Bearing engineering
6 Checks I Run on Every SKF Bearings Delivery (Before I Approve It)
-
Step 1: Check the Logo and the Fine Print
-
Step 2: Confirm the Exact Designation, Not a 'Similar Equivalent'
-
Step 3: Verify the Actual Dimensions and Rotational Smoothness
-
Step 4: Inspect Storage and Packaging (the Step Almost Everyone Skips)
-
Step 5: If It's a Ball Screw Actuator, Ask Speed Questions Differently
-
Step 6: Demand Traceability and Treat Small Orders Like Real Orders
-
Common Mistakes to Avoid
If you've ever had a bearing fail in the first 200 hours, you know the frustration. It's usually not the bearing itself. It's something invisible: wrong internal clearance, mismatched seal, a repackaged part from another batch, or a speed spec that didn't match the duty cycle.
I'm a quality and compliance manager at a small OEM that uses industrial bearings and linear components in every product line. I review roughly 120 incoming deliveries per month before they go to production. In 2024, I rejected 4.6% of first deliveries. Not because the parts were all bad—most were usable—but because the paperwork, markings, or storage condition didn't match what we approved.
This is a practical checklist. Use it when you're accepting an order of SKF bearings, precision ball bearings, or a ball screw actuator. I broke it into six steps.
Step 1: Check the Logo and the Fine Print
A printed SKF bearings logo on a box means less than the wording on the bearing itself. I've seen boxes with the logo and a completely different part number inside. The logo is the marquee; the part number, suffix, and date code are the contract.
What I check:
- The bearing end face marking. It should have the SKF name or trademark, the designation, and a batch/code mark. The die stamp should be crisp, not blurry.
- The box label. It should include the full designation (for example, 6205-2RS1/C3), the quantity, and a production traceability code.
- The packaging. Genuine SKF bearings have a distinct yellow/blue box style and protective paper. If a vendor repackages into plastic bags with a handwritten label, that's not automatically a counterfeit, but it is a reason to demand batch documentation.
One thing people miss: the logo on the website or quote is not the logo on the product. Ask for a photo of the actual bearing face before the order ships. If it's too hard for the supplier to provide, that's a red flag.
Step 2: Confirm the Exact Designation, Not a 'Similar Equivalent'
A 6205 and a 6205-2RS1/C3 are not the same bearing. The first is a deep groove ball bearing that's suitable for certain applications; the second is a different internal configuration with two rubber seals and a higher-than-normal C3 clearance. In my opinion, the most common source of premature failure is substituting one for the other without a spec review.
Your checklist:
- Match the full suffix on the bearing face to the purchase order. 2RS means two moulded rubber seals; Z means metal shields; C3 means greater radial internal clearance for higher temperatures or interference fits. These details matter.
- If you're using precision ball bearings for a spindle or a servo motor, confirm the tolerance class (P6, P5, P4, and so on). ISO 492 covers radial bearing tolerances, and P5/P4 are much tighter than standard. But tightness alone isn't the goal—the matching housing, shaft fit, and preload also have to be right.
- Check the type of actuator. A ball screw actuator can use precision ball bearings internally, but its spec sheet is different. You're looking at lead, pitch, stroke, dynamic load, and backlash. Treat it as a separate component, not an accessory to a bearing order.
To be fair, a 'similar equivalent' can sometimes work. But if the project depends on it, the person who accepts that substitution should own the failure. That's not engineering; that's gambling.
Step 3: Verify the Actual Dimensions and Rotational Smoothness
This is the step where most buyers focus on price and completely miss runout, clearance, and noise. You don't need a metrology lab for an initial incoming inspection. But you do need a micrometer, a dial indicator, and a feel for what's normal.
Basic checks:
- Measure bore and outside diameter at a few positions. The bearing should sit roughly in the middle of the standard tolerance range. If the measurements are near the limit, it's still acceptable, but it can complicate an interference-fit assembly.
- Spin the bearing by hand. A sealed precision ball bearing should rotate smoothly and quietly, with no catching. If you feel a slight harmonic grinding on a brand-new bearing, set it aside and investigate rather than assuming it will wear in.
- For angular contact or cylindrical roller bearings, check the height or width more carefully. Mismatched stack heights are a common hidden problem in multi-bearing arrangements.
One nuance: a bearing with out-of-spec dimensions can still pass an eyeball test. That's why I reject based on measurement, not appearance.
Step 4: Inspect Storage and Packaging (the Step Almost Everyone Skips)
If you've ever opened a package and found a perfectly good bearing with a thin film of condensation on it, you've seen the result of storage problems.
Bearings are not potatoes. They come with a specified preservation coating, and the packaging is engineered to keep moisture and contamination out.
What I look for:
- Is the bearing still in its factory-designed wrapper or sleeve? If it arrived loose in the box, it may have been removed from the original packaging for inspection or counterfeiting.
- Are there signs of rust or fingerprint corrosion? Fingerprint salts can etch a precision surface within days, especially in high-humidity climates.
- Is the box damaged or punctured? A torn package doesn't necessarily mean a bad bearing, but for open bearings, the risk is much higher.
Honestly, I have sent back more boxes due to packaging and preservation failures than due to actual bearing defects. That's because a good bearing, stored badly, is no longer a good bearing.
Step 5: If It's a Ball Screw Actuator, Ask Speed Questions Differently
One of the most common questions I get is: "How fast can a linear actuator move?" The answer is usually "it depends," which annoys buyers. Let me give you a more useful way to think about it.
For a ball screw actuator, linear speed comes from a simple relationship:
Linear speed = screw lead × rotational speed
If a ballscrew has a 20 mm lead and the servo motor runs at 1500 rpm, the theoretical linear speed is 20 × 1500 = 30,000 mm/min, or 500 mm/s. Increase the lead to 40 mm and you double the speed, assuming the screw's critical speed and the motor's torque permit it.
But speed is not the real spec. You also need:
- Duty cycle: how many cycles per minute? The motor and screw can both overheat if you push a fast cycle without pause.
- Thrust: a small lead gives higher thrust per motor revolution. If everyone asks for speed and no one asks for thrust, you'll get a fast actuator that stalls under load.
- Critical speed: a long, unsupported ballscrew can whip at high RPM. You can't just dial the speed up beyond a certain limit.
- Lubrication and seals: faster actuators need proper grease, wipers, and sealing. It's not a one-time build-and-forget component.
The better question is: "At what specific speed, load, and duty cycle do you need this actuator to run?" The answer to "how fast" only makes sense inside that envelope. A decent supplier will ask you these questions. If they quote you a speed before asking about load, be careful.
For a quick SKF company overview: industrial bearings and linear systems are related but different. SKF makes both, but your purchasing team should not handle a ball screw actuator like a box of bearings. The actuator has its own test report, preload, and stroke. Ask for those documents.
Step 6: Demand Traceability and Treat Small Orders Like Real Orders
I'll get on a small soapbox here.
When I was starting out, I tried to order a trial pack of precision ball bearings for a prototype. Some vendors didn't even reply after hearing the quantity. The suppliers who treated my $200 order seriously are the ones I still use for $20,000 orders. That's not sentimentality. It's a practical test: if a supplier can't handle a small, low-complexity order properly, they won't magically improve when the order is bigger.
For every SKF bearings order—large or small—you should receive:
- A packing list with the exact designation, quantity, and batch number.
- A certificate of inspection or a declaration of conformity for traceable industrial bearings.
- For precision ball bearings and ballscrew actuators, a dimensional or runout data sheet where available.
If the supplier says that's only available for large orders, that's a conflict of interest. A 20-piece order deserves the same quality data as a 2,000-piece order. It costs the supplier almost nothing to copy-paste the inspection report. The only reason to withhold it is that they don't have it.
Small doesn't mean unimportant. It means the buyer is still paying attention. In my experience, that's exactly the kind of customer who will catch a problem later.
Common Mistakes to Avoid
- Trusting the logo without verifying the code. The SKF bearings logo is a brand mark, not a specification. The part number suffix is the spec.
- Ignoring internal clearance. C3 is not a universal upgrade. It creates more internal play and can ruin a high-precision spindle assembly.
- Choosing speed over thrust. A ballscrew actuator that's fast but can't hold the load is a very expensive lemon.
- Buying bearings based on a photo. If you're sourcing online, ask for a photo of the actual item and its packaging. A stock photo means nothing.
- Forgetting the storage trail. A bearing can be genuine and factory-new, then sit in a damp warehouse for a year and turn into a problem.
Bottom line: a solid acceptance process takes about 15 minutes per line item. On a small order, that might look like overkill. But in my opinion, that's the wrong comparison. The question is not whether the bearing costs $30 or $3,000. The question is what the bearing is attached to when it lets go.