A bead mill and a three roll mill both reduce particle size in a liquid. They do it through completely different physics, and the physics — not the price, not the brand — decides which one your product will tolerate. Below roughly 10,000 to 15,000 mPa·s a media mill circulates the product, grinds it continuously and reaches finer than any roll ever will. Above that, the product stops moving through the chamber at all, and the rolls become the only route that still works.
This guide puts the two routes on the same axis: viscosity, achievable fineness, heat, contamination, failure modes and the numbers a quote actually depends on. It is written for the buyer who has to choose one of them, not for someone collecting machine descriptions.
The choice is decided by one measurable property, and it is your product’s apparent viscosity at the shear rate it sees in the machine. A bead mill needs the product to flow; a three roll mill only needs it to be spreadable. If the slurry circulates, the media mill grinds it continuously and reaches finer. If it does not, media milling is not a tuning problem, it is the wrong machine.
| Apparent viscosity | Does it circulate through a chamber? | Route that works |
|---|---|---|
| Up to about 5,000 mPa·s | Yes, with margin | Bead mill — usually the fastest way to a fine result |
| 10,000 to 15,000 mPa·s | Marginally, and only with the right pump and screen | Bead mill sized for it, or a basket mill on a batch basis |
| 50,000 to 1,000,000 mPa·s | No | Three roll mill |
| Above 1,000,000 mPa·s | No | Three roll mill with hydraulic loading; planetary mixer upstream |
The ceiling is not a single number. Our own documentation quotes both 10,000 and 15,000 mPa·s as the pumpable limit depending on machine and pass strategy, so treat it as a band and confirm it on your own material rather than buying to the edge of it.
The other half of the answer is what you are trying to achieve. A media mill is the only one of the two that reaches below a micron, because fineness there is set by bead diameter rather than by a mechanical gap. Our PZB and PZL series run media from 0.2 mm and are specified down to 50 nanometres. A production three roll mill works in the 1 to 10 micron band and is chosen for what it does to the paste, not for the number on a gauge.
Read the rest of this guide in that order: viscosity first, because it eliminates one route outright; then fineness, because it eliminates the other if you need nano; then heat, contamination and cost, which decide the configuration once the family is fixed.
A media mill is a loop, not a vessel. Product enters the chamber, is thrown against the bead charge by a rotating pin or disc, leaves through a separation screen and returns to the tank. Every one of those steps is a hydraulic requirement, and the paste has to satisfy all of them at once: it must be pumpable, it must be turbulent enough in the chamber for the beads to do work, and it must pass the screen without blinding it.
Thicken the product and the three requirements fail in sequence. First the screen blinds, because the flow that used to sweep it clean is gone. Then the chamber starves, because the pump delivers pressure instead of volume and the bead charge compresses against the outlet. Finally the temperature climbs, because the mill is now putting its drive power into shearing a nearly static mass instead of into a flowing film. Operators usually see this as a machine "losing power", and the next thing they do is add passes, which makes it worse.
A three roll mill has no loop to satisfy. The paste is fed into the first nip, transferred as a film on the middle roll, and taken off the fast roll by an apron. Nothing has to be pumped and nothing has to pass a screen; the only requirement is that the material can be made to adhere to a roll. That is why the two families do not overlap much: one is a flow machine, the other is a film machine.
Inside a bead mill the energy is carried by the media. A pin rotor or a disc accelerates a charge of zirconia beads, and the beads collide with each other and with the product in the gaps between them. Grinding happens in a volume, and the smallest particle the volume can produce scales with the bead size, which is why a media mill is the only route to nanometres.
Inside a three roll mill the energy is carried by a velocity step. Three rolls are stacked horizontally and turn at different speeds — our machines use a helical gearbox with a nominal 1 : 3 : 9 ratio across the slow feed roll, the middle transfer roll and the fast discharge roll. Paste is drawn through a feed nip and then a narrower discharge nip, and at each nip the two surfaces move at different velocities. The film between them is stretched and torn in laminar flow rather than struck by media.
The consequence for the product is not cosmetic. Media collision is a statistical process — some particles are hit many times and some are hit few times, so the distribution is wide and passes are needed to tighten it. Nip shear is a deterministic one: every element of paste that goes through the nip sees the same velocity gradient, so the whole batch is treated the same way and a small number of passes is enough. This is why a roll mill can reach a repeatable Hegman reading in one to three passes while a media mill is optimized by residence time and pass count.
Ball mills are included because buyers often arrive with a ball mill quote in hand and a bead mill price next to it. The three are not interchangeable, and the row that usually settles the discussion is the second one: only one of them puts a wearing medium into your product.
| Criterion | Bead mill | Three roll mill | Ball mill |
|---|---|---|---|
| How it grinds | Media collisions and shear in a chamber | Laminar shear at two nips between rolls turning at different speeds | Tumbling balls in a rotating drum |
| Media in contact with product | 0.2 to 2.0 mm zirconia beads | None — rolls and apron only ✓ | Steel or ceramic balls, large |
| Viscosity it can handle | Pumpable, up to about 10,000 to 15,000 mPa·s | 50,000 mPa·s and above ✓ | Wide, but slow at high solids |
| Fineness it reaches (D90) | 50 nm to 15 µm, by series ✓ | 1 to 10 µm | 1 to 50 µm |
| Running mode | Continuous | Batch passes | Batch |
| Throughput | Hundreds to thousands of L/h ✓ | Scales with roll length, not flow | Low per unit volume |
| Heat removal | Jacketed chamber plus rotor cooling | Chilled cores through the rolls ✓ | Poor |
| Color change | Flush and circulate | Scrape the apron and re-feed ✓ | Empty and clean the drum |
| Product hold-up | Chamber volume plus pipework | Film on rolls and apron ✓ | Drum charge |
| Wear parts | Screen, rotor, chamber liner | Roll surface, apron blade | Liners, balls |
| Main shutdown risk | Mechanical seal leak | Roll surface damage | Liner failure |
| Where it is still the right answer | Nano slurries, low viscosity, high volume ✓ | Inks, silver paste, lipstick, high-solids paste ✓ | Minerals and coarse reduction |
✓ marks the strongest of the three on that criterion, not an absolute claim: a bead mill reaches finer than a roll mill, and a roll mill is cleaner because it adds no media, but neither is better on every row.
The table below is the same information as the criteria table, cut the other way: start from your product and read across to the machine. The bead mill rows are the four architectures in our own range, and the numbers are the published figures for those series rather than general industry claims.
| Machine and series | Viscosity it handles | Fineness (D90) | Media | Typical product |
|---|---|---|---|---|
| PZB pin type (horizontal) | Pumpable, up to about 10,000 to 15,000 mPa·s | 50 nm to 5 µm | 0.2 to 1.5 mm | Inkjet ink, automotive OEM coating, battery paste |
| PZP static disc | Pumpable | 1 to 15 µm | 0.6 to 2.0 mm | Architectural paint, flexo ink, agrochemical SC |
| PZD dynamic disc | Pumpable | 2 to 15 µm | 0.8 to 2.0 mm | High-solids primer, marine barrier paint, dye paste |
| PZL vertical nano | Pumpable, small batch | 50 nm to 5 µm | 0.2 to 1.5 mm | Small-batch and trial production of nano material |
| PTM / PHM three roll | 50,000 mPa·s and above, past 1,000,000 on hydraulic models | 1 to 10 µm | None | Offset and gravure ink, silver paste, lipstick mass |
Media and fineness here follow our product database, which states those two figures at series level: PZB and PZL run 0.2 to 1.5 mm media to 50 nm to 5 µm, and PZP is listed at 0.6 to 2.0 mm with a 1 to 15 µm working range. Where a series can physically accept a finer bead than its stated band, that is a separation-system question, not a marketing one, so ask before you specify it. The three roll window comes from our three roll mill guide, which gives the working viscosity range as above 50,000 mPa·s and quotes 3,000,000 mPa·s as the upper end for roller configurations; the 1,000,000 mPa·s figure is the one our specification guide uses for production machines.
The fastest way to settle a machine argument is to run the numbers in front of both parties. This is the chain we walk through with a buyer who is making a conductive or high-solids paste: every input is stated so you can substitute your own and reproduce the result.
| Step | Input or formula | Result for this paste |
|---|---|---|
| 1. Measure viscosity | Apparent viscosity at the shear rate the machine will impose | 60,000 mPa·s |
| 2. Test against the media ceiling | Pumpable limit of a bead mill chamber and screen | About 10,000 to 15,000 mPa·s — exceeded by roughly four times |
| 3. Route decision | Above the ceiling, media milling is not a tuning problem | Three roll mill |
| 4. Convert the fineness target | Hegman 7.5 on a grindometer | Roughly 2 to 5 µm |
| 5. Passes | Published: 1 to 3 passes reach Hegman 7.5 to 8 | Plan 2 passes, verify by gauge after each |
| 6. Slow roll surface speed | v = π × D × n / 60, D = 0.260 m, n = 24 rpm | 0.327 m/s |
| 7. Middle roll surface speed | Same formula, n = 68.5 rpm | 0.932 m/s |
| 8. Shear rate at the feed nip | γ̇ = (v2 − v1) / gap, gap = 3 µm | About 2.0 × 10⁵ s⁻¹ |
| 9. Check the drive size | PTM260: 7.5 kW on a 675 mm roll face | Move to PTM406 at 15 kW only when throughput climbs |
Steps 6 to 8 are arithmetic on the PTM260 geometry and the factory nominal roll speeds published in our three roll mill guide; they show how the shear rate is generated and they are not a performance figure. Steps 1 to 5 are the decision itself, and step 1 is the only one that needs a measurement from you before a supplier can size anything.
Two things in that chain are worth reading twice. Step 3 is a route decision, not a tuning step — once the paste is four times over the pumpable ceiling, no rotor design or bead size fixes it. And step 9 is where buyers normally overspend: the drive size on a roll mill follows the roll face and the throughput, so buying the largest model up front costs money that the first year of production will not recover if the volumes never arrive.
If the viscosity test sends you down the media route, the next three decisions are media diameter, chamber volume and flow rate, and they are not independent. Media diameter sets the fineness ceiling; chamber volume and flow together set how long the product stays in the grinding zone; and the separation system has to be able to retain whatever bead you chose. The PZB pin type that runs the middle band is on our pin type bead mill page with its own chamber and power table.
| Media diameter | Fineness this band reaches | Series that runs it | What it costs you |
|---|---|---|---|
| 0.2 to 0.5 mm | 50 nm to 1 µm | PZB pin type, PZL vertical nano | Needs a fine separation system; a standard static screen will not retain the bead |
| 0.6 to 1.2 mm | 0.5 to 5 µm | PZB pin type, PZP static disc | The widest working window in the range, and the usual starting point for pigment and ink |
| 1.2 to 2.0 mm | 2 to 15 µm | PZD dynamic disc | Fast and tolerant of coarse feed, but will not reach the sub-micron range |
Bands derived from the published media ranges in our product database, not from measurements on your material. The point of the last column is that the finest bead is not automatically the right choice: it is only usable if the separation system is built for it.
Residence time is the number buyers ask about least and complain about most. It is simply chamber volume divided by flow, so a large chamber run at a low flow gives a long single-pass time and a narrow distribution, while the same chamber run at maximum flow gives short, repeated exposures. Neither is correct in the abstract — it depends on how much energy the product can absorb before its temperature or its color strength gives out.
| Model | Chamber volume | Published flow range | Residence per pass |
|---|---|---|---|
| PZB-20 | 20 L | 300 to 600 L/h | 2.0 to 4.0 minutes |
| PZB-60 | 60 L | 600 to 1,500 L/h | 2.4 to 6.0 minutes |
| PZB-150 | 150 L | 800 to 2,000 L/h | 4.5 to 11.3 minutes |
Calculated as chamber volume divided by flow, from the published PZB parameters. It is single-pass residence, before any recirculation, and it is the figure to quote when a supplier asks how long the product may be in the mill.
On the roll route the two settings that decide the product are the nip gap and the roll speed. The speed ratio is fixed by the gearbox, so what an operator actually adjusts is the gap and, on hydraulic machines, the contact pressure. The working range is roughly 1 to 5 microns at the discharge nip, and at that scale the resolution and the parallelism of the adjustment matter more than the number on the dial.
| Model | Roller Ø (mm) | Working length (mm) | Roll speeds (rpm) | Motor (kW) | Weight (kg) |
|---|---|---|---|---|---|
| PTM150 | 150 | 300 | 26 / 78 / 205 | 2.2 | 500 |
| PTM260 | 260 | 675 | 24 / 68.5 / 195 | 7.5 | 2,000 |
| PTM406 | 406 | 810 | 16 / 48 / 145 | 15.0 | 5,300 |
| PTM406L | 406 | 1,000 | 16 / 48 / 145 | 22.0 | 5,800 |
Verbatim from our published PTM table. Roll speeds are the factory nominal values at full drive speed. Voltages, hydraulic pressure systems, variable frequency drives, ceramic rollers and explosion-proof configurations are engineered on request.
Read that table as three capacity classes rather than four models. PTM150 is a bench or pilot machine at 2.2 kW; PTM260 is the mid-volume production unit and the one most ink plants start with; PTM406 and PTM406L are the same roll diameter with different face lengths, so the jump from 15 to 22 kW buys throughput and nothing else. Going from 260 to 406 mm changes the geometry, the film length and the whole installation — it is not an upsizing of the same machine. The individual configurations are listed on the PTM triple roller mill page.
The fineness limit of a media mill is the bead. Grinding happens in the gaps between beads, so the smallest particle a charge can produce scales with the bead diameter; run 0.2 mm media in a PZB pin mill and the series is specified down to 50 nm. Run 0.8 to 2.0 mm media in a PZD and the floor moves to 2 microns, not because the machine is weaker but because the beads are too big to reach into smaller gaps.
A roll mill has a mechanical floor instead. The two rolls cannot be brought into contact — they would destroy each other — so there is always a finite gap, and the film thickness at that gap is the smallest dimension the shear field can impose. In production that means roughly 1 to 5 microns at the discharge nip and typical output of 1 to 10 microns. Hegman 7.5 to 8, which is the reading most ink specifications actually call for, sits comfortably inside that window in one to three passes.
This asymmetry matters commercially because it is often quoted the wrong way round. Buyers ask which machine is "better" and expect the answer to be the one that goes finer. But almost no ink, paste or cosmetic product needs 50 nanometres, and specifying a media mill for a paste that only rolls will produce is how plants end up with a machine they cannot feed.
Both machines put nearly all of their drive power into the product as heat, and both remove it through water, but the geometry of the removal is different and that decides which products each can protect.
| Point of comparison | Bead mill | Three roll mill |
|---|---|---|
| Where the heat is generated | Through the bead charge, in a volume | In the film at each nip |
| Cooling surface | Jacketed chamber plus internal rotor cooling channels | Chilled water through the core of every roll |
| Published performance | Spiral-baffled jacket and rotor cooling hold the slurry below 35 °C | Chilled cores keep the roll face cold through the run |
| What fails first at the limit | Solvent loss, flocculation, shade drift | Viscosity drift and drying on the apron |
| Practical consequence | Heat-sensitive formulations need the cooling specification fixed up front | The paste never sits in a hot chamber, so brief excursions are survivable |
Bead mill figures are from the cooling specification on our bead mill guide. The roll description reflects the chilled-core construction on our PTM and PHM lines; we do not publish a temperature drop figure for the roll route, so none is claimed here.
The trade is subtler than "rolls run cooler". A media mill can hold a thin slurry below 35 °C indefinitely because the jacket and the rotor both have cold surface in contact with moving product. A roll mill cools through the roll body, which is very effective at the film but does nothing for the paste sitting in the feed trough. What the roll route really buys is absence of dwell time in a heated chamber — the product passes through in seconds rather than circulating for minutes.
Most "we bought the wrong mill" conclusions are actually four faults, and three of them are cheaper to fix than a new machine. The pattern below is the one we work through on the phone before anyone discusses a purchase order.
| Symptom | Actual cause | Fix before replacing anything |
|---|---|---|
| The chamber starves or blocks on a thick paste | Product is above the pumpable ceiling, so the pump delivers pressure instead of volume | Cut the solids or move the step to a three roll mill; this is a route problem, not a setting |
| Fineness plateaus around 15 microns | Media too coarse for the target, or agglomerates never wetted out upstream | Step the media down a band and confirm the separation system retains it; pre-disperse first |
| Color drifts darker across the run | Over-grinding past the strength peak, with the extra energy arriving as heat | Shorten the pass and lower the chamber temperature before adding more passes |
| Beads appear in the product | Separation gap is larger than the smallest bead in the charge | Match the screen to the bead you actually run, and stop mixing two bead sizes in one charge |
The first row is the only one that ends in a different machine. The other three are media choice, separation setup and pass discipline, and all three are cheaper than a capital purchase.
Comparing purchase prices between the two families is misleading, because what recurs is different. A media mill consumes beads and a chamber liner; a roll mill consumes roll surface and apron blades. Over five years the cheaper machine to buy is often the more expensive one to own, and which of them that is depends on whether your product is abrasive.
| Cost line | Bead mill | Three roll mill |
|---|---|---|
| Drive size across our range | 15 kW on a 6 L chamber to 160 / 200 kW on a 150 L chamber | 2.2 kW on PTM150 to 22 kW on PTM406L |
| Recurring consumable | Grinding media, chamber liner, separation screen | Roll re-grinding, apron blades, gap calibration |
| Failure that stops the line | Mechanical seal leak under chamber pressure | Roll surface damage |
| What you should hold as spares | A bead charge and a spare screen | A spare apron blade and a feeler gauge set |
| Where the money is saved | A correctly banded bead and a screen that matches it | Hydraulic loading instead of handwheels on high-volume runs |
Drive sizes are the published figures for the PZB and PTM ranges. We are not quoting prices here: on both routes the configuration — chamber lining, seal type, separation system, roll material — moves the number more than the model does, which is why both are quoted against a specification.
A supplier who receives these nine lines can price the right machine the first time. A supplier who receives only a product name will quote the model they sell most of, which is how mismatches get manufactured.
| # | What to state | Why the quote depends on it |
|---|---|---|
| 1 | Apparent viscosity, and the shear rate you measured it at | Decides whether a media mill is allowed at all |
| 2 | Target fineness and the gauge you use | Hegman, grindometer and laser diffraction give different numbers for the same paste |
| 3 | Solids loading and vehicle | Sets wetting behavior and whether the product can be pumped |
| 4 | Batch size and required output per hour | Sizes the chamber on one route and the roll face on the other |
| 5 | The temperature ceiling the product tolerates | Decides whether standard cooling is enough |
| 6 | Solvent system and site classification | Explosion-proof configurations available upon request where required |
| 7 | Metal contamination limit | Decides chamber lining, or ceramic rolls on the roll route |
| 8 | How often you change color | Decides whether the cleaning strategy is a flush cycle or an apron scrape-down |
| 9 | Utilities actually available on site | Power, chilled water temperature and flow all change the specification |
Line 1 is the one that gets skipped and the one that decides the family. If you send nothing else, send the viscosity and the fineness target.
Some products genuinely belong to neither family, and saying so early saves a bad purchase. Three cases come up often enough to write down.
The product is a dry powder. Neither machine is a dry process. A bead mill has no way to convey dry solids and a roll mill needs the material to adhere to a roll. This is a different machine family and specifying either of these two for a dry route is a specification error, not a sizing error.
The step is dispersion, not size reduction. If the powder still needs wetting out and the agglomerates are large, the right next step is a high speed disperser or a basket mill, and the mill comes after. Putting an un-dispersed powder straight into a bead mill multiplies passes without improving the reading, which is the second row of the failure table above.
The volume is small and the color changes constantly. A single-vessel basket mill is often the better answer for short runs: it grinds in the same tank it disperses in, so hold-up and cleaning are a fraction of a circulation loop. Our basket mill guide covers where that trade stops being worth it.
This guide covers the choice between two machine families. The questions below each need a page of their own, and each is answered by the single best hand-written page we have on it.
POLYC Technology manufactures wet grinding and dispersion machinery — bead mills, basket mills, three roll mills and high speed dispersers — at its 9,000 m² facility in Wuxi, Jiangsu, China, and has been building process machinery since 2006. We build both of the machine families this guide compares, which is the only reason a comparison between them is worth reading from a supplier: we have no interest in steering every enquiry to one route.
The series and model tables in this guide are taken from our own published product data, and the shear rate and residence time figures are arithmetic on those published dimensions and speeds with the inputs printed beside each result. Where our own documents quote a range rather than a single value — the pumpable viscosity ceiling is the clearest example — the range is stated as a range. Performance on your material will depend on its hardness, solids loading and target fineness, so we recommend a trial run on your own sample before you commit to a specification.
Where this guide refers to a measurement standard, these are the documents concerned. Machine specifications come from POLYC product data; the calculated figures are marked as calculations and their inputs are shown.
Tell us the apparent viscosity and the shear rate you measured it at, the target fineness with the gauge you use, and the batch size or hourly output. Our engineers will tell you which of the two routes your product allows — including when the answer is neither — and then size the chamber, the beads or the roll face to match. Quoted directly from the Wuxi factory, with a 3-year global warranty. Reply within 24 hours.
Request a configured quoteChat on WhatsApp
Nip shear explained, with the PTM model table and the small-batch configurations.
Pin, static disc, dynamic disc and vertical nano frames on one parameter set.
What to compare between roll mill suppliers once price is set aside.
The application view: which pigment class belongs on which machine route.