FIELD GUIDE / coatings
One Scratch in a Painted Ship
The painted ship is not saved by paint alone. It survives by a coordinated substrate, pretreatment, film thickness, edge treatment, and repair doctrine, all of which arrive at once when a scratch opens.
At a glance
- A paint system protects through its continuous stack, adhesion, thickness, and substrate preparation rather than colour alone.
- Sharp edges, holes, recesses, and damage need explicit coverage and inspection because average thickness can conceal the governing thin point.
- A coating specification needs repair, inspection locations, gauge calibration, adhesion classification, and records as well as a nominal film thickness.
A small steel ship has been painted exquisitely.
Not merely brightened for the harbour photographer. Not merely given the cheery blue one associates with boats that have never met a winter tide.
Its hull had received preparation, pretreatment, primer, build coat, topcoat, measurements, records, and the sort of inspection at which a stray fingerprint begins to feel morally compromised.
It was now a painted ship.
This is an object of great public confidence. The sea can be seen. The steel cannot.
The conclusion arrives with the assurance of a man who has confused a curtain with a wall: the steel has therefore ceased to be a subject of chemistry.
Then, at a fender, a quay edge, an inattentive chain, or one of the innumerable small collisions by which ships maintain their acquaintance with the world, the paint acquires a scratch.
The scratch is small.
The ship is not.
And from that moment the largest part of the paint is the least informative part of the ship.
What Has Actually Been Put to Sea
The popular noun is paint, which is a little like calling a clock a face.
The useful thing is a coating system: steel made receptive by preparation, perhaps conditioned by pretreatment, then a primer, an intermediate layer that supplies build, and a top layer for its exposed duty.
The substrate provides stiffness and shape. The outermost few micrometres decide how water and ions reach it.
A coating works by material, continuity, interface and geometry, which is an inconvenient number of responsibilities for something sold in a tin.
Paint itself is normally an organic polymer network with pigments and additives. It becomes a film by evaporation, by reaction, or by both.
A two-component epoxy or polyurethane forms a crosslinked network after mixing. A waterborne or lacquer system leaves its water or solvent behind and coalesces.
A film may be firm enough to measure and still not be fully cured.
Dry to touch, ready for another coat, and fully cured are not three poetic descriptions of one afternoon.
They are different states, with different permissions.
The hull has therefore not been dressed in a colour.
It has been entrusted to a sequence of interfaces.
Oil, oxides, salts, dust, moisture and weak old layers are not small domestic untidinesses at such a sequence. They prevent intimate contact, or make an unstable boundary.
Paint durability depends heavily on surface preparation, and the sea has no obligation to overlook a contaminant because the finish looked handsome at launch.
There is a useful unit for seeing how little romance the matter contains.
A one-mil paint film is 25.4 µm thick.
It is a barrier only for as long as it is a barrier.
An interruption reaching steel is not an intact system with a regrettable mark on it.
It is a different system, one in which the substrate has been admitted to the conversation.
But Does Paint Not Make Steel Immortal?
No.
This is the point at which popular imagination generally puts on a sou’wester and walks straight into a lamppost.
It supposes that a painted hull has become a plastic boat with some steel concealed inside for ballast.
A coating does not abolish the metal beneath it.
It reduces access to that metal, under the conditions for which its preparation, chemistry, thickness, cure and geometry were chosen.
Nor is every coating meant to behave the same way at a defect.
A damaged sacrificial metallic coating can consume itself to protect the steel. Beside a damaged noble coating, the exposed steel is the anode and loses its protection instead.
One cannot settle that question by observing that something more noble has been placed on the outside.
The coating family, its continuity, the exposed areas, drainage, electrical isolation and the electrolyte all retain votes.
The scratch carries the same rude lesson. A very small area of exposed steel, beside a much larger intact noble-coated surface, is an electrochemical arrangement rather than a cosmetic defect.
This does not mean every scratch instantly produces a theatrical orange wound.
Its behaviour depends on the substrate, the pretreatment, the coating chemistry and thickness, the cure, the edge form, seawater access, cathodic effects, the size of the damage, and the maintenance interval.
The point is more sober.
The coating cannot be judged as though the defect were absent, because the defect has changed what the coating is doing.
Thickness Is a Promise With a Number
Thickness is not an adjective applied after the painting has happened.
It is a functional variable, and a ship’s coating must possess it in the places where the sea will ask for it.
Wet-film control is the immediate lever, because dry film is what remains once the volatile material has left.
A wet film of 100 µm at sixty percent volume solids predicts a dry build of 60 µm.
The wet gauge is a predictor. The dry-film reading is the direct check.
But “25 µm of paint” is not a specification.
It is a small fog bank.
Is it the primer only, or the total dry film? A local point, or an average over a panel?
Does it include the edge, the weld seam, the hole, the recess, the boot-top, the place where a mooring line has already begun to form opinions?
A requirement must say whether the thickness belongs to each layer or to the whole stack, and where it is to be measured.
Otherwise a dispute waits inside the number like a worm inside an apple.
The gauge must also know what it is being asked.
A magnetic-induction gauge measures the separation between a nonmagnetic coating and a ferromagnetic substrate, which suits many paints on steel.
An eddy-current gauge answers a different question, through a conductive nonferromagnetic substrate such as aluminium.
Calibration belongs to the substrate class, the curvature, the roughness and the thickness range of the actual work, and a calibration range published for flat coupons is not a promise about a built-up hull detail.
Even a magnetic gauge cannot separate the individual paint coats once they all exist together.
Intermediate thickness must be checked at hold points, or by a method able to resolve the interface.
A large reading on the completed field cannot travel backwards in time and certify the primer beneath it.
Paint, like a ship’s log, is most useful when it is written during the voyage.
The Edge Refuses to Behave Like the Middle
Because the middle is broad, flat and shamelessly easy.
Pigment, solvent balance, airflow, film build and substrate temperature all affect how paint flows before it sets.
A coat can look smooth across the broad field and still have low build at a sharp edge.
It can be generous overall and retain solvent in a locally heavy place.
Sharp geometry gives nonuniform fields, local build, and troublesome measurement.
The edge has not rebelled.
It has merely declined to be a flat panel.
This is why integrity is a separate question from adhesion.
Visual inspection searches for runs, pinholes, craters, dry spray, contamination, edge coverage and uncoated marks.
Holiday detection can find a through-film discontinuity, provided its voltage suits the film thickness, the geometry and the governing method.
A glossy finish can be imperfect in every way that matters to seawater and perfect in every way that matters to a brochure.
A documented sampling plan makes the same point in the plainest possible arithmetic: five locations in every hundred square feet, three readings at each location, each individual reading falling within a window around the specified thickness.
That is an example, not a universal shipyard commandment.
Yet it reveals the principle.
Averages conceal thin, hard-to-reach regions, so a sampling plan needs the locations where the design expects trouble, rather than the places at which an inspector can lean comfortably.
The scratch, seen in this light, is an edge made suddenly and badly, in service.
It has no application record.
It has no declared feathering procedure.
It is the arrival of a new geometry, a new exposed area, and possibly a new route beneath the paint.
What Begins Beneath a Small Scratch
Not necessarily a broad, democratic thinning of the hull.
Underfilm corrosion is not a superficial stain with ambitions.
It is a question of whether a defect, an interface, a deposit or a retained liquid has made one particular route more active than the rest of the painted field.
The paint above such a route may still look entirely respectable.
That is precisely the difficulty. The coating hides the substrate from the sea, which is the intention, and hides the substrate from the inspector, which is not.
A pleasingly low average figure for wall loss may tell a shipowner a great deal about the average, and nothing sufficient about the scratch.
The controlling quantity is the remaining metal at the deepest credible site, together with its location, its morphology, and whatever sheltered geometry or deposit enabled it.
A ship cannot convert a local defect into uniform loss by dividing two pleasant numbers.
Adhesion Cannot Be Seen From the Pier
Adhesion is resistance to separation at an interface.
Cohesion is the coating’s internal resistance to splitting.
The distinction becomes embarrassing only after a failure, when a pull-off fixture may reveal coating on both the fixture and the substrate, or a clean substrate, or merely the fixture adhesive as the weakest party of all.
A number without its failure location is a report which has misplaced its verb.
Cross-hatch and X-cut tests are rapid qualitative screens, reported on a small rating scale. A pull-off test reports a tensile stress instead.
These are not rival ways of obtaining the same fact, and their figures cannot be compared as though they belonged to a single ladder.
The requirement needs the method, the test environment, the locations, the replicates, the cure state and the classification of the failure.
A coating does not adhere in general.
It adheres, or fails to adhere, somewhere, in a stated system.
Surface roughness carries its own comic temptation.
More roughness appears to promise more grip, as a larger beard appears to promise more wisdom.
In fact roughness can improve mechanical keying, while a sharp profile may be incompletely covered by thin paint and concentrate stress in a brittle film.
A profile recorded for one blasted thermal-spray surface is evidence about that process condition. It is not a universal hull-preparation command.
It does show why a profile must be selected together with the system intended to sit on it, rather than celebrated in the abstract.
Somebody Must Repair the Scratch
And it is not the person who merely restores its colour.
A repair is a controlled return to the intended surface state.
Isolate the affected condition as required. Remove contamination and unstable material. Prepare the exposed steel. Restore the necessary layers and their thickness. Protect the edges. Check continuity. Record what happened.
The record should say what was removed, and how the protective surface was restored.
Otherwise the next inspector receives a painted silence, and must guess whether a sound patch or the beginning of underfilm spread lies beneath it.
No single observation proves the absence of every corrosion form.
Thickness trending, close visual examination, deposit inspection, metallography, electrochemical methods and crack-sensitive examination all have different capabilities and different blind spots.
Nor is a calendar by itself a maintenance doctrine. A twelve-month trend can be meaningful for one mechanism and far too coarse for another.
What matters is mechanism-specific evidence: coating continuity and edge condition after a repair; observed drying and the absence of retained deposits; a changed environment kept under surveillance where cracking is credible.
Paint has no respect for a maintenance interval that never looks at the places where paint has failed.
And So the Ship Is Maintained
Imagine the vessel alongside on an ordinary wet morning.
Nothing luminous occurs. No heroic sheet of paint rises up to throw back the ocean.
There are gauges, calibrated standards, defined locations, records of cure and repair, and inspection at the edges and the penetrations.
The broad blue field may remain beautiful, and it may be doing excellent work.
But the proof was never the beauty of the field.
The proof is the treatment of the scratch: whether the steel was prepared, the barrier restored, the edge examined, the defect geometry understood, and the next look arranged before the sea had time to make the decision on everybody’s behalf.
AT THE BENCH
Questions, with the conditions attached.
Does paint make a steel ship corrosion-proof?
Paint does not make steel immune to corrosion. A paint system protects by providing a continuous, adherent, sufficiently thick barrier over a prepared substrate, often as pretreatment, primer, intermediate build, and topcoat. A scratch, thin edge, poor interface, retained salt, or failed repair changes the system because seawater can then reach the steel and sustain localized corrosion.
Why is dry-film thickness checked instead of trusting the amount of paint applied?
Dry-film thickness is the direct record of the cured barrier that remains on the steel. Wet-film readings help control application, but volatile material leaves during drying or curing: a 100 µm wet film at 60% volume solids predicts 60 µm dry build. The requirement must also state whether thickness means one layer, the whole stack, a local reading, or an average.
Why do painted edges and scratches need special attention?
Sharp edges, holes, recesses, and scratches can receive less effective coverage than a broad flat field or can expose steel directly. Coating requirements therefore identify mandatory locations and inspect edge coverage, pinholes, craters, contamination, and through-film discontinuities. An average thickness over a large area cannot prove that the thin or damaged location which admits seawater remains protected.
What does surface preparation have to do with a paint failure?
Surface preparation determines whether a coating can make stable intimate contact with steel. Oil, oxides, salts, dust, moisture, and weak prior layers can prevent adhesion or create an unstable boundary. Roughness can improve mechanical keying, but an excessively sharp profile can remain incompletely covered by a thin film and can concentrate stress in a brittle coating.
Can a tiny scratch really be worse than a large area of bare steel?
A small steel break in an intact noble coating can corrode faster than broadly exposed bare steel because the exposed steel can act as the anodic member beside a larger cathodic coated area. This outcome depends on coating type, continuity, exposed-area ratio, drainage, electrical contact, and electrolyte access. A sacrificial metallic coating behaves differently because it is intended to consume itself to protect steel.
Why is a repaired scratch not finished when the colour is restored?
A repair is a controlled return to the intended surface state: isolate the affected area, remove contamination and unstable material, prepare the exposed steel, restore the required layers and thickness, protect the edges, check continuity, and record what was done. Without that record the next inspector receives a painted silence and must guess whether a sound patch or the beginning of underfilm spread lies beneath it.
Which gauge measures paint thickness on a steel hull?
A magnetic-induction gauge measures separation between a nonmagnetic coating and a ferromagnetic steel substrate, making it suitable for many paint systems on steel. Its calibration must match the substrate class, curvature, roughness, and thickness range of the work. Readings also need defined locations because an instrument result from a flat panel does not prove edge or recess coverage.
Is a cross-hatch adhesion rating the same thing as a pull-off strength?
A cross-hatch or X-cut rating is a qualitative adhesion screen, while a pull-off test applies tensile load to a bonded fixture and reports stress. FHWA describes a 1 to 5 rating scale in the cited ASTM cross-hatch method, whereas 1 MPa equals 1 N mm⁻² in a pull-off result. The methods are not interchangeable, and failure location must be recorded.
How should a painted area be sampled for thickness acceptance?
A sampling plan defines the inspection unit, mandatory locations, number of readings, acceptance window, calibration record, and treatment of inaccessible areas. FHWA’s archived bridge-painting manual gives one documented example: five locations in each 9 m², three readings at each location, and an individual reading window of 80 to 150% of specified thickness. A contract must adopt its own governing rule.
What makes a coating repair more than paint applied over a scratch?
A coating repair restores a controlled surface state, not merely colour. It identifies the substrate condition, removes contamination and unstable material, restores the required layer architecture and thickness, protects edges, and checks continuity. The close-out record states what was removed and how the protective surface was restored, so later inspection can distinguish a stable repair from renewed underfilm damage.
Can a paint system be accepted just because it looks smooth and glossy?
A smooth glossy field can still have low build on a sharp edge, retained solvent in a heavy area, a weak interface, or a holiday that exposes steel. Acceptance separates appearance, thickness, adhesion, and integrity. Visual inspection looks for runs, pinholes, craters, dry spray, contamination, edge coverage, uncoated marks, and colour variation; suitable holiday detection finds through-film discontinuities.
Sources and further reading
- FHWA, Field Manual for Bridge Painting Inspectionfhwa.dot.gov
- NIST, Coating Thickness Calibration Standardsmetallurgy.nist.gov
- HSE, Corrosion / selection of materialshse.gov.uk
- HSE, Inspection/Non-Destructive Testinghse.gov.uk