An Iron Ore Separator is a magnetic beneficiation device that pulls iron-bearing minerals out of crushed and ground ore slurry using a rotating drum or belt fitted with permanent or electromagnets. Unlike gravity spirals or flotation cells that exploit density or surface chemistry, it works on magnetic susceptibility — magnetite responds strongly, hematite weakly, and silica gangue not at all. The purpose is to lift Fe content from a run-of-mine grade of 25-35% up to a saleable concentrate of 65-68% Fe. A modern wet drum LIMS at 1,200-2,000 gauss can recover over 98% of liberated magnetite at throughputs of 100-300 t/h per metre of drum width.
Iron Ore Separator Interactive Calculator
Vary slurry flow, drum size, speed, and magnet arc to see separator loading, residence time, surface speed, and capacity utilization.
Equation Used
The calculator divides slurry flow by drum width to get volumetric loading per metre of separator. It also estimates how long particles remain under the fixed magnetic arc from drum speed and arc angle, then compares loading with the 120 m3/h/m upper limit discussed in the article.
- Wet drum separator treating magnetite slurry near the article reference condition of 35% solids.
- Recommended loading is compared with the article's 80-100 m3/h/m sweet spot and 120 m3/h/m upper limit.
- Residence time is based on drum surface travel through the fixed magnetic arc.
The Iron Ore Separator in Action
The separator passes a slurry of ground ore — typically 30-40% solids by weight, with particles below 150 µm — across the face of a rotating drum. Inside the drum sits a stationary arc of permanent magnets, usually neodymium-iron-boron in modern designs, generating a surface field of 1,200-2,000 gauss for low intensity magnetic separation (LIMS) or up to 20,000 gauss for wet high intensity machines (WHIMS) used on weakly magnetic hematite and goethite. Magnetic particles pin to the drum shell, ride out of the slurry pool, and discharge into the concentrate launder once they pass beyond the magnet arc. Non-magnetic gangue — quartz, chert, clay — falls through into the tailings stream.
Design matters because magnetic susceptibility varies wildly across iron minerals. Magnetite (Fe₃O₄) is ferromagnetic and lifts cleanly at LIMS field strengths. Hematite (Fe₂O₃) is paramagnetic with susceptibility roughly 100 times lower, so it needs WHIMS or a roasting step to convert it to magnetite before LIMS will touch it. Get the field strength wrong and you either lose recoverable iron to tails or you grab gangue along with the magnetite, dragging concentrate grade down.
The two failure modes you see most often: undersized magnetic gap, where the drum-to-tank clearance is too tight and magnetic flocs bridge the gap and choke the discharge; and oversized particles, where ore above 1 mm tumbles through too fast for the field to hold it, especially at drum speeds above 30 RPM. Davis Tube tests on a representative sample tell you whether the ore will respond before you commit to a circuit design.
Key Components
- Magnetic Drum Shell: A non-magnetic 304 stainless shell, typically 1.0-1.2 m diameter and 1.5-3.0 m long, rotates around a fixed internal magnet array. Shell thickness sits at 6-10 mm — too thick and you waste field strength across the air gap, too thin and abrasive slurry wears through inside 12 months.
- Internal Magnet Arc: A stationary arc of NdFeB or ferrite magnets covering 90-120° of the drum's lower face. Surface field measured at the shell ranges 1,200-2,000 gauss for LIMS. The arc must be aligned within ±2° of the slurry pool centreline or you get premature discharge of concentrate back into the feed.
- Slurry Tank: Concurrent, counter-current, or counter-rotation tank geometry sets the residence time. Counter-current tanks suit fine feeds below 200 mesh; concurrent tanks handle coarser 0.5-1.0 mm feeds at higher throughput but lower recovery.
- Concentrate Scraper or Spray Bar: A low-pressure water spray, 1-2 bar, washes the magnetic product off the drum into the concentrate launder once it clears the magnet arc. Spray nozzle spacing of 100-150 mm keeps wash water at 0.5-1.0 m³/h per metre of drum length.
- Drive Motor and Variable Speed Reducer: Drum surface speed runs 0.4-1.2 m/s. Slower for cleaner concentrate grade, faster for higher throughput. A 5.5-15 kW gearmotor with VFD lets the operator tune speed against grade in real time.
- Tailings Launder: Discharges the non-magnetic fraction. A magnetic susceptibility check on the tails — using a hand magnet or a Satmagan reading — tells you immediately whether the separator is set correctly. Tails reading above 2-3% Fe magnetic means you're losing recoverable iron.
Where the Iron Ore Separator Is Used
Iron ore separators sit in nearly every iron beneficiation flowsheet on earth, but they also show up in scrap recovery, foundry sand reclamation, and food and pharmaceutical safety lines where any ferromagnetic contamination is unacceptable. The choice between LIMS and WHIMS depends entirely on the mineralogy — magnetite ores get LIMS, hematite and goethite ores get WHIMS or a magnetising roast plus LIMS. Tailings rejection, concentrate grade, and recovery percentage are the three numbers that get watched on every shift report.
- Iron Ore Mining: LKAB's Kiruna underground magnetite operation in Sweden runs LIMS wet drum separators downstream of autogenous mills to produce 67-69% Fe concentrate for pellet feed.
- Hematite Beneficiation: Sino Iron at Cape Preston in Western Australia uses Metso WHIMS units to upgrade weakly magnetic hematite-goethite ore from the Pilbara into a magnetite concentrate.
- Steel Slag Recovery: ArcelorMittal slag yards run cross-belt overband separators to pull ferrous fines and skull from cooled BOF slag for re-melting.
- Foundry Sand Reclamation: Disa and Sinto foundry sand return lines use rare-earth roll separators to strip iron fines and chromite contamination before sand recirculates to the moulding line.
- Mineral Sands: Iluka Resources operates rare-earth roll and induced-roll magnetic separators in their dry mill circuits to split ilmenite, rutile, and zircon based on magnetic susceptibility.
- Coal Preparation: Heavy-media coal washing plants in West Virginia use LIMS drums to recover magnetite media from clean coal and reject streams, returning it to the cyclone feed.
The Formula Behind the Iron Ore Separator
Sizing a wet drum separator comes down to volumetric throughput per metre of drum width and the residence time the slurry spends inside the magnetic field. Push the loading too low and you've overbuilt the circuit and wasted capital. Push it too high — above roughly 120 m³/h per metre of drum width on a standard 1.2 m drum — and recovery collapses because magnetic particles never reach the drum surface before the slurry sweeps past. The sweet spot for magnetite at 35% solids on a 1.2 m diameter drum sits around 80-100 m³/h per metre, which is where most plant designs land.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Qm | Solids mass throughput per metre of drum width | t/h/m | tph/ft |
| Qv | Volumetric slurry feed rate | m³/h | gpm |
| ρs | Slurry density | t/m³ | lb/ft³ |
| Cw | Solids concentration by weight | fraction (0-1) | fraction (0-1) |
| W | Effective drum width | m | ft |
Worked Example: Iron Ore Separator in a Labrador Trough magnetite concentrator
A magnetite concentrator in the Labrador Trough is sizing a primary LIMS wet drum to handle the cobber duty downstream of a 5.5 m primary ball mill. Feed is ground to 80% passing 75 µm at 32% solids by weight. The plant needs to process 280 t/h of solids and they're specifying a 1.2 m diameter drum. The question is what drum width to order from the OEM and whether one drum will do the job or whether they need two in parallel.
Given
- Solids throughput target = 280 t/h
- Cw = 0.32 fraction
- ρsolids = 5.0 t/m³ (magnetite ore)
- Drum diameter = 1.2 m
- Target loading = 90 t/h/m (industry nominal for 1.2 m LIMS drum)
Solution
Step 1 — calculate slurry density at 32% solids by weight, magnetite specific gravity 5.0:
Step 2 — at the nominal industry loading of 90 t/h per metre of drum width, the required width is:
That doesn't fit a single drum — standard OEM drums from Eriez, Metso, and Longi cap at around 3.0 m face length on a 1.2 m diameter shell. So one drum at nominal loading won't quite do the duty. Step 3 — at the low end of the typical operating range, 70 t/h/m, where you'd run for premium concentrate grade on a fine cobber feed:
That's two drums of 2.0 m face each, which is a clean parallel arrangement and gives the operator headroom to drop loading further if grade slips. Step 4 — at the high end of the typical operating range, 120 t/h/m, where some plants run primary cobbers when grade requirements are looser:
One 2.4 m drum would cover it on paper, but recovery typically drops 2-4 percentage points at that loading because the particle bed on the drum gets thick enough that inner-layer magnetite is shielded from the field by outer-layer gangue.
Result
The recommended specification is two 1. 2 m diameter × 2.0 m face LIMS drums in parallel, each loaded at 70 t/h/m for a combined 280 t/h. At the low-end 70 t/h/m loading you'll see concentrate grades around 65-67% Fe with recovery above 97%; at the 90 t/h/m nominal you trade roughly one point of grade for capacity headroom; push to 120 t/h/m on a single drum and recovery falls to 93-95% with visible magnetite in the tailings launder — the operator sees a hand magnet pulling black grains out of the tails sample. If the plant measures recovery below the predicted 97% at nominal loading, the three usual culprits are: (1) feed coarser than the design 80% passing 75 µm, where unliberated composite particles report to tails, (2) slurry density drifting above 35% solids and creating a viscous bed that traps gangue in the concentrate, or (3) a worn shell with field strength dropped below 1,000 gauss at the surface — check with a gauss meter, and replace the shell if reading is below the 1,200 gauss design spec.
When to Use a Iron Ore Separator and When Not To
Magnetic separators are not the only way to upgrade iron ore. The choice between LIMS, WHIMS, gravity spirals, and reverse flotation comes down to mineralogy, particle size, and the grade of feed you're working with. Get this wrong at the flowsheet stage and no amount of operating tweaks will fix it.
| Property | LIMS Wet Drum Separator | WHIMS (Wet High Intensity) | Reverse Flotation |
|---|---|---|---|
| Field strength required | 1,200-2,000 gauss | 5,000-20,000 gauss | Not applicable |
| Suitable feed mineral | Magnetite, pyrrhotite | Hematite, goethite, ilmenite | Hematite, magnetite (after grinding) |
| Throughput per unit | 100-300 t/h per drum | 30-100 t/h per machine | 20-80 t/h per cell row |
| Concentrate grade achievable | 65-69% Fe on magnetite | 60-65% Fe on hematite | 66-68% Fe on either |
| Capital cost (relative) | Low (1×) | High (3-5×) | High (4-6×) |
| Operating cost per tonne | $0.30-0.60/t | $0.80-1.50/t | $2-5/t (reagents dominate) |
| Particle size range | 10 µm - 1 mm | 5 µm - 500 µm | 20-150 µm only |
| Recovery on liberated mineral | 97-99% magnetite | 70-90% hematite | 85-95% either |
Frequently Asked Questions About Iron Ore Separator
Permanent magnet decay is rare with NdFeB at typical slurry temperatures, so the usual cause is mechanical. Check shell wear first — abrasive magnetite slurry erodes 304 stainless at roughly 0.5-1.0 mm per year, and once the shell drops below 6 mm thickness the air gap between magnet face and slurry grows enough to cut surface field strength by 10-15%. Measure with a gauss meter directly on the shell at the lowest point of the magnet arc.
The second cause is magnet array shift. Vibration over time can rotate the internal arc by a few degrees, moving the discharge point relative to the slurry pool. If concentrate is washing back into the feed before reaching the launder, you'll see grade drop and recovery hold steady — that's the diagnostic signature.
Davis Tube tests run on fully liberated material at ideal residence time, so they always report optimistic. The 6-point gap usually comes from incomplete liberation — if your grind is 80% passing 106 µm but the magnetite is locked in 50 µm grains within silica, those composite particles either report to tails or drag silica into concentrate. A microscope check of the tails will show you locked particles immediately.
Slurry feed distribution across the drum width is the second common gap. Uneven feed boxes create dead zones where loading exceeds 150 t/h/m locally even when the average is 90 t/h/m. Pull a feed sample from each end of the drum and the middle — if solids concentration varies by more than 3 percentage points, the feed box needs rework.
Run both in series. A LIMS cobber pulls the magnetite at low cost and high recovery, then a WHIMS scavenger picks up the hematite from LIMS tails. This is the standard flowsheet at operations like Sishen in South Africa and parts of the Carajás circuit. Trying to do the whole job on WHIMS alone wastes capital because WHIMS units cost 3-5× more per tonne of capacity and you'd be using that expensive field strength on magnetite that a $200k LIMS drum could handle.
The only exception is if a Satmagan reading on the head sample shows magnetite below 8-10% — at that point the LIMS stage doesn't justify itself and you go straight to WHIMS or magnetising roast.
Drum surface speed sets the centrifugal force on particles pinned to the shell. Below about 0.4 m/s, particles aren't slung free of the slurry pool fast enough and you carry water and entrained gangue into the concentrate. Above about 1.2 m/s, the centrifugal force on coarse magnetite exceeds the magnetic holding force at the top of the arc and particles fly off prematurely, falling back into the feed.
The rule of thumb: start at 0.6 m/s for fine feeds (-75 µm), 0.8-1.0 m/s for coarser cobber duty. If you're chasing grade, slow down 10% and accept the throughput hit. If you're chasing throughput, speed up but expect grade to fall 0.5-1.0 points per 0.1 m/s above 1.0 m/s.
Run a hand magnet through a dried tails sample. If black grains pull out, the separator is leaving recoverable magnetite behind and you have a setup problem — usually slurry density too high, drum speed too fast, or magnet field weakened. If the iron in the tails is non-magnetic (it stays in the dish), it's hematite or goethite that LIMS physically cannot recover and you need a WHIMS scavenger or to accept the loss.
A Satmagan reading on the tails closes the loop. Magnetic Fe above 1% in tails means separator problem; magnetic Fe below 0.5% with total Fe at 5% means it's locked or non-magnetic iron and the separator is doing its job.
Yes for some duties, but the trade-offs are sharp. Dry drum and roll separators work well on coarser feeds above 1 mm where particles flow freely, and they're standard in mineral sands and slag recovery. Below 200 µm, dry separation collapses because fine particles agglomerate, dust becomes a hazard, and the magnetic force-to-particle-mass ratio drops too low to overcome interparticle friction.
For typical iron ore beneficiation grinds of 80% passing 75 µm, wet processing is the only practical option. Saving water has to come from thickeners and tailings dewatering, not from running the separators dry. The Iron Bridge magnetite project in Western Australia is one of the only operations attempting dry processing at fine sizes, and it required ground-up flowsheet redesign.
References & Further Reading
- Wikipedia contributors. Magnetic separation. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
