The Pump That Sounded Like Gravel
By 9:15 on the last Tuesday of March, I was leaning over a workbench in Forney, Texas, looking at a pressure washer pump in pieces. A few feet away, Ray was pulling off his work gloves and telling me exactly what was wrong. 'It's the unloader,' he said.
Ray runs a two-person shop that handles air conditioner repair in Forney and the surrounding towns. The pressure washer on his bench is company equipment, and it sees hard duty. His crew uses it to clean condenser coils, wash filter racks, and knock years of grime off rooftop units before they start a repair. When the pump started making a sound like gravel in a coffee can, Ray filled out a warranty claim.
I'm the quality compliance manager for Forney, and I review returned tools before anyone approves a replacement. I've been doing it long enough to know that the first diagnosis from the field is often wrong. Ray's unloader valve looked fine. The pump head did not. Its valve seats had small craters, and the oil had a metallic shimmer to it. Pitting plus metal in the oil usually means one thing: cavitation.
Ray asked what cavitation had to do with a noisy unloader, so I gave him the short version.
What Causes Hydraulic Pump Cavitation?
Hydraulic pump cavitation starts at the inlet, not inside the pump. When a pump wants more fluid than the supply line can deliver, the pressure at the inlet drops. If it drops low enough, the fluid flashes into tiny vapor bubbles. The pump carries those bubbles into a high-pressure area, and they collapse with enough force to erode metal. That collapse is the 'gravel' sound Ray heard.
Ray's pressure washer pump is a positive-displacement hydraulic pump, but it moves water. An oil hydraulic pump works the same way. The usual causes are all on the supply side: a plugged suction strainer, a line that's too small or too long, a kinked hose, a filter that's too fine, or fluid that's too thick to move into the pump fast enough.
On Ray's machine, the inlet screen was packed with silt. The crew had been feeding the pump through a long, undersized hose because the nearest spigot was too far from the condenser unit. The pump was starving. No pump design can fix that.
One distinction worth learning: if air leaks into the suction side through a loose clamp or a worn shaft seal, that's aeration, not cavitation. Aeration sounds similar and damages pumps too, but the fix is different. Cavitation is a flow problem. Aeration is a leak problem.
Pump engineers use the term net positive suction head to describe the same thing. According to pumps.org, the Hydraulic Institute, inlet pressure needs to stay above the fluid's vapor pressure. For a field tech, that simplifies to: stop starving the inlet.
The Repair
Once Ray trusted the diagnosis, the repair was straightforward. We cleaned the inlet filter, swapped the undersized feed hose for a larger one, and replaced the worn valve kit. Then we put everything back together with proper torque.
Before reassembly, Ray reached for the twin grip pliers set on his bench. The brass inlet adapter was rounded from an earlier repair, and those parallel jaws grabbed it around the flats instead of the corners. It came loose without damaging the manifold. I've watched too many techs hammer a fitting like that and turn a small job into a replacement part.
One stripped cover screw also needed to come out. Ray drilled a careful pilot with the Forney mini drill bit set 60239 and then used an extractor. The bit is small, but because he let it cut at low speed instead of leaning on it, the screw came out and the threads survived. That's the thing about mini bits: they punish careless hands, but they're perfect for precision work.
I also noticed the pump mounting bolts had been vibrating loose. At some point, someone had replaced the original metric split lock washers with plain flat washers. A loose mount lets the pump rock, which makes seal wear worse. We replaced all four bolts with proper metric split lock washers and torqued them evenly. That's a boring sentence, but it's how a repair stays fixed.
When we fired the pressure washer up, the gravel sound was gone. Ray ran it for twenty minutes with the wand open and watched the pressure gauge hold steady. The pump was not dead. It was starving, and starvation has a sound.
What I'd Tell a Technician
Before you replace a hydraulic pump, look at the inlet first. Check the filter, the strainer, the hose diameter, every clamp, the oil level, and the reservoir vent. Most pump failures I see are supply failures wearing a pump-shaped mask.
It took me about four years and more warranty teardowns than I can count to understand that. The pump is the messenger. It will keep failing until you fix what feeds it.
That said, this was one field diagnosis on one pressure washer setup. If you're working on an oil hydraulic system, there are more variables—viscosity, reservoir size, breather ratings, suction line design. When in doubt, talk to someone who has the design curves. Don't just throw another pump at a problem and hope.
Maren Jorgensen
Maren Jorgensen is an independent hand tool and torque applications analyst covering wrenches, pliers, screwdrivers, hammers, sockets, ratchets, hex keys, and tool sets. She applies ISO 6789-1 torque-tool conformance principles while examining jaw capacity, leverage, fastener engagement, torque range, accuracy, handle geometry, and material hardness. Her practical guides help tradespeople and procurement teams select suitable tools, plan controlled tightening, and compare durability without relying on brand reputation alone.