The Centrifuge

Shake a jar of muddy water and let it stand undisturbed. Log the clarity and the height of the settled layer every fifteen minutes.

Expect it to take hours, and the boundary to stay indistinct.

This is the septic tank's mechanism from the sanitation batch — stillness sorting by density — and it is the same physics the centrifuge accelerates.

Record the time to reach a clear supernatant. That figure is what the next steps are going to compress.

Put the same suspension into a sealed tube and spin it in a salad spinner, or on a hand-cranked rig, for two minutes. Compare with the standing sample.

Expect separation in minutes rather than hours, with a much sharper boundary.

Then measure the rotor radius and estimate the speed, and compute a = ω²r in multiples of g.

Even a kitchen spinner produces tens of g. Write down the ratio between your computed g-force and the time saved — they should track each other, because it is the same sedimentation running faster.

Seal tubes properly and never spin glass in an unbalanced rig.

Run the same sample at half speed, then full speed, keeping radius constant. Then, if your rig allows, at the same speed but half the radius.

Compare separation times.

Expect halving the speed to hurt roughly four times as much as halving the radius.

That is ω² against r, measured.

It explains the shape of every centrifuge ever built: compact rotors spun very fast, rather than large slow ones. An ultracentrifuge reaching hundreds of thousands of g is a small rotor turning extremely quickly, and its whole design problem is holding itself together.

Spin your rig with tubes balanced opposite each other, then deliberately unbalanced by a small mass, and observe the vibration. Stop as soon as you have seen it.

Expect a dramatic difference from a small imbalance.

An unbalanced mass at radius r generates a rotating force of mω²r — the same ω² that gives the machine its power now works against its bearings, thousands of times a second.

This is why every centrifuge instruction begins with balancing, and why laboratories keep water-filled blank tubes: an unbalanced rotor is the one genuine hazard of the machine.

Layer a dense sugar or salt solution in the bottom of a tube and a lighter one above it, then add a mixture of particles of differing densities on top and spin gently.

Expect particles to stop at the level where the surrounding liquid matches their own density, rather than all going to the bottom.

This is density-gradient centrifugation, and it is a different instrument in the same machine: the first mode sorts by how fast things sink, this one by where they stop.

It is how cell types are separated, and how the isotopes of uranium are told apart in gas centrifuges — by a density difference of under one percent.

Dairy got there first. The De Laval cream separator elsewhere in this corpus is a continuous industrial centrifuge, spinning milk to throw the denser skim outward and take cream from the centre. It transformed dairying decades before centrifuges became laboratory instruments — and the laboratory version is the same machine with a smaller batch and a scientific question.

Then it made biology quantitative. Theodor Svedberg's ultracentrifuge could spin fast enough to sediment proteins, and how fast a molecule sedimented became a measurement of its size — the svedberg unit, still used to name ribosomal subunits. A separation technique became a measuring instrument, which is a promotion few methods achieve.

It settled how DNA replicates. The Meselson-Stahl experiment used density-gradient centrifugation to show that new DNA contains one old strand and one new one, distinguishing between competing models by a difference in density from a heavy nitrogen isotope. It is often called the most beautiful experiment in biology, and the instrument doing the work is the one in step 5.

The same physics, at industrial scale and geopolitical stakes. Gas centrifuges enrich uranium by exploiting a mass difference of about one percent between isotopes, cascaded through thousands of machines. The reason such machines are hard to build — and therefore controllable — is step 4: a rotor spinning fast enough to separate isotopes is at the limit of what materials will survive, and balancing it is the whole art.

Honest limits. It separates by density and size only, so components that match in both stay mixed however long you spin. Delicate structures are damaged by high g. Batch machines process a tube at a time, which is why continuous designs exist for industry. And the energy and engineering cost climbs steeply with speed — every extra g is paid for in rotor strength, which is why the ultracentrifuge is one of the most expensive ordinary objects in a laboratory.