There is one page in an aggregate test report that determines how the material will behave on site more than any other item: the particle size distribution. The water demand of concrete, the void content of asphalt, whether a sub-base layer will compact and whether the pump will block are largely written into that curve. Even so, on most sites the grading curve is glanced at only to check whether it falls inside the band. Yet the curve describes both the product and the setting of the crushing and screening line that produced it on that day. This guide follows sieve analysis from the laboratory bench through to reading the curve and to the decision taken on site.
Why the particle size distribution is the identity of an aggregate
Concrete, asphalt and sub-base aggregate all work on essentially the same logic: the particles build a skeleton and a binder or fine material fills the voids between them. The volume of that void comes directly from the particle size distribution. If the void is large, more cement paste, more bitumen or more fine material is needed, and that means both cost and durability. If the void is small, the same performance is reached with less binder. Grading is therefore the single property that determines the price and the behaviour of the material at the same time.
The second point is this: two loads coming from the same quarry, the same rock and the same crusher can be completely different materials because the screen setting has changed. The rock named at the top of the test report is limestone in both cases; yet one works faultlessly in concrete while the other segregates, bleeds and strains the pump. When buying aggregate, therefore, the first question to ask is not what the rock is, but what the particle size distribution is and how consistently it stays there.
This is why sieve analysis is the first and most frequently repeated test in a quality control laboratory. It does not only check the product; it also reports on the state of the production line, from the jaw crusher setting to blinded screen apertures, from the feed rate to the washing flow. For the general framework of which fraction is used for which job, see the guide to choosing size fractions; here we deal with how those fractions are measured and interpreted.
How a sieve analysis is done: from sample to weighing
The most critical step of the test is taken not in the laboratory but in the stockyard: sampling. An aggregate stockpile is segregated by its very nature; coarse particles roll to the foot of the cone while fine material stays at the top and in the centre. A shovelful taken from a single point does not represent that pile. The correct method is to take a large number of increments from different areas and different levels of the pile and combine them, or, where possible, to take a full cross-section sample from a moving conveyor belt. The combined sample is reduced to the test quantity by coning and quartering or with a sample splitter. What represents the material is the sample reduced through the chutes of a splitter, not one scooped from the middle with a shovel.
The sample is dried in an oven to constant mass and weighed; this first weighing is the denominator of every percentage calculation. Then, for materials containing fines, wet sieving is carried out: the sample is washed with water over the 0.063 mm sieve, the fine dust is removed with the water, and the remaining material is dried and weighed again. Skip this step and the result is systematically wrong, because in dry sieving the fine dust sticks to the moist surface of the coarse particles and never passes the 0.063 mm sieve. One of the most common misconceptions on site is that a dry-sieved sample shows a lower fines content than the material really has.
Sieving is carried out with a nest of square-mesh sieves. The sieves are stacked with the largest aperture at the top and the smallest at the bottom, with a pan underneath. The common series has apertures of 63, 31.5, 16, 8, 4, 2, 1, 0.5, 0.25, 0.125 and 0.063 mm; intermediate sieves are added or removed according to the size of the product. The nest is placed on a shaker or shaken by hand. The duration must be long enough for all the material that can pass each sieve to do so; short sieving makes the material look coarser than it is. The material retained on each sieve is weighed separately and these masses are expressed as proportions of the total sample mass.
The calculation has two steps. First the mass retained on each sieve is expressed as a percentage of the total. These percentages are then added from the top downwards to give the cumulative percentages retained; subtracting them from a hundred gives the cumulative percentage passing each sieve aperture. The grading curve is drawn from these passing percentages. At the end of the test, the sum of the masses collected from the sieves is compared with the initial mass; if the difference exceeds the acceptable limit, material has stuck in the sieves or been spilled and the test must be repeated.
The grading curve: reading the semi-logarithmic graph
The grading curve is a graph with the sieve aperture on the horizontal axis and the cumulative percentage passing that sieve on the vertical axis. The horizontal axis is logarithmic, and the reason is a practical necessity. An aggregate mix spans a thousandfold size range between 0.063 mm and 63 mm. On a linear axis the fine end is pressed against the left edge of the graph and becomes unreadable; on a logarithmic axis every decade occupies the same width and the fine and the coarse end are seen in equal detail.
Reading the shape of the curve gives information faster than reading the numbers. A steeply rising curve shows that the material is concentrated in a narrow size range, that is, that it is close to single-sized; this is the typical picture of a washed drainage material or a single-fraction crushed stone. A flatter, evenly spread curve says that particles of every size are present, that is, that the distribution is continuous and well graded. A section that flattens out in the middle of the curve points to a shortage of particles in that size range, that is, to a gap grading.
A few characteristic points on the curve make the work easier. The size below which ten per cent of the material falls is read as D10, the size below which half of it falls as D50, and the size below which sixty per cent falls as D60. The ratio of D60 to D10 is the uniformity coefficient: the larger the value, the wider the distribution and the better the compaction behaviour; the closer it is to one, the more single-sized the material. These indicators, which come from soil mechanics, are used directly in the selection of drainage and filter material; for the details see the article on choosing drainage and filter aggregate.
Continuous, gap and single-sized gradings; void content
In a continuous grading there are particles in every size range. Small particles settle into the voids between larger ones, still smaller ones fit between those, and the chain runs down to 0.063 mm. The result is the lowest possible void content and the highest packing density. Structural concrete, plant-mix base and dense asphalt mixes are designed on this logic, because the aim is a denser and more stable body with the same amount of binder.
In a gap grading, an intermediate size is deliberately reduced or removed altogether. The aim is for the coarse particles to touch one another directly and build a stone-to-stone skeleton; with the intermediate size absent, the coarse particles rest on one another and the load is transferred directly through the particles. The voids of this skeleton are then filled with a mortar rich in fine material and binder. Stone mastic asphalt is the best known application of this principle. The price of a gap grading is that the mix is more prone to segregation; laying and haulage require care.
A single-sized grading, as the name suggests, is concentrated in one narrow range. The void content is high and permeability is very good, but the stability gained through compaction is low and the particles can move against one another under load. Single-sized material is therefore used not as a structural layer but as a drainage body, a filter layer or a pipe surround fill.
The void content can also be measured directly. Comparing the loose and compacted bulk densities of the material with its particle density shows what percentage of the volume is air. The difference between the void contents of two materials produced from the same rock comes entirely from their particle size distribution. This measurement is a practical tool for estimating binder demand in mix design; it turns the continuity you read off the curve into a number. For how many tonnes a given volume amounts to and what that means for haulage, see the article on m3-tonne conversion.
The fineness modulus and particle size indicators
The fineness modulus is a practical indicator that reduces the coarseness of a fine aggregate to a single number. On a given series of sieves, in common practice the 0.125, 0.25, 0.5, 1, 2 and 4 mm sieves, the cumulative percentages retained are added together and divided by a hundred. The result gives a rough indication of which sieve range the average particle size of the sand falls into. A small value indicates a fine sand, a large one a coarse sand. For concrete sands this value usually moves within a band between two and three; specifications often set a permitted deviation from a target value.
Where the fineness modulus really earns its keep is in production monitoring. Even if the curve of the sand is not plotted every day, the fineness modulus can be calculated quickly and compared from day to day. If the value is falling, the sand is getting finer: water demand will rise, the mix will become sticky and more admixture will be needed for the same slump. If the value is rising, the sand is getting coarser: the mix begins to segregate and bleed, finishing the surface becomes harder and the risk of blockage in the pump increases. By following this single number, a plant operator can correct the mix early.
The fineness modulus has one important limitation, however: the same value can come from two very different curves. A balanced sand rich in medium sizes and a sand with plenty of very fine and very coarse particles but an empty middle can give the same modulus. The fineness modulus therefore does not replace the curve; it is read alongside it. The modulus is also useful when blending two different sands: the modulus of the blend comes out close to the weighted average of the individual moduli, so the proportion needed to reach a target value can be estimated quickly. We discussed when such a blend makes sense in the article on crushed sand or natural sand.
What does staying between the specification limit curves mean?
Concrete, asphalt and road specifications define the particle size distribution not with a single target curve but with a lower and an upper limit curve. The band between these two curves is the range of behaviour acceptable for that use. A material close to the lower limit of the band is coarser, one close to the upper limit is finer; both comply with the specification, but they behave differently in the mix. Staying inside the band is therefore only half the job.
The other half of the job is consistency. If the product curve leans against the upper limit one week and slides to the lower limit the next, both reports look compliant; but the plant produces two different concretes from the same recipe. The slump does not hold, strength fluctuates and the admixture dosage has to be changed constantly. In good production the curve settles at a particular place within the band and stays there; the real criterion of quality control is not being inside the band but being able to stay in the same place.
When the band is breached, the sieve at which it is breached is itself a diagnosis. Exceeding the upper limit at 0.063 mm means excess fines: washing may have been inadequate or clay may have got into the material. A deviation around 4 mm shows that the balance between sand and crushed stone has been upset and usually comes from the screen setting. Exceeding the upper limit at the coarse end means a tear in the screen mesh or a leak past it; it reports that the fraction above has found its way into the product.
A single fraction rarely has to fit the band on its own. In concrete and asphalt mixes several fractions are blended in set proportions and it is the blend that has to fall inside the band. The blend is calculated by multiplying the percentage passing each sieve for each fraction by its proportion and adding the results. This calculation also shows which part of the curve will move, and how, when you increase the proportion of a given fraction; it is the most practical tool in mix design.
Interpreting the curve on site: where a deviation points
A grading curve is a production report as much as a product certificate. A trough in the middle of the curve, that is, a shortage of particles in a particular size range, usually shows that a gap has been left between the apertures of the screening stages or that this size is being returned continuously in the closed circuit. A hump in the same region indicates that a crushing stage is producing that size in excess; the setting of the cone crusher or the fullness of its feed should be reviewed.
There are several typical reasons for the fine end swelling more than expected: overfeeding the crusher so that the particles grind one another, screen apertures blinded with damp fine material, and the washing unit going out of service or its flow dropping. Swelling at the coarse end is usually a mechanical leak; a tear in the screen mesh, a loosened seal or a wrongly fitted screen deck lets the size above find its way into the product. In both cases the curve tells you where to look before you walk the line.
The symptoms of the product on site match the curve as well. Segregation and bleeding when placing concrete point to a coarsening sand and a shortage of the middle sizes. A rise in pump pressure and blockages herald a mix that is either too coarse or too rich in fines. A sub-base layer that will not close under the roller, with coarse particles moving at the surface, means the fine fraction is inadequate. A layer that turns muddy and heaves shows an excess of fines. We set out the link between these symptoms and the measured properties of the material in detail in the article on how aggregate quality affects concrete.
When you see a deviation, the first thing to do is not to change the production setting but to take a second sample by the correct method and repeat the test. A sampling error and a production error look exactly the same on the same graph; the only way to tell them apart is repetition. Once a deviation is confirmed, the order of intervention is usually this: first screening, that is, the apertures, the mesh, the inclination and the shaking; then the crusher setting; and only last the feed and the quarry face.
Common errors that spoil the test, and the right habits
Most of the errors made in sieve analysis begin not in the laboratory but where the sample is taken. Material taken from the foot of a stockpile cone leans coarse, material taken from the top leans fine. A sample taken from the top of a truck body represents the coarse particles that have worked their way to the surface during haulage. The correct approach is to combine increments taken from several points and to reduce them with a splitter. Skip this step and all the care that follows is wasted.
The second common error is not drying the sample sufficiently. Damp fine material balls up, sticks to the surface of the coarse particles and does not pass the sieve; the result is that the fine fraction appears lower than it is. The third is overloading the sieve. When the material building up on a sieve becomes too thick, the particles underneath never reach the mesh, and the test counts material that should have passed as retained. In that case the sample should be split and sieved in two batches.
The fourth is the sieving time; too short a shaking period ends the test before the fine particles have finished passing. The fifth is sieve maintenance: sieve wires stretch over time, the apertures open up and the corners deform. A worn sieve systematically passes coarser material, and because the difference is small it escapes notice for a long time. Sieves should be checked periodically and replaced if there is visible deformation. Finally, the mass balance must be checked in every test; if the difference between the sum of the collected masses and the initial mass exceeds the acceptance limit, the result must not be used.
All these habits serve a single purpose: making the curve show the true state of the material. At its limestone quarry in Milas, İTS Maden treats the particle size distribution of the fractions it produces not as a random output of the production line but as a controlled parameter. To choose the fraction and the blend that match the limit curves in your specification, you can review our products, look at our approach to quality control and contact us.
