A corrosion protection coating is the cheapest insurance on a sheet metal enclosure and the easiest thing to specify wrong. A SanJun customer once shipped 200 telecom cabinets to a Mediterranean port with a powder coating rated for 500 h salt spray. The coating performed exactly to its data sheet; the data sheet was the wrong one for that site. Eighteen months later the field team was sending photographs of red rust blooming along the bottom seam. This guide covers the corrosion mechanisms that actually destroy enclosures, how to rate the site before you pick a coating, what each of the four workhorse coating systems really buys you, why the salt-spray number on the data sheet will not tell you a service life, and where coated enclosures fail first — which is almost never the middle of a flat panel.
Key takeaways
- Most field failures are galvanic or crevice corrosion, not uniform corrosion. A continuous film across the whole part matters more than the salt-spray number on a flat coupon.
- Rate the site before you pick the coating. The ISO 9223 / ISO 12944 corrosivity categories run from C1 (heated indoor) to CX (offshore splash zone), and the category — not the alloy name — decides the coating system.
- Salt spray hours are not service years. ASTM B117 and ISO 9227 are method standards: neither sets a duration or a pass/fail criterion. Those come from your drawing.
- Coated enclosures fail at edges, welds, seams and fasteners. Powder coat pulls thin at a sharp edge, and a weld that is not cleaned will rust under any coating you put on it.
- SanJun supplies the metal, the machining and the supporting evidence. We are a sheet metal fabricator: we do not issue listings or certificates of compliance, and we do not run UL 9540A.
Why does a coating fail when the data sheet was followed?
Because the data sheet answered a question nobody asked. A supplier certificate that says "passed 1,000 h salt spray" is a statement about a flat test coupon in a cabinet at 35 °C, exposed to a continuous 5% sodium chloride fog with no drying stage, no ultraviolet light and no mechanical damage. A cabinet on a coastal substation sees none of those conditions and all of the ones the test omits.
Three gaps explain almost every premature corrosion failure we are asked to investigate:
- The specimen was not the part. A flat coupon measures the coating system. A production-representative part — with its edges, welds, hems and fastener holes — measures the manufacturing system. Only the second one predicts the field.
- No failure definition was written. "1,000 h salt spray" without a rating method and a threshold is not a requirement. It is a hope. Our corrosion testing guide goes through what to write instead.
- The environment was never rated. The same coating system that lasts twenty years indoors fails in nine months on a North Sea turbine. Nobody wrote down which one this site was.
The fix is not a thicker coating. It is a specification that names the site category, the coating system, the surface preparation, the test method and the failure definition — in that order.
Which corrosion mechanisms actually matter for sheet metal?
Four mechanisms account for nearly all sheet-metal corrosion failures, and only the first is what people picture when they hear the word "rust".
- Uniform atmospheric corrosion — slow, predictable oxidation across an exposed surface. Salt fog and humidity accelerate it. It is the easiest to see and the least likely to kill a product.
- Galvanic (bimetallic) corrosion — two dissimilar metals in contact in the presence of an electrolyte. Carbon steel bolts into a stainless enclosure in a wet site is the textbook case, and it destroys the fastener, not the panel.
- Crevice corrosion — under gaskets, inside fastener heads, and at folded seams where moisture sits and cannot evaporate. Often invisible until the part is disassembled.
- Poultice corrosion — the same mechanism under a deposit: dust, salt crust, or a label that holds damp against the surface.
The practical consequence is that film continuity beats film thickness. A continuous coating across the whole part — including the edges, the inside of the hem and the area under the gland plate — outperforms a thicker coating with a gap at the seam. That is also why the details in our condensation control guide matter as much as the coating choice: time-of-wetness is the dominant variable, and a cabinet that dries quickly between wet cycles corrodes far less than one that stays wet.
How do you rate the environment before choosing a coating?
Rate the site first, then pick the substrate, then pick the coating. The standard severity ranking comes from ISO 9223 and ISO 12944, and it is the shared language worth putting on a drawing:
| Corrosivity category | Typical environment | What it means for the enclosure |
|---|---|---|
| C1 (very low) | Heated indoor spaces, dry atmosphere | Direct contact of dissimilar metals acceptable; standard paint is adequate |
| C2 (low) | Unheated indoor, temperate rural outdoor | Direct contact acceptable; galvanized steel is fine |
| C3 (medium) | Urban or industrial outdoor, coastal with low salinity | Insulate dissimilar-metal joints; drainage design starts to matter |
| C4 (high) | Industrial or coastal with moderate salinity | Insulation mandatory; powder coat or anodize aluminium |
| C5-I (very high, industrial) | Heavy industrial, high humidity | Insulation plus coating plus drainage; consider 316L |
| C5-M (very high, marine) | Coastal or offshore with continuous salt exposure | 316L body; insulation mandatory; consider a sacrificial anode |
| CX (extreme) | Offshore splash zone, subsea, chemical immersion | Specialty alloys or non-metallic composite |
Two engineering rules apply across every category. First, time-of-wetness dominates: drainage design matters more than coating choice at C3 and above, because a joint that dries in an hour corrodes far less than one that stays damp for a week. Second, the galvanic risk scales with the category — a pair of metals that is harmless at C2 becomes a maintenance item at C4. The coastline case, including which stainless grade to use at which distance from the shore, is worked through in our coastal enclosure guide.
An EV charging customer specified 304 stainless for a site 400 m from a Mediterranean shoreline and got eighteen months before tea-staining appeared on the door. The alloy was right for C3. The site was C5-M. 316 would have cost 8% more on the bill of materials and nothing at all on the tooling — the decision was made once, and it was made from the drawing rather than from the site.
Which coating suits steel enclosures?
Four systems cover almost every steel enclosure we build. The two numbers that separate them are film thickness — how much barrier you have — and salt spray behaviour — how that barrier ages under a corrosive atmosphere. Neither one gives you a service life.
| Coating system | Substrate | Typical thickness | Salt spray (ASTM B117) | Best environment |
|---|---|---|---|---|
| Powder coating (TGIC polyester) | Steel / aluminium | 60–120 µm | 500–1,000 h | Indoor and outdoor non-coastal |
| Zinc plating, clear chromate | Steel only | 5–25 µm | 96–120 h to white rust | Indoor, mild humidity |
| Hot-dip galvanizing | Steel only | 50–100 µm | 1,000–2,000 h | Outdoor, structural |
| E-coat (electrodeposition) | Steel / aluminium | 15–30 µm | 500–1,000 h as primer | Primer under powder, sharp edges |
Powder coating is the default for a reason: it is thick, it covers edges better than liquid paint, and the overspray is reclaimable. Its weakness is geometry — it pulls thin at sharp edges and cannot reach into an unvented recess. Zinc plating gives sacrificial protection rather than barrier protection, which is why it works on threaded fasteners and small brackets but not as a standalone outdoor finish. Hot-dip galvanizing puts a metallurgically bonded zinc-iron layer under the coating, which survives the handling damage that would breach a paint film — at the cost of a rougher surface and a thicker dimension to design around. E-coat is the one people forget: it is applied by immersion, so it reaches into seams and recesses that powder cannot, which makes it the best primer under powder on a part with a lot of folded geometry.
For liquid paint as an alternative, and where each finish wins on cost and appearance, see our powder coating versus liquid painting guide. For the full surface-treatment menu including conversion coatings and passivation, see the surface finishing guide.
Which coating suits aluminium enclosures?
Aluminium corrodes differently. It forms its own oxide film, which is why bare aluminium survives outdoors far better than bare steel — and why the failure mode is usually pitting or galvanic attack rather than general wastage.
| Treatment | Thickness | What it buys | Watch out for |
|---|---|---|---|
| Anodizing, Type II | 5–25 µm | Hard, abrasion-resistant, dyeable surface; good on HMI bezels | Non-conductive — mask bond zones for EMI |
| Anodizing, Type III (hard) | 25–75 µm | Wear surface, thickest oxide | Requires sealing; dimensional growth; can craze on sharp bends |
| Chromate conversion (Alodine) | < 1 µm | Paint adhesion and mild corrosion resistance; conductive enough for bonding | Not a standalone outdoor finish |
| Polyester powder coat | 60–120 µm | Colour, UV stability, thick barrier | Needs a conversion pre-treatment to adhere |
Two things surprise people on aluminium. The first is that anodizing is an electrical insulator — an anodized panel bolted to an anodized enclosure body is not bonded, and if that seam was carrying your EMI shield the assembly fails emissions testing. The second is that anodizing needs sealing to deliver its corrosion resistance; an unsealed Type II layer will hold dye and then let salt in. Everything about cleaning and passivating stainless and aluminium surfaces, including what not to use, is in our passivation and weld cleanup guide.
What does the salt-spray number actually mean?
Less than most specifications assume. ASTM B117 and ISO 9227 are method standards: they define the chamber, the solution, the temperature and the collection rate. Neither one sets a test duration and neither one sets a pass/fail criterion. Those come from your drawing or your product specification, which is why "1,000 h salt spray" on a data sheet is not a claim about your product.
| Test family | Typical methods | Best used for | Weak at |
|---|---|---|---|
| Neutral salt spray (NSS) | ASTM B117, ISO 9227 NSS | Batch comparison, supplier screening, quick triage | Life prediction, real-world ranking |
| Acidified (AASS / CASS) | ISO 9227 AASS / CASS, ASTM B368 | Decorative plated finishes, fast screening | Outdoor barrier systems — frequent misuse |
| Cyclic corrosion (CCT) | ASTM G85, ISO 11997-1, IEC 60068-2-52 | Coastal, de-icing salt, long outdoor service | Cheap screening; slower and costlier |
Cyclic testing is the more honest tool for a coastal site, because it alternates a wet salt stage with humidity and a drying stage — and the drying stage is where the mechanism that actually destroys painted steel happens. It costs more and takes longer per data point, and for a twenty-year outdoor asset it is worth it.
Whatever the method, a duration is meaningless without a failure definition. Attach one of three: no red rust at all; red rust below a stated percentage of area rated per ASTM D610 or ISO 4628-3; or a maximum creep width from a scribe, rated per ASTM D1654. Blistering is rated separately under ASTM D714 or ISO 4628-2. Without one of those, almost any result can be called a pass.
Where do coated enclosures actually corrode first?
Almost never in the middle of a flat panel. Coating failures start at geometry, and the same six locations account for most of what comes back:
| Location | Why it fails | Prevention |
|---|---|---|
| Sharp edges and laser-cut corners | Powder pulls thin during cure and leaves the edge barely covered | Radius or break the edge before coating; see the edge break guide |
| Welds and heat-affected zones | Oxide scale and spatter sit under the coating; the weld is electrochemically different from the parent metal | Grind, clean and passivate before coating |
| Folded seams and hems | Trap pretreatment chemicals and water; the crevice never dries | Vent and drain the hem, or seal it after coating |
| Fastener holes and fasteners | The fastener is usually the anode; the hole edge is uncoated | Match fastener material to the site; insulate at C4 and above |
| Gland and connector cutouts | Machined after coating, or masked badly, leaving bare metal at the seal face | Mask before coating, or seal every penetration at assembly |
| Under gaskets and labels | Crevice with a long time-of-wetness | Seal the crevice; avoid upward-facing horizontal flanges |
A washdown equipment builder came to us after 40 stainless cabinets came back with rust staining around every weld. The material was 316, the coating was fine, and the problem was in the fabrication: welds had been left with heat tint and cleaned with a carbon steel brush, which embedded free iron in the surface. The cabinets were not corroding — the embedded iron was. ASTM A380 and ASTM A967 exist to prevent exactly this, and the fix cost an afternoon of training.
How does galvanic corrosion change the coating choice?
Galvanic corrosion is a design problem that a coating can only paper over. When two dissimilar metals touch in an electrolyte, the less noble one becomes the anode and corrodes at a rate set by the area ratio — a small anode next to a large cathode is the dangerous geometry, which is why steel fasteners into a large stainless panel fail so quickly.
Four controls, in the order we apply them:
- Break the circuit. Insulate the joint — a nylon washer, a gasket, an isolating sleeve. Cheapest and most effective where geometry allows.
- Coat the cathode, not the anode. Painting the noble metal reduces the cathode area and slows the reaction. Painting only the anode is worse than doing nothing once the film is scratched.
- Move the metals closer together on the galvanic series, or add a sacrificial anode that is meant to be consumed.
- Design for drainage so the joint does not stay wet. Sloped surfaces, drain holes, drip edges, and no upward-facing horizontal flanges.
The full ranking of pairs, the area-ratio arithmetic and the severity table are in our galvanic corrosion guide, and fastener material selection is covered in the fastener selection guide.
What surface preparation decides whether the coating lasts?
Preparation is where coating life is actually determined, and it is the step that gets value-engineered out. Three things matter:
- Cleanliness. Degrease and rinse. Oil left on the surface prevents the conversion coating from forming and the powder from wetting out.
- Conversion coating. A zirconium or zinc-phosphate layer gives the powder something to bond to chemically rather than mechanically. It is the single biggest difference between a coating that lasts and one that peels in sheets.
- Surface profile and grade. For blast-cleaned steel, ISO 8501-1 defines the preparation grades that coating suppliers specify their systems against. If your drawing does not name a grade, the applicator is choosing for you.
Then there is the step nobody puts on the drawing: edge and weld preparation. Grinding spatter and putting a small radius on every sharp edge costs minutes per part and removes the most common failure site in the table above. On aluminium it also removes the burrs that anodizing will highlight rather than hide.
How do you specify a coating on a drawing?
A coating callout that survives contact with a supplier has five lines. Anything less and the decisions get made for you:
- Site corrosivity category — C2, C3, C4, C5-I, C5-M per ISO 9223 / ISO 12944. This is the line that everything else follows from.
- Surface preparation — degrease plus conversion coating, or a named ISO 8501-1 blast grade.
- Coating system and thickness — for example, 80–120 µm TGIC polyester over a zinc-rich primer.
- Test method, duration and failure definition — for example, ISO 9227 NSS 1,000 h with no red rust and scribe creep below 2 mm per ASTM D1654.
- Mask and no-coat zones — bond areas for EMI, sealing faces, grounding points, datum surfaces.
Masking is the one most often missed, and it is the one that creates an assembly failure rather than a cosmetic one. A masked bond zone costs nothing to specify before coating and is very expensive to create afterwards by machining.
Those three bullets decide whether a coating adheres; they do not decide which conversion chemistry delivers them on a mixed-metal assembly, or what to require instead of a quoted salt spray figure. That layer is set out in which conversion coating sits under that coating system.
Frequently Asked Questions
How many salt spray hours do I need for an outdoor enclosure?
The number means nothing on its own. ASTM B117 and ISO 9227 define the method, not a duration or a pass/fail criterion — those come from your drawing. Rate the site against ISO 9223 first, then write a duration with a failure definition: no red rust, red rust below a stated area per ASTM D610, or scribe creep below a stated width per ASTM D1654.
Is 1,000 hours of salt spray equivalent to ten years outdoors?
No. Salt spray hours are not service years, and both ASTM B117 and ISO 9227 are explicit that the test is not intended for life prediction. Continuous fog with no drying stage, no UV and no mechanical damage does not reproduce an outdoor cycle. For coastal or long-service assets use a cyclic method such as ASTM G85, ISO 11997-1 or IEC 60068-2-52 instead.
Powder coating or anodizing for an aluminium enclosure?
Powder coat for colour, UV stability and a thick barrier; anodizing for abrasion resistance and a hard, dyeable surface. Anodizing is an electrical insulator, so any seam that has to carry an EMI bond must be masked before anodizing or machined afterwards. Anodizing also needs sealing to deliver its corrosion resistance.
Why does my powder-coated enclosure rust at the edges first?
Powder pulls thin at a sharp edge during cure, leaving the edge barely covered. Break the edge or put a small radius on it before coating, grind weld spatter, and consider an e-coat primer — it is applied by immersion, so it reaches edges and recesses that powder cannot.
Do I need e-coat under powder coating?
Not always, but it is the right call on parts with a lot of folded geometry, deep recesses or tight seams. E-coat is applied by immersion so it covers surfaces powder cannot reach, and it dramatically improves edge coverage. On a simple flat-panel enclosure a good conversion pre-treatment plus powder is usually sufficient.
Can I bolt stainless fasteners into a coated steel enclosure?
Yes at C1 and C2. At C4 and above the steel becomes the anode and corrodes, so insulate the joint or match the fastener material to the site. Remember the area ratio: a small steel anode against a large stainless cathode is the geometry that fails fastest.
Which stainless grade should I use near the coast?
304 is acceptable for sheltered inland installations with a documented surface finish. For shoreline-adjacent sites, 316 is a decision you make once. At C5-M the standard severity tables call for a 316L body with insulation mandatory, and a sacrificial anode is worth considering.
What does SanJun supply and not supply on a coated enclosure?
We build the enclosure, do the machining and supply supporting evidence, and we build to the agreed procedure. We do not issue listings or certificates of compliance, and we do not run UL 9540A. Where a project needs a witnessed third-party test or a certificate naming a specific standard, that is agreed at RFQ and run by the named laboratory.
Specification is the deliverable
The coating itself is a commodity. What is not a commodity is the specification that says which site category the product is built for, which preparation the surface gets, which system goes on it, and what counts as a failure. Write those five lines on the drawing and the rest of the supply chain can be managed. Leave them out and every supplier fills the gap with a different assumption.
Send us the drawing, the installation environment and the service-life target. We will return a coating specification — site category, preparation, system and test plan — with a quotation. For how the test gets written and run, start with our qualification test plan guide; for what the cosmetic side has to look like, see the cosmetic surface class guide.
Need a coated enclosure that survives its site? Send your drawings and environment for a coating specification and quotation within 24 hours.