What a Dissimilar Metal Joint Needs From You

Why this matters

Every shop knows that putting two different metals together in a wet place causes trouble, and most shops respond by trying to avoid the pairing. In the field you often cannot: the equipment is one alloy, the structure is another, and something has to hold them together.

The useful move is to stop treating the metal pairing as the decision. In a joint you are actually building, the pairing is usually fixed and the things that decide whether it lasts five years or five months are three other variables that nobody argues about at the counter. This card is about those three, and about the one you do not control, so you stop spending effort there.

The one you do not control

The driving force in a galvanic couple is the potential difference between the two metals in the electrolyte they are sitting in. That difference is a property of the metals and the environment, and it is not something a technician adjusts. Two sibling articles cover it properly: one explains the mechanism, the other explains how to read a galvanic series table and what the gap between two entries does and does not tell you. Use those for the pairing question and do not re-derive it here.

What matters for this card is a single consequence. Potential difference sets whether current flows. It does not set how fast anything is destroyed. Two joints with the identical pairing, identical potential difference, and identical wetting can corrode at rates that differ by orders of magnitude. What separates them is geometry.

The first thing you control: area ratio

Galvanic corrosion is an electrical circuit. Current leaves the anode, travels through the electrolyte to the cathode, and returns through the metal-to-metal contact. Two things about that circuit decide the damage:

• The total current is limited mostly by how much cathode surface is available to support the reduction reaction. Big cathode, more current.
• The damage is that current concentrated on the anode's surface. Small anode, higher current per unit area, faster metal loss at any given point.

Put them together and you get the rule that matters more than the metal selection: damage scales roughly with the ratio of cathode area to anode area. A large cathode feeding a small anode is the worst configuration there is, and it is exactly the configuration a fastener creates when the fastener is the less noble part.

So the design rule, which sounds backwards until you see the arithmetic: the fastener should be the more noble metal, and the large member should be the less noble one. You want the sacrificial role assigned to the part with a lot of surface to spare, not to the part holding the assembly together.

The second: electrolyte continuity

No electrolyte, no circuit, no galvanic corrosion, regardless of the pairing. Everything in this category is about breaking or interrupting the path:

• A dielectric fitting or an isolating sleeve-and-washer set, which has to isolate the fastener from the hole as well as from the face. A shoulder washer with no sleeve leaves the bolt shank touching the hole and the isolation does nothing.
• Drainage and geometry. A joint that sheds water is a joint that spends most of its life with no electrolyte. A crevice, an upward-facing pocket, or a lap that traps a film keeps the cell alive between rains, and crevices bring their own oxygen-starvation corrosion on top of the galvanic one.
• Sealing the joint so the electrolyte cannot reach the interface at all, using a sealant that stays flexible through the temperature range and does not itself hold moisture against the metal.

The third: which member you coat

This is the counterintuitive one and it is worth getting right, because the instinct is exactly backwards.

Coat the noble metal, the cathode. Do not coat the active metal alone.

Every coating has holidays: pinholes, edges, handling damage. If you coat the anode and leave the cathode bare, the coating on the anode reduces its exposed area to a few pinholes, while the cathode keeps its whole area. You have just built the worst possible ratio on purpose, and the entire galvanic current now concentrates into those pinholes, which perforate.

Coat the cathode and the effect runs the other way: cathode area falls, total current falls, and the anode's exposure is unchanged. Coating both is better still, and if you can only reach one, reach the noble one.

The fourth thing, and it is a safety item

An isolating fitting does exactly what its name says: it breaks electrical continuity. Where metallic piping or structure is serving as a bonding path, that is a hazard you just created.

Before you separate a metallic water or process line, or install an isolating fitting in one, establish whether it is part of the bonding or grounding path and install a bonding jumper across the break in accordance with NEC Article 250 in the edition your authority having jurisdiction has adopted. Fault current that was returning through that pipe will otherwise look for another route, and hands are a route.

Two configurations, one pairing

A bronze component bolted to a plain carbon steel bracket on an outdoor pad. The pairing is fixed by the equipment. Note the bracket is plain rather than galvanized, and that matters: zinc is anodic to steel, so a galvanized bracket would be the sacrificial member protecting the fasteners rather than the cathode driving them, which is a different example with the opposite symptom. The wetted footprint of the bronze component and the bracket it sits against is about 60 in2. The wetted area of the four fasteners together is about 0.6 in2. Illustrative geometry, and the point is the ratio, not the absolute numbers.

Configuration A: steel fasteners. Steel is the more active member here, so the fasteners are the anode and the bronze acts as the cathode. Cathode to anode is roughly 60 to 0.6, about 100 to 1. The entire galvanic current supported by 60 in2 of cathode is being delivered to 0.6 in2 of fastener, and the fastener is also the part carrying the load and the part whose cross-section you cannot spare.

Configuration B: fasteners of the more noble alloy. Now the fasteners are the cathode at 0.6 in2 and the large bracket area is the anode at 60 in2. The ratio inverts to about 1 to 100.

To first order, current density on the anode scales with that ratio, so between the two configurations the anode's rate of attack differs by a factor on the order of 10,000. Same two metals. Same site. Same weather. The only change is which member is small.

Hedge that number honestly: this is a first-order relationship for a fully wetted couple with a continuous electrolyte and good metallic contact. Real rates also depend on electrolyte conductivity, how far apart the members are in the electrolyte, and how each alloy polarizes, and for anything structural or safety-critical the number belongs to a corrosion engineer rather than to this ratio. What survives all of that is the direction and the scale: this is not a 10 percent effect, and no fastener coating rescues configuration A.

Now the coating decision on configuration B. Suppose someone paints the bracket, the anode, and leaves the fasteners bare. A single handling scratch exposing a 1/8 in diameter spot is an area of about 0.012 in2. Against 0.6 in2 of bare cathode that is a ratio near 50 to 1 concentrated into one spot, which is how a painted assembly develops a perforation at a scratch while the rest of the panel looks new. Paint the fasteners instead, or paint both, and the failure mode does not exist.

What the field sees when this is wrong. In configuration A the heads look sound and the shank inside the hole is gone, so the first symptom is a fastener that spins or shears at a torque nowhere near its rating, sometimes with no visible corrosion at all. In configuration B done correctly, the visible symptom is mild general dulling and staining spread across a large bracket surface, which is the outcome you designed for: distributed, slow, and inspectable.

Opening one of these joints later

A dissimilar-metal joint that has been in service is often seized, and the ways people free it are where the injuries are.

• Expect the fastener to shear. Corroded shanks lose section invisibly. Load it steadily rather than with a jerk, and keep your body out of the arc your wrench will travel when it lets go.
• Do not apply heat to a plated or galvanized fastener without knowing the plating and controlling the fume. Heated zinc coatings give off zinc oxide fume, which causes a flu-like illness, and cadmium plating gives off cadmium oxide, which is acutely dangerous at low concentrations and is a carcinogen. That is an inhalation route: local exhaust that captures at the source and respiratory protection under a written program per 29 CFR 1910.134, not a face shield. Heating anything also brings hot work fire prevention requirements under 29 CFR 1910.252 in general industry or 29 CFR 1926.352 in construction, including the fire watch.
• Isolate stored energy first. If the joint is holding pressure, spring load, or the weight of the equipment, lock out under 29 CFR 1910.147 and support the load before the last fastener moves. Where conductors land on the equipment, that half is electrical work under 29 CFR 1910.333(b)(2), proved live-dead-live per NFPA 70E-2021, 120.5, which reaches you through your employer's electrical safety program rather than on its own.
• Grinding a seized fastener releases whatever the coating is, plus the base metal. Same respiratory logic as heating.

What changes the answer

• A dry, conditioned indoor location with no condensation removes the electrolyte, and with it most of this analysis. The pairing that would be reckless outdoors is routine inside a dry space, which is why the same two metals are perfectly normal in one place and a repeat failure in another.
• A joint that is inspectable and serviceable can be designed to corrode slowly and be remade, which is often more practical than designing corrosion out entirely. A joint that is buried, encased, or behind finished work has to be designed to outlast the access, and that raises the standard for isolation.
• Where the two metals are the same but the joint still seizes, this is not galvanic at all. Some alloy pairs, notably stainless on stainless, gall by cold welding of the thread flanks under load, and the fix is a thread lubricant or anti-seize compatible with the service rather than isolation.

How to verify you got this right

• State which member is the anode before you buy fasteners. If the answer is the fastener, change the fastener.
• Check the isolation set is complete: sleeve through the hole and washers on both faces, and confirm with a continuity check that there is actually no metallic path, which takes seconds and catches the missing sleeve every time.
• Confirm the bonding jumper is in place wherever you interrupted metallic piping that may serve a bonding function, before the joint goes back into service.
• Inspect the small member first at the next visit. The large member shows you cosmetic staining; the small one shows you the actual rate.

References

• 29 CFR 1910.134 for respiratory protection under a written program where heating or grinding coated metal releases metal fume
• 29 CFR 1910.252 (general industry) and 29 CFR 1926.352 (construction) for hot work fire prevention, including the fire watch
• 29 CFR 1910.147 for stored energy isolation and 29 CFR 1910.333(b)(2) for electrical isolation, with NFPA 70E-2021, 120.5 for live-dead-live proving where your employer's program adopts it
• NEC Article 250, in the edition your authority having jurisdiction has adopted, for bonding continuity across an interrupted metallic path
• See related: The Galvanic Series as a Field Tool; Dissimilar Metals and Corrosion; What a Transition Fitting Is Actually Managing