← Back to Archive

Corrosion, and the trouble with two metals touching

Galvanic pairs, and the design rules that follow.

Two different metals touching each other in the presence of moisture corrode faster than either metal would alone because the pair forms a small, unintended battery, with one metal giving up electrons preferentially and dissolving away at an accelerated rate while the other metal is left largely protected, and the design trouble that follows is entirely about which metal ends up cast in which of those two roles.

What is really going on

Every metal has some underlying tendency to give up electrons and corrode, a tendency that differs from one metal to the next and is well documented as a ranked order called the galvanic series, the same underlying chemistry an ordinary battery is built around deliberately, two different metals and an electrolyte, wired so the resulting current does something useful instead of simply eating away at the more reactive metal. When two different metals are placed in electrical contact with each other and both are also in contact with a shared electrolyte, anything from seawater to ordinary rainwater carrying a little dissolved salt, the pair completes an electrical circuit, and the metal further along that ranked order gives up electrons preferentially, corroding faster than it would have done sitting alone, while the metal that received those electrons is left corroding more slowly than it otherwise would, protected at the direct expense of its partner. Neither metal has to be inherently poor at resisting corrosion for this to happen, since the entire effect is about the relative standing of the two metals against one another rather than either one's absolute resistance measured in isolation. This is why galvanic corrosion is treated as its own distinct failure mode in design guidance, since a material's corrosion rating alone says nothing about how it behaves once bolted or welded directly against a different metal.

The foil-and-baking-tray comparison

Aluminium foil wrapped around a dish of leftovers and left touching a stainless steel baking tray develops small pinholes exactly where the two metals were in contact by the next morning, especially if the food underneath was salty, the moisture and salt together forming precisely the electrolyte the pairing needs. The foil is eaten away at the points of contact while the tray beneath it remains completely unaffected, not because stainless steel is immune to every kind of corrosion but because, in this particular pairing, aluminium is the one left playing the sacrificial role, giving up its own material to protect the steel it happens to be touching. Wrap the same leftovers against a plain aluminium tray instead, with no second metal in the circuit, and the same foil in the same salty conditions shows nothing like the same damage, since a metal cannot form a galvanic pair with itself.

Why the size of each metal's exposed area matters as much as which metals are used

The rate at which the sacrificial metal is consumed depends heavily on how much of it is exposed relative to the protected metal sharing the same electrolyte, since the corrosion current the pairing generates has to be delivered entirely through whatever area of the sacrificial metal is available to supply it. A small area of a reactive metal in contact with a large area of a more noble one concentrates that current into a small patch, corroding it away rapidly, while the same reactive metal spread across a much larger area dilutes the same total current enough that the damage in any one place is far less severe. This is why a small steel bolt through a large aluminium panel is a far worse combination than a small aluminium rivet through a large steel panel, even though the same two metals are involved either way, since the punishing case is always the one where the smaller area belongs to the metal being sacrificed. Boat fittings make this rule visible quickly, since a small steel screw holding an aluminium fitting to a hull sitting in seawater can corrode away to nothing within a single season, the surrounding aluminium acting as the large, protected partner concentrating the entire corrosion current onto that one small fastener.

The one number worth remembering

The more reactive metal in a mismatched pair can corrode several times faster than it would in isolation, and considerably faster still when it forms only a small area against a much larger area of the more noble partner, severe enough to eat a visible hole through a thin aluminium panel in roughly the time an identical, isolated panel would show barely a mark.

What this changes in practice

Avoiding galvanic corrosion means either keeping dissimilar metals from touching directly, an insulating washer, a plastic bush or a coating breaking the electrical path between them, or deliberately choosing which metal is allowed to be sacrificial and sizing it generously enough to absorb the loss over the design's intended life, the same principle behind a sacrificial anode bolted onto a ship's steel hull to corrode away in the hull's place. Where dissimilar metals cannot be avoided entirely, keeping moisture away from the joint, sealing it, draining it, or simply keeping it dry, removes the electrolyte the whole effect depends on and stops the pairing from doing any real damage regardless of which two metals are involved. Household plumbing runs into the same issue wherever a copper pipe is joined directly to a steel fitting, which is why a dedicated dielectric fitting, an insulating sleeve breaking the electrical path between the two metals, sits between them in plumbing built to last.

Where this stops being true

Galvanic corrosion needs a genuine electrolyte to function, so two dissimilar metals bolted together somewhere permanently dry, indoors and away from condensation, can sit in direct contact for years without meaningful corrosion, since there is no path for the current to flow. The risk also shrinks sharply between metals that sit close together on the galvanic series, aluminium alloys touching one another rather than aluminium touching a strongly dissimilar metal like copper, since a small gap in reactivity drives only a small current even when moisture is present. This is why an assembly built entirely from one family of alloys, aluminium fasteners on an aluminium structure, stainless fasteners on a stainless structure, sidesteps the whole problem automatically, without anyone needing to consult a chart of galvanic standing at all.

More on Materials that misbehave