A design-stage screening of pedestrian wind comfort at urban / building-block scale. It solves the wind flow around your massing for each wind direction, combines the results with your site's wind rose, and colours the ground plane by how comfortable it is to sit, stand or walk there across the year — plus a wind-safety check for dangerous gust spots.
Urban Wind Simulation
Runs in the browser — nothing to install.
| Input | .3dm / .obj massing + EPW weather file |
| Engine | Browser GPU Lattice-Boltzmann CFD |
| Output | Annual comfort map at 1.5 m + safety check |
How it works
A real CFD solve runs in your browser on the GPU, then the directional results are weighted by your local wind climate into an annual comfort statistic per location.
- WebGPU Lattice-Boltzmann (D3Q19, TRT collision, Smagorinsky LES) solves each wind direction client-side.
- Per direction it extracts the 1.5 m amplification factor γ = local speed / reference wind speed.
- Your EPW is fitted to a 16-sector Weibull wind rose; calm hours are tracked separately.
- Annual exceedance F = (1 − calm) · Σ p_θ · exp[−((U_thr/γ_θ)/c_θ)^k_θ] per pedestrian cell.
- Cells are classified by Lawson LDDC / NEN 8100 / Davenport, with a shared 15 m/s danger check.
- No GPU? A fast parametric estimator (wakes, corner acceleration, channelling) gives an approximate map.
Using it
Upload your massing
Drop a .3dm (or .obj) of your building plus the surrounding context.
Load the site climate
Add an EPW weather file — the wind rose and speeds come from it.
Pick fidelity & directions
Choose draft / standard / fine and how many wind directions to solve.
Run the study
Each direction solves in turn; a draft map appears first, then sharpens.
Read the comfort map
Colours show sit / stand / stroll / walk comfort; click a spot for detail.
Export
Download the comfort map, CAD contours, or a PDF report.
What to trust it for
- Comparing massing or podium options for street-level windiness.
- Finding uncomfortable entrances, plazas and gap-between-buildings spots early.
- Orienting screens, canopies and planting before detailed design.
- Scoping whether a site needs a formal wind-tunnel study.
What not to use it for
- A substitute for an accredited wind-tunnel test or certification-grade CFD.
- Thermal / buoyancy-driven (stack) wind effects — the solve is isothermal.
- Sign-off on a marginal site from a single coarse 4-direction run.
Known limits
- Accuracy depends on grid resolution and how complete the surrounding context model is.
- Gust / safety speeds are approximated from turbulence intensity, not directly resolved.
- The no-GPU fallback is qualitative — clearly labelled 'Approximate'.
Tips
- Do — Include neighbouring buildings within roughly a few building-heights — context drives local wind.
- Avoid — Don't treat a draft / 4-direction result as final — refine before any decision.
Urban Wind Simulation
Runs in the browser — nothing to install.