Surface Finish Calculator
Convert between Ra, Rz, and other surface roughness values used in machining, including ISO 1302 N-grade designations, right in your browser. Check a measured Ra in micrometers against a drawing callout like N11 or 0.4 μm, free and without sign-up.
N8
Ra ≤ 3.2 μm
≈ 125 μin
Standard as-machined CNC finish
| Grade | Ra (μm) | Ra (μin) | Typical process |
|---|---|---|---|
| N1 | 0.025 | 1 | Superfinishing, lapping |
| N2 | 0.05 | 2 | Lapping, honing |
| N3 | 0.1 | 4 | Fine honing / polishing |
| N4 | 0.2 | 8 | Fine grinding |
| N5 | 0.4 | 16 | Precision grinding |
| N6 | 0.8 | 32 | Fine grinding / fine turning |
| N7 | 1.6 | 63 | Precision CNC milling or turning |
| N8 | 3.2 | 125 | Standard as-machined CNC finish |
| N9 | 6.3 | 250 | General or rough milling / turning |
| N10 | 12.5 | 500 | Rough machining, heavy cuts |
| N11 | 25 | 1,000 | Coarse machining, as-forged / as-cast |
| N12 | 50 | 2,000 | As-cast, flame-cut, sand casting, forging |
Usage examples
Reading an ISO 1302 N-grade off a drawing
Look up the N6 grade called out on a legacy drawing to see its Ra ceiling in both μm and μin and the kind of process it typically implies.
Comparing a measured Ra to a spec grade
Check whether a measured Ra of 2.1 μm still falls under N7 or has to be classified as N8, since grades are upper-limit ceilings, not midpoints.
Getting a ballpark Rz from a known Ra
Convert a datasheet's Ra value into an approximate Rz range when a quick sanity check is all that is needed, not a certified conversion.
Checking a ground bearing seat against a 0.4 μm Ra callout
A precision-ground shaft seat specified at Ra 0.4 μm (N5) converts to an estimated Rz of roughly 1.6-2.8 μm — enough to sanity-check a profilometer readout without needing a certified lab conversion.
Judging an as-cast housing surface against N11
Look up N11 to confirm that an as-forged or as-cast housing face, with no secondary machining, falls within the expected Ra ≤ 25 μm (1000 μin) ceiling before deciding whether a finish pass is needed.
Quick reference
ISO 1302 N-grade reference table
| Grade | Ra (μm) | Ra (μin) | Estimated Rz range (μm) | Typical process |
|---|---|---|---|---|
| N1 | 0.025 | 1 | ≈ 0.1 – 0.175 | Superfinishing, lapping |
| N2 | 0.05 | 2 | ≈ 0.2 – 0.35 | Lapping, honing |
| N3 | 0.1 | 4 | ≈ 0.4 – 0.7 | Fine honing / polishing |
| N4 | 0.2 | 8 | ≈ 0.8 – 1.4 | Fine grinding |
| N5 | 0.4 | 16 | ≈ 1.6 – 2.8 | Precision grinding |
| N6 | 0.8 | 32 | ≈ 3.2 – 5.6 | Fine grinding / fine turning |
| N7 | 1.6 | 63 | ≈ 6.4 – 11.2 | Precision CNC milling or turning |
| N8 | 3.2 | 125 | ≈ 12.8 – 22.4 | Standard as-machined CNC finish |
| N9 | 6.3 | 250 | ≈ 25.2 – 44.1 | General or rough milling / turning |
| N10 | 12.5 | 500 | ≈ 50 – 87.5 | Rough machining, heavy cuts |
| N11 | 25 | 1,000 | ≈ 100 – 175 | Coarse machining, as-forged / as-cast |
| N12 | 50 | 2,000 | ≈ 200 – 350 | As-cast, flame-cut, sand casting, forging |
Frequently asked questions
What exactly is the ISO 1302 N-grade system?
It's a standardized set of twelve roughness grades (N1 through N12), each defined by an Ra ceiling value that roughly doubles from one grade to the next. N1 is the smoothest (superfinishing/lapping level), N12 the roughest (as-cast or flame-cut level). Note that ISO 1302 itself has since been superseded by ISO 21920, and N-grades are technically a legacy notation — but they remain extremely common on existing drawings and in shop-floor conversation, which is why this tool still covers them.
Is the Ra value for a grade the typical value or the maximum?
It's the maximum (ceiling) allowed for that grade, not a midpoint or average. For example, a measured Ra of 2.1 μm exceeds the N7 ceiling of 1.6 μm, so it has to be classified as N8, even though it's much closer to N7's value than to N8's 3.2 μm ceiling.
Can I convert an Ra value to an exact Rz value?
No, and we deliberately don't pretend otherwise. Ra (arithmetic average roughness) and Rz (mean roughness depth) measure different statistical properties of the same surface profile, and the ratio between them depends on the shape of that profile — which varies by machining process, tooling, and material. Commonly cited references put the ratio roughly between 4x and 7x, but treat any single fixed multiplier you see elsewhere (e.g. "Rz = 7.2 × Ra") with suspicion — it is not a universal physical law, just a convenient approximation for certain surfaces.
Why don't you break the Ra-to-Rz ratio down by process (grinding vs. turning, etc.)?
We looked into this, and found that different published sources disagree on which direction the ratio goes for which process — some claim ground surfaces have a lower multiplier and turned surfaces a higher one, others claim the opposite. Rather than hard-code a directional claim we couldn't independently confirm, we show one honest, process-agnostic range instead of a falsely precise, possibly backwards, per-process number.
The process description for my grade doesn't match what I'd expect — why?
The process labels (lapping, grinding, turning, etc.) are common industry convention, not part of the ISO specification itself. Any given process can produce a range of surface finishes depending on feed rate, tooling, and setup, so treat these labels as a general guide rather than a strict rule.
Is Ra the same thing as RMS roughness?
No, though they're often confused and the numbers are close for many real surfaces. Ra is the arithmetic average of the absolute deviations from the mean line, while RMS (Rq) is the root-mean-square of those same deviations. Because RMS squares each deviation before averaging, it weights larger peaks and valleys more heavily, so Rq is typically about 11% higher than Ra for a random (Gaussian-like) surface profile — but that relationship isn't fixed either, and some older U.S. drawings use "RMS" loosely to mean what is actually an Ra measurement, so always check which one a callout actually means.
Why might two measurements of the same part give different Ra readings?
Ra depends heavily on measurement setup, not just the physical surface. The sampling length (cutoff, often denoted λc), the direction of the trace relative to machining lead marks (measuring across lathe turning marks gives a much higher Ra than measuring along them), the instrument type (contact stylus vs. optical), and even stylus tip radius can all shift the reported value. A single Ra number on a drawing is meaningless without an implied or stated measurement standard and direction, which is part of why real quality processes specify measurement conditions alongside the roughness value.