Why powder coating fails on outdoor enclosures

A canopy that blisters at fourteen months is rarely a powder problem. Powder is a commodity; the same polyester grade is used by every coater in the country. The answer to why powder coating fails on outdoor enclosures is almost always upstream of the booth: what happened to the steel in the twenty minutes before it went in, and what the metal temperature was when it came out of the oven. Those two variables decide whether the film survives two monsoons or lifts in patches.

Diagnose from the pattern on the part, not from the powder brand. It is written for the design engineer specifying a finish on a genset canopy, an outdoor electrical enclosure or a telecom cabinet, and for the quality engineer who has to write an acceptance criterion that can actually be measured. Every mode below has a physical cause, a number that controls it, and a callout that belongs on the drawing.

Five ways powder coating fails on an outdoor enclosure

Most site complaints fall into one of five patterns. Identifying the pattern first saves a lot of arguing about powder brands.

Blistering in patches on a flat face. The film lifts in domes, often 2–5 mm across, with liquid or rust under it. This is soluble salt or moisture at the interface. The salt came from the pretreatment rinse, from handling with bare hands, or from a part that was coated damp.

Rust creeping from a laser-cut edge or a hole. The face is intact and the edge is orange. Edge coverage is the weakest point on any coated part: powder is applied as a charged cloud and the field is weakest at a sharp corner, so film build drops there.

Pinholes clustered around welds and corners. Small craters, sometimes with a grey ring. Gas is escaping from under the film during cure. Welds, castings and galvanised surfaces trap air and moisture in pores; heat drives it out through the powder before the film gels.

Chalking, colour shift and a matt finish after one summer. The film looks sound but has lost gloss and the colour has moved. This is undercure or the wrong resin for the UV load. The powder never fully cross-linked, so the surface degrades.

Cracking and chipping at fixings and bends. The film is too thick or too brittle, or the part was coated before bending, and the film fractures at the deformation. A 150 µm film on a 2 mm bend will crack where a 70 µm film survives.

The table is the diagnostic order to follow on a site visit. Take a photo, measure the DFT at the failure and 200 mm away, and check whether the failure is on a face, an edge or a weld. That narrows it to one row in under ten minutes.

Salt is the enemy, and it arrives in the rinse

A dried salt film is invisible. Sodium chloride from a rinse tank, or sulphate from a drag-out of the degreasing stage, sits at a few tens of nanometres to a few microns and is enough to start an osmotic blister. Water vapour passes through the cured film, dissolves the salt, and the resulting pressure lifts the coating from the steel.

The control is rinse quality, not powder choice. In a multi-stage pretreatment line the last rinse should be demineralised water with a conductivity low enough that nothing crystallises on drying. A working limit for the final rinse is around 30 µS/cm; many lines run lower. Above that, dissolved solids are being deposited on the part every time it leaves the tank.

Two process habits defeat a good rinse. The first is a weak dry-off: a part that enters the booth with moisture in a hem or a folded flange will blister there, because the oven turns that moisture into steam under the gelling film. The second is bare-hand handling after pretreatment. Skin salts on a cleaned part produce fingerprints that blister in exactly the shape of the hand, which is a memorable and expensive lesson.

Edge coverage: the number that predicts rust lines

Film thickness on a sharp edge is not the same as film thickness on the face. The electrostatic field that pulls powder to the part concentrates on flat areas and falls away at corners, so a face reading 70 µm might read 35–45 µm on a sheared edge. That is below the point where the film provides a continuous barrier.

The geometry makes it worse. Take a 1,500 mm × 1,000 mm canopy panel. Both faces give 3.0 m² of coated area, and the perimeter is 5,000 mm. If a 1 mm strip of that perimeter is effectively bare because the film is too thin to cover it, the exposed area is 5,000 mm², or 0.005 m². That is 0.17% of the total area, and it is where every rust line on the panel will begin.

The practical controls are:

• Specify a radius on any edge that will be handled, typically 0.5–1.0 mm, and deburr sheared edges before coating.
• Blast or mechanically prepare edges where the service environment is severe. The 16 ft × 16 ft chamber takes full canopy panels.
• Ask for edge coverage to be checked on the edge as well as the face. A thickness gauge reading on an edge is difficult, but a visual check under magnification at 10× will show a thin or discontinuous film immediately.
• Avoid specifying a film so thick that it bridges and then cracks at the edge. Thicker is not better past about 100–120 µm.

Laser-cut edges deserve a specific note. The cut face carries a thin oxide layer that pretreatment chemistry does not fully remove, and oxide under a film is a weak boundary. Where the edge is visible or the part is outdoor, blasting the edge before the seven-tank line gives the conversion coating something to bond to.

Undercure, and why the oven dial is not the answer

Powder cures by cross-linking at a metal temperature, not an air temperature. A polyester powder might specify 180 °C for 15 minutes or 200 °C for 10 minutes at the substrate. The oven display tells you the air in the chamber; it does not tell you what the thickest section of the part reached, or when.

Mass matters. A 1 mm panel reaches oven temperature within a couple of minutes. A 6 mm bracket on the same hang, or a frame with a 10 mm base plate, can lag the oven by four to six minutes. If the line is set for a ten-minute cycle based on the panel, the bracket is undercured. The film will pass a fingernail test, pass a visual check, and chalk within a year because the cross-link density never reached its design value.

The controls are part temperature measurement with a data logger on the first article of each part family, a cure schedule written against the heaviest part on the hang, and a booth discipline that does not put a 10 mm plate on the same hook as a 1 mm skin.

Overcure is a failure too. Too long at temperature makes the film brittle, yellows light colours and can degrade adhesion on some substrates. A cure window is a window, in both directions.

Film thickness: a range, not a minimum

A drawing that says "powder coat 80 microns minimum" invites 150 µm, because the coater protects himself against a thin reading. Thick films chip at fixings, crack at bends and cost more powder. Specify a range: 60–80 µm for a standard outdoor canopy, with 60 µm as the floor at any measured point and 120 µm as the ceiling.

Measure on the flat faces where the gauge is reliable, at a defined number of points per part, and record the lowest reading rather than the average. One thin point is where corrosion starts, and an average hides it.

Writing a coating specification that can be inspected

A finish specification is only useful if it can pass or fail. Four elements make that possible.

1. Substrate and preparation. Name the steel, the thickness, and whether blasting is required before pretreatment. A galvanised part needs a different preparation from CRCA and should not be treated as the same line item.
2. Coating system. Powder or wet paint, resin family where it matters for UV, colour to a RAL number or an approved sample, and gloss level.
3. Film build. Target range plus minimum at any point, and the faces covered: exterior only, or both faces.
4. Acceptance tests. DFT gauge, cross-hatch adhesion, visual standard, and a salt spray duration if the environment demands it.

Salt spray deserves care. A requirement of "500 hours" without a rating scale is not a specification, because every laboratory will report it differently. State the test method (ASTM B117 or ISO 9227 neutral salt spray), the duration, and the acceptance rating on a recognised scale such as the ISO 4628 blister and rust ratings. That turns a claim into a criterion.

For environment classification, ISO 12944 defines corrosivity categories from C1 indoors to CX offshore. A canopy in an inland industrial city is typically C3; coastal installations move to C4 or C5 and need a thicker system, a primer, or both. Matching the system to the category is the fastest way to avoid over-specifying a shed in Nagpur and under-specifying a cabinet in Mumbai.

What Ganesh runs

Coating at Ganesh is a single in-house route with no transport between steps. Parts go from fabrication to shot blasting in a 16 ft × 16 ft chamber where the specification calls for it, then into seven-tank pretreatment, then to either the 18 ft × 14 ft electrostatic powder booth or the wet paint booth for primer and topcoat work, then to the curing oven. The paint shop handles canopies, frames, enclosures and furniture parts on the same line, which is why a 2,000 mm frame and a small bracket can be specified together.

In-process checks are dry film thickness with a gauge and cross-hatch adhesion testing. Both are recorded against the batch, so a question about a specific panel can be answered from the record rather than from memory. The effluent treatment plant runs under ISO 14001:2015, and the quality system under ISO 9001:2015, across five units in Mumbai (Vasai) and Nagpur.

The paint shop capability page lists the tanks, booths and oven in detail. The powder coating cost estimator gives an order-of-magnitude figure based on part area and film build, which is useful before a drawing exists.

When powder is the wrong finish

Powder is not the answer to every outdoor problem, and saying so is cheaper for everyone.

If the part is galvanised and will be powder coated, the zinc surface needs a sweep blast and usually a primer, because zinc outgasses and reacts with some pretreatment chemistries. If the environment is C5 coastal with salt fog and mechanical abrasion, a multi-coat wet paint system with a zinc-rich primer is a better fit than a single powder film. If the customer needs a specific aerospace or defence-approved coating scheme, that scheme names products we do not run, and the part should go to a specialist coater.

We do not offer cathodic electrocoating, hot-dip galvanising or fluoropolymer architectural finishes in-house. Where a specification calls for those, they are bought in or the part is redirected, and the drawing should say which.

Drawing fields to release with the part

• Steel grade and thickness, and whether the part is CRCA, mild steel, stainless or aluminium.
• Corrosivity category or a plain description of the service environment.
• Powder or wet system, colour reference and gloss.
• DFT range with a minimum at any point, and which faces are coated.
• Blasting requirement, and edge radius or deburr instruction.
• Masking plan for earth pads, threads, machined faces and hinge lands.
• Adhesion test method and frequency.
• Salt spray requirement with method, duration and rating scale, if any.
• Packing method, because a coated canopy that arrives with strapping marks has failed before installation.

A coating quotation is mostly a function of area, preparation level and film build, not of colour. Two quotes that differ by a wide margin are usually quoting different preparation, and the pretreatment process guide explains what the tanks actually do. If the failure is already visible on your parts, the degreasing guide and the phosphating guide diagnose the earlier stages, and the surface treatment hub collects the cluster.

Send the enclosure drawing with the environment and the acceptance test written on it, and the route can be quoted against a real specification through request a quote.