Guide
Which powder coating defects should a receiving inspector reject, and what causes each?
In receiving inspection, powder coating defects are departures from a continuous, evenly cured organic film that you can see or measure: pinholes, bubbles and blisters, fish eyes and craters, chips and poor adhesion, thin or bare edges, orange peel, dirt inclusions, and discoloration from over-baking. Some start on the coating line, but many start in the part itself, in casting porosity, sharp machined edges, trapped coolant or weld gas, so naming the defect correctly is the first step to deciding who fixes it and whether the lot can be used.
By Saola Manufacturing engineering team · Published · Updated
This guide is for quality engineers, design engineers and buyers who have a lot of powder-coated parts on the receiving bench and need to answer three questions: is this a defect under the requirements we actually wrote, where did it come from, and do we accept, rework or reject it. It gives a defect atlas that maps each symptom to its usual cause and owner, a bench diagnosis sequence using published ASTM test methods, a disposition table, and the drawing and purchase order wording that stops the same argument on the next lot. Pinholes on die castings are covered in depth in the die-cast outgassing guide; this page covers the whole family of defects a receiving inspector meets.
On this page
What are the common powder coating defects, and how do you recognize each one?
The defects that come up most at receiving fall into eight groups: pinholes, blisters or bubbles, fish eyes and craters, chipping or flaking, thin or bare edges, orange peel, embedded dirt, and over-bake discoloration. Each has a recognizable look and location pattern, and the pattern is the best first clue to the cause.
A pinhole is a tiny hole, often with a raised rim, that may or may not reach the metal. A blister is a dome of film lifted off the substrate or off a lower coat; if it is already broken it can look like a pinhole with a ring around it. A fish eye or crater is a round depression where the melted powder pulled away from a spot of contamination, usually wider and shallower than a pinhole and sometimes with a visible dot in the center. Chipping is film that breaks away under impact or at an edge, and if large flakes lift cleanly from bare metal, the real problem is adhesion rather than handling.
Edge defects deserve their own line on the inspection sheet. Melted powder is pulled away from sharp edges by surface tension, so a knife edge left by machining often ends up with thin or no film, and inside corners and deep recesses tend to be thin for a different reason: the electrostatic field concentrates on outer edges and does not reach into the recess, which coaters call the Faraday cage effect. Orange peel is a dimpled texture like the skin of an orange; it is normal to some degree in powder, so it is a defect only when it exceeds an agreed sample. Dirt or “seeds” are raised specks of foreign matter or gelled powder trapped in the film, and over-bake shows as yellowing, loss of gloss or a color shift, most visible on white and light colors.
| Defect | What it looks like | Most likely causes | Usual owner | How to confirm |
|---|---|---|---|---|
| Pinholes | Tiny holes, often rimmed, repeating in the same spots part after part | Gas from casting porosity, welds or galvanizing; liquid trapped in pores or seams; film applied too thick | Substrate (casting, weld) more often than the coater | Magnifier; wet-sponge holiday test (ASTM D5162 Method A) on functional faces |
| Blisters and bubbles | Domes of lifted film, sometimes appearing only after humidity or salt-spray exposure | Soluble salts or poor rinsing under the film; moisture; outgassing; weak pretreatment | Coater (pretreatment); substrate for gas blisters | Rate size and frequency against ASTM D714 reference photographs |
| Fish eyes and craters | Round depressions where powder pulled back, scattered rather than in a fixed spot | Low-surface-tension contamination: oil, silicone, coolant, wax, cross-contaminated powder, oil or water in compressed air | Coater, or machining and cleaning if oil comes in on the parts | Location varies by rack or shift, not by part geometry |
| Chipping and flaking | Film breaking off at edges or under impact; clean bare metal under large flakes | Poor pretreatment; under-cure; excessive film build at edges; packing and transit damage | Coater, unless damage pattern matches packaging | Tape adhesion (ASTM D3359); solvent rub for cure (ASTM D5402) |
| Thin or bare edges and recesses | Metal or primer showing on sharp edges; low film in inside corners and pockets | Sharp machined or stamped edges; Faraday cage effect in recesses; gun settings and racking | Design and machining for edges; coater for recesses | Dry film thickness (ASTM D7091) at named points; holiday test on edges |
| Orange peel | Dimpled texture, varying across the part or between lots | Film too thick; slow heat-up so powder gels before it flows; powder or cure schedule | Coater | Compare with an approved texture sample under agreed lighting |
| Dirt, seeds and inclusions | Raised specks or fibers embedded in the film | Booth or oven debris; unsieved reclaim powder; poor cleaning; falling particles from conveyors | Coater | Count per surface against the agreed limit |
| Over-bake discoloration | Yellowing, gloss drop or color shift, often on thick sections or parts that sat in the oven | Oven temperature or time beyond the powder supplier’s cure window; recoating | Coater | Compare with the color standard; instrument color if the PO names a tolerance |
What causes powder coating defects, and do they start in the part or on the coating line?
The usual causes are contamination on the part, weak pretreatment, gas or liquid trapped in the substrate, wrong film thickness, wrong cure, and part geometry such as sharp edges and deep recesses. Contamination, pretreatment and cure belong to the coater; trapped gas, sharp edges and oil carried in from machining start before the parts reach the coating line.
Contamination is the first suspect for fish eyes and craters because melted powder has a higher surface tension than oil, silicone or wax, so it pulls away from any spot that carries them. The source can be on the line, such as oil or water in the compressed air or another powder mixed in the booth, but it can also arrive on the parts: coolant and cutting oil soaked into the pores of a casting, silicone from a release spray or sealant, or fingerprints after cleaning. A tumble-deburred part that was not rinsed well can bring in compound residue the same way. Powder Coated Tough magazine puts silicones and penetrating oils at the top of the contaminant list and names poorly filtered compressed air, contaminated blast media and cross-contamination from other powders as common line sources (Powder Coated Tough).
Pretreatment controls adhesion and much of the blister resistance. ASTM D1730 groups the ways to prepare aluminum for painting into solvent cleaning, chemical treatments, anodic treatments and mechanical treatments (ASTM D1730-09(2020)), and a chemical conversion coating is the usual choice under powder. When a step is skipped or a rinse is dirty, the film can look perfect at receiving and still flake under a tape test or blister later in humid service, which is why ASTM D714 exists to describe blister size and density with photographic references (ASTM D714-25).
Trapped gas is a substrate problem that shows up in the coating. Die castings hold porosity, welds can hold gas and flux, and hot-dip galvanized steel releases gas when heated, which is why ASTM D7803 includes a thermal step when preparing galvanized steel for powder (ASTM D7803-25). When the part heats in the cure oven, that gas pushes out through the melting powder and leaves pinholes or blisters in the same places on every part.
Film thickness and cure are the last two. Too thin, and edges and high points show through; too thick, and the film traps air, develops heavy orange peel, chips at edges and interferes with fits. Under-cure leaves a soft film with poor solvent and impact resistance, while over-bake discolors it. Both are invisible to a quick look and need the tests in the next section.
How do you diagnose a powder coating defect at the receiving bench?
Start with the location pattern, then measure: a defect that repeats in the same place on every part points to the substrate or the geometry, while one that moves around points to contamination or the line. Film thickness, cure and adhesion checks then separate the remaining causes.
Work through the lot in this order before writing a rejection:
- Map the defect. Mark where it occurs on several parts. Same spot every time, on thick sections, machined faces of castings or near welds: suspect trapped gas. Random spots, clusters on one rack position or one date code: suspect contamination or line conditions. Only on outer edges: suspect sharp edges and thin film.
- For craters, trace the contaminant. Ask the coater when the compressed-air filters and moisture separators were last serviced, whether blast media is reused, and whether another powder ran in the booth just before your parts; on your side, check whether the parts were sprayed with a silicone or penetrating-oil product, or packed in oiled paper, before they shipped to finishing. A crater pattern that changes with the coating date points to the line; one that follows a machining batch points to the parts.
- Look closely. Use a low-power magnifier. A small round hole with a raised rim, recurring in the same place on each part, is usually gas; a wide, shallow, saucer-shaped depression is usually a fish eye; a raised speck is dirt.
- Measure film thickness. Take readings at the defect area and at agreed reference points per ASTM D7091, which covers magnetic gauges on steel and eddy-current gauges on aluminum (ASTM D7091-22). Very high readings near pinholes or orange peel point to excessive build; very low readings at edges confirm thin coverage.
- Check whether holes reach the metal. On faces where corrosion protection matters, a low-voltage wet-sponge holiday detector finds discontinuities that the eye misses; per ASTM D5162, Method A is generally used for coatings 0.5 mm (20 mil) thick or less, and the allowable number of discontinuities should be decided before the test (ASTM D5162-24).
- Check cure. A solvent rub test per ASTM D5402 is a quick way to separate an under-cured film from a well-cured one; the practice does not set the solvent, rub count or pass level, so take those from the powder supplier’s data sheet (ASTM D5402-19). If the rub test fails, ask the coater for a part-temperature record from the oven for that lot: heavy, thick-walled parts need longer dwell than light-gauge parts to reach cure temperature, so an oven set-point alone does not prove the part was cured.
- Check adhesion. If parts are chipping or flaking, run a cross-cut tape test away from the defect; per ASTM D3359, Method B is for coating systems less than 125 µm (5 mil) in total thickness and Method A for thicker ones, so measure the film first, since many single-coat powder films fall under that limit but heavy builds and two-coat systems may not (ASTM D3359-23). Clean metal under the removed film points to pretreatment; failure only on machined faces points to coolant residue.
- Check the packaging. Chips that line up with straps, dividers or part-to-part contact are transit damage, which is a packing instruction problem, not a coating problem.
Keep the evidence: photographs with a scale, the part numbers and lot or date code, thickness readings and test results. That record is what lets the supplier trace the defect to a line, a rack or a casting cavity, and it is the core of any 8D corrective action report you ask for.
Which powder coating defects should be rejected, and which can be accepted or reworked?
Reject defects that break the film on a functional surface, such as holes to metal, blisters, flaking or bare edges where corrosion protection is required, and defects that exceed the cosmetic limits written on your drawing or purchase order. Purely cosmetic marks on surfaces you did not classify are rejectable only if a written standard or limit sample says so.
The split matters because no ASTM method sets a pass level for appearance by itself. ASTM D714, for example, states that the permissible degree of blistering is to be agreed between purchaser and seller, and ASTM D5162 says the allowable number of discontinuities should be decided before the test is run. If your documents set no limit, a cosmetic rejection is an opinion, and the coater will usually win the argument. The table below is a starting point to adapt into your own acceptance criteria.
| Finding | Functional surface (corrosion or sealing) | Cosmetic significant surface | Non-significant surface |
|---|---|---|---|
| Pinhole to bare metal | Reject or recoat with degas | Reject if visible at the agreed distance | Accept unless the corrosion requirement covers it |
| Blisters | Reject | Reject | Reject above the agreed ASTM D714 rating |
| Fish eyes and craters | Reject if the crater reaches metal | Reject above the agreed count per surface | Usually accept |
| Flaking or tape-test failure | Reject the lot; strip and recoat | Reject the lot | Reject the lot (adhesion is a lot property) |
| Bare or thin edges | Reject if below the minimum thickness | Reject if visible | Accept unless corrosion requirement applies |
| Orange peel, dirt, minor gloss variation | Accept | Judge against the limit sample | Accept |
| Film thickness out of range | Reject if low; reject if high where fits or threads are affected | Reject if low | Judge against the drawing range |
| Chips matching packaging contact | Touch-up only if the drawing allows it | Reject; correct the packing method | Touch-up if allowed |
Rework deserves caution. Touch-up paint seldom matches a cured powder film in gloss or durability, so allow it on the drawing only for named non-significant areas. Stripping and recoating restores appearance but repeats the cure cycle, which can shift color and does nothing about a substrate that is still releasing gas unless the parts are degassed first. A use-as-is decision belongs to your design authority; record it as a deviation tied to the lot so the next lot is not accepted on the same basis by default.
What should the drawing, RFQ and purchase order say so the same defects do not come back?
Write down the surface classes, the viewing conditions, the allowed count of each defect per surface, the film thickness range and where it is measured, the adhesion and cure tests, and the edge condition you need before coating. Then approve a limit sample at first article so both sides judge the same thing.
- Surface classes. Mark significant (cosmetic) surfaces and functional surfaces on the drawing; everything else is non-significant. Inspectors cannot reject against a class that was never defined.
- Viewing conditions. State the viewing distance, angle, lighting and inspection time for each class. Typical industrial practice is to inspect cosmetic faces at a fixed distance under diffuse light with a limited time per face, so that a defect a trained eye cannot find under those conditions is not grounds for rejection.
- Defect counts. For each class, give the maximum number and size of pinholes, craters and inclusions per face, and state that blisters, flaking and bare metal are not permitted on functional surfaces.
- Film thickness. Give a minimum and maximum, the measurement method per ASTM D7091, and the number and location of readings, as explained in the powder coat thickness guide. A single nominal value invites both thin edges and fit problems.
- Adhesion and cure. Name the ASTM D3359 method and minimum class, and a cure check such as a solvent rub per ASTM D5402 with the supplier’s rub count. The adhesion guide covers how to pick the grade.
- Edges and masking. Require a broken or radiused edge on coated edges that need protection, because a knife edge cannot hold film; name masked threads, bores and grounding points as in the masking guide; and say where hang points and contact marks may go.
- Substrate notes. For die castings, galvanized steel or weldments, ask for a degassing pre-bake on the routing. For machined parts, require cleaning to remove coolant before pretreatment.
- Packing. Specify separators or individual wrapping for cosmetic parts, since chips from part-to-part contact are the cheapest defect to prevent.
A finish note can then read, for example: “Powder coat, color and gloss per note 5. Film thickness per ASTM D7091 at points A to D, minimum and maximum per table. Adhesion per ASTM D3359 Method B, class 4B minimum. Surfaces marked S: no pinholes, craters, blisters or inclusions visible at the stated viewing conditions; limit sample approved at first article. Edges marked E: break all edges before coating. No touch-up on surfaces marked S.” Adjust each value to your product; the methods come from the standards, but the limits are yours to set.
How are powder coating defects handled on a Saola order?
On a Saola order, metal parts are machined at our lead CNC plant or made at a partner plant we manage, and powder coating is applied by a partner finishing plant we manage; the quote names the production site and country of origin for each step. Defects are judged against the surface classes, limits and limit samples on your drawing and purchase order.
CNC turning and milling run at our lead CNC plant, T.K Viet Nam Precision Company Limited, in operation since 2017, described on the CNC milling page. Powder coating is done by partner plants we manage, as listed on the surface finishing page; partner plants are not named and are not necessarily in Vietnam, which is why each quote states where every step is done. That split maps directly onto the atlas above: edge breaks, deburring, and removing coolant before parts leave for finishing sit on the machining side, while pretreatment, film build, cure and booth cleanliness sit on the finishing side. Castings and stampings come from partner plants as well, so pinholes from casting porosity are taken up with the plant that cast the part.
In practice, send the finish callout with surface classes, viewing conditions, defect limits and the test methods you want, and attach any limit samples or written cosmetic standard; the quality page explains that inspection is built around the cosmetic acceptance sample or written standard you supply. You review the first article and any cosmetic samples before production. If a lot shows a defect, send photos with a scale, the location on the part and the lot identity, and we take it up with the plant that did the work, using the checks in this guide to separate a substrate cause from a coating cause. ISO 9001 certification is in progress, not held. Send the drawing and finish notes through the quote form.
Frequently asked
What is the difference between a pinhole and a fish eye in powder coating?
A pinhole is a small, deep hole usually caused by gas escaping from the substrate during cure, and it tends to recur in the same spot on every part. A fish eye is a wider, shallower crater where the powder pulled away from contamination such as oil or silicone, and its location varies from part to part.
Why does powder coating bubble or blister?
Bubbles that appear in the oven usually come from gas or liquid in the part, such as casting porosity, weld seams or water trapped after pretreatment. Blisters that appear later, in humidity or salt spray, usually point to salts or contamination left under the film by weak cleaning or rinsing.
Can powder-coated parts with defects be recoated over the old film?
Sometimes, for cosmetic defects, if the first coat is fully cured, well adhered and the powder supplier approves recoating. It does not fix adhesion failures or outgassing; those parts need stripping, and gassing substrates need a degassing pre-bake before the new coat.
Why is powder coat always thin on the edges of my parts?
Melted powder pulls back from sharp edges as it flows, so a knife edge keeps very little film. Specify a broken or radiused edge on the drawing for edges that need protection, and check edge film thickness at named points.
Is orange peel a reason to reject powder-coated parts?
Only if it exceeds what your documents allow. Some orange peel is normal for powder, so approve a texture limit sample at first article and judge later lots against it under the same lighting.
Sources
- ASTM D714-25, Standard Test Method for Evaluating Degree of Blistering of Paints, ASTM International, 2025
- ASTM D5162-24, Standard Practice for Discontinuity (Holiday) Testing of Nonconductive Protective Coating on Metallic Substrates, ASTM International, 2024
- ASTM D7091-22, 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
- ASTM D3359-23, Standard Test Methods for Rating Adhesion by Tape Test, ASTM International, 2023
- ASTM D5402-19, Standard Practice for Assessing the Solvent Resistance of Organic Coatings Using Solvent Rubs, ASTM International, 2019 (reapproved 2024)
- ASTM D1730-09(2020), Standard Practices for Preparation of Aluminum and Aluminum-Alloy Surfaces for Painting, ASTM International, reapproved 2020
- ASTM D7803-25, Preparation of Zinc (Hot-Dip Galvanized) Coated Iron and Steel Product and Hardware Surfaces for Powder Coating, ASTM International, 2025
- Kevin Biller, Don’t Blame the Powder, Powder Coated Tough magazine, February 1, 2014