Every machined part carries three pieces of information that engineers tend to handle separately: the roughness number on the drawing, the production method capable of reaching it, and the price of demanding it. A surface finish chart binds all three together. Interpreted correctly, it reveals whether a callout represents routine shop work or a serious expense—long before any supplier quotes the part.
The sections below unpack what the chart contains, how the two dominant parameters diverge, how to decode the roughness symbol, and which values each process genuinely delivers in production.
What a Surface Finish Chart Actually Does
Most charts organize roughness values against the processes that produce them, typically across four columns covering micrometers, microinches, process names, and value ranges. Yet the chart's deeper purpose is verification. When a print demands Ra 0.2 micrometers inside a milled pocket, one glance confirms that milling alone cannot get there. A secondary operation must enter the plan, along with its setup and inspection costs.
How Roughness Is Quantified: Ra, Rz, and Beyond
Measurement starts with a profilometer, which drags a fine stylus across the surface and logs every rise and fall. Software then compresses that profile into parameters, and each parameter emphasizes something different.
Ra, the roughness average, expresses the arithmetic mean deviation of the profile from its centerline. Its worldwide popularity comes from stability and repeatability. The trade-off is blindness to anomalies: averaging allows a surface marked by one deep scratch to report the same Ra as a uniformly textured one.
Rz approaches the profile from the opposite direction, averaging the largest peak-to-valley heights across several sampling lengths. Because it responds to extremes rather than means, it exposes exactly the flaws Ra conceals. European drawings favor it, and engineers specify it on sealing faces and bearing surfaces where a single defect can compromise function.
A common field approximation places Rz at four to seven times Ra, depending on the process. Treat this as orientation, not conversion—a drawing that calls for Rz should be inspected against Rz.
Two supplementary parameters appear occasionally. Rmax records the single largest peak-to-valley height within the evaluation length, while Rt spans the total height from the profile's tallest peak to its lowest valley.
Standard Values in CNC Machining
Across the range most machined parts occupy, three thresholds shape cost conversations. Ra 3.2 micrometers (125 microinches) is the default as-machined condition, requiring no extra operations. Ra 1.6 becomes available through dedicated finishing passes and marks the point where added cost starts becoming noticeable. Dropping below Ra 0.8 almost always means introducing grinding, honing, or lapping as a separate operation, and pricing climbs accordingly.
Every Process Has Its Ceiling
Each manufacturing method owns a natural roughness band. Pushing beneath it is possible only through slower feeds, repeated passes, or a different operation entirely.
Where a result lands within the band depends on cutting speed and feed rate, tool nose radius, edge wear, machine rigidity, coolant, and the workpiece material. Soft, ductile alloys such as aluminum and copper tend to tear rather than shear cleanly at low speeds, which is why their finishes improve as surface speed rises. Shops running modern CNC machining with rigid setups reliably hold the lower end of each range, but the process limit still applies—finish is decided by the process first and the operator second.
Decoding the Roughness Symbol
On a drawing, the symbol takes the form of a checkmark whose additions carry specific meaning. The parameter, value, and sampling details all attach to this basic shape.
One addition deserves particular attention: the lay direction mark, which indicates whether tool marks run parallel or perpendicular to the surface. On sealing faces and sliding contacts, this orientation influences leakage and wear behavior, making it functional rather than cosmetic.
The Cost Curve Behind the Numbers
A dependable rule when specifying: every step down the chart raises finishing expense, and steps below Ra 0.8 frequently bolt an entire new process onto the job. The economical strategy is simple—specify the loosest finish the function allows.
When Secondary Processes Enter
Treatments applied after machining reshape the surface in ways the original number may not anticipate. Grinding, honing, and lapping refine texture, while blasting, plating, and anodizing alter both appearance and measured values. Because of this, the drawing should state whether the roughness requirement applies before or after finishing. Aligning the callout with the intended surface finishing route prevents a part from being rejected over a value the coating itself caused.
From Instrument to Inspection
Industrial verification runs primarily on contact profilometers, with optical instruments reserved for surfaces too delicate to touch. Suppliers operating a controlled metrology room produce more repeatable readings than shop-floor spot checks. Comparator plates—physical reference samples judged by eye and fingertip—work well for quick sorting, but they cannot substitute for a measured value when the requirement is tight.
The Mistake That Costs the Most
The most expensive habit in specification is stamping one tight roughness callout across the whole part when only a surface or two truly needs it. Identify the functional faces, tag them individually, and leave everything else at the standard as-machined condition.
Three further habits prevent friction with suppliers: state the finish in the units the shop actually works in, name the parameter explicitly rather than assuming Ra, and confirm the chosen material can hold the value—soft aluminum and gummy plastics behave very differently from steel under identical tooling. Teams uncertain whether a callout is achievable in their alloy can review the drawing with XC Machining before release.
Frequently Asked Questions
Q: What is the standard surface finish for CNC machined parts?
Ra 3.2 micrometers, or 125 microinches, is the standard as-machined condition. It requires no additional operations and suits most functional surfaces. Ra 1.6 is achievable with finishing passes at modest added cost, while tighter values usually demand a secondary process.
Q: What is the difference between Ra and Rz?
Ra averages the profile's deviation from its centerline, smoothing over occasional defects. Rz averages the largest peak-to-valley heights, so it captures isolated scratches. Rz is preferred on sealing and bearing surfaces where one defect matters.
Q: Can Ra be converted to Rz?
Only approximately. Rz typically runs four to seven times Ra depending on the process, but the relationship is not fixed. A surface specified in Rz should be measured against Rz, not converted from an Ra reading.
Q: Does a smoother surface finish cost more?
Yes, and the increase is not linear. Moving from Ra 3.2 to Ra 1.6 adds finishing passes. Going below Ra 0.8 usually requires grinding, honing, or lapping—a separate operation with its own setup and inspection.
Q: How is surface finish measured?
Most commonly with a contact profilometer drawn perpendicular to the tool marks. Optical instruments handle delicate surfaces. Comparator plates allow quick visual checks but lack the accuracy for tight requirements.
