Richard Kraaij

Gas laboratory in 3D

Rotate a chamber of colliding spheres and measure how many fall inside a sampling cube.

Particle chamber

Running
Interactive three-dimensional gas chamber

Drag to rotate the chamber.

Mint ring · inside sampling cube
SlowFast
Particles
160
Temperature
300 K
Mean speed
—
Wall pressure
—
Elapsed time
0.0 s

Experiment controls

160
300 K

Particle distribution

Particle distribution histogram

Sampling cube

Inside
—
Expected
—
Volume fraction
12.5%
50%
50%
50%
50%

Position is a percentage of the available travel along each axis. At 50% the cube is centred.

Count in cube · last 20 s

Cube count history
ObservedExpected

Cube count distribution

Waiting for the first observation.

Cube count probabilities
ObservedBinomialPoisson
Observations0
Observed mean
—
Predicted mean
—
Observed variance
—
Binomial variance
—
Poisson variance
—

The sampler counts particle centres every 0.5 s of simulation time. Nearby observations can be correlated. Changing the cube or physical settings starts a new histogram.

Both predictions use the selected volume fraction and assume independent, uniformly distributed particle centres. The Poisson approximation works best for small regions. Finite particle size and crowding can shift the measured distribution.

?
Observed
—
Binomial
—
Poisson
—

What changes in three dimensions?

A cube with half the chamber’s side length takes up one eighth of its volume. In a uniform gas of 160 particles, its expected count is 20. Halve its side length again and the expected count drops to 2.5.