Put two different metals in electrical contact with moisture between them and you've built a small battery — one metal corrodes faster than it would alone, and the other corrodes slower. Which one loses depends on a ranking most designers have never seen, and getting it backwards is a specific, well-documented, entirely avoidable mistake.
Every metal has a natural electrochemical potential. Connect two different metals with an electrolyte (rain, condensation, salt spray) bridging them, and the less noble one becomes the anode and corrodes preferentially — sacrificed to protect the more noble cathode. Aluminium sits well toward the active/anodic end of the galvanic series; mild steel sits closer to the middle; stainless steel (with its passive film intact) and copper sit toward the noble end — which is exactly why aluminium in contact with steel corrodes, and aluminium in contact with copper corrodes badly.
| Position | Metal | Behaviour when paired with something below it |
|---|---|---|
| Most active (anodic) | Zinc, aluminium alloys | Corrodes to protect the more noble metal |
| Middle | Mild/carbon steel | Anodic to stainless and copper, cathodic to zinc and aluminium |
| Noble | Stainless steel (passive), copper alloys | Protected — at the expense of whatever's paired with it |
| Most noble | Titanium, gold, graphite | Essentially always the protected member |
A small anodic area paired with a large cathodic area corrodes far faster than the reverse, because the same total galvanic current concentrates into whatever anodic area is available. Small aluminium fasteners holding a large steel panel is a bad pairing made worse by area — the fasteners are the anodic member, and the same total galvanic current is forced through their tiny surface, producing fast, deep, localised attack right at the fastener. The same pairing with the areas reversed — a large aluminium panel with a few steel fasteners — is far more forgiving, because that current spreads thin across the much bigger anodic aluminium surface.
Zinc plating and hot-dip galvanizing protect steel sacrificially — zinc is more anodic than steel, so it corrodes first, and it keeps protecting a small scratch because the surrounding zinc is still sacrificing itself. Anodising thickens aluminium's own natural oxide layer electrochemically — it's not a coating so much as a deliberately grown, thicker version of what aluminium already does on its own, and it's why anodised aluminium can be dyed. Passivation for stainless steel is a chemical treatment, not a coating at all — it strips free iron off the surface so the alloy's own chromium can form its passive layer properly. Powder coating is different from all three: it's a barrier coating only, with no sacrificial chemistry behind it — scratch it, and the exposed spot isn't protected the way a scratch in galvanized steel still is.
The bolts didn't fail — the aluminium around them did, because nobody checked what sat next to what before spec'ing the fasteners. I check the pairing and the area ratio on any assembly with dissimilar metals before I check anything else about the joint, because the fix — a nylon washer, a different fastener alloy, keeping the anodic member the larger one — costs nothing at the design stage and is very expensive to retrofit once the part's in the field.