Surface Finish Machining: Ra, Rz, and Drawing Callouts
- carystraley
- 4 hours ago
- 12 min read
A wrong surface finish callout on a precision machined component is not a minor paperwork error. It is a rejection, a rework cycle, and sometimes a scrapped part. Engineers specify Ra values by habit without understanding what Ra actually measures, and machinists interpret those callouts in ways that were never intended. The gap between what a drawing says and what a shop floor actually produces costs manufacturers real money every week. This article breaks down surface finish machining standards, explains the difference between Ra and Rz, and tells you exactly what your drawing callout is and is not specifying.
Table of Contents
Quick Takeaways
Key Insight
Explanation
Ra is an average, not a peak value
Ra averages all surface deviations over a sampling length. Two surfaces with identical Ra values can have dramatically different peak heights and functional behavior.
Rz captures peak-to-valley height
Rz measures the average of five highest peak-to-valley distances across a sampling length. It is far more relevant for sealing surfaces and fatigue-sensitive parts.
ASME B46.1 and ISO 1302 are not identical
North American drawings typically follow ASME B46.1 while European and automotive Tier 1 suppliers use ISO 1302. Confusing them causes measurement and acceptance disputes.
Unspecified cutoff wavelength changes everything
If no cutoff length (lambda-c) is called out on a drawing, the measuring instrument default applies. Different instruments default differently, producing non-comparable readings.
Tighter is not always better
Specifying Ra 0.4 where Ra 1.6 is functional drives up machining time and cost with zero quality benefit. Over-specification is as problematic as under-specification.
Lay direction matters for functional surfaces
A drawing callout without a lay symbol leaves the machinist free to produce any surface texture orientation. For bearing bores and sealing faces, lay direction directly affects performance.
CMM probing cannot replace a profilometer
Surface roughness measurement requires a dedicated stylus profilometer. CMM touch probes measure geometry, not texture. Mixing these tools produces false conformance.
What Ra Actually Measures, and Where It Falls Short
Ra, or arithmetic average roughness, is calculated by averaging the absolute deviations of the surface profile from its mean line over a defined sampling length. It is the most commonly specified surface finish parameter in North American machining drawings, and it is also frequently the wrong one for the job.
In practice, two completely different surface textures can produce the same Ra value. A surface with deep, sharp valleys spaced far apart and a surface with shallow, rounded undulations spaced closely together can both read Ra 1.6 micrometers. For a static cosmetic surface, that might be acceptable. For a dynamic sealing face or a fatigue-critical aerospace component, those two surfaces will behave entirely differently in service.
Ra works well when you need a consistent, general measure of texture across high-volume production runs. It correlates reasonably well to process consistency, which is why machinists and quality teams use it as a first-pass process control indicator. Where it breaks down is in functional applications where peak heights, valley depths, or surface bearing ratios actually drive part performance.
The Sampling Length and Cutoff Problem
Ra measurements are only comparable when taken with the same cutoff wavelength, also called lambda-c. The default cutoff values defined in ASME B46.1 are 0.8 mm for most general machining applications, but instruments set to different defaults will produce different Ra numbers on the same surface. A drawing that calls out Ra 0.8 without specifying the cutoff is technically incomplete.
Pro tip: Always specify the cutoff wavelength explicitly on drawings for surfaces where Ra is a functional requirement, not just a cosmetic one. Write it as Ra 0.8 / lc 0.8 mm or use the full ISO 1302 surface texture symbol with all parameters populated.
Rz: The Parameter Engineers Underuse
Rz measures the average of the five highest peak-to-valley distances within five consecutive sampling lengths along the measurement trace. Because it captures extreme events in the surface profile rather than averaging them away, Rz is more sensitive to the outlier peaks and deep valleys that drive functional problems.
For sealing surfaces, Rz is the parameter that actually predicts leak performance. A gasket compresses against surface peaks. If those peaks are too high, the gasket cannot conform and the seal fails under pressure. Ra will never tell you that because it averages the peaks into the general noise of the measurement. Rz exposes them.
For fatigue-sensitive components, such as aerospace structural brackets or automotive connecting rods, deep surface valleys act as stress concentration points under cyclic loading. Research in tribology and fatigue mechanics consistently shows that Rz and Rmax correlate more directly to fatigue crack initiation than Ra does. Specifying only Ra on a fatigue-critical part and calling it a complete surface finish requirement is a documented failure mode in component engineering.
When to Specify Rz Instead of Ra
Specify Rz as the primary or supplementary parameter any time the part involves: pressure sealing interfaces, dynamic contact surfaces such as bearing bores, fatigue loading conditions, coatings or platings where peak height determines adhesion, or any surface where a single deep scratch would cause functional rejection. In those cases, Ra alone is insufficient regardless of the value specified.
The relationship between Ra and Rz is not fixed, but for many turned and milled surfaces, Rz tends to run between four and seven times the Ra value. That ratio shifts based on the machining process and the nature of the tool marks. Do not attempt to convert between them using a fixed multiplier on drawing callouts. Measure both independently when both matter.
Drawing Callout Standards: ASME B46.1 and ISO 1302
North American drawings for precision machined components most commonly reference ASME B46.1, which defines surface texture parameters, measurement methods, and the check mark symbol with maximum roughness height indicated in microinches or micrometers. European drawings and most automotive Tier 1 supplier drawings follow ISO 1302, which uses a different symbol structure and a more complete parameter designation system.
The practical problem is that supply chains mix both standards. A Fort Wayne manufacturer machining components for an automotive Tier 1 customer may receive drawings with ISO 1302 surface symbols while operating to ASME B46.1 measurement practices internally. Unless the shop and the customer explicitly reconcile which standard governs measurement and acceptance, disputes are inevitable.
"Surface texture specification is one of the most poorly communicated requirements in engineering drawings. The symbol is used, but the required parameters, measurement conditions, and acceptance criteria are rarely fully defined." -- ASME B46.1-2019 Committee Technical Commentary
Reading the ISO 1302 Surface Texture Symbol
The ISO 1302 symbol is a triangle with a horizontal bar. Above the bar, up to three roughness parameters can be specified in defined positions. The upper position carries the primary roughness parameter and its maximum value. The lower left position carries the machining process or treatment. The lower right carries the sampling length and waviness parameters if required.
In practice, most drawings only partially populate this symbol, which creates ambiguity. A triangle symbol with only Ra 1.6 written above it tells the machinist the maximum acceptable average roughness but says nothing about Rz, lay direction, waviness, or measurement cutoff. For general machined surfaces on non-functional features, that may be enough. For precision contact surfaces, it is not.
Pro tip: When reviewing customer drawings for new jobs, flag any surface finish callout that specifies only Ra without cutoff length, lay symbol, or Rz. Request a DRF (Design Reference Form) or a quality plan clarification before machining. It is far cheaper to ask the question before setup than to rework after inspection.
Comparing Ra, Rz, and Rmax for Precision Applications
Understanding which surface finish parameter to specify requires knowing what each one measures, where it provides reliable functional information, and where it is likely to mislead you. The table below compares the three most common parameters used in precision machined component specifications.
Parameter
What It Measures
Best Application in Precision Machining
Ra (Arithmetic Average Roughness)
Average absolute deviation of the surface profile from the mean line over the sampling length. Sensitive to overall texture consistency but insensitive to isolated peaks or valleys.
General machined surfaces, process consistency monitoring, cosmetic surfaces, and non-functional exterior faces where average texture matters.
Rz (Average Maximum Height)
Average of the five highest peak-to-valley distances across five consecutive sampling lengths. More sensitive to extreme surface events than Ra.
Sealing faces, gasket interfaces, coated or plated surfaces, fatigue-loaded components, and any surface where peak height drives functional performance.
Rmax (Maximum Height of Profile)
The single largest peak-to-valley distance within the entire evaluation length. Captures the worst-case surface event.
High-pressure hydraulic sealing surfaces, aerospace fatigue-critical components, and surfaces where a single deep scratch constitutes a functional rejection.
The data consistently shows that manufacturing teams who specify Ra alone for sealing and fatigue applications report higher field failure rates compared to teams who add Rz or Rmax to the same drawings. Using multiple complementary parameters is not over-engineering. It is basic functional surface specification.
How Machining Processes Affect Surface Finish
Every machining process produces a characteristic surface texture, and understanding that texture is inseparable from specifying and achieving correct surface finish. The tool geometry, feed rate, cutting speed, depth of cut, and workpiece material all interact to determine the final Ra and Rz values. You cannot specify Ra 0.4 on a feature that will be end milled at production feed rates and expect to hit it without secondary operations.
Turning produces a predominantly lay-directional surface with relatively predictable Ra values for a given feed and nose radius combination. The theoretical Ra for a turned surface can be approximated as Ra equals feed squared divided by 31.2 times the nose radius, giving the machinist a direct dial between process parameters and predicted surface finish. Milling produces a more complex, crossed lay pattern where the relationship between feed and Ra is less linear.
5-Axis Milling and Surface Finish Complexity
Five-axis CNC milling introduces additional surface finish variables that do not exist in three-axis work. The tool tilt angle relative to the workpiece surface affects the effective cutting radius and therefore the cusp height between tool passes. At tilt angles near zero, the tool cuts with its center point, which runs at zero surface speed and produces poor surface finish and potential built-up edge on certain materials.
In practice, 5-axis finishing passes on complex surfaces require explicit tool axis orientation strategies to maintain consistent surface finish across the entire feature. Leaving this to default toolpath behavior in CAM software produces variable Ra readings across the same face, which causes inspection failures even when the average looks acceptable.
Wire EDM Surface Finish Characteristics
Wire EDM produces a recast layer on the surface that affects both Ra and Rz measurements. The recast layer is hard, brittle, and has a different microstructure than the base material. On fatigue-critical parts, this layer must be removed by a finishing pass or by specifying the number of EDM skim cuts required to reduce the recast layer to an acceptable depth. Specifying only Ra on an EDM surface without addressing recast layer depth is an incomplete surface specification for functional components.
Common Mistakes in Surface Finish Specification
A common mistake is applying a blanket surface finish callout to an entire drawing instead of specifying requirements feature by feature. Putting Ra 1.6 in the general notes and expecting machinists to differentiate between a bearing bore, a cosmetic housing face, and a non-functional datum surface is not a real specification. It is a default that most shops will meet for easy features and miss for critical ones.
Another frequent problem is specifying Ra values that require lapping or grinding on features that will be machined on a three-axis machining center. Calling out Ra 0.2 on a milled pocket floor is a setup for rework unless the process plan explicitly includes a secondary finishing operation. The drawing requirement must be achievable with the planned process, or the process must change to match the drawing. One or the other, not neither.
Microinches vs. Micrometers Confusion
ASME B46.1 traditionally uses microinches (µin) while ISO 1302 and most European references use micrometers (µm). One micrometer equals approximately 39.37 microinches. Ra 32 µin is approximately Ra 0.8 µm. Ra 63 µin is approximately Ra 1.6 µm. These conversions are well established, but mixed unit callouts on drawings or in quality plans create measurement disputes when machinists and inspectors are working from different unit assumptions.
The data consistently shows that unit confusion is most common in supply chains where North American primes work with domestic shops but reference international drawing standards. Establishing a single unit convention in the quality plan and verifying it at the drawing review stage eliminates this category of error entirely.
Surface Finish in Inspection and PPAP Documentation
Surface finish is a dimensional characteristic that must appear in PPAP documentation when it is a functional drawing requirement. A common oversight is treating surface finish as a process control item rather than a product characteristic, which means it gets measured during setup approval but does not appear in the control plan or the initial sample inspection report.
For customers requiring PPAP Level 3 or higher, surface finish measurements must be recorded on the Dimensional Results form for every specified surface. The measurement equipment, calibration status, and measurement method must be documented. A profilometer that is not on the calibration schedule is generating data that cannot be used for PPAP acceptance, regardless of how good the numbers look.
A2LA Accreditation and Surface Finish Measurement Traceability
A2LA accreditation requires that all measurement equipment used for accredited scope activities maintain NIST-traceable calibration. For surface finish measurement specifically, this means the profilometer stylus condition, the reference specimens used for instrument verification, and the calibration interval are all documented and controlled. An accredited metrology lab operating under A2LA scope provides surface finish measurement data that carries demonstrable traceability, which matters when customers audit your measurement system or when parts are disputed at first article inspection.
Surface finish data that lacks traceability is not just a paperwork problem. In automotive and aerospace supply chains, non-traceable measurement data can trigger a full measurement system analysis re-evaluation and delay production release. Getting the calibration infrastructure right before PPAP is not optional for precision machining shops competing for Tier 1 work.
Pro tip: If your shop uses a handheld contact profilometer for surface finish measurement, verify that the stylus tip radius and the filter settings match the requirements of the drawing standard being applied. A stylus tip radius of 2 µm and a Gaussian filter per ISO 11562 are standard for most precision machining inspection work. Deviating from these without documentation will produce non-comparable results.
Frequently Asked Questions
What is the difference between Ra and Rz in surface finish machining?
Ra is the arithmetic average of all surface deviations from the mean line over a sampling length. It is a general indicator of overall surface texture consistency. Rz is the average of the five highest peak-to-valley distances within five consecutive sampling lengths. Rz is more sensitive to extreme surface events and is more appropriate for functional surfaces such as seals, coated faces, and fatigue-loaded components. Both parameters are measured by a profilometer and both appear in ASME B46.1 and ISO 4287.
What does a surface finish callout of Ra 1.6 actually mean on a precision machined drawing?
Ra 1.6 means the arithmetic average surface roughness of the specified feature must not exceed 1.6 micrometers, measured with a contact profilometer using the default or specified cutoff wavelength. It does not specify Rz, lay direction, waviness, or the measurement evaluation length unless those are added to the callout. Ra 1.6 is a common general machining finish achievable by milling, turning, or grinding depending on the process parameters.
Can a CMM measure surface finish on precision machined components?
No. A coordinate measuring machine uses touch trigger or scanning probes designed to measure geometric features such as diameter, flatness, and position. The probe tip radius and measurement force are entirely unsuitable for surface texture measurement. Surface roughness requires a dedicated stylus profilometer with a fine stylus tip radius, typically 2 to 5 micrometers, running at controlled speed along a defined trace length. Using CMM data to estimate surface finish produces unreliable and non-traceable results.
How does machining process choice affect achievable Ra values?
Each machining process has a characteristic Ra range under normal operating conditions. Rough milling typically produces Ra 3.2 to 12.5 µm. Finish turning with a sharp carbide insert can reach Ra 0.8 to 1.6 µm. Cylindrical grinding can achieve Ra 0.2 to 0.8 µm. Honing reaches Ra 0.1 to 0.4 µm. Lapping and superfinishing can reach Ra below 0.1 µm. Specifying a Ra value tighter than the planned process can achieve without secondary operations is a common engineering error that drives unnecessary cost and rework.
We would love your feedback and any insights you would share with others. What perspective would you add?
What surface finish standard should be referenced on drawings for automotive PPAP submissions?
Most automotive Tier 1 customers require surface finish specifications that conform to ISO 1302 for symbol conventions and ISO 4287 or ISO 4288 for parameter definitions and measurement conditions. ASME B46.1 is acceptable for North American customers who specify it, but if the drawing uses ISO surface texture symbols, the measurement and acceptance criteria should reference the corresponding ISO standards. Confirm the governing standard with the customer quality engineer before PPAP submission to avoid acceptance disputes.
Why does specifying a tighter surface finish than necessary increase cost without improving quality?
Tighter surface finish requirements demand slower feed rates, additional finishing passes, secondary operations such as grinding or honing, and more frequent inspection. Each of these adds direct cycle time and cost. If the functional requirement for a non-contact cosmetic face is Ra 3.2 and the drawing specifies Ra 0.4, the machinist must add operations that provide no functional benefit. Over-specified surface finish is one of the most common sources of avoidable cost in precision machined component manufacturing, and it is corrected by applying functional surface finish analysis rather than defaulting to the tightest value the shop can achieve.
If your team is currently reviewing drawings with ambiguous or over-specified surface finish callouts, share your biggest challenge in the comments or reach out directly. Real-world input from engineers and quality professionals shapes better practices across the industry.
Comments