A finish callout on a drawing is a number you hit on purpose, not luck. Most of it comes from two levers you already control, feed and nose radius, plus a sharp edge and a setup that does not chatter. Here is what Ra, Rz and RMS actually mean, the typical finish each process can reach, the turning formula that predicts it, and the method to dial in a target. These are reference ranges and standards, not a promise; confirm the verified cutting data for your exact insert with the advisor.
A roughness spec is a parameter and a value. Reading the wrong parameter is the most common way to fail an inspection on a part that was actually fine, so know which one the drawing calls out before you cut.
| Parameter | What it measures | Use it when |
|---|---|---|
| Ra | Arithmetic mean deviation from the mean line, the everyday roughness average | The default callout; good general measure, but it hides single deep scratches |
| Rz | Average maximum peak-to-valley height over several sample lengths | When peaks and valleys matter, sealing faces, fatigue-critical surfaces |
| Rq (RMS) | Root-mean-square roughness, like Ra but weights big deviations more | Often on optical and older US prints; slightly higher than Ra for the same surface |
| N grade | ISO roughness grade number, N1 to N12, mapped to an Ra value | Shorthand on drawings; N6 is Ra 0.8, N8 is Ra 3.2 |
Ra and Rz measure different things, so there is no exact factor between them. A common rough rule is Rz around 4 to 7 times Ra for typical machined profiles, but the real ratio depends on the surface. If the drawing calls out Rz, measure Rz; do not convert from Ra and hope.
These are typical Ra ranges from common practice, not guarantees and not from any one catalog. The finish you get depends on feed, nose radius, edge condition, rigidity and material. Use the band to sanity-check a callout against the process, then confirm with a measurement.
| Process | Typical Ra (µm) | Approx. µin | Notes |
|---|---|---|---|
| Rough turning / milling | 3.2–12.5 | 125–500 | High feed, depth and material removal; finish is not the goal |
| Finish turning / milling | 0.4–3.2 | 16–125 | Light feed, larger nose radius or wiper, sharp edge |
| Drilling | 1.6–6.3 | 63–250 | Bore wall; ream or bore after if a finer finish is needed |
| Reaming | 0.8–3.2 | 32–125 | Sizing and finishing an existing hole |
| Grinding | 0.1–1.6 | 4–63 | Standard route to a fine, controlled finish on hard parts |
| Honing | 0.05–0.4 | 2–16 | Bores, cylinder walls; crosshatch for oil retention |
| Lapping / polishing | 0.012–0.2 | 0.5–8 | Mirror and sealing surfaces; slow and dedicated |
Typical achievable ranges from common production practice. Microinch values are rounded. Subgroup, tooling, rigidity and material shift the result; confirm with a profilometer.
For a turned finish, the ideal roughness left by the tool is mostly geometry: the feed marks the nose radius leaves behind. The standard theoretical relationship is worth knowing because it tells you which lever to pull.
Example: feed 0.1 mm/rev with a 0.8 mm nose radius gives about (0.01 / 25.6) × 1000, roughly 0.39 µm theoretical Ra. The two takeaways from the f² term: halving the feed cuts theoretical roughness to about a quarter, and a bigger nose radius lowers it at the same feed. Real Ra is always rougher than the formula because of edge wear, built-up edge, vibration and material, so treat it as a floor, not a promise.
A larger nose radius or a wiper insert finishes better at a given feed, but it raises radial cutting force and chatter risk on slender or long parts. If the part rings, a smaller nose radius and a stiffer setup can beat a bigger radius that is vibrating. Finish is a system: tool geometry, feed, rigidity and the edge, not a single number.
Change one variable, re-cut, re-measure. A profilometer reading beats eyeballing it, and the chip and wear tell you why the finish moved.
You set the finish with feed and nose radius, but holding it across a run comes down to the right grade and a sharp edge for your material. That is the brand-neutral problem this tool solves: it grounds the grade, geometry and cutting-data recommendation in real catalog data and tells you when there is no verified match, instead of inventing one.
Free, no strings: 8-brand grade cross-reference (PDF) · ISO material-group cheat-sheet (PDF)