Choosing a Metal Surface Finish for Machined Parts
Why the finish is a design decision, not a cosmetic afterthought
The surface finish on a machined part decides how it resists corrosion, how it wears, how it reads to the eye and how much it costs. Picking one because it "looks industrial" is how parts fail in the field or come back over budget. The right way is to start from the failure mode: is the enemy moisture, abrasion, galvanic corrosion, or just fingerprints on a show part? Once that is clear, the shortlist of finishes gets very small. A finish is a specification like any other - it has a thickness, a process, an area to mask and a cost - and treating it as "whatever the shop defaults to" is how budgets and tolerances both slip at the same time.
Anodising (aluminium)
Anodising grows a controlled oxide layer on aluminium. Type II (decorative) is typically 5-25 micrometres and takes dye; Type III (hardcoat) runs 50 micrometres and up, is harder and more wear-resistant, and is the choice for wear faces and harsh environments. It is electrically insulating and dimensionally grows the part, so masked features and tapped holes need planning. If you want a durable, colour-stable aluminium look, metal surface finishing chosen by function is where the trade-offs are laid out plainly. Specify the type and thickness class, not just "anodise it", or you will get whatever the shop defaults to. A hardcoat layer is also brittle in thin sections, so corners and edges benefit from a generous radius in the design stage.
Powder coating
Powder coating deposits a thicker (roughly 50-100 micrometres) polymer layer that is baked on. It beats liquid paint on chip resistance and coverage of edges, and it hides minor surface marks. The cost is thickness control - it is not for press-fit bores unless they are masked and re-machined. For enclosures and handles it is often the best value, because the film builds a uniform skin that tolerates shop handling. The catch is dimensional: an 80-micrometre coat on both sides of a slot is 160 micrometres of lost clearance, so design the gap before finishing, not after. If you coat a threaded fastener, plan to chase the threads or mask them, because powder will fill the root of the thread and the nut will not start.
Plating: zinc, nickel, chrome
Zinc plating is the cheap corrosion shield for steel fasteners and brackets. Nickel adds wear and a brighter base; chrome (often over nickel) is the bright, hard top layer people recognise on fixtures. Each adds thickness and can affect thread fit, so call out the plating and the class if the part assembles. A threaded rod that is plated without masking will not thread a nut afterwards - that is a specification error, not a plating error. For high-visibility hardware, our notes on PVD coating for shower hardware walk through the same decision for a demanding indoor environment where the finish is also the product. And for the hardware itself, a supplier such as Zimmor Hardware lives in that world of finish-as-product every day.
Passivation and electropolish (stainless)
Passivation removes free iron from the surface of 300-series stainless so the chromium oxide layer can do its job; it is not a coating, just a chemistry step. Electropolish takes material off and smooths the surface, which helps in food, medical and high-purity contexts. Neither changes the colour much - they protect, they do not decorate. If the part sees a sterile or hygienic environment, electropolish is often worth the step because a smoother surface is easier to clean and less likely to harbour residue. The trade is dimensional loss: electropolish removes stock, so hold the pre-finish tolerance accordingly if a feature is critical.
Bead blasting and black oxide
Bead blasting is a texture, not a protection; pair it with a coating if the part sees weather. Black oxide is a thin, decorative, mildly corrosion-resistant finish for steel - fine indoors, not a substitute for plating outdoors. Used alone on an outdoor bracket it will rust at the first humid week. A shower-hardware buyer weighing PVD against the alternatives is already in this world: the question is always service environment first, appearance second. Bead blast before anodise is a common combo - the blast gives a uniform matte key for the oxide - but the blast itself adds no corrosion resistance.
The selection flow: match finish to failure mode
A simple way to choose is to name the dominant threat. Moisture on aluminium that must stay colourful - Type II anodise. Wear on aluminium in a harsh environment - Type III hardcoat. Indoor steel that must not rust in storage - zinc plate. A visible steel fixture that must stay bright - nickel then chrome. A hygienic stainless part - electropolish. An enclosure that takes knocks - powder coat. Start from the threat, not the catalogue, and the choice almost makes itself. The mistakes all come from starting at the other end: "what looks good" leads to a bright finish that fails in six months outdoors.
Environmental and regulatory notes
Some finishing chemistries are restricted or watched in certain markets, and that affects both cost and whether a finish is even available to you. Hexavalent-chrome processes, for example, are heavily limited in many regions, which is why trivalent and PVD options get specified instead. If you sell into a regulated market, ask the shop what chemistry class they use and whether it clears the destination's rules, because a finish that is fine at the factory can be a compliance problem at the border. This is the same discipline as getting the customs documentation set right before shipping - the paperwork and the process both have to match the destination.
How to specify it without overspending
Name the finish, the thickness class and the areas to mask. "Anodise, Type II, black, mask threads" is a buildable instruction; "make it look tough" is not. The material grades and machining tolerances you choose upstream also decide which finishes are even sensible, so decide finish while you still have a chance to change the alloy. A finish that fights the base material is a recurring problem; a finish matched to the alloy is a non-event. And put the finish on the drawing at the feature level, not in a note at the bottom, so the shop cannot "default" the wrong process on one critical face.
Rework, and what it actually costs
The expensive finish is rarely the one on the quote - it is the one you strip and redo. A wrong anodise colour means stripping (which etches the part), re-machining to size, and re-anodising; a wrong powder colour means grit-blasting back to bare metal. Every rework step removes material and adds handling, so a finish chosen carefully the first time is cheaper even when it looks more expensive on paper. The cheapest finish is the one applied once. This is why the selection flow above is worth doing formally: a ten-minute decision prevents a part that comes back as scrap.
| Finish | Typical thickness | Best for | Watch out |
|---|---|---|---|
| Anodise Type II | 5-25 micrometres | Indoor aluminium, colour | Dimension growth on threads |
| Anodise Type III | 50+ micrometres | Wear faces, harsh use | Masking essential; brittle on thin edges |
| Powder coat | 50-100 micrometres | Enclosures, handles | Clears bores unless re-machined |
| Zinc plate | 5-15 micrometres | Steel corrosion shield | Thread fit after plating |
| Nickel / chrome | varies | Bright, hard top layer | Thickness on threads; cost |
| Passivation | none (chemistry) | Stainless hygiene | Not a coating |
FAQ
Will anodising change my dimensions? Yes, slightly - the layer grows outward. Mask threaded holes or plan to chase them.
Which finish for outdoor steel? Zinc or a powder coat over zinc; black oxide alone will not survive weather.
Can I plate a tight-tolerance bore? Only if you mask and re-machine it afterwards, or hold the tolerance to the plated size.
Is powder coat stronger than paint? For chip resistance and edge coverage, yes - but it is thicker, so it is not free on precision fits.
Why specify finish at the feature level? So the shop cannot default the wrong process on one critical face buried in a global note.
Finish and how it meets assembly
The finish is rarely the last thing to think about - it is part of the assembly. A threaded fastener that is anodised grows in the root of the thread; a press fit that is powder-coated loses clearance on both sides; a snap feature that is bead-blasted may no longer snap. Design the mating features to the finished size, not the machined size, and the assembly works the first time. This is why finish belongs in the design review, not the purchasing note: the engineer who drew the fit is the one who can move it to compensate for the coating, and by the time it reaches a buyer the only lever left is hope.
Finish versus tolerance: they interact
A tight surface finish and a tight dimensional tolerance are not the same thing, and chasing both when only one matters wastes money. A hidden internal bore may need a tight diameter but not a polished surface; a visible face may need a smooth finish but not a tight tolerance. State the functional requirement - "seals against fluid" or "slides without galling" - rather than a numeric finish that may be stricter than needed. The tolerance guidance and the finish choice should be decided together, because the cheapest part is the one where each spec matches exactly the function it serves and no more.
Quick reference by base material
| Base material | Usually right | Usually wrong |
|---|---|---|
| Aluminium | Type II/III anodise | Bare, or a steel process |
| Steel (indoor) | Zinc plate, powder coat | Unprotected |
| Stainless | Passivation, electropolish | Paint that hides the grade |
| Show hardware | PVD / nickel-chrome | Anything that rubs through |
Masking is the hidden finishing cost
Masking is the quiet labour in finishing. Every thread, bore or face you do not want coated must be masked by hand before anodise or plating, and masking is slow, fiddly and charged by the operation. Designing the part so the masked areas are few and easy to reach - or so the coating can be machined off after, on a simple feature - turns a costly masking job into a cheap one. A designer who specifies "mask these three holes" without thinking about access can add more finishing cost than the coating itself. The rule: minimise what must be masked, make what must be masked easy to reach, or plan to machine the coating off a simple face rather than mask a complex one. That single habit often matters more to the finishing line than the choice of coating.
A thirty-second finish decision
If you remember one thing: name the enemy first. Moisture on aluminium that must stay colourful - Type II anodise. Wear in a harsh environment - Type III hardcoat. Indoor steel that must not rust - zinc plate. A visible steel fixture that must stay bright - nickel then chrome. A hygienic stainless part - electropolish. An enclosure that takes knocks - powder coat. Start from the threat, write the finish, the thickness class and the masked areas on the drawing, and you have a buildable instruction instead of a hope. The finish is the cheapest insurance on a part when it is chosen before the metal moves, and the most expensive surprise when it is chosen after the part comes back wrong and must be stripped, re-machined and re-finished at your cost. The selection flow in this article exists only to make that thirty-second decision repeatable for every part, not just the one you thought about.
Standards and references. Engineering tolerance; ASME Y14.5; anodizing
评论
发表评论