The Galvanic Series as a Field Tool

Why this matters

Most people meet the galvanic series as a poster in a supply house and use it as a compatibility chart: find both metals, see how far apart they are, decide. That works often enough to be dangerous, because the table is a ranking produced under one specific set of conditions and your joint is not in those conditions. This card is mostly about what the table leaves out, since the omissions are where the field failures come from and the table itself is the easy part.

What the series actually is

A galvanic series is a list of metals ordered by the electrical potential each one settles at when it sits in a stated electrolyte, measured against a stated reference electrode. Nothing more. It is a measurement of one property, in one liquid, at one temperature.

The classic table is measured in flowing seawater against a saturated calomel electrode, and typical published values run roughly like this, from active to noble: magnesium near -1.6 volts, zinc near -1.0, aluminum alloys around -0.75 to -0.95, mild steel and cast iron around -0.6 to -0.7, lead and tin near -0.5, brass and copper around -0.3, silicon bronze near -0.25, passive 304 and 316 stainless within about -0.1 of zero, titanium near -0.05, graphite positive of zero. Treat those as approximate and read the range from the table in front of you, because different sources publish different values for the same alloy and they are all telling the truth about their own electrolyte. ASTM G82 is the standard guide for developing and using a series like this, and it is worth knowing it exists because it frames the series as a prediction of direction rather than as a compatibility verdict.

The practical consequence of the definition: the series predicts direction, not rate. It tells you which metal will corrode. It tells you nothing about whether that will take four months or forty years.

Why your table and mine disagree

Before you compare two numbers, check that they came from the same table. Three things vary between published series and any of them will change your answer:

• The electrolyte. Seawater, fresh water, soil, and a condensate film all produce different potentials and sometimes different orderings.
• The reference electrode. Values against a saturated calomel electrode and values against a copper/copper-sulfate electrode differ by a fixed offset. Mixing two tables shifts every comparison.
• The alloy state. Whether the stainless entry is listed passive or active, whether the aluminum is a specific wrought alloy or a generic entry, whether the steel is bare or coated.

Never mix rows from two tables. If you only have one metal on the table you are holding, find a table that has both rather than importing a number.

MIL-STD-889 handles this by publishing an anodic index instead of raw potentials and by stating an allowable index difference that tightens as the environment gets harsher. Use its own tables and its own environment classes together; the allowance is not portable to a different index.

The gap is the number you use

Position on the list is trivia. The difference between two positions is the working number, and it is the input to whatever allowance your standard sets. Two metals both near the noble end can still be a bad couple if the gap between them is wide relative to the exposure, and two metals both near the active end can be a fine couple because the gap is small.

Compute the gap, then judge it against the environment: an indoor, conditioned, dry-most-of-the-time joint tolerates a far wider gap than a joint that gets wind-driven rain or washdown, and a joint near salt or pool chemistry tolerates the least of all.

What the table leaves out, on purpose

This is the section worth keeping. Five things the series does not contain, each of which has flipped a real decision:

It leaves out area ratio, which usually dominates it. A wide couple with a large active member and a small noble member corrodes slowly and visibly. A narrow couple with a small active member and a large noble member can fail faster. The table cannot see the geometry of your joint, so a table-only decision is half a decision at best.

It leaves out the electrolyte's conductivity. A galvanic cell needs current to flow, and current needs a conductive path through the liquid. Distilled or very soft water conducts poorly and limits the cell; salted water, softened water with high chloride, or a wet mortar bed conducts well and lets the cell run at full strength. Same two metals, same gap, order-of-magnitude difference in rate.

It leaves out coatings and plating, which are what is actually in the cell. The surface participates, not the substrate. A zinc-plated steel screw behaves as zinc until the plating wears through at the contact face, then it behaves as steel, and the whole comparison changes mid-service. Any table decision involving a plated part carries a hidden expiry date set by the plating thickness.

It leaves out the third metal. Solder, brazing filler, an anti-seize compound, and a thread sealant all put a fourth material into the joint. Graphite-loaded anti-seize is the one that catches people, because graphite sits at the noble end of the same table they just read, so smearing it on a steel or aluminum flange in a wet location adds a large, permanent cathode to the joint they were trying to protect.

It leaves out state changes in a single alloy. Stainless is listed twice on any honest table, passive and active, because the alloy moves between those states depending on whether its oxide film has oxygen available. A joint can contain both states of the same alloy at once, which the table has no way to express, and which is the mechanism behind most stainless failures in the field.

Worked comparison: four fasteners into one bracket

A 316 stainless bracket mounts equipment on an exterior wall. You need fasteners. Four are on the shelf. Use the approximate seawater values above and take passive 316 at about -0.05 volts as the base.

Check the arithmetic on the two that matter: -0.05 minus -0.80 is 0.75 volts, and -0.05 minus -1.0 is 0.95 volts. The ranking is unambiguous and the table earned its place.

Now apply the part the table does not know. Every one of the last three puts the active metal in the small member. The fastener is a fraction of the wetted area of the bracket, so all three run the cell in the concentrating direction, and the gap only sets how fast. The aluminum and the plated steel are not "acceptable at 0.75 and 0.95 volts if the environment class allows," they are the wrong architecture regardless of the gap, because the part being consumed is the part carrying load.

That leaves 316 into 316 at a gap of zero as the first choice, and silicon bronze as the fallback if the stainless is not available in the size, on the grounds that a 0.20 volt gap is small enough that the unfavorable area ratio has very little to drive it. The table ranked four options. The area-ratio rule is what rejected the two widest gaps outright and demoted the third to a conditional fallback, and that condition is a number the table cannot supply.

What would change the answer: if the bracket were the small member and the fastener the large one, which happens with a heavy through-bolt carrying a thin clip, the ranking reverses and the gap becomes the governing number again. The rule is not "stainless fasteners always," it is "the small member is the noble one," and the table tells you which is which.

Where the same table inverts

Two documented cases where reading the series straight gives the wrong answer.

Hot water reverses zinc and steel. In cool water, zinc is active to steel and a galvanized coating protects the steel underneath by corroding first. That is the entire premise of galvanizing. Above roughly 140 degrees F, in many domestic water chemistries, the polarity of the zinc-steel couple can reverse, so the zinc becomes noble to the steel and the coating starts driving corrosion of the steel it was applied to protect. Any pinhole in the coating then becomes a small anode facing a large cathode, and the pipe pits through from the inside. This is why galvanized steel on a domestic hot water line behaves worse than the same pipe on the cold side, and no room-temperature table shows it. Before opening any hot water line to inspect for this, close the isolating valve, relieve pressure to zero at an open drain or a gauge, and let the line cool; water at storage temperature scalds on contact and pressure holds it against you when the joint releases.

A single alloy fights itself in a crevice. Take the 316 bracket from the worked example, bolt it flat against a painted surface with stagnant chloride-bearing water trapped in the joint face, and the stainless inside that crevice loses its oxygen supply, cannot maintain its passive film, and drops toward its active potential. Now the metal inside the crevice is active and the metal outside is passive, on the same part, with a gap of several tenths of a volt and an extremely unfavorable area ratio because the crevice is small and the exposed bracket is large. It pits under the joint face where nothing is visible until the fastener pulls through. The table has both rows. It has no way to tell you they can both be your part.

How to verify you got this right

Four checks, and every one of them takes seconds:

1. Both numbers came from the same table, same electrolyte and same reference electrode. If you cannot name the electrolyte your numbers came from, you have a ranking, not a measurement.
2. You wrote down the gap, not the two positions. A decision that names two metals without naming the difference between them has skipped the only arithmetic in the method.
3. You named the small member and it is the noble one. If the small member is active, the gap is not the deciding factor and no allowable-difference table will save the joint.
4. You checked whether anything in the joint is plated, coated, or has a graphite-bearing compound on it. Those change which surface is in the cell, and two of the three change it partway through the part's service life.

If a joint you built to this method comes back corroded anyway, the first thing to re-examine is not the table. It is whether the electrolyte turned out to be something other than what you assumed, because a condensate drip you did not know about is the most common way a correct material decision produces a failed joint.

References

• MIL-STD-889, Dissimilar Metals, for anodic index values and allowable index differences by environment class
• ASTM G82, Standard Guide for Development and Use of a Galvanic Series for Predicting Galvanic Corrosion Performance
• Manufacturer and supplier documentation for plating thickness, coating service life, and approved fastener alloys for a given base metal
• See related: Galvanic Corrosion and the Metals That Fight; How to Avoid a Dissimilar Metal Joint You Will Regret