The short version
- A new bearing failed again means something is still present. Bearings do not fail twice by chance.
- Four causes cover almost all of it: an oversize bore, installation damage, a live contamination source, or the wrong corner.
- The timing tells you which. Weeks points at installation, months points at the bore or water.
- Keep the second failed part. It carries the evidence you need for a warranty argument.
The first failure was bad luck. A new bearing failed again within a year is not bad luck. Treating it as such is how people reach a third one.
Something at that corner is still doing the damage, and it is usually identifiable.
Why a new bearing failed again
A quality bearing fitted correctly into a good bore lasts a very long time. Two failures in quick succession is a pattern, not a coincidence.
The useful question is not which brand to try next. It is what has not changed between the two failures.
What normally ends a bearing is on the failure causes page.
The four causes, and how the timing separates them
| Lasted | Most likely cause | Confirm by |
|---|---|---|
| Days to weeks | Installation damage | Denting marks on the raceway |
| Weeks to a few months | Oversize knuckle bore | Measuring the bore |
| Months, with water history | Failed seal or deep water | Milky grease, rust inside |
| Never quiet from day one | Wrong corner replaced | Road test the other side |
| Months, heavy modification | Spacers, tires, load | What changed on the car |
If the previous bearing spun in the knuckle, the bore is oversize. Every bearing fitted after it will loosen.
No brand of bearing fixes this, and the second failure usually enlarges the bore further.
Measuring it is on the bore measurement page, and the repair options are on the bore repair page.
Installation damage, which hides well
Pressing on the wrong race dents every contact point the load crossed. So does an impact gun on the axle nut.
Those dents are evenly spaced and they produce noise within a few thousand miles rather than immediately.
Reading those marks is on the failure reading page, and the pressing error is on the race driver page.
The one that is not a failure at all
If the noise never actually went away, the original diagnosis was wrong and the new bearing is fine.
Turning left loads the right side, and that inversion is where most people identify the wrong corner.
That case is covered on the still humming page, and finding the corner is on the which corner page.
What to do before fitting a third
- Keep the failed part and look at the raceway before anything else.
- Measure the bore, because this is the cause nobody checks.
- Ask how it was fitted, specifically what the axle nut was tightened with.
- Look for water history, milky grease or rust staining inside.
- Road test both sides before assuming the new part is the problem.
Where you stand on a warranty
A parts warranty covers a defective bearing. It does not cover one destroyed by an oversize bore or a fitting error.
That is exactly why the old part matters, because the damage pattern is what decides which of those it was.
How those claims get judged is on the warranty page.
Common questions
Why has my new bearing failed again so quickly?
Almost always an oversize knuckle bore, damage during fitting, a live water source, or the wrong corner having been replaced.
Does the timing tell me anything?
A great deal. Weeks points at installation damage, months points at the bore or water ingress.
Should I try a better brand?
Not yet. Two failures in a row is a corner problem. A better bearing in a bad bore fails the same way.
How do I check the bore?
Measure it against the bearing outer diameter. If the old bearing came out easily and the bore is polished, that is the answer.
Can an impact wrench have caused it?
If it was used to tighten the axle nut, yes. The shock dents the raceways while the assembly is unclamped.
What if the noise never went away at all?
Then the bearing probably was not the fault, or the wrong corner was replaced. Road test both sides.
Will the warranty cover the second one?
Only if the part was defective. Damage from a bad bore or poor fitting is normally declined, which is why the old part is evidence.