A grinding mill is a rotating drum that reduces ore, clinker, or grain to a target particle size by tumbling the feed against grinding media. The grinding media — steel balls, rods, or autogenous rock — is the active component that delivers the impact and abrasion energy that fractures each particle. The mill exists to liberate valuable mineral from gangue or to produce a controlled fineness for downstream processes. A modern 7 m × 12 m SAG mill at a copper concentrator can chew through 2,500 tonnes per hour of run-of-mine ore.
How the Grinding Mill Works
A grinding mill is, at its core, a horizontal rotating shell loaded with feed and grinding media. As the shell turns, lifter bars on the inside cast the charge upward in a cataracting arc, and the falling media impact the ore at the toe of the charge. Below that you get cascading — media rolling over each other in shear — which delivers attrition grinding for the finer fractions. The split between cataract and cascade is set by mill speed, expressed as a percentage of critical speed. Critical speed is the rotational speed at which centrifugal force pins the outermost media to the shell and grinding stops dead. Most ball mills run at 70-78% of critical, SAG mills at 70-80%, and rod mills lower at 60-68% to keep the rods aligned and prevent tangling.
Why is it built this way? Because comminution — the breaking of solids into smaller pieces — is brutally energy-inefficient. Less than 5% of the electrical input ends up as new fracture surface. Everything else dissipates as heat, noise, and liner wear. Drum geometry, lifter profile, and charge fill (typically 30-40% of mill volume) get tuned to maximise the energy fraction that actually lands on a particle at the right angle to break it. If you under-fill the mill, the media throws against bare liner and you tear out lifters in weeks. Over-fill past about 45% and the toe of the charge climbs too high, the cataracting arc collapses, and throughput drops while power draw spikes.
Tolerances and timing matter more than people expect. A worn lifter bar — say, profile height down from 150 mm to 60 mm — flattens the cataract trajectory and you'll see oversize in the cyclone overflow within a shift. Bolt-up torque on the shell liners must be re-checked after the first 24 hours of new-liner operation; a loose liner backing plate destroys the shell weld in days. Common failure modes are trunnion bearing oil-film collapse from cooling-water loss, ring-gear pinion misalignment showing up as audible clatter, and ball-charge segregation where small media migrate to the discharge end and starve the feed-end of impact energy.
Key Components
- Mill Shell: The pressure-bearing rotating cylinder, fabricated from rolled steel plate 50-90 mm thick on large mills. The shell carries the charge weight (often 200-600 tonnes on a large SAG) and transfers torque from the ring gear to the liners.
- Liners and Lifter Bars: Replaceable wear plates bolted to the inside of the shell. Lifter heights of 100-200 mm and face angles of 14-30° determine how aggressively the charge is thrown. Liners are chrome-moly steel or rubber-and-steel composites and last 6-18 months depending on ore abrasiveness.
- Grinding Media: Forged or cast steel balls (25-125 mm diameter) for ball mills, rolled steel rods (75-100 mm diameter, slightly shorter than the mill length) for rod mills, or in SAG mills the ore itself plus a 6-15% steel-ball charge. Media hardness is typically 60-65 HRC.
- Trunnion Bearings: Hydrostatic or hydrodynamic plain bearings that support the mill ends. They float the mill on a 50-100 µm oil film at 4-6 MPa. Lose the high-pressure pump for more than 30 seconds at full load and you wipe the bearing.
- Ring Gear and Pinion Drive: A girth gear bolted around the shell, driven by one or two pinions geared to slow-speed motors. Backlash must hold to 0.4-0.8 mm; outside that range you'll hear gear chatter and tooth-flank pitting begins.
- Discharge Grate or Overflow: Controls product residence time. Grate discharge mills hold a lower charge level and produce a coarser product faster; overflow mills hold ore longer for a finer grind.
Who Uses the Grinding Mill
Grinding mills sit at the front end of nearly every operation that processes a hard, granular feedstock into a controlled-size powder. The economics of mining, cement, power generation, and even pigment production depend on getting the right grind at the lowest kWh per tonne. The choice between ball mill, rod mill, SAG mill, or vertical roller mill comes down to feed size, target product size, ore hardness expressed as Bond work index, and tonnage. Where the customer needs P80 of 75 µm for flotation, a ball mill in closed circuit with hydrocyclones is the standard answer. Where they need 200 µm for heap leach, a SAG mill alone often does the job.
- Copper Mining: The 40 ft × 26 ft SAG mill at the Escondida concentrator in Chile, driven by a 28 MW gearless ring motor, processes copper porphyry ore to roughly 2 mm before secondary ball milling.
- Cement Production: Two-compartment ball mills at Holcim's Lägerdorf plant grinding clinker and gypsum to 3,200-4,000 cm²/g Blaine fineness for OPC.
- Gold Processing: Rod mill followed by ball mill at Newmont's Carlin operation in Nevada, grinding refractory gold ore to P80 of 75 µm ahead of pressure oxidation.
- Coal-Fired Power: Vertical roller pulverisers — like the Loesche LM 23.2 D — grinding bituminous coal to 70% passing 75 µm before injection into boiler burners.
- Iron Ore Pelletising: Wet ball mills at LKAB's Kiruna concentrator producing the 45 µm fines needed for green-pellet balling drums.
- Industrial Minerals: Imerys ground calcium carbonate plants using stirred bead mills for sub-2 µm pigment grades in paper coating.
- Flour Milling: Bühler MDDK roller mill stands at a North American hard red wheat mill, reducing wheat to 75 µm flour at extraction rates above 75%.
The Formula Behind the Grinding Mill
Critical speed sets the entire operating window of a tumbling mill. Run too far below it and the charge just slumps and rolls — you get attrition but almost no impact, and coarse particles never break. Run at or above 100% and the media centrifuges against the shell, grinding stops, and power draw collapses. The sweet spot for a ball mill sits around 72-75% of critical, which gives you the parabolic cataract trajectory that lands media on the toe of the charge with maximum kinetic energy. The formula below tells you the absolute speed limit; what you actually run is a fraction of it tuned to your ore and liner profile.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Nc | Critical rotational speed of the mill shell | rpm | rpm |
| D | Internal diameter of the mill measured liner-to-liner | m | ft (use 76.6 / √(D−d) form) |
| d | Diameter of the largest grinding ball | m | ft |
Worked Example: Grinding Mill in a phosphate rock ball mill
A phosphate beneficiation plant in Khouribga, Morocco is commissioning a 4.0 m internal-diameter wet overflow ball mill charged with 80 mm forged steel balls to grind apatite ore to P80 of 106 µm ahead of flotation. Engineering needs the operating speed for the variable-frequency drive setpoint, plus an idea of how throughput and grind feel will change across the typical 65-78% of critical operating window.
Given
- D = 4.0 m
- d = 0.080 m
- Target operating fraction = 72 % of critical
Solution
Step 1 — compute the effective grinding diameter, accounting for ball size:
Step 2 — apply the critical speed formula to get the absolute upper limit, where centrifugal force would pin the outermost ball to the shell:
Step 3 — at the design nominal of 72% of critical, where most ball mills hit the sweet spot of cataracting impact plus cascading attrition:
Step 4 — at the low end of the typical window, 65% of critical:
At 13.9 rpm the charge cascades more than it cataracts. You get a finer product because residence time goes up and impact energy drops, but throughput falls noticeably — a Khouribga-class mill would lose roughly 8-12% in tonnes per hour at this speed. Operators often pull back to this region when the ore feeds harder than design.
Step 5 — at the high end, 78% of critical:
At 16.7 rpm the cataract arc is steep and balls land hard at the toe. Throughput climbs, power draw rises with it, and liner wear accelerates. Push past 80% and you start seeing balls pitching directly onto the opposite shell wall instead of the charge toe — the audible signature changes from a steady rumble to a sharp clack-clack and you'll be replacing shell liners in 9 months instead of 14.
Result
The VFD setpoint is 15. 4 rpm at the 72% nominal operating point, sitting comfortably below the 21.4 rpm critical limit. In practice the operator hears a steady deep rumble and sees mill power draw track within ±3% of the design figure for that ore hardness. Across the 13.9 to 16.7 rpm operating window the mill trades throughput against grind fineness — slow for finer product, fast for tonnage — with the sweet spot around 15.0-15.6 rpm for typical Khouribga apatite. If your measured power draw is 15% below the design value at 15.4 rpm, suspect (1) charge fill below 32% letting media impact bare liner and waste energy as heat, (2) excessive water addition flooding the mill so the charge slurries rather than cataracts, or (3) trunnion seal leakage dropping the slurry density and shifting the toe position forward of the lifter strike zone.
Choosing the Grinding Mill: Pros and Cons
The choice of grinding mill is rarely about which mechanism is best in the abstract — it's about feed size, product size, ore competence, and capital available. Ball mills, SAG mills, and vertical roller mills each occupy a different niche, and picking wrong costs millions in throughput shortfall.
| Property | Ball Mill | SAG Mill | Vertical Roller Mill |
|---|---|---|---|
| Feed size (top size) | 25 mm | 300 mm | 100 mm |
| Product P80 range | 20-300 µm | 1-3 mm | 30-100 µm |
| Specific energy (kWh/t for hard ore) | 10-20 | 6-12 | 5-9 |
| Capital cost (relative) | 1.0× baseline | 2.5-3.5× baseline | 1.8-2.2× baseline |
| Liner replacement interval | 6-12 months | 4-9 months | 12-24 months |
| Suitable application | Secondary grinding for flotation feed | Primary grinding from ROM ore | Cement, coal, slag dry grinding |
| Throughput per unit (large size) | up to 600 t/h | up to 4,000 t/h | up to 800 t/h |
| Sensitivity to feed-size variation | Low — protected by SAG ahead | High — needs steady ROM feed | Moderate |
Frequently Asked Questions About Grinding Mill
Power draw on a tumbling mill is proportional to the mass of the charge and the position of its centre of gravity relative to the mill axis. If you've verified speed and volumetric fill but power is short, the charge density is wrong. Two real causes dominate: ball-charge wear has dropped your top-size from 80 mm to 60 mm (smaller balls give a tighter packed charge but a lower lever arm), or slurry density is below 70% solids by weight so the charge is sloshing rather than tumbling as a coherent mass.
Quick check — drop a sample from the mill discharge into a Marcy density cup and confirm slurry density. If it's at 62-65% solids, cut dilution water at the feed and watch power climb back within an hour.
The crossover point is ore competence, measured by SAG Mill Comminution (SMC) test or A×b value. For competent ores (A×b below 40, like a hard quartz porphyry) a three-stage crushing circuit ahead of ball mills uses 15-20% less total energy than SAG, but doubles the building footprint and adds dust handling. For softer ores (A×b above 70) SAG dominates economically because crushers struggle with the clays and the crusher liners eat themselves.
Tonnage matters too — below about 8,000 t/d, three-stage crushing is usually cheaper to build and operate. Above 30,000 t/d almost every greenfield project goes SAG-ball because the crusher count and conveyor complexity become unmanageable.
Broken or chipped balls (called "scats") in the discharge mean either ball quality is off-spec or the mill is running too fast. Forged balls should pass a 10× drop test from 5 m onto a steel anvil without spalling; if your supplier shipped under-tempered media at 55 HRC instead of 62 HRC, they'll fragment in service. The other cause is over-speeding above 80% of critical, where balls strike shell-on-shell at the 12 o'clock position instead of landing on the charge toe.
It's not immediately dangerous but it's expensive — broken scats jam the discharge grate, drop throughput, and can damage downstream pump impellers. Pull a sample, check Brinell hardness, and verify your VFD setpoint matches the calculated nominal speed.
Rods don't tolerate cataracting motion. Above about 70% of critical the rods stop staying parallel — they cross over each other, tangle, and bend into bananas within hours. A bent rod jams against the shell, smashes liner bolts, and in a worst case punches through the discharge trunnion liner.
If you find rods bent into a U shape during a reline, the mill has been over-sped at some point. Rod mills want a steady cascade, not impact, which is why the design rule is 60-68% of critical and operators are trained to never push them.
Use the Bond ball-size formula as a starting point: top-size ball diameter (mm) ≈ (F80 in µm / K)0.5 × constant, where K depends on ore work index. For typical hard rock with Wi around 14 kWh/t and an F80 of 12 mm, the calculation gives roughly 75-80 mm top size.
For make-up, charge only the top size — say 80 mm balls — and let natural wear produce the full ball-size distribution inside the mill. Adding a mixed-size make-up does not help; the mill self-classifies media as it runs, and you'll just pay for grinding work the smaller balls were going to do anyway after one month of wear.
SAG throughput is governed by the coarsest fraction in the feed, not the average. A pit phase change often shifts the rock mass structure — joint spacing, alteration, blast fragmentation — without changing the chemical assay. If the new ore is more competent in the 50-150 mm range, that fraction builds up inside the mill as a critical-size pebble load and chokes the grate.
Two diagnostic moves: pull a belt cut and screen for the 50-150 mm fraction, and check pebble-port discharge tonnage — if it's spiked from 8% to 15% of feed, you're seeing critical-size buildup. The fix is either coarser pre-crushing of the SAG feed or a pebble crusher in recycle.
Yes, with caveats. Rubber and rubber-steel composite liners typically last 1.5-2× longer than plain chrome-moly in wet ball mills below 4.5 m diameter, and they cut noise by 8-12 dB. The catch is heat — rubber liners can't run hotter than about 80 °C, so dry grinding or high-temperature applications are out, and you must verify your slurry temperature.
The other constraint is ball size. Rubber liners struggle with balls above 75 mm because impact energy at the toe exceeds what the rubber can absorb, and lifter bars get torn out at the rubber-to-steel interface. If your top-size charge is 80 mm or larger, stay with steel or use a steel-capped composite design.
References & Further Reading
- Wikipedia contributors. Mill (grinding). Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
