Guide
How much does anodizing change part dimensions for Type II and Type III?
Anodizing is an electrochemical conversion process that turns the outer skin of an aluminum part into aluminum oxide, so the coating is partly grown out of the old surface and partly eaten into it. That split is why an anodized part does not simply get bigger by the coating thickness, and why a bore, a shaft and a thread each need a different machining target before the part goes into the tank.
By Saola Manufacturing engineering team · Published · Updated
This guide is written for mechanical engineers who have to put a machining size and a coated size on the same drawing, and for buyers who need to know which of those two sizes a supplier will be held to. Every figure in the tables is computed from the design rules in MIL-A-8625F, the anodic-coating specification most US drawings still cite, so treat them as a starting allowance to confirm on first articles.
On this page
How much does anodizing add to part dimensions?
Per MIL-A-8625F paragraph 6.10.1, each coated surface grows by about one half of the coating thickness, so a diameter or a wall measured across two coated faces changes by roughly one full coating thickness. The other half of the film forms below the original surface, in metal that was consumed to make the oxide.
The rule is simple to apply once you know which way each feature moves. An external feature, such as a shaft, a boss or the width of a tab, gets larger. An internal feature, such as a bore, a slot or the gap between two walls, gets smaller. A step between two coated faces pointing the same way barely moves, while a distance from a coated face to a masked face moves by one surface’s growth.
The specification uses its own worked example: per MIL-A-8625F paragraph 6.10.1, a 0.004 inch hard coat on a close-tolerance part calls for a pre-machining allowance of 0.002 inch on each surface. The same paragraph warns that build-up on close fits, especially with Type III, can turn a slip fit into an interference fit at assembly.
How does the dimensional change differ between Type II and Type III?
Type III hard coat is usually several times thicker than Type II, so it moves dimensions several times as much and with a wider spread. Type II is often thin enough to ignore on ordinary tolerances, while Type III almost always needs its own machining dimension.
Per MIL-A-8625F Table IV, the typical thickness range is 0.00007 to 0.0010 inch for Type II and 0.0005 to 0.0045 inch for Type III. For Type III, per MIL-A-8625F paragraph 3.7.2.1, the nominal thickness is 0.002 inch unless the drawing or order says otherwise, and the thickness may vary by plus or minus 20 percent up to that value, or by 0.0004 inch for thicker coatings. Type II has no default thickness in the specification at all, so a Type II callout with no thickness gives the finisher a wide window and gives you no dimensional control.
That tolerance on the coating matters as much as its nominal value. Because a diameter changes by one full coating thickness, the whole coating tolerance lands on the diameter too. Per MIL-A-8625F, a nominal 0.002 inch hard coat can therefore shift a bore or shaft diameter anywhere from about 0.0016 to 0.0024 inch, a spread of 0.0008 inch before any machining variation is added. The table below turns these rules into machining targets for common coating thicknesses.
What should the pre-anodize size be for a bore or a shaft?
Machine a shaft undersize and a bore oversize by one full coating thickness on the diameter, so that the half-thickness growth on each side brings the feature to its finished size. Width and wall features measured across two coated faces follow the same arithmetic.
Pick the coating thickness you will write on the drawing and read across. The last column is the pitch-diameter change on an unmasked 60-degree thread, covered in the next section.
| Coating (type, thickness) | Growth per surface | Diameter change | Machine shaft OD | Machine bore ID | 60° thread pitch diameter |
|---|---|---|---|---|---|
| Type II, 0.0002 in (5 µm) | 0.0001 in (2.5 µm) | 0.0002 in (5 µm) | Finished OD − 0.0002 in | Finished ID + 0.0002 in | +0.0004 in (10 µm) |
| Type II, 0.0005 in (13 µm) | 0.00025 in (6 µm) | 0.0005 in (13 µm) | Finished OD − 0.0005 in | Finished ID + 0.0005 in | +0.0010 in (25 µm) |
| Type II, 0.0010 in (25 µm), top of Table IV range | 0.0005 in (13 µm) | 0.0010 in (25 µm) | Finished OD − 0.0010 in | Finished ID + 0.0010 in | +0.0020 in (51 µm) |
| Type III, 0.0015 in nominal (38 µm), coating 0.0012–0.0018 in | 0.00075 in (19 µm) | 0.0012–0.0018 in | Finished OD − 0.0015 in | Finished ID + 0.0015 in | +0.0030 in (76 µm) |
| Type III, 0.0020 in default nominal (51 µm), coating 0.0016–0.0024 in | 0.0010 in (25 µm) | 0.0016–0.0024 in | Finished OD − 0.0020 in | Finished ID + 0.0020 in | +0.0040 in (102 µm) |
| Type III, 0.0030 in nominal (76 µm), coating 0.0026–0.0034 in | 0.0015 in (38 µm) | 0.0026–0.0034 in | Finished OD − 0.0030 in | Finished ID + 0.0030 in | +0.0060 in (152 µm) |
The allowance only moves the center of the tolerance band; it does not remove the coating variation. The worked example below shows what happens to a real bore when the coating tolerance is added to the machining tolerance, and why the answer is sometimes to size the bore after coating instead.
| Step | Value | Comment |
|---|---|---|
| Finished bore (after coating) | 0.7500–0.7510 in | Band of 0.0010 in |
| Coating | 0.0016–0.0024 in (0.0020 in nominal) | Default Type III nominal and tolerance |
| Diameter change | 0.0016–0.0024 in | One full coating thickness on the diameter |
| Machine the bore to 0.7520–0.7530 in | Coated result 0.7496–0.7514 in | Band of 0.0018 in: fails at both ends |
| Machining band that still fits | Machine to 0.7524–0.7526 in (band of 0.0002 in) | 0.0010 in drawing band minus 0.0008 in coating spread |
| Alternative | Machine undersize, coat, then hone to 0.7500–0.7510 in | Per paragraph 6.7 |
How does anodizing change threads and small holes?
Threads move far more than plain diameters, because per MIL-A-8625F paragraph 6.10.2 the pitch diameter of a 60-degree thread changes by four times the surface growth. Small and tapped holes are the opposite problem: the coating inside them is unpredictable.
Per the same paragraph, the major and minor diameters of a thread change by twice the surface growth, and for thread forms other than 60 degrees the pitch diameter changes by twice the growth divided by the sine of half the included angle. In practice that means a default hard coat on an external thread moves the pitch diameter by roughly twice the coating thickness, as the table above shows, which is more than the whole pitch-diameter tolerance of many fine thread classes.
The usual choices are to mask the thread so it stays bare, to cut the thread to an allowance that the coating brings back into class, or to chase the thread after coating. Masking is the most predictable for Type III. For Type II, the growth is small enough that a thread cut toward the favorable side of its class often gauges after coating, but check it with the go and no-go gauges on first articles rather than assuming it.
For holes, per MIL-A-8625F paragraph 6.10.1.1, the film in small holes and tapped holes can vary from none to a full normal coating, and the specification says holes over 1/4 inch should be anodized. Below that size, treat the coating in the hole as unknown and call such holes out as masked or as “coating not required” so neither side has to guess.
Should drawing dimensions apply before or after anodizing?
By default, MIL-A-8625F paragraph 3.9 holds parts to the drawing dimensions after the coating is applied, and paragraph 6.10.1 recommends showing both a machining dimension and a coated dimension. The safest drawing states which one each critical dimension means, in words.
Leaving it to a default causes disputes because the machinist and the finisher each read the drawing for their own step. If the part is machined in one shop and anodized in another, the machinist will usually work to the numbers printed on the drawing, and a note that says dimensions apply after coating only helps if someone has converted those numbers into machining sizes. ASME Y14.5-2018 sets how dimensions and tolerances are expressed, but it does not choose a coating allowance for you; that is a design decision to write on the drawing. Notes that work well:
- ALL DIMENSIONS APPLY AFTER ANODIZE UNLESS MARKED ◊. Then mark the few features that are controlled before coating, such as datum faces that will be masked.
- ◊ DIMENSION APPLIES BEFORE HARD ANODIZE. MACHINE TO THIS SIZE. Use it beside the machining dimension of a bore or shaft, with the coated size shown as a reference dimension next to it.
- HARD ANODIZE PER MIL-A-8625F, TYPE III, CLASS 1, [THICKNESS AND TOLERANCE]. MASK THREADS AND BORE A. A thickness with a tolerance turns the growth into a number you can plan around.
- BORE A: HONE TO SIZE AFTER HARD ANODIZE. MINIMUM REMAINING COATING [VALUE]. Use it when the coating spread is larger than the finished tolerance, as in the worked example.
Either way, inspect machining sizes before finishing and coated sizes after it. The machining RFQ guide covers how to package these notes with the rest of the drawing set.
When should a hard-anodized bore be honed or lapped after coating?
Hone or lap after coating when the finished tolerance is tighter than the coating’s own thickness spread, or when the bore is a sliding or sealing surface that needs a smoother finish than hard coat leaves. MIL-A-8625F paragraph 6.7 describes exactly this route: coat oversize, then lap or hone down to size.
The specification notes that hard coat is rougher and less dense at its outer surface than near the base metal, so removing some of the top layer gives a better running surface without losing the hard core. That makes honing a normal finishing step, not a repair. To leave room for it, write a coating thickness that includes the stock to be removed and state a minimum thickness that must remain after sizing.
For Type II, per MIL-A-8625F paragraph 6.10.1.1, parts with a total tolerance of 0.0004 inch or less that are lapped, honed or stoned to size after anodizing should get a supplementary protective treatment on the sized surface, and discoloration there is acceptable.
Compare the coating spread with the machining tolerance quoted for the feature. Per MIL-A-8625F, a default hard coat alone adds a diameter spread of 0.0008 inch, so a finished bore tolerance of a thousandth or less is usually a hone-after-coat feature regardless of how well the bore was turned or milled.
Which alloys and part features make anodizing growth harder to predict?
Sharp edges, high-silicon or high-copper alloys and very thick hard coats all make the actual growth drift from the one-half rule. Design for them before the part is machined, because none of them can be corrected in the tank.
Per MIL-A-8625F paragraph 6.10.3, anodic coatings develop voids at sharp corners and edges, so edges and inside corners should be radiused before anodizing; Table III of the specification links the minimum radius to the coating thickness. Break every edge that will be coated, and say on the drawing whether the break is before or after coating.
Alloy matters too. Per MIL-A-8625F paragraph 3.4.3, Type III is not applied without approval to alloys with a nominal copper content above 5 percent or a nominal silicon content above 8.0 percent, and per paragraph 6.16 a hard coat of 2 mils or more is extremely difficult to obtain on high-silicon die-casting alloys such as 380 and 383. ADC12 sits in the same high-silicon family, and the ADC12 vs A380 guide compares it with A380. On those alloys the achievable thickness, and so the growth, is lower and less uniform, which is one reason die castings are more often finished another way; the finish comparison guide covers the alternatives.
Thickness itself has limits. Per MIL-A-8625F paragraph 6.17, the abrasion resistance of Type III drops as the coating approaches 3 mils, and per paragraph 6.10.7 thicker coatings reduce fatigue strength more. Pick the thinnest coating that does the job and size the part around it.
For parts made on CNC turning and CNC milling and then finished, the drawing notes above are what keep the two steps aligned; the surface finishing page lists the finishes our partner plants apply, and the quote names where each step is done. You can send a drawing for a quote with the coating note as written, and any before-or-after question will be raised before machining starts.
Frequently asked
Does dyeing (Class 2) change the dimensions compared with Class 1?
The class only says whether the coating is dyed. The dimensional change follows the coating thickness, which is set by the type and by any thickness you write on the drawing.
Is the one-half growth rule exact?
No. The specification describes it as the increase that can be expected, and the real split varies with alloy and process. Use it for the first machining target, then correct the target from before-and-after measurements on first articles.
How is anodize thickness measured on a finished part?
Usually with an eddy-current gauge per ASTM B244, one of the methods MIL-A-8625F names; the specification averages several readings. On a diameter, a before-and-after micrometer or bore-gauge reading also shows the growth directly.
What happens to dimensions if a part is stripped and re-anodized?
Stripping removes the oxide, including the half that replaced base metal, so the bare part comes back smaller than it was machined. A second coat then grows from that smaller size, so tight features may no longer hold after rework.
Should I specify a thickness for Type II anodizing?
If any dimension on the part is close to its limit, yes. MIL-A-8625F gives no default Type II thickness, so a thickness range on the drawing is the only way to bound the growth.
Sources
- MIL-A-8625F, Anodic Coatings for Aluminum and Aluminum Alloys, US Department of Defense, 10 September 1993, with Amendment 1 of 2003 (paragraphs 3.4.3, 3.7.2.1, 3.9, 6.7, 6.10.1 to 6.10.3, 6.10.7, 6.16, 6.17 and Tables III and IV)
- ASME Y14.5-2018, Dimensioning and Tolerancing, American Society of Mechanical Engineers, 2018 (reaffirmed 2024)
- ASTM B244-09(2021), Measurement of Thickness of Anodic Coatings on Aluminum with Eddy-Current Instruments, ASTM International, reaffirmed 2021