Building geometry strongly influences pedestrian comfort and safety, especially in dense urban environments. As urbanization increases and cities continue to develop, understanding how building arrangement affects pedestrian-level wind conditions becomes increasingly important for creating safe and comfortable outdoor conditions. In this thesis, the TU Delft campus was used as a case study to investigate how modifications in the spatial arrangement of buildings affect pedestrian-level wind conditions. By relocating groups of buildings within the campus area, a set of four hypothetical modified layouts was created and steady-state RANS simulations were performed for each layout. To assess pedestrian wind comfort, a combined exceedance criterion based on wind velocity and turbulent kinetic energy was used rather than the standardized wind comfort guideline NEN 8100. The results show that building rearrangement mainly redistributes discomfort zones, following the regions of high wind velocity and turbulence kinetic energy. The strongest effects occur in the places where layout modifications took place. The relocation of high-rise buildings is the dominant factor that determines the probability and the extent of the discomfort zones, with more exposed placements generally leading to a larger area of discomfort. While most layouts mainly redistribute the zones of high discomfort risk, one modified configuration shows the clearest improvement in pedestrian wind comfort in the main central open area of the campus. For a critical wind direction that produces the highest wind speeds in the main open space of the campus, an additional blockage-ratio analysis was performed. The results indicate that local wind velocity in the region responds to upstream geometric blockage, with higher frontal blockage generally associated with lower wind velocity. Overall, these findings highlight the important role of building design in shaping pedestrian-level wind flow and provide useful insight for improving pedestrian comfort in urban spaces.

Accurate urban Computational Fluid Dynamics (CFD) simulations require volumetric meshes whose boundary conforms to complex city geometry while maintaining cell quality and enabling adaptive mesh sizing. This thesis develops a Voronoi-based polyhedral meshing methodology tailored to the output of City4CFD, where the simulation domain is provided as a triangulated Piecewise Linear Complex (PLC) whose faces are grouped into semantic boundary patch types. The central goal is to generate a boundary-conforming polyhedral mesh without clipping Voronoi cells against the boundary, and to preserve patch information so that CFD boundary conditions can be transferred consistently to the mesh.

The method constructs a set of boundary (surface) Voronoi sites by intersecting triplets of spheres centered at the vertices of a refined surface triangulation. Building on the sphere-based sampling conditions of VoroCrust, sphere radii are initialized and iteratively adjusted to satisfy smooth-coverage, smooth-overlap, and Lipschitz-type size-transition constraints, while a shrinking step resolves "half-covered" seed configurations. The surface triangulation is refined by splitting triangles and protecting edges, then regularized with centroidal smoothing until the triangles associated with each facet intersect in two points, yielding paired sites on opposite sides of the boundary. These sites induce Voronoi facets that coincide with the input surface, producing triangular boundary faces and avoiding cell clipping. Patch-boundary preservation is enforced by identifying protected edges not only via dihedral-angle sharpness but also via changes in patch groups, ensuring that 1D interfaces between patch types are explicitly represented in the resulting Voronoi boundary.

To populate the mesh interior with Voronoi sites, the thesis evaluates several strategies (uniform random scattering, adaptive distance-based refinement, and structured lattices) to regulate cell density and promote larger cells away from geometric features. A prototype implementation in C++ using CGAL demonstrates feasibility for 2-manifold inputs and produces boundary-conforming Voronoi meshes compatible with OpenFOAM-style polyhedral representations. The approach assumes a valid 2-manifold boundary and does not repair non-manifold or overlapping input geometries.

Computational Fluid Dynamics (CFD) is widely used to analyse wind flow around buildings; however, creating detailed input geometries and corresponding meshes can be a time-consuming process. This thesis investigates voxelization as a means to simplify building models for their use in CFD and analyses the impact of voxel resolution on simulation accuracy.

Three building geometries with varying roof shapes and footprints were converted from detailed continuous models into voxel models with increasingly finer voxel resolutions. The voxelized models were compared to a non-voxelized LoD 3.2 model to assess accuracy under four key wind directions (90°, 45°, 22.5°, and 0°).

The CFD simulations were performed using OpenFOAM’s RANS solver with a $k–\epsilon$ turbulence model. Due to its higher computational efficiency compared to other turbulence-resolving frameworks, the RANS approach enabled a large number of simulations while maintaining sufficient accuracy for urban CFD applications. A grid-independence test was conducted using the Grid Convergence Index (GCI) method for one model. The resulting grid-independent mesh was then scaled for the other models, ensuring that all simulations remained grid-independent.

The results show that coarse voxel resolutions (1 m and 0.5 m) significantly increase the size of the building geometry and leads to large velocity differences compared to the non-voxelized model. Sloped roofs were most affected by voxelization, as these models showed greater velocity differences than those with rounded roofs.

Wind direction also plays a significant role in voxelization accuracy. While the 90°, 22.5°, and 0° wind directions showed similar results across voxel resolutions, the 45° direction produced notable velocity differences. An exception was observed for the model with a rounded roof, which showed more consistent results across all wind directions.

Overall, the velocity difference between non-voxelized and voxelized models decreases as voxel size decreases. However, below a voxel size of 0.1 m, the reduction in velocity difference stagnates, indicating that smaller voxel sizes offer limited additional benefit to CFD accuracy.

This thesis introduces the “city stack” framework and the concept of the “typology grid” to analyze and procedurally generate cities using publicly available geospatial data. It builds upon existing urban morphology research and combines existing metrics with new metrics to examine road networks and building footprints across 43 cities worldwide. The resulting unsupervised clustering of road patterns produced mixed outcomes: some global road typologies were consistently identified, while others lacked clarity or validity. Supervised classification of building typologies showed potential for city generation applications but faced challenges related to data quality and validation methods.
To encode the captured urban form, the typology grid was introduced, which allows for straightforward comparison between cities and links the analysis phase with the generation phase. While this approach simplifies complex urban patterns for better understanding, it may oversimplify details needed for effective city regeneration, needing future research into parameterizing the grid.
The simulated annealing optimization technique was applied to generate new city models to produce typology grids resembling those of actual cities. It proved a promising method, with relatively plausible results from just a few shape-based rules in the objective function. Computation time and grid size were a limiting factor, ruling out real-time use. The road and building generation methods based on the typology grid and city stack framework demonstrated the approach’s feasibility but indicated that further refinement is necessary.
Further contributions by this thesis are an open source command line tool for analyzing the urban form real-life cities based on publicly available geospatial data, a proof of concept tool for generating cities, and a publicly available dataset of the analyzed cities.
In conclusion, the city stack framework and typology grids offer a viable method for captur- ing and generating urban form and can be used as a starting point for future research.

While three-dimensional coral reef models are valuable for various applications, existing approaches like photogrammetric scanning and manual modeling require substantial time and expertise, limiting their scalability. Previous algorithmic approaches, particularly Agent-Based Models (ABM), have relied heavily on complex ecological simulations and deep domain knowledge. This thesis explores an alternative data-driven approach to automated coral reef modeling, investigating whether empirical data sources can provide a scalable method for generating ecologically plausible 3D models. Rather than simulating ecological processes from first principles, we develop a pipeline that leverages observational data to inform and constrain procedural generation techniques. Through systematic evaluation of available data sources, including the Global Biodiversity Information Facility (GBIF), CoralNet, the Allen Coral Atlas, the Coral Traits Database and the Smithsonian Institution's 3D coral collection, we identified both opportunities and significant limitations in current data availability. The research developed a modular pipeline implemented in Blender that combines procedural terrain generation with the placement of 3D coral models, integrating species occurrence data aggregated over geomorphic zones. To ensure robust data integration across sources and maintain compatibility with evolving taxonomic standards, the pipeline implements automated species name verification through the World Register of Marine Species (WoRMS) (WoRMS - World Register Of Marine Species, n.d.) API. While the resulting pipeline successfully establishes a foundation for automated coral reef modeling, limitations in available structural data necessitated the use of manually configured parameters for critical aspects such as terrain characteristics and population density. The pipeline's modular structure, standardized taxonomy handling, and integration with standardized classification systems position it well for future iterations as improved data sources become available. This research demonstrates the potential of data-driven approaches to coral reef modeling while highlighting the need for more comprehensive, fine-scale structural data to enable fully automated, ecologically plausible modeling of coral reef environments.

Urban microclimate significantly affects people’s experiences and activities in urban environments by a series of phenomena, among which urban flow is an important factor to be considered. Computational Fluid Dynamics (CFD) method has become a popular tool for studying urban airflow because of its low cost compared with experiment methods. However, flows over urban areas exhibit turbulence nature of being three-dimensional, unsteady, and multi-scale. Additionally, the large computational domain that should be covered and the inherent inhomogeneity of the urban structures make it challenging to do full-scale modelings. Large Eddy Simulation (LES), with the development of computing power, becomes a promising tool to study such flows.

In the campus of Delft University of Technology (TU Delft), a crossroad near the EWI building (the main building of Faculty of Electrical Engineering, Mathematics and Computer Science) is constantly complained for its strong wind. This research tackles such problem using LES, and takes TU Delft campus area itself as case study. The development of this research is composed of three stages.

In the first stage, Vreman eddy viscosity model is implemented into Canonical Navier-Stokes (CaNS), a massively-parallel Navier Stokes solver developed by Costa, (2018). Based on a structured three dimensional Cartesian grid, the subgrid scale (SGS) eddy viscosity model is inserted into the Navier Stokes equation by adding an extra diffusion term. The inserted diffusion term is discretized with second-order difference scheme along with interpolation of the velocity field due to the staggered grid arrangement. The implementation is validated with a turbulent channel flow with friction Reynolds number 𝑅𝑒𝜏 = 360. The good agreement is found and the discrepancy is small.

In the second stage, the solver employs a direct-forcing Immersed Boundary Method (IBM) and is further validated with the flow over periodic cube arrays. Signed-Distance Field (SDF), as a convenient tool, is generated and functions as read-in data for IBM. The IBM processes the effect of the boundary as an added force on the fluid points at the interface. The stair-step approach approximates the structure boundary with the cuboid cells faces. The results match well with the wind tunnel test data from Castro et al., (2006) and a previous LES study by Tomas et al., (2016).

In the last stage, the validated solver is applied to a scaled-down TU Delft campus model. The simulation setup is designed by considering the achievability of a possible future wind tunnel measurement. Three grids are used for a grid convergence analysis by comparing the total IBM force, mean velocity, and Reynolds stress at certain locations. The flow converges with the finest grid with grid number 𝑁𝑥 × 𝑁𝑦 × 𝑁𝑧 = 960 × 880 × 240. Around the EWI building, a high-speed region is found at the cross road location. Behind the building, a wake area is observed, and a clear shear layer is on the top of the building.