Particle Size Distribution Interpretation: Reading D10, D50, D90, Fines, and Oversize in Process Context

Particle size distribution data looks precise, the graph is clean, the percentile values are exact, and the report usually gives a clear D10, D50, and D90. That clarity can be helpful, but it can also hide the underlying question. A powder can meet its D50 target and still behave differently in a hopper, feeder, press, mixer, filling line, dispersion tank, or screen deck. The reason often sits outside the median value. That’s where particle size distribution interpretation comes in.

A small increase in fines can change dust release, air retention, cohesion, adhesion, electrostatic behavior, or wetting. A heavier coarse tail can affect dissolution time, sieve loading, visible defects, granule uniformity, or blockage risk. A wider curve can alter packing, segregation tendency, and batch consistency.

This Know How article explains how to read particle size distribution data as process information. The goal is not to treat PSD as a standalone answer. The goal is to connect the curve to the behavior that matters in production, development, or quality control.

A PSD curve only becomes meaningful when it is connected to a process condition. The same powder distribution can work well in one operation and create problems in another.

A broad distribution, for example, may improve packing in a die or container because smaller particles occupy the voids between larger ones. The same distribution may create separation during transfer when coarse particles roll outward and fines move through the bed. In a filling line, the fine fraction may dominate the dust signal. In a dispersion step, the coarse tail may control the final undissolved residue.

That is why the first question is operational. What changed in the process?

In a hopper or feeder, the relevant signs are discharge rate, surging, rat holing, flooding, refill behavior, and weight variation. In a blending or transfer line, the relevant sign is composition drift at the point of use. In wet processing, the key signal may be wetting delay, dry cores, slow dissolution, or sediment. In screening, the question becomes screen load, retained oversize, blinding, or misplaced material. In product quality, the signal may appear as specks, surface roughness, texture change, compaction behavior, or appearance defects.

Once the process question is clear, the curve becomes easier to read. D50 may show that the batch shifted overall. D10, D90, fines percentage, oversize percentage, and span usually explain where the shift sits and why it matters.

D10, D50, and D90 describe points on a cumulative particle size distribution.

D10 means that 10 percent of the reported distribution lies below that particle size. D50 is the median value. D90 means that 90 percent of the reported distribution lies below that size, with the coarsest 10 percent above it.

The phrase “reported distribution” matters. Laser diffraction commonly reports volume-based distributions. Sieve analysis usually reports mass-based fractions. Image analysis can report number-based or volume-based values, depending on the instrument, software, and method settings.

Those reporting bases are not interchangeable. A volume-based distribution gives large particles more influence because they contribute more volume. A number-based distribution gives fine particles more visibility because each particle is counted. A mass-based sieve result describes how much material sits above or below an aperture by weight.

So D10, D50, and D90 are not universal properties of the powder in isolation. They are values produced by a method, a preparation route, and a reporting basis. When those conditions stay consistent, they become powerful comparison points.

D10 describes the fine side of the curve. It is relevant when the problem involves dust, cohesion, air retention, poor deaeration, adhesion, wetting, dissolution rate, or surface-driven behavior.

D50 gives the midpoint of the reported distribution. It is useful for tracking general batch movement, milling performance, classification, supplier consistency, or process drift.

D90 describes the coarse side of the curve. It becomes important when the problem involves oversize, retained agglomerates, slow dissolution, visible particles, screen loading, nozzle blockage, or surface defects.

The fines fraction gives the amount of material below a defined cutoff. That cutoff should come from the process question. A dust release investigation may use one limit. A wetting problem may use another. A flow problem may need a different definition again.

The oversize fraction works the same way. A screen aperture, nozzle dimension, coating requirement, product texture limit, or dissolution target may define the relevant cutoff.

Span describes the width of the distribution. It is commonly calculated as:

Span = (D90 minus D10) divided by D50

Span helps track whether the distribution is becoming broader or narrower. Its meaning depends on the process. A broader distribution may improve packing in one system and increase segregation risk in another.

Curve shape gives the context that single values miss. Shoulders, long tails, or separate peaks often reveal mixed populations, weak agglomerates, attrition, incomplete classification, or measurement effects.

D10 is a useful starting point when the process issue sits in the fine fraction. A lower D10 often points to a stronger fine population, but the full curve still matters. A lower D10 with a narrow distribution is different from a lower D10 with a long fine tail.

Fine particles carry a large surface area relative to their mass. That changes how the powder interacts with air, moisture, liquid, contact surfaces, and neighboring particles.

In dry handling, fines often increase cohesion and reduce permeability. They can hold air after filling, release dust during transfer, or build up on walls and contact surfaces. In blends, fines can move through voids between larger particles when the bed is vibrated, dropped, or discharged. In wet processing, they may dissolve quickly, form lumps, or demand more liquid depending on surface chemistry and dispersion conditions.

D10 becomes more useful when it is paired with a defined fines fraction. A statement such as “more material below 10 micrometers” or “more material below 45 micrometers” usually gives a clearer process signal than D10 alone.

D50 is the most familiar PSD value because it gives a compact summary of the distribution. It helps compare supplier batches, milling settings, classification performance, agglomeration behavior, or process drift.

Its limitation is also straightforward. Two powders can share the same D50 and behave very differently.

One batch may have a tight distribution around the median. Another may have the same D50 with more fines and more coarse particles. In production, those two powders are not equivalent. The second batch may release more dust, pack differently, segregate more readily, dissolve unevenly, or load a screen differently.

D50 works best when the shape of the curve remains stable. Once the tails, shoulders, or span change, the median becomes only a reference point.

D90 describes the upper side of the reported distribution. A higher D90 usually points to a stronger coarse fraction, larger granules, retained agglomerates, incomplete milling, insufficient classification, or contamination by a coarser stream.

The coarse tail becomes important when large particles create a specific process effect. In screening, they increase retained material or screen load. In wet processing, they may dissolve or hydrate more slowly. In filled or coated products, they can appear as specks, roughness, or surface defects. In spray, dosing, or nozzle based operations, they may increase blockage risk. In compaction, they can change packing and stress distribution.

For many decisions, the percentage above a defined limit is more useful than D90 itself. That limit may be a mesh size, product specification, visual threshold, dissolution requirement, or equipment constraint.

Fines and oversize sit at opposite ends of the curve, but both deserve close attention because they often create disproportionate process effects.

The main body of the powder may look stable while one tail shifts enough to change production behavior. A modest increase in fines can alter dust release, flow stability, liquid demand, air retention, or adhesion. A modest increase in oversize can raise screen rejection, extend dissolution time, create visible defects, or signal incomplete breakage of agglomerates.

The cutoff matters. “Fines” is not a universal size class. In one process, fines may mean material below 10 micrometers. In another, it may mean material passing a 45 micrometer sieve. For dust, exposure, cohesion, or wetting, the useful limit depends on the mechanism being investigated.

The same applies to oversize. A particle becomes oversize when it matters to the operation. That may be the screen aperture, nozzle clearance, layer thickness, final texture requirement, or product specification.

Span gives a quick view of distribution width. It helps show whether a material is becoming broader or narrower over time.

A wider span means the powder contains a larger spread between fine and coarse particles. That can affect packing because smaller particles can fill spaces between larger ones. In some systems, this improves bed density or reduces voidage. In other systems, the size difference increases separation during transfer, vibration, filling, or discharge.

A narrow span creates a more uniform particle population. That can improve consistency in some products and reduce certain segregation mechanisms. It can also change packing density, permeability, flow regime, or dissolution profile.

Span therefore needs process context. A lower span is not automatically better. A higher span is not automatically worse. The useful question is whether the distribution width supports or disturbs the operation under review.

D10, D50, D90, and span reduce the curve to a few numbers. That is convenient, but it can hide useful evidence.

A single smooth peak usually points to one dominant particle population. A shoulder can suggest a secondary fraction, partial agglomeration, recycle material, incomplete milling, or a change in classification. A bimodal curve often deserves attention because it may represent two real populations rather than one broad material.

The tails also matter. A longer fine tail can reflect attrition, overmilling, breakage during conveying, fragile agglomerates, or fines generated during drying and handling. A longer coarse tail can reflect retained agglomerates, incomplete dispersion during measurement, insufficient classification, or contamination by larger material.

The curve does not explain itself. Microscopy, sieve checks, image analysis, dispersion trials, bulk density, flow data, or process history can show whether the feature reflects real particles, weak agglomerates, sampling variation, or method conditions.

This is especially relevant when supplier data and internal measurements disagree. A certificate may show acceptable D10, D50, and D90 values while the curve shape reveals a shoulder or tail that matters in the plant.

Particle size influences flow, but it does not define flow on its own. Flow behavior also depends on particle shape, surface texture, moisture, electrostatic charge, consolidation, bulk density, wall friction, permeability, and process stress.

Fines often increase the influence of surface forces compared with particle weight. That can reduce flow stability, increase hang-up, change feeder behavior, or make the powder more sensitive to humidity. However, a fine powder with dense, rounded, dry particles may still flow well. A coarser powder with angular, rough, moist, or fibrous particles may flow poorly.

For flow questions, PSD gives an important part of the material picture. It becomes more useful when read with bulk density, shear data, wall friction, permeability, moisture sensitivity, and observations from the actual equipment.

A higher fine fraction often raises dust potential, but dust release also depends on how easily those fine particles detach from the bulk.

Some powders contain many fines but release little airborne material because the fines remain bound to larger particles, form agglomerates, or sit in a cohesive structure. Other powders release dust readily because the fine fraction separates during filling, transfer, or impact. Moisture, oil films, electrostatic charge, particle shape, and handling energy all influence the result.

For a dust investigation, the useful PSD value is usually a defined fine fraction rather than the median. The PSD result shows the available fine material. A dustiness test shows how that material releases under a defined condition. Both pieces of information are useful because they answer different questions.

Particle size differences contribute to segregation when the process gives particles a chance to separate. Larger particles may roll outward or travel further after a drop. Fines may move through voids in the bed. Airborne fine material may migrate during filling or discharge.

A wider PSD can raise segregation risk when the coarse and fine fractions respond differently to movement. That does not mean the powder will always segregate. Cohesion, moisture, particle shape, density differences, agglomeration, equipment geometry, and handling energy all affect the outcome.

PSD therefore helps identify a possible segregation driver. Sampling across process locations shows whether the composition actually changes between mixer discharge, intermediate storage, transfer, feeding, and the point of use.

Particle size affects how powders interact with liquids. Smaller particles provide more surface area, which often supports faster dissolution or hydration. That surface area only helps when the liquid can reach it.

Wetting behavior depends on surface chemistry, entrained air, agglomerate strength, mixing energy, liquid properties, and powder addition method. A fine powder may float, clump, or form dry cores. A coarser powder may wet cleanly but dissolve more slowly. A broad distribution may create fast dissolving fines alongside coarse particles that remain visible or settle.

PSD interpretation in wet processing should therefore stay close to the observed behavior. Floating, lumping, slow hydration, sediment, and undissolved residue each point to a different part of the powder and liquid interaction. For deeper context, see the guide on <a href=”https://powdertechnology.info/the-complete-guide-to-powder-dispersion-and-data-interpretation/”>powder dispersion and data interpretation</a>.

Agglomerated powders need careful interpretation because the reported PSD may describe granules, primary particles, weak clusters, or particles broken during measurement.

In production, agglomerates may improve handling, reduce dust, support instantization, or stabilize feeding. They may also break during conveying, milling, compression, packaging, or sampling. After handling, a shift toward fines often points to granule attrition. A growing coarse shoulder can point to excessive agglomerate growth, incomplete classification, or retained lumps.

The dispersion setting used in the measurement is central here. Gentle dispersion may show the granule size present in the process. Strong dispersion may break the structure and reveal the primary particles. Both results can be valid, but they answer different questions.

Microscopy, sieve analysis, image analysis, and process trials help connect the reported curve to the structure seen in production.

Particle size is method-dependent. The instrument does not simply reveal one fixed size. It reports size through a measurement principle.

Laser diffraction estimates particle size from light scattering. It is widely used for full PSD curves because it is fast, repeatable, and suitable for many powders, suspensions, sprays, emulsions, and related systems. ISO 13320 gives guidance for laser diffraction measurements and instrument qualification. Results depend on optical settings, dispersion state, sample preparation, and the model used to calculate the distribution.

Sieve analysis separates particles by passage through defined apertures. It aligns well with screening, retained oversize, granule control, and coarse particle applications. ASTM E11 defines requirements for woven wire test sieve cloth and test sieves. The method reports aperture-based fractions by mass, which can make it very practical for plant questions involving screens and oversize.

Image analysis measures particle dimensions from images. It can reveal shape information that equivalent sphere methods simplify. Elongated, flaky, fibrous, or irregular particles may behave differently from rounded particles with a similar equivalent size. ISO 13322 covers image analysis methods, including dynamic image analysis.

Different methods can give different values for the same powder. That difference is not automatically an error. It usually reflects a different definition of size.

PSD interpretation also depends on the reporting basis.

A volume based distribution gives more weight to larger particles because they contribute more volume. A number based distribution gives stronger visibility to fines because each particle counts individually. A mass based distribution, common in sieve analysis, reports how much material sits in each size fraction by weight.

These bases answer different questions. A volume based D50 from laser diffraction should not be compared casually with a number based image analysis value or a mass based sieve result. The powder may be the same, but the data describes it through different logic.

For supplier comparison, method transfer, incoming QC, and troubleshooting, the method and reporting basis should remain visible next to the numbers. Otherwise, a change in measurement logic can look like a material change.

Sample preparation can shift the reported PSD as much as the material itself.

A weakly agglomerated powder may look coarse under gentle dispersion and much finer under stronger dispersion. A fragile granule may break inside the measurement method. A poorly wetted powder may clump during wet dispersion and produce an artificially coarse result.

The right preparation depends on the question. A raw material identity check may require a stable routine method. A troubleshooting study may compare gentle and strong dispersion to understand whether the process sees agglomerates or primary particles. A dissolution study may need wet dispersion behavior that reflects the real liquid system.

The measured PSD should represent the material state relevant to the decision. Maximum dispersion is not always the most informative condition.

Batch comparison is one of the most useful applications of PSD data. It can show supplier drift, milling variation, classification changes, drying effects, agglomeration shifts, attrition, or contamination by a second fraction.

The comparison only works when the method is stable. Sample location, splitting method, moisture state, handling history, dispersion setting, obscuration, sonication, sieve time, reporting basis, and replicate strategy all influence the result.

A strong batch comparison includes the full curve, D10, D50, D90, span, fines percentage, oversize percentage, and a short note on sample history. Replicate measurements are useful when the material is variable or the decision carries production risk.

The interpretation becomes stronger when the PSD shift aligns with a process observation. A new fine tail together with increased dust during filling gives a clearer signal than a fine tail alone. A higher oversize fraction together with more screen rejection points to a different investigation than the same oversize shift without any plant symptom.

Start with the operation. Storage, feeding, conveying, blending, compaction, screening, coating, dispersion, and dissolution all place different demands on the powder.

Then check the method. Laser diffraction, sieving, and image analysis define size differently. The reporting basis, dispersion route, and sample preparation determine what the values represent.

Read the tails before relying on the median. Fines and oversize often explain the practical problem more clearly than D50.

Study the full curve. Shoulders, tails, bimodal patterns, and changes in peak shape can reveal process history, agglomeration, attrition, mixed populations, or classification issues.

Connect PSD to another observation. Flow testing, bulk density, microscopy, moisture data, dustiness, sieve checks, dispersion behavior, and product performance all help confirm the interpretation.

Finally, connect the result to a decision. The decision may involve a supplier limit, incoming QC check, milling setting, screen aperture, classifier adjustment, dust control measure, feeder setup, dispersion protocol, or rework route.

The companion PDF for this Know How article is the PSD Interpretation Field Sheet. It is designed as a working document for reviewing D10, D50, D90, span, fines, oversize, and curve shape in process context.

The field sheet supports incoming material checks, supplier reviews, troubleshooting, product development, and process qualification. It helps the reader move from a reported PSD curve to a practical interpretation: which part of the distribution changed, what process behavior it affects, and which supporting check can confirm the reading.

For numerical review, the PowderTechnology.info Calculator Suite includes a Particle Size Distribution Analyzer and Plotter. It can help calculate D10, D50, D90, and span from particle size data. The interpretation still depends on the material, method, and process condition.

PSD data can answer direct size questions. It can show whether a batch became finer, whether the coarse fraction increased, whether a milling step shifted the curve, or whether a material moved outside an agreed size window.

Other powder behaviors require more information. Flow, caking, adhesion, dust release, electrostatic charging, compaction, and dispersion also depend on shape, density, roughness, porosity, moisture, surface chemistry, storage stress, and process geometry.

PSD works best as a starting point for interpretation. It describes the particle population. The process behavior shows which part of that population matters.

Particle size distribution interpretation becomes valuable when the curve is read as process evidence.

D50 gives a useful midpoint. D10, D90, fines, oversize, span, and curve shape show where the distribution has changed and where the process risk may sit. Fines often affect dust release, cohesion, air retention, adhesion, wetting, and feeding. Oversize often affects screening, dissolution, visible defects, and blockage. Distribution width often affects packing, segregation potential, and batch consistency.

The strongest PSD interpretation links the curve to method conditions, sample history, process observations, and supporting material data. That connection turns particle size data into a practical tool for quality control, troubleshooting, product development, and process design.