Anodizing is not a paint that sits on top of the metal — it is a ceramic layer that grows out of the aluminum surface itself, and roughly half of its thickness builds outward while the other half penetrates inward. The practical consequence is that every anodized surface ends up slightly larger than the surface you machined, and if you cut the part to final print size and then anodize it, it will no longer be at final print size. A bore that was dead-on becomes tight, a threaded hole that was class-accurate becomes stiff, and a press-fit diameter that was sized for a slip becomes an interference. The fix is not to anodize thinner — it is to machine the part with the coating already budgeted in. This guide breaks down how anodizing changes dimensions, how much to leave, and how to size threads, bores, and fits so the part assembles cleanly after coating.
How anodizing changes a dimension
An anodic coating grows by converting the aluminum surface into aluminum oxide, and that oxide occupies more volume than the metal it replaces. The growth is not one-sided: depending on the process, somewhere between a third and a half of the coating builds out from the original surface, while the rest grows into the metal. The outward growth is the part that matters for fit, because it is what changes the outside dimensions of the part. A bore in an anodized part shrinks in diameter by the amount of coating that built inward, and an external diameter grows by the amount that built outward. On a shaft-and-bore pair, both halves move toward each other, and the clearance closes by the sum of the two coatings.
The thickness you get depends on the type of anodizing. A cosmetic or "Type II" sulfuric acid anodize used for color and corrosion resistance is thin — on the order of 5 to 15 microns per surface. A hardcoat, or "Type III," anodize built for wear resistance is far thicker, commonly 25 to 50 microns and sometimes more. The dimensional impact scales directly with that thickness: a thin cosmetic coat is a tolerance headache on a tight bore, while a thick hardcoat is a design input that has to be planned from the first cut. The rule of thumb that keeps most shops out of trouble is to budget roughly half the specified coating thickness as the change on each surface — enough to be safe without over-cutting.
Threads — the first place anodizing bites
Threaded holes are where anodizing most often turns a good part into a scrap one, because a thread has almost no clearance to spare. A standard tapped hole has a small gap between the internal and external thread forms, and a coating that fills even a fraction of that gap can turn a finger-tight screw into one that seizes halfway in. The coating builds on both flanks and on the crest, so the effective minor diameter and the thread flank clearance both close up. The same logic applies in reverse to an externally threaded feature that gets anodized: the male thread grows slightly, and a part that fit its mating nut before coating will not start after it.
There are two practical ways to handle this. The first is to machine the thread slightly oversize before anodizing — using a slightly larger tap or a controlled oversize cut so that the coating brings the thread back to the intended class of fit. The second is to mask the thread entirely so no coating lands on it, which is common on tapped holes that must stay conductive or that mate with a fastener torqued to a precise value. Masking costs more per part and adds a step, so it is reserved for threads that cannot tolerate any coating — a grounding boss, a bearing shoulder, or a thread that must remain electrically functional. For ordinary assembly threads, the oversize-machining route is usually cheaper and just as reliable.
Bores, press fits, and the parts that must still slide
Bores and press-fit diameters are the second place the coating shows up, and they are often more expensive to fix because they carry a tolerance that the anodize line does not naturally hold. A reamed bore sized for a bearing press fit will end up too small after hardcoat, and the bearing will not seat — or worse, it will seat with enough force that the coating cracks. A shaft sized for a running clearance will bind. The planning rule is the same as for threads: machine the bore oversize by the expected coating buildup so the finished, coated dimension lands back in tolerance. For a hardcoat on a bearing bore, that means a deliberate, measured oversize on the order of the coating's inward growth, verified with a go/no-go after coating rather than before.
Where this gets tricky is that the coating thickness on a bore is not perfectly uniform — it varies with how well the electrolyte flows into the hole, how long the part is, and how the rack and contact points are arranged. A deep blind bore anodizes thinner at the bottom than at the mouth because the electrolyte spends more time and carries more current near the opening. That is why the safest approach on a critical bore is to leave it undersized for machining and re-work it after anodizing — either a light hone or a controlled finish pass on the coated surface — so the final dimension is set on the coated part rather than predicted through the coating. It adds a step, but on a bore that must hold a press fit it is the only way to be certain.
| Feature | What anodizing does | Planning move before coating |
|---|---|---|
| External diameter | Grows by the outward coating buildup | Machine undersize by roughly half the specified thickness |
| Internal bore | Shrinks by the inward coating buildup | Machine oversize, or re-work (hone) after coating |
| Internal thread | Flanks and crest close up, screw seizes | Oversize tap before coating, or mask the thread |
| External thread | Male thread grows, will not start | Cut undersize before coating, or mask |
| Press / bearing fit | Interference increases, coating can crack | Budget coating into the fit, verify go/no-go after coating |
| Deep blind bore | Coats thinner at bottom than mouth | Account for non-uniform buildup, re-work critical bores |
Masking — where the coating is not wanted
Masking is the deliberate choice to keep anodizing off a specific surface, and it matters wherever the coating would cause a functional problem rather than a cosmetic one. The classic candidates are threaded holes that must stay conductive, bearing seats and press bores that are sized to a finished tolerance, flat sealing surfaces where the coating would disrupt a gasket or an O-ring, and any surface that must remain bare aluminum for electrical grounding or for a subsequent operation like bonding or welding. Masking is done with plugs, caps, tape, or a resilient coating that the anodize line removes afterward, and it always costs money and cycle time, so it is applied selectively, not as a blanket.
The trade-off to understand is that masking protects the dimension but leaves the surface uncoated, which means it does not get the corrosion resistance or the hard wear skin that the rest of the part gets. On a part that lives in a wet or corrosive environment, a masked thread becomes the weak point. The decision is therefore a balance: mask the surfaces where a coating would break the function, and accept that those surfaces trade the coating's protection for their dimensional and electrical integrity. A good drawing calls this out explicitly — the masking areas, the threads to leave bare, and the surfaces that must hold a coated tolerance — rather than leaving the shop to guess which surfaces matter.
Sizing the allowance into the print
The most reliable way to keep anodized parts assembling cleanly is to build the coating allowance into the drawing before the first chip is cut, so the machinist is not reverse-engineering it from the finish spec at the end. That means the print states the final, coated dimension as the requirement, and the machining dimension as the undersize or oversize value that accounts for the coating. A bore that must finish at 12.000 mm after a 25-micron hardcoat, for example, gets machined oversize by the inward growth so the coating brings it back to 12.000. The key is to specify the coating type and thickness in the same place the tolerance is specified, because the machinist cannot budget for a coating he does not know about.
There is also a sequencing question: whether to dimension the part for machine-then-anodize, or to plan a post-coat operation. The simplest parts are sized to be machined to a coating-adjusted dimension and anodized last, with no further cutting. The more demanding parts — tight bearing bores, high-precision fits, flatness-critical sealing faces — are often better served by machining, anodizing, and then a light finishing operation on the coated surface to bring the final dimension home. The difference is confidence: a coating-adjusted dimension is a prediction, while a post-coat finish pass is a measurement. Pick the prediction for the features that can tolerate a few microns of uncertainty, and the measurement for the ones that cannot.
- State the coating type and thickness on the print — the machinist cannot budget for an anodize he does not know is coming.
- Budget roughly half the coating thickness per surface — machine oversize on bores, undersize on external diameters.
- Oversize internal threads or mask them — a coating on both flanks and the crest seizes a standard screw.
- Verify go/no-go after coating, not before — the finished, coated dimension is the one that matters.
- Mask threads and faces that must stay conductive or bare — grounding bosses, sealing surfaces, weld zones.
- Re-work critical bores after anodizing — a post-coat hone sets the dimension on the coated part instead of predicting through it.
- Account for non-uniform buildup in deep blind bores — the mouth coats thicker than the bottom.
"Anodizing does not change a dimension by accident. It grows the surface by a predictable amount, and the part that no longer fits after coating was simply machined as if the coating would not exist. The whole discipline is to machine the part with the coating already in the math — oversize the bores and threads, undersize the shafts and diameters, mask the surfaces that must stay bare, and verify the fit on the coated part rather than the bare one. When the drawing states the finish and the tolerance in the same breath, the part comes back from the anodize line the size it was always meant to be. When it does not, the machinist is left guessing — and a guessed allowance is a seized screw waiting to happen."
Anodizing and fit tolerance are not competing requirements; they are two halves of the same dimension that have to be planned together. Know the coating type and thickness up front, budget its growth into every bore, thread, and diameter, mask the surfaces that must stay bare, and verify the fit on the coated part. Do that and the part comes back from anodizing the size the drawing called for — not because the coating happened to come out thin, but because the allowance was in the math from the first cut.
If you are machining an aluminum part that has to be anodized and still assemble — threaded holes, bearing bores, press fits, or sealing faces — send the drawing over and we will plan the coating allowance into the process, from the machining dimensions to the anodizing spec and the post-coat checks that keep the part assembling cleanly.