Menu
Hydraulic systems move excavators, lift aircraft landing gear, steer ships. They work under pressures that would rupture ordinary pipes. When they fail, the consequences range from costly downtime to fatal accidents. Testing is how manufacturers and maintenance crews catch problems before the fluid starts spraying.
The phrase “testing is the soul of hydraulics” comes up often in the industry. Nobody seems to know who said it first. It might have started in Germany. It might have come from an old Bosch Rexroth training manual. Either way, it stuck. Another saying that circulates in valve workshops: a good valve isn’t something you can calculate. The math gets you close. The test bench tells you if you actually got there.
A hydraulic pump rated for 3,500 psi doesn’t just stop working one day. Wear happens gradually. Internal clearances increase by microns over thousands of hours. Efficiency drops. The system runs hotter. Output pressure falls. By the time an operator notices something wrong, the pump may be weeks or months past the point where a test would have caught the problem.
Greg Haldeman manages the hydraulic repair shop at Bernell Hydraulics in Riverside, California. The company has been rebuilding pumps and motors since 1977. He talked about what he sees coming through the door.
“Most of the stuff we get is not catastrophic failures,” Haldeman said. “It’s gradual. A guy brings in a cylinder and says it’s not holding like it used to. We put it on the bench, pressurize it to 2,800 psi, and watch. Sometimes you see the rod creeping back right away. Sometimes you have to wait. Fifteen, twenty minutes. If it moves more than maybe an eighth of an inch in that time, the seals are going.”
He mentioned that the worst cases are the ones where someone ignored the early signs. “Had a pump come in last spring, completely scored inside. The guy said it had been making noise for a month. A month. If he’d brought it in when the noise started, we could have replaced the bearings. Instead the whole thing is scrap.”
Pressure testing is the most basic check. The idea is simple. You pressurize a component to its rated working pressure, or above, and see if it holds. Leaks show up as drops on gauges or as fluid pooling where it shouldn’t.
Static pressure tests check for leaks when nothing is moving. Dynamic tests run the system through its cycles while monitoring pressure. Both matter. Only testing can reveal the true performance of hydraulic valves. A valve that meets every dimension on the drawing might still chatter at certain flow rates or leak past the spool under sustained pressure.
The test pressure depends on the application. Industrial systems often test at 1.5 times working pressure. Aerospace components might go to 2 times or higher. Standards like SAE J343 and ISO 10100 specify procedures for different component types.
A basic test setup for a hydraulic cylinder includes:
The cylinder gets filled with oil and pressurized slowly. Most shops bring it up in stages. Half the test pressure first, hold for a few minutes, check for leaks. Then full test pressure. Hold times vary. Ten minutes is common for industrial cylinders. Aerospace specs often require 30 minutes or longer.
Temperature affects results. Cold oil is thicker and might mask a marginal seal. Most standards call for testing at operating temperature, usually somewhere between 100°F and 130°F.
Pumps and motors need more than pressure checks. Volumetric efficiency tells you how much fluid actually moves versus how much should move based on the pump’s displacement. A new pump might run at 95% efficiency. As it wears, that number drops.
Efficiency testing requires measuring flow at a known pressure and speed. A worn pump might still hit its rated pressure if you dead-head it against a closed valve. But open that valve and try to move fluid, and the flow falls short.
Karl Becker worked at Eaton’s hydraulics division in Eden Prairie, Minnesota for 22 years before retiring in 2019. He spent most of that time in product testing.
“We had pumps come back from the field that still made pressure,” Becker said. “Customer says the cylinder is slow. We put the pump on the dyno, load it up, and it’s putting out maybe 80% of what it should. All that missing flow is going right back to the tank through internal leaks. Pressure looks fine. Flow is garbage.”
His team used test stands that could simulate field conditions. Variable loads, duty cycles that matched actual equipment, temperature swings. “You can’t just run it at one point and say it’s good. Real world doesn’t work that way.”
Flow meters for hydraulic testing come in different types. Gear meters are common and relatively cheap. They work well up to maybe 4,000 psi. Turbine meters handle higher pressures but cost more. Ultrasonic meters don’t obstruct the flow path at all. They’re expensive. Labs use them. Most field shops don’t.
Testing during product development is different from testing in a repair shop. R&D testing tries to find the limits. Engineers push prototypes until something breaks. They want to know where the design fails, not just whether it meets spec.
Becker described some of the development work at Eaton. “We’d take a new pump design and run it at 120% of rated pressure for thousands of hours. Heat it up past normal operating range. Contaminate the fluid on purpose to see how fast it wore. The goal was to find weak points before customers did.”
Validation testing follows a protocol. Life cycle tests might run a component through millions of cycles. A directional valve might get actuated 10 million times. A pump might run 10,000 hours at full load. The test plan comes from application data. How will this thing actually be used. Then you simulate that, usually with some margin added.
Prototype testing generates data that feeds back into the design. A pressure spike in the transition curve might mean a port needs reshaping. Excess heat at high flow might mean the clearances are too tight. The engineers study the test curves and make changes. Then they test again.
Production testing happens after the design is finalized. Every unit coming off the line gets checked. The goal is different from R&D. You’re not looking for design flaws. You’re looking for manufacturing defects. A machining error. A bad seal. A contaminated assembly.
Production tests are shorter than development tests. Time is money. A factory making 200 pumps a day can’t run each one for a week. The test has to catch problems fast without slowing the line too much.
Typical production tests for a hydraulic pump might include:
Some manufacturers do more. Some do less. The level of testing often depends on the application. A pump going into a mining truck might get a 15-minute test. The same pump going into an aircraft might get two hours of testing plus documentation that fills a binder.
Haldeman mentioned that production testing doesn’t always catch everything. “We’ve gotten brand new pumps in here, still in the box, that failed on our bench. Not often. But it happens. Either their test missed it or the pump got damaged in shipping.”
Dirt kills hydraulic systems. Particles smaller than what the eye can see will score valve spools, wear pump surfaces, plug orifices. A system can be running fine and fail within hours if contaminated fluid gets in.
Cleanliness standards use codes like ISO 4406. A rating of 18/16/13 means there are up to 2,500 particles larger than 4 microns per milliliter, up to 640 particles larger than 6 microns, and up to 80 particles larger than 14 microns. Lower numbers mean cleaner fluid.
New oil from a drum is not clean. It might come in at 20/18/15 or worse. Most hydraulic systems need fluid in the 17/15/12 range or cleaner to hit their design life. Servo systems and aerospace applications often require 15/13/10 or better.
Particle counters pull a sample and run it past a laser. The machine counts shadows. Results come back in a few minutes. Portable counters exist but they’re finicky. Lab analysis is more reliable. Shops send samples out and get results in a day or two.
Water is the other contaminant that matters. More than about 500 ppm and you start getting problems. Corrosion. Additive breakdown. Reduced lubricity. Some oils can hold more water than others before it causes trouble. But the safest approach is to keep it out.
A proper test stand is a controlled environment. Known fluid, known temperature, calibrated instruments. Field conditions are messier.
Mobile testing equipment exists. Pressure gauges, flow meters, particle counters in ruggedized cases. Technicians use them to diagnose problems on site. The results are rougher than shop data but often good enough to decide whether a component needs to come out.
Haldeman from Bernell Hydraulics mentioned the tradeoff. “Sometimes a guy can’t take his machine down. Big crane, middle of a job. We’ll go out there with portable stuff and get some numbers. It’s not as accurate. You’re testing in the dirt with the sun beating down. But if the pump is bad, you’ll see it. Maybe you can’t say exactly how bad. But you’ll know.”
Some operations do regular sampling programs. Fluid analysis every 250 hours or every month. They track trends over time. A spike in iron particles means something is wearing. Copper means bronze bushings. Silicon usually means external contamination.
Predictive maintenance based on testing data costs money upfront. Sampling, analysis, record-keeping. But catching a failing pump before it contaminates the whole system saves more than the program costs. Most large fleet operators figured that out years ago.
Equipment matters. So does the person running it. A gauge that’s out of calibration gives bad data. A technician who doesn’t follow the procedure gives bad data too.
Becker talked about training at Eaton. “We had guys who’d been doing this for 20 years and still made mistakes. Set up the test wrong. Didn’t wait long enough for the temperature to stabilize. Read the wrong scale on the gauge. We ran refresher training every year. Some people thought it was a waste of time. It wasn’t.”
Understanding the test curve is an indispensable part of hydraulic training. A pressure-flow curve tells a story. The shape matters. A bump at a certain flow rate means something. A flat spot means something else. Technicians who can read curves catch problems that technicians who just check numbers will miss.
Documentation matters as much as the test itself. A pressure reading means nothing without knowing the conditions. What was the fluid temperature. What was the inlet pressure. How long did you hold it. Shops that keep good records can spot trends. Shops that don’t just react to failures.
Testing costs money. Time on a test stand is time the component isn’t generating revenue. Customers push for faster turnaround. The temptation is to cut corners. Run a shorter test cycle. Skip the flow check if the pressure looks okay.
Haldeman said his shop doesn’t negotiate on test procedures. “We’ve had guys ask can you just check it quick. No. We do the full test or we don’t put our name on it. Had one guy complain about the bill for bench time. I told him the test costs less than one service call when that cylinder blows out on the job.”
The industry doesn’t have universal testing standards that cover everything. Different manufacturers specify different procedures. OEM requirements don’t always match aftermarket practices. Some shops test to the standard. Some test to the customer’s spec. Some do whatever is fastest.
There’s a saying that floats around engineering departments. “Engineering hydraulics” should not be “formulaic hydraulics.” The formulas give you a starting point. They tell you what the theory predicts. But hydraulic systems operate in the real world. Fluids compress a little. Seals wear. Tolerances stack up. Temperatures change. The only way to know if the theory matches reality is to test.
That variation is part of why the phrase exists. Testing is the soul of hydraulics because it’s the one step that separates good work from guesswork. A rebuilt pump looks the same whether it was tested properly or not. The difference shows up later, out in the field, when it matters.
“A leader in hydraulic technology, A convenient choice around you”
