Grit Number vs Micron Size: Particle Grade or Finish Requirement?

“Use 600 grit,” “use a 15-micron abrasive,” and “achieve Ra 0.8 micrometers” are three different instructions. The first names a grade, the second concerns the abrasive particles, and the third concerns the finished surface. A grit conversion chart cannot turn all three into the same requirement.

To compare grit numbers and micron sizes, ask what is being measured before asking which number to buy.

One unit, different objects

A micron is a micrometer, written µm: one thousandth of a millimeter. That unit can describe an abrasive particle, a scratch depth, a coating thickness or a surface-roughness value. Sharing a unit does not make those quantities interchangeable.

Within a conventional grit system, a higher grade number generally means finer abrasive. When comparing particle diameters on the same basis, a smaller micron value means smaller particles. Neither statement predicts an exact finished Ra.

A micron label is not a collection of identical grains

Ask whether the supplier's number is nominal, a mean, a median or a limit. A median describes the middle of a distribution under the stated measurement basis; it does not say that every particle has that size or that none is larger. Do not relabel a nominal number as a maximum simply because a drawing requires a maximum.

FEPA's standards explanation makes this distinction useful even for ordinary sandpaper: its macrogrit diameters are defined as ranges rather than single exact values. A table that compresses each grade to one number is a simplified comparison, not the whole particle specification.

For a demanding purchasing requirement, obtain the relevant distribution and measurement basis from the abrasive manufacturer. For an ordinary workshop process, preserve the complete product and grade identity. More decimal places copied from an unrelated chart do not improve either decision.

Ra belongs to the workpiece

KEYENCE defines arithmetic mean roughness, Ra, from the measured roughness profile. It is an average height-related parameter, not the diameter of the abrasive that made the surface and not a limit on every individual scratch.

Consider a hypothetical job drawing requiring Ra 0.8 µm and an abrasive labeled 15 µm. Those two numbers cannot be compared as if the abrasive were nineteen times “too rough.” The grain does not have to penetrate the work to its full diameter. Contact, support, material response and the preceding surface all affect the result.

The reverse shortcut is also wrong: selecting a nominal 0.8-micron abrasive would not prove that the finished part meets Ra 0.8 µm. If the drawing specifies a measured finish, the finished part needs the agreed measurement. Visual gloss or a smooth feel can be useful observations, but neither certifies a numerical roughness requirement.

Turn the requirement into a usable process

Keep the desired result separate from the consumable used to achieve it. If the job specifies an abrasive, verify its identity. If it specifies roughness, clarify the parameter, units and inspection method before trial work. A note saying only “micron finish” needs clarification; silently treating it as a grit request could produce a convincingly wrong result.

For a process already suited to 600-grit sheet sanding, eQualle's ten-sheet 600-grit pack supplies 9 × 11-inch silicon-carbide sheets. It is a concrete consumable to evaluate, not a certificate of particle maximum size or final surface roughness.

Short overview: coarse and fine sanding roles

Grit and particle size help describe the tool. Roughness describes the result. Keeping those roles separate is what makes a finishing specification actionable.

Write an acceptance line that another person can repeat

Replace shorthand such as “600 grit finish” with the actual control. A consumable-controlled instruction can name the manufacturer, product family, grade system, backing, method, pressure or machine settings, and replacement rule. A result-controlled drawing can name the roughness parameter, cutoff or filter, sampling direction, instrument, and acceptance limits. The exact fields depend on the process, but the principle is stable: record enough information that purchasing, production, and inspection are not silently measuring three different things.

Also define what happens when the named abrasive is unavailable. An acceptable substitution process may require a documented equivalence from the supplier, a trial coupon of the actual workpiece material, and the same final inspection used for production. It should not rely on a web chart copied without its grading-system assumptions. Store the coupon result and instrument report with the revision that approved the substitute. Then, if gloss, adhesion, or roughness changes later, the team can trace the consumable instead of arguing about whether two familiar-looking numbers “should have been close enough.”