Flexural shear failure is a brittle failure mode that can occur in reinforced concrete (RC) beams without stirrups due to the combination of flexural and shear stresses. The failure mode begins with vertical flexural cracks at the bottom of the RC beam central span area due to flexural tensile stresses, followed by diagonal cracks. During stabilization, a diagonal crack enlarges, leading to flexural shear failure. The failure mode is brittle due to the significant bearing capacity reduction making it more difficult to predict.

Accurately predicting the capacity of concrete structures is important for ensuring their safety, especially in the case of brittle failures. Various design codes are available to design and assess such structures, but an advanced numerical method called the Non-Linear Finite Element Analysis (NLFEA) is an alternative to these codes. NLFEA allows for more detailed and accurate modeling of the structure behavior by considering material, geometry, and boundary conditions nonlinearity. By using NLFEA, engineers can optimize their design and gain a deeper understanding of the behavior of RC beams without stirrups. The NLFEA model requires several modeling decisions to accurately simulate the structures’ behavior.

Sensitivity analysis on different modeling aspects is crucial to obtain a numerical model that can accurately simulate the RC beam. To be considered accurate, the numerical model should simulate approximately the same damage progression, failure mode, and failure load compared to the experiments. The sensitivity analysis is performed to modeling aspects with uncertainties identified during the literature review. These uncertainties are in the constitutive model, finite element discretization, and analysis procedure modeling aspects. Sensitivity analysis on various modeling aspects is per-formed using four experimental beams with distinct geometrical sizes, while some material configura-tions differ. This research investigates whether, using sensitivity analysis, a numerical model can be obtained that accurately simulates flexural shear failure for RC beams without stirrups.

The total strain crack models’ crack orientation sensitivity analysis shows that the rotating crack orientation can suffer from over-rotation, which causes delamination of the concrete cover. Over-rotation also shows a strong correlation with many non-converged steps. In addition, the fixed crack orientation simulates a more realistic representation of the experimental failure mode. The compression-compression confinement sensitivity analysis shows that this modeling aspect does not influence simulations for cases with flexural shear failure much and can thus be excluded. A slightly lower failure load is simulated with the confined numerical model for one of the four cases. The sensitivity analysis on the FIB bond-slip relation and Shima bond-slip relation reveals that the former has a lower initial stiffness when using the same material configurations for their modeling assumptions. Due to the lower initial stiffness, there is a higher relative displacement between the concrete and reinforcement. In some cases, this results in either increased convergence problems, a higher possibility of dowel failure, a lower failure load, or a combination of them.

For the fourth sensitivity analysis modeling aspect, the full Newton-Raphson (NR) iteration scheme simulations are slightly more representative of the experiment than the Secant iteration scheme. This result is obtained despite the full NR scheme having more convergence problems during the initial crack. In addition, for a few cases, the Secant iteration scheme simulates symmetrical flexural shear failure due to failing to include material nonlinearity.

Sensitivity analysis of the reinforcement elements shows that simulations with truss elements are more accurate than beam elements. The beam elements models show compatibility issues when combined with plane stress elements. The interface elements fail to correctly tie the beam elements’ extra rotational degree of freedom to the transitional degree of freedom. This incompatibility results in convergence problems. Also, higher relative displacements and a higher stiffness after the initial crack is noticed in some cases compared to the experiment. The final sensitivity analysis reveals that the element size sensitivity increases with an increase in the geometrical beam size. Too-large element sizes decrease the accuracy of simulations. In contrast, too-small element sizes increase the computational cost but can also simulate irregular crack patterns not representative of the experiment. A formula is introduced from the sensitivity analysis for beams up to a depth of 1200 mm to predict an appropriate element size.

The sensitivity analysis reveals that the most accurate numerical model is a fixed crack orientation and the Shima bond-slip relation combined with truss elements using the full NR iteration scheme. The sensitivity analysis is followed by a quantitative analysis of 76 experimental cases to verify the accuracy of the obtained numerical model for a broad range of differently configured experimental cases. Analysis shows that dowel failure can get captured due to an excessive change in the dam-age-based shear retention factor using the obtained numerical model. However, decreasing sensitive load step sizes to very small ones results in flexural shear failure. Also, the quantitative simulations show that the numerical model simulations are largely accurate, with 62 simulated cases below a failure load percentage difference of 10 % compared to the experiment. The average percentage difference is 6 % between the simulations and the experiment.

Analysis shows that this research successfully obtains a numerical model that accurately simulates flexural shear failure for RC beams without stirrups. The information obtained from this research can be used to make modeling choices. In addition, some uncertainties for other modeling aspects are introduced for future research. These modeling aspects are the shear retention model, concrete elements compatibility with the reinforcements beam elements, and the global element size for beams deeper than 1200 mm.

With the rise of hydrogen as a green alternative for fossil fuels, the demand for storage capacity of hydrogen increases significantly. Due to the high risk of explosions in urban environments it brings along, the blast analysis of buildings within in range of the explosions becomes more relevant.

The goal of this research is to develop a quick method to analyse the roof of a building subjected to blast load. This is done by first clarifying the blast load definition on a reinforced concrete (RC) structural element through existing design standards and literature. Next, the material behaviour of the concrete and the reinforcement steel is scrutinised by an extensive literature study. Materials behave differently under dynamic loads. The dynamic material properties are increased by the strength increase factor (SIF) and the dynamic increase factor (DIF).

In blast analysis, most energy is dissipated though plastic deformation. Therefore, it is of great importance to accurately describe the nonlinear behaviour of RC elements. The nonlinear behaviour of RC elements is translated in the moment-curvature relationship. This relationship is calculated on cross-sectional level and serves as input for the global beam or slab model. The global structural behaviour of the beam or the slab is calculated using the finite difference method (FDM). The FDM model generates a force-deflection (F-u) relationship which can be used in the single degree of freedom (SDOF) mass-spring system. The SDOF mass-spring system is used in this research to predict the dynamic behaviour of RC elements.

The research method is validated by published experiments and finite element analysis. Three experiments are reported, where the following results are obtained:
• Flexural stiffness may be assumed when the scaled distance is above 1.2 m/kg1/3. This is labelled as the ‘far field design range’.
• When choosing the DIFs carefully, the dynamic behaviour of RC elements can be predicted well.
• The FDM model can provide a good estimation of the nonlinear F-u relationship. The method of incorporating cracks in the FDM model is not previously presented in published literature.
• The unloading stiffness requires additional care. This research briefly covers the unloading stiffness.
• According to the UFC 3-340-02 (Department of Defence, US, 2008), RC elements without shear reinforcement and without the possibility of membrane action, fail at a support rotation of 2 degrees. This is where crushing is supposed to happen. This research shows that this is rather conservative and that the support rotation can go up to 6 degrees before failure.

Finally, the validated research method is applied on a case study. The case study contains a slab supported on two stiff beams on opposing sides. This results in a main span (weak direction) and a secondary span (stiff direction) due to the flexural stiffness of the supporting beams. In most cases, the slab supported by beams can be approached as a SDOF mass-spring system. After occurrence of cracks, the slab reinforcement in the main span direction will absorb most of the energy and is therefore the dominating member in the two degrees of freedom (2DOF) mass-spring system.

Dynamic behaviour often is a topic of concern in the design of pedestrian bridges. It is time-consuming to assess and little is known about the way it is influenced by the characteristics of a bridge. In many cases, a bridge needs testing after is has been finished to determine the specifications of a damper.

Striving to reduce the environmental impact of bridges, there lies a great potential in using materials with a low environmental impact, such as timber. This research combines the lack of knowledge about dynamic behaviour of footbridges with the need for using timber instead of other materials. It consists of two parts. The first part, the parameter study, investigates the influence of three preliminary design parameters on the dynamic behaviour of a long-span timber footbridge, namely the pylon height, the pylon shape and the amount of cables. The second part, the optimisation study, examines to what extent it is possible to design a long-span timber footbridge that does not need dampers to control excessive vibrations.


To this end, a parametric model of a bridge was made in which parameters can be varied and optimised to create realistic design variants. To be able to optimise taking into account dynamic behaviour, a python script was written to automatically determine the type of modes. The results of the parameter study show that the dynamic behaviour can be influenced by the parameters, although the results depend on the specific model, dimensions, parameter values and damping value. The results of the second part show that a with a 14% increase in mass a design variant that does not need dampers to control excessive vibrations can be obtained.

Cracks in masonry structures are a cause for concern as they signal a potential lack of functionality and/or aesthetics. It thus becomes important to identify the cause of damage in order to mitigate it and to prevent its occurrence in the future. Similarities in crack patterns may correlate to similarities in the damage cause. Currently, the assessment of similarities in crack patterns and their corresponding damage causes is done by masonry experts and structural engineers. This process is often expensive and subjective. The use of a Convolutional Neural Network (CNN) may offer an alternate robust and dependable means to automate the assessment of masonry crack patterns by processing their images.

The main research goal of this MSc thesis is to answer how accurately can the CNN -- fitted to data generated from finite element models -- estimate masonry crack pattern similarities. To develop a neural network that can perform such an automated assessment of masonry crack patterns with a high degree of accuracy, a large number of crack patterns with similarity ratings given by human experts are required. This data is collected in increasing complexity, first from a statistics-based approach by generating synthetic crack patterns from Markov walks. This is followed by a computational physics-based approach, such as the Finite Element Method (FEM), that generates crack patterns on 2D masonry façades subjected to differential settlements and out-of-plane loads. Finally, real-world data is also collected. This data is used to fit and test a convolutional neural network developed by Kleijn (Kleijn, 2022). Continuing along the previous line of research done at TNO (where 12 crack patterns were chosen and developed using the statistics-based approach), this thesis focuses on developing parametric finite element models of 8 out of these 12 Pattern IDs. Additionally, real-world images are also collected from Gouda in The Netherlands. This data is then used to form crack pattern image pairs that can be assessed for their similarities by 28 raters using three similarity label categories: crack pattern similarity label, damage severity label, and the overall similarity label. Using these labels, the raters assessed 2587 image pairs generated from the statistics-based approach, 500 image pairs from the computational physics-based approach, and 50 image pairs from the combination of images from the statistics-based approach, computational physics-based approach, and the real-world cases.

An inter-rater agreement analysis is performed on the similarity assessments using Krippendorff’s alpha measure. Additionally, the agreement of each rater with a chosen standard rater is studied using Lin’s Concordance Correlation Coefficient (CCC). Using Lin’s CCC, the intra-rater agreement is also assessed for the standard rater to see how consistent a rater is with their own annotations. These labelled image pairs are then used to fit and test the regression neural network to evaluate its accuracy in predicting the similarity labels. The neural network is also fitted to and tested with various combinations of labelled data to study its generalisability.

It is found that in all three sets of data, Krippendorff’s alpha is less than 0.80 for all the labels, which indicates an insufficient agreement among the raters. It is also seen that, in general, agreement among the raters increases with their experience level, i.e. the descending order of agreement within the rater group is: industry experts, PhD students, and MSc students. Studying the Lin’s CCC of each rater’s performance compared to that of the standard rater helps to choose the raters who can be considered as reliable as the standard rater. Additionally, the intra-rater agreement analysis of the chosen standard rater shows that the highest self-consistency (agreement) is achieved for the crack pattern similarity label, followed by the overall similarity label and finally the damage severity label, with corresponding Lin's CCC values of 0.96, 0.86 and 0.72, respectively.

The neural network is tasked to predict the similarity level in each similarity rating for each image pair in the test sample. The ground truth of this neural network is established by averaging the similarity ratings given to each image pair by multiple raters. It is found that the neural network is able to achieve a sufficiently high degree of accuracy when fitted to and tested with all the image pairs generated from the computational physics-based approach. The crack pattern similarity label, the damage severity label, and the overall similarity label achieve an accuracy of 87%, 82%, and 69%, respectively. However, the generalisability experiments on the neural network that consist of predicting the similarity of a type of crack pattern image pair that is not included in the fitting data set, show very poor performance with respect to the prediction accuracy of the similarity labels. When the neural network attempts to predict the similarity of Pattern ID or a façade geometry that it did not see in the fitting procedure, it predicts all three labels with an accuracy that varies from 40% to 50%. Additionally, the neural network is also fitted to images generated from the computational physics-based approach and then tested with a pool of image pairs generated from the statistics-based approach, computational physics-based approach, and real-world images. The average accuracy with which the three similarity labels are predicted is even lower, lying between 25% and 40%.

This MSc thesis concludes that the neural network fitted to data generated from the computational physics-based approach and assessed by all the raters is able to predict the crack pattern similarity label, the damage severity label and the overall similarity label with sufficiently high degrees of accuracy. However, the generalisability experiments on the neural network show very poor results. This indicates that in order to achieve a greater prediction accuracy, the neural network may need to be fitted to a considerably larger sample of crack patterns that covers all of the relevant situations. Furthermore, the substantial inter-rater variability in the labelling of crack pattern image pairs suggests that even an ideal neural network architecture may not be able to overcome the inconsistencies in the fitting data.

The implementation of timber as a load-bearing and stabilizing material for high-rise structures has significantly increased over the last decades due to its potential of reducing the environmental footprint of a structure. Timber structures present a reduced global stiffness and self-weight compared to structures incorporating traditional construction materials, making them prone to high accelerations caused by wind load which affects the structural integrity and user comfort. The acceleration of a structure is a highly complex parameter that must be determined using modal analysis computed by a full-scale finite element model, and is significantly influenced by the magnitude of the wind and vertical loads, as well as the global stiffness of the structure. During the preliminary design phase, it is the responsibility of structural engineers to determine the feasibility of the design and estimate the amount of material required and the distribution of the structural elements. However, this process must be done under a limited amount of time, forcing engineers to rely on rules of thumb and previous experience, which are not currently available for the design of timber high-rise structures. Therefore, the goal of this investigation is to perform a parametric study to determine the influence of various parameters on the design of timber high-rise structures and assist structural engineers in making well-argued decisions during the preliminary design phase. The preliminary design parameters to be studied in the parametric study are: stability system design, connection stiffness and building height. The ranges for the parameters were selected based on extreme values such that clear trends regarding their influence on design could be visualized. This decision limits the design feasibility of some configurations studied in this investigation. The selected stability systems to be studied are glulam frame, CLT core and glulam diagrid. The design verification for each design alternative is done using ULS and SLS criteria provided by Eurocode, as well as a simplified method to estimate the dynamic response of the structure. Finally, the sizing of the structural elements and data collection was done by the implementation of evolutionary algorithms. From the data collected it was determined that the most influential design parameter for the design of timber high-rise structures is the stability system selection given that it determines the global stiffness, which showed a significantly higher influence on the dynamic response of the structure than its self-weight. Moreover, the efficiency of the different stability systems was assessed based on the maximum slenderness they could achieve compared to the amount of material they required. Based on this definition, it was determined that the glulam diagrid is two times more efficient than the CLT core and three times more efficient than the glulam frame. The efficiency of the glulam diagrid is caused by its high global stiffness provided by the triangular configuration of the diagonal elements. These observations prove that the effect of timber’s low density can be mitigated by the implementation of efficient stability systems. The influence of connection stiffness was determined to be directly related to the ability of the stability system to provide lateral stability to the structure. The influence of this parameter on form stable structures such as the CLT core and the glulam diagrid, proved to be nearly negligible, while for non-form stable structures such as the glulam frame it proved to be highly influential. However, it was also determined that the addition of rotational stiffness in the connections causes an exponential increase of the costs and environmental footprint of the structure, which decreases their design feasibility. Finally, the influence of building height on the design is visualized by its effect on the dynamic response of the structure caused by the logarithmic increase of the wind speed as this parameter increases, as well as a decrease of the global stiffness proportional to the efficiency of the stability system.

Anticipating the benefits and potential of using Wire and Arc Additive Manufacturing (WAAM) in the construction field for structural connections to enhance automation, energy efficiency, and material utilization rate, the process to provide a reliable and efficient prediction for the strength of WAAM components is the focus of this thesis. Finite Element Analysis (FEA) is proven reliable and can be used as a method for strength prediction. ABAQUS has been chosen as a simulation tool due to its automatic and adaptive ability for computation and the dominancy in academic research. Current WAAM components research builds WAAM models with the heat source and printing path, which took an unacceptable time (could be several days) to simulate the behavior (distortion and residual stress) for meter-scale structural WAAM steel components. A semi-analytical model, which implicitly incorporated the behavior of the WAAM material by the direct use of components' mechanical test data as the FEA inputs, can speed up the strength prediction process. This research is built with a feasible process with unique steps for predicting WAAM components' strength. First, WAAM components’ lab tests from Van Bolderen [1] have been reviewed and numerically examined before being used for designing WAAM joints , which makes the data used more reliable. Four lab tests that have been reviewed and simulated are the tensile test, the compression test, the bending test, and the buckling test. Examined material non-linearity, anisotropy of the material, and the surface roughness have been applied to predict the strength of the joints. Five aims and five actions have been set and implemented in the design of WAAM printed gridshell joints. Moment-rotational curves for three joints are presented to acquire flexural stiffnesses and moment capacities. Second, the inclusion of the defined geometrical imperfections (the surface roughness and the lack of straightness) and material imperfections (anisotropy) to get better simulation results to yield closer results to lab experiments. The surface roughness is derived from the bending test, the lack of straightness is incorporated with the failure mode found from Linear Buckling Analysis (LBA), and the anisotropy is evaluated by building orthotropic material with the young's modulus from two WAAM printing directions (transverse and longitudinal) in the elastic range. Last, the codified guidance for WAAM printed components, which are defined as shell structures in EN 1993-1-6: Eurocode 3: Design of steel structures - Part 1-6: Strength and stability of shell structures, have been implemented. From the numerical simulation of four lab experiments, the tensile test simulation results show only maximum 0.5% errors (while compared yield strength, ultimate strength, and young's modulus to the lab result). Therefore, the material non-linearity from the lab test has been correctly used in the numerical simulation. The compression test results indicate the material non-linearity used in the three-dimensional objects is valid since the compression stress-strain curve fits the tensile engineering stress-stain curve well before the yield point (0.41% of errors with the comparisons of yield strength and young's modulus). The bending test can not precisely infer the errors of derived effective thickness mainly due to the lack of friction coefficient. However, the inclusion of the effective thickness reduces the error from 22.0% to 16.8%, and limited to 16.8% when compared to the lab test results, despite the fact that the value is from assumptions. The buckling test shows large buckling load errors (35%) that might due to the hand measured lack of straightness is too less, the effective thickness derived are too thick, or the defined geometrical imperfections are too simple that can't reflect the real printed conditions, which has local geometrical imperfections and deviations. The high sensitivity of the geometrical imperfections to the buckling load has also been found out from the analysis of scale factors with buckling loads. Few fraction increase of lack of straightness (0.2% to 0.4%) leads to tens of percentage error drop (19.6%) of buckling load compared to the lab results. Examined data that have been used in the analysis of joints are material non-linearity, the anisotropy of the material (build-in orthotropic manners), and the surface roughness. The target values of the flexural stiffness and moment capacity are insensitive to the lack of straightness due to the fact that the instability of the joint does not come from the buckling effect, the WAAM printed parts are not in slender shape, and the lack of straightness is small (L/791). Three WAAM printed joints have been designed, and the strength prediction have been carried out with the joint classification. At the end of this thesis, the conclusion is that there is potential to provide reliable strength prediction of the designed WAAM joints with the use of EN 1993-1-6: Eurocode 3 as guidance for building semi-analytical models for WAAM components and joints with the lab test material data and defined imperfections for two reasons. First, the simulation results yield closer values to the lab test results with the shell analysis process in EC3. In the compression test simulation, the inclusion of the possible geometrical imperfections shape found from LBA reduces the percentage errors decrease from 52.27% to 13.75%. In the bending test simulation, the inclusion of effective thickness leads to a closer value to the lab test (from 22% to around 16.8%). In the buckling test simulation, the implementation of effective thickness for surface roughness reduces around 10% to 15% error, and the inclusion of effective thickness combined with lack of straightness reduces roughly 40% of the error. Second, the parameters are examined before being used in the strength analysis of designed WAAM joints. The material non-linearity data is proven valid in the FEA with 3D geometry in tensile test and compression test simulations. The orthotropic behavior shows a minor effect in all four numerical simulations. The surface roughness combined with the use of assumed friction coefficient and the assumed equipment properties in the bending test simulation show limited error (around 17%). However, limited errors exist and undermine the strength prediction precision due to the assumptions of lacked lab test measurements and the simplification (parameterized) imperfections from WAAM components. Therefore, It is expected that more tests and works could be made to improve the strength prediction precision of WAAM components. Also, the FEA setup of the designed joints is referred to as the gridshell joints' lab experiments, therefore, providing a methodology for possible future examination of the strength of WAAM printing joints with lab experiments.

This research investigates if confining rubble stone masonry by timber bands and columns increases resistance against earthquake loads, by performing a number of numerical analyses in the finite element software program Diana. In order to establish a reliable model, the input parameters are investigated by means of a literature study and sensitivity study. Additionally, the numerical model is validated by comparing the results to an analytical study. From the analytical study it is determined that the failure mechanisms are correctly estimated by the numerical analysis. However, differences between the values of ultimate strength and ductility were observed. The effect of the confinement is investigated by a pushover analysis on a shear wall with two different masonry tensile strengths: ft = 0.01 N/mm2 and ft = 0.03 N/mm2. These values are selected to show how such a small difference in tensile strength results in a different failure mechanism of the wall, and therefore results in a vastly different displacement capacity and ultimate strength. Additionally, a shear wall with and without a window opening is studied. If the building is constructed with extremely low-strength masonry (ft = 0.01 N/mm2), the timber frame confinement will increase the resistance of the wall (with or without window opening) against the pushover load. If the building is constructed with masonry having a tensile strength of 0.03 N/mm2 or higher, the confinement has a negative impact on the ductility of the closed wall. For the wall with window opening the confinement triples its ultimate strength. Weather a strong or a ductile structure is more desirable, depends on the demand with respect to the seismic spectrum. If the ground motion demands a strong structure, it is advised to confine the masonry with a timber frame consisting of four timber bands, and columns at each wall junction. The design of the timber frame as recommended by the Nepali building codes is determined to not be sufficient and must be altered in order to provide this positive impact on the structure’s resistance. Firstly, the columns must be placed at both sides of the band, instead of on the inner side only, to avoid eccentric loads on the bands. Secondly, the cross-sectional dimensions of the bands and columns must be increased avoid failure of the connections and splitting of the timber. Taking these aspects into account, a new design for the confinement method is presented in this study. The limitations of the conclusions of this research follow from the investigation of a single, in-plane wall only. One of the goals of the use of bands is to improve the box behaviour, for which the out-of-plane performance must be investigated. Moreover, an analysis on a three-dimensional structure is needed to fully answer the research question. In a three-dimensional study, the closed walls and walls with opening give a combined response to the load, therefore, the advantages and disadvantages of the confinement are combined as well.

Increased traffic loads and ageing of concrete bridges and overpasses in the Netherlands make it necessary to reassess these existing structures. Consequently, the current condition and capacity of many concrete structures need to be evaluated. Residual capacity could be discovered during reassessments of concrete slabs due to a phenomenon called compressive membrane action (CMA). CMA is the formation of internal compressive arches caused by the lateral restraint. As a result, the load is not only transferred by bending action but also by arching action. Research has shown that the ultimate capacity can be significantly increased due to the occurrence of CMA. The goal of this study is to examine the influence of geometrical nonlinearity on this increase in capacity. Also, accurate quantification of the capacity enhancement for a variety of concrete slab variants can be scientifically useful and increases the knowledge on CMA. The study is confined to one way reinforced and restrained concrete slabs.

A new analytical model is presented to quantify the capacity enhancement due to CMA and the geometrical nonlinear (GNL) effect on the capacity. Calibration of the analytical model is performed with a finite element model in DIANA FEA. Also, the finite element model validates the analytical results and is used to study the failure mode of a restrained concrete slab in detail.

The enhancement factor – defined as the enhanced capacity divided by the conventional capacity – turned out to be varying between 1.35 and 4.7 for a large variety of restrained concrete slabs. Thus, the ultimate capacity of restrained one way slabs is significantly increased due to CMA. However, the capacity enhancement would have been even greater if geometrical nonlinearity was not accounted for. Geometrical nonlinearity reduces the increase in capacity because the formed compressive arches will tilt as a result of deflections, therefore leading to a relative decrease of the resisting arching moments. This GNL reduction effect varies between 3% and 37% according to the finite element model. The calibrated analytical model sufficiently estimates this effect with a maximum deviation of about 12%. An important finding was that the enhancement factor is larger for deep slabs than for slender slabs, while the reduction of the ultimate load due to geometrical nonlinearity is larger for slender slabs than for deep slabs.

Thin-tile vaults are a a type of vaults that went out of fashion in the early twentieth century. Its origins are around the Mediterranean, but modern interest is mostly due to Guastavino, and the research done at MIT and ETH. The thin-tile vaults have a unique construction method without any temporary support. Eventually the increase in labour costs and the advancements in concrete and steel made the structure non-competitive.Robotics are a type of machines that can perform (semi)-automated tasks. In the past decades the development of robots have led to their implementation in the construction industry. Robots developed specifically for masonry show a high promise where they're able to lay much more bricks than even the most skilled mason.This research aims to investigate the time it takes for a robot to construct a thin-tile vault, and thereafter to advance the possibilities of research into and construction of the thin-tile vault. This is researched by answering the following question:How does a robotic construction of a parametrically designed thin-tile vault perform based on step-wise structural analyses?To answer this question a parametric model has been made to include the design of the thin-tile vault, the structural analysis and the robotic construction. This model is made in Grasshopper, but is supported by Python and RoboDK. Python performs any calculation necessary and analysis of this output. RoboDK is a robotic simulation program aimed to give users a tool to translate their design to robotic instructions. The structure of this report is based on the three aspects of the parametric model.In the first part the state of the art and the relevance of this research is stated. Thereafter the objectives, questions and methodology are noted.The first part of the model is related to the design. In the first chapter a literature review is provided on thin-tile vaults and similar vault structures. The second chapter of this part describes how the design model has been made. First the global shape of the vault is set as a barrel vault formed as a catenary arch. Then the bricks are placed on this vault surface using a similar approach as map projection. This is done with the introduction of the centre point approach. This ensures that proper information is maintained, like the course and the position within the course per brick. With this approach it is possible to fill this surface even with courses misaligned to the primary curvature. Bricks at the edge of the vaults surface are cut to fit within its domain. The last step is to thicken the bricks, which thus far had been represented as surfaces, into volumes.The structural analysis consists of three major parts. In the first chapter literature is shown to justify the use of a linear elastic analysis on masonry. Additionally the behaviour of this uncommon type of masonry is tried to be found in literature. The materials brick, mortar and epoxy are looked into as well. Bricks have time-independent properties, making these similar as found in the Eurocode, by manufacturers and in research. Mortar and gypsum plaster especially (also known as Plaster of Paris), are described in their material properties as well. Due to uncertainty in these properties, epoxy is investigated as well and found to have more theoretical values useful for this research. A small-scale experiment with bricks and epoxy has been done to verify what values from epoxy should be used.After the literature review on material properties, the structural analysis is done. The flow of this chapter starts with the form from the design model, through the forces of a cantilever, to the stresses occurring in the structure. The vault has been simplified to an arch during construction. Thrust is not present. The sectional forces are found by first considering the (partial) arch as a cantilever with a vertical shear force and rotational moment. Next this shear force is converted into the normal and shear force in the cross-section of the arch. This calculation has been worked out in a Python script.In the last chapter a possible stress distribution is shown. The analysis is based on a phased structural analysis. Each time step is equivalent to the placement of one row of bricks, which have a similar cantilever length. The stresses of the bottom and top side are shown, as well as the stresses between the wythes of the vault. In a DIANA model it is shown these wythes behave more like a monolithic material than a layered one. The last part is related to the robotic construction. First literature is shown related to masonry robotics. In the second chapter the stations the robot visits, are described. At the pallet station three instructions are described. The adhesive station has a number of instructions which is dividable by four. The vault station again has three instructions. This is similar to a pick & place work order, but with the adhesive station in between. Two configurations are considered. The first is when the robot is outside the vault, where it is closer to the support than to the apex. In the second configuration the robot is placed within the vault, where it is positioned close to the apex. In the last chapter the robot is chosen. Additionally the path is described with boundary conditions.The results are shown for 8 computations: a basis computation; one with the robot changed; one with the shape of the vault changed; one with the orientations of the wythes changed; one with an increase of the adhesive hardening time; one with a small course length of the vault; one with a different work space configuration; and lastly one with a different preferred construction sequence.In conclusion the models and results show that the construction of a thin-tile vault is similar to other masonry robotics, if the hardening time is excluded. Included, the number of bricks built per day is similar to two masons building a thin-tile vault prototype. The hardening time is the primary influence of the total construction time and any reduction here is advised. Further optimisation of the movement of the robots is possible with a more in-depth optimization of the work space configuration and an optimization where the shortest movements are considered across instructions. The linear-elastic structural analysis has been found to be possible to use for the calculation of the thin-tile vault during construction. Additionally, two construction sequence preferences work well with the multiple wythes, but one preference shows better results when more bricks with more cantilever length need to be placed, due to reducing the stresses on the edge of the extrados wythe.Further research into thin-tile vaults is still required to fully understand and model the thin-tile vault. Related to the three models (design, engineering and robotics) each have been improved in the models made for this research, but are still a long way away from full implementation.

The city of Amsterdam contains about 1600 bridges and 600 kilometres of quay walls. Of these walls, about 200 kilometres are of masonry walls placed on a timber floor and founded on timber piles. These quay walls are sometimes over 100 years old. Due to the increasing loads in the past century and the degrading of material properties in the masonry wall and timber elements, the quay walls are in bad shape. When designing a quay wall, a cross-sectional analysis is used to calculate the desired dimensions to withstand the loads. When this calculation is performed on a quay wall over a 100 years old, containing a failing pile foundation, the quay wall should fail. However, many of the quay walls under the condition of a partly failing foundation, are deforming, but still standing. During recent years, at 16 locations in Amsterdam, the risk of collapse appeared imminent, and emergency structures are put into place. Possible practical measures are removing trees on the quay walls, traffic limitations in the city centre and placing temporary struts and sheet piles to provide stability. Since the scale of the problem in Amsterdam is large, the time to renovate all the quay walls is lengthy. Therefore, there is a need for knowledge on the state of quay walls at the end of their life phase when partly failing. Different failure mechanisms occur, and various measures are developed to control those and provide (temporary) stability. This research answers the question: How can 2D analyses of quay walls, in multiple directions, under the condition of a partly failing foundation, provide insight into the hidden structural capacity within the masonry work? The study focuses on the severeness and scale of the foundation defects in the quay wall's cross-sectional and longitudinal direction. For the longitudinal models, the effects of the masonry material qualities are studied by using different material properties. Also examined is the effect of the failing foundation pile's post-peak behaviour, modelled as brittle and checked for plastic behaviour. Finally, the relevancy of the timber floor is studied for a stiff continuous floor and most notable, the full removal of the floor. To study this, two 2D regular plane stress, nonlinear elastic, finite element models are created in Diana FEA. The foundation piles are modelled as nonlinear elastic springs via a force-displacement diagram. The foundation piles' defects are modelled by assuming a smaller pile diameter, resulting in a weaker force-displacement diagram and larger displacements in the quay wall system. The foundation defects can be scaled over a small or big area by adapting multiple foundation piles over the length of the quay wall. The masonry's behaviour is researched by using a macro material model using smeared material properties for the brick and mortar, resulting in a continuous material. The material model used is the Total Strain Rotating Crack Model, which can be used in a 3D analysis of the quay wall system in future research. Finally, the interface between the timber floor and masonry is modelled using a coulomb friction interface criterion. This simulates the effects of the mortar layer connecting the timber floor and masonry work in a quay wall. The results conclude that analysing a foundation defect in the cross-sectional direction of the quay wall results in instability of the wall without further horizontal and vertical constraints to keep the quay wall in place. Modelling the pile foundation defects using a reduced pile diameter and consequently, a decreased force-displacement diagram as spring input provides the model with temporary stability. Ultimately, the cross-sectional analyses contribute little knowledge on residual strength and hidden structural capacity. Separately, the longitudinal model implements a vertical constraint in the masonry by using the bending capacities of the material. The results present an expected correlation between the scale of the foundation defects and the vertical displacements. Similar to the cross-sectional analyses, the reduced pile capacity of the foundation piles provides the model residual strength compared to the situation where the total failure of a timber pile is used. The timber foundation pile's failure mechanism needs to be researched in-depth since the results present a notable difference for crack patterns and force-displacement curvatures when modelled brittle or plastically. For brittle failure, a horizontal crack forms at the tip of the central, vertical crack, due to the abrupt enlargement in vertical displacement of the quay wall. The functionality of the timber floor in the longitudinal analyses presents itself when the crack patterns are analysed. The presence of the timber floor results in multiple smaller cracks instead of a single large crack when foundation defects of the quay wall system are analysed without a timber floor. It can be concluded that the masonry quality, most notably the tensile strength, affect the results significantly in terms of maximum values in the force-displacement diagrams and crack development. The material properties are based on Groningen masonry experiments, and it is recommended to perform experiments to the masonry quality of Amsterdam quay walls. Finally, the observed displacements related to the intervention points of the municipality conclude that foundation defects result in cracks for displacements below the marking points of 20 and 25 millimetres. For weaker masonry, the quay wall fails before the indication values. It is recommended to perform more measurements to the quay walls in Amsterdam and study the reliability of the intervention points.

Each year, the province of Groningen experiences many induced earthquakes due gas extraction, which has been ongoing since 1963. The earthquakes cause damage to the buildings situated in the Groningen area, and they constitute a potential danger for the safety of the residents. These buildings are typically unreinforced masonry structures which are designed without knowledge of the presence of seismic activity in this area.

It is therefore essential develop and use assessment methods that are on one end reliable and accurate, but on the other hand allow to perform a large number of assessments of the vulnerability of the buildings in a short time. In other words, the assessment of all the buildings requires a quick and reliable assessment method. Such an assessment method should offer a strong understanding of the occurring failure mechanism during an earthquake, an acceptable prediction of the ground acceleration at which the collapse of the building may occur (maximum base shear force) and the displacement capacity of unreinforced masonry (URM) building.

The NPR9998 recommends four seismic assessment approaches, which differ in complexity and assessment time needed to be performed. The most comprehensive and time-consuming assessment method is the NLTHA (nonlinear time history analysis), which includes both the dynamic and nonlinear effects. In practice, this method is used only in special cases, such as in the case of monumental buildings. A simpler approach is the NLPO (nonlinear pushover) analysis, which is static and considers the nonlinear properties of the structure. An NLPO is less time consuming than an NLTHA, even when the finite element method (FEM) is considered.

A more simplified approach is the Simple Lateral Mechanism Analysis (SLaMA). This method is a simplified mechanism-based analytical approach. If the SLaMA method predicts realistically conservative global capacities, it could serve as an effective alternative assessment method for URM buildings, and especially to the NLPO FEM analysis. This study focusses on the comparison between the SLaMA method and the NLPO FEM analysis. Therefore, this study aims to answer the following research question:

Could the SLaMA method be a realistically conservative and effective alternative to the NLPO FEM analysis in making a seismic assessment for two-storey unreinforced masonry buildings?

In conclusion, the SLaMA method could be a realistically conservative and effective alternative to the NLPO FEM analysis in predicting the maximum base shear force. The displacement capacity predicted using the SLaMA method is validated only for buildings with RC floors. This predicted SLaMA method was realistically conservative compared with the ultimate displacement achieved using the NLPO FEM analysis. The SLaMA method is overall suitable for obtaining a quick understanding of the behaviour of an URM building. However, it requires a proper evaluation of the analyses to identify properly the type and the location of the failure mechanisms. For this reason, this method could be valuable to be applied before using a more complex assessment method.

The seismic assessment of unreinforced masonry structures in Groningen is still ongoing. The assessment is vital to determine whether a building must be strengthened or not. Different assessment approaches have been followed in recent years. First, the Non Linear Time-History analyses (NLTHA) were initially the only approach used for the seismic assessment. They are still overall the most accurate type of assessment, but they are also the most time-consuming one. Nowadays, NLPO analyses are more frequently used. This assessment procedure presents some limitations of application and may be less accurate for complex structures but it requires less computational time. The NLPO analyses can be performed by means of different tools, such as analyses based on the finite element method (FEM), equivalent frame (EF) or macro-element based analyses and, eventually, also analytical mechanism based analyses. The SLaMA method belongs to this last category: this method is an analytical approach already tested and validated in New Zealand for RC structures.
This research aims to answer the following research question:
• How is the in-plane behaviour of single-storey URM wall facades affected in simplified calculation methods compared to FEM when geometrical irregularities are present?
The walls have been modelled in 2D with three different methods: FEM, EF and SLaMA. Material properties and modelling assumptions were maintained as consistent as possible within the three different methods. For researching the influence of the geometrical irregularities on the accuracy of EF and SLaMA when compared to FEM, the variation of geometrical irregularities, each quantified by an index value, have been studied. The influence of these indices on the accuracy of the calculation methods has been researched with a sensitivity analysis.
The objective has been pursued by looking into single-floor URM façades, and the conclusions of this research can be applied to this typology of walls in Groningen made of solid clay brick masonry (pre 1945). The study focuses specifically on the base shear capacity of the walls.
The differences observed when comparing the in-plane behaviour of a wall analysed with 3MURI and DIANA are not significantly affected by the presence of geometrical irregularities. The ratio between the base shear capacity computed with the two approaches and the predicted failure mechanisms remains consistent for all geometrical irregularities defined in this report.
Similarly, the differences observed when comparing the in-plane behaviour of a wall analysed with SLaMA and DIANA are not largely affected by the presence of geometrical irregularities, since the base shear computed according to SLaMA is consistently lower than that obtained with DIANA. However, the base shear capacity obtained with SLaMA showed large variations between 0.34 and 0.75 with respect to DIANA when implementing geometrical irregularities. The largest variation is obtained when more than a single pier is considered, due to the inability of SLaMA to define the re-distribution of the vertical axial forces in the piers, nor correct boundary conditions at the top of the piers since the constraining action of the spandrel appear underestimate. This affected also the prediction of the failure modes, which differed for the two methods. However, in most of cases flexural failure mode was obtained, and the study should be extended to consider also geometries and loading conditions that cause also the shear failure of the walls.

In light of the global attempts to reduce material use by the construction industry, this research focuses on combining topology optimisation with additive manufacturing of steel. It is investigated whether an automated procedure can be developed to generate reliable strut and tie models for reinforced concrete elements, while satisfying the constraints that apply to 3D-printing using the Wire and Arc Additive Manufacturing(waam) technique.
Additive manufacturing offers a fully automated production process where a large freedom in form can be achieved. Topology optimisation concerns with finding a good material distribution within a prescribed domain. A literature review was performed on current developments regarding both subjects. It was found that the waam-technique is very suitable for printing reinforcement designs. Sufficiently large models can be printed, and material properties can be achieved that match the properties of traditional reinforcement steel. This manufacturing process is expected to produce functional structures that can readily be used as reinforcement steel in buildings. Two main manufacturing constraints should be accounted for during design of the model. A minimum member inclination and a minimum member diameter are both expected to be necessary to ensure a smooth printing process.
Several different topology optimisation algorithms are discussed in the second part of the literature review, and it is determined which algorithm is most suitable to continue with in the rest of this research. Examples are presented that explain the functionality of three important optimisation schemes: Bi-directional Evolutionary Structural Optimisation(beso, Solid Isotropic Material with Penalisation(simp) and Ground Structure Optimisation(gso). It was found that all three can be used to analyse reinforced concrete. Each algorithm has advantages and disadvantages, so there is no obvious best choice. However, motivated by the easy access to member forces and availability of a very good Python implementation, it is chosen to use gso for the remainder of this research.
This Python script was modified to include the constraints that come with an additive manufacturing process. It was found that the minimum member inclination can straightforwardly be included. The new function that was proposed allows the user to specify a minimum inclination, and ensures that no members are generated within the design domain that violate this minimum angle. Experimenting with this new function revealed cases where material use increased significantly when this function was used. This lead to development of an alternative procedure to ensure a printable design. In this alternative procedure, an optimisation without any angle constraint is performed first. Then, in the form of a post-processing script, The complete model is rotated around two separate axes in an attempt to find a suitable printing orientation.
The third and final proposition that was done in this part, consists of a post-processing script for the minimum member diameter. Including this minimum diameter in the optimisation would require rigorous changes to the optimisation script. Therefore it was chosen to investigate the performance of this post-processing script first.
The case study that was performed in the third part of this research, proved that this post- processing script for member diameter is sufficiently efficient for practical implementation. Together with the ability to slightly suppress the amount of members that are generated in the design domain, the printing constraint for minimum diameter could relatively easily be enforced. A bigger challenge lies within ensuring the minimum member inclination. The 60◦ minimum that was set, proved to be very harsh on the solution space. In the example from the case study, no printable model could be generated without significantly reducing the material efficiency. However, it is argued that this minimum inclination constraint can possibly be relieved by recent developments in additive manufacturing techniques. An example of this could be a rotating printing surface, that has the potential to remove this angle constraint completely.
Overall, the experience of combining topology optimisation, additive manufacturing and strut and tie modelling has been predominantly positive throughout this research. The combination of a state of the art manufacturing technique and a more performance driven design process with a labour intensive traditional calculation procedure has shown promising first results. In the example in this research, 30% less material was required to accommodate the tensile forces in the concrete.

Since the 60’s the Nederlandse Aardolie Maatschappij (NAM) has been extracting gas from the province of Groningen in the Netherlands. This resulted into seismic activity. However, the buildings in Groningen are not designed to withstand any seismic loading. The assessment of the seismic behavior of the building stock of Groningen is required to verify whether a structure causes any life safety risk during an earthquake. In this thesis two main objectives are studied. Firstly, it is studied whether the simplified analysis approaches: Simplified Lateral Mechanism Analysis (SLaMA), as described in the NPR9998-2018, and the Equivalent Frame Method (EFM) as implemented into the software package 3Muri are able to describe the seismic behavior of an Unreinforced Masonry (URM) terraced house. Secondly, the influence of the geometry of the piers in a facade is assessed. Both the objectives are studied by means of two case studies. The first case study represents a typical but idealized URM structure of a terraced house from which the seismic capacity was determined in the TU Delft lab by cyclic pushover tests. The second case study concerns a specific two-storey URM terraced house located in Groningen. The case studies are characterized by large daylight openings, slender piers and a low lateral capacity into the x-direction.

Geometric patterns inspired in historical Islamic ornamental art have attracted the attention of contemporary designers worldwide. The large variety of shapes, symmetries and combinations is a source of inspiration, but at the same time makes it difficult to provide general rules when used in a technological application. Structural small-scale applications have been object of research, but there are potential applications at larger scales that deserved attention. In view of some recent proposals of Islamic patterns as structural grids for tall building skins, the question of their structural efficiency, especially compared to conventional grids, arises. The purpose of this research is to assess the performance of structural grids based on geometric Islamic patterns as outer skins of tall buildings. For this purpose, several historic patterns have been classified. An equivalent meta-material has been defined for each pattern, based on the homogenization method for a series of saturations or beam sizes. Their response in different orientations has been studied to identify their structural behaviour based on the pattern geometry. Their relative performance has been assessed for all patterns against themselves and against the conventional diagrid system. All this process has been collected and summarized in a predesign tool made of graphs, pictures and tables Finally, the predesign tool accuracy has been assessed and applied to three tall buildings. All those steps have been structured in three distinctive levels: At the method level, the conclusion is that the developed predesign tool is a success as it provides a higher level of accuracy than modelling all the beams. It is also faster and easier to implement, than modelling all the beam elements, to compare alternatives in early stages as the complexity of modelling the patterns is postponed to later stages. As the saturation decreases and the effective beam length influence in the beam model results diminishes, the beam model will become more reliable than the predesign tool and vice versa. At the pattern level, the most interesting finding is that the patterns with square symmetry (symmetry directions at 90º) display a perpendicular isotropic behaviour, whereas the patterns with pentagonal symmetry (symmetry directions at 72ª) display an orthotropic behaviour, and the patterns with hexagonal symmetry (symmetry directions at 60ª) display an isotropic behaviour. It has also been studied the effect that would have filling the stars as an alternative to building the patterns as an assembly of beams. At the building level, it has been found a few geometric Islamic patterns that could be suitable alternatives to the conventional diagrid systems, a pattern with a similar performance and even a pattern with a higher structural performance than the conventional diagrids. This highly performing pattern is currently been used for some architects such as Shigeru Ban in their parametric designs. In this regard, it can be concluded that the objective of finding suitable alternatives to conventional diagrid systems has also been a success and it can affect some designers engineering judgement. The homogenization process obtained an equivalent ideal material corresponding to a plane infinite panel that will not correspond with the built structural grid. The use of complex geometries and its application to tall buildings introduce effects not considered in the homogenization that will disrupt the expected structural performance. Those effects are minimized in the case of other shells structures such as domes but can be important in the case of tall buildings. It is not advisable to account for the squeezing effect by adapting the saturation with the change of the modulus size in the x-direction as the relative beam depth has a greater impact in the overall stiffness than the change of geometry due to the squeezing effect. The distortion effect cannot be accounted for directly and it depends on the angle of the distortion and the pattern. However, in the studied case it has been found a required correction factor of 1.2-1.3, in line with other uncertainty factors used in practice. Finally, the intermediate supports can have a great influence in the final drift. It depends on the pattern used and the number of diaphragms inside the module. Nevertheless, the use of intermediate supports is always beneficial and not considering them will always lead to more conservative solutions. In conclusion, this document successfully bridges the knowledge gap regarding the structural behaviour of historic Islamic patterns, with comparative tables. It identifies the best performing patterns and their best orientation, and it provides a useful tool for the decision making in the design process of in-plane bearing geometric Islamic patterns.

Masonry is one of the most commonly used construction materials for residential buildings and historic buildings around the world. Some of these buildings are located at seismic zones, while unreinforced masonry structures are vulnerable to seismic loads. To assess the existing masonry buildings and to design new masonry structures, nonlinear seismic simulations are conducted with macro modelling or micro modelling approach. The macro modelling approach, which smears out the details of the bricks and joints as a homogenous material, can efficiently and robustly model complete masonry structures. A commonly used orthotropic constitutive model is the Engineering Masonry Model of the DIANA FEA, which is based on the Total Strain Method that eliminates the mapping-back process in conventional elastoplastic constitutive models. The micro modelling approach, which explicitly models the bricks and joints, can better represent the mechanical behaviours of masonry. However, most of the constitutive models used in micro modelling are based on elastoplasticity that usually causes numerical difficulties due to its mapping-back process. The lack of a robust constitutive model has severely hindered the application of this accurate analysis approach.

So, this thesis proposes a sub-increment based iterative constitutive model for interface elements, based on Multi-surface Plasticity Criterion. This model aims to enhance the robustness and accuracy of the constitutive model used for micro modelling. It eliminates the conventional mapping-back process in elastoplastic constitutive models by introducing the ideas of sequential uni-axial loading algorithm and an extra damage iterative calculation algorithm. These algorithms are robust even when the stress state is at the corners of the yield surface. The model also introduces the concept of sub-increments to consider the path dependency in plastic process. All the formulations of this constitutive model are derived based on a simple mechanical model. Formulas and examples are provided for obtaining the input parameters from material tests. The proposed constitutive model is tested on a single integration point level and found to be stable and reliable. It is further applied on the component level, by modelling three masonry walls of different dimensions and boundary conditions, under cyclic loading. For the verification of these wall models, the numerical results are compared with the experimental results in terms of force-displacement curve and crack pattern. Finally, the thesis presents a brief study on parameter sensitivity to provide guidelines for the level of accuracy needed for each input parameter, in order to get satisfactory numerical results.

The constitutive model is found to be robust for all the wall analyses conducted, without encountering divergence. The comparison between numerical results and experimental results shows that this constitutive model can cover the majority of shear and flexural failure mechanisms and mimic the crack patterns well. It is capable of modelling shear failure with high accuracy. It can also model flexural failure well with a few parameters calibrated. The fact that the model is little sensitive to parameters that are hard to be measured from experiments, such as tensile strength and tensile fracture energy, ensures its feasibility in engineering practices.

The city of Amsterdam has a large number of old quay walls with rotten foundation piles. These foundation piles need to be identified and measures need to be taken. The urban quay walls are supported by two types of foundation: pinewood piles, which are easily affected by bacterial decay and spruce piles. To understand the mechanical behaviour of quay walls better, it is needed to know the type of wood used for each pile foundation along the 200 km of quay walls currently showing signs of damage. For that reason, specialized diving teams are hired to identify the rotten piles and foundation defects, to know which foundation piles need replacement. Since the area is very large and diving inspections are costly and lengthy in time, there is a need to correlate the foundation defect to the masonry damage above the water level. The masonry above the water level could give lots of information about the condition of the foundation, due to cracks or deformations in the masonry. This research could help to relate foundation defects with damage patterns in the masonry. Understanding this relation helps to identify foundation defects at an earlier stage and helps the municipality to prioritize the replacement of foundation piles. The thesis aims to find indicators above the water line to identify foundation problems by studying the crack patterns in a typical unreinforced masonry quay in Amsterdam. From the point of view of the masonry structures, failure of foundation piles results in a settlement deformation causing cracking. This research will support the current work by Sweco in helping to find foundation defects from above the waterline via masonry damage patterns in quay walls. This will be achieved by performing a parametric study, bases on 2D nonlinear finite element analyses, varying the extent of the pile defects, the material properties of the masonry and lateral boundary conditions for a selected representative base case. To simulate the damage in masonry, a smeared crack approach was used. The foundation defects were simulated by applying a settlement deformation to the quay wall. A Gaussian settlement deformation profile was imposed and the ratio between the length of the profile and the length of the quay wall was varied to simulate the failure of single or multiple piles. To capture the influence of the material properties of masonry (especially related to tension failure), three types of masonry were defined: weak, average and strong. The influence of the boundary conditions at the edges was checked by performing analyses with horizontally free lateral sides and with horizontally fixed lateral side. This is done to simulate the effect of arching in the structure. Eventually, the influence of the location of the foundation defect was analyzed by comparing a symmetric Gaussian settlement deformation with an asymmetric settlement. The analyses show correlations between the vertical displacement at the top of the structure and the length of the settlement profile. As expected, this can be interpreted with the fact that if several piles are damaged simultaneously, a larger portion of the quay wall is cracking. The material properties of the masonry influence the development of crack patterns. The stronger is the masonry, meaning increasing the values of Young’s modulus, tensile strength and fracture energy, the larger is the settlement displacement needed to obtain the same crack pattern. The lateral constraints contribute mostly to the development of the horizontal crack since no horizontal cracks appeared in situations without these constraints. Since the influence of additional loads is not considered in the analysis and the model is modelled in 2D it is recommended to analyze the influence of both in further studies. It is also recommended to validate the model against field measurements since no verification has been done.

Gas-induced earthquakes are a major problem in the north of the Netherlands due to many unreinforced masonry (URM) buildings being located in this area. To improve the seismic performance of the URM buildings, existing structures must be retrofitted and if the aesthetic of the building is important, the implemented retrofitting technique must not alter the appearance of the property. An experimental campaign was conducted at Delft University of Technology to investigate whether bed joint reinforcement (BJR) can be used for seismic retrofitting of URM walls. In this thesis, the influence of BJR, diagonal anchors and reinforcement layout on the in-plane seismic behavior of the retrofitted wall is studied. To achieve this goal several nonlinear static analyses using DIANA software were performed. First, the case experimentally tested is adopted as a benchmark and a validation of the numerical model is performed. Afterwards, the validated numerical model is adopted to perform a parametric study considering different reinforcement layouts. The numerical study showed that BJR was able to improve the seismic performance of the retrofitted wall. The Peak-load was increased slightly (13%) in the retrofitted wall compared with the URM wall. BJR acts in tension to restrains the crack opening so, the maximum crack width is another affected factor: a difference of -110% in maximum crack width was observed after retrofitting. Finally, the crack pattern and failure mechanism of the structure was changed due to BJR. The sensitivity analysis showed that the results were sensitive to variation in tensile fracture energy and modulus of elasticity, while tensile strength, compressive strength, and compressive fracture energy were the other variations that did not affect the results considerably. According to an extensive parametric study, it was concluded that double BJR was slightly more effective than single BJR. By applying only 4 layers of double BJR almost the same peak-load and maximum crack width were obtained comparing the results of the retrofitted wall with the original layout (with 12 BJR layers). The location of the BJR, however, was the governing concern. In the URM wall cracks mostly developed diagonally from the window corners and by applying horizontal bars below and above the window opening, the diagonal cracks were restricted and failure modes of the wall changed. The absence of single BJR next to the opening (in piers) might lead to the shear mechanism of the piers. The number of BJR is a function of the opening’s dimension. According to a proposed reinforcement layout it is believed that the mentioned 4 layers of double BJR (above and below the opening) and 3 layers of single BJR (next to the opening) were a wise choice for retrofitting of the wall. Furthermore, diagonal anchors could be ignored as they did not affect the performance of the wall. Conservatively, a layer of double BJR far above and below the window level could be applied. In this way reinforcement was decreased by 30%, however, the results were comparable with the retrofitted wall with original reinforcement layout.