The short version
- It is not rust welding. It is oxide swelling, and the numbers are startling.
- Aluminum oxide occupies 28 percent more volume than the metal it replaced.
- Converting one thousandth of the bore wall can roughly double the interference fit.
- Which is why a hub that came apart easily at 40,000 miles fights you at 120,000.
Wheel bearing corrosion gets described as the hub rusting into place, which is close but misses the mechanism that actually matters.
It also welds axle splines into hubs. The bearing is not glued in by rust. Wheel bearing corrosion squeezes it in with metal that grew.
The number that explains it
When a metal oxidizes, the oxide takes up a different amount of space than the metal it came from. The ratio has a name.
| Reaction | Volume ratio | Effect |
|---|---|---|
| Aluminum to aluminum oxide | 1.28 | Swells 28 percent |
| Iron to red rust | 2.14 | Swells 114 percent |
| Magnesium to magnesium oxide | 0.81 | Shrinks, flakes away |
A bearing bore is a closed space with nowhere for that extra volume to go. It goes into the joint.
Magnesium is the interesting contrast. Its oxide shrinks and falls off, which is why magnesium corrodes away rather than seizing.
What that does to the fit
A pressed bearing is held by an interference of roughly one to two thousandths of an inch. That is the whole grip.
| Bore wall converted | Added interference |
|---|---|
| 0.2 thousandths | About 0.11 thousandths |
| 0.5 thousandths | About 0.28 thousandths |
| 1.0 thousandths | About 0.56 thousandths |
| 2.0 thousandths | About 1.1 thousandths |
Read the bottom row against an original fit of one to two thousandths. The grip has roughly doubled.
None of that is visible. The bore looks like a bore, and the bearing that slid in with a press now needs several times the force to come out.
Why aluminum knuckles are the common case
Most modern knuckles are aluminum and most bearings are steel. Two dissimilar metals with salt water between them form a battery.
The aluminum is the one that gives up material in that pairing. So the bore corrodes rather than the bearing.
Road salt turns intermittent moisture into a continuous electrolyte. The reaction runs all winter, every winter.
A ten year old car from a salt region can need many times the removal force of the same car from a dry one. Nothing looks different.
How that plays out during removal, and the techniques that work, is on the stuck hub page.
Preventing the next one
This is the part worth doing while the corner is already apart, because it is the only time it is easy.
- Clean the bore back to bright metal. Not shiny, but free of the white powdery oxide and any loose scale.
- Do not remove metal. A wire wheel is fine. A grinding stone or emery on the bore changes the fit permanently.
- Check the bore is still round and unscored before anything goes back in.
- Apply a thin anti-seize suited to dissimilar metals, sparingly, on the mating surfaces only.
- Keep it off the bearing races and the encoder, where it does harm rather than good.
Anti-seize is a barrier, not a lubricant for the fit. A heavy coat can affect how the bearing seats and how the fit behaves.
A film you can barely see is what is wanted, and it is enough to break the metal to metal contact that starts the reaction.
When the bore is already damaged
The mechanism that loosens a fit rather than seizing it is on the spun bearing page.
Sometimes wheel bearing corrosion has gone past prevention. The oxide came away with the metal it grew from, and the bore is now oversize.
That means no interference is left to hold a new bearing. The fix is a knuckle rather than more cleaning.
How to tell, and what it adds to the bill, is on the bearing or whole hub page.
The bolt-on case
Bolt-on hubs suffer the same chemistry elsewhere. Wheel bearing corrosion attacks the pilot register and the flange face instead of a bore.
The result is a hub that will not release from the knuckle even with every bolt removed. Same cause, same prevention, different surface.
Which design you have is on the design comparison, and it decides which surfaces to treat.
Common questions
What causes wheel bearing corrosion to seize a hub?
Oxide swelling. Aluminum oxide takes 28 percent more volume than the aluminum it replaced, and that extra volume goes into the interference fit.
How much does that tighten the fit?
Converting one thousandth of the bore wall adds roughly half a thousandth of interference on the diameter, against an original fit of one to two.
Why does aluminum corrode rather than the steel bearing?
Because the two metals in contact with salt water form a cell, and aluminum is the one that gives up material in that pairing.
Does road salt really make that much difference?
Yes. Salt turns occasional moisture into a continuous electrolyte, so the reaction runs all winter instead of briefly.
How do I prevent it next time?
Clean the bore to bright metal without removing any, then apply a thin anti-seize on the mating surfaces only, keeping it off the races and encoder.
Can I grind the bore clean?
No. Removing metal changes the fit permanently, and there is only one to two thousandths of it to lose.
What if the bore is already oversize?
Then no interference is left and the corner needs a knuckle. More cleaning cannot restore material that has gone.