You order a bag of filled PTFE suspension resin, crack it open, and there are lumps in it. The instinct is to grab a sieve or a lump breaker and fix it. I get it. That feels like the obvious move. The problem is it is also one of the quicker ways to ruin a batch.
Taking agglomerates out of a filled PTFE blend is not one job. It is a mix of diagnosis and process control that touches how you store the material, how you screen it, how gently you de-lump it, and whether you bring air into it at all. Get the order wrong and you can fiberize the resin, split the filler away from the polymer, or grind good granules down into dust that flows badly and packs worse.
So this is a walk through what actually causes those clumps, how to spot the ones that will hurt you, and a process route you can bend to fit your own formulations.
Why screening alone does not cut it
PTFE suspension resin usually gets blended with somewhere between 5% and 40% of an inorganic or organic filler. The point is better wear resistance, conductivity, creep resistance, or stiffness. Before that powder hits a compression mold or a ram extruder, it has to stay uniform in composition, flow steadily, stay free of big lumps, and not be over-ground into fines.
The word “agglomerate” hides at least three different problems:
Soft agglomerates. You can break these and often reuse them, as long as you check what they are made of.
Hard agglomerates. You should not just smash these back in.
Segregation-type agglomerates. Here the material left on the screen is not even the same recipe as the bulk. This is the one that bites people.
Screening, crushing, and air dispersion each fix a different layer. Stack them in the wrong order and they make things worse. Too much shear and the PTFE goes fibrous. Too much air and the filler drifts off the resin. Too fine a screen and you throw away good particles while the fines pile up.
The order that tends to hold up looks like this: balance the temperature and humidity first, run a gentle screen to see what is actually in there, break only the soft oversize, re-screen in a loop, bring in low-energy air only if you need it, and send any hard lump, contamination, or filler-heavy chunk back upstream instead of forcing it through.
Where the clumps come from
Soft agglomerates: breakable, but check first
Most soft lumps are boring in origin. Packaging gets compressed and transport vibrates particles into a mechanical lock, but you can crush them with your fingers and the color matches the bulk. Static does the same thing, and PTFE is a stubborn insulator, so powder picks up charge during conveying and blending and fine dust clings to screens and hopper walls. A little moisture or a temperature swing lets hygroscopic fillers form liquid bridges, so small blocks show up after storage and clear up after you dry or condition the lot. And plain old packing after blending, where filler settles into the gaps between PTFE particles, raises the contact points and gives you loose clumps.
You can screen, lightly de-lump, or air-fluff these. But before any oversize goes back in, confirm its filler content matches the bulk. Skip that and you are quietly feeding a local segregation straight into the product.
Hard agglomerates: do not just smash them
Hard lumps show up as a few types. Filler-rich blocks come from poor pre-dispersion, dead corners in the mixer, or the wrong addition order, and they leave hard spots, weird wear, and black or white specks in the part. Wet clumps happen when glass fiber, mineral, carbon, or bronze picks up moisture or oil, and after sintering you get pores, bubbles, and weak spots. Over-sheared PTFE goes fibrous from high-speed mixing or rough pneumatic conveying, then bridges and flows poorly with uneven preform density. And then there is plain contamination: worn screens, equipment debris, packing material, oil. Those are defects on sight and get pulled.
Freshly milled PTFE has poor bulk density and flow, so it packs in hoppers and small barrels and bridges on its own. The tempting fix, more shear, can damage the molded part you are trying to make. A hard lump is not a break-it-and-forget-it situation.
Segregation-type agglomerates: the real hazard
The thing I would flag hardest in any filled PTFE line is oversize that is not the same stuff as the bulk.
Glass or carbon fiber clumps lose their length when you crush them, so the reinforcement drops off. Graphite, carbon black, and carbon fiber form dark rich blocks that turn into black specks and patchy conductivity. Bronze is dense, so a long vibrating screen or a puff of air separates the PTFE from the metal. Flake fillers like molybdenum disulfide or graphite reorient and concentrate under fast air and hard shear. And organic fillers like PI, PPS, or aramid can soften, cake, or coat the equipment when temperature and friction climb.
With these, crushing is not the answer. You identify, you isolate, and you walk the problem back to mixing or filler pre-treatment.
What screening is actually for
A screen does four things here. It keeps big transport and storage blocks out (safety screening). It tells you how bad the batch is by measuring the oversize ratio (identification). It gives soft agglomerates a little shear from the vibration or rotation (light de-agglomeration). And it keeps chunks out of the hopper, cavity, ram extruder, or press (downstream protection).
For compression molding powder, pulling blocky material and loading evenly is a control step, not a nice-to-have.
Picking the mesh
Do not choose a screen size in a vacuum. Filled PTFE granule powder can run from tens of microns out past several hundred, sometimes coarser. Too fine a screen rejects good particles, loses formulation, and piles up fines.
A staged approach works better:
A coarse safety screen around 3 to 5 mm knocks out obvious transport blocks, packaging trash, and hard lumps.
A 1 to 2 mm screen before loading is the common control point for molding powder.
Parts that demand surface quality, thin sheet, skived film billet, seals, high-dielectric pieces, can justify 0.6 to 1.0 mm.
A 40 to 80 mesh or finer screen only belongs in systems that are genuinely fine powders, and you have to watch the usable-particle loss.
For suspension filled material, the opening should come out of D90 or D97, final part thickness, surface-defect needs, and filler type, all at once.
Screening equipment
Round vibrating screens are fine for routine safety screening of small batches. They are cheap, but long vibration stratifies and segregates heavy fillers. Linear screens eat large volume and coarse work, with dust and segregation to manage. Rotary or centrifugal screens screen continuously and de-agglomerate gently, though high speed adds shear and heat. Ultrasonic screens help fine, blinding-prone, high-static powders, but filled granule powder does not always need them. Screens with de-lumping blades lift yield on soft clumps yet can fiberize or crush filler.
The rule I keep coming back to: prefer low-energy screening, and never run the whole batch through aggressive crushing.
Crushing and de-lumping: oversize only, and only when soft
De-lumping should not touch the whole batch. It should act on what the screen rejected. Screen everything, sample the oversize, decide if it is soft or hard and what it contains, send the qualifying soft stuff through a low-speed de-lumper, re-screen, and feed the qualifying undersize back in a limited amount. That keeps the total shear low and leaves PTFE particle shape alone.
Equipment choices
Low-speed lump breakers are the first pick for soft agglomerates and compacted chunks: low speed, high torque, short residence. Oscillating granulators are gentler on soft blocks if you keep the plate holes large. Cone granulators give good control at low speed and large mesh without making fines. Pin, hammer, and turbo mills are poor choices for routine work. They shear hard, heat up, crush filler, and throw off fines. Cryogenic crushing fits hard PTFE blocks that need size reconstruction, but it costs more and brings condensation and safety issues. Jet mills are for micronization, not for ordinary molding-powder de-agglomeration.
The moment crushing shifts from “soft de-agglomeration” to “rebuild the particle size,” it stops being a post-treatment and becomes a powder-making process that needs its own validation.
What to watch while de-lumping
The goal is not as fine as possible. Keep an eye on these:
Oversize passes after one de-lumping pass, but the fines ratio should not jump.
Filler content after de-lumping matches the bulk.
Bulk density, flow, angle of repose, or hopper behavior do not get worse.
Temperature stays low so you avoid frictional heat.
Glass and carbon fiber length hold up.
Graphite, bronze, and molybdenum disulfide do not enrich or break free as dust.
Set a maximum number of de-lumping passes. A hard block still sitting on the screen after one pass gets isolated and studied, not recycled. Crush it again and again and you turn a rejected lump into tiny defects that are far harder to find.
Air dispersion: know which one you mean
The word “air” covers three different operations, and people mix them up.
Low-energy air loosening fluffs soft agglomerates and knocks off adsorption with dry air. That fits fine-powder sticking and mild false agglomeration.
Air classification uses air and centrifugal force to split fine from coarse. Good for controlling fines or oversize, but it can pull filler off the resin.
Air jet milling smashes particles by high-speed collision. That is micronization or size rebuild, and it makes fines, crushed filler, and a changed gradation.
Jet milling can reject oversize and give a narrow distribution with a sharp top cut, which suits some heat-sensitive or high-purity materials. For filled PTFE molding powder it cuts both ways: it removes coarse agglomerates but also wrecks particle shapes you wanted to keep.
When air dispersion fits
Reach for it when screen blinding comes from static fines, when the agglomerates are loosely fluffy and air breaks them, when you need to pull excess free fines to help flow, when cleanliness matters and you want less mechanical contact, or when you need closed-loop classification to control D97 or oversize residue.
Leave it alone on dense filler systems like bronze-filled PTFE that separate easily, on fiber fillers like glass or carbon that lose length or distribution, on hard, wet, or oily agglomerates that air only dusts on the surface, and when your goal is granular molding-powder flow while jet milling would flood you with fines.
Air system controls
Use dry, oil-free, clean compressed air or nitrogen. Keep velocity low, paths short, bends few, to cut charge buildup and particle impact. Ground every metal part, soft connection, screen frame, bin, and dust collector. Re-test filler content after air classification, especially for high- or low-density fillers. Do not return bag filter, cyclone, or recovery dust to the batch unless its composition checks out. And never blow open dust off the floor or equipment with compressed air. That just scatters it and exposes people.
A route that holds up
Routine filled PTFE molding powder
For glass fiber, graphite, carbon, bronze, and molybdenum disulfide blends:
- Condition the raw material in closed containers at balanced temperature and humidity.
- Run a coarse safety screen to drop obvious blocks and foreign matter.
- Sample the oversize: look, finger-crush, ash or filler check, microscopy.
- Send soft agglomerates through a low-speed lump breaker.
- Re-screen on the same mesh.
- Feed the qualifying undersize to the finished-goods buffer.
- Use low-energy air loosening or dust removal if needed.
- Re-screen, metal-detect, and keep a sample before packing.
Screening leads. De-lumping is a closed loop on the oversize. Air is the helper for dispersion or dust, nothing more.
High-surface-quality or high-dielectric parts
Thin sheet, skived film billet, seals, high-voltage insulation:
Use two-stage screening, coarse to drop blocks and fine to cap the top size. Add micro-defect counts and sintered small-sample checks. Cap the oversize reuse ratio hard. Do not force hard blocks back in. Use air classification only after you prove it leaves filler content and dielectric behavior alone.
Powder state and part performance move together. Flow, bulk density, surface roughness, breakdown voltage, tensile strength, and elongation all track each other, so powder condition is not just about feeding convenience.
High-fill or large-density-difference systems
Bronze-filled, high-carbon, and ceramic-filled blends need short vibrating-screen residence so heavy components do not sink, low air velocity so light and heavy do not split, filler-content testing on all oversize, small-ratio same-batch reuse, and a straight call of “mixing failure” instead of “de-lumping problem” when oversize filler content is clearly off.
Fiber-filled systems
Glass, carbon, and aramid fibers need no high-speed pin, hammer, or turbo mills, a preference for low-speed de-lumping and large-hole screens, fiber-length distribution testing, attention to anisotropy, wear, tensile, and compressive creep in the part, and no strong-shear blades inside the screening gear.
How you know it worked
Powder metrics
Track screen oversize residue for agglomeration level, oversize filler content for segregation, D10 through D97 for size change after de-lumping or air work, fines ratio for over-crushing, bulk and tapped density for packing stability, angle of repose and flow time for bridging risk, moisture or volatiles for wet-agglomerate risk, electrostatic potential or decay for adhesion and blinding risk, and microscopy or SEM for fiberization, filler clumps, and foreign matter.
Sorting the oversize
Four buckets. Soft agglomerate with matching composition gets low-speed de-lumped, re-screened, and limited reuse. Hard agglomerate with matching composition gets evaluated on its own, not defaulted back in. Filler enriched or depleted gets isolated, no reuse, and a trace back to mixing. Foreign matter, oil, discoloration, metal debris gets scrapped or written up as non-conforming.
Verifying the part
Powder work has to be confirmed with a molded sample, not just a particle-size pass. Check preform density distribution, sintered density with pores and bubbles, surface specks and hard spots and flow marks, tensile and elongation and compression set, wear and friction, dielectric strength and volume resistivity, high-temperature dimensional stability, and machined surface roughness.
Building your own parameters
For a small DOE, look at screen opening in three steps (coarse, target, tight), screening method (vibrating, rotary, ultrasonic), screening time or feed rate (low, medium, high load), de-lump speed (mostly low, then up), de-lump hole size (large first), de-lump count (0, 1, 2), air pressure or flow (low, medium, high), powder temperature (stable room range, no rise), relative humidity (dry, medium, high), and reuse ratio (0%, small, medium).
Watch the responses: oversize residue down, fines ratio not clearly up, filler-content RSD not up, bulk density stable, flow improved or steady, forming defects down, mechanical properties not below baseline, wear and dielectric not below baseline, dust controlled, equipment temperature controlled.
At scale-up, check whether continuous screening stratifies, whether screens clog or wear or break wire, whether the de-lumper heats, whether filler lands in the dust after air work, whether packing re-agglomerates at 7, 14, and 30 days, and whether transport vibration pushes oversize residue back up.
Safety and cleanliness
PTFE is thermally stable, but high-temperature processing can give off fumes you do not want to breathe, so thermal work needs ventilation. Follow the SDS for the powder and compounds, and use dust masks, goggles, and gloves where they are called for. Metal fillers like bronze and polymers like polyimide can lower the thermal stability of some fluoropolymer systems, which matters for high-fill PTFE and for abnormal temperature control.
Set up closed feeding with local dust removal, ground all equipment and bins and soft connections and collectors, keep metal debris out of product, forbid blowing open dust with shop air, interlock de-lumping gear on temperature, torque, and current, collect dust separately without default reuse, clean thoroughly between formulations especially for carbon black, graphite, and bronze, and add magnetic separation or metal detection for high-cleanliness parts.
PTFE does not form a flammable dust cloud under normal plant conditions and sits in dust-explosion class St 0. That still does not replace control of filler dust, static, inhalation, and local contamination.
Mistakes I keep seeing
Returning all oversize after crushing. Wrong. Oversize can be filler-rich, wet, foreign, or fibered PTFE. Analyze it first.
Finer screen is better. Wrong. Over-fine screens waste good particles, raise fines, lower flow, and can segregate the recipe.
Jet mill instead of ordinary de-lumping. Wrong. A jet mill micronizes and classifies. It is not a loosening device and has no place in routine filled molding powder.
High-speed crushing fully solves agglomeration. Wrong. High shear fibers the PTFE, adds heat, makes fines, and can damage fiber filler.
Judging by particle size, not molded samples. Wrong. Final judgment of PTFE molding powder has to reach preforming, sintering, and part performance. A clean screen pass does not mean a clean part.
Where to put your effort
Control the source first. Drier filler, better pre-dispersion, correct addition order, controlled mixing energy and temperature, less storage compaction, gentler conveying. Then screen to identify and remove. Then low-energy closed-loop de-lumping on the soft oversize only, never on the whole batch. Use air as a helper for loosening or classification, not as a stand-in for de-lumping or for re-making powder. And isolate the hard lumps, filler-rich blocks, foreign blocks, and fibered blocks, then trace them to mixing, filler pre-treatment, storage, and conveying.
Do that and “get the lumps out” stops being a guess and becomes a step you can actually trust.