Surface-texture engine for machinists and quality engineers: bidirectional conversion between Ra (CLA), Rz, Rq (RMS) and ISO N-grades in µm or µin, with selectable Rz:Ra process ratio, Gaussian RMS relationship, theoretical turning finish Ra = f²/(32·r) from feed and tool nose radius, process-capability cross-checks and conversion-validity warnings. Every input with selectable universal units and reference-standard values, every report with the full audit trail.
SectorCalc SC-028 converts surface-roughness specifications between the systems found on international drawings: Ra (CLA), Rz (ISO ten-point height), Rq (RMS) and the ISO 1302 N-grades, in metric µm or inch µin. It cross-checks against the theoretical turning finish from feed and nose radius and flags when a conversion leaves its validity band. All cross-parameter conversions are statistical relationships, not identities — the report states the assumed ratios so a drawing reviewer can audit exactly what was claimed.
Ra is the arithmetic mean of profile deviations — robust, but blind to single deep scratches. Rz averages the five highest peaks and five deepest valleys, catching the extremes Ra hides; sealing surfaces and fatigue-critical parts specify Rz for exactly this reason. Rq (RMS) weights large deviations quadratically. N-grades (N1–N12) are the ISO 1302 shorthand: each grade is an upper Ra limit doubling per step from 0.025 to 50 µm.
None of these is an identity. The Rz:Ra ratio is set by the process: fine grinding gives ~4, turning with visible feed marks gives ~7. Choosing the ratio is an engineering statement — the engine forces you to make it explicitly instead of hiding a default.
The cusp geometry of an ideal round insert sets a floor no machine can beat: halving the feed quarters the theoretical roughness, doubling the nose radius halves it. Real results run 1.5–3× worse from built-up edge, vibration and tool wear — if measured Ra beats theory, suspect the filter cutoff or the instrument.
Typical achievable Ra bands used by the capability check: honing/lapping 0.025–0.4 µm, grinding 0.1–0.8, reaming 0.4–3.2, turning/milling 0.4–6.3, drilling 3.2–12.5, EDM 0.8–6.3, sand casting 12.5–25. Specifying N4 on a drilled hole is a cost error; specifying N10 on a ground shaft is a function error.
Cross-parameter conversion is defensible only inside 0.05–12.5 µm Ra and only for random-process profiles. Below N5 the profile is often deterministic (honing cross-hatch, grinding lay) and the Gaussian assumptions fail; above N9 the profile wavelength approaches the evaluation length. The engine warns when the result leaves the band. Rt is deliberately not converted — its peak-to-valley definition makes any fixed ratio to Ra misleading.
No. It is the most quoted member of a 4–7 band. On a drawing dispute, measure both parameters — never accept a converted value as inspection evidence.
Usually a wrong filter cutoff (λc) or a skidless probe flattening the profile. Occasionally a wiper insert genuinely beats round-insert theory — the formula assumes a simple nose radius.
Yes: CLA (centre-line average) and AA (arithmetic average) are historical names for Ra. Older US drawings in µin CLA convert directly: divide by 39.37 for µm.
Older DIN Rz averaged individual peak-to-valley heights and can differ ~10 % from ISO Rz on irregular profiles. Ry (Rmax) is a single maximum — always ≥ Rz. This engine uses ISO Rz; flag legacy drawings for measurement rather than conversion.
