Every finishing line has a reject bin, and almost every argument about powder coating defects starts in the same place: the coater says the powder is bad, the powder supplier says the pretreatment is wrong, and the purchasing manager just wants to know who is paying for two hundred parts. In twelve years of sending parts through both our own line and outside coaters, we have found that one question settles most of these disputes faster than any laboratory analysis: does the defect follow the part, or does it follow the batch? Once you know which, you also know whose drawing has to change. This guide walks through the common defects by root-cause family rather than alphabetically, because the fix depends on where the defect enters the process and not on what it looks like.
Key takeaways
- Ask one question first: does the defect appear on the same features of every part, or did it start appearing partway through a run? Feature-repeating defects are geometry problems; run-dependent defects are process problems.
- Geometry-driven defects — Faraday cage shadowing in corners, thin coverage on sharp edges, back ionisation in deep recesses — are designed into the part and cannot be sprayed away reliably.
- Substrate-driven defects begin before coating exists: outgassing from porous metal, trapped moisture after pretreatment, weld porosity, and laser-cut edge scale that survives cleaning.
- Cure defects — orange peel, brittle film, poor adhesion — are usually a temperature-at-the-part problem, not a nominal oven setpoint problem. Profile the part, not the air.
- Contamination defects need a factory-wide answer. Silicone from a hand cream three metres away will find a way into a booth.
- Earthing and hanging defects are usually a bad path to ground, not a bad gun: check continuity from rail through hook to bare metal before you touch the settings.
- Half of this is preventable at the drawing stage: edge radii, drain holes, hang points, recess depth, and sealing every place moisture can hide.
- Write acceptance into the PO before the first piece: viewing distance, lighting, which faces matter, and signed limit samples.
A powder coating defect is any break in film continuity, uniformity, adhesion or appearance that falls outside the agreed acceptance criteria for the part. In practice it is classified by where it enters the process — geometry, substrate, cure profile, contamination, or earthing and hanging — because the fix depends on the origin, not on what the blemish looks like. That last clause matters more than the defect list — without acceptance criteria, everything is arguable and nothing is rejectable. Our surface finishing guide sets out where powder sits among the options before you start diagnosing it.
The one question that splits the blame
Before you open a single bag of powder, sort the symptom:
| Observation | What it means | Who owns the fix |
|---|---|---|
| The defect appears in the same place on nearly every part | A feature of the geometry or of the substrate — consistent input, consistent output | The designer, or whoever specified the material |
| The defect appeared partway through a run | Something moved in the process: gun, oven, humidity, contamination, or powder batch | The coater |
| The defect follows a powder batch but not a run | Contaminated, degraded, wrongly stored or wrongly specified powder | Supplier, with the coater's records as evidence |
| The defect appeared one season and vanished the next | Usually humidity or substrate temperature | Coater environment control |
Most disputes we sit in resolve in about ten minutes once someone photographs three parts from three different batches under the same light.
Master table: appearance, origin, first checks
| Defect | What you see | Most likely origin | First three things to check |
|---|---|---|---|
| Orange peel | Bumpy, orange-skin texture in an otherwise continuous film | Cure profile or film thickness outside the powder's window | Part temperature profile through the oven, measured film thickness, whether the line stopped mid-cure |
| Pinholes / craters | Tiny holes or blow-outs, often after cure | Gas leaving the substrate during cure | Substrate (cast, galvanised, porous weld), drying after pretreatment, whether a pre-bake would have released it |
| Fisheyes | Small circular craters where powder refuses to wet out | Surface contamination — oil, silicone, release agents, fingerprints | Compressed air cleanliness, handling gloves, anything silicone within metres of the booth |
| Faraday shadowing | Thin or missing coverage in corners and deep recesses | Electrostatic field repelling powder from enclosed areas | Recess depth versus opening, whether the same feature fails on every part |
| Edge pull / picture framing | Thicker film at the edge with thinning immediately beside it | Charge concentration at edges | Edge radius, gun angle at edges, whether parts hang with edges square to the guns |
| Back ionisation | Grainy, sandpaper texture or tiny craters in thick film areas | Excess film thickness and poor charge dissipation | Measured film build, part earthing continuity, gun-to-part distance |
| Poor adhesion | Coating flakes at a scratch or a bend | Pretreatment, contamination, or under-cure | Pretreatment chemistry and its maintenance records, part temperature through the full dwell, bar-to-hook contact cleanliness |
| Thin overall coverage | Colour translucent, substrate showing through | Transfer efficiency, poor earthing, wrong gun parameters | Earthing continuity from rail to part, hook cleanliness, kV and powder flow settings |
| Colour or gloss drift between batches | Parts do not match earlier deliveries | Different powder batch, different cure, different film thickness | Batch numbers, whether film thickness lands in the same window, measured gloss against a reference panel |
| Contamination specks | Fibres or foreign colour in the film | Booth hygiene, recovery system filters, environment | Booth cleaning between colours, filter condition, operator clothing, adjacent processes |
Note the pattern in the third column: five of the ten rows are decided before the part ever reaches a spray gun.
Geometry-driven powder coating defects
The difference from every other family in this guide: these repeat on every part, including the ones your best operator sprayed.
Faraday cage shadowing
The electrostatic field that attracts powder to the part also prevents it from entering a recess whose opening is small relative to its depth. The deeper and narrower the pocket, the less powder reaches the internal surfaces — and the field gets worse, not better, if the operator raises the voltage to compensate.
The countermeasures are all aimed at the geometry first: open the recess, widen the entry, add a second opening, or accept that the inside of that pocket needs a manual touch-up pass or a different coverage strategy agreed in advance. Our design-for-the-powder-line guide goes through the line station by station, including which recess geometries we warn customers about at quoting stage.
Edges are where film goes to die
Sharp edges are excellent corrosion starters and poor coating receivers. A defined edge radius costs nothing at the laser and saves rework at the booth. Trace enough "the coating failed early" investigations back and a large number of them end at the same place: a ninety-degree edge nobody radiused. A defined edge break and deburring requirement on the drawing is the whole fix. If you need a reference for how much appearance you are buying with which specification, our cosmetic surface class guide explains how the same edge gets graded differently depending on who sees it.
Back ionisation: when thick film fights itself
Thick film on a small-radius feature is not extra protection; over-thickness can be its own defect family, producing graininess, pinholing and poor mechanical performance. The means are the process settings and the part earthing, but the trigger is usually a local geometry that the reciprocator cannot treat evenly.
Substrate-driven powder coating defects
Outgassing from porous metal
Castings, hot-dip galvanised steel and certain porous welds hold gas — the same substrate question you meet when choosing a corrosion-protection coating system. Put them in an oven and the gas leaves through the film while the film is molten, leaving pinholes — small blow-outs through the molten film — behind. The remedy is upstream of the gun: a pre-bake above the coating's cure temperature to drive the gas out, plus a powder chosen for outgassing tolerance where the supplier offers that option. Neither is something an operator can invent on the day.
Moisture and pretreatment carry-over
Water trapped in a folded seam, a blind hole or a hollow section does exactly what the casting does — except it arrives later in the cure cycle, which is why the evidence never quite matches the outgassing pattern. This is why our own hang-point and drainage rules are non-negotiable on enclosures: every place that can hold rinse water must have a way to let it out, and the hang orientation has to let it out while the part is still moving.
Weld and cut-edge condition
Weld spatter and laser-cut edge scale are both coating defects waiting to happen — the first is a welding quality issue and the second a cutting-parameter issue, and neither is the coater's to fix. The first shows as discrete pops in the film; the second as a line of adhesion failure exactly along the cut. Both are removed mechanically before pretreatment — which means the fabrication drawing has to require it, not merely the quality plan, and somebody has to be named. The same rule applies to the joint surface condition on machined and plated parts that arrive for coating.
Process-driven powder coating defects
Orange peel is a curve problem, not a temperature problem
The powder has to melt, flow and level before it gels. Whether it does depends on how fast the part reaches its cure window, which can be very different from what the oven controller displays. A slow ramp, a heat spike, too thick a film, powder stored past its life in a hot warehouse, and a line that stopped while parts were inside the oven all produce a version of the same appearance.
The diagnostic we recommend is simple and cheap: put a temperature logger on a representative part and run it through the cycle. Air temperature profiles are useful for oven commissioning; they will not tell you what a two-millimetre section reached while a six-millimetre boss forty millimetres away was still climbing.
Line stoppages
Every stop inside a curing oven damages a batch in ways that will not be visible until the film cools. Ask for the stoppage log correlated to batch identity. If your coater cannot produce one, that is itself the finding — and you should say so, in writing, before the analysis bill is agreed.
Contamination: the factory-level answer
Fisheyes and cratering nearly always trace to silicone, oil or some other contaminant sitting at a lower surface tension than the molten film around it — the powder pulls away from that spot and leaves a crater. The sources are distributed: compressed air lines, operator hand products, release agents used in an adjacent shop area, even certain lubricants on the conveyor chain. Fixing it means auditing the whole route from compressor to part, filtering air at the gun, separating compatible processes, and — unglamorously — managing consumables.
Earthing and hanging defects
Thin overall coverage is far more often an earthing failure than a gun failure. Powder transfer depends on the electrostatic charge on the part, and that charge only exists if there is a continuous conductive path from the rail through the hook to bare metal. Insulating accumulation on hooks is the usual culprit, and it builds up silently over a shift. Check the path with a meter rather than trusting that yesterday's reading still holds.
Hook marks also belong here: whatever the hook touches is a hairline of bare metal unless a touch-up strategy or a sacrificial contact point was agreed. On a steady programme that is the point at which dedicated hanging and racking fixtures stop being a cost and start being a defect reduction. Where those marks sit is decided at the drawing stage, which is the same message as our surface class article — put the contact decisions in writing before the first part hangs.
What the designer can still change
Once a part is sitting in the reject bin there are only two moves left: change the process and hope, or rework. Everything cheaper than those two was available at drawing release:
- Edge radii — every external edge gets a defined minimum radius.
- Drainage — every cavity that can hold pretreatment rinse has a drain path, sized and located so how it drains in the hang orientation is predictable.
- Recess depth — every pocket is shallow relative to its opening, or it is flagged as needing manual coverage.
- Hang points — contact locations are chosen where the mark is invisible in use, and they are shown on the drawing.
- Material callout — castings and galvanised parts are identified up front so the pre-bake step is quoted rather than discovered.
- Finish specification — film thickness window, gloss target reference panel, and what the coater must log per batch. If you are still choosing between processes, our powder coating versus liquid painting comparison covers what each one forgives.
Acceptance criteria: write them before the first piece
None of this is rejectable unless you said what "acceptable" means. Three items settle it:
- Viewing conditions — distance, time, lighting and angle. Without these any defect can be argued away by taking one step backwards.
- Zoned drawing — which faces carry which class, so nobody inspects the concealed plinth as though the customer were looking at it.
- Limit samples — one best-acceptable part and one worst-acceptable part, signed and photographed, held on both sides. Our QC inspection guide has a template that accommodates appearance alongside dimensional checks, and our salt spray testing guide explains how to read a corrosion rating once you start testing rather than looking.
Add a requirement for a retained first piece per batch with the coater's parameters logged — the finishing equivalent of a first-article inspection record — and the next investigation takes an afternoon of filing instead of a fortnight of reconstruction.
Where you want the criteria to be testable rather than argued, name the method as well as the limit: dry film thickness by ASTM D7091 on ferrous substrates or ISO 2360 on non-ferrous, cross-cut adhesion by ASTM D3359 or ISO 2409, gloss by ASTM D523, impact by ASTM D2794 and bend by ASTM D522. ASTM D3451 is the umbrella guide that collects the powder-coating test procedures, and it is the reference to cite when you and your coater disagree about what a test means. Film thickness and cure windows are powder-specific: quote your supplier's technical data sheet, not a figure from an article — including this one.
Frequently Asked Questions
Who is responsible when powder coating defects appear — the coater or the designer?
It depends on whether the defect tracks the part geometry or the production run. If the same feature fails on nearly every part regardless of who coated it and when, the cause is in the design or the substrate, and the drawing has to change. If the defect started at a particular time or follows a particular batch, it is a process or material variable and belongs to the coater's or supplier's records. Sorting the symptom this way usually resolves the blame question faster than any laboratory test, and you can do it with photographs before anyone pays for analysis.
Can a deep recess actually be powder coated properly?
Not reliably by automatic spraying alone, because the electrostatic field that attracts powder to the part also prevents it entering a deep, narrow pocket. The durable fixes are geometric: open the recess, widen its entry, reduce its depth relative to the opening, or add a second access. Failing that, agree a manual touch-up strategy and a reduced coverage expectation for that area before production starts. Spraying harder usually makes it worse rather than better.
Why does my powder coating look like orange peel?
Orange peel means the film did not level before it gelled, and the usual suspects are all downstream of the setpoint: part heating too slowly or too suddenly, film thickness above or below the powder's specified window, degraded powder that has lost its flow, or a line stoppage while parts were inside the oven. The useful diagnostic is to profile the part temperature through a real cycle rather than trusting the oven controller's air reading, because different sections of the same part reach the cure window at very different times.
How do I stop pinholes from appearing after coating?
Pinholes usually mean gas or moisture left the substrate during cure. Check whether the substrate is porous — castings, hot-dip galvanised steel and porous welds are the usual candidates — whether pretreatment rinse has been fully dried and drained, and whether the part has any folded seam or hollow section that can trap water. A pre-bake above the powder's cure temperature, plus selecting a powder specifically formulated for outgassing forgiveness, addresses most of it. Both need to be priced at quotation, not discovered on the day.
What film thickness should I specify for powder coating?
Do not specify a number from an article — specify the window on the powder supplier's technical data sheet for the exact product you are buying, and state it as a range measured on the part, not an oven setpoint. Then say where it is measured, because thickness on a flat face and thickness at an edge or in a recess are different numbers on the same part, and the thin spot is the one that fails.
What should I put in a purchase order to make coating defects rejectable?
Three things. Stated viewing conditions (distance, time, lighting and angle); a zoned drawing showing which face carries which appearance class so concealed surfaces are not inspected as though they were customer-facing; and signed limit samples held on both sides — one best-acceptable part, one worst-acceptable part, with the defects on the second one ringed and explained. Add a requirement to retain a first-piece per batch with the coater's parameters logged, and root-cause investigations become an afternoon's work.