Surface Finish for Metal Micro Parts: What Ra Values Actually Mean

Surface finish is one of those specifications that tends to receive less attention than it deserves until something goes wrong. The print calls for a finish value, the supplier delivers parts within that value, and the program moves on. At macro scale, this is usually fine. The finish is a quality attribute, the process produces it within tolerance, the conversation ends there.

At micro scale, surface finish stops behaving like a quality attribute and starts behaving like a functional one. The texture of a sub-millimeter component can determine whether the part articulates cleanly, whether it resists bacterial growth on a medical device, whether it integrates with surrounding assemblies, and whether the manufacturing process used to make it is even cost-viable. For engineers specifying or sourcing precision metal parts at micro scale, understanding how surface finish works — and how it gets measured, controlled, and priced — is part of the job.

Why Surface Finish Matters More at Micro Scale

At macro scale, surface texture affects how a part looks, how it feels in the hand, and how cleanly mating surfaces seat against each other. These are real concerns, but they are usually secondary to dimensional accuracy and material properties.

At micro scale, the relationship inverts. The same physics that makes micro-precision manufacturing a different discipline from macro-precision manufacturing — friction scaling differently than mass, surface interactions dominating bulk properties, contact areas becoming a larger fraction of total volume — also makes surface finish disproportionately important. A surface texture that was acceptable at centimeter scale can change how a sub-millimeter part articulates, whether bacteria can colonize it, how it interacts with biological fluids, or whether it produces unexpected wear at contact points.

For medical micro-components in particular, surface roughness affects bacterial colonization, fluid dynamics, biocompatibility, and the assembly behavior of tiny parts that need to mate with other tiny parts. Surface finish at this scale is not aesthetic. It is functional.

The Vocabulary: Lay, Waviness, and Roughness

Surface texture is described at three different scales, and each one matters for different reasons.

Lay describes the largest surface features, typically the directional marks left by the manufacturing process — tool marks from machining, grain from rolling, layer lines from additive manufacturing. Lay is easy to see and easy to specify. It matters for how a part looks and for some directional friction characteristics, but it is rarely the dominant concern at micro scale.

Waviness describes the medium-scale surface variations that are not immediately visible but show up under measurement — gentle undulations across a surface caused by machine vibration, workpiece deflection, or thermal effects during processing. Waviness affects how a part seats against a mating surface and can influence sealing performance, but it is also not typically the binding constraint at micro scale.

Roughness describes the smallest surface features, the texture at the level of individual peaks and valleys. Roughness is what most surface finish specifications actually control, and it is the parameter that matters most for micro-manufacturing applications.

In practice, when an engineer specifies a surface finish for a micro-precision metal part, they are almost always specifying roughness.

How Surface Roughness Gets Measured

Several measurement methods exist, and the differences between them matter when reading specifications across suppliers or comparing data sheets.

Ra (roughness average) is the most commonly specified value. It represents the average distance between the peak heights of the surface and the mean line — essentially, the average deviation from a perfectly flat reference. Lower Ra means smoother. Higher Ra means rougher.

Rz is the average of the five highest peaks and five lowest valleys within the measurement area. Rz captures extremes that Ra averages out, which makes it useful for surfaces where outlier features matter (sealing surfaces, sliding contacts).

RMS (root mean square) is the root mean square average of the height variation, mathematically related to Ra but weighting larger deviations more heavily.

Most micro-manufacturing specifications use Ra, and most data sheets in the industry quote Ra values. Trio Labs, the Haldeman & Frazier principal that produces metal micro-components through their proprietary Resin Infused Powder Lithography (RIPL) process, specifies surface finish in Ra and delivers Ra32 native finish — the same surface quality level as CNC machining.

The Microinches vs. Microns Confusion

This is worth its own section because it produces real and expensive program errors.

Ra values can be expressed in two different unit systems: microinches (µ-in) or micrometers, also called microns (µm). The two are not close in magnitude. One micron equals approximately 39.4 microinches. So an Ra value of 32 µ-in (a fine finish, equivalent to good CNC machining) is fundamentally different from an Ra value of 32 µm (a very rough finish, equivalent to a low-grade casting).

A spec that says "Ra32" without unit specification is ambiguous, and engineers who assume the wrong unit can end up either over-finishing parts that did not need it or accepting surfaces that are nowhere near the intended quality. The conversion: Ra32 µm ≈ Ra1,260 µ-in. The difference is two orders of magnitude.

The practical fix is to specify the unit explicitly in every drawing and quote. When evaluating supplier capability claims, confirm the unit before comparing numbers. A supplier quoting "Ra32" should be asked which unit they mean, and if the answer is unclear, the answer matters more than the supplier admits.

How Process Choice Determines Native Finish

The surface finish a manufacturing process produces natively — without additional finishing operations — depends on the process physics. Each method has a characteristic finish range it can hit without rework.

CNC machining produces Ra16-32 µ-in natively on most metals, with finer finishes achievable through specialized tooling, slower feeds, or polishing operations. This is the reference point most engineers anchor to.

Metal injection molding produces surfaces in roughly the same range, depending on the mold finish and the material.

Most metal additive manufacturing processes — DMLS, SLM, binder jetting, and similar — produce rougher native surfaces, typically in the Ra100-400 µ-in range or worse. To bring these parts down to CNC-equivalent finish, additional operations are required: tumbling, abrasive blasting, polishing, sometimes machining. These operations add time, add cost, and at micro scale introduce their own risks (handling damage, dimensional drift from material removal, edge rounding that affects functional features).

Trio Labs' RIPL process is unusual in the additive manufacturing category because it produces Ra32 µ-in native finish without secondary operations. The mechanism is in the process — the resin-infused powder lithography approach produces denser, more homogeneous parts with finer surface features than conventional powder-bed metal additive methods, which means the as-built part is already at the finish quality CNC machining achieves. No tumbling, no blasting, no follow-up machining required.

For a micro-precision part where surface finish matters functionally, this is the difference between a single-step manufacturing process and a multi-step one. The economic implications compound at volume: secondary finishing operations on micro parts are expensive per unit, particularly when handling fragile small geometries through tumbling or blasting introduces yield loss.

What This Means for Specification

A few practical takeaways for engineers working on metal micro-components.

Specify the surface finish explicitly, in the right units, using the measurement method that matches how it will be inspected. Ra32 µ-in is meaningfully different from Ra32 µm, and from Rz32, and from RMS32. The specification has to be unambiguous.

Match the finish requirement to the application. Not every micro part needs Ra32. Some functional requirements demand it (sliding contacts, sealing surfaces, blood-contacting surfaces). Others can accept rougher finishes. Over-specifying surface finish adds cost without performance benefit; under-specifying it causes functional problems that are expensive to fix later.

Evaluate the supplier's native finish capability separately from their finishing operations capability. A shop that produces rough native parts and finishes them through secondary operations is operationally different from one that produces fine native parts directly. Both can deliver to spec, but the cost structure, lead time, and yield behavior are different.

For programs evaluating micro-precision manufacturing options, the native surface finish question is one of the more useful discriminators between process choices. Trio Labs is a Haldeman & Frazier principal for micro-precision metal parts specifically, and the RIPL process is built around native Ra32 finish at micro scale — which makes it a fit for applications where surface quality matters and the cost of secondary finishing operations does not.

To discuss a micro-precision manufacturing requirement or evaluate Trio Labs through Haldeman & Frazier, contact us.