Cement Particle Size Distribution: PSD Guide

The particle size distribution of cement is the hidden specification behind every bag sold: it decides how fast the cement reacts with water, how much water the concrete needs, how strong it becomes at 2, 7 and 28 days, and how it behaves in the mixer, the pump and the formwork. Yet for decades it was controlled only indirectly — through the Blaine fineness and the sieve residue — because the instruments for measuring the full distribution were too slow for process control. The laser diffraction revolution changed that: the modern plant measures the complete distribution in minutes, and the particle size distribution has moved from a research curiosity to a production control parameter and a commercial differentiator. This article is a complete technical treatment of the cement particle size distribution: its physics, its mathematics, its measurement, its control in the grinding process, its effects on the cement and the concrete performance, and the optimization practice that uses it to balance strength, workability and cost.

1. What the Particle Size Distribution Is and Why It Matters

Cement is a powder of particles ranging from below one micrometer to above 100 micrometers, and the way those particles are distributed across the size range defines the cement’s behavior. The hydration of the cement minerals proceeds from the particle surface: a 1-micrometer particle hydrates almost completely within hours, a 30-micrometer particle takes weeks and contributes the late strength, and a particle above 60 to 90 micrometers hydrates so slowly that it behaves almost as an inert filler. The distribution therefore has a direct engineering meaning: the fine end drives the early strength and the heat of hydration, the middle range carries the main strength development, the coarse tail determines the ultimate strength and the efficiency of the cement, and the balance of the ranges sets the water demand and the workability.

The practical consequences are the reason the distribution matters commercially. Two cements with identical Blaine fineness can have completely different distributions: one with a steep, narrow distribution and one with a flat, broad one, and their strengths at 2 days can differ by 5 to 10 MPa and their water demands by several percent. The cement user — the ready-mix producer, the precast plant, the construction contractor — experiences the difference in the setting time, the strength development and the finishability, and the cement producer who controls the distribution controls the consistency of the customer’s experience. The distribution is also an economic variable: producing it costs energy, because the fine fraction is the expensive fraction, and the optimal distribution is the one that delivers the required performance at the minimum energy.

2. The Mathematics: The Rosin-Rammler Equation

The mathematical description of the cement distribution is the Rosin-Rammler equation, derived from the physics of grinding and classification and valid for cement over the size range of 1 to 200 micrometers. The equation states that the fraction R retained on the sieve size x is R = exp(-(x/x0)^n), where x0 is the characteristic size at which 36.8 percent of the material is retained and n is the distribution slope or uniformity constant. The two parameters have distinct meanings: x0 sets the position of the distribution on the size axis, and n sets its shape — the steepness of the cumulative curve. A high n, typically 1.0 to 1.2 for a well-classified ball mill product, means a narrow distribution with a small coarse tail; a low n, 0.7 to 0.9, means a broad distribution with a long tail. The equation converts into the logarithmic-linear form log(log(100/R)) versus log(x), where the data of a real cement plots as a straight line with the slope n, and the deviation of the real data from the line is itself diagnostic — a curved plot signals a bimodal distribution, typically from a grinding circuit with poor classification or a blend of components with different grindability.

The practical use of the Rosin-Rammler parameters is the compact description of the product: the cement’s whole distribution is summarized by two numbers, which the quality system compares with the target distribution of the cement type and the production history. The parameters are also the input to the performance models: the strength prediction, the water demand prediction and the hydration models all take the distribution as their material input, and the modern quality department’s optimization is done on the x0-n plane rather than on the single Blaine axis.

3. The Measurement: From Sieves to Laser Diffraction

The measurement technology of the distribution has evolved through three generations. The sieve analysis — the classical method of the weight retained on a series of sieves down to 32 or 45 micrometers — is simple, cheap and standardized, but it stops at the fine end where the interesting behavior lives, and its resolution at the coarse end is limited by the sieve tolerances and the clogging. The sedimentation methods — the Andreasen pipette and the X-ray sedimentometer — measure the fine range by Stokes law, but they take hours and are laboratory tools rather than production instruments. The revolution came with the laser diffraction: the sample is dispersed in a liquid or a dry air stream, the laser beam is scattered by the particles, and the angular scattering pattern is inverted into the size distribution in minutes, with a range from 0.1 to 2,000 micrometers and a resolution of 100 channels or more. The modern plant measures the distribution on every routine sample, and the online instruments in the separator or the air slide loop the measurement into the process control.

The measurement quality depends on the sample and the dispersion, and the discipline is the same as for any quality instrument: the sampling point and the frequency follow the standard, the sample is split and dispersed reproducibly, the instrument is calibrated with the standard reference material, and the results are verified against the Blaine and the sieve residue for consistency. The laser diffraction result is an equivalent-sphere distribution: the real irregular particles are represented by their optical equivalents, and the correlation with the performance is established per plant, because the optical properties of the cement — the refractive index and the absorption — affect the inversion and are calibrated against the microscopy.

4. How the Grinding Process Makes the Distribution

The distribution is made by the combination of the grinding and the classification, and each element of the circuit contributes its signature. The ball mill’s grinding mechanism — the impact for the coarse particles and the attrition for the fines — produces a characteristic distribution whose tail grows with the overgrinding, because the residence time distribution of the mill lets the fast-passing particles escape coarse and the slow-passing ones overgrind. The classification in the separator shapes the product: the cut size and the sharpness set the coarse boundary, and the bypass — the coarse particles passing directly to the fines — is the main source of the coarse tail. The two interact: the mill grinds the reject that the separator returns, and the reject’s composition determines the mill’s duty. The result is that the same product fineness can be made in many ways, and the distribution shape is the fingerprint of how it was made.

The control of the shape is the modern practice: the separator speed sets the cut, the air flows set the sharpness, and the bypass is minimized by the maintenance of the vanes, the seals and the dispersion. The mill’s contribution is controlled by the charge: the fine grinding of the second compartment with the small balls produces the fine fraction that the separator then removes, and the overgrinding is minimized by the ventilation, which carries the fines out of the mill before they are overground. The optimized circuit holds the distribution slope n within its target band while the Blaine is held by the separator speed, and the quality data of the distribution is reviewed with the mill data in the daily and the weekly loops.

5. The Distribution of the Cement Types

Each cement type has its characteristic distribution target, set by its composition and its market use. The ordinary Portland cements are ground to a distribution with a Blaine of 3,000 to 3,800 and an n of 0.9 to 1.1, balancing the early strength and the water demand. The rapid-hardening cements are ground finer, to a Blaine of 4,000 to 5,000, with the distribution shifted to the fine end, and their higher early strength costs the higher energy and the higher water demand. The sulfate-resisting and the low-heat cements are coarser, with a flatter distribution that reduces the heat of hydration at the cost of the early strength. And the blended cements with the slag, the fly ash and the limestone carry the distribution of the mixture: each component grinds to its own distribution in the mill, and the mixture’s distribution is the superposition, whose shape the recipe and the classification control.

The additions bring their own distributions into the mixture, and their effects are significant. The interground limestone, at 5 to 15 percent, grinds finer than the clinker and fills the fine end of the distribution, which is part of its strength mechanism: the limestone fines pack between the clinker particles and nucleate the hydration. The slag, harder than the clinker, stays coarser in the same mill, and the distribution of a slag cement has a coarser effective character at the same Blaine. The fly ash, with its spherical porous particles, distorts the equivalent-sphere measurement and brings its own water demand. The quality department’s distribution data must therefore be interpreted per cement type, with the component distributions measured separately when the recipe is optimized.

6. The Effects on the Fresh Concrete

The fresh concrete behavior — the workability, the water demand and the setting — is where the distribution’s effects are felt first. The water demand of the paste is set by the particle surface: the fine particles need water to wet their surface and to fill the space between them, and the water demand of the cement rises with the fine fraction and with the flatness of the distribution. The classic result is the packing concept: the optimal distribution is the one that fills the space between the coarse particles with the intermediate and the fine ones, minimizing the void volume and therefore the water demand at a given strength. The coarser tail of the distribution contributes the inert strength — the unhydrated cores that act as micro-aggregate — and the balance between the reactive fines and the inert tail is the workability-strength compromise.

The setting behavior follows the fine end: the hydration of the aluminate and the silicate fines drives the initial and the final set, and the cements with the fine-shifted distributions set faster, other things being equal. The pumpability and the finishability of the concrete follow the same physics: the fine end lubricates the paste and carries the finish, while the coarse tail makes the mix harsh when it is excessive. The concrete producers’ complaints — the sticky paste, the bleeding, the hard finish — are distribution effects as often as they are chemistry effects, and the cement producer who can see the distribution can answer the complaints.

7. The Effects on the Hardened Concrete

The hardened concrete performance is the commercial heart of the distribution question. The strength development follows the hydration kinetics of the size fractions: the fine fraction below 3 to 10 micrometers hydrates in the first days and drives the early strength; the intermediate fraction of 10 to 40 micrometers hydrates over the weeks and carries the main strength; and the coarse fraction above 40 micrometers hydrates over months and contributes the ultimate strength — or remains as the inert core that limits the cement’s efficiency. The practical relationship, documented across the industry’s research, is that the maximum strength for a given fineness is achieved when the distribution has the maximum surface for reaction with the minimum void space, and the empirical optima cluster around a distribution with about 10 to 20 percent below 3 micrometers, 50 to 60 percent below 30 micrometers, and a controlled tail below 10 percent above 60 micrometers.

The durability effects follow: the finer cement hydrates faster and consumes the capillary water earlier, improving the early density but increasing the early heat and the shrinkage cracking risk in the massive sections; the coarser tail leaves the unhydrated cores that can react later, and the late hydration of the coarse slag and fly ash particles is part of the blended cements’ long-term strength. The heat of hydration, the autogenous shrinkage and the sulfate resistance all have distribution components, and the modern concrete standards and the performance specifications increasingly reference the distribution indirectly through the strength development requirements. The cement producer’s distribution control is therefore a durability instrument, not only a strength instrument.

8. The Distribution and the Grinding Energy

The distribution is the interface between the quality and the energy of the cement, because the fine fraction is the expensive fraction. The energy to grind increases steeply with the fineness: the Bond energy for the product 80 percent passing size scales with the inverse square root of the size, so a cement at a 25-micrometer characteristic size costs about 15 percent more energy than one at 32, and the fine end below 5 micrometers costs disproportionately because it is produced by the low-efficiency attrition grinding. The economic optimum of the distribution is therefore not the finest product but the product that meets the strength and the workability specifications at the minimum energy, and the optimization finds the target distribution that balances the two.

The optimization levers are the separator and the mill: the separator’s sharpness and bypass set the coarse tail and the overgrinding, the mill’s charge and ventilation set the production of the fine fraction, and the grinding aids set the efficiency of the fine grinding by reducing the agglomeration. The documented result of the distribution optimization is a 5 to 15 percent energy saving at the same performance, or a strength improvement at the same energy, because the typical circuit overgrinds the fine fraction while leaking the coarse fraction through the bypass — both of which the distribution data exposes. The plant that measures the distribution controls the energy better than the plant that measures only the Blaine, because the Blaine hides the shape.

9. The Distribution in Process Control

The distribution has entered the process control loop with the online instruments. The online laser diffraction analyzers, sampling the separator fines or the air slide, deliver the distribution every few minutes, and the control system uses the distribution parameters — the characteristic size, the slope and the coarse tail — as the controlled variables, with the separator speed and the air flows as the manipulated variables. The advantage over the Blaine-only control is the speed and the resolution: the online analyzer sees the bypass events and the classification drift that the daily Blaine hides, and the fineness control becomes a distribution control. The documented experience of the plants with online analyzers is a marked reduction of the quality variability and of the energy at the same average quality, because the control acts on the shape as well as the position.

The offline loop remains the weekly practice: the laboratory laser diffraction on the standard samples, the comparison with the online analyzer, and the review of the distribution trends against the strength results. The quality report now carries the distribution parameters for every cement type, and the weekly review correlates them with the strength at 2 and 28 days, the water demand and the customer feedback, closing the loop between the distribution and the market.

10. The Distribution as a Commercial Tool

The distribution has become a commercial instrument in the differentiated cement markets. The cement types sold on performance — the rapid-hardening, the low-heat, the low-water-demand, the sulfate-resisting — are distinguished as much by their distributions as by their chemistry, and the producer’s quality certificate now reports the distribution parameters for the informed customers. The concrete producers and the precast plants select their cement partly on the distribution, because their processes — the early stripping, the fast cycling, the self-compacting mixes — are tuned to the cement’s hydration kinetics, and the distribution is the specifiable proxy for the kinetics. The export markets, where the cement must perform identically in a different climate and with different aggregates, depend on the distribution consistency, and the producer who controls it protects the brand.

The commercial consequence cuts both ways: the distribution is also the place where the production cost and the market price meet, and the producer’s competitive position is set by the efficiency of producing the required distribution. The producer who can deliver the specified distribution at 25 kWh per tonne has a cost advantage over the producer who delivers a variable distribution at 32, and the advantage is measured in the market share of the performance products. The distribution is therefore a strategic variable, and the quality department’s distribution capability is a commercial asset.

11. The Common Errors and the Diagnostic Value

The distribution data has its diagnostic value and its errors, and the experienced engineer reads both. The diagnostic value: a rising coarse tail at constant Blaine points to a rising separator bypass, a worn cage or a dispersion fault; a falling slope points to overgrinding or a feed change; a bimodal plot points to a grindability difference between the mixture’s components or a classification instability; and a drift of the whole distribution points to the separator speed drift or the feed moisture. The errors: the sampling errors from the non-representative samples, the dispersion errors from the agglomeration, and the optical errors from the uncalibrated refractive index, each of which produces a distribution that the data system cannot distinguish from the real one. The discipline of the quality system — the standard sampling, the calibration and the cross-verification with the Blaine and the sieves — is what separates the useful distribution data from the misleading.

12. The Road to Distribution-Based Quality Control

The road to the distribution-based quality control is a staged implementation. Stage one is the measurement: the laboratory laser diffraction installed, calibrated and validated against the Blaine and the sieves, with the distribution reported on the routine samples. Stage two is the understanding: the distribution trends correlated with the strength, the water demand and the customer feedback, and the target distributions established for each cement type. Stage three is the process: the distribution parameters added to the daily quality report, the separator and the mill operated against the target shape, and the weekly review run on the distribution data. Stage four is the automation: the online analyzer and the distribution-based control loop, closing the loop on the shape as well as the fineness. Stage five is the commercialization: the distribution in the quality certificate, the performance products specified on the distribution, and the customer relationships built on the consistency. The completed road is the state where the particle size distribution is what it should be — the real specification of the cement, measured, controlled and sold.

Frequently Asked Questions

Why are two cements with the same Blaine different in performance?

Because the Blaine measures only the total surface area, not its distribution. Two cements at the same Blaine can have different shares of fine and coarse material, and the fine share drives the early strength and the water demand while the coarse tail drives the ultimate strength. The distribution is the missing information.

What is the Rosin-Rammler equation?

It describes the cement distribution as R = exp(-(x/x0)^n), where R is the retained fraction at size x, x0 is the characteristic size and n is the slope. The two parameters summarize the whole distribution, and the modern quality practice controls them rather than the single Blaine number.

How is the distribution measured in the plant?

By laser diffraction: the sample is dispersed and illuminated by a laser, and the scattering pattern is converted into the size distribution in minutes. The online analyzers measure the separator fines continuously, and the laboratory analyzers verify the routine samples and the calibration.

What is the optimum distribution for strength?

The empirical optimum balances the reactive fine fraction, the packing and the coarse tail: roughly 10 to 20 percent below 3 micrometers for the early strength, 50 to 60 percent below 30 for the main strength, and a controlled coarse tail, with the exact target set per cement type and market. The optimum also minimizes the grinding energy at the required performance.

How does the separator control the distribution?

The separator’s cut size sets the position of the distribution, its classification sharpness sets the slope, and its bypass adds the coarse tail. The distribution-based control manipulates the cage speed and the air flows against the measured distribution parameters, holding the shape as well as the fineness.

Summary

The particle size distribution is the real specification of the cement: it decides the hydration kinetics, the strength development, the water demand, the workability and the durability, and it is the interface between the quality and the energy of the product. The modern measurement technology — the laser diffraction with its online instruments — has made the distribution a production parameter, and the modern practice controls the Rosin-Rammler parameters, the cut and the slope, rather than the Blaine alone. The optimization of the distribution balances the strength, the workability and the energy, delivering the performance at the minimum cost, and its commercialization distinguishes the performance products in the differentiated markets. The road is staged — measure, understand, control, automate, commercialize — and the completed road is the state where the cement is specified, produced and sold by its true quality: its particle size distribution.

13. The Measurement of the Particle Size Distribution

The particle size distribution of the cement is measured by the modern laser diffraction instruments: the sample is dispersed in the air or the liquid, the laser beam scatters on the particles, and the scattering pattern is converted into the size distribution by the Mie theory: the measurement covers the range of the 0.1-1000 microns with the reproducibility of the 1-3%, and the report provides the full distribution curve, the mean sizes (the d10, the d50, the d90) and the specific surface computed from the distribution. The alternative methods include the sieve analysis (the coarse fractions above the 32-45 microns), the sedimentation methods (the Andreasen pipette, the X-ray sedimentometer) and the air permeability (the Blaine surface, an indirect index of the distribution). The laser diffraction is the reference method of the modern quality laboratories, and the instrument calibration and the sampling discipline determine the reliability of the results.

14. The Rosin-Rammler Parameters and the Distribution Shape

The Rosin-Rammler distribution describes the cement PSD with the two parameters: the characteristic size (the x’) where the retained fraction is the 36.8% and the uniformity coefficient (the n) that defines the steepness of the distribution: the high n values (the 1.0-1.2) indicate the steep distribution with the narrow size range produced by the efficient closed-circuit grinding, the low n values (the 0.7-0.9) indicate the wide distribution with the significant fine and the coarse fractions typical of the open circuits. The distribution shape controls the cement behavior: the steep distributions hydrate quickly and develop the early strength but demand the higher water for the workability, the wide distributions improve the particle packing and the workability at the cost of the early strength, and the modern cement specifications increasingly include the distribution targets beyond the Blaine: the two-parameter description is the standard language of the PSD control.

15. The Strength Correlation and the Quality Application

The strength correlation of the particle size distribution is the practical quality application: the fraction below the 3 microns drives the early strength (the 1-7 day development) and the heat of hydration, the fraction between the 3 and the 30 microns contributes the strength over the weeks, and the fraction above the 60 microns reacts slowly and reduces the 28-day strength potential: the optimum distribution for the strength places the majority of the particles in the 3-30 micron range with the minimum of the over-fine and the coarse material. The quality laboratories correlate the PSD parameters with the mortar strengths of their cement, establish the plant-specific models and use the distribution control (through the separator and the mill adjustments) to manage the strength class with the minimum energy: the PSD is the link between the grinding energy and the cement performance.

13. The Measurement of the Particle Size Distribution

The particle size distribution of the cement is measured by the modern laser diffraction instruments: the sample is dispersed in the air or the liquid, the laser beam scatters on the particles, and the scattering pattern is converted into the size distribution by the Mie theory: the measurement covers the range of the 0.1-1000 microns with the reproducibility of the 1-3%, and the report provides the full distribution curve, the mean sizes (the d10, the d50, the d90) and the specific surface computed from the distribution. The alternative methods include the sieve analysis (the coarse fractions above the 32-45 microns), the sedimentation methods (the Andreasen pipette, the X-ray sedimentometer) and the air permeability (the Blaine surface, an indirect index of the distribution). The laser diffraction is the reference method of the modern quality laboratories, and the instrument calibration and the sampling discipline determine the reliability of the results.

14. The Rosin-Rammler Parameters and the Distribution Shape

The Rosin-Rammler distribution describes the cement PSD with the two parameters: the characteristic size (the x’) where the retained fraction is the 36.8% and the uniformity coefficient (the n) that defines the steepness of the distribution: the high n values (the 1.0-1.2) indicate the steep distribution with the narrow size range produced by the efficient closed-circuit grinding, the low n values (the 0.7-0.9) indicate the wide distribution with the significant fine and the coarse fractions typical of the open circuits. The distribution shape controls the cement behavior: the steep distributions hydrate quickly and develop the early strength but demand the higher water for the workability, the wide distributions improve the particle packing and the workability at the cost of the early strength, and the modern cement specifications increasingly include the distribution targets beyond the Blaine: the two-parameter description is the standard language of the PSD control.

15. The Strength Correlation and the Quality Application

The strength correlation of the particle size distribution is the practical quality application: the fraction below the 3 microns drives the early strength (the 1-7 day development) and the heat of hydration, the fraction between the 3 and the 30 microns contributes the strength over the weeks, and the fraction above the 60 microns reacts slowly and reduces the 28-day strength potential: the optimum distribution for the strength places the majority of the particles in the 3-30 micron range with the minimum of the over-fine and the coarse material. The quality laboratories correlate the PSD parameters with the mortar strengths of their cement, establish the plant-specific models and use the distribution control (through the separator and the mill adjustments) to manage the strength class with the minimum energy: the PSD is the link between the grinding energy and the cement performance.

This file is part of the Complete Cement Technical Package (931 files) available from cementequipment.org. Respective rights holders; library copy for the licensed single user.