Galvanic corrosion: the smallest part in a joint loses

Galvanic corrosion needs three things present at the same moment, and a fourth to keep it running. Two metals of different natural potential have to be in electrical contact, and a liquid that conducts has to bridge them. The UK's National Physical Laboratory lists those conditions and adds one more. A reaction has to keep going on the nobler metal, which in ordinary conditions means dissolved oxygen being consumed at its surface.

The electrolyte does not have to be a puddle. That same guidance describes it as a bulk volume of solution, a condensed film, or a damp solid such as soil, a salt deposit or the corrosion product itself. Overnight condensation on a bolted joint qualifies. So does a crust of dried sea salt that goes damp again with the morning humidity.

Contact is looser than the word suggests. Direct physical touching is the usual route, but continuity established through structural metalwork or through an earth connection counts as well, and the guidance notes the junction itself need not be immersed. Two components can be joined and still form one working cell with only a film of water across them.

This is a different fault from the one that tunnels through a copper tube wall. Formicary corrosion works on a single metal and needs an organic acid carried in the air around it. The mechanism here needs no acid at all. It needs a second metal, and it will run in water that is harmless to either metal on its own.

A galvanic series ranks metals by the potential each takes up in one stated environment. Set two of them against each other on that list, and the one nearer the active end is the one that corrodes. That single reading is what the series exists for. The United States defence standard that publishes such a table footnotes it to say the table identifies the anodic member and must not be used for anything else.

For aircon hardware the order is short and consistent. Aluminium sits at the active end, copper sits below it, and stainless steel sits below copper again. Aluminium therefore loses to both of them. Copper loses to stainless steel. Stainless steel loses to almost nothing you will find on a domestic system.

What the ranking will not tell you is how fast. That defence standard was rewritten in 2021 to stop using the gap between two potentials as its test and to use a measured corrosion rate instead. Its own worked example is aluminium. Paired with titanium the potential gap looks worse than with stainless steel, yet in service the titanium pairing corrodes more slowly. Direction and speed are separate questions.

Position on that list is not permanent either. Stainless steel reads as noble because it carries a passive oxide film, and published guidance records polarity reversals when that film breaks down or the chemistry shifts. Temperature can flip a pair. So can acidity. None of that appears in a table read at room temperature.

A standard coil pairs copper tube with aluminium fins, so the couple is built in before anyone installs anything. Manufacturer patent literature places it precisely. The couple is made where tube meets fin, at the collar each fin punches around the tube, and because aluminium is the anodic member it is the fin that wastes away.

The reason a coil survives its first year is arithmetic and water. Those same engineering documents state that corrosion durability of an aluminium fin and copper tube exchanger runs inversely to the exposed area of copper in the fin pack. The collars cover much of the tube, so the copper working as a cathode is a small fraction of the surface the fins present. Cutting that copper area cuts the corrosion current in proportion.

Clean condensate barely conducts, which settles the water side. Corrosion guidance from the national measurement institute notes that this form of attack is seldom a problem in pure water, and that severity climbs with the ionic conductivity of whatever bridges the two metals. Condensate is condensed water vapour, so it starts out close to distilled. It touches both metals and holds little that could carry a current.

Change the water and the same couple wakes up. Patent literature on coil corrosion names the two outdoor electrolytes that matter: salt water in coastal regions, and the acid formed when sulphur dioxide and other industrial pollutants meet moisture. Both lift conductivity sharply. That is the mechanism underneath coastal aircon servicing, and it is why one coil design ages at two different rates across a single city.

Manufacturers price this couple in. The patent record puts a copper tube and aluminium fin exchanger at roughly ten times lower corrosion durability than an all-aluminium one. That figure is a manufacturer's own comparison, and it is worth reading as one. The same documents describe engineering around the couple directly. Exposed tube gets plated with a metal closer to aluminium, and the collar is tightened so less water sits in the crevice.

The counter-intuitive rule is that a large anode is safe and a small one is not. Corrosion guidance puts it plainly. The larger the cathode compared with the anode, the more oxygen reduction can occur, so the greater the current and the greater the corrosion. Under still conditions the total quantity of attack is fixed by the cathode area. Shrink the anode and identical damage concentrates into less metal.

That principle has a name and an immediate consequence. Published guidance calls it the catchment area principle, and the consequence it draws is that adverse ratios turn up at fasteners and at joints. A rivet, a bolt or a screw is small by definition, and it is nearly always holding something far larger than itself.

Fastening metal should sit at the same potential as the parts it joins, or better, be the nobler of the two. The defence standard offers the clearest illustration of it. Stainless bolts and screws may be used to fasten aluminium sheet. The reverse arrangement is unacceptable, because stainless steel is cathodic to aluminium. The same guidance warns against mild steel rivets in copper or stainless plate.

A condenser bracket carries that geometry exactly. Where the support is stainless steel, the joint contains a large noble surface. Anything less noble threaded into it becomes a small anode facing a large cathode, outdoors, wetted by every storm. It is also the part holding the machine up. A stainless steel aircon bracket in sound condition can be carried on fixings that are not.

Inspecting or correcting a support is not work an owner should take on. Who may touch a support outside a building is settled by the trained installer requirement, and the choice of fixing metal sits with whoever specifies the job. Knowing the rule buys one answerable question at quotation stage. What are the fixings made of, and does that match the support they go into.

Every fix works by removing one of the conditions. Break the electrical path with insulating bushes, washers or gaskets. Break the water path by sealing the joint so nothing can stand in it. Or delete the pairing at source by specifying metals from the same group. Corrosion guidance lists all three, and adds that insulation should be verified as effective once installation is finished, because later modifications and earthing arrangements quietly reconnect things.

Drainage matters more outdoors than the arithmetic does. The catchment area principle describes metal immersed in still, aerated water. In open air the same guidance redirects the effort to design for drainage, so condensation and rain cannot accumulate at joints and in hollow sections. A joint that dries out between showers spends most of its life switched off.

Coating is the step most often done backwards. Guidance is explicit that both members should be painted where that is possible. Where only one of them can be, it should be the cathode. Treating the anodic metal alone raises the risk of severe localised attack at any defect in the layer. A pinhole in that coating is a very small anode, facing a cathode that has lost nothing.

That reads directly onto a coil. An aircon fin coating protects the aluminium, and the aluminium is the anodic member of the pair, so the integrity of the layer is the whole of its value. A uniformly bare fin corrodes slowly across its entire face. A mostly coated fin with a scraped patch has funnelled the same driver into that patch.

Contact is not even strictly required for the last of these. Corrosion guidance describes a noble metal dissolving in one place and depositing on a less noble metal somewhere else, forming small intense cells wherever it lands. Copper is the metal named for it, and water running off copper onto aluminium is a documented route. That is why a checklist for mixed-metal equipment covers parts that never touch each other.

• •Which metal the fixings are, and whether it matches the support they enter.
• •Whether anything sits between two unlike metals to keep them apart.
• •Whether the joint is sealed, or open to water that can stand in it.
• •Whether a treated coil is washed with a product and a pressure that leave the coating whole.
• •What was found on an existing support before the decision to reuse it.