Topography Factor (Bolund Hill)

Terrain features - hills, escarpments, ridges, valleys - accelerate or decelerate the wind relative to the flat-ground ABL. Standards such as EN 1991-1-4 and ASCE 7 address this through a topography factor \(c_0\) (or \(K_{zt}\)), but these analytical corrections are limited to idealized geometries. For complex terrain, CFD is the standard engineering tool.

The Bolund Hill experiment (Berg et al., 2011) is the canonical benchmark for terrain-induced wind speed-up. Bolund is a small coastal hill in Denmark - height \(H\) = 12 m, steep cliff on the western and northern faces - with a dense network of 10 measurement masts (\(M0\)–\(M9\)). Its compact size and high-quality experimental dataset make it the standard test case for terrain flow models.

This guided case walks you through setting up a single-direction terrain simulation in AeroSim using the Bolund geometry. The output is the mean speed-up factor across the mast positions. Because Bolund is surrounded by open sea on the approach side, this is also the simplest possible AeroSim setup: Category 0 terrain requires no roughness fins and no profile development section.

Case Description

Parameter

Value

Hill height

\(H\) = 12 m

Wind direction (this guide)

270° (West - directly into the cliff)

Terrain category

0 (open sea, \(z_0\) = 0.003 m)

Roughness fins

Not required

Measurement masts

\(M0\)–\(M9\) (10 masts)

Reference mast

\(M0\) (upstream, open-sea inflow)

The primary output is the mean speed-up factor \(U/U_H\) at each mast position, where \(U_H\) is the mean wind speed at height \(H\) measured at the reference mast \(M0\).

Prerequisites

Before setting up this case you must have a completed ABL case configured for Category 0 (open sea). Refer to ABL (Atmospheric Boundary Layer) and confirm:

  1. The inlet profile CSV uses a Category 0 logarithmic profile with \(z_0\) = 0.003 m and \(u_*\) matched to the reference wind speed at \(M0\).

  2. No roughness fins are present in the domain - Category 0 is the only terrain category where fins are omitted entirely.

  3. The mean velocity profile at the intended terrain location is consistent with the target open-sea profile.

The inflow for the Bolund simulation is calibrated against mast \(M0\) measurements, not against an analytical standard profile. If you are using the experimental dataset, extract the mean velocity and Reynolds stress profiles from \(M0\) to construct the inlet CSV.

Setting Up the Case in AeroSim

Import the Terrain

Import the Bolund terrain as a closed surface mesh (STL or equivalent). The terrain mesh should include the seabed floor so that the IBM can resolve the cliff edge and the hill surface correctly. Orient the geometry so that the 270° wind direction aligns with the positive \(x\)-axis (west-to-east flow).

Todo

Add UI steps for importing the terrain geometry, orienting it to the 270° wind direction, and confirming placement within the domain. Include screenshots.

Boundary Conditions

Category 0 setup is the minimal AeroSim configuration:

  • Inlet: Synthetic Eddy Method (SEM) using the Category 0 profile CSV calibrated against mast \(M0\) (columns: z, Ux, Rxx, Ryy, Rzz, Rxz, Rxy, Ryz).

  • Ground and terrain surfaces: IBM EqLog wall model with \(z_0\) = 0.003 m. No roughness fins are needed.

  • Top and lateral faces: Free-slip Neumann condition (RegularizedNeumannSlip).

  • Outlet: Zero-gradient Neumann condition (RegularizedNeumannOutlet).

Todo

Add UI steps for assigning boundary conditions. Include screenshots of the BC panel.

Body Refinement on the Hill

Add a body refinement region around the hill to resolve the flow over the cliff and the wake behind the hill. The refinement should be concentrated at the hill crest and the steep western face, where speed-up gradients are largest.

Todo

Add UI steps for configuring the body refinement region around the hill. Include screenshots of the mesh refinement panel.

Domain Guidelines

Todo

Write the internal guidelines for domain sizing for terrain cases: upstream fetch from the terrain feature, downstream length, domain height, lateral extent, and blockage considerations for topographic features.

Refinement Guidelines

Todo

Write the internal guidelines for mesh refinement for the Bolund case: finest cell size at the hill surface, number of refinement levels, cell count in the body refinement region, total node count, and expected GPU memory budget.

Exports

Todo

Describe what to export from the terrain case: velocity fields, mast probe time series, surface velocity colormap, convergence history. Mention any AeroSim export presets for terrain cases.

Post-Processing and Visualizing Results

Speed-Up Factor at Mast Positions

Extract the time-averaged velocity magnitude at the height of each mast sensor position (\(M0\)–\(M9\)). Compute the speed-up factor relative to the reference mast \(M0\) at the same height:

\[ S = \frac{U_\text{mast}}{U_{M0}} \]

Plot the simulated speed-up factor against the Bolund experimental data (Berg et al., 2011) at each mast. The flow at the hill crest (mast \(M8\)) typically shows the largest speed-up; the eastern wake masts show speed reduction and elevated turbulence.

Surface Velocity Colormap

Generate a horizontal slice of mean velocity magnitude at 2 m above the terrain surface (sensor height in the Bolund experiment). This colormap shows the spatial extent of the speed-up zone over the crest and the recirculation zone in the lee of the cliff.

The AeroSim validation results for the Bolund Hill benchmark, including mast comparisons for all three wind directions (239°, 255°, 270°), are available at docs.aerosim.io/validation.

Todo

Add AeroSim-specific instructions for extracting mast velocities and generating terrain colormaps in the post-processing interface. Include screenshots of the result panel.

Turbulence intensity profiles at each mast and advanced analyses (separation length, recirculation height) are covered in a separate guide.

Next Steps