Why surface texture specs cause surprises in RFQs
Surface texture notes look simple on a drawing, but they can quietly drive cost, lead time, and risk. A single line like “Ra 32” may seem like a minor quality preference, yet it can force different tooling, extra finishing passes, alternative processes, or more inspection time. The surprise often shows up late, either as a quote that is higher than expected, or as a quality debate when parts are delivered and measured.
A better approach is to treat surface requirements the same way you treat tolerances: specify what function needs, verify what is realistically measurable, and understand what the process naturally produces. The goal of this guide is to help U.S. engineers and buyers write surface requirements that are clear, testable, and aligned with how parts are actually machined and inspected.
Quick perspective: A roughness value is not a full description of how a surface behaves. Roughness is one slice of surface texture, and texture also includes longer-wave waviness and the directionality of the machining pattern [3].
Table of Contents
Surface finish CNC machining: Ra, Rz, and what the numbers really mean
Ra is the arithmetic mean of the absolute deviations of the roughness profile from its mean line over a defined evaluation length [7], [16]. Put plainly, Ra tells you the average height of the small peaks and valleys after filtering out longer-wave form and waviness. It is widely used because it is simple, stable, and works reasonably well for many functional surfaces, from brackets to general housings.
Rz is a height parameter that emphasizes extremes more than Ra. In common usage, it relates to peak-to-valley behavior over defined sampling rules, which makes it more sensitive to deep scratches, torn material, or isolated valleys that Ra might average away [7], [16]. That sensitivity is useful in some applications, but it also means Rz can create confusion if the drawing does not align on the definition and measurement rules [2]. When in doubt, specify the parameter, the measurement conditions, and the final state of the surface you care about.
What “Ra 32” actually means in U.S. practice: It typically refers to a maximum Ra of 32 microinches, which is about 0.8 micrometers. A value near this range is common for many finish-machined surfaces, but whether it is achievable without secondary finishing depends on material, toolpath, and stability [15].
A fast unit map for common callouts
Most U.S. prints use microinches (µin). Many global suppliers use micrometers (µm). The rough conversions below are useful for RFQs.
| Common callout | Approx. in µm | Typical intent |
|---|---|---|
| Ra 125 µin | 3.2 µm | standard machined finish |
| Ra 63 µin | 1.6 µm | smooth machined finish, often with careful parameters |
| Ra 32 µin | 0.8 µm | fine machined finish, may require optimized finishing strategy |
| Ra 16 µin | 0.4 µm | very fine, often pushes toward grinding, honing, or polishing depending on geometry |
These are not guarantees. They are planning anchors that should be validated against the specific feature, material, and inspection method [15].
Ra, Rz, Rq, and the parameter choices that matter most
Ra is not the only option you will see. Rq is the root-mean-square roughness, which weights larger deviations more strongly than Ra because the deviations are squared before averaging [7]. Rt is the total height of the roughness profile over the evaluation length, which can be very sensitive to a single defect [7]. Some applications use these to protect against one-off spikes that could cause leakage, fatigue initiation, or friction changes.
Practical selection: If you are buying a part with a sealing surface, a surface that mates under load, or a surface that slides, you usually want a parameter set that reflects both average behavior and risk of isolated extremes. For many general surfaces, Ra is enough. For high-risk interfaces, adding an extreme-sensitive parameter can reduce ambiguity, but only if the print also defines measurement rules [7], [5].
Parameter comparison table (for buyers and print reviewers)
| Parameter | What it represents | Strength | Common pitfall |
|---|---|---|---|
| Ra | average absolute deviation | stable, easy to compare | can hide rare but important defects |
| Rq | RMS deviation | highlights larger peaks/valleys | can be specified without realizing it is stricter than Ra |
| Rz | peak-to-valley based height metric | catches deep scratches and torn material | definition varies across standards if not specified [2] |
| Rt | total height over evaluation length | sensitive to one big defect | can reject parts for a single localized anomaly |
If you need Rz, treat it as a precision requirement and fully define how it will be assessed, including cutoff and evaluation length [7], [6].
How roughness is measured, and why the settings matter
Most production roughness checks use a contact stylus that drags a diamond tip across a surface and records height versus distance. The instrument then applies filters to separate longer-wave form and waviness from shorter-wave roughness [5]. The result depends on the traverse length, evaluation length, sampling length, and the long-wavelength cutoff used to define what counts as roughness versus waviness [5], [6].
A national metrology lab report highlights that roughness measurement conditions commonly include a long-wavelength cutoff (λc), a short-wavelength cutoff (λs), and a specified stylus tip radius, because those settings change what the instrument can resolve and what it filters out [6]. If you do not define measurement conditions, two competent inspectors can measure the same part and report different values while both are acting correctly.
Cutoff, sampling length, evaluation length (plain-language definitions)
Cutoff (λc): A filter setting that decides what wavelengths count as roughness versus waviness. A commonly used λc in roughness measurements is 0.8 mm, but it is not universal [6], [7].
Sampling length: The base segment over which many roughness parameters are computed and then combined by the instrument rules [7], [5].
Evaluation length: The total length used for the calculation after end effects are discarded; it usually spans multiple sampling lengths [5].
Measurement reality checklist (include in RFQs when finish is critical)
- Surface measured direction relative to lay (parallel, perpendicular, or specified direction) [4]
- Parameter(s) required (Ra, Rz, Rq, etc.) [7]
- Cutoff and evaluation length (or reference to the applicable rules) [6], [5]
- Contact stylus versus optical method (if optical is required or prohibited) [7]
- Final state of the surface (as-machined, after blasting, after anodize, after plating) [1]
When suppliers ask “How will you measure it?”, they are trying to avoid disputes and ensure the quote reflects real inspection requirements.
What drives finish in turning operations
Turning finish is heavily shaped by feed per revolution, tool nose radius, tool geometry, and stability. A simple planning model treats the surface as a series of cusps formed by the tool nose as it advances. That model is useful because it shows why feed changes can dominate finish, even when speed stays the same [12].
Stability matters as much as math. Chatter, built-up edge, and tool wear can introduce tearing and random defects that raise Rz and sometimes Ra. Tooling guidance notes that insert style and edge preparation influence finish and chip control, and that wiper-style geometries can maintain surface quality at higher feeds in appropriate conditions [10].
Compact text chart: turning levers and their typical effect
Feed per rev: higher feed usually increases roughness height
Nose radius: larger radius can improve finish but may increase cutting forces and chatter risk
Tool wear: dull edges raise tearing and can spike Rz
Built-up edge: more likely in some alloys, causes roughness instability
Rigidity: better support reduces vibration-related defects
Mini-summary: If the print calls for a fine finish on a turned diameter, the fastest path is often a stable finishing pass with the right insert geometry and a feed matched to the tool radius, rather than simply slowing everything down [10], [12].
What drives finish in milling operations
Milling finish is often governed by cutter runout, tool sharpness, feed per tooth, and the effective scallop height left by the stepover strategy. Even with a sharp tool, large stepovers on finishing passes leave visible cusps, especially on 3D contoured surfaces. Reducing stepover reduces scallop height but increases cycle time because more passes are needed [13].
Runout is a frequent hidden culprit. If one tooth cuts more than others due to runout, the surface can show periodic marks and roughness spikes. Toolholder quality, balancing, and tool length all influence this. For fine finishes, the best results come from a controlled finishing strategy: appropriate tool selection, minimal runout, stable workholding, and a stepover matched to the surface requirements.
Table: milling finish risks and mitigations
| Issue | What it looks like | Common cause | Mitigation direction |
|---|---|---|---|
| Scallops | uniform ridges | stepover too large | reduce stepover, use appropriate cutter radius [13] |
| Tooth marks | periodic pattern | runout, imbalance | improve toolholding, shorten stickout |
| Tearing | smeared surface | dull tool, wrong chipload | sharp tool, correct chip thickness |
| Chatter | wavy pattern | rigidity limits | adjust strategy, support part, change engagement |
Mini-summary: For milling, the finish you see is often the toolpath you chose. When a drawing calls out a fine surface on a milled face, the quote needs to reflect stepover and toolpath strategy, not just a generic “finish pass.”
Material effects: why the same Ra is easier in some alloys than others
Material behavior can shift finish outcomes dramatically. Some alloys are prone to built-up edge, which can cause tearing and unstable roughness. Others harden during cutting, which can degrade surface quality if the tool is not sharp and the parameters are not tuned. Softer materials can smear, while harder materials can show brittle micro-chipping if the edge or coating is not appropriate.
Practical guidance: If you are specifying a very fine finish, include the material grade on the RFQ and note any heat treatment state. A finish callout that is easy in one alloy can be expensive in another due to tool wear, instability risk, and the need for secondary finishing [15].
Material notes that often matter in RFQs
Aluminum: can achieve fine finishes, but some grades can smear; tool sharpness is critical.
Stainless steels: can work-harden; finish may vary with tool wear and coolant control.
Tool steels: heat treated conditions can require different strategies; fine finishes may move toward grinding.
Titanium: can generate heat and chatter sensitivity; stable setup and tool choice matter.
Engineering plastics: surface texture can change with cutting heat; burrs and fuzzing can affect perceived finish.
When finish is function-critical, a short supplier discussion about material behavior can prevent both over-specification and underperformance.
Secondary processes and coatings: how final state changes the number
Many parts are not used in the as-machined condition. Grinding, honing, lapping, polishing, and burnishing can reduce roughness. Bead blasting can change appearance and tactile feel, but it can also alter roughness values in ways that do not align with a simple Ra target. Coatings and conversion layers can add thickness, change microtexture, and affect how a surface reads under a stylus or optical method [1].
Key rule: Specify finish relative to the condition that matters in function. If the part will be anodized or plated, clarify whether the requirement is pre-coating, post-coating, or both. If a sealing surface must meet a value after coating, the process plan and inspection plan change.
Good practice matrix (before and after)
| Process step | Surface appearance | Roughness impact (typical) | RFQ note to include |
|---|---|---|---|
| As-machined turning | directional tool marks | predictable, feed-driven | specify lay direction if it matters [4] |
| As-machined milling | scallops, tooth marks | toolpath-driven | identify the functional face |
| Grinding | uniform fine lines | often lowers Ra | clarify whether grinding is allowed |
| Polishing | mirror-like | can lower Ra but may alter edges | define edge condition and any rounding limits |
| Bead blasting | matte, uniform | can increase or change texture | specify if cosmetic only or functional requirement |
| Coating | altered microtexture | can change measurement response | specify post-process measurement condition [1] |
How to specify surface requirements on prints without overpaying
Over-specifying finish is one of the most common ways to inflate machining cost without improving function. A blanket note that applies a tight Ra value to every surface forces conservative strategies everywhere, even where the surface does not matter. The most cost-effective practice is location-based specification: assign the tight requirement only to functional surfaces, and allow a looser default elsewhere.
Surface texture symbols can also include direction of lay, which matters for sealing, friction, and wear behavior [4]. If the direction matters, specify it. If it does not, avoid adding unnecessary constraints.
Good versus risky callouts (examples)
| Better callout style | Why it helps | Risky callout style | Why it causes trouble |
|---|---|---|---|
| “Functional face A: Ra 32 µin, measure perpendicular to lay” | ties requirement to a surface and method | “All surfaces Ra 32” | forces cost everywhere, invites disputes |
| “Seal OD: Ra 16 µin after final process” | defines final condition | “Ra 16” with no context | unclear measurement, unclear process |
| “Cosmetic faces: uniform texture, blast allowed” | separates cosmetic from functional | “No tool marks” | subjective and hard to inspect |
Inline note: If you require Rz, add enough definition so inspection is repeatable. If you do not, use Ra and keep the callout focused on the critical surfaces [7], [6].
Cost and lead-time drivers tied to surface requirements
Surface requirements influence cost through cycle time, scrap risk, inspection time, and the need for secondary processes. The cost jumps are rarely linear. Going from a standard machined finish to a moderately smooth finish may only require a better finishing pass. Going from moderately smooth to very fine can force grinding, polishing, or extremely conservative parameters, especially on difficult geometries [15].
Table: what typically increases quote cost for finish
| Cost driver | Why it increases cost | What you can do as a buyer |
|---|---|---|
| Tight finish across entire part | more finishing time everywhere | apply tight finish only where functional |
| Very fine finish on milled 3D surface | reduced stepover increases cycle time | allow a functional limit, not cosmetic perfection |
| Finish after coating | requires post-process inspection and control | specify final condition clearly, avoid ambiguity |
| High inspection burden | roughness checks take time, can be position-sensitive | specify what must be checked and where |
| Geometry that is hard to access | special probes or fixtures may be needed | discuss measurement access early |
Mini-summary: The cheapest surface requirement is the one you do not need. The best surface requirement is the one that directly protects function and is easy to inspect.
RFQ-ready checklist and a practical spec template
If you want quotes that are accurate and comparable, include surface data the same way you include tolerances and material. Use the checklist below as an RFQ attachment or as fields in your RFQ form.
RFQ surface requirement checklist:
- List the functional surfaces and their roughness parameter(s)
- Provide units (µin or µm) and confirm the target is a maximum unless stated otherwise
- Clarify measurement direction relative to lay when needed
- Clarify final condition (as-machined, after blasting, after anodize, after plating)
- State any prohibited processes (example: no blasting on seal faces)
- Confirm measurement approach if it is constrained (contact stylus, optical allowed, etc.)
- Provide a sample part or functional description when the surface is performance-critical
Spec template (copy into notes):
Functional surface(s): [identify feature or datum reference]
Requirement: [parameter and value]
Units: [µin or µm]
Measurement direction: [relative to lay]
Final condition: [as-machined or post-process]
Exceptions: [cosmetic-only surfaces, blast allowed, etc.]
CTA placement (mid-article): Offer “Surface Finish Spec Checklist for RFQs” as a downloadable one-page attachment that buyers can reuse across suppliers.
FAQ
What does Ra 32 mean? Ra 32 usually means the arithmetic average roughness must be 32 microinches (about 0.8 µm) or better, measured under defined filtering and evaluation settings [7], [5]. On many parts it is achievable with a stable finishing strategy, but on some geometries or materials it may require additional time or secondary finishing [15].
Ra vs Rz, which should I specify? Ra is a stable average metric and works for many surfaces. Rz is more sensitive to extremes and may better reflect risk of deep valleys or scratches, but it can be easier to misunderstand if measurement rules are not explicit [2], [7]. If you choose Rz, define measurement conditions so results are repeatable [6].
What is “lay” and why should I care? Lay is the predominant direction of the surface pattern caused by the manufacturing process [5], [4]. It can affect sealing direction, friction, and wear. If the direction matters, specify it; if it does not, avoid adding constraints.
Can milling hit Ra 16? Sometimes, but it depends on material, cutter selection, runout, and toolpath strategy. On flat surfaces it may be possible with careful finishing, but on complex 3D surfaces it can require small stepovers that increase cycle time, or secondary finishing [13], [15].
How is roughness measured in production? Many shops use a contact stylus instrument with defined cutoffs and evaluation lengths, then compute parameters like Ra, Rq, and Rz from the filtered profile [5], [6], [7]. Results can vary if the measurement direction, cutoff, or evaluation length changes, which is why clear measurement rules matter.
When does finish affect sealing or friction? Finish often matters when the surface is part of a sealing interface, a bearing surface, a sliding contact, or a fatigue-sensitive feature. In these cases, specifying only an average metric can be incomplete, because isolated defects may dominate performance [7], [16].
Key Takeaways
- Ra is an average metric, while Rz is more sensitive to extremes; choose parameters that match function and inspection reality.
- Measurement settings such as cutoff and evaluation length can change results, so critical finishes should include measurement conditions.
- Apply tight finish requirements only where they protect function, and clarify final condition after coatings or secondary processes.
- Use an RFQ checklist so suppliers quote the same requirement, with fewer assumptions and fewer downstream disputes.
- Want a quick surface finish sanity check before you release an RFQ, contact Progressive Turnings for a print and spec review
References
Standards and terminology
[1] “ISO 21920-1:2021 Geometrical product specifications (GPS) – Surface texture: Profile – Part 1: Indication of surface texture,” International Organization for Standardization, 2021 [PDF].
[2] “ISO 4287:1997 Geometrical Product Specifications (GPS) – Surface texture: Profile method – Terms, definitions and surface texture parameters,” International Organization for Standardization, accessed 2026 (page notes replacement by ISO 21920-2).
[3] “B46.1 – Surface Texture (Surface Roughness, Waviness, and Lay),” ASME, publish date listed as 2020 for B46.1-2019.
[4] “ISO 1302:2002 Technical drawings – Method of indicating surface texture,” International Organization for Standardization, 2002 [PDF].
Measurement and cutoffs
[5] “B46.1 – 2019 Poster: Surface Roughness, Waviness, and Lay Basic Terms,” ASME, May 21, 2022 [PDF].
[6] Vorburger, T. et al., “Appendix A for web 18mar14: Surface roughness and step height calibrations,” National Institute of Standards and Technology, 2014 [PDF].
[7] “Quick Guide to Surface Roughness Measurement,” Mitutoyo, 2017 [PDF].
[8] “Cutoff (Cutoff Value) – Surface Roughness Terminology,” KEYENCE, accessed 2026.
Machining guidance and tools
[9] “How to achieve good component quality in turning,” accessed 2026.
[10] “Turning Handbook,” Sandvik Coromant, 2019 [PDF].
[11] “Surface Finish Calculator,” Kennametal, accessed 2026.
[12] “Milling Step-over Distance Calculator,” CustomPartNet, accessed 2026.
Guides and typical ranges
[13] “Exploring Surface Texture,” University of Southampton, 2014 [PDF].
[14] “Roughness measurement poster (parameters and definitions),” Jenoptik, 2024 [PDF].
[15] “Arithmetical Mean Height (Ra, Pa, Wa) and related parameters,” KEYENCE, accessed 2026.
[16] “The profile standards ISO 21920: parameters and definitions,” Digital Surf Surface Metrology Guide, accessed 2026.