Technology research
Improving the world through magnetics
A research platform for frequency-tunable nonferrous metal separation. We build a special electromagnet that attracts copper and aluminum, and tune its field frequency toward the harder prize: telling nonferrous metals apart from one another.
The recovery gap
The metal is already above ground, sitting in mixed scrap. The bottleneck is the sorting stage: once nonferrous metal is set aside, telling copper from aluminum from titanium is still done largely by hand or by chemical density separation. Both are slow and labor-intensive.
Copper mining carries a real cost to the land around it and the people working it. A faster way to sort nonferrous metal at industrial throughput leans the whole system away from extraction, without asking anyone to use less metal.
How it works
Nonferrous metals are not magnetic, yet this magnet pulls them in. A shorted copper secondary inside the core, driven by an alternating field, induces eddy currents that produce a lagging field and a horizontal force toward the magnet face. The design follows L.R. Crow's 1951 work. Change the frequency and the response changes, which is the whole idea.
The alternating field induces eddy currents in the nonferrous test piece. By Lenz's law they oppose the field, and the shaded-pole geometry turns that into an attractive force.
An H-bridge inverter and DSP drive the magnet from 150 to 1000 Hz, with a switching capacitor bank correcting power factor as the frequency moves.
Each metal should respond strongest at its own frequency, so one magnet could sort them. The 2017-2018 testbed was built to prove that. Showing it is the next step.
Origins
Magnetocs began in 2017 as an electrical engineering senior design project at UC Santa Cruz, built by Evan Zaro, Justin Yau, and Tyler Shewbert (the self-styled "Metal People"). The magnet design is based on L.R. Crow's 1951 research on attracting nonferrous metals with modified electromagnets. The project was sponsored by Dr. Keith Corzine of the UCSC Electrical Engineering department and IDEA Hub of UCSC CIED.
The team delivered a working testbed and demonstrated attraction of copper and aluminum. Proving the magnet can tell the metals apart was left for future work. This platform carries that forward.
The report
The complete 2017-2018 report, from Crow's principle and the magnet builds through the inverter, capacitor bank, sensor network, and what the testbed did and did not prove. Now readable in full on the web.
Assessment
Where Magnetocs stands as an artifact versus a business, who actually owns the recovery and differentiation markets right now, and a SWOT that does not flatter. Published on purpose.
Research threads
Proving one frequency-swept magnet can attract copper, aluminum, and titanium selectively. The unfinished goal.
Read the report → HardwareCrow-geometry magnet, H-bridge inverter, DSP control, and a switching capacitor bank for power factor.
Read the report → EconomicsRetention points near 2 kW at 208 VAC and 600 VDC, common industrial voltages, and what still stands in the way.
Read the report → Built byThe engineer behind the build. Standards work, power systems, and the rest of it.
tylershewbert.com →Media
Build logs, magnet tests, and teardowns are going on the Magnetocs channel. Watch the research as it happens, rather than after it is polished.
Archive