Further Reading

For a broader understanding of motion transfer system design and the role of gears and couplings in overall system behavior, the following articles provide additional engineering insights:

• Gears & Couplings: An Engineering Guide to Precision Motion Transfer

• How to Choose the Right Coupling Without Guessing

• Common Coupling Failures and How to Prevent Them

• Gear Material Selection Guide: Strength, Wear, Corrosion & Environment – How to Choose Correctly

• Backlash Is Not a Number: Understanding What Really Determines Accuracy, Stability, and System Life

• Spur, Helical and Worm Gears – Engineering Differences and How to Choose the Right One

• Backlash in Gears – From Geometry to System Behavior: Understanding what really happens between gear teeth
• Small Spur Gears: Why Miniaturization Creates Hidden Mechanical Problems
• Gear Hardening Explained – Why Case Hardened Gears Dominate Heavy Duty Power Transmission
• Why a Million-Dollar Medical System Still Uses Rack & Pinion
• Why Most Engineers Use Bevel Gears for Only 10% of What They Can Actually Do
• Why the Number of Starts in a Worm Gear Matters More Than the Gear Ratio

Every time a gearbox fails, the same process begins.

Replace the gearbox.

If the problem returns, replace it with another brand.

If that doesn’t solve it either, people start questioning the manufacturing quality.

But there’s one problem.

In most cases…

The gearbox was never the real problem.

It was simply the first component that couldn’t survive the operating conditions created by the system itself.

In other words:

It wasn’t the cause of the failure.

It was the first victim.

---

The Mistake Begins Long Before You Choose a Gearbox

Ask ten engineers how they select a gearbox.

Most will answer with something like:

• Gear ratio
• Torque
• Speed

These are certainly important parameters.

But they are nowhere near enough.

Because a gearbox is not just a device that reduces speed.

A gearbox is a component that absorbs loads.

If you don’t calculate the actual loads, you haven’t really selected a gearbox.

You’ve simply selected a number from a catalog.

---

“But It’s Rated for 40Nm…”

That may be one of the most expensive sentences in mechanical engineering.

The catalog says:

Maximum Output Torque: 40Nm

And that’s where the mistake begins.

Because the real question isn’t:

How much torque can the gearbox transmit?

The real question is:

What torque will the gearbox actually experience throughout its service life?

Those are two completely different things.

---

The Gearbox Doesn’t Know What’s Written in the Catalog

Let’s say your motor normally operates at 8Nm.

Great.

But during startup…

Emergency braking…

Direction reversal…

Or aggressive servo acceleration…

The torque may suddenly jump to 50Nm.

Sometimes 80Nm.

Sometimes even higher.

“For only 20 milliseconds.”

The gearbox doesn’t know it was “just for a moment.”

It simply experienced a load it was never designed to withstand.

And if this happens thousands of times every day…

The first microscopic crack forms long before the first broken tooth appears.

---

When the Gears Look Perfect – But the Gearbox Still Fails

It sounds strange.

But it happens all the time.

The gearbox is opened.

The gear teeth look almost new.

Yet the bearings are severely damaged.

Why?

Because nobody calculated the radial load.

A timing pulley.

A chain sprocket.

A large gear.

A long lever arm.

Each of these applies a side load to the output shaft.

In many applications, that radial load is actually greater than the load carried by the gear teeth themselves.

And yet…

Most people only look at torque.

---

The Load Many Designers Forget Completely

Axial load.

Ball screws.

Helical gears.

Linear actuators.

Push mechanisms.

All of them generate forces along the shaft axis.

If the bearings were never designed to carry those loads…

It doesn’t matter how strong the gear teeth are.

---

Then Comes the Load Nobody Calculates

Shock load.

It’s not simply another load.

It’s an impact.

A servo stops a heavy mechanism.

A robot changes direction.

A shaft suddenly locks.

The system hits a mechanical stop.

Instantaneous torque can rise three times.

Sometimes five times.

Sometimes even ten times.

The catalog wasn’t wrong.

It simply assumed conditions that no longer exist in the real machine.

---

Maybe the Coupling Is the Real Culprit

This is one of the most frustrating failure scenarios.

The gearbox is replaced.

Again.

And again.

Eventually, someone discovers that the real problem wasn’t the gearbox at all.

It was the coupling.

Misalignment.

Vibration.

Cyclic loading.

Shaft bending.

The bearings fail.

The gearbox gets blamed.

The coupling keeps working as if nothing happened.

---

Backlash Isn’t Always the Problem

Many projects spend thousands of dollars on a low-backlash gearbox.

Then…

The machine frame deflects.

The shaft twists.

The coupling flexes.

The bearings move.

The entire system deflects more than the gearbox backlash the engineer worked so hard to eliminate.

In many cases…

The money was invested in the wrong place.

---

So How Should a Gearbox Really Be Selected?

It’s not a shopping list.

It’s a risk assessment.

Before selecting a gearbox, you should know:

• What is the Peak Torque?
• What is the Continuous Torque?
• How many start-stop cycles occur every hour?
• Are there any Shock Loads?
• What radial loads act on the output shaft?
• What axial loads are present?
• Is Backdriving required, or should the gearbox be Self-Locking?
• What is the operating temperature?
• What service life is required?
• What positioning accuracy is needed?
• What is the maximum allowable Backlash?

If you don’t have answers to these questions…

It’s too early to select a gearbox.

---

The Bottom Line

Whenever a gearbox fails, everyone looks at the gearbox.

Experienced engineers look everywhere else first.

Because in most cases…

The gear teeth aren’t the problem.

Neither is the manufacturing quality.

And often…

Neither is the gearbox itself.

The real problem is that the system demanded the gearbox perform a job it was never designed to do.

So the next time you’re selecting a gearbox, don’t just ask:

“What gear ratio do I need?”

Ask a completely different question:

“What will this gearbox actually experience during the next ten years of operation?”

Because gearboxes are not selected by gear ratio alone.

They are selected by understanding reality.

Case Study

The Gearbox Wasn’t Too Small.

It Was Simply Carrying a Load That No One Calculated.

Application

A machine manufacturer developed an industrial indexing conveyor for transferring aluminum components between machining stations.

The system operated at 18 cycles per minute, with each cycle consisting of:

• Rapid acceleration
• Approximately 90° of rotation
• Precise positioning
• A dwell time of about one second
• Return movement

The drive system included:

• 0.75kW AC motor
• HPCE60-40 worm gearbox with a 40:1 reduction ratio
• Input speed: 1,500 RPM
• Output speed: approximately 37 RPM
• Average operating torque: approximately 28 Nm

According to the gearbox specifications, the application appeared to be well within the recommended operating limits. The HPCE60-40 is designed for reduction ratios up to 40:1, input speeds up to 1,500 RPM, and output torques of approximately 62 Nm, depending on operating conditions.

---

The Problem

After approximately eight months of operation, the following symptoms began to appear:

• Unusual mechanical noise
• Increased vibration
• Elevated temperature around the output shaft
• Reduced positioning accuracy of the indexing conveyor

A few weeks later, the system failed completely.

The initial diagnosis was straightforward:

“The gearbox has worn out prematurely.”

---

The Investigation

After disassembling the gearbox, the findings were unexpected.

The worm and worm wheel showed almost no visible wear.

The gear tooth surfaces remained in excellent condition.

However, the output shaft bearings exhibited severe fatigue damage, with radial play significantly exceeding the original specification.

This was a clear indication that the failure had not originated in the gear set.

---

What Really Happened?

A closer inspection of the drive assembly revealed a seemingly minor design detail.

The 210 mm timing pulley was mounted approximately 85 mm away from the gearbox output bearing.

As a result, belt tension generated a continuous radial load on the output shaft.

In addition, every indexing stop introduced a short-duration dynamic load caused by the inertia of the moving assembly.

The average transmitted torque remained around 28 Nm.

However, during acceleration and braking, peak torque values of approximately 65-70 Nm were recorded for several tens of milliseconds.

In other words:

According to the catalog…

The gearbox selection appeared correct.

From the bearings’ perspective…

They had been subjected to millions of load cycles that were never considered during the design phase.

---

The Solution

Instead of replacing the gearbox with a larger model, the engineering team made three mechanical modifications:

• Reduced the distance between the timing pulley and the gearbox housing from 85 mm to 25 mm.
• Added an external support bearing adjacent to the pulley.
• Replaced the coupling with a flexible coupling capable of compensating for minor shaft misalignment.

No modifications were made to the HPCE60-40 gearbox itself.

---

The Result

Following the redesign:

• System vibration was reduced by approximately 45%.
• Bearing operating temperature decreased by approximately 12°C.
• Positioning accuracy improved noticeably.
• After more than 20 months of continuous operation, no additional gearbox failures were reported.

---

What Can We Learn?

When a gearbox fails, the first instinct is usually to check:

• Gear ratio
• Torque
• Motor power

But these represent only a small part of the complete picture.

In this case, the gearbox did not fail because the transmitted torque exceeded its rating.

It failed because of a combination of:

• Unaccounted radial loading
• Repeated peak torque events during acceleration and braking
• Incorrect positioning of the timing pulley

Each factor alone appeared acceptable.

Together, they created the conditions that ultimately caused the failure.

---

Engineering Conclusion

When a gearbox fails, replacing it should not be the first step.

Instead, ask the following questions:

• Were peak loads calculated, or only average torque?
• Are radial or axial loads acting on the output shaft?
• Does the location of the pulley or gear create excessive bending moments?
• Are shock loads generated during acceleration, braking, or direction changes?

In many cases, the gearbox is not the real problem.

The mechanical design surrounding it is.