Most powder coating defects are designed in, not sprayed on. A welded frame with no vent hole, a blind cavity that traps pretreatment fluid, a threaded hole the drawing never said to mask — each one is a geometry decision made weeks before the part reaches the coating line, and each one costs more to fix at the rack than it would have cost to draw. This guide walks the part through the five stations of a finishing line and states what the geometry has to provide at each: somewhere to hang it, somewhere for liquid to leave, somewhere for air to escape, a path for the powder to reach, and a clear definition of what must stay bare. We give real numbers where we have them and name the source. Two places in this article genuinely depend on your coater's rack and oven, and we say so instead of inventing a figure.
Key takeaways
- Five stations, five geometry requirements: hang (contact and ground), drain (liquid out), vent (air and vapour out), reach (electrostatic access), mask (bare where required).
- A closed cavity that traps pretreatment fluid will blow out through the film in the cure oven. The failure looks like a coating defect and is actually a missing hole.
- Masking on the drawing costs minutes; re-tapping coated threads costs far more and usually chips the surrounding film. Specify plugs, not repair.
- Deep recesses and narrow channels coat thin because of the Faraday cage effect. If uniform coverage inside a pocket is functional, say so early enough to change the geometry.
- Powder coating is an insulator. Any grounding or bonding path has to be a designed bare-metal area with a masking callout, not an assumption that a screw will cut through.
- Hot-dip galvanizing has its own vent and drain rules, and they are stricter than powder coating because molten zinc has to flow in and out.
A drain hole lets liquid out of a cavity. A vent hole lets air and vapour out. Both are needed, and both are positioned in the hanging orientation — not the orientation the part is used in.
Walk the part through the line, not through the drawing
Stop reviewing the finished part. Follow it through the line. A coating line has five stations, and each one makes a demand on geometry.
If a part fails at any station, the symptom appears at the end and gets blamed on the finisher. It is almost always upstream.
Station 1 — Hang: contact, load and ground
The part has to be suspended from a hook or a rack. That contact does three jobs at once: it carries the weight, it positions the part for the spray gun, and it provides the electrical ground that makes electrostatic deposition work.
Three consequences for the designer:
- There must be somewhere to hang from. Existing holes often work — mounting holes on a non-cosmetic face, a flange slot, a hole that will later be covered by another part. If none exists, add a dedicated hole, a break-off tab, or agree a rack fixture with the coater. Do not assume the coater will improvise well.
- The contact point will be uncoated. Powder cannot deposit where metal touches metal. That leaves a bare dot or line. Put it where nobody looks: inside a flange, under a gasket land, behind a cover, or on a face that will be hidden in assembly. Where a permanent hole is unacceptable, a break-off tab on a hidden edge leaves a small mark that can be touched up.
- The contact point must carry the load hot. At typical powder cure temperatures, aluminium — particularly heat-treatable alloys such as 6061 in the T6 temper — loses far more strength than mild steel. The hanging feature therefore needs to be sized for the part at oven temperature, not at room temperature. Shop guidance we have seen gives quite different load figures for aluminium and for mild steel at the same hole size, which is why you should confirm the limit with your own coater — alloy temper and oven dwell both move it — rather than adopt a number from an article. If the part is heavy or the hang feature is small, use multiple points or a dedicated bracket.
If your enclosure has a cosmetic class applied to its visible faces, the hang point has to be agreed against that class. Our guide to cosmetic surface classes explains why a rack mark that is acceptable on a side panel is a reject on a door skin.
Station 2 — Drain: liquid has to leave
Before coating, the part goes through cleaning and pretreatment stages involving liquids. Any geometry that can hold liquid will hold it, carry it into the next bath, and then into the oven.
Where liquid hides:
- Closed or near-closed cavities formed by overlapping seams or by welding a cap onto a tube.
- Lapped joints and hemmed edges, which hold fluid by capillary action and then weep out during cure, contaminating the film below — see our hem bend, relief and corner joint guide for the joint forms that do this worst.
- Deep pockets with the opening facing up in the hanging orientation.
- Blind internal volumes in welded box sections and frames.
The fix is a drain hole at the lowest point in the hanging orientation — not the lowest point in the assembly orientation, which is a distinction that gets missed whenever the part is hung differently from the way it is used. Place drain holes at the corners or edges of the volume, and size them so they cannot be blocked by the coating itself.
Station 3 — Vent: air and vapour have to escape
Drainage and venting are different problems and both are needed. Drainage removes liquid. Venting lets air out so liquid can get in, and lets vapour out so the film is not blown apart during cure.
The failure mode is dramatic and entirely predictable. A sealed welded tube goes through the wash, holds liquid inside, then enters the cure oven. The trapped liquid boils, expands, and forces its way out through the softening film, leaving craters or bare patches. One missing hole can ruin an entire rack.
The general rules:
- Every enclosed volume needs a vent at the highest point in the hanging orientation and a drain at the lowest.
- Both ends of a tube are better than one. Leaving tube ends open is the cleanest solution where the design allows it, and the same geometry thinking drives custom perforated vent patterns.
- More than one bay in a frame means more than one vent path. A frame with internal cross-members can have isolated volumes that each need their own hole.
For hot-dip galvanizing the requirement is stricter, because molten zinc has to flow into, through and out of the fabrication. Published galvanizing design guidance is tiered, and the tiers matter. No vent hole should be smaller than 8 mm and no drain hole smaller than 10 mm; the preferred minimums are 12 mm for vents and 25 mm for drains. Beyond those floors the hole scales with the job: hollow sections need vent and drain area equivalent to roughly 25% of the section's diagonal cross-section (about 12 mm on a small section, 25 mm around 100 mm, 50 mm around 200 mm), and closed vessels need around 1250 mm² of vent area per cubic metre of enclosed volume. Confirm the final sizes with the galvanizer. The reason they are larger than powder-coating vents is twofold: roughly 200 g of zinc ash is produced per square metre galvanized and, being a solid powder, it will not pass through a small opening — and because steel is only about 15% heavier than molten zinc, a small amount of trapped air is enough to stop the section sinking at all.
The same venting logic applies to sealed enclosures for a different reason. Our guide to condensation control covers what happens to a sealed box in service when it cannot breathe — because the two failures run on different clocks. Service condensation is vapour moving over years. A cure-oven blowout is trapped liquid expanding in about ninety seconds.
Station 4 — Reach: the powder has to get there
Electrostatic powder deposition needs a path from the gun to the surface. Geometry that blocks that path produces thin or missing coverage, and the resulting corrosion failure appears months later at the edge nobody could see.
Two mechanisms:
- Line of sight. Surfaces that the spray cannot see will not be coated properly. The inside of a long tube, the bottom of a deep blind pocket, and the back of a closely spaced pair of panels are the usual offenders. Where internal coverage is functionally required, the process may have to change — dipping rather than spraying — and that decision has to be made before the geometry is frozen.
- The Faraday cage effect. In deep recesses, narrow channels and tight inside corners, the electrostatic field concentrates at the exposed edges instead of drawing powder into the recess. The result is thin coverage exactly where corrosion starts. This happens on geometry that is otherwise perfectly reasonable for fabrication, so it cannot be designed out by a blanket rule; it has to be reviewed.
Related to both: sharp edges coat thin. Charged particles do not build evenly on a knife edge, and the edge is also the place where corrosion begins and where handling damage concentrates. Edge preparation therefore links directly to finishing quality — see our edge break and deburring guide.
Table — geometry requirements by station
| Station | Question the geometry must answer | If the answer is no | Typical evidence of failure |
|---|---|---|---|
| Hang | Is there a non-cosmetic contact point that carries the load at oven temperature? | Racking is improvised | Bare mark on a visible face; dropped or distorted parts |
| Drain | Does every cavity empty in the hanging orientation? | Fluid carries over between baths | Contamination, staining, poor adhesion below the weep path |
| Vent | Can air out and vapour escape from every enclosed volume? | Pressure builds during cure | Craters, blowouts, bare patches over welds |
| Reach | Can the gun see and reach every surface that needs coverage? | Electrostatic shadowing | Thin film in recesses and corners; early corrosion at edges |
| Mask | Are the bare-metal areas defined on the drawing? | Threads and lands get coated | Rework, chipped film, failed continuity testing |
Station 5 — Mask: define what must stay bare
Powder coating is a functional layer, and a functional layer in the wrong place is a defect. Anything that must conduct, seal, thread or locate precisely should not be coated.
The list of things designers forget to call out:
- Threads — tapped holes, studs, and the threads of installed hardware.
- Grounding and bonding lands — the bare-metal pads that carry fault current and provide EMC continuity.
- Gasket and sealing lands — coating build changes compression and can bridge a seal face; see our gasket and sealing selection guide.
- Precision bores and locating surfaces — film thickness is a dimensional change, and it is not uniform on edges and in corners, so it lands inside your tolerance stack-up.
- Mating faces that must sit flat — coating on both sides of a joint adds thickness twice.
- Hardware interfaces — the seating face of a self-clinching fastener, a hinge leaf, a latch strike.
Masking is cheap when it is specified and expensive when it is not. In the comparison published by one fabricator, plugging a batch of holes costs a few minutes of labour against roughly ten times that to re-tap them — and running a tap through a coated hole frequently chips the surrounding film, turning a threading job into a refinishing job. Threads are the clearest case, but the same arithmetic applies to every item on the list.
The insulation problem
Because the coating is an electrical insulator, the grounding clause is the one with safety consequences. A screw driven through a coated surface is not a reliable bonding path: the film is hard, it is continuous, and the contact area is uncontrolled. Where an enclosure needs protective bonding or EMC continuity, the bare-metal area has to be designed, dimensioned, masked and verified. Our grounding and bonding design guide covers the detailing.
Note the interaction with hardware as well. Whether self-clinching fasteners go in before or after coating changes both the masking requirement and the risk of coating interfering with installation; the right order depends on the hardware, the material and the function, so identify the exact part number and the sequence on the drawing. Our fastener selection guide covers the hardware side.
What to put on the drawing
Finishing requirements are cheapest when they are part of the drawing rather than an email — and they have to travel with the same revision control as everything else on that drawing. The following set covers most of what a coater needs:
- Colour and finish system — manufacturer and product code where it matters, plus a RAL reference and gloss level and texture. Budget impact sits with the rest of your enclosure cost drivers. A RAL number alone does not define the system; two powders of the same colour differ in gloss, texture, chemistry and durability.
- Hang points — location, and a note that rack contact is not permitted on the designated cosmetic faces.
- Drain and vent holes — location and size, dimensioned in the hanging orientation, with a note that they are functional and must not be filled by coating.
- Masking — every area that must remain uncoated, shown with a boundary and a leader, and listed.
- Grounding lands — location, size, and the requirement that they be tested for continuity after coating.
- Film thickness at functional features — where coating build affects assembly, say what the finished dimension is.
- Hardware sequence — installed before or after coating, with the part number.
- Reference standard — an approved colour and gloss panel, the standard the visual appraisal follows, and where the nameplate and labelling sits relative to the masked areas.
Also worth stating: which coating specification applies beyond colour. If the enclosure has a corrosion requirement, coating choice interacts with the environment and the substrate — our corrosion resistance coatings guide covers the selection, and our comparison of powder coating and liquid painting covers where each is the right answer.
One more interaction is easy to miss. Masked bare-metal areas and coated areas in contact, in a wet environment, create a galvanic couple in which the small bare-metal area becomes the anode against a large coated cathode, which is the ratio that drives rapid localised attack. See our guide to galvanic corrosion between dissimilar metals before you mask a large area inside a wet enclosure.
Send the drawing with the finishing requirements attached and we will review the geometry against all five stations before quoting. It is a free check, and it is much cheaper than the alternative.
Frequently Asked Questions
Why does a welded tube frame need vent holes before powder coating?
The part goes through liquid cleaning and pretreatment before coating. If the tube is sealed, liquid stays inside, and when the part enters the cure oven that liquid boils and expands, forcing its way out through the softening film. The result is craters, blowouts or bare patches over the weld. Every enclosed volume needs a vent at the highest point and a drain at the lowest point in the orientation the part is hung.
Where should a powder coating hang point be placed?
On a surface that is not cosmetic and that will be hidden in assembly — inside a flange, under a gasket land, behind a cover. The contact area will be bare, and the feature has to carry the part's weight at oven temperature, where metal is softer than at room temperature and aluminium loses more strength than steel. If no suitable hole exists, add a dedicated hole, a break-off tab, or agree a fixture with your coater.
How big do drain and vent holes need to be?
For powder coating there is no universal number; size them so they cannot be blocked by the coating and confirm with your coater. There is no universal number for powder coating; size holes so the coating cannot block them and confirm with your coater. For hot-dip galvanizing the published design guides are tiered: absolute minimums of 8 mm for vents and 10 mm for drains, preferred minimums of 12 mm and 25 mm, and beyond that the hole scales with the section — about 25% of the diagonal cross-section for hollow sections, and roughly 1250 mm² of vent area per cubic metre for closed vessels. Always confirm with the galvanizer.
What is the Faraday cage effect in powder coating?
It is the electrostatic shadowing that occurs in deep recesses, narrow channels and tight inside corners. The field concentrates at the exposed edges rather than drawing powder into the recess, so coverage ends up thin exactly where corrosion starts. It can occur on geometry that is otherwise perfectly sound, so it has to be reviewed rather than designed out by a blanket rule.
Why should threads be masked rather than re-tapped after coating?
In the comparison published by one fabricator, plugging a batch of holes costs a few minutes of labour against roughly ten times that to re-tap them, and running a tap through a coated hole frequently chips the surrounding film, turning a threading operation into a refinishing operation. The same arithmetic applies to every other item on the list. The same argument applies to grounding lands, gasket faces, precision bores and any mating surface where film build changes a dimension.
Does powder coating need to be kept off grounding points?
Yes. Powder coating is an electrical insulator, and a screw driven through a coated surface is not a reliable bonding path. Where an enclosure needs protective bonding or EMC continuity, the bare-metal land has to be designed, dimensioned, masked and then verified for continuity after coating.
What does a RAL number actually specify?
It communicates colour, and that is all. Two powders with the same RAL reference can differ in gloss level, texture, chemistry, durability and appearance. A complete callout needs the powder manufacturer and product code where it matters, plus gloss level and texture, alongside the colour reference.
Should self-clinching hardware be installed before or after coating?
It depends on the hardware, the material and the function. Installing before coating risks coating interfering with threads and seating; installing after coating risks damaging the film and leaves a bare interface. Identify the exact part number and the required sequence on the drawing rather than leaving it to the shop.