The defects I see most often on anodized parts
Anodizing is the last step, so every defect shows up when the part is already machined, inspected and expensive. In our shop the same five or six problems come back again and again, and almost none of them are caused by the anodizer being lazy. They start earlier - in the alloy choice, the drawing notes, or the way the part was racked.
Here is what each defect actually looks like, what causes it, and what stops it from happening on the next batch.
Defect, cause, and what really fixes it
| Defect | What you see | Root cause | Fix |
|---|---|---|---|
| Color variation | Same part number, two visibly different shades between batches | Alloy lot and copper content, bath chemistry drift, different rack positions and current density | Fix the alloy and temper on the drawing, keep one supplier lot per batch, use a color reference panel agreed in advance |
| White spots / bleeding | Chalky spots near blind holes, threads or tapped holes | Electrolyte trapped in holes bleeding out during sealing, poor rinsing | Add vent or drain holes, mask tapped holes, or specify a proper rinse and de-ionized water step |
| Burned edges | Dark, matte or pitted corners and edges | Current density concentrates on sharp edges | Break all edges 0.3-0.5 mm before coating, no razor-sharp corners on anodized features |
| Blooming / smut | Hazy white film that wipes away and comes back | Poor sealing, dye bleed-out, contaminated seal bath | Sealing quality check per ISO 2143 dye stain test, keep the seal bath in spec |
| Dimensional growth | Shafts and bores that no longer meet tolerance | The oxide layer builds outward as well as inward | Pre-machine undersize or split the tolerance: plan for roughly half the coating thickness per side |
| Crazing / micro-cracks | Fine cracks on the surface, usually on hard anodize | Thick coating on sharp corners or high-stress geometry | Radius corners, reduce thickness, or machine the feature after coating |
Why 7075 goes blotchy while 6061 stays clean
If you anodize 7075 and 6061 side by side, 6061 almost always comes out uniform and 7075 almost always shows some variation. The reason is copper: 7075 contains roughly 1.2-2.0% copper and 2024 even more, and copper-rich phases dissolve at a different rate during anodizing. The result is uneven coating thickness and visible color shifts, especially with clear anodize where nothing hides it.
That does not mean you cannot anodize 7075. It means you should not promise a cosmetic finish on it. If appearance matters, use 6061. If you need the strength of 7075, expect a slightly darker or mottled look and specify black dye if you want consistency - a dark dye hides far more variation than clear.
Dimensional growth: the tolerance killer nobody plans for
Anodizing does not only build on top of the surface - the oxide grows into the substrate as well. As a rule of thumb, the dimension grows by about half the coating thickness per side. That is enough to scrap a part that was machined exactly to the nominal size.
| Coating | Typical thickness | Growth per side | What it means for your tolerance |
|---|---|---|---|
| Type II clear (decorative) | 5-15 µm | about 2.5-7.5 µm | A ±0.01 mm bore can go out of tolerance |
| Type II, heavy | 20-25 µm | about 10-12.5 µm | Pre-machine undersize on all coated fits |
| Type III hard anodize | 25-50 µm | about 12-25 µm | Plan the allowance, or mask the fit and coat elsewhere |
The practical answer is to decide on the drawing which surfaces are coated and which are masked, and to give the machinist an allowance on the coated fits. If a bore must stay at H7 after coating, say so explicitly instead of leaving it to the anodizer.
What to put on the drawing so parts come back right
- Specify the type and thickness, not just the word anodize: Type II clear, 10-15 µm, ISO 7599; or Type III hard, 25-40 µm, hardness 400 HV minimum
- Name the alloy and temper, for example 6061-T6 rather than aluminium - alloy drives both color and thickness uniformity
- Mark coating and masking zones: threads, sealing faces, bearing bores and any datum you will measure against
- Allow for growth on every coated fit, or state that the fit must hold after coating
- Break sharp edges 0.3-0.5 mm, especially on hard anodize, to avoid burning and edge cracking
- Say what the surface has to look like: a color reference panel, gloss level, or a note that cosmetic variation is acceptable
Incoming inspection: catching bad anodizing in ten minutes
- Measure coating thickness with an eddy-current gauge at five or six points per part, including an inside surface if the part is a housing
- Compare color against the agreed reference panel under daylight, not under workshop fluorescents
- Check sealing with a dye stain test if the parts will be exposed to weather or washdown
- Look at edges and blind holes at a low angle - burns and trapped-electrolyte spots show up best when light rakes across the surface
- Verify the critical fits again after coating, not just before - this is where most dimensional surprises appear
TruPart Precision is an ISO 9001:2015 shop in Dongguan. We machine 6061 and 7075 aluminium to ±0.005 mm on critical features, plan the anodizing allowance into the process, and supply a coating thickness record with the batch. When a finish is cosmetic-critical, we say so at the quote stage rather than after the parts are coated.
Frequently Asked Questions
How much does an aluminium part grow during anodizing?
Plan on roughly half the coating thickness per side. Type II clear at 10 µm adds about 5 µm per surface, and Type III hard anodize at 30 µm adds about 15 µm per surface. Anything tighter than ±0.02 mm on a coated fit needs an allowance written into the drawing.
Can a part with a bad anodized finish be re-anodized?
It has to be stripped first, usually in a caustic bath, which removes material and changes dimensions slightly. On a tight-tolerance part that is often worse than the original defect, so in practice most parts are machined oversize, stripped and re-coated, or simply scrapped.
Why does 7075 anodize unevenly compared with 6061?
7075 contains 1.2-2.0% copper and 2024 contains even more. Copper-rich phases react at a different rate, so coating thickness and color vary across the part. Clear anodize shows it immediately; black dye hides most of it.
Do I need to mask threads before anodizing?
If the thread has to hold a gauge or a class fit after coating, yes. Either mask it, or specify the thread tolerance as measured after anodizing so the shop knows to cut it undersize. Leaving it unsaid is how a 6H thread becomes too tight.
How do I know if the sealing step was done properly?
Ask for a dye stain test per ISO 2143, or test it yourself: a poorly sealed part bleeds dye and later blooms with a white film. If the parts will see outdoor or washdown service, poor sealing is the defect that shows up months later.