Guide
Powder coat thickness: what mil range should you specify, and how much clearance do fits need?
Powder coat thickness is the dry film thickness of the cured coating, measured on the part after it leaves the oven, and it is the number that decides whether a coated shaft still enters its bore, a tab still drops into its slot and a bolt still passes through its hole. A typical single coat lands somewhere in the 2–4 mil (51–102 µm) band per side, which is several times the build of a zinc plate or a Type II anodize, so it has to be designed in rather than discovered at assembly.
By Saola Manufacturing engineering team · Published · Updated
This guide is for design and quality engineers who have to write a film-thickness requirement on a drawing or purchase order and leave the right clearance on the features that mate. It does not repeat the comparison between finishes, which is covered in anodizing vs powder coating vs zinc plating, and it does not cover pinholes on castings, which are in powder coating die cast aluminum without outgassing. The focus here is narrower: what range to write, why a single nominal value is the wrong callout, how the film is measured, and how to do the fit arithmetic before the part is ever quoted.
On this page
What is a typical powder coat thickness, and where does the range come from?
Most single-coat powders are applied to a dry film somewhere between 2 and 4 mil, and the exact window for a given product comes from the powder manufacturer's technical data sheet, not from a general standard. Textured and specialty powders have their own windows, and a primer plus topcoat system adds the two films together.
Start with units, because drawings get this wrong more often than you would expect. A mil is one thousandth of an inch, not a millimeter, and NIST lists the conversion as exactly 25.4 micrometers per mil in its guide to SI units (SP 811). A note that says “2–3 mil” on a metric drawing, read as millimeters, asks for a coating roughly forty times too thick, and some suppliers will stop and ask while others will simply quote something odd. Write the unit out in full, or give both units, as in this typical note: “0.002–0.004 in (51–102 µm)”.
There is no single ASTM or ISO standard that tells you how thick a powder coat must be on a general industrial part. Architectural aluminum specifications set minimums for building products, and some end-use specifications set their own, but for a machined bracket or a die-cast housing the governing number is the window printed on the data sheet of the powder you approve. That window exists because the powder needs enough film to flow out, hide the substrate and reach its rated properties, and an upper limit, because a film built well past the window tends to show more orange peel and to chip more easily at edges. As industry guidance, many single-coat powders, smooth and some textured ones alike, list a window of roughly 2–3 mil (51–76 µm), while a two-coat system totals the primer and topcoat windows. Treat that as a starting point to check against the actual product, not as a value to copy onto the print.
How much does powder coat thickness change a fit?
Every coated surface grows by the full film thickness, so a diameter or a width changes by twice the film, and a coated tab in a coated slot loses four times the film from its clearance. Unlike anodizing, nothing grows into the substrate; the powder sits entirely on top.
The arithmetic is simple, but it has to be done per feature, because the effect doubles each time two coated surfaces face each other. An external diameter gets bigger by two films. A hole gets smaller by two films, if powder reaches the bore at all, which on a small blind hole it often does not reach evenly. A slot and the tab that drops into it each change by two films in opposite directions, so their clearance closes by four. Two coated plates bolted face to face add four films to the stack height, which matters for a spacer, a shim pack or a clamp that relies on bolt preload. The table below runs those cases at the two ends of the typical single-coat band. Compare the result with the minimum clearance of your fit, not the nominal, because the worst case is a bore at its smallest size meeting a pin at its largest, both at the top of the film range. For the base fit values themselves, see clearance, transition and interference fits.
| Feature | How the film acts | Change at 2 mil (51 µm) per side | Change at 4 mil (102 µm) per side |
|---|---|---|---|
| Shaft or pin diameter | Grows by 2 films | +0.004 in (+0.10 mm) | +0.008 in (+0.20 mm) |
| Through-hole diameter, coated inside | Shrinks by 2 films | −0.004 in (−0.10 mm) | −0.008 in (−0.20 mm) |
| Tab thickness or rib width | Grows by 2 films | +0.004 in (+0.10 mm) | +0.008 in (+0.20 mm) |
| Tab-in-slot clearance, both coated | Closes by 4 films | −0.008 in (−0.20 mm) | −0.016 in (−0.41 mm) |
| Stack of two plates clamped face to face | Grows by 4 films | +0.008 in (+0.20 mm) | +0.016 in (+0.41 mm) |
Why is a single nominal powder coat thickness the wrong callout?
Spray-applied powder never lands as one value across a part, so a nominal with no range gives the coater nothing to pass or fail against and gives you nothing to stack against your fits. Write a minimum and a maximum, and say where they apply.
Electrostatic powder follows the field lines from the gun to the grounded part. Flat faces facing the gun build most evenly. Sharp outside corners tend to end up thin, and the inside surfaces of cup and box shapes are shielded by the Faraday cage effect, so they come out thin or poorly coated, as the International Zinc Association warns casting designers. Recoats, which are common when a lot is touched up after a cosmetic reject, add a second full film on top of the first. The result is that one part can carry noticeably different film on a flat face, a sharp edge and an inside corner.
Three things follow for the drawing. First, the minimum is set by protection and appearance, so take it from the powder data sheet and from your corrosion requirement. Second, the maximum is set by your fits, by edge chipping and by the assembly, so derive it from the table arithmetic above, not from the data sheet alone. Third, a window that is too narrow is a cost and a scrap risk: a range about as wide as the powder's own data-sheet window, applied to flat, accessible surfaces, is a realistic request for a single coat, while a window much narrower than that needs a conversation with the coater before it is quoted. If a feature really cannot tolerate that variation, it should be masked or machined after coating rather than squeezed into an impossible film tolerance.
A workable drawing note reads like this: “Powder coat per [powder product and color]. Dry film thickness 0.002–0.004 in (51–102 µm) on significant surfaces marked ‘S’, measured per ASTM D7091. Inside corners and recesses: full coverage, thickness not controlled. Surfaces marked ‘M’: mask, no coating.” The marked significant surfaces matter, because a thickness range that applies everywhere will fail in every inside corner of every part.
How is powder coat thickness measured, and what should the acceptance plan say?
Cured powder is normally measured with a handheld magnetic gage on steel or an eddy-current gage on aluminum and zinc, following ASTM D7091. The method is quick and nondestructive, but it only reads reliably on flat or gently curved surfaces away from edges.
ASTM D7091 covers both gage types: magnetic-induction or magnetic-pull gages for nonmagnetic coatings on ferrous metals, and eddy-current gages for nonconductive coatings on nonferrous metals such as aluminum and zinc die castings (the full test plan for coated zinc castings is in powder coating zinc die castings). It describes how to verify gage accuracy against reference standards and how to adjust the gage, and that matters on parts like these, because a gage zeroed on a flat steel block can misread on a thin stamping, a small radius or a cast surface whose alloy and roughness differ from the block. Readings taken close to an edge, on a tight curve or inside a small bore are not trustworthy with a handheld probe, which is another reason to tie the thickness range to marked surfaces that a probe can actually reach.
The acceptance plan should answer four questions on the purchase order: which surfaces are measured, how many readings make one result, how many parts per lot are checked, and what happens when one reading falls outside the range while the average is inside. A common approach is to average a few readings in a small area as one spot measurement, then require every spot to fall inside the range on a sampled set of parts. Agree the plan before the first article, then put it in writing; otherwise the coater and your receiving inspector will each pick a different reading and both will be right. If you already sample other features by AQL, the same plan can carry film thickness; see AQL sampling for machined parts.
Where a gage cannot reach, measure the function instead. A coated bore can be checked with a GO pin sized for the minimum coated diameter, a slot with a GO tab, and a threaded hole with a GO thread gage after coating, which tells you directly whether the part assembles, regardless of what the film did in that recess.
How should the drawing handle dimensions that will be powder coated?
State whether each critical dimension applies before or after coating, and for a fit that must hold after coating, size the bare feature so that the coated feature lands inside its tolerance. Leaving it unstated forces the machinist and the coater to guess which of them owns the film.
There are three workable patterns, and they suit different features. The first is to give the bare dimension and a separate coated dimension, the way MIL-A-8625F recommends for anodized parts; per MIL-A-8625F paragraph 6.10.1, drawings should carry both a machining and a coated dimension, and the same logic carries over to powder. This is clearest for a shaft or a tab: the machinist works to the bare value, and inspection checks the coated value. The second is to mask the feature and dimension it bare, which is usually right for bearing bores, press-fit diameters, ground contacts and threads, because no film range is tight enough for those fits. The third is to coat everything and machine the critical features afterward, which gives a clean bare surface and an exact size but costs a second setup and leaves a raw edge where the film ends.
For clearance holes, the cheapest fix is often to open the hole. If a bolt hole is sized for a close fit and the hole is coated inside, two films at the top of a typical range can take up much of the clearance and push a hole pattern out of its position tolerance at assembly; the bonus tolerance arithmetic in MMC bonus tolerance shows how quickly that disappears. Moving to the next larger standard clearance hole, or masking the hole with a plug, usually costs nothing at the machine. For threads, use the reasoning in threads after anodizing or plating; a powder film is too thick to absorb by cutting the thread oversize, so mask or chase. When several coated features stack up across an assembly, carry the film into the stack as its own toleranced contributor, using both ends of its range rather than the nominal.
At Saola, powder coating is one of the eight finishes on the surface finishing menu, applied by a partner plant that Saola manages; the quote names where it is done. The surface finishing page asks buyers to mark threads, ground contacts and press-fit bores that must stay bare, so that masking is done correctly the first time, and the first-article inspection report described on the quality page covers every toleranced dimension on your drawing, which includes any coated dimension you tolerance. Our quality page does not list a dry-film-thickness gage, so if film thickness is an acceptance criterion rather than only a fit result, ask in the RFQ how it will be checked and reported before you rely on it.
What should a powder coat RFQ include, and what mistakes cause rework?
Send the powder product or color standard, the film-thickness range with its unit, the surfaces it applies to, the masked features and the measurement method, all on the drawing rather than in an email thread. Most rework on coated parts traces back to one of those five items being missing.
A complete RFQ for a powder-coated part carries the following:
- Powder identification. The product, or at least the resin family, color and gloss you will approve, so the coater can quote the data-sheet film window rather than invent one.
- Thickness range with a unit. Minimum and maximum, in mil or micrometers, written out, never a bare “2–3”.
- Where it applies. Significant surfaces marked on the drawing, with recesses and inside corners called out as coverage-only.
- Masked features. Threads, bores, ground points and datum faces that stay bare, shown on the drawing.
- Before or after coating. A general note, plus a coated dimension on every fit that must hold after finishing.
- Measurement and sampling. ASTM D7091, the number of readings per result, and the parts per lot.
The mistakes that cause rework repeat from part to part. A drawing that gives one nominal film value with no range. A metric drawing that says “mil” and gets read as millimeters. A fit sized for the bare part and coated anyway, discovered when the first pin will not enter. A thickness range applied to the whole part, so every inside corner fails. A recoat after a cosmetic reject that doubles the film on a fit nobody rechecked. And a corrosion requirement written as a salt-spray duration with no film minimum, so the thinnest part of the coating, usually an edge, decides the result.
Frequently asked
Is thicker powder coat always better for corrosion protection?
No. Above the data-sheet maximum the film gets brittle at edges and impacts, and a chip lets moisture under the coating. Corrosion performance depends mostly on pretreatment and on reaching the minimum film on edges and corners, which is why the minimum on the drawing matters more than a generous maximum.
Can I specify the film thickness only in micrometers?
Yes, and on a metric drawing it is the safer choice because it removes the mil-versus-millimeter misreading. On an inch drawing, giving both units in the note costs nothing and stops the same error in the other direction.
Does powder coat go inside holes and threads?
Partly and unevenly. The Faraday cage effect keeps film thin inside small bores and blind holes, but it is rarely zero, and an uneven film inside a thread is worse than either a full or a bare one. Mask holes and threads that must fit, and check them with a GO gage after coating.
How do I handle a part that may need a touch-up recoat?
Say on the drawing or purchase order whether recoating is allowed and whether the thickness range still applies after it. If fits are tight, require that recoated parts are rechecked at the fit features, or that they are stripped and recoated rather than coated over.
Should the coater or the machinist own the coated dimension?
Split it explicitly. The machinist owns the bare dimension, the coater owns the film range, and the coated dimension is the sum that inspection checks. When one supplier does both, the split still belongs on the drawing so the inspection report can show where a deviation came from.
Sources
- ASTM D7091-22, Standard Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals and Nonmagnetic, Nonconductive Coatings Applied to Non-Ferrous Metals, ASTM International, 2022
- NIST Special Publication 811, Guide for the Use of the International System of Units (SI), National Institute of Standards and Technology, 2008 edition
- MIL-A-8625F, Anodic Coatings for Aluminum and Aluminum Alloys, US Department of Defense, 1993 (Amendment 1, 2003)
- Powder coating, International Zinc Association, 2016