Surface Finish Is a Functional Specification, Not the Last Step
The Line Item That Causes Assembly Failures
On most drawings, surface finish occupies a single line in the notes block. "Anodise per MIL-A-8625 Type II." Eleven words, added late, rarely reviewed.
Those eleven words decide whether the housing grounds to chassis, whether the bearing bore still accepts its bearing, whether the part survives 500 hours of salt spray, and whether the anodised colour matches the sample the customer approved. Each of those is a functional requirement. None of them is cosmetic.
The recurring pattern in aluminium casting programmes is that finish gets specified by habit — copied from a previous drawing, or chosen because it looked right — and the consequences surface at assembly, months later, when the tooling is cut and the alloy is fixed.

Start With the Alloy, Because It Limits Everything After
A surface finish is a reaction between a process and a substrate. Change the substrate and the same process produces a different result.
High-Copper Casting alloys do not anodise cosmetically. A380 and ADC12 carry 3–4% copper. During anodising, copper-rich intermetallic phases dissolve preferentially rather than converting to oxide, producing a mottled, dark grey film with visible blotching. This is not a plating shop error. It is metallurgy, and no amount of process tuning corrects it.
| Alloy | Cu % | Anodised Appearance | Notes |
| A380 / ADC12 | 3.0–4.0 | Dark grey, mottled | Functional anodise only |
| A360 | ≤0.6 | Reasonably uniform, light grey | Best HPDC choice for anodising |
| A413 | ≤1.0 | Fair, slight variation | Good compromise |
| A356 (gravity, T6) | ≤0.25 | Closest to a clean silver | Preferred where appearance matters |
Silicon has its own effect. Every casting alloy carries 7–13% silicon for fluidity, and silicon does not anodise at all. The particles remain as unconverted grey specks embedded in the oxide layer, which is why even A360 will never match the bright clarity of anodised 6061 extrusion. Buyers who benchmark a cast housing against a machined billet sample are comparing outcomes the process cannot deliver.
The practical rule: if bright cosmetic anodising is a requirement, decide that before the alloy is fixed, and expect to trade some castability for it. If the alloy is already committed to A380, the honest options are Powder Coating, e-coat or wet paint — all of which hide the substrate rather than revealing it.
Conductivity: The Detail That Fails Silently
This distinction causes more field problems than any other finishing decision, and it appears nowhere on most drawings.
Anodising is an insulator. The aluminium oxide layer it grows is dielectric. A Type II film at 15 µm will not pass a continuity test. If your Die Cast housing relies on a bolted joint to carry chassis ground, or on wall continuity for EMI shielding, a fully anodised part breaks that path.
Chemical film conversion coating conducts. Chromate or trivalent-chromium conversion coating (MIL-DTL-5541) forms a thin 0.5–2 µm layer that provides corrosion protection while remaining electrically conductive — typically under 5 milliohms per square inch when correctly applied.
Powder coat, wet paint and e-coat are all insulators, and thick ones.
Selective masking
- — anodise or coat the part, mask the grounding pads so they remain bare or chemical-film treated only
Chemical film overall
- — accept lower corrosion resistance in exchange for full-surface conductivity
Post-coat machining
- — coat everything, then machine the grounding pad back to bare metal, which requires the pad to be a machined feature anyway
Whichever route, it must appear on the drawing as a defined zone with a stated resistance requirement, not as a verbal instruction. "Ground pad to remain uncoated, contact resistance ≤ 5 mΩ" is testable. A note saying "mask as required" is not.
Coating Thickness Consumes Your Tolerances

Every coating adds material. On a flat external face this rarely matters. On a bore, a shaft, a threaded hole or a mating flange, it decides whether the assembly goes together.
Anodising grows both inward and outward. Roughly half the oxide layer penetrates into the substrate and half builds outward. A 25 µm hard anodise therefore adds about 12–13 µm per surface — and on a bore, that applies to both walls, reducing diameter by roughly 25 µm total.
Powder coating does not deposit evenly. Electrostatic attraction concentrates powder on edges and corners while starving internal corners and recesses. Nominal 80 µm can mean 120 µm on an edge and 40 µm in a pocket. For any dimension that must be held after coating, that variation is the real tolerance.
| Process | Typical Build | Grows Into Substrate? | Effect on a Ø20 mm Bore |
| Chemical film | 0.5–2 µm | Negligible | Negligible |
| Anodise Type II | 5–25 µm | ~50% | −5 to −25 µm |
| Hard anodise Type III | 25–100 µm | ~50% | −25 to −100 µm |
| Electroless nickel | 5–50 µm | No | −10 to −100 µm |
| Powder coat | 60–120 µm | No | −120 to −240 µm |
| E-coat | 15–35 µm | No | −30 to −70 µm |
Specify whether the drawing dimension is before or after coating. This single clarification prevents a large share of first-article disputes. Where a bore must hold tolerance after a thick coating, the standard solution is to mask it, coat the rest, then machine or ream the bore last.
Threads deserve particular attention. A powder-coated M6 tapped hole will not accept an M6 screw. Threads must be masked, or tapped after coating, or fitted with an insert.
Masking: The Cost Line Buyers Never Anticipate

Masking is manual labour applied to every single part, every single cycle. It does not benefit from volume in the way casting does.
Typical masking requirements on a die cast housing: silicone plugs in tapped holes, caps on precision bores, high-temperature tape over grounding pads and sealing faces, and hooks or racking points that themselves leave uncoated marks.
Masking is per-part manual time
- — twelve masked features can add more cost than the coating itself
Racking marks are unavoidable
- — the part must hang from something, and that contact point will be uncoated or damaged. Agree the racking location during DFM and place it on a hidden surface
Masking tolerance is coarse
- — a masked boundary holds roughly ±0.5 mm, not ±0.1 mm. If a coated-to-bare transition must be sharp, machine it after coating
A design decision that removes masking cost entirely: locate all grounding pads and precision bores on one face, so a single mask covers them. Scattering them across four faces multiplies handling.
Corrosion Performance: What Salt Spray Hours Actually Mean

Neutral salt spray testing to ASTM B117 is the standard corrosion benchmark, and it is widely misread.
What it measures: relative performance between coating systems under one accelerated, continuous condition.
What it does not measure: service life. There is no reliable conversion from salt spray hours to years in the field. A coating passing 1,000 hours is better than one passing 500, but neither figure predicts calendar life in a specific environment.
Indicative performance to first white corrosion:
| System | Typical Hours (ASTM B117) |
| Bare as-cast aluminium | 24–96 |
| Chemical film (Cr³⁺) | 168 |
| Anodise Type II, sealed | 336–500 |
| Powder coat over conversion pre-treatment | 500–1,000 |
| E-coat + powder topcoat (duplex) | 1,000+ |
Pre-treatment matters more than topcoat. Powder applied directly to bare cast aluminium fails at the interface regardless of film thickness. The conversion coating underneath is what creates adhesion and prevents undercreep. When comparing quotes, confirm the pre-treatment stages — a supplier omitting conversion coating will always be cheaper and always be worse.
Edges and recesses fail first. Salt spray panels are flat; real castings have edges, fins and pockets where powder thins. E-coat exists precisely because it deposits uniformly into recesses that electrostatic powder cannot reach, which is why duplex e-coat plus powder is standard for automotive underbody components.
Sealing is not optional on anodise. Unsealed anodic film is porous and offers a fraction of the protection. Confirm the sealing method — hot water, nickel acetate or dichromate — is specified, not assumed.
Matching Finish to Application
| Requirement | Recommended Finish | Why |
| Electrical grounding / EMI continuity | Chemical film (Cr³⁺) | Only conductive corrosion coating |
| Outdoor, colour required | Powder coat over conversion | Best colour range plus UV and impact resistance |
| Complex geometry, deep recesses | E-coat, optionally + powder | Uniform deposition into cavities |
| Wear surface, sliding contact | Hard anodise Type III | 1,000+ HV surface, but check clearance loss |
| Marine / salt exposure | A360 alloy + duplex system | Alloy and coating must both be selected for it |
| Cosmetic bright metallic | Plating (Ni/Cr) — better still on Zinc Casting | Cast aluminium cannot anodise bright |
| Thermal dissipation housing | Black anodise or thin black powder | Emissivity gain; thick coatings insulate |
| Paint adhesion base only | Chemical film | Designed as a pre-treatment |
For decorative plated parts, zinc beats aluminium. ZAMAK's as-cast surface is dense, smooth and plating-ready. Aluminium castings require zincate pre-treatment before plating and rarely achieve the same cosmetic result. If a small component must look like polished chrome, zinc casting is the correct process choice, not an aluminium casting with more finishing effort.
Thermal parts need care. A black anodised or thin black powder finish increases radiative emissivity and can improve heat rejection. A thick powder coat does the opposite — it insulates. For Heat Sinks, specify maximum coating thickness, not just colour.
Getting It Onto the Drawing
Most finishing disputes trace to under-specified drawings. A complete callout answers six questions:
Process and standard
- — "Anodise per MIL-A-8625 Type II Class 2" rather than "anodise"
Thickness range
- — minimum and maximum, since maximum controls fit
Colour and reference
- — a signed physical boundary sample, not a Pantone code, because cast substrate shifts colour
Masked zones
- — identified on a drawing view with tolerances on the boundary
Functional test
- — contact resistance for grounding zones, adhesion class per ASTM D3359, salt spray hours per ASTM B117
Dimension timing
- — whether stated dimensions apply before or after coating
Also worth resolving early: whether finishing is in-house or subcontracted. Most foundries subcontract plating and anodising. That is normal and not a concern in itself, but it affects lead time, adds a transport leg where parts can be damaged, and creates a second party in any quality claim. Ask which metal casting service operations happen on site and which do not — and confirm who owns the non-conformance if a coating fails.
For hybrid assemblies combining cast bodies with custom metal fabrication brackets, one further trap: cast aluminium and welded 6061 accept coatings differently. Anodising a mixed assembly produces two visibly different shades on the same product. Powder coat or e-coat the assembly instead, or finish the components separately before joining.
Frequently Asked Questions
Q1: Which surface finish should I use on aluminium die castings?
Work backwards from function. If the part must carry electrical ground or maintain EMI continuity, chemical film conversion coating is the only conductive corrosion-resistant option. For outdoor use in a specific colour, powder coat over a conversion pre-treatment. For complex geometry with deep recesses, e-coat deposits where electrostatic powder cannot reach. For sliding wear surfaces, hard anodise Type III — but check the clearance it consumes. Bright cosmetic metallic appearance is not achievable on cast aluminium and should be routed to plated zinc or a machined component.
Q2: Is anodising electrically conductive?
No. Anodising grows an aluminium oxide layer that is dielectric, so an anodised surface will not pass a continuity test. Chemical film conversion coating (MIL-DTL-5541) is the conductive alternative, typically below 5 milliohms per square inch. Where a part needs both corrosion protection and grounding, mask the grounding pads during anodising, treat them with chemical film only, or machine them back to bare metal after coating. Specify the requirement as a defined zone with a stated resistance limit on the drawing.
Q3: How much clearance does anodising add to a bore?
Anodising grows roughly half into the substrate and half outward, so a 25 µm coating adds about 12–13 µm per surface. In a bore, both walls coat, reducing diameter by around 25 µm for Type II at 25 µm, and up to 100 µm for a heavy Type III hard anodise. Powder coating is far more severe at 60–120 µm nominal, and it builds unevenly, concentrating on edges. Always state on the drawing whether dimensions apply before or after coating, and mask precision bores, then ream them last.
Q4: Why does my anodised die casting look grey and blotchy?
Alloy chemistry. A380 and ADC12 contain 3–4% copper, whose intermetallic phases dissolve rather than convert during anodising, producing a mottled dark grey film. Silicon at 7–13% compounds this because it does not anodise at all and remains as grey specks in the oxide. This is inherent to cast alloys, not a plating fault. For a more uniform anodised appearance, specify A360 or A413 during design; for anything approaching bright silver, use gravity-cast A356 or accept powder coating instead.
Q5: What do salt spray hours actually tell me?
They allow relative comparison between coating systems under one accelerated test condition per ASTM B117. They do not convert to service life in years, and no reliable correlation exists. More decisive than the headline figure is the pre-treatment: powder applied directly to bare cast aluminium fails at the interface no matter how thick it is, so confirm that a conversion coating stage precedes it. Also remember that test panels are flat while real castings have edges and recesses where coating thins first.
Q6: How much does masking add to the part cost?
More than most buyers expect, because masking is manual labour repeated on every part and does not scale down with volume. A housing with a dozen tapped holes, two precision bores and a grounding pad can incur more masking cost than coating cost. Reduce it at design stage by grouping all features requiring masking onto a single face so one operation covers them, and by agreeing racking point locations during DFM so the inevitable uncoated contact mark sits on a hidden surface.
Q7: Should the foundry handle finishing, or should I manage it separately?
Most foundries subcontract anodising and plating, which is normal. The questions that matter are which operations occur on site, how long the additional transport leg adds to lead time, and — most importantly — who owns the non-conformance when a coating fails. A single supplier accountable for casting through to finished part removes the situation where the foundry blames the coater and the coater blames the substrate. Get that accountability stated in the purchase order rather than assumed.

