The Pain Point: The Radiator That Failed From the Water Side

An aluminium radiator returned from service usually shows its damage on the inside. The external surface may be intact - paint holding, fins clean - while the tube walls are pitted through from the coolant side. The vehicle owner reports no abuse; the coolant was changed at the specified interval, or so they believe.

Internal corrosion is the least visible and most expensive failure mode in aluminium heat exchangers, because it attacks the pressure boundary from a direction that no external inspection can reach. It is also the failure mode most dependent on decisions made outside the radiator plant: coolant chemistry, system materials, manufacturing residue and air ingress.

Aluminium is not inherently fragile in water. It is fragile in the wrong water.

The Principle: Passive Film, Galvanic Couples and Crevices

The passive film is thin and self-repairing - within limits. Aluminium in neutral water forms an oxide layer a few nanometres thick that stops further attack. The film repairs itself when damaged, provided the local chemistry stays in a range where the oxide is stable. Outside that range - too acidic, too alkaline, high chloride - the film breaks down locally and pitting begins.

Chloride is the primary aggressor. Chloride ions penetrate the oxide film at weak points and initiate pits. Once a pit starts, the local chemistry inside it becomes acidic and self-sustaining. A pit that is 0.1 mm deep in a 0.25 mm tube wall is a leak in waiting.

Galvanic coupling drives attack at specific locations. Aluminium is anodic relative to copper, brass, steel and most other metals used in cooling systems. When a brass fitting, a copper core or a steel bracket is in electrical contact with aluminium through the coolant, the aluminium becomes the sacrificial anode. The attack concentrates at the junction - which is usually a brazed joint or a thread, the two places where the wall is already thinnest.

Crevices concentrate corrosion. Any gap where coolant can enter but cannot flow freely becomes a corrosion site. Flux residue under a brazed fillet, a poorly sealed gasket groove, a thread root - all are crevices.

The oxide film also matters at high temperature. Aluminium is strongly oxyphilic, and at elevated temperature the film grows. In a brazing furnace this is a problem; in service it is protection, provided the coolant does not chemically remove it.

Temperature accelerates everything. Reaction rates roughly double for every ten degrees of temperature rise, and a radiator's coolant side runs well above ambient. The same water chemistry that is harmless in a cold circuit will pit aluminium at operating temperature. This is why corrosion validation has to be run hot, and why a laboratory immersion test at room temperature proves very little about a real cooling system.

Brazing flux residue is a built-in corrosion source. Fluoride-based flux residues are hygroscopic and acidic when wet. A core with excessive residue will corrode from the inside regardless of how good the coolant is.

Equipment Classification: System Materials and Their Interactions

Brazed aluminium cores. The dominant construction. Susceptible to internal pitting where coolant chemistry is poor and to galvanic attack where dissimilar metals are present.

Cast aluminium tanks and covers. Porosity can create internal crevices that hold stagnant coolant.

Copper and brass components. Radiators, heater cores, fittings and miniature end covers. In an aluminium system these become cathodes and drive galvanic attack on the aluminium.

Steel and stainless fittings. Lower potential difference than copper, but still cathodic to aluminium.

Elastomer seals and hoses. Permeation through hose walls admits oxygen, which sustains corrosion by continuously depolarising the cathodic reaction.

Coolant itself. Ethylene glycol or propylene glycol base with an inhibitor package - silicate, phosphate, organic acid or hybrid chemistry. The inhibitor package determines whether the aluminium is protected.

Practical Guide: Controlling the Water Side

Specify coolant chemistry, not just coolant type. Glycol concentration, inhibitor chemistry, chloride limit, pH range and conductivity all belong in the specification.

Control flux residue at the source. Choose vacuum brazing where the application demands maximum cleanliness, or define and verify a residue limit for flux-brazed cores.

Avoid copper and brass in the coolant circuit where possible. If a brass fitting is unavoidable, isolate it electrically or ensure the inhibitor package is formulated for mixed-metal systems.

Eliminate crevices by design. Seal gasket grooves fully, avoid blind tapped holes open to coolant, and specify fillets that do not leave a capillary gap.

Keep air out of the system. A vent at the highest point and a properly designed expansion tank prevent air pockets, which both cause hot spots and supply oxygen.

Control the internal surface. Some cores use a zinc-rich internal coating or a sacrificial cladding layer to provide cathodic protection to the tube interior. Specify it where chloride exposure is expected.

Define the service interval and the test. Coolant degrades: inhibitors deplete, glycol oxidises to glycolic acid, and the pH drifts. A defined interval with a measurable criterion is more useful than a general instruction.

Verify with a corrosion test, not just a pressure test. A cyclic corrosion test with representative coolant and temperature profile reveals pitting susceptibility that a hydrostatic test cannot.

Check the system, not only the part. The radiator can be perfect and still fail because of a brass fitting added downstream. Ask what else touches the coolant, and where the electrical paths run.

Field Experience: The Brass Tee That Ate the Core

A fleet of light trucks began reporting heater core leaks at around 18 months, always in the same region of the core, always on the tube row nearest the inlet. Coolant analysis showed acceptable glycol concentration and acceptable pH. Pressure testing of returned cores showed pinholes in the tube wall.

Sectioning revealed severe localized pitting on the coolant side of the first two tube rows, with no attack elsewhere. The cause was traced to a brass tee fitted in the heater circuit by a body builder, installed about 80 mm upstream of the core inlet. The tee was in direct electrical contact with the aluminium through the coolant, with no isolating hose section.

The galvanic couple concentrated the attack at the point where current density was highest: the inlet region, where flow was turbulent and the surface was freshly exposed. The rest of the core, acting as a remote cathode, was unaffected.

The fix was an isolating rubber section upstream of the core and a change in the coolant specification to a hybrid organic acid package formulated for mixed-metal systems. Return rates dropped to near zero within two model years.

The lesson: an aluminium radiator can only be designed for the system it is told about. A component added downstream by someone else can still destroy it.

Summary

Corrosion in aluminium radiators is a system property, not a material property. The passive film, the chloride level, the galvanic couples, the crevices and the coolant inhibitor package interact continuously, and the failure appears where the wall is thinnest or the geometry is worst. Manufacturers who control only their own process - alloy, braze, test - are controlling perhaps half of the variables.

Outlook

The direction of travel is toward more mixed-material cooling systems, not fewer: aluminium cores with stainless fittings, brazed cold plates with copper busbars nearby, and dielectric coolants with very different chemistry. Each addition creates new galvanic pairs and new compatibility questions, and each one moves the failure mode further from something the radiator supplier can control alone. At the same time, coolant technology is moving toward longer-life organic acid packages with lower silicate content, which is generally good for aluminium but requires that the system be clean at fill. Expect corrosion validation to become a system-level requirement in contracts, with documented coolant chemistry and periodic sampling - because the alternative is discovering the problem at 18 months in the field.