The Fastener That Was the Wrong Material

Why this matters

A fastener that fails on material does not look like one that fails on load. It does not shear, it does not strip, and it usually does not fail while anyone is watching. It thins, it cracks in a plane you did not expect, or it welds itself into the hole and takes the part with it on removal. It keeps happening because the bin on the van is organized by size and grade, the two properties that tell you nothing about whether the bolt survives where you are putting it.

Before removing corroded fasteners from anything carrying load or spring tension, support the load independently and de-tension any spring, counterweight, or loaded member first, then confirm it is de-tensioned before the last fastener comes out. A corroded fastener can be the only thing holding an assembly that was never designed to be held by one.

A fastener carries four material decisions, not one

When you pick a fastener you are making four separate calls, and the bin usually only labels the third one.

1. The base alloy family. Carbon steel, one of the stainless families, aluminum, brass, silicon bronze. This is the corrosion behaviour that remains after every coating is gone, which is the behaviour you eventually get.
2. The coating system and its thickness class. Not "zinc" but which zinc process and how much of it. This sets how long you get before the base alloy is on its own.
3. The strength grade or property class. What load it holds and, past a hardness threshold, failure modes the lower grades do not have.
4. The galvanic pair it forms with what it clamps. A fastener is never alone, and that relationship decides which of the two metals corrodes.

Get all four right and the fastener outlives the joint. Get the fourth wrong and it does not matter that the other three were perfect.

What is deliberately not on the list

This is where most of the wrong calls come from, because each of these feels like it belongs on the list and none of them does.

• The grade number. Grade and property class are strength ratings and they say nothing about corrosion. A higher grade in the same base steel with the same plating corrodes at the same rate as the lower one. Reaching for a stronger bolt to fix a corrosion failure is the most common wrong move here, and the worked example below shows it making the joint worse rather than neutral.
• Whether a magnet sticks to it. Austenitic stainless is usually weakly magnetic or non-magnetic and gets more magnetic after cold working, which is exactly what forming a thread does. A magnet on a threaded fastener is not a grade test.
• The colour of the finish. Bright, yellow, black, and olive are conversion-coating colours and the convention is not standardized across suppliers. Colour is a housekeeping aid inside one bin, not a specification.
• "It is the same size and thread." Correct fit is the entry requirement, not the decision. An assortment bin is organized entirely around this property, which is exactly why it cannot carry a material decision.
• The torque value. Torque belongs to assembly, not to selection. A material-mismatched fastener torqued perfectly still fails.
• What came out of the hole. The removed fastener is evidence about what was installed, not about what should have been. If it failed on material, the identical replacement fails on the same schedule, and now the failure is yours.

The pairing rule: make the small part the noble one

Two different metals in electrical contact with moisture between them form a cell. The more active metal corrodes and protects the more noble one. Zinc protects steel, steel protects copper, aluminum protects stainless.

What decides how fast is area ratio, and that is what makes fasteners a special case. Corrosion current spreads across the whole cathode but concentrates on the anode, so a small anode against a large cathode drives the entire attack into a small piece of metal.

• A noble fastener in an active member is usually survivable. A stainless screw in an aluminum sheet makes the aluminum the anode, but its area is huge compared to the screw, so the attack spreads thin: local staining and slow pitting around the hole rather than a failure.
• An active fastener in a noble member is the one that disappears. A plain steel or aluminum fastener into a stainless or copper member makes the small part the anode, and it thins visibly in a season or two.

So the working rule is to make the fastener the more noble member of every pair, and to isolate where you cannot: an insulating washer and sleeve, a dielectric fitting, an intact coating, or a sealant that keeps water out of the interface. Cleaning a galvanic joint and reassembling it without isolation buys one cycle of the same clock.

Where the coating lives and where it does not

A coating is a consumable, and its service life scales with how much of it is there.

Electrodeposited zinc under ASTM B633 is specified by service condition class, and ordinary commercial hardware typically carries the thinnest of them, on the order of 0.0002 inch. Hot-dip zinc on hardware under ASTM A153 is a different process producing a far heavier layer, on the order of 0.002 inch for fasteners. That is roughly a factor of ten, and because zinc consumes at a fairly steady rate for a given exposure, the heavier coating buys roughly a factor of ten in time before the base steel is exposed - other things equal, and only where the coating is actually intact.

The "actually intact" clause is where field practice breaks it. Cutting a bolt to length leaves bare steel at the cut, on a horizontal end where water sits. Field drilling leaves a bare bore. A washer dragged under a power driver scrapes the coating off the bearing face, which is the tightest and wettest interface in the joint. And hot-dip coatings are thick enough that the fasteners are made with the coating allowed for, so running a standard tap through a hot-dip nut to make it turn easier removes the protection you ordered.

If you are specifying rather than grabbing, the coating class is the field that carries the corrosion decision, and it is the field most orders leave blank.

Where strength turns against the fastener

Two mechanisms make the strongest fastener the wrong fastener, and both are environment-dependent rather than load-dependent.

Hydrogen embrittlement. High-strength steel fasteners can absorb hydrogen during acid cleaning or electroplating, and hydrogen in a hardened steel produces delayed brittle cracking under sustained tension, sometimes hours or days after assembly, at loads well below rating. Susceptibility rises sharply with hardness and the commonly cited threshold is around 35 HRC, which brings the higher common grades and property classes into scope and leaves the ordinary mid-grades below it. The mitigation is a post-plating bake, which the supplier either performs or does not and which you cannot see on the part. That is why "use the stronger bolt in a wet location" is not a neutral choice.

Chloride stress corrosion cracking. Austenitic stainless is vulnerable to cracking under the combination of tensile stress, chlorides, and heat, with published guidance generally putting the risk in play above roughly 140 degrees F in chloride-bearing service; at ambient temperature the more likely stainless problem is crevice corrosion in the tight gaps under heads and in threads, where the passive film cannot re-form. Both say the same thing: "stainless" is a family, not an answer, and chlorides move the decision within it.

Worked example: the stand bolts replaced with stronger ones

An outdoor condensing unit on an aluminum stand near a pool deck. The mounting bolts, plated steel, are heavily rusted at three service years and two of them shear at the head during removal.

The move that feels right and is not. The tech reaches for the same size in a higher grade, same plated finish, on the reasoning that the old ones failed so use stronger ones. Check it against the four decisions. Base alloy: unchanged, still carbon steel. Coating: unchanged, still thin commercial plating. Galvanic pair: unchanged, still steel against an aluminum stand. Strength grade: raised, into the hardness range where an electroplated fastener under sustained tension carries hydrogen embrittlement risk. Three of the four decisions are untouched and the fourth got worse. The corrosion clock is identical and a new failure mode was added.

What the coating arithmetic says. The plating went in three years in this exposure. Specify hot-dip hardware instead and the coating is roughly ten times thicker, so at a steady consumption rate the coating alone should protect for roughly an order of magnitude longer in the same environment. That is the single change with the largest effect, and it costs nothing but a line on the order.

What the pairing says. A steel fastener into an aluminum stand makes the aluminum the anode. The stand is large and the bolts are small, so this is the survivable direction, and the damage should show as local staining at the stand's bolt bosses rather than as destroyed bolts. Which means the bolts did not fail galvanically at all. They failed on plain wet corrosion of an under-coated base steel in a chlorinated, splash-exposed location, so the fix is coating and base alloy, not isolation.

What would flip it. Put the same unit on a stainless stand and the picture inverts: plated steel bolts become the small anode against a large noble member, they thin fast, and the answer moves to a stainless or silicon bronze fastener with an isolating washer at the interface. Move it indoors into a dry mechanical room and the original plated hardware was always adequate.

The failure mode already happened here: two heads sheared during removal. Corrosion-thinned fasteners break rather than back out, so a twenty-minute swap becomes a drilling-and-tapping job on a stand that now has to be supported before the last bolt is touched.

Reading a fastener that failed on material

Pull the failed hardware and look before you throw it away, because the failure surface names the mechanism.

• Uniform thinning with general rust is ordinary wet corrosion of a base steel with the coating gone. Fix the coating class.
• Attack concentrated at one interface, clean elsewhere, is galvanic. Fix the pair or the isolation.
• A bright, granular, flat break with no necking down in a high-grade fastener is brittle fracture, and in a plated high-strength bolt under sustained load, hydrogen embrittlement is the first candidate.
• Fine branching cracks in a stainless part, often otherwise bright and clean, points to stress corrosion cracking rather than corrosion loss.
• A stainless fastener seized solidly in a stainless hole is galling, not corrosion - two matching alloys cold-welding under load. Anti-seize on assembly is the answer, and because it changes the torque-to-preload relationship, the torque value has to change with it.

Checking you got this right

• Name all four decisions out loud for the fastener you are about to install. If you cannot name the coating class, you did not make decision two, you inherited it.
• Identify what the fastener touches at both ends and say which member is the anode. If the fastener is the anode and it is the small part, stop.
• Check whether anything in the install removes coating: cutting to length, field drilling, running a die, or driving a washer under power.
• Where you used anti-seize, confirm the torque value you are applying was set for a lubricated joint. A dry-thread value on lubricated threads overloads the fastener.
• On a repeat corrosion failure, check the record for what went in last time. If nobody wrote down the coating class, the next failure is as undiagnosable as this one.

References

• ASTM B633, electrodeposited zinc coatings on iron and steel, for service condition classes and coating thickness
• ASTM A153, zinc coating (hot-dip) on iron and steel hardware, for fastener coating weights
• SAE and ASTM fastener grade standards for property class, hardness ranges, and post-plating baking requirements for high-strength fasteners
• Manufacturer and standards-body guidance on galvanic series and dissimilar-metal isolation practice
• See related: Dissimilar Metals and Corrosion; Common Fastener Types Reference; Reading Rust and Corrosion Patterns