What Happens When a Linear Actuator Fails? Usually It Was the Setup, Not the Actuator

What happens when a linear actuator fails? In my experience, the machine doesn't scream. It just stalls mid-cycle. The next time you command it, you hear a grinding sound that wasn't there yesterday.

My first reaction years ago was simple: bad actuator. I ordered a replacement, installed it, and watched that one fail too. The second failure took about three weeks longer to happen, which made it feel less like a pattern and more like bad luck. It wasn't bad luck.

I'm the person who handles Danfoss VFD and motion control orders, so I've seen both sides of this problem. Back then, I was also the person comparing timing belt price and lead time because the machine had a belt drive I wanted to replace while I was in there. I thought that was being efficient. It was the same mistake wearing a different hat: I focused on the part, not on why the old part died.

The failure that looked like a bad part

What exactly goes wrong when a linear actuator fails? Sometimes it stalls mid-stroke. Sometimes the motor runs but the rod doesn't move. In my first serious brush with this, the actuator was on a small indexing mechanism and had run for about six weeks. One morning it stopped in the wrong position. The operator cycled it again, and the motor ran but the rod stayed still. I remember thinking: a $220 actuator should not do that after six weeks.

I replaced it with an almost identical unit. That one lasted four weeks. Only then did I stop blaming the hardware and start asking questions I had skipped.

Why the actuator really failed

The questions were uncomfortable because the answers were obvious in hindsight. What was the actuator's rated duty cycle at the load I was using? I had to dig out the datasheet. The rating was 25%. How often did that machine move? Nearly continuously. The datasheet assumed a cycle of roughly 75 seconds of motion followed by a rest period. Industrial duty classes exist for exactly this reason: IEC 60034-1 separates S1 continuous running duty from S3 intermittent periodic duty. The actuator was rated for something closer to S3-type use. The machine was effectively demanding S1. The motor never got enough rest, heat built up, and the thermal protection eventually shut things down. The actuator didn't die from a defect. It died from being applied outside its specification.

There was also a side load. The mounting holes were close but not perfectly aligned. You wouldn't notice it by eye, but the rod bushing noticed. Side load increases friction. Friction is not a footnote in an electric linear actuator; it becomes heat inside a device that already had no time to cool.

That's the pattern that matters. Every comparison I did—timing belt price, actuator price, delivery lead time—assumed the old part was the problem. The failed component was actually a messenger. The load profile and the mounting design were the problem.

The Danfoss VFD parameters problem

A few months later, I repeated the same mistake with motor controls. We supplied a motor and a Danfoss VFD for a small conveyor. The wiring was done, and the drive was set up quickly. The motor ran, so everyone assumed the setup was finished.

Two days later, the drive tripped on overcurrent during acceleration. Someone lengthened the ramp time, and then it tripped on overvoltage during deceleration. It was easy to say the Danfoss VFD was causing problems. It wasn't. The drive was still running on factory defaults plus whatever numbers were already in the display. It had not been configured for that specific motor.

A VFD doesn't just vary frequency. It uses motor data to decide how much voltage to apply and how to handle slip. If the Danfoss VFD parameters don't match the motor nameplate, the drive can command excessive current at low speed or trip at the worst possible moment. That's the electrical version of the actuator problem.

The fix was not a different drive. It was a proper commissioning session. We entered the nameplate data, ran the drive's motor adaptation routine, and the same drive ran smoothly. That's why this article exists: if you catch yourself searching for a Danfoss VFD VLT manual after the fact, download it. Better yet, download it before you connect power. The commissioning section tells you which parameters to check, and when automatic motor adaptation makes sense. That step is not paperwork. It is what makes VFD control protective instead of destructive.

The Arduino servo motor version of the same trap

The same logic appears in smaller projects. I see people who built an Arduino servo motor test rig at their desk and now want to scale it into something industrial. Bench prototypes are genuinely useful. I don't want to discourage anyone from experimenting.

The trap is assuming a demo has the same requirements as production. An Arduino servo motor experiment usually runs on a lab supply for a minute at a time, with no load, no dust, no duty cycle, and no safety chain. Industrial servo control is a system: motor, drive, feedback, cabling, tuned gains, protective functions. It can look like the same motion, but it is not the same engineering problem.

I've made enough of these mistakes to know that "it worked on the bench" is not a specification. It's a starting point.

What the mistake actually costs

Let me put some rough numbers on the first actuator failure. The replacement actuator and hardware came to maybe $530. Service time and troubleshooting added more. If I remember correctly, the total direct cost was somewhere around $1,400. I might be off on the exact invoice amount, but the part I remember clearly is the delay. The machine sat still for two days, and people associated me with that delay.

Since then, I've looked at returns differently. When a component comes back as dead, I ask whether the application killed it. More often than you'd think, the part tests fine on the bench. It was misapplied, misconfigured, or selected for the wrong duty.

Why this is a brand problem, not a parts problem

Here's the lesson that changed how I work. The customer doesn't buy a linear actuator or a VFD. They buy a machine that moves a product from one place to another without stopping. When that machine stops, they remember the last name involved in the chain. If a Danfoss VFD trips because the setup was skipped, someone says Danfoss is unreliable. If a linear actuator fails because the duty cycle was wrong, the actuator manufacturer gets the blame. The component's quality is judged by the project's last failure.

Quality isn't only about whether a part meets its specification. It is also about whether the part is used the way it was designed to be used. A $50 difference in component quality can matter. But a $0 difference in setup discipline can matter much more.

A shorter way to check before you buy

I don't have a perfect system. I still make calls too fast sometimes. But I now keep a short pre-buy list:

• Start with duty cycle. How many cycles per hour, how many hours per day, what ambient temperature, and how much rest does the mechanism get?
• Check the whole motion profile, not just the peak force. Side loads and acceleration matter more than the highlighted max load number.
• Pair the motor with the correct drive setup. If it's a Danfoss VFD, read the Danfoss VFD VLT manual before commissioning and enter the nameplate Danfoss VFD parameters. Run motor adaptation when the manual recommends it.
• If a prototype used an Arduino servo motor, treat that as proof of concept, not as a production spec.
• Buy from a supplier who asks questions. If they quote a timing belt price and don't ask what it is driving, you're talking to a warehouse, not a partner.

The final step is the one I still have to force myself to do: admit when a component failure is my fault. Usually, it isn't the actuator. It's the hour I didn't spend reading the manual, or the decision to save a few dollars on a part that was never the real issue.