Here's the question I get more often than you'd think: What happens if a ball bearing goes out?
It sounds simple. It isn't. I'm a quality compliance manager at Thomson Linear. I review a couple hundred unique product configurations each year—linear actuators, ball screws, linear bearings, motion control components—and I reject about 12 percent of first-article runs because the delivered spec doesn't match the intended spec. Maybe 14 percent, I'd have to check our log. The point is: in motion control, a bearing failure is rarely the root cause. It's the first visible casualty.
The Surface Problem: What a Failing Bearing Does
At the system level, a ball bearing that has 'gone out' shows up in ways that feel obvious in hindsight:
- Noise. A low growl or high-frequency chatter from the motor end or the carriage.
- Heat. The housing gets hot to the touch before any thermal alarm trips.
- Lost motion. The actuator doesn't repeat to the same position; homing offsets start to drift.
- Backlash. Reversing load produces a visible click or play.
- Current rise. On a small stepper motor, drive current climbs because friction is doing some of the work the magnetic field should be doing.
That's the surface problem. If you run a Thomson linear actuator, the first sign may be nothing more than a rough feel during a manual jog. On a machine with an AC motor, the first sign might be a slightly longer ramp time. The bearing hasn't failed yet—but the system is already losing performance.
The Deeper Problem: Why Bearing Life Is Shorter Than the Catalog Says
Most engineers calculate bearing life from the L10 rating, apply a load factor, and get an answer that looks like years. Then the bearing fails in months. I made the classic spec error in my first year: I assumed the catalog L10 number meant no bearing would fail before that number. Cost me a $600 prototype redo and a week of debugging.
The truth is that rating life assumes correct mounting, clean conditions, and adequate lubrication. Real installations rarely give a bearing a clean life. Here are the causes I see most often.
1. The load on the bearing is not the load you calculated
Linear motion systems produce dynamic loads that are easy to miss at the drawing stage. Acceleration torque, cable drag, vacuum forces, and clamping loads all add to the basic payload weight. A cantilevered load can double the equivalent load on a bearing. Remember: for a ball bearing, life is proportional to load cubed. A 10 percent load increase cuts L10 life by roughly 25 percent if you follow the standard calculation.
2. Misalignment sends the bearing into an impossible geometry
When a ball screw and motor shaft are not aligned, the angular misalignment is absorbed by the first bearing in the load path. I still kick myself for not checking screw-to-motor alignment on a prototype lead-screw assembly. If I'd measured that, we'd have avoided three bearing failures in a week. The machine thought the bearing was defective. The bearing was just the pivot point for an alignment error.
3. Contamination is the quiet killer
In a dusty or wet environment, a small particle can get between the balls and the raceway, create a dent, and then eat away at the surfaces around it. In our Q1 2024 review of returned units, about 70 percent of the wear patterns indicated contamination, not classic fatigue. That is why a sealed or shielded bearing is not a luxury decision; it is a life calculation input.
4. Lubrication breaks down before the bearing does
A grease with the wrong base oil viscosity, or a system that never gets relubricated, can start the spalling process long before the steel reaches its fatigue limit. So glad I insisted on sealed linear bearings in a dusty application. Almost went with open ones to save cost; I'd have spent the savings on returns.
5. Preload assumptions don't survive contact
Bearings in a linear actuator often use preload to remove play. Too much preload raises temperature and shortens grease life; too little preload allows ball skidding, especially on a small stepper motor shaft near zero speed. The right preload has to match the actual bearing temperature, not just the initial stiffness.
I'm not 100% sure every failure fits these five categories, but in my experience, these are the usual suspects. The common thread is that the bearing is not the first failure; it's the last symptom.
What a Bearing Failure Really Costs
At the component level, a ball bearing might cost $35. At the system level, it can cost an entire shift. A customer of ours had a food packaging line stop for eleven hours because a small stepper motor bearing seized. The replacement motor cost about $120. The downtime cost more than $9,000 in labor, missed output, and the cost of washing down the line again. Don't hold me to the exact figures—the point is the ratio. The bearing was a rounding error in the bill.
In 2022, a bearing failure on one of our in-house test fixtures ruined a spindle and a precision housing. The redo cost $18,000 and delayed a validation program by three weeks. (Should mention: the root cause was a lubrication mismatch, not the bearing manufacturer.) The bearing itself was $37. The cost wasn't the part. It was the trust lost in the machine.
At Thomson Linear, motion systems are not just assembled; they are specified around the bearing's real load and life. That is why our design reviews spend more time on bearing selection than on the brand name on the actuator. A poorly sized bearing, delivered on time, still fails early.
What a Quality-First Approach Looks Like
I'm not going to give you a thirty-step checklist. Start with three questions before you configure your next linear motion system:
- What is the real load, including acceleration and shock loads? Not just the payload mass. If the moving carriage is cantilevered, calculate the moment it puts into the bearing.
- What is the real environment? Dust, washdown, temperature, humidity. These should change the bearing seal and grease choice, not just the IP rating of the motor.
- What is the motor doing to the bearing? An AC motor with an inline gearbox can create overhung loads if the coupling is not aligned. A small stepper motor with a lead screw on its shaft can turn a radial motor bearing into a thrust bearing—a job it was never designed for.
When I evaluate a Thomson Linear motion system—whether it's a ball-screw stage, a belt-driven unit, or a complete actuator—I look for the same thing I'd look for in any bearing application: a clear load path, a calculated life that matches the duty cycle, and a maintenance plan that matches the environment.
The basic rating life standard (ISO 281) and the linear motion rolling bearing life standard (ISO 14728-1) both assume clean, correctly aligned, properly lubricated conditions. Neither one promises that a bearing will live forever.
The Bottom Line
If someone calls you and says a ball bearing went out, ask them what was going on at the moment it happened. That answer will teach you more than the bearing failure itself. Was the motor running hot? Was there a squeal on startup? Was there a recent maintenance event? Those clues matter.
Then ask one more question: what bearing load does the machine actually produce? If nobody knows, the next bearing is already running on borrowed time.
A bearing that fails early is a symptom. Treat it as evidence, not as the culprit. Do that, and you'll stop fixing bearings and start fixing machines.