the NLPO method for low-rise URM apartment buildings in Groningen. However, it should be taken into account that the conservativeness of 3MURI could lead to a significantly larger amount of required retrofitting, in comparison with DIANA. The applicability of the NLPO method is reviewed by comparing the results of the MRS, NLPO and NLTH methods. Similar behaviour of the structure according to the NLPO and NLTH method was captured, which suggests that the NLPO method is a suitable analysis method for the studied typology. The maximum admissible seismic load using the target displacement method according to NPR 9998 is in-line with the NLTH analysis. However, the governing load case made use of a uniform load pattern, which returns a structural behaviour different than that obtained by NLTH analyses, as can be seen from the force-displacement behaviour. If only the capacity curve according to the modal load pattern would be considered, then the allowable seismic load according to NLTH is similar to that of NLPO using the target displacement method of Eurocode 8. Furthermore, from the results of the case study can be concluded that the choice of target displacement method has a significant influence on the maximum admissible seismic load. For the case study, NPR 9998 is more conservative in the strong direction, and Eurocode 8 is more conservative in the weak direction. Regarding the MRS method, very conservative results were found. A reason that was found for these conservative results is that the prescribed behaviour factor by NPR 9998 is too low for the case study when compared to that derived from the NLPO analysis. However, even with a larger behaviour factor, the results according to the MRS method would still be conservative in the weak direction.

Researchers of TU Delft have been designing and engineering with columns created out of multiple glass rods that are bundled together to form a single column. These aesthetically attractive and slender columns were tested to ensure safety before implementation in the real structure: the Glass Truss Bridge located at the campus of the university. However, the ultimate capacity and structural behaviour of such columns remained unknown. In this studies the structural behaviour of two designs of such columns is researched. It appeared that, among others, the differences in length between the rods within the columns caused an unequal distribution of stresses. This resulted in a large range of ultimate capacities of bundled glass columns that were designed according to the same design principles. Using a special approach in finite element modelling of the glass until failure, furthermore, resulted in more thorough knowledge regarding the failure behaviour of glass and the best way to model this behaviour.

This thesis describes the creation of the Equivalent Shear Masonry Model (ESMM). This is an orthotropic material model for low bond strength masonry that uses a smeared cracking approach. It was developed as an improvement of the Engineering Masonry Model (EMM) and adopts that model’s constitutive relations for the behaviour normal to the bed joints and for the compressive behaviour normal to the head joints. However, the shear failure criterion and the tensile head joint failure options are replaced by a new failure criterion for diagonal staircase cracks that is based on the observation that these cracks often open up horizontally. This failure criterion, the equivalent shear failure criterion, is derived from horizontal force equilibrium and evaluates both the shear stress and the tensile stress normal to the head joints. The combined constitutive relation is based on the Coulomb friction shear behaviour of the bed joints, that consists of linear loading, linear softening and a residual stress plateau. The softening is dictated by both the total shear strain and the total horizontal strain. This thesis first describes the development of this theory and its implementation into a user supplied subroutine for Diana FEA. This code was then verified for a single integration point and compared to the EMM’s Diagonal stair-case cracks option for a selection of load paths, among which combinations of shear and horizontal extension. This verification showed that the ESMM is more stable and provides more probable stress-strain diagrams. Subsequently, the model was validated against a micromodelled masonry unit cell for the same selection of load paths. This validation showed that the ESMM’s stress-strain diagrams are realistic, save some details that could be improved. Finally, the model was validated at a structural level with a macromodelled prediction of shear wall experiments. This validation showed that the ESMM is able to model post-peak behaviour, with both softening and a residual force plateau. The model featured inelastic deformation, hysteresis and even peak force reduction. Compared to the EMM’s Diagonal stair-case cracks option, it had less convergence issues and more realistic crack localisation. Some adjustments to details of the theory and the code are suggested. Also, further validations are required, for more load cases and other applications. Altogether, the Equivalent Shear Masonry Model shows promising characteristics and it is recommended that it is further improved and developed into a convenient, practical material model for low bond strength masonry.

This thesis is a study towards in-plane behaviour of masonry shear walls modelled through a predefined crack pattern at macro level. The Semi-Lumped method, proposed by Messali (2015), is analysed in detail to confirm the consequences of the SLM assumptions. This is achieved via literature and numerical validations. The final part of the thesis contains a pilot study towards a combined procedure of the Semi-Lumped Method and the Sequentially linear analysis.