I'm a plant maintenance engineer. I've been handling SKF bearing orders for six years, and I've personally made—and documented—14 significant mistakes that cost us about $27,000 in wasted budget. Now I maintain our team's bearing selection checklist. This article is that checklist, but not as a one-size-fits-all guide. Because there isn't one.

The question has no single answer

Should you buy SKF bearings? Which ones? The answer depends on what you're doing. Before I recommend anything, figure out which bucket you're in:

  • Replacing a failed bearing in equipment that has been running—your goal is to avoid changing the problem.
  • Designing a new machine or upgrade—your goal is to calculate loads and select a bearing with enough capacity.
  • Assembling a linear actuator kit or linear motion axis—your goal is to integrate a motor, screw, and bearing system so they don't fight each other.

These are different jobs. The bearing that is perfect for one could be the wrong choice for another.

Scenario 1: You're replacing a failed bearing

In my first year (2017), I made the classic same-size-must-be-the-same-part mistake. A ball thrust bearing on a conveyor failed. I measured the bore, grabbed a deep groove bearing with the same bore from the parts room, and installed it. It lasted two weeks. The problem was that the application was pure axial load—exactly what a ball thrust bearing is for. The deep groove bearing couldn't handle the continuous thrust, so it ran hot and then broke.

My only excuse is that I was in a hurry. The real lesson: read the old bearing's markings before you even think about price. The letters after the SKF bearing number—C3, P5, E, the contact angle suffix—tell you about clearance, accuracy, and internal design. SKF Explorer ball bearings often have a specific suffix or designation. If you don't recognize it, look it up. Don't assume.

And before you ask: knowing how ball bearing are made helps here. It sounds like trivia, but when you understand that the steel rings are turned, heat-treated, ground, and honed, you start to appreciate why the raceway finish matters. A bearing with the same size but a lower finish grade can lose oil film faster, which means shorter life. That's why replacing like-for-like matters.

If the machine ran fine for years with the original bearing, don't improve it. I've made that mistake too: I upgraded to a beefier option once, thinking more load capacity was better. Instead, the different internal geometry shifted the load path, and the adjacent component failed. Looking back, I should have asked why the old bearing failed before changing anything. The machine was telling us something—like a lubrication problem—and I just covered the symptom.

Scenario 2: You're designing something new

This is where the numbers matter. If you're specifying a bearing for a new machine, you need to calculate the equivalent load. For angular contact ball bearings, the SKF catalog gives you three factors: X, Y, and e. They're not arbitrary. They decide how radial load and axial load combine into a single equivalent load for life calculations.

When you look up SKF angular contact ball bearings factors X Y and e in the catalog, you're seeing a decision rule. The standard formula per ISO 281 is P = X Fr + Y Fa, but you only apply the Y term when the axial load is significant relative to the radial load. That's what e is for. If Fa/Fr ≤ e, the bearing behaves differently than when Fa/Fr > e.

I used to ignore e because I found it confusing. That's how I ended up with a $3,200 order of SKF angular contact ball bearings that were undersized for a spindle. We assembled everything, it ran hot, and we had to disassemble the unit two days later. The total cost was around $8,000 after labor and downtime—over twice the bearing purchase price.

The exact X and Y values depend on the bearing series and contact angle, so don't take someone else's spreadsheet numbers as gospel. Use SKF's engineering calculator or a genuine implementation of ISO 281. But calculate. Don't guess.

Also, don't choose bearings on price per unit. The cheapest bearing with enough static load might be fine in theory, but in practice, a slightly more expensive SKF Explorer ball bearing with a higher dynamic rating can be cheaper over the life of the machine if the application has variable loads. This is the total cost thinking I wish I'd had earlier. Price is the first number, not the last number.

Scenario 3: You're building a linear actuator kit

Linear motion is a different animal. A linear actuator kit usually includes a motor, a lead screw or ball screw, guide rails, and some kind of thrust bearing or angular contact bearing arrangement. The screw converts rotation into linear movement, so the bearing has to handle the thrust of the screw, while the guide rails handle the side loads. If you treat the whole thing as just some bearings, you'll miss where the loads are going.

I once bought a linear actuator kit from a supplier because the price was unbeatable. The screw was supported by a cheap thrust bearing that was visibly smaller than the end of the screw. It failed within a week. The supplier said the bearing was the one that came with it. In the end, I replaced it with a proper SKF ball thrust bearing, and the actuator has been running for two years. The extra $30 cost was nothing compared to the $450 in freight and lost production I paid because I tried to save at purchase.

The lesson: when you assemble a linear motion system, check the bearing arrangement before you power it up. Ask yourself: Where is the axial thrust absorbed? Are the thrust and radial loads handled separately? What clearance or preload is in the bearings? For a high-accuracy axis, angular contact bearings set in pairs—back-to-back or face-to-face—are often the right answer, not a single deep groove bearing.

How to tell which scenario you're in

Ok, here's the practical part. If you're holding a failed bearing in your hand and there's a machine waiting to run, you're in Scenario 1. Stop reading the internet and go find the part number. The best answer is the exact original or the true equivalent from SKF's cross-reference.

If you're starting with a blank machine design and a load requirement, you're in Scenario 2. You need a calculation, not a recommendation from a forum. Get the radial load Fr and axial load Fa, and use the X, Y, e factors properly.

If you're putting together a linear actuator kit, or designing a linear stage, you're in Scenario 3. The motor and the screw are part of the bearing decision. The cheapest kit isn't cheaper if you have to rip it apart before the first PM cycle.

Final lesson: total cost, not price

I can only speak to my own world. My experience is based on about 300 replacement orders, 25 new designs, and a dozen linear actuator projects. If you're dealing with ultra high speed spindles, severe vibration, or temperatures above 100°C, the calculus might be different. But the one thing I've learned is universal: bearing purchasing decisions are total ownership cost decisions.

Looking back, I should have built a simple pre-order checklist sooner. At the time, I thought it was overkill. It wasn't. Now our team has one, and we've caught 47 potential errors in the past 18 months using it. None of those were clever engineering wins. They were just us checking the part number, reading the failure, and calculating the load before ordering.

The right SKF bearing for your situation depends on whether you're replacing, designing, or assembling. But in every case, the right process is the one that keeps you from repeating my mistakes.