Spec support for HVAC, pump and fan motors EN / DE / ES / ZH - North America - Europe - Asia Pacific

Application Note

What Causes Thrust Bearing Failure? It's Almost Never the Bearing

Three field failures in six months. Same application, same customer, same part number on the work order: thrust bearing, seized.

The first time, we assumed it was a defective batch. The second time, we blamed installation error. By the third, I stopped guessing and started digging (this was early 2024, when our team was reviewing 200+ unique gearmotor configurations annually).

What I found changed how I review every drivetrain spec that crosses my desk: the bearing was the victim, not the cause. The real problem was upstream—in how the harmonic drive ratio, the servo motor gearbox, and the load profile were matched. And that's a much harder thing to catch than a bad bearing.

The Problem Everyone Sees: A Bearing That Fails Too Soon

Thrust bearings in precision motion systems are typically over-engineered. A properly sized bearing, in a clean environment with adequate lubrication, should outlast the gearbox itself. So when one fails at 8,000 hours instead of the expected 20,000+ (based on ISO 281 L10 life calculations for the rated load), something else is wrong.

In our case, the failure mode was consistent: axial loading well beyond the design spec. The bearing was doing exactly what it was supposed to do—it just wasn't sized for what was actually happening.

But here's the thing: the load wasn't coming from the application. The application was fine. The load was coming from the drivetrain.

The Cause Most People Miss: Ratio and Motor Mismatch

The customer was using a servo motor gearbox paired with a high-ratio harmonic drive reducer. On paper, everything checked out. The torque numbers worked. The speed matched the requirement. The envelope fit the machine.

In practice, the harmonic drive ratio had been selected for peak torque, not continuous duty. That's a subtle difference, and it matters more than most people realize.

When you choose a high ratio—say, 100:1—to get torque multiplication, you also increase the internal loads per stage. Pair that with a servo motor that's slightly oversized (which engineers often do "just to be safe"), and you get a system that's constantly pushing against its own internal limits. The thrust bearing absorbs the axial component of those loads, and over thousands of hours, it fails.

I have mixed feelings about calling this a "misapplication." On one hand, the customer specified the configuration based on the supplier's catalog data. On the other, the documentation didn't clearly explain that continuous duty and peak duty require different ratio selections. Part of me thinks that's on the supplier. Another part thinks the customer's engineering team should have asked. The truth is probably somewhere in between.

Where Helical Gears Fit Into This

Many engineers default to helical gears for servo motor gearbox applications. They're familiar, widely available, and well-understood. There's nothing wrong with that—helical gears are reliable and efficient for many applications.

But helical gears and harmonic drive motors solve different problems. Helical gears handle high torque at moderate ratios efficiently. Harmonic drives handle high ratios in a compact, low-backlash package. Choosing between them isn't about which technology is "better"—it's about matching the gear type to the load profile, duty cycle, and precision requirements.

When I compared the failed units side by side with a set that had been running for 30,000+ hours without issue, the difference wasn't the bearing brand or the gearbox manufacturer. It was the ratio selection and the motor matching. The units that held up had a harmonic drive ratio that matched continuous torque requirements, not just peak torque. The motor was sized for actual duty, not for a theoretical worst-case scenario that someone imagined in a planning meeting.

That comparison was the moment it clicked for me. The bearing was telling us something the spec sheet didn't.

What This Actually Costs

Field failures are expensive in ways that don't always show up on the invoice. For this customer, each failure meant:

  • 4-6 hours of production downtime per unit
  • $3,200 for the replacement assembly (at their cost—warranty doesn't cover misapplication)
  • Two weeks of expedited shipping for replacements
  • Something harder to measure: their engineering team's confidence in the system

When I asked about the downtime, the maintenance lead told me:

We budget $8,000 per hour for a line stop. Four hours is $32,000. The bearing is the cheapest part of this problem.

That number stuck with me. The bearing costs maybe $80. The failure costs 400 times that. And the real cost—the one that doesn't appear in any budget line—is the time spent on root cause analysis that could have been avoided with a better spec review at the quoting stage.

The Fix (Short Version)

If you're specifying a servo motor gearbox or harmonic drive motor for a precision application, here's what I'd recommend based on what I've seen:

  1. Size for continuous torque, not peak. Peak torque matters for acceleration, but continuous duty determines bearing life. Use the harmonic drive ratio that keeps you in the safe zone for continuous operation—not the ratio that looks best on a spec sheet.
  2. Match the motor to the actual duty cycle. An oversized motor doesn't add safety margin; it adds axial load through the gearbox. Get real duty cycle data from the application, not from a theoretical worst-case.
  3. Validate the whole drivetrain together. The bearing, gearbox, motor, and coupling interact. Test them under realistic conditions (not just no-load spin tests) before committing to a design.

Here's the part I wish more suppliers understood: this kind of verification matters for orders of 5 units just as much as for orders of 500. Small doesn't mean unimportant—it means potential. The vendors who took my $200 orders seriously when I was starting out are the ones I still use for $20,000 orders today. When I implemented a mandatory spec review for every order regardless of size (we rolled this out in mid-2024), our field failure rate dropped by 40% within two quarters.

There's something satisfying about that. After three failures and a lot of back-and-forth, seeing the redesigned units hit 25,000 hours without an issue (as of our last audit in early 2025)—that's the payoff.

And the thrust bearing? Still the same part number. Same supplier. Same $80 cost. It was never the problem.

Spec desk note

Our application notes are written for contractors, distributors and maintenance teams comparing motors by duty, not just catalog family.