This thesis is part of a larger ongoing research project at the TU Delft regarding structural glass as a novel construction material. Here, the focus is directed towards the casting of glass to create complex free-form geometries. Thus far, only smaller components have been created, but a multitude of previous student work explored how the current limitations in the fabrication process could be overcome by using topology optimization to design larger cast glass components. This thesis continues this research by integrating the limitations and remarks concluded from previous work to further explore the potential of designing large cast glass components using topology optimization.

This thesis intends to contribute to the existing research by creating a tool for the design of 3D optimized cast glass components, which can take into consideration the structural properties of glass as well as manufacturing, annealing and specific design criteria. The tool is created in Matlab, using the SIMP optimization method with 8-noded hexahedral finite elements for the structural model.

From the literature review it was concluded that the SIMP methodology is best suited the requirements of this project. For the structural model, 8-noded hexahedral elements were selected. Two algorithms were developed: one with a compliance based objective and a second with a volume based objective.

The algorithm’s performance was tested on different domains. To facilitate comparison with existing literature, the algorithm was initially calibrated with a domain width of one element, effectively creating a two-dimensional optimization. Subsequently, its performance was assessed through the application of a small three-dimensional problem. Finally, the algorithm is used to design a case study involving the structural slab of a small pedestrian bridge located inside the British Museum.

The results validate the benefits of the 3D algorithm. The geometries optimized in three dimensions have a higher structural stiffness and offer more spatial insight. Moreover, the volume optimizations results in a larger material reduction than those achieved in two-dimensional optimizations. For this reason, the volume-optimized component was selected for post-processing and implementation in the final design.

Glass has become ever more popular in the building industry. Glass is already used in facades, staircases, fins, floors, and beams. Nevertheless, free-standing load-bearing glass columns are rarely used. It needs to be robust, which means that it needs to give a warning before failure so that people can evacuate. Glass is a brittle material and has a sudden failure, which means that it is not directly logical to use solid glass as a load-bearing structural column. Rather despite the high compressive strength, due to complications with manufacturing, spontaneous failure, and lack of satisfactory of fireproof performance, more research needs to be done on the manufacturing, the fire safety, and the robustness of glass columns. In the literature study, the layered tubular glass column was found to be most promising for further investigating. The aim of this research is to design and engineer a transparent tubular glass column as a structural element, which is robust and fireproof. In the literature study a few aspects emerged: the tubular shape, the column needs to be sealed to avoid the column from becoming dirty, end connections, the geometric tolerances in the glass tubes and the robustness. The designs were engineered for the case study Bouwdeel D in Delft, but the designs can also be applied to other buildings: the fire resistance, demountability, the dimensions and the calculated compression loads. With these aspects in mind, three designs were engineered, the MLA (Multi Layered with Air) (2x) and the SLW (Single Layered with water) (1x) which fulfil the defined design criteria/main concerns. Six small samples with a length of 300 mm were produced to test the lamination process with regard to bubble formation and possible breakage of the glass by internal stresses. Three laminated DURAN (annealed) samples and three laminated DURATAN (heat-strengthened) samples were tested. The glass tubes were manufactured by extruding and were made by SCHOTT. H.B.Fuller Kӧmmerling carried out the lamination process. Furthermore, the connection components (steel and POM) were produced and arranged by Octatube, the steel hinges were arranged by Techniparts, and the Hilti mortar was arranged by Hilti. Afterwards, these samples were tested on compression strength, to investigate the behaviour of the interlayer material, the post-failure behaviour of the designs, the differences between annealed and heat-strengthened glass samples, the behaviour of the connections under pressure, and the capacity of the glass tubes and the connections. The first cracks appeared earlier than expected. Nevertheless, the samples seemed to be really robust, because the samples had a large load-bearing capacity even after the first cracks occurred. The first crack appeared between 95-160 kN in the DURAN samples, and after that the samples were still able to carry a load of 700-750 kN. So, after the first crack, the specimens were able to have around 4-5 times more load. The heat-strengthened samples first cracked at 120-160 kN. The maximum force of these samples was 390-490 kN. This means that after the first crack, the specimens were able to have around 3 times more load.

The search for a transparent, reversible, and reusable consolidation system for monuments led to the ambition to test possible cast glass interlocking brick designs using numerical calculations. The focus of this research is hence the design, simulation and evaluation of a possible cast glass interlocking geometry using Finite Element Analysis (FEA). For this purpose, design criteria are formulated from literature; considering glass, polyurethane (used as interlayer) and interlocking systems. As interlocking systems are determined by their boundary conditions, a case study of a monument is chosen to provide additional design criteria. The goal of the case study is to provide a reversible and reusable restoration and consolidation alternative for the current invalid restorations.
The design criteria obtained then are combined into an initial geometry, whose parameters are varied to test their sensitivity to its shear capacity, using FEA. Christensen’s failure criterion is used to locate prone areas in the geometry, and to evaluate the theoretical moment of failure. This output value combines the three principal stresses into a failure envelope, hence can generate contour plots to envision peak-stress-prone areas. This is important especially for glass structures, as they are prone to peak tensile stresses.
From the results design diagrams are created and applied on a conceptual cast glass interlocking consolidation design for the monument chosen as case study: The Lichtenberg Castle ruin.
The initial design is moreover prototyped to check its interlocking capabilities, residual stresses and deviations introduced by shrinkage.

Being able to evaluate possible geometries using FEA can decrease costs and time when searching for a new interlocking geometry. Prone areas are easily highlighted using the Christensen’s failure criterion output. Hence peak stress sensitive or invalid geometries can be discarded before reaching the prototyping stage, which is time consuming and costly.
The creation of a methodology to predict this behaviour is hence valuable for further research on other cast glass geometries and can moreover be applied in any other field when analysing solid complex geometries.
Another goal is to find a cast glass brick design which not only can consolidate the monument of the case study, but is moreover applicable in other projects or configurations. The brick then is not a one-solution design, but can be reused in other projects.

The geometry hence is varied using Grasshopper plug-in for Rhinoceros. By exporting the geometry using a STEP-file, a solid can be loaded into DIANA FEA, where it can be analysed using their newly implemented output value of the Christensen’s failure criterion.
The geometry of the monument is gained through a 3D laser scan, resulting in a point cloud. The point cloud is adapted using Autodesk Recap, then further processed in Rhinoceros.

The Christensen’s failure criterion output is a proper and fast way to evaluate possible cast glass brick designs. Any compressive stresses on the interlocking brick geometry are beneficial for its shear capacity, as is an increase in interlocking amplitude or brick height. Increasing the amplitude however affects the allowable tolerance negatively, which is also the case for a decrease in brick height. Decreasing the brick height hence results in both negative effects.
The conceptual design for consolidation of the Lichtenberg Castle tower can replace the current interventions with equal or higher capacity, even for all conservative assumptions and simplifications. The design can still be altered less conservative after more experimental results and simulations come available.

The methodology applied can now be further developed and performed on other complex geometry designs. The presented multifunctional cast glass interlocking brick design, and its variations can be further investigated and applied in other projects.