Why You Shouldn't Reuse a Bearing After You Pull It
The SKF 6208-2RS1/C3 you just pulled still spins on the bench. The seals look intact. There is no spall you can see with a flashlight. Putting it back feels like saving a couple hundred dollars and a trip to the crib. It is how that motor, fan, or gearbox becomes a two-outage job.
Quick answer: A pulled bearing is a used bearing that has also been overloaded in a way it was never designed for. Puller jaws, a hammer, a torch, or a three-jaw that cocked on the way off will brinell raceways, smear tracks, stretch 2RS lips, and oval a ring that still "spins fine" with no load on it. Scrap the pulled unit. Put a new one in. If the shaft or housing is already chewed up, the new bearing is not enough either — fix the seat, and if the seal track is grooved, an SKF Speedi-Sleeve is the usual field repair.
What a puller actually does to a bearing
A rolling bearing is built for Hertzian contact that rolls. Extraction is a static overload applied through the wrong path.
On a motor or pump, the inner ring of a 6208 or 6308 is an interference fit. The outer ring is usually a clearance or light transition fit in the end bell. The correct pull is axial force on the tight ring, collinear with the shaft, with the rolling elements carrying as little of that force as possible. That is what a proper Kukko external or internal puller is for.
What usually happens on the floor:
- Jaws on the outer ring while the inner ring is still locked on the shaft. Extraction force goes through the balls or rollers into the raceways. That is true brinelling.
- A two-jaw puller cocked 5–10°. One raceway takes a smear, the opposite side takes an indent. The bearing still rotates by hand.
- Jaws that bite the 2RS lip instead of the outer-ring face. The seal stretches, the garter spring jumps, and the lip never tracks again. The bearing still "looks sealed."
- A hammer on the inner-ring face, or a punch walked around the race. Local brinell plus a rolled-over corner that will not sit against the shaft shoulder.
- Heat on the shaft instead of the inner ring. The journal grows, the fit gets tighter, and you pull harder. The ring comes off oval.
Spherical inserts (UC208, UK208, 1000-series) fail the same way, plus one extra: set screws that were torqued into a commercial shaft leave a burr and a local bore bruise. Pull that insert and reuse it and you have a damaged bore trying to clamp a damaged shaft.
Damage you will not see without a load
| What happened during the pull | What you see on the bench | What happens under load |
|---|---|---|
| True brinelling (static overload through the rolling elements) | Smooth-looking race; maybe a dull spot if you clock it under a light | Balls/rollers drop into the indents once per rev. Vibration starts. Spall follows the indent pattern. |
| Smeared / skewed raceway | Slight roughness, "it has a little catch" | Edge-loaded contact, heat, grease breakdown in hours to days |
| Stretched or rolled 2RS / 2RS1 lip | Seal still in the groove, spins quietly dry | Contaminant path is open. Grease walks out. Water and dust walk in. |
| Inner ring pulled off an interference fit | Bore "looks round" | Ring is out-of-round enough to eat residual C3 clearance. Runs hot. |
| Cage nicked by a jaw or a punch | Cage looks intact from the side | Roller/ball control is gone. Skew, smear, then a locked bearing. |
| Set-screw insert reused after pull | Collar still tightens | Bore is bruised; shaft is burred. Insert walks, frets, and cooks. |
Brinell marks from a puller are not the same as false brinelling (fretting from vibration at standstill). False brinelling is a wear scar with red-brown debris. True brinelling is a plastic indent with raised edges. Both kill the raceway. Neither is something you "run in."
Why "it spins fine on the bench" is a useless test
Hand-spinning a 6208 with no radial load tells you the cage is not in pieces and the grease has not turned to cement. That is the entire list.
A brinelled raceway has indentations on the order of the elastic deflection the balls already make under catalog load. With no load, the balls roll over those dents the same way a car tire rolls over a pothole at walking speed — you barely feel it. Put belt tension, rotor weight, or gear reaction on that bearing and the balls fall into the pits. You get a once-per-rev knock that a vibration pen will read as a bearing defect because it is a bearing defect.
The other things a bench spin misses:
- Residual clearance. An inner ring that came off a k6 motor shaft has been expanded and then yanked. It does not spring back to the roundness it left the factory with. C3 was there to survive the mount and the heat, not a second plastic deformation.
- Grease condition. The pull heated the ring. The grease next to the race is already sheared and darkened. A 2RS bearing is filled for life. You cannot flush it.
- Seal interference. A 2RS1 lip that was stretched by a jaw may still wipe the inner ring on the bench. At speed it flutters. That is a leak, not a seal.
- Geometry against the shoulder. A mushroomed inner-ring face will not sit square. The bearing runs at a small angle. Deep-groove balls hate that. A spherical insert can take a little static misalignment; a 6208 in a motor end bell cannot.
If you needed a puller to get it off, you needed a new bearing to go back on. The puller did its job. The old bearing is evidence, not inventory.
A new bearing is not enough if the seat is already damaged
This is the other half of the reuse argument, and it is the one that turns a $80 bearing into a second unplanned outage.
Shaft journal (inner-ring seat)
Fretting rust — the cocoa powder everyone calls "the shaft is rusting" — means the inner ring was creeping. The journal is now undersize, full of pits, or both. Press a new 6208-2RS1/C3 onto that and you have built a slip fit on a rotating inner ring. The new bearing will walk, fret, and cook. Same PN, same failure, shorter life.
Witness lines, a step where the inner ring sat, or a journal that mikes small of the k5/k6 you needed are not "close enough." Options, in order of how much shaft you still have:
- Measure. If the journal is still in the manufacturer's tolerance for that fit, clean it, stone the burrs, and mount the new bearing correctly.
- If it is worn under size but the seal track is the only groove, that is a Speedi-Sleeve job for the seal — and a different conversation for the bearing seat (weld/turn, metal spray, or a proper repair sleeve). Do not pretend a 0.25 mm Speedi-Sleeve is a new k6 journal.
- If the shoulder is rolled or the fillet is beat up from a punch, machine it. A new bearing against a damaged shoulder is already misaligned.
Housing bore (outer-ring seat)
A bell-mouthed motor end bell, a spun outer ring, or a housing that was pried on with a screwdriver will not hold a new 6308. The outer ring will spin, fret, and walk into the cover. Loctite in a housing that has already spun is a temporary assembly aid, not a restored H7/J7. Measure the bore. If it is out, bore and sleeve the housing or replace the end bell.
Pillow-block housings with a spherical-OD insert are more forgiving of a little angular error. They are not forgiving of an egg-shaped bore or a cracked base. Reusing the insert on a damaged housing repeats the failure the insert just came out of.
Seal track vs bearing seat — do not mix them up
The groove you see just outboard of the bearing is almost always the radial-shaft-seal track, not the bearing journal. A new 2RS bearing on a grooved seal journal will leak past the new lip on the first shift. That leak is what people then blame on the bearing.
An SKF Speedi-Sleeve presses over that worn counterface and restores original diameter so you keep the standard seal size. Measure the shaft at the seal position (not down in the groove), pick the sleeve from the chart, and fit it before the new bearing and seal go on. Example: a 40 mm motor shaft at the seal is in Speedi-Sleeve territory around SKF 99157; a 2.125 in seal journal is SKF 99212. Use the chart — do not guess from the bearing bore.
Speedi-Sleeves are a seal-surface repair. They are not a substitute for a worn inner-ring seat. If the bearing journal is undersize, say so on the work order and repair that seat. Putting a new 6208 and a Speedi-Sleeve on a fretted journal is how you get a dry seal and a walking inner ring at the same time.
Unused spare vs pulled unit
| What you have | Reuse? | Why |
|---|---|---|
| New bearing, original SKF/NSK/Timken box, unopened, stored dry and flat | Yes | Factory geometry, factory fill, factory seals. |
| New bearing, box opened, still in the wrap, never mounted | Usually yes | Inspect seals and the bore/OD for corrosion or handling brinell. If it was dropped, treat it as damaged. |
| New bearing that was pressed on, then pulled because the fit was wrong or the job got aborted | No | It has already taken a mount load and an extraction load. That is a pulled bearing. |
| Bearing pulled during this repair, "spins fine" | No | See everything above. |
| Bearing pulled last year, oiled, sitting on a shelf as a "spare" | No | Unknown damage, unknown grease, unknown storage (false brinell from vibration on a shelf next to a punch press is real). |
| Spherical insert pulled off a commercial shaft, set screws still in it | No | Bore is bruised; shaft is burred. New insert, dress the shaft, new set-screw locations or a shaft repair. |
| Open bearing from a gearbox, washed and "inspected" | No, unless you are a rebuild shop with raceway inspection, new cage/rolling elements, and a controlled re-lube. Field wash-and-pack is not that. | A 22218 that came out of a gearbox has running damage even if it is not spalled yet. |
The cost of the bearing is not the cost of the outage. A 6208-2RS1/C3 is cheap next to a second motor pull, a lost shift, and a shaft that now needs a sleeve because you ran the pulled unit for three more weeks.
Buyers: if the PO says "repair kit" and someone is planning to reuse the pulled 6308 to save a line item, kill that line item. Spec the new bearing, and spec the Speedi-Sleeve if the seal track is worn. That is still cheaper than a repeat.
Pull it like you intend to replace it
You are not trying to save the bearing. You are trying to save the shaft, the housing, and the next bearing.
- Pull the tight ring. Inner ring on a shaft → grip the inner ring (or use a bearing separator behind it). Outer ring in a housing → internal extractor on the outer-ring bore. Do not drag the load through the rolling elements if you can avoid it.
- Keep the spindle on center. A cocked two-jaw is how you smear a race and raise a burr on the journal.
- Heat the inner ring, not the shaft, if heat is required. Even, controlled, and only enough to break the interference.
- Stop as soon as something yields that is not the fit — a cracked housing, a rolled shoulder, a jaw that slipped onto the seal.
- The Kukko puller guide covers external jaws, internal extractors, and hydraulic pullers. We stock them on apollousa.com/collections/all-kukko-products. A decent puller is how the replacement bearing gets a round journal.
A puller that saves the shaft is doing more for bearing life than any "inspection" of the unit you just yanked off.
Field sequence that actually sticks
- Pull the old bearing with a centered puller. Do not reuse it.
- Measure the shaft journal and the housing bore. Write the numbers down. Compare them to the fit the new bearing needs — not to "it used to go on."
- Inspect the seal track. If it is grooved, pick an SKF Speedi-Sleeve by shaft diameter and fit it before the new seal.
- Stone burrs. Do not file a new taper into the journal.
- Mount the new bearing the way the fit requires: induction heat or a proper press on an interference inner ring, never a hammer on the outer ring of a 6208 you are putting on a motor shaft.
- If this was a belt-drive 6208 that died young, the replacement conversation is a 6308 or a spherical insert — same rule, new part, not the one in the drip pan.
FAQ
Can't I just swap a pulled bearing onto a light-duty spare machine?
No. Light duty still loads the raceway. Brinell plus a stretched 2RS lip is how a "spare blower" becomes the next call. Scrap it.
What if I only pulled it a few thousandths and it never ran after the first mount?
It still took a full interference expansion and an extraction. Rings do not have a free "almost mounted" cycle. New bearing.
Do sealed bearings survive a pull better than open ones?
No. 2RS / 2RS1 lips are the first thing a jaw wrecks, and you cannot re-lube or re-pack a sealed bearing after the race is brinelled. Open bearings at least show you the damage. They still should not go back in.
Is a Speedi-Sleeve enough if the inner ring was creeping?
No. Speedi-Sleeve repairs the seal diameter. Creep damage is on the bearing journal. Repair or replace that seat, then sleeve the seal track if it is worn too.
We have a rebuild shop. When is reuse acceptable?
When you are actually rebuilding: raceway inspection under magnification, measured residual clearance, new rolling elements and cage if needed, controlled clean-pack, new seals. That is not a pull-clean-grease-and-go on a 6208 in a mechanic's truck. If you do not have that process, you do not have a reused bearing. You have a spare failure.
Need the new bearing — and a shaft that will hold it?
Send the PN on the shield, a mike reading on the journal, and a photo of the seal track. We will tell you whether the next part is another 6208-2RS1/C3, a 6308, or a spherical insert, and whether that shaft wants an SKF Speedi-Sleeve before the seal goes on.
Apollo Industries (ApolloUSA.com) — Oxford, MI. Bearings, SKF Speedi-Sleeves, and Kukko pullers.
Call 1 (248) 429-9073.