Particle size distribution

A distribution specifying relative particle amount by size for a powder, granular solid, or dispersion.

Version

v1 · 2026-09-28 · History

Domain-specific #

7718

Domain group

Natural Sciences

Origin domain

Chemistry & Materials Science

Subdomain

Granulometry → Chemistry & Materials Science

Core Idea

A particle-size distribution (PSD) describes how the particles in a powder, granular material, aerosol, suspension, or other particulate sample are apportioned across size.^[1] It may be given as fractions in discrete size intervals or as a cumulative or continuous function.^[2] Each value states a relative amount—commonly by mass, number, volume, or another declared weighting basis—rather than merely listing the smallest and largest particles.^[3]

The distribution is inseparable from its operational definition of size.^[4] Irregular particles do not have one unique diameter: sieving classifies by passage through an aperture, sedimentation by settling behavior, imaging by measured geometry, and scattering methods by a model-dependent equivalent diameter.^[5] The measurement method, preparation and dispersion state, weighting basis, and bin or cumulative convention can therefore change the reported PSD for the same material.^[6]

The invariant is: a representative particulate sample is assigned sizes by a stated measurement model, and nearly all of its declared amount is accounted for across ordered size classes or a normalized size function.^[7] Change the instrument, bin widths, or mathematical fit and the identity can remain if the basis and conservation of amount are explicit. A single mean diameter, an unweighted list of observed particles, or a histogram whose fractions do not refer to a defined sample and weighting basis is not a complete PSD.^[8]

Range and cumulative forms answer different questions but can encode the same distribution. A range analysis gives the amount between successive limits; a cumulative analysis gives the amount smaller than or larger than each limit.^[9] Representative sampling and prevention of segregation or loss of fines are constitutive to interpreting either form, because a precise instrument cannot recover the distribution of material absent from the sample.^[10]

Structural Signature

Sig role-phrases:

• particulate population — the powder, granular solid, aerosol, suspension, or other material whose particles are to be apportioned by size
• representative sample — a sample taken and reduced without size segregation or selective loss relative to the declared population
• preparation state — the dispersion, disaggregation, or handling condition under which particles enter the measurement
• sizing method — sieving, settling, imaging, scattering, counting, or another procedure that assigns particle size
• operational diameter — the aperture, settling-equivalent, image-derived, scattering-equivalent, or other method-defined size variable
• weighting basis — mass, number, volume, or another declared amount used to calculate relative fractions
• ordered size support — the bins, thresholds, or continuous size variable over which the amount is distributed
• distribution values — normalized interval fractions or a mathematical function accounting for the declared particle amount
• range–cumulative branch — interval amounts and passing or retained totals are alternative presentations linked by summation and differencing
• normalization guarantee — the reported fractions account for nearly all of the declared sample apart from explicit losses or exclusions
• method boundary — results from different size definitions, preparations, or weighting bases are not interchangeable merely because they share units
• information boundary — the distribution does not recover particle shape, density, composition, or agglomeration history discarded by the measurement model

What It Is Not

• Not a single mean, median, percentile, or size range. Those statistics summarize parts of a PSD but do not account for the declared particle amount across its ordered size support.
• Not an unweighted list of measured diameters. A distribution requires a defined particulate sample, weighting basis, size classes or function, and normalization rather than a collection of observations alone.
• Not a generic probability distribution. A PSD apportions the amount of a material population by operational particle size, with sampling, preparation, and mass-, number-, or volume-weighting obligations.
• Not one intrinsic diameter for every irregular particle. Sieve aperture, settling behavior, image geometry, and scattering models can assign different equivalent sizes to the same particle.
• Not interchangeable across weighting bases. A number-weighted curve can emphasize fines while a mass- or volume-weighted curve emphasizes larger particles; equal size units do not make the fractions comparable.
• Not interchangeable across sizing methods or preparation states. Sieving, imaging, settling, and scattering apply different size models, and dispersion can shift agglomerates into constituent-sized classes.
• Not a cumulative percentage confused with an interval fraction. Amount below or above a threshold and amount between two limits are related by summation or differencing but answer different questions.
• Not proof that the analyzed sample represents the source population. Segregation, loss of fines, breakage, dissolution, or incomplete sampling can distort the population before the instrument measures it.
• Not a description of particle shape, density, composition, or history. The size distribution cannot recover attributes discarded by its operational diameter and weighting model.
• Not complete when its fractions do not close. Unexplained missing or excess amount means the normalization guarantee has failed, even if the plotted curve appears smooth.

Scope of Application

A particle-size distribution applies wherever a defined particulate population is sampled, assigned an operational size, weighted by a declared amount basis, and normalized across ordered size classes or a continuous size function. Its literal reach spans many particulate substrates, but only when sampling, preparation, sizing method, equivalent-diameter model, and weighting basis remain explicit. A lone mean or range, an unnormalized diameter list, or a curve with an unknown sampled population or amount basis falls outside the instrument.

• Powders and granular industrial products — toner, ceramics, metals, pigments, cosmetics, pharmaceuticals, and other formulated powders are specified and controlled by the relative amount in required size classes.
• Soils, rocks, and sediments — grain-size distributions support geotechnical, geological, and sedimentological descriptions of packing, load-bearing behavior, transport, and material history under stated disaggregation procedures.
• Aerosols and emission particles — atmospheric and industrial particulate samples are apportioned by aerodynamic, optical, or other operational diameter for monitoring and control.
• Suspensions, colloids, and fluid dispersions — particles dispersed in liquids or gases are sized by counting, settling, scattering, acoustic, or imaging methods while dissolution and agglomeration are controlled.
• Mining, crushing, milling, and comminution — range and cumulative PSDs describe product from size-reduction processes and support fitted forms such as Rosin–Rammler distributions.
• Filtration and particulate collection — settling chambers, centrifugal collectors, fabric filters, scrubbers, electrostatic precipitators, and filter cakes are selected or assessed against the sizes and fractions they must capture.
• Agricultural, forestry, and bulk-material handling — photoanalysis, sieving, and representative stream sampling characterize materials while minimizing segregation, breakage, contamination, and loss of fines.
• Granulometry laboratories — sieve analysis, microscopy, electroresistance counting, sedimentation, laser diffraction, obscuration, and ultrasound each produce a method-defined PSD rather than an interchangeable intrinsic diameter.
• Process and product quality control — continuous or repeated PSD measurement detects shifts in grinding, mixing, dispersion, agglomeration, and collection performance when the same operational definition is maintained.
• Distribution modeling and specification — interval fractions, cumulative passing or retained curves, percentiles, and parametric fits summarize the same accounted population only when conversion, units, and amount basis are conserved.

Clarity

A particle-size distribution is interpretable only when the sampled material, preparation and dispersion state, sizing method, operational diameter, weighting basis, size units, and reporting form are stated. “Twenty percent below 10 μm” changes meaning if the percentage is by number rather than mass or volume, or if 10 μm denotes sieve aperture, settling-equivalent diameter, image geometry, or scattering-model diameter. The fractions should also account for the declared sample apart from documented losses or exclusions.

Range and cumulative presentations must not be mixed: a range value is the amount between two limits, whereas a cumulative value is the amount passing or retained at a limit. A mean or median diameter summarizes a distribution but does not replace it. Comparisons across instruments require compatible size definitions, sampling, and weighting, because irregular particles and agglomerates can yield different distributions under different procedures. The useful practitioner question is: whose amount is distributed across which operational particle size, and do sampling, preparation, measurement, and normalization support that comparison?

Manages Complexity

A bulk powder, aerosol, suspension, or soil sample may contain enormous numbers of irregular particles across many size decades. A particle-size distribution makes that population tractable by assigning an operational diameter, choosing a mass, number, volume, or other weighting basis, and accounting for the sample in ordered bins or a normalized function. The retained record includes sampling and dispersion state, method, size units, weighting, bin edges, and normalization. It makes range versus cumulative reporting, passing versus retained fractions, and sieve-, settling-, image-, or scattering-equivalent size branches readable while supporting comparison of characteristic percentiles and controlled size fractions.

The distribution is only as representative as the sample and only as physical as its size model. It does not preserve each particle’s shape, orientation, density, composition, or agglomeration history, and a mass-weighted curve can differ sharply from a number-weighted one. Segregation, loss of fines, dissolution, breakage, or incomplete dispersion can change the measured population before analysis. Results from different methods therefore cannot be merged merely because their horizontal axes use micrometres. The compression boundary is crossed when sampling, operational diameter, weighting basis, or accounted fraction is omitted.

Abstract Reasoning

The distribution licenses quantitative moves from thresholds to material fractions. A cumulative curve gives the declared fraction finer or coarser than any selected operational size; differences between successive cumulative values give the fraction in a size interval. Conversely, summing normalized interval fractions reconstructs the cumulative form. Percentile sizes and comparisons between samples are therefore legitimate only after the weighting basis, size definition, units, preparation, and accounted total have been aligned.

Diagnostic and intervention reasoning focuses on how the observed population was produced. Loss of fines, segregation during sampling, or incomplete dispersion predicts a distorted curve; changing dispersion can move apparent mass or number from agglomerate-sized classes toward constituent-sized classes without changing the primary particles. A disagreement between sieve, settling, imaging, and scattering results first directs attention to their different equivalent diameters and weighting models rather than to an unexplained material transformation. The inference boundary is strict: a PSD can reveal the amount assigned to size classes, but it cannot by itself recover particle shape, density, composition, or agglomeration history that its measurement model did not retain.

Knowledge Transfer

Within granulometry and materials analysis, a PSD transfers literally across powders, soils, aerosols, suspensions, and granular products by preserving a representative sample, operational diameter, weighting basis, ordered size classes or function, and normalization. Sieving, settling, imaging, and scattering carry different equivalent-size models, so method, dispersion state, bin convention, and mass-, number-, or volume-weighting must accompany the curve. Re-dispersing agglomerates, checking sample splits, reconciling cumulative and interval forms, or comparing methods diagnoses whether a shift reflects the material population or the measurement procedure.

Beyond particulate materials, the honest reach is a mix of (C) instrument or measure, and (B) a shared abstract mechanism: other composition analyses can carry the practice of partitioning a conserved amount across an operationally defined ordered variable, with explicit weighting and cumulative conversion. What carries is the distributional accounting; physical particles, equivalent diameter, sample dispersion, sieves, settling, and scattering models remain home-bound. Calling any collection of “large and small” items a PSD is only (A) analogy. Transfer stops when only a mean or range is reported, the weighting basis is unknown, or the fractions do not account for a defined sampled population.

Examples

Canonical

Consider a conventional sieve analysis of a carefully divided 100 g powder sample. Suppose stacked sieves with openings of 150 μm, 75 μm, and 45 μm retain 10 g, 35 g, and 40 g, while 15 g reaches the pan. On a mass basis, the range distribution is therefore 10% above 150 μm, 35% from 75–150 μm, 40% from 45–75 μm, and 15% below 45 μm; the four fractions sum to 100%. The same result can be written cumulatively: 90% passes 150 μm, 55% passes 75 μm, and 15% passes 45 μm. These values describe aperture-defined size classes, not unique geometric diameters for irregular particles. If fines were lost while splitting the sample, or if the four recovered masses did not close to the analyzed amount, the smoothness of the plotted curve would not rescue the distribution's representativeness.

Mapped back: The powder is the particulate population, and careful sample division supplies the representative sample in its declared preparation state. The sieve stack is the sizing method; aperture width is the operational diameter; recovered mass is the weighting basis; the three apertures and pan define the ordered size support. The four percentages are the distribution values, their conversion between interval and passing forms instantiates the range–cumulative branch, and their 100% total supplies the normalization guarantee. The aperture definition and possible fines loss make the method boundary and information boundary explicit.

Applied / In Practice

In powder processing, wet sieving of milled limestone is a real bulk-material application of PSD measurement. A representative process sample is washed through a selected sieve series with a non-reacting liquid until the retained fractions stabilize, then each recovered fraction is measured by mass. The resulting range or cumulative curve lets the operator see whether milling is leaving too much coarse product or producing an excessive fines fraction. It remains a sieve-defined, mass-weighted PSD: it cannot be compared directly with a laser-diffraction volume-equivalent curve unless the size definition, preparation, and weighting basis are reconciled. Excessive sieving energy may abrade particles and shift material toward finer classes, while insufficient energy may leave loose agglomerates in coarser classes, so the procedure can change the population it purports to measure.

Mapped back: Milled limestone is the particulate population; the process sample and wet handling specify the representative sample and preparation state. Wet sieving is the sizing method, aperture width the operational diameter, recovered mass the weighting basis, and the sieve series the ordered size support. Retained or passing percentages furnish the distribution values, with the range–cumulative branch supporting either process view and mass closure providing the normalization guarantee. Attrition, incomplete deagglomeration, and incompatibility with laser-diffraction outputs mark the method boundary and information boundary.

Structural Tensions

T1: Instrument precision versus sample representativeness.

A sizing method can measure its aliquot with fine resolution while the aliquot poorly represents the source population because segregation or loss of fines occurred during sampling and reduction. More extensive sampling improves population coverage but introduces handling steps that can themselves sort, abrade, or contaminate the material. The PSD therefore joins two distinct evidentiary problems: obtaining the right particles and measuring the obtained particles consistently. Diagnostic: Does the reported uncertainty cover only the instrument's repeatability, or also the sampling and reduction process that determines which size fractions reached it?

T2: Operational diameter versus particle geometry.

Sieving, settling, imaging, and scattering assign size through different physical or model-based relations. For irregular particles, each may produce a legitimate equivalent diameter without any one being the unique intrinsic size. Treating method-defined values as interchangeable erases shape and orientation effects; refusing all comparison prevents the use of PSDs across established techniques. Comparability must be earned by declaring what each diameter operationally means. Diagnostic: Is the same size variable genuinely being compared, or do equal numerical units conceal different aperture, settling, image, or scattering definitions?

T3: Number weighting versus mass or volume weighting.

A population containing many fine particles and a few coarse ones can look fine-dominated by number but coarse-dominated by mass or volume. Converting or comparing curves without retaining the weighting basis can therefore reverse the apparent importance of size classes. Yet no basis is universally correct: the relevant amount depends on whether the application concerns counts, conserved material, occupied volume, or another declared quantity. Diagnostic: Which particle amount is being distributed, and would the conclusion change if the same physical sample were expressed on a different weighting basis?

T4: Disaggregation versus population alteration.

Preparation seeks to expose the particles intended for measurement by dispersing loose agglomerates, but added energy or liquid can abrade, break, or dissolve particles and thereby change the population. Insufficient preparation reports agglomerate-sized classes where constituent particles were intended; excessive preparation manufactures a finer distribution. The correct endpoint depends on the declared object of measurement rather than a universal demand for maximum dispersion. Diagnostic: Does the preparation reveal the target particle units while preserving them, or has it either left unintended aggregates intact or created sizes absent from the sampled material?

T5: Interval resolution versus cumulative readability.

Range distributions show how much material lies between successive limits, making local modes and specification bands visible. Cumulative curves make passing or retained thresholds easy to read but can conceal how material is divided within a broad rise. Finer bins add resolution while increasing sensitivity to noise and procedural differences; coarser bins stabilize summaries while hiding structure. The two forms can encode the same accounted amount only when their limits and arithmetic are conserved. Diagnostic: Does the chosen binning or cumulative presentation preserve the size feature relevant to the decision, and can one form be reconstructed from the other without lost or double-counted material?

T6: Normalized closure versus measurement exclusion.

Fractions summing to the declared total provide an internal accounting guarantee, but normalization can mask particles that were lost, censored, dissolved, or outside the instrument's range before percentages were rescaled. Refusing normalization leaves no coherent distribution; accepting it without an explicit denominator can give false completeness. The analyst must distinguish closure over the analyzed sample from representativeness of the source population. Diagnostic: What material defines one hundred percent, and are exclusions, unrecovered fractions, and out-of-range particles documented rather than silently normalized away?

T7: Size-distribution compression versus discarded particle attributes.

A PSD turns an enormous particulate population into an actionable allocation across one ordered size variable, enabling compact comparison and threshold decisions. That compression necessarily discards particle-level shape, orientation, density, composition, and agglomeration history except insofar as the operational diameter indirectly reflects them. Refusing compression would preserve detail at the cost of losing the distribution's tractable summary; treating the curve as a complete material description licenses conclusions its measurement model cannot support. The omitted attributes must be restored when the application depends on more than size allocation. Diagnostic: Can the decision be made from the amount assigned to operational size classes alone, or does it require particle attributes that the PSD has collapsed or excluded?

T8: Particle-size-distribution autonomy versus reduction to Measurement. The exact parent Prime Measurement strictly subsumes the result: every qualifying particle-size distribution ties a particulate target population and size attribute to an operational size scale through a declared sampling-and-preparation chain, sizing instrument, procedure, unit and weighting frame, calibration or traceability chain, and uncertainty account. The mapping also retains bidirectional instrument–target coupling: sampling, preparation, and dispersion can segregate material, lose fines, abrade or break particles, or dissolve a fraction, thereby changing the population whose sizes are assigned. The distribution remains in situ because it additionally requires representative sampling, method-defined diameter, normalized allocation across ordered size classes, and range–cumulative closure. Reduction gains portable target–instrument–procedure–traceability structure but erases granulometric allocation; complete autonomy hides both why the curve is an evidential result and how the measurement process can alter its target. Diagnostic: if operational diameter, weighting basis, normalized allocation, and range–cumulative form are removed while a target attribute remains assigned through a calibrated or traceable procedure with declared instrument–target coupling, does Measurement survive while Particle Size Distribution does not; and if that measurement chain or coupling is removed, does the PSD lose its evidential status even when a binned curve remains?

Structural–Framed Character

Particle-Size Distribution is mixed-structural. Its smallest portable skeleton is Measurement: a target population and attribute are coupled to an instrument and procedure, an operational scale and unit, an observation frame, and stated uncertainty. The PSD specializes that structure by assigning a representative particulate sample an operational diameter and weighting basis, then normalizing its amount across ordered size support with linked range and cumulative forms. That portable reach belongs to the Measurement Prime; sampling, dispersion, equivalent-diameter models, and mass-, number-, or volume-weighted granulometry remain the domain accent.

Its evaluative_weight is low because a distribution describes accounted particle amount rather than ranking a material, although normalization and representativeness govern validity. Its human_practice_bound character is moderate because particle populations exist independently, while every reported size and weighting depends on sampling, preparation, and an operational measurement procedure. Its institutional_origin is low to moderate because laboratory and industrial standards stabilize methods without constituting the particulate population. Its vocab_travels result is partial: target, scale, procedure, uncertainty, and normalization carry, whereas sieving, equivalent diameter, dispersion, and retained or passing fractions remain specialized. Under import_vs_recognize, Measurement is recognizable elsewhere, but a PSD must be imported with its particulate carrier, sizing model, weighting basis, and closure requirement.

Its character: mixed-structural because Measurement owns the portable instrument–attribute skeleton while granulometric sampling, operational size, and distributional accounting determine literal qualification.

Structural Core vs. Domain Accent

Particle-Size Distribution is a domain-specific granulometric abstraction rather than a prime and is a strict kind of Measurement. Its complete signature fixes a representative particulate sample, an operational model of particle size, a sizing instrument and procedure, preparation and dispersion state, ordered size support, a declared mass-, number-, volume-, or other weighting basis, normalized allocation across intervals or a cumulative function, uncertainty, and closure accounting for nearly all declared material.

What is skeletal (could lift toward a cross-domain prime). Measurement supplies a target and attribute, coupling to an instrument, procedure, scale and units, calibration and observer frame, uncertainty, and an interpretable result. That complete skeleton recurs in clinical blood-pressure measurement, cadastral surveying, and electrical power metering—three unrelated domains. PSD instantiates it with a distribution-valued result rather than one scalar.

What is domain-bound. Particles, equivalent-diameter conventions, sieving, sedimentation, imaging and scattering branches, sampling representativeness, dispersion and segregation, ordered size bins, weighting bases, normalization, range–cumulative conversion, and loss of unmeasured particle attributes are granulometric accents. Remove the domain accent and Measurement remains; remove the structural core of instrument-mediated, framed, uncertainty-bounded attribution while retaining a histogram shape, and there is no valid PSD measurement.

Why this does not clear the prime bar. The complete named signature cannot recur literally in three unrelated domains without importing particulate samples, operational diameter, and conserved allocation across size. Measurement already owns the portable evidential architecture. Prime promotion would either duplicate that parent or turn a specialist distributional output and its sampling obligations into requirements for measurement generally.

Instantiates / Related Primes

This entry is a kind of Measurement.

Instantiates — Measurement (Measurement). A particle-size distribution is a strict granulometric measurement whose target attribute is the apportionment of a particulate population by operational size. Ordered size support supplies the scale; sieving, settling, imaging, scattering, or counting supplies the instrument and procedure; the operational diameter, size unit, weighting basis, sampling and preparation state supply its unit and observer frame; and uncertainty includes both instrumental error and representativeness of the sample. Handling and dispersion also realize Measurement’s coupling boundary because the procedure can segregate, abrade, dissolve, or disaggregate the particles it measures. The distribution-valued result adds the specialist residual: normalized material amount across size classes, with explicit range–cumulative conversion and closure.

Related to — Representation (Representation). Interval fractions and cumulative curves are alternative media for the same accounted particle amount when size limits, weighting, and normalization are conserved. Representation explains this encoding relation but not the upstream sampling, sizing, and measurement obligations that make the values a PSD.

Decline — Pattern (Pattern). A PSD may be flat, irregular, or unique to one sample. It need not exhibit a recurrent organization under admissible variation or exceed a null baseline; normalized allocation across operational particle size is sufficient. Pattern may be found in a series of PSDs, but it is not their genus.

Relationships to Other Abstractions

Current abstraction Particle size distribution Domain-specific

Parents (1) — more general patterns this builds on

• Particle size distribution is a kind of Measurement Prime

  A particle-size distribution is a strict granulometric measurement whose target attribute is the apportionment of a particulate population by operational size.

Hierarchy path (1) — routes to 1 parentless root

• Particle size distribution → Measurement

Neighborhood in Abstraction Space

Particle size distribution sits in a sparse region of the domain-specific corpus (87^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Measurement Standards & Material Properties (10 abstractions)

Nearest neighbors

• Measurement Method — 0.82
• Radial Immunodiffusion — 0.81
• Flocculation — 0.81
• Characteristic Property — 0.81
• Growth curve (biology) — 0.80

Computed from structural-signature embeddings · 2026-10-08

Not to Be Confused With

• Mean particle size. Mean size reduces a sample to one central value, whereas a PSD accounts for the declared particle amount across ordered size classes or a normalized function. Tell: ask whether the report preserves the full allocation by size or only one summary statistic.
• Particle-size analysis. Particle-size analysis is the measurement and inference process used to obtain a distribution; the PSD is the resulting apportioned description. Tell: distinguish the sampling, preparation, instrument, and model steps from the curve or class fractions they produce.
• Size range. A minimum-to-maximum range reports endpoints without showing how much material lies between them. Tell: require normalized interval or cumulative amounts across the support rather than two extreme diameters.
• Cumulative size distribution. A cumulative distribution gives the amount smaller or larger than each threshold and is one representation of a PSD, not a different material property. Tell: inspect whether values accumulate across thresholds or give fractions within successive intervals before comparing curves.
• Particle-size histogram. A histogram is a binned graphical representation whose shape depends on binning and weighting; it constitutes a PSD only when tied to a defined sample, size model, and normalized amount basis. Tell: verify closure of mass, number, or volume fractions and declare the bin convention.
• Particle shape distribution. A shape distribution apportions elongation, roundness, or another morphology descriptor, information not recoverable from operational diameter alone. Tell: identify the measured variable on the horizontal axis—size versus a shape attribute.

References

[1] Jillavenkatesa, Dapkunas, and Lum, Particle Size Characterization, NIST Special Publication 960-1 (2001) (accessed 2026-09-13). registry ↩ Show verification details

[2] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[3] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[4] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[5] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[6] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[7] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[8] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[9] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩

[10] Unverified encyclopedia synthesis; no authoritative source located for the claim as written. ↩