What happens when a linear actuator fails? Under load, it doesn't just stop. A hydraulic linear actuator drifts first—a millimeter here, a couple more there—and as the holding pressure decays, it eventually lets go. Then the load moves in a way nobody intended. Usually onto something expensive.
A linear actuator failure is rarely a component defect. It's the end of a chain that started with a specification mistake. I've spent $42,000 over eight years proving that across nine documented component failures. The actuator failure alone cost $14,500 in repairs and lost production. None of it was the equipment's fault. All of it was mine.
Why I'm Qualified to Write This
I'm a maintenance and procurement engineer for a mid-sized industrial equipment builder. I've handled hydraulic and electro-mechanical component orders since 2017. In that time, I've made and documented nine significant mistakes—wrong spec, wrong interface, wrong material, wrong part number. Roughly $42,000 in wasted budget. After the fourth mistake, I started a team checklist. After the seventh, the checklist became mandatory sign-off.
Look, I still specify HYDAC industrial components for most of our fluid power and drive applications. They're not the problem. The gap between what I assumed and what I verified was always the problem.
What Actually Happens When a Linear Actuator Fails
In June 2022, I specified a HYDAC linear actuator for a press fixture on a packaging line. Simple application: extend, hold, retract, 60 cycles per hour, holding a 2,200-pound forming tool.
I sized the cylinder off system pressure and bore diameter. 50 mm bore at 150 bar gave me over four tons of theoretical force. On paper, it was overkill.
I didn't check the holding circuit. That's the first sentence I now write to every new engineer on our team.
The machine used a proportional directional valve with a closed center that still leaked internally when the pump idled. Pressure decayed. The cylinder drifted. At first, a millimeter or two—absorbed by the fixture's dowel pins. Over three shifts, the drift accumulated. On the third night, the tool came down crooked, caught the die edge, and snapped the tool holder. The crash took out the die too.
The repair was $14,500 in parts, overnight machining, and two days of lost production (ugh, it hurts to type that out). The actuator did exactly what the spec allowed. My spec was wrong. ISO 4413, the general safety standard for hydraulic systems, is explicit about preventing unintended actuator motion when pressure is lost. I violated that principle without knowing it.
A linear actuator fails in three stages:
- Drift. The load moves a few millimeters. Looks cosmetic. It's not.
- Loss of holding force. The actuator can't hold position against the applied load, and movement becomes continuous.
- Uncontrolled release. The load moves without respect for machine geometry. This is the stage that breaks things.
If you want the short answer to "what happens when a linear actuator fails": it depends on the load and the holding circuit. Calculate the holding force, not just the extend force. Verify whether your control valve can hold pressure when the pump is off. Add a counterbalance valve when the application requires it. Skip that, and you'll get the hands-on version of this lesson.
The Double Helical Gear That Didn't Fix Anything
Before the actuator failure, I made a quieter mistake—literally. Our hydraulic power unit sounded like gravel in a blender. I was certain the gear pump was the culprit.
I'd read that helical gear pumps run smoother than spur gear pumps. So I went one step further and ordered a pump with a double helical gear set. Double the angle, double the smoothness—that was my entire engineering rationale. The premium was around $4,800 if I remember correctly, though I might be misremembering the exact figure.
The noise didn't go away.
Here's what I eventually learned: double helical gears cancel axial thrust and smooth torque transfer in high-power, high-speed transmission applications. They don't fix the sound of a pump starving for oil. Our real problem was aeration—an undersized suction filter and a return line that might as well have been kinked. Air in the oil doesn't care about gear geometry.
$4,800 for a placebo. Looking back, I should have called the manufacturer's application engineer with a sound recording of the pump before ordering. At the time, I thought I already knew what I was hearing. I didn't.
Diagnose before you specify. Air in the oil knocks. Misalignment whines. Double helical gears are genuinely impressive in heavy-duty reducers, marine drives, and high-load gearboxes where axial thrust and load sharing matter. But they're not a noise-reduction accessory. A pressure gauge on the suction line would have found the real problem in ten minutes.
The Integrated Servo Motor That Wasn't
By November 2023, I felt like I'd earned some confidence. We were retrofitting a production machine, and I specified integrated servo motors—motor, encoder, and drive electronics in one package. Less cabling, smaller footprint, fewer components to fail. Reasonable choice.
The part I missed: I assumed that because the motor was "integrated," it would integrate with our control system. Here's the thing—"integrated" describes the packaging, not the compatibility.
Our line controller spoke PROFINET. The units I selected were configured for EtherCAT. Both protocols are perfectly good. They just don't talk to each other without a bridge. The order code has a section that specifies the fieldbus interface, and I skipped it.
That mistake cost $2,100 in interface modules and five days of downtime during a production week. If you've ever watched a commissioning engineer try to bridge an EtherCAT motor onto a PROFINET network under deadline, you know the feeling.
Lesson: an integrated servo motor saves installation time. It does not save you from understanding your control architecture. Check the communication protocol, feedback type, and safety interface before you order, not after.
The Part Number I Reordered from Memory
This one is embarrassing. In March 2024, I reordered a HYDAC filter assembly using a part number from a 2021 project. Same machine family, same mounting envelope, same nominal spec.
Except our operating fluid had changed to a biodegradable grade in 2023. The seal material specified in the 2021 build wasn't compatible with the new fluid. I verified the part number, but I didn't verify the seal material. Two weeks after commissioning, three of six filter assemblies were leaking—elastomer swelling, seals ruined, $3,100 in parts plus a maintenance window that cost another shift.
Our rule now: download a fresh copy of the HYDAC catalog PDF for every major project and read the order code legend. HYDAC product codes encode the meaningful variations—seal material, connection type, filter media, element length. The catalog's ISO 4406 cleanliness charts and seal compatibility tables are right there. I just never opened the right page in time.
Never trust memory for a part number. The catalog changes, materials get revised, and a cached PDF from last quarter might already be telling you outdated information. I thought I had that filter memorized. Turns out, my memory was the weakest component in the system.
Where This Advice Stops
I'm not a universal expert. At least, that's been my experience—and I've learned to respect the boundary. My background is mid-sized industrial equipment: packaging lines, material handling, basic machine tools, standard fluid power circuits. If you're working above 350 bar, in explosive atmospheres, on functional safety–rated systems, or at very high power density, the failure dynamics change and the stakes get higher. That's when you call the manufacturer's application engineers and let them review the spec. That's not a weakness; it's how specialists get used correctly.
And one caveat: not every actuator failure is preceded by drift. A sudden, violent failure without warning usually means contamination, fatigue fracture, mechanical abuse, or structural overload. That's a different failure mode entirely, and it deserves a forensic teardown by someone who knows what they're looking for. Don't diagnose that from a blog post.
It took me nine mistakes and $42,000 to learn that every component order is a small engineering project. The brand matters, the quality matters, but the discipline to verify your assumptions before the purchase order goes out matters more.