A chip in car paint spreads. A scratch on a galvanised gate does not. The difference is worth ₹ lakhs on a real structure

Look at a chip in the paint on a car door that has been left alone for a year. The bare spot has rusted, but that is not the interesting part — the interesting part is that the paint around it has lifted, and the rust has crept sideways underneath the film, well beyond where the damage was. Now look at a galvanised steel gate that has been scraped by a vehicle. The scrape is bright and bare, and a year later it is still, more or less, not rusting.

Same metal, same air, same rain, opposite outcomes. The reason is that "protective coating" describes three genuinely different mechanisms that happen to share a shelf in the hardware shop, and knowing which one you are buying is the difference between an asset that lasts twenty-five years and one that gets repainted every four.

Rust is a battery, not an attack

The habitual mental picture of corrosion is chemical — something in the air eating the metal. It is closer to the truth to say the metal builds a small battery and then runs it down.

At one spot on a wet steel surface, iron gives up electrons and dissolves: Fe → Fe²⁺ + 2e⁻. That spot is the anode. The electrons travel through the metal itself to another spot where oxygen and water accept them: O₂ + 2H₂O + 4e⁻ → 4OH⁻. That spot is the cathode. The water film on the surface, especially if it has any dissolved salt in it, closes the circuit as the electrolyte. The iron ions and hydroxide meet, and the hydrated iron oxides that precipitate are what we call rust.

Four things are therefore required: an anode, a cathode, an electrical connection between them, and an electrolyte. Remove any one and corrosion stops. Every anti-corrosion technology ever invented is an attack on one of those four.

One more property of rust matters more than it looks. The oxides occupy several times the volume of the iron they came from. That expansion is why rust jacks paint off a surface from underneath, and why chloride reaching a steel reinforcing bar can crack the concrete around it from the inside — the corrosion product does not fit in the space the steel used to occupy. Rust is not just loss of metal; it is loss of metal plus a wedge.

Three quite different jobs a coating can do

Barrier. Most paints, epoxies and clear coats work by keeping the electrolyte and oxygen away from the steel. Nothing is chemically clever here — the film is a wall, and its performance is a matter of how little water and oxygen pass through, how well it sticks, and how thick it is.

The weakness is written into the mechanism: a wall works until it has a hole in it. When a barrier film is breached, corrosion starts at the exposed metal, and then two things make it worse than a fair fight. Water creeping along the coating-metal interface sustains the cathodic reaction under the film, and the hydroxide it generates is alkaline enough to destroy the adhesive bond — cathodic disbondment. On thin painted sheet in humid air the same process runs as narrow threads that visibly crawl outwards, called filiform corrosion. That is the car door.

Sacrificial, or galvanic. Galvanising, and zinc-rich primers, do something else entirely. Zinc is electrochemically more active than iron, so when the two are in contact and wet, the zinc becomes the anode and corrodes instead of the steel. The steel is forced to be the cathode, and a cathode does not dissolve.

This is why the scratched gate does not rust. The zinc does not have to physically cover the scratch — it protects the exposed steel across a short distance through the surrounding moisture film. The protection is finite: it lasts as long as there is zinc left to consume, which is why galvanising is specified by coating mass or thickness and why a design life can be estimated from it. But within that life, a defect is not a failure. That is a fundamentally different relationship with damage than a barrier coating has.

Inhibitive. The third route puts pigments in the primer that slowly release species which passivate the steel at a defect — reinforcing the thin protective oxide layer instead of walling it off or feeding a sacrificial metal to it.

This category carries an important piece of live history. Hexavalent chromium did this job superbly for decades and is a confirmed carcinogen, and its phase-out under REACH and equivalent regimes forced the coatings industry into a search for replacements — phosphates, ion-exchanged silicas, rare-earth compounds such as cerium salts. Several work well. Matching chromate's combination of performance, tolerance of poor application and price across every substrate remains, honestly, unfinished, and it is one of the reasons corrosion science still has a lot of open work in it.

The rule that explains most surprises: area ratio

Here is the single most useful idea in the subject, and it explains several things that otherwise look perverse.

The rate at which an anode dissolves depends on how much cathode is driving it. A large cathode connected to a small anode concentrates the entire cathodic reaction onto a tiny patch of metal, and that patch corrodes fast.

That is why a pinhole in an otherwise perfect coating can be more dangerous than a broadly indifferent coating: the pinhole is a small anode with a large area of coated, oxygen-supplied steel around it acting as cathode. It is why joining dissimilar metals in the wrong ratio is a design fault — steel bolts in an aluminium plate is a bad idea, and an aluminium plate with steel bolts is worse, because the small aluminium anode would be devoured. And it is why the pairing rule of thumb runs: if you must mix metals, make the less noble one the large area, and insulate the joint if you can.

Given a choice between a superb coating with a few defects and a merely adequate coating applied without any, the second one usually outlives the first. Corrosion protection is a discipline of defect control, not of product selection.

Salt-spray hours are a poor thing to buy on

Almost every anti-corrosion product is sold with a salt-spray number: 500 hours, 1,000 hours, sometimes more. The test behind it, ASTM B117, holds a panel in a continuous warm salt fog and records when rust appears.

It is a useful screening and quality-control tool. As a predictor of how long a coating will last on a real structure it is famously weak, and this is not a controversial opinion inside the field — it is stated in the standard's own scope. The reason is mechanistic: a real coating outdoors goes through wet-dry cycling, ultraviolet exposure, temperature swings and, in much of India, a monsoon followed by months of dust. Constant wetness is not a harsher version of that; it is a different loading altogether, and it suppresses some failure modes while exaggerating others.

The better answers are cyclic corrosion tests — ASTM D5894 and its relatives, which alternate salt fog, drying and UV exposure — and, for specification rather than testing, ISO 12944. That standard does the sensible thing and starts from the environment: it classifies sites by corrosivity, from C1 (dry heated interiors) through C3 (urban, moderate humidity), C4 (industrial, coastal) to C5 and CX (severe marine and tropical industrial), with separate categories for immersion. Then it defines durability ranges and specifies coating systems — primer, intermediate, topcoat, with dry film thicknesses — that meet a given durability in a given category.

That is the correct shape of the question. "Which coating is best" has no answer. "Which system reaches high durability in C5-M on blast-cleaned carbon steel" has several.

Most coating failures are preparation failures

If there is one line from corrosion engineering worth carrying away, it is that one, and it is repeated so often precisely because clients keep spending the money in the wrong place.

Three preparation variables dominate. Cleanliness — mill scale, old coating and rust must go, with blast-cleaning grades from ISO 8501-1 (Sa 2½ being the common specification for industrial work) defining how thoroughly. Profile — the blasted surface's roughness gives the primer something to key into; too smooth and adhesion suffers, too coarse and the peaks stand proud of the film. And soluble salt contamination, which is the one most often skipped: chloride left on the steel from a coastal atmosphere draws water osmotically through the film and blisters it from beneath, and no film thickness fixes that.

A mid-range coating over properly prepared steel will comfortably outlive a premium coating sprayed onto a surface that was merely wiped down. The money is in the preparation, and the specification should say so in enforceable terms.

Where it costs money in India

Corrosion is one of the largest unbudgeted costs in any industrial economy — global studies have put the total at a few per cent of GDP, and the striking part of those studies is not the size but the estimate that a substantial share of it is avoidable with practices already known.

The Indian exposure profile is unusually demanding. A long coastline puts chloride into the air for a very long stretch of infrastructure. High humidity extends the time of wetness, which is the variable atmospheric corrosion actually tracks. The monsoon delivers repeated wet-dry cycling, which is the aggressive case. Industrial clusters add sulphur dioxide. Reinforced concrete carries its own version of the problem, where chloride reaching the rebar depassivates it and the expanding rust cracks the cover from within — a failure that is invisible until it is structural, and which links directly to how engineers predict where a structure will fail.

There is also a biological route that surprises people. Microbially influenced corrosion — sulphate-reducing bacteria in particular — attacks pipework, tanks, cooling systems and marine structures by creating local electrochemical conditions under a biofilm. Here corrosion control and surface hygiene turn out to be the same engineering problem approached from two directions.

What to ask before specifying

1. What is the corrosivity category and the required durability? Answer this before looking at any product. ISO 12944 C-category plus a durability range is the specification; a product name is not.
2. Is this a system or a can? Primer, intermediate, topcoat and dry film thickness per coat, on a named substrate. Anti-corrosion performance is a property of the stack, not of the top layer.
3. What is the surface preparation specification? Blast grade, profile range, and a maximum soluble-salt level with a test method. If nobody will commit to this in writing, the coating warranty is decorative.
4. Cyclic test data, not just B117 hours. Ask which cyclic standard, how many cycles, and what the scribe creep was — the measured sideways spread from a deliberate scratch is far more informative than a time-to-first-rust number.
5. Any galvanic pairings in the design? Fasteners, brackets and dissimilar-metal joints are where clean designs corrode, and area ratio decides the severity.
6. What is the touch-up and repair procedure? Every structure gets damaged during erection. A system without a defined repair method has a hole in its service life.

To make the categories concrete: Smart Warrior Coatings, part of Reinste Nano Ventures — which, in the interest of disclosure, is part of the same group as this publication — keeps its industrial corrosion clear coats as a separate line from its antimicrobial and surface-shield products rather than presenting one coating as the answer to every surface problem. That separation is what a buyer should expect, and the technical data sheet is where the questions above get answered: qualified substrates, application method, film thickness, test basis. A product page tells you what a coating is for; the TDS tells you whether it fits your case.

Why it matters for students and researchers

Corrosion is an unusually honest field to enter, because it is old, enormously consequential, and still full of problems nobody has closed.

The chromate replacement problem is the clearest of them: an entire industry needs a pigment chemistry that matches a banned material's performance, and the search has produced good partial answers rather than a general one. Self-healing coatings are an active and genuinely elegant line of work — microcapsules that rupture when the film is damaged and release a healing agent or inhibitor into the defect, or polymer chemistries that re-form bonds across a scratch. Smart and sensing coatings that change colour, fluoresce or report electrochemically when corrosion begins under the film address a real diagnostic gap, because the failure everyone fears is the one that is invisible until it is structural.

Then there is the measurement problem, which is the theme this subject keeps returning to. Accelerated tests correlate poorly with service life, everyone knows it, and constructing tests that correlate well is unglamorous, slow work of exactly the kind that makes a solid thesis. Alongside it, atmospheric corrosion mapping and data-driven service-life prediction have become tractable now that environmental datasets are dense enough to support them.

The disciplinary spread is the attraction. Electrochemistry, polymer science, surface engineering, microbiology for the biofilm route, structural engineering for the consequences, and the regulatory dimension around banned actives. A student can enter from any of those and find work that matters within a few years rather than a few decades.

Frequently asked questions

Does stainless steel rust?

Yes, under the right conditions, and the misunderstanding causes real failures. Stainless resists corrosion because chromium forms a thin passive oxide film that repairs itself in the presence of oxygen. Deprive it of oxygen — inside a tight crevice, under a gasket, beneath a deposit — and the film cannot regenerate, and crevice corrosion proceeds. Chlorides attack the passive layer directly, causing pitting, which is why seawater is hard on the cheaper grades. Choosing a stainless grade is choosing which environment its passive film can survive, and that is an alloying question, not a guarantee.

Why does galvanising protect a scratch when paint does not?

Because they intervene at different points in the electrochemical cell. Paint is a wall: it stops the electrolyte reaching the steel, and a hole in a wall is simply a hole. Zinc changes which metal dissolves: being more active, it becomes the anode and corrodes in place of the exposed steel nearby. The protection is finite — it lasts while zinc remains — but within its life a small defect does not propagate.

Is rust-converter paint over rust a real solution?

It has a genuine but narrow use. Converters, usually tannic or phosphoric acid based, react with rust to form a more stable compound that can be overcoated, and on light surface rust with loose material removed they are a reasonable repair. They do not remove the underlying cause, they do not deal with soluble salts embedded in the rust, and on heavy or actively wet corrosion they seal the problem in rather than solving it. For anything structural, preparation back to sound steel is the answer.

Do nano-coatings stop corrosion better?

Sometimes, and for specific reasons rather than by category. Nanoscale fillers such as platelet clays can lengthen the path a water molecule must take through a film, improving barrier performance at low loading; nanostructured inhibitor carriers can release actives on demand at a defect; and thin functional layers can add corrosion performance without changing appearance. Those are real mechanisms. "Nano" on a label, unattached to a mechanism and a test result, is a marketing claim, and the questions above apply unchanged.

How long should a coating system last?

That is a specification, not a property. Under ISO 12944 you state the corrosivity category and the durability range you want, and the standard tells you which systems and thicknesses reach it. A high-durability system in a moderate C3 environment can be a decades-long proposition; the same nominal product in C5-M marine conditions, on poorly prepared steel, may need attention in a few years. The environment, the preparation and the system are one answer, and quoting any of the three alone is how disappointments get bought.