Lattice materials have been studied extensively for their advantageous mechanical properties like high stiffness-to-weight and strength-to-weight ratios. These properties make them suitable for a multitude of applications like load-bearing structures, energy absorption and impact attenuation. Amongst several planar microarchitectures, the Kagome topology is found to be optimal for effective mechanical properties like macroscopic elastic modulus. However, it has been shown that lattice defects such as nodal misalignments are detrimental to the macroscopic stiffness of the Kagome lattice. From a realistic standpoint, any manufactured lattice is expected to inevitably have these defects, which naturally makes the favorable stiffness properties of the Kagome inaccessible.

This work takes a step towards designing planar lattices with defect insensitive macroscopic stiffness. A novel design strategy is proposed for generating stiff, lightweight and robust lattice materials. The Kagome lattice is chosen as base architecture for the developed designs due to its excellent stiffness-to-weight property and extreme sensitivity towards defects. The generated designs involve appending the Kagome bulk with two unique and carefully designed microarchitectures. The first is a lattice developed using combinatorial design strategies, which demonstrates properties like perturbation-dependent switching between mechanism-states and load-bearing states. The second is a polarised lattice that enables localisation of mechanisms at specific sites of the lattice. This arrangement forms a "composite" lattice which when appropriately designed demonstrates robustness against manufacturing defects.

Development of such an architecture is carried out through an iterative design and analysis process which is based on reformulated matrix methods. A mass-spring model is used in contrast to regular pin-jointed models and a dynamical matrix method is developed as an alternative to traditional matrix analysis. Suitable designs formed out of this iterative process are verified for their stiffness properties using Finite Element models. Robustness against nodal defects is observed in well-designed lattice models. Satisfactory performance of such designs indicates the efficacy of the adopted design strategy which could potentially be applied to any lattice microarchitecture other than Kagome for robustness.

Printing patterns in additive manufacturing have been extensively studied regarding their effects on the microstructure and mechanical properties of the materials. However, weaving, a welding technique that produces good quality welds in large areas and ensures good weld penetration has not been applied in additive manufacturing. This study focuses on comparing four weaving patterns to two traditional line printing patterns in Wire arc additive manufacturing of austenitic stainless steel 316L. This thesis covers the groundwork of weaving pattern characterization, starting from optimization of printing parameters and printing path, followed by pattern characterization through measurements of voltage, current, thermal history, and thermal profile, and investigation of surface quality of the samples. Microstructure characterization was then conducted by optical microscopy, electron backscatter diffraction, and X-ray diffraction; and mechanical properties were obtained through microhardness and tensile testing. The weaving patterns showed an improved deposition rate and avoided lack-of-fusion defects compared to the line printing patterns. Two out of four weaving patterns achieved superior surface quality over the line printing patterns and are free of macro-defects such as side wall collapse and spattering. The weaving patterns have excessive heat accumulation and lower cooling rates compared to line printing patterns, which led to coarser microstructures and inferior microhardness and tensile properties. The thermal gradient is also more uniformed and aligned to the build direction, increasing the degree of grain alignment and texture of the weaving patterns across the samples.

Bolted steel joints are critical components in steel structures, influencing force transfer, deformation and energy dissipation mechanisms. In traditional analysis of steel structures, beam-column joints are often simplified as either pinned or rigid. While this simplification eases analysis and reduces computational demands, it fails to accurately capture the semi-rigid joint behaviour, underestimating designs, especially in multi-story steel structures where the cumulative effects of joint flexibility and frictional slip can be relevant. The complexity of accurately modelling semi-rigid joints in finite element analysis (FEA) arises from challenges such as geometric and material nonlinearities, frictional contact interactions, and detailed meshing, all of which increase computational demands, necessitating the use of simplified joint behaviour models. This study identifies and attempts to address a persistent gap in efficiently modelling these semi-rigid joints during structural analysis.

The primary goal of this MSc thesis is to investigate the behaviour of semi-rigid joints and propose a relatively cost-efficient approach to incorporating more realistic joint behaviour into structural analysis. To achieve this, the research explores the frictional behaviour and potential relaxation of bolted joints, focussing on the relative sliding of plates in frictional contact. The study specifically examines how these factors influence load transfer, energy dissipation, and overall structural performance. In multi-story constructions, the cumulative effect of small joint relaxations at each story level can lead to increased lateral sway and amplified second-order effects, raising significant concerns about structural integrity.

The research begins with a detailed examination of an individual beam-column joint assembly consisting of top, bottom and web plates welded to the column flange and bolted to the beam flanges and webs. Through numerical simulations in ANSYS, the study analyzed the joint under both monotonic and cyclic loading conditions, focussing on the force transfer mechanisms and relative deformations between various components of the connection. A parametric study, varying friction, bolt pretension, and loading magnitude, was performed to further explore how these parameters influence the stiffness, energy dissipation, and overall behaviour of the joint assembly. Based on the force-displacement response of different components of a beam-column connection, a Component-Based Spring Method (CBSM) is proposed to simplify the modelling of semi-rigid joints in 2-D multi-story frame structures. This method involves creating equivalent spring elements that simulate the load transfer and deformation behaviours observed in various components of a connection. The study then extended the analysis to the global behaviour of a three-story steel frame structure, comparing the performance of rigid and semi-rigid frames. The simplified CBSM approach was validated by applying it to a three-story steel frame structure and comparing the results with those obtained from detailed solid connection modelling.

The investigation into the local behaviour of the beam-column connection identified key components, such as the Top Plate & Top Beam Flange, Web Plate & Beam Web, and Bottom Plate & Bottom Beam Flange, that govern the force transfer and deformation within the joint assembly. The force-displacement curves of each component provided insights into the pre-yield and post-yield behaviour of the connection. The results revealed that the friction coefficient and bolt preload significantly influence the stiffness and energy dissipation capacity of the joints. Higher friction coefficients and bolt pretension enhanced joint stiffness but reduced energy dissipation, indicating a trade-off that must be balanced in design if dynamic vibration is a concern. The Component-Based Spring Method was validated against detailed solid models, achieving accuracy in frame deflection results of 93% to 96% while reducing computational time by 99.5%. The analysis of the global behaviour of joints within the frame structure revealed that semi-rigid frames exhibit significantly higher lateral deflections, up to 40.63% more when compared to rigid frames, highlighting the risks of oversimplified rigid joint models. These findings highlight the importance of accurately modelling the behaviour of semi-rigid joints to ensure realistic structural performance and the safety of steel frame structures.

This research addresses the implementation of learning algorithms and generative design in string-based topology exploration methods. It aims to generate diverse structural patterns for shells and surface structures that align architectural, engineering, and construction objectives. By integrating reinforcement learning (RL) and quad-mesh grammars, surface topology is explored through quantitative metrics, demonstrating the strength and generality of this approach. The research ultimately promotes creative exploration during the conceptual stages of structural design, emphasizing collaboration between form-designers and form-analyzers to harness emerging computational techniques.

The quad-mesh grammar was first formulated within a Markovian decision framework to integrate with open-source RL Python packages. States, actions, and rewards were defined with sufficient generality to avoid over fitting while evaluating the RL agent's ability to navigate between two specified string-action sequences and their associated mesh layouts. Initially, simple tasks involving four design steps were tested, followed by more generalized target terminal states with longer design sequences. The impact of different reward structures and varied model parameter setups on convergence and accumulated rewards was also analyzed.

The findings indicate that reward functions based solely on topological and grammatical characteristics did not fully guide the agent from an initial coarse mesh to a target state. However, extended design episodes demonstrated potential for improved RL outcomes. The DQN struggled with non-optimal policies due to negative rewards and sparse positive reinforcement, suggesting that customized model architectures or alternative RL algorithms could enhance performance. The exploration phases yielded suboptimal but diverse mesh configurations, highlighting the need for additional structural and geometric parameters, as well as more complex grammar operations to improve diversity while mitigating computational challenges. These insights underscore the importance of balancing feasibility, exploration, and optimization in computational design workflows.

Numerous traffic bridges have been constructed over the last 150 years. Because bridges are commonly built based on an expected lifetime of 75 to 100 years, a large quantity of bridges are reaching the end of their design lifetime, or even are far overdue. To ensure these bridges remain operational without experiencing catastrophic failures, they have to be recalculated, and if need be, repaired, strengthened or replaced. Particularly older bridges are commonly constructed utilizing rivets. Their often overly complex geometries, the fact that riveting has become largely obsolete as a construction process, and old bridges are commonly not design for fatigue loading, engineers regularly face significant challenges when reassessing such bridges.

While fatigue phenomena have been extensively investigated throughout the years, studies pertaining to the fatigue of riveted connections are relatively limited. The Eurocode on fatigue, EN-1993-1-9 includes only two detail categories. Additional guidelines, like RBK Steel expand upon these detail categories, but focus primarily on built-up beam cross-sections in riveted structures, rather than riveted connections. In order to attempt to more accurately assess the complex joints present in ancient steel bridges, this thesis attempts to answer the following question: What would be a suitable approach to model complex riveted joints and assess their fatigue life considering a balance between the level of complexity and applicability in design practice?

A literature study is performed to identify the different factors that affect the fatigue resistance of riveted connections, as well as to highlight several of the available methods to perform a fatigue assessment. Through the investigation of experimental studies complemented with Finite Element (FE) Analyses, a design methodology for a full scale riveted model is drawn up, and finally a FE model of a joint of the John S. Thompsonbridge is constructed. A critical location within the joint is identified. On this critical locations, several stress- and strain-based fatigue life analyses are performed, namely the use of Stress Concentration Factors (SCF), Smith-Watson-Topper’s (SWT) strain-life equation, and the multiaxial shear strain criterion (MSSC) method, to investigate the effects of incorporating mean stress effects and multiaxiality.

From these analyses, it is concluded that SCF appears to provide overly conservative fatigue life estimates, whereas SWT and MSSC provide more probable results. The increased life estimate through MSSC suggests a limited degree of multiaxiality present in the critical location. All three methods require a detailed FE model, complicating the fatigue assessment of the joint. While the SCF method is slightly simpler to use than SWT and MSSC, it does not weigh up to the conservativity of its life estimation. SWT is deemed the most suitable approach for the fatigue assessment of riveted joints, given that it is more widely applicable and relevant than MSSC.

Steel is a widely used material in construction because of its availability, high strength-to-weight ratio, and recyclability. However, steel manufacturing consumes a lot of energy, and there is an increasing demand for more environmentally friendly building materials. Consequently, high-strength steel is becoming more popular as it reduces the mass of a structure. Although high-strength steel has been available for many years, its use in offshore topsides is limited due to stability and deflection issues. This research aimed to assess the feasibility of utilizing high-strength structural steel in offshore topsides, and investigated how the use of high-strength steel can be optimized in topside design.

Two different screening tools were constructed to assess the feasibility of high-strength steel within an entire topside. It was concluded that the length of a particular beam can tell an engineer if it is worth further investigating the potential of high-strength steel, while columns showed potential in all cases. The methods were tested with a case study in which an topside was assessed for its feasibility of utilizing high-strength steel beams and columns. Only hot-rolled primary and secondary beams combined with the columns and bracings were considered for this topside. When S460M steel was used for strength-governing beams and seamless tubulars, in combination with S690Q steel for welded tubular columns, the highest benefits were found and a maximum steel weight reduction of 15% was found for the considered components. At the same time, the material costs were reduced by 10%, the welding costs by 13% and the embodied carbon savings equalled 14%. When comparing these results with the total topside steel weight, 5% of the topside steel weight was reduced by using a combination of S460 and S690 steels. It was concluded that high-strength steel is feasible for offshore topsides and is more environmentally friendly and cost-effective, providing a promising alternative to conventional steels within certain components on offshore topsides.

This research presented a screening tool that is simple to use and assessed the feasibility of high-strength steel. Other engineers can easily extend this tool. As more detailed calculations are included in the screening tool, it is expected that additional cost reductions and embodied carbon savings can be found. Furthermore, including additional steel components in the assessment, such as plate girders, may result in finding much higher total weight reductions.

The field of wind energy has experienced significant expansion and development driven by the need to reduce reliance on fossil fuels. Offshore wind turbines have become increasingly popular leading to larger turbines with greater energy output to match the growing demand. The most common foundation type of offshore wind turbines is the monopile foundation, consisting of two large diameter steel tubulars, the monopile (MP) and the transition piece (TP).
As wind turbine generators (WTGs) increase in size, the transportation and installation of MP foundations becomes more complex and expensive, requiring specialized vessels with adequate lifting capacity. This research investigates a new connection – the underwater wedge connection – as a method of connecting an MP to a TP underwater. By having an underwater connection, the size of the MP can be reduced as it no longer has to exceed the waterline.
The wedge connection consists of a dowel with an inclined plane and two flanges. A number of dowels will be placed around the circumference of the connection, fitted onto flanges on both the MP and TP. As the dowels are pushed into position, the inclined plane creates a vertical preload between the two flanges. This allows a tensile load on a segment of the connection (caused by the bending moment in the foundation) to be transferred to the foundation via two load paths: reduction of the preload, and by loading the dowel itself in shear.
First an analysis on the current state of art and relevant design codes was carried out. This analysis highlighted the requirements the connection needed to satisfy, in order to serve as starting points to make decisions regarding the design. The preliminary design of the connection is made using analytical calculations. The structural integrity of the flanges and the dowel at the Ultimate Limit State (ULS) was verified. A 3D model of the connection was then created to perform numerical analyses using ANSYS Static Structural. The behaviour of the connection under ULS and Fatigue Limit State (FLS) loading is studied. The opening and failure point of the connection along with the sensitivity of various parameters are also investigated.
The results show that connection was able to effectively transfer both tension and compression loads. The connection has a mechanical advantage of 1.95, meaning that it achieves the same preload as the ULS load by applying only 51.3% of the ULS segment load during installation. The connection opens gradually and only opens at a load higher than the ULS load. Even after opening, it continues to transmit loads effectively, with ultimate failure governed by the yielding of the lower flange. Additionally, the connection exhibits good fatigue resistance, with a low fatigue damage level of 3.8%.
It is recommended to conduct experiments to validate the numerical model employed in this study. Further studies should also be carried out to investigate the impact of structural imperfections on the behaviour of the connection.