D10 D50 D90 particle size values can make a powder specification look more complete than it really is. Three numbers summarize a particle-size distribution, but they do not show whether the sample was dispersed correctly, whether the distribution has a problematic tail, or whether another supplier measured the same material the same way.

For procurement, the useful question is not simply “What is the D50?” It is whether the full distribution supports the process. Particle-size data should be read alongside physical properties such as bulk density, moisture, and flow behavior because powders sharing the same median size can still handle very differently.

Why a Single Particle Size Number Can Mislead Buyers

A powder is rarely made up of particles that are all the same size. Even a tightly controlled material contains a distribution of finer, middle-sized, and coarser particles. That means one reported value can summarize the material while still hiding changes at either end of the distribution.

For procurement teams, this distinction matters because different parts of the particle-size distribution can influence different process outcomes. Finer particles may affect dusting, dissolution, surface area, or dispersion, while coarser particles may create filtration, screening, nozzle, or surface-finish problems. The full distribution matters more than one isolated number.

Two supplier lots can also have almost identical median particle sizes while behaving differently in production. One batch may contain more fines, another may have a heavier coarse tail, and both can still report a similar D50. Looking at D10, D50, and D90 together gives buyers a clearer view of where those differences are occurring.

That is why particle-size specifications should be treated as process-control information, not just descriptive laboratory data. Before using the numbers to approve a supplier or compare batches, buyers need to understand what each percentile represents, how the distribution was measured, and whether the reported values relate to the actual failure modes of their process.

D10 D50 D90 Particle Size Values Are Percentiles, Not Averages

D10, D50, and D90 are points on a cumulative particle-size distribution. In a volume-based result, D10 is the size below which 10% of measured particle volume falls, D50 is the size below which 50% falls, and D90 is the size below which 90% falls. D50 is the median, not necessarily the arithmetic mean or the most common size.

The basis matters. Results can be reported by volume, mass, or number depending on the analytical technique and reporting convention. Buyers should not compare D-values until the distribution basis and measurement method are known.

Laser diffraction is widely used for particulate materials. The current laser diffraction standard explains that the method derives an equivalent-sphere size distribution from light scattering and that non-spherical particles can produce results different from methods based on other physical principles.

That is a reminder that particle size is method-dependent.

Each Percentile Answers a Different Procurement Question

The three values become useful when buyers stop treating them as interchangeable specifications.

D10 can expose movement in the fine fraction, which may matter for dispersion, dusting, dissolution, surface area, or filtration depending on the product. D90 gives more visibility into the coarse side, where larger particles or agglomerates may affect screens, nozzles, coatings, or downstream texture.

D50 is useful, but the median can hide change. Two powders can have nearly identical D50 values while one contains more fines and the other has a broader coarse tail.

Distribution Width Can Matter More Than a Small D50 Difference

A buyer comparing suppliers may be tempted to rank whichever product has the preferred D50. That can miss the characteristic that operations actually notice: distribution width.

One common descriptor is span, calculated as (D90 − D10) / D50. A larger span generally indicates a broader distribution relative to the median. It is not a universal quality score; acceptable breadth depends on how the powder must disperse, feed, filter, pack, or react.

For a process sensitive to coarse material, D90 may deserve a tighter limit than D50. Where excessive fines create dust or rapid dissolution, the lower end may deserve more scrutiny. A narrow distribution can be desirable in some systems, while a broader distribution may be useful in others.

Particle-size specifications should therefore connect to actual process failure modes, not simply copy numbers from a supplier certificate.

Measurement Conditions Can Move the Numbers

D-values are only comparable when analytical conditions are comparable. Laser-diffraction results can be influenced by sample preparation, wet or dry dispersion, dispersant choice, deagglomeration, concentration, optical inputs, and instrument settings.

The powder laser-diffraction method guide addresses development of wet or dry methods and volume-based distribution points such as Dv10, Dv50, and Dv90. For purchasing, the practical lesson is that the test method belongs beside the acceptance limit.

Agglomeration is especially important. A weak dispersion procedure may measure clusters rather than the intended dispersed particles. An aggressive preparation can break weak agglomerates that remain intact in the buyer’s real process.

The useful method is one that is defined, reproducible, and relevant to how the material will actually be used.

D90 Is Not the Same as an Oversize Limit

Buyers sometimes read D90 as though it means no particle exceeds that size. It does not. By definition, part of the measured distribution remains above D90.

That matters where a small amount of coarse material can block a filter, obstruct a nozzle, damage a finish, or create visible defects. A D90 requirement can characterize the coarse side of the main distribution without adequately controlling rare oversize material.

Depending on the product and failure mode, a separate sieve, residue, image-analysis, or maximum-particle control may be more appropriate. Procurement should separate distribution control from oversize control because they answer related but different questions.

Batch Trends Reveal More Than a Single Certificate

One certificate can show that a lot falls inside an agreed range; a sequence of certificates can show whether the supplier process is drifting toward finer, coarser, or broader material.

Trend D10, D50, and D90 together rather than watching the median alone. If D50 remains stable while D90 moves upward, the coarse tail may be changing. If D10 falls while the other values remain steady, the fine fraction may be increasing.

Those movements should be interpreted with related physical data. A shift in particle-size distribution may coincide with a change in bulk and tapped density, moisture, flowability, dispersion, or filtration performance. Looking across properties can expose a supplier or process change before the material fails outright.

D10 D50 D90 particle size data is most useful when it becomes a procurement control rather than three isolated numbers. Buyers should define the distribution basis, test method, preparation conditions, acceptable percentiles, and any separate oversize requirement, then trend those values against process performance. That turns particle-size data into evidence that the powder arriving today should behave like the powder originally qualified.

Frequently asked questions

What does D50 mean in particle size data?

D50 is the median of the stated particle-size distribution. In a volume-based result, 50% of the measured particle volume lies below that size and 50% lies above it.

Is a lower D90 always better?

No. The preferred D90 depends on the application. Some processes need tight control of coarse particles, while others can tolerate or even benefit from a broader particle-size distribution.

Can D10, D50 and D90 from different suppliers be compared directly?

Only when the measurement method, distribution basis, sample preparation, dispersion conditions, units, and other relevant test settings are sufficiently comparable. Otherwise, numerical differences may partly reflect the analytical method rather than the material.