p✳ PopFun.org↖ All experiments

EXPERIMENT 004 / FLUID DYNAMICS

Wind Tunnel v1.0.0

A small change in shape. A whole new flow.

A living, 2D experiment
DESIGN · OBSERVE · REFINE
2D CROSS-SECTION
AIRFLOW ⟶ ⟶ ⟶
Flow developing
LIVE FLOW FIELD 0 steps
SPEED / INLET SPEED
0×1×2×
REYNOLDS NUMBER202Inlet speed × chord / viscosity
WAKE ROTATION0.00RMS vorticity × chord / inlet speed
FLOW DEVELOPMENT0.0 passesInlet distance / tunnel length
↳

Read the flow. Brighter regions move faster. Trails follow the computed velocity field; slow or looping trails reveal the wake.

REAL EQUATIONS. ROOM FOR CURIOSITY.

Inside the wind tunnel

Choose a shape, adjust its angle and size, then watch the air respond. Custom outlines can be edited directly in the tunnel, or with the vertex selector and arrow keys. Changes reshape the obstacle immediately; the wake takes time to develop.

A computed flow field

This is an original 220 × 100 cell, two-dimensional D2Q9 lattice-Boltzmann simulation with BGK collision and no-slip, half-way bounce-back obstacles. The inlet and top/bottom edges are equilibrium reservoirs; the outlet uses a zero-gradient approximation. Time advances in fixed lattice steps. Slower devices advance less simulation time per second.

Velocity, relative pressure (density difference / 3), and signed vorticity come from the solver. Flow trails integrate the changing velocity field and are pathlines, not guaranteed steady streamlines. Reynolds number uses the shape’s nominal chord; wake rotation is measured in a region downstream of the shape. Values use lattice units, not metres per second or pascals.

What this experiment can tell you

Explore qualitative flow separation, wakes, rotating eddies, and the effect of geometry or viscosity. It does not solve 3D flow or use a turbulence closure model. Vorticity also occurs in smooth shear and does not prove turbulence. The coarse, stair-stepped grid, finite boundaries, simplified outlet, and numerical transients limit accuracy. No validated drag, lift, or engineering predictions are provided. Shape edits reset the development counter; compare designs after similar development times.

Method reference: S. Succi, Lattice Boltzmann Methods, Scholarpedia (2009). Geometry: symmetric NACA 0018 thickness polynomial for the airfoil preset. Geometry is normalized and rasterized, not a certified airfoil analysis.

Changelog

WIND TUNNEL v1.0.0 · NOVEMBER 1, 2026