The EPG loop, drawn

The Electrical Particle Generator as the archive can model it: a medium — a magnetic slurry, plain water, the HHO off a cell, an argon aerosol — pumped round a non-magnetic tube past a drive coil, and a pickup coil downstream. The question underneath every version Meyer built is whether the medium can be polarised and induce anything in a coil it passes through, and how much. The ceiling is stated first; then what the remanence, the forgetting and the flow's smearing leave of it; then what that is into a load, against what the drive and the pump cost. Click a coil to change it, drag it along the tube, and stop the fluid to see what the drive alone does — that checkbox, on a bench, is the experiment.

The medium · 5 % Iron (carbonyl, soft) (1 µm) in water

Density: 1.34e+3 kg/m³
Viscosity: 1.14 mPa·s
Susceptibility: 0.15
Saturation: 8.5e+4 A/m
Remanence: 12 A/m
Holding time: 0.394 s
Settling: 3.74e-6 m/s
Flow: 50 in/s · Re 3.73e+4 · turbulent
Ceiling at the first pickup: 1.64 V

Soft Iron (carbonyl, soft) keeps almost nothing: 240 A/m per particle against a saturation of 1700 kA/m — the slug that leaves the coil is nearly blank, at a remanence ratio of 0.00014. A hard particulate is the lever: the remanence of a multi-domain sphere is about three times its coercivity, so a material with a hundred times the coercivity carries a hundred times the magnetisation out of the coil.
At 1 µm this Iron (carbonyl, soft) particle is above its single-domain limit of 15 nm, so it is many domains and the Néel picture does not apply: a multi-domain particle changes its magnetisation by domain-wall motion, which needs a field above its coercivity of 80 A/m. With no field on it, the only way its orientation changes is the whole particle rotating — the effective time is the Brownian one, 393.6 ms.
At the holding's 50 in/s (1.27 m/s) the medium travels 499.8 mm in one relaxation time — the distance downstream of the drive coil over which a magnetised slug still exists.
The particle falls one bore in 6687 s, which is about 8,493 laps of a one-metre loop at the holding's 50 in/s (0.79 s a lap). In a running loop the flow keeps it up for that long; the moment the pump stops, the clock starts.
Re = 37292: turbulent. The front stays flat to within 148.7 mm, but the eddies turn the fluid over every 369.2 ms, which randomises the particles' orientation as fast as it flattens the front — the relaxation calculation says whether that costs the magnetisation.

The loop needs script to draw and move. Without it, the same chain with every formula shown is the composite calculation at this point.

Built on Design an EPG loop: medium, drive, transit, pickup and the energy accounting, which runs the same chain at one operating point with every formula shown — the studio's "check this point" link lands there pre-filled. Every material number the browser moves about was derived on the server through a named calculation in the EPG medium, flow and coils groups. Nothing here is Meyer's except the speeds and the pump the EPG #1 holding gives; the recipes, the coils and the particulates' coercivities are engineering choices, and every preset says so.