# Galvanic corrosion

Dissimilar conductive materials in wet service: electrical and electrolyte paths, galvanic-series limits, wetted areas, isolation, coatings and investigation.

Reviewed: 2026-11-02

Tags: production, process-effects, corrosion, dissimilar-materials

In a wet couple with **100 cm²** of cathode and **10 cm²** of anode, an imposed **2 mA** yields **0.2 mA/cm²** average anodic current density. If coating damage exposes just **1 cm²** of anode and the same cathode stays wetted, density becomes **2 mA/cm²** only if current remains **2 mA**. Geometry can also change current. This model gives neither local penetration nor service life.

ASSDA identifies an electrolyte, electrical contact and sufficient potential difference. The active member is anodic. A fastener, pipe or wire can close the electrical circuit. [ASSDA, Technical FAQ 1: Galvanic/Dissimilar Metal Corrosion](https://www.assda.asn.au/images/2020/FAQs/ASSDA_Technical_FAQ_1_-_Galvanic_Dissimilar_Metal_Corrosion.pdf)

Check conductive paths and common wetting even when the exposed surfaces are separated.

## Assembly and environmental boundaries

Alloys, coatings, exposed substrate, fasteners, washers, inserts and conductive gaskets or composites can participate in the circuit. Intended and accidental electrical paths both matter. A coating specification on the drawing does not establish that the coating remains intact at a thread, cut edge or bearing surface after assembly.

| Input | Question to resolve |
| --- | --- |
| Conductive members | What materials and surface states are actually exposed? |
| Electrical path | Are they connected through fasteners, wiring or another structure? |
| Electrolyte path | Can rain, condensation, process fluid or retained deposits connect them? |
| Wetted areas | How much anodic and cathodic surface participates locally? |
| Service history | What changes during drying, wear, cleaning or maintenance? |

Do not confuse galvanic corrosion with every form of attack at a joint. Crevice conditions, pitting, general corrosion, fretting and contamination can coexist with galvanic effects. A photograph of rust near a stainless fastener does not by itself identify the cause or quantify remaining strength.

## A galvanic series is environmental data

A galvanic series ranks measured corrosion potentials in a specified environment; it is not a universal material compatibility chart. [ASSDA, galvanic-series discussion](https://www.assda.asn.au/images/2020/FAQs/ASSDA_Technical_FAQ_1_-_Galvanic_Dissimilar_Metal_Corrosion.pdf)

 Surface condition matters, particularly whether stainless steel remains passive. Outokumpu notes that passive stainless steels can accelerate attack on less noble coupled materials and conditions its discussion of different stainless grades on their remaining passive. [Outokumpu, corrosion resistance: Galvanic corrosion](https://www.outokumpu.com/en/expertise/stainless-basics/corrosion-resistance)

The data depend on electrolyte, temperature, aeration, flow and surface state. Potential separation alone is not a quantitative corrosion-rate model. It does not capture the available cathodic reaction, electrical resistance, electrolyte access or localized breakdown.

For an edition-specific example, the official **MIL-STD-889** record identifies **Revision D, July 21, 2021**, validated by **Notice 1, April 29, 2026**. Its scope says the underlying data were collected under **static immersion in artificial seawater according to ASTM D1141 without heavy metals**. That scope is a boundary on using its data, not a universal prediction for an atmospheric joint or process liquid. The validation notice is not a new technical revision. [DLA ASSIST, MIL-STD-889 document record](https://quicksearch.dla.mil/qsDocDetails.aspx?ident_number=36032)

## Wetted area changes the severity

Outokumpu warns that a small less-noble surface coupled to a large noble surface can corrode rapidly, using carbon-steel bolts in stainless sheet as an example. It also identifies defects in coatings on the less-noble member as a way to create a small exposed anodic area. [Outokumpu, area ratio and coating defects](https://www.outokumpu.com/en/expertise/stainless-basics/corrosion-resistance)

The relevant quantity is **participating wetted area**, which can differ from the total area in a CAD model. ASSDA describes how drying can change the local wetted-area ratio around a stainless fastener in aluminum. It also notes that paint can be damaged during assembly and that stainless steel's passive film is not an electrical insulator. [ASSDA, wetted areas and electrical isolation](https://www.assda.asn.au/images/2020/FAQs/ASSDA_Technical_FAQ_1_-_Galvanic_Dissimilar_Metal_Corrosion.pdf)

A small cathode/large anode arrangement can reduce one concern without establishing immunity. Retained liquid, local deposits and coating damage can change the effective geometry. Avoid declaring a joint safe from nominal fastener size alone.

## Original area/current-density example

Assume a hypothetical wet couple has **100 cm²** of participating cathodic surface and **10 cm²** of exposed anodic surface. Its cathode-to-anode area ratio is **10:1**. If, purely as an imposed analytical assumption, total galvanic current is **2 mA**, the average anodic current density is:

`2 mA / 10 cm² = 0.2 mA/cm²`

Now assume coating damage leaves only **1 cm²** of exposed anode, while the same 100 cm² cathode remains wetted. The area ratio becomes **100:1**. **If the current remained 2 mA**, average anodic current density would become **2 mA/cm²**, ten times the earlier value.

This original example illustrates concentration of a given current onto a smaller area. It does not predict that current remains constant when geometry changes, that attack is uniform or that the thickness-loss rate increases exactly tenfold. No service life or allowable defect size follows from it. Those require the material reaction, actual current distribution and exposure history.

## Interrupt the circuit without breaking the product

Design options include compatible material selection, electrically isolating the members, excluding the electrolyte, drainage and protective surface systems. Isolation effectiveness depends on all conductive contact paths, including bolt shanks and washers, and on assembly damage, compression and environmental aging. A washer under the head does not isolate a conductive bolt shank touching both members.

Coating strategy needs the whole couple. Outokumpu recommends considering protection of the nobler member to reduce the galvanic effect; coating only the more active member can concentrate attack at its defects. This is a design principle, not a universal instruction to paint every electrical interface. [Outokumpu, galvanic design measures](https://www.outokumpu.com/en/expertise/stainless-basics/corrosion-resistance)

Required grounding or bonding changes the decision. DOE's solar-PV guidance gives an example of a bonding device deliberately piercing anodizing to establish electrical continuity and explains that corrosion at such connections can increase resistance. An insulating barrier can defeat a required safety-current path; corrosion protection and the qualified bonding function therefore need a compatible design. [DOE FEMP, Managing and Mitigating Solar PV Corrosion](https://www.energy.gov/cmei/femp/managing-and-mitigating-solar-pv-corrosion)

Sacrificial protection intentionally uses an active member to protect another material. That is an engineered system with consumption and maintenance requirements, not evidence that any zinc-coated fastener will protect an arbitrary assembly indefinitely.

## Evidence limits for a proposed solution

Attack location, wetting patterns, actual material/coating conditions and electrical continuity help distinguish competing explanations. Corrosion-product volume is not a direct measure of local metal loss or remaining function.

Evidence from nonrepresentative materials, area ratios, assembly conditions or exposures may not describe the intended joint. A generic salt-spray duration or room-temperature immersion result does not independently establish field life in cyclic wet/dry service. Section loss, coating condition, mechanical integrity and electrical resistance are different performance measures. Maintenance that changes a fastener or coating can also change the electrical circuit and required bonding function.

## Source versions

Reviewed **2026-11-02 PST**. MIL-STD-889 dates and test-environment scope above come from the official DLA record; this page does not claim to reproduce its full technical requirements. No separate technical issue date was verified in the retrieved ASSDA FAQ PDF, Outokumpu page or DOE guidance; the ASSDA URL's “2020” directory is not treated as an edition. The numerical example and explanatory synthesis are original and bounded.

## Related reference

[Manufacturing process references](../../processes/index.md)