Grit-to-Surface-Finish (Ra) Chart: Which Grit Gives Which Finish

Quick Answer

On the FEPA-P scale, P40–P60 leaves roughly Ra 3–5 µm (stock removal), P150–P240 about 0.8–1.5 µm (pre-finish), and P800–P1200 about 0.1–0.3 µm (polish prep). These are indicative bands, not guarantees: material, pressure, lubrication, and abrasive wear all shift the measured Ra, so confirm a critical finish with a profilometer.

What a grit-to-surface-finish chart actually tells you

A grit-to-surface-finish chart maps a grit band to the surface roughness it typically leaves and the job it suits, from coarse stock removal through polish prep. It answers the recurring shop-floor question — "which grit gives which finish?" — but it is a starting point for picking a grit, not a guarantee. Actual roughness depends on the material, the applied pressure, lubrication, the machine, and how worn the abrasive is, so a measured finish is verified with a profilometer, not assumed from the grit number (Alves, 2015; see Surface Roughness Parameters).

Surface roughness is most often reported as Ra — the arithmetic-average roughness, the mean of the absolute height deviations from the profile mean line over a defined evaluation length. A lower Ra means a smoother surface. Ra is so dominant that it is frequently the only roughness symbol on a drawing, even though it averages away the deep isolated scratches that other parameters (Rz, Rmax) would catch. Ra is quoted in micrometres (µm) or microinches (µin); the conversion is 1 µm = 39.37 µin, commonly rounded to ≈ 40 µin, so an Ra 0.8 µm finish ≈ 32 µin (Astro Pak, 2026).

Two ground rules before you read any chart:

• Name the scale. All grit numbers in the chart below are on the FEPA P-scale for coated abrasives (the "P" prefix, standardised as ISO 6344). Above roughly P220 the FEPA and ANSI/CAMI scales diverge, so a "400" on one scale is not the "400" on the other. Always confirm the scale before reading the table.
• Name the cutoff. Ra is meaningless without its cutoff (λc). For most abrasive-finished metal — Ra ~0.1–2 µm — the default cutoff is λc = 0.8 mm, which is why "0.8" is baked into most gauge presets (Mitutoyo, per ISO 4288 / ASME B46.1, 2026).

The grit-to-Ra chart (FEPA-P)

This is the consolidated crosswalk: a FEPA-P grit band, the indicative Ra it leaves in both microns and microinches, and the use band it suits.

The Ra figures above are approximate and overlap at the edges. The same grit can leave a markedly different finish depending on workpiece material (hardwood versus stainless), contact pressure, feed or belt speed, lubrication, and abrasive wear. In a controlled belt-sanding study, belt speed, pressure, and grit size each independently shifted the measured surface roughness (Alves, 2015). Treat the table as a selection guide and confirm a critical finish by measurement against the spec, not by the grit number.

The grit-to-finish ladder: each step up in grit drops the achievable Ra, from coarse stock removal at the bottom to polish prep at the top.

A finer per-grit ladder for metal

For shops working to a specific Ra rather than a band, single-grit landings on steel follow a steeper ladder. These are typical surface-finish-machine values (Hanningfield micron-to-grit chart, 2026); read them with the same caveats — they are floors, not promises.

It is rarely practical to hit a fine target with one grit. Operators step through three to four progressively finer grades — see Grit Progression and Sequencing below for why.

Why the chart is indicative, not a guarantee

The honest framing matters because Ra is set by the process, not the grit alone. Three forces pull a real result away from the chart:

• Material and pressure. Hardwood, mild steel, and stainless do not respond the same way to the same grit. In a controlled study, grit size was the only factor that significantly affected roughness on softwood, while a sibling study on eucalyptus found belt speed, pressure, and grit all influenced Ra, with the best finishes obtained at higher contact pressure (Alves, 2015; Varasquim, 2012). The variables that matter shift with the material.
• Abrasive wear. A worn or glazed abrasive stops cutting and only polishes the peaks, so the same grit number can leave a coarser-than-expected finish late in its life.
• The coarsest surviving scratch. The floor Ra of a finish is governed by the deepest scratch that survives the sequence, not the last grit used. Skip too many grits and the chart's prediction breaks.

That last point connects the chart to grit sequencing. The governing rule is never skip more than one grit step — 80→120 is fine, 80→220 is not. A finer abrasive physically cannot reach the bottom of a coarser grit's scratch valleys, so skipping leaves buried "ghost" scratches that telegraph through stain or clearcoat. Pick the band that brackets your target Ra, then reach it through a disciplined ladder rather than a jump. See our guide to grit progression and sequencing — and why you should never skip more than one step for the full method.

Ra is not the only number on the drawing

Ra averages every point, so a few isolated deep scratches get diluted across the trace and can still produce an in-spec Ra. Rz (maximum height of profile) and Rmax/Rt (the single worst peak-to-valley) key off the extremes and expose what Ra hides — which is why Rz is always higher than Ra. A rule-of-thumb often cited is Rz ≈ 7.2 × Ra for typical ground or abraded surfaces, though the real ratio scatters widely with the process (Fractory, 2025). Conversion factors between Ra, Rz, and RMS are approximations that only hold for similar surfaces produced by similar processes; you cannot reliably back-convert a measured Rz into a guaranteed Ra. If a customer specs Rz, measure Rz.

The standards behind these numbers are also in transition. ISO 4287 (parameter definitions) and ISO 4288 (measurement rules) were the long-running international pair; the newer ISO 21920 family, published December 2021, modernizes and replaces them, computing most parameters once over the whole evaluation length rather than averaging (Digital Surf; ZEISS, 2026). The US parallel is ASME B46.1, whose 2019 edition is current. The parameter symbols are mostly unchanged, but how Rz and the peak/valley parameters are computed changed, so two "Rz 6.3" callouts under different standards are not strictly interchangeable. Confirm which standard a spec references before comparing.

Reading the chart against finish names buyers actually order

Most fabrication and stainless customers do not think in FEPA grit — they buy to a finish name or an Ra band. The stainless mill-finish designations map to Ra like this (Astro Pak, 2026):

Note the gulf: a No. 4 brushed finish (a flap-disc or belt job, reached with roughly 180–240 grit non-woven or coated belt) is about 40 times rougher than a No. 8 mirror, which requires successive polishing through buffing compounds well beyond any single bonded grit. If you are working toward one of these targets, our walkthrough on how to achieve a #4 brushed or #8 mirror finish on stainless steel ties the grit ladder to the finish name. For a broader job-by-job grit map, see the abrasive grit chart for metalworkers.

One precondition for any of this to hold: the grit must be honestly graded to standard. A "180" sheet that carries scattered 80-sized grains leaves rogue deep scratches the next step cannot remove — silently breaking the ladder and the chart. Tight grading to a published standard is not a spec nicety; it is the foundation the whole crosswalk stands on.

The Whitby Abrasives recommendation

Whitby Abrasives spans both ends of this crosswalk, from coated belts and sheets for the fine bands to flap discs and non-woven products for coarse stock removal and final scuff. We publish the grit-to-Ra chart with the scale named (FEPA-P) and the indicative Ra stated because spec honesty is the wedge — we would rather tell you a band is indicative and process-dependent than print a bare grit number with no finish context. The objection we hear is "value-tier means low quality," and the answer is the opposite: honestly graded, stamped grit is exactly what makes your progression hit the chart, and premium grain is wasted if a mislabeled coarse grain hides in the lot.

• Step into the fine bands with sanding belts for the coated-abrasive ladder.
• Finish and de-tint with non-woven discs for the low-Ra scuff step.
• Reach the mirror end with polishing wheels for buffing well beyond bonded grit.

Frequently asked questions

What Ra does P120 grit leave?

On the FEPA-P scale, P80–P120 typically leaves an Ra of roughly 1.5–3 µm (about 60–120 µin), suited to blending, coarse shaping, and scratch removal. Single-grit charts put 120 grit closer to 1.32 µm (~53 µin) on steel, but the actual figure depends on material, pressure, and abrasive wear, so confirm with a profilometer for a critical spec (Hanningfield, 2026).

How do I convert Ra from microns to microinches?

Multiply microns by 39.37 to get microinches; 1 µm = 39.37 µin, commonly rounded to ≈ 40 µin. So an Ra 0.8 µm finish is about 32 µin, a common drawing callout, and an Ra 0.4 µm finish is about 16 µin (Astro Pak, 2026).

Why does the same grit give different finishes?

Because Ra is set by the process, not the grit alone. Workpiece material, contact pressure, feed or belt speed, lubrication, and how worn the abrasive is all shift the measured Ra. In controlled studies, belt speed, pressure, and grit size each independently affected surface roughness (Alves, 2015; Varasquim, 2012). Treat any grit-to-finish chart as a selection guide, not a guarantee.

What grit gives a #4 brushed stainless finish?

A standard No. 4 brushed stainless finish sits around Ra 1.0–1.3 µm (about 40–50 µin) and is typically reached with roughly 180–240 grit non-woven or coated belt. A finer dairy or sanitary No. 4 lands near 0.8 µm (~32 µin). A No. 8 mirror (~0.025 µm) requires buffing compounds well beyond bonded grit (Astro Pak, 2026).

Does FEPA-P and CAMI grit give the same finish at the same number?

Not above about P220. The FEPA P-scale and ANSI/CAMI scale track closely up to ~180–220, then diverge: Washington Mills puts P400 at ~35 µm versus CAMI 400 at ~23 µm, so a "P400" sheet is finer than a CAMI "400" (Washington Mills, 2026). Reading a grit-to-Ra chart off a CAMI number above ~220 will overstate fineness. Confirm the scale first.

Can I rely on a grit chart instead of measuring?

For non-critical work, the chart is a sound starting point. For a critical or contractual finish, no — Ra averages away deep isolated scratches, so verify the result with a profilometer against the spec, including the cutoff (λc) and, where function depends on peaks, Rz or Rmax (Fractory, 2025).

Sources

• Single-grit Ra ladder (36–400 grit + mirror) — Hanningfield, Surface Finish Chart (Micron/Grit Conversion), accessed 2026 — https://www.hanningfield.com/resources/glossary/surface-finish-chart-micron-grit-conversion/
• Ra definition, µm↔µin conversion (1 µm = 39.37 µin), and stainless #2B–#8 finish Ra bands — Astro Pak, Surface Roughness Average (Ra), accessed 2026 — https://astropak.com/surface-roughness-average-ra/
• Ra↔Rz relationship (Rz ≈ 7.2 × Ra rule-of-thumb; Ra masks isolated scratches) — Fractory, Surface Roughness Explained, 2025 — https://fractory.com/surface-roughness-explained/
• Cutoff (λc = 0.8 mm) default for Ra ~0.1–2 µm — Mitutoyo, Basic knowledge of surface roughness, per JIS B 0633 / ISO 4288, accessed 2026 — https://www.mitutoyo.co.jp/about-metrology/knowledge/roughness/
• ISO 21920 (Dec 2021) replacing ISO 4287/4288; ASME B46.1-2019 current US standard — Digital Surf, ISO 4287 vs ISO 21920, accessed 2026 — https://www.digitalsurf.com/blog/what-are-the-differences-between-iso-4287-and-iso-21920/
• FEPA-P vs CAMI divergence above ~220 (P400 ~35 µm vs CAMI 400 ~23 µm) — Washington Mills, FEPA Particle Size Conversion Chart, accessed 2026 — https://www.washingtonmills.com/resources/guides/fepa-particle-size-conversion-chart
• Standards bodies: FEPA P-scale standardised as ISO 6344; ISO 21920:2021 (profile roughness); ASME B46.1-2019 (US surface texture).
• Alves, M. C. de S., Santiago, L. F. F., Gonçalves, M. T. T., Valarelli, I. D., Varasquim, F. M. F. de A. (2015). Effects of belt speed, pressure and grit size on the sanding of Pinus elliottii wood. CERNE. DOI: 10.1590/01047760201521011216 (gold OA) — https://doi.org/10.1590/01047760201521011216
• Varasquim, F. M. F. de A., Alves, M. C. de S., Gonçalves, M. T. T., Santiago, L. F. F., Souza, A. J. D. de (2012). Influence of belt speed, grit sizes and pressure on the sanding of Eucalyptus grandis wood. CERNE. DOI: 10.1590/s0104-77602012000200007 (diamond OA) — https://doi.org/10.1590/s0104-77602012000200007