News
Inside A High-Pressure Pump
The line in the specification read: working pressure 50 MPa. Twenty characters, and they describe a number that a pump reaches for a fraction of a second at a time, on a good day, with everything in the circuit working.
What the drawing does not say is how much the pressure is allowed to vary, how long it takes to reach the figure, and what happens when it drops. Those three questions are answered by the internal parts of the pump, and each part fails in a way that produces a different cleanliness result on the part.
1. The Plunger And Its Packing
A high-pressure pump is a positive displacement device: a plunger reciprocates in a cylinder, and each stroke pushes a fixed volume of water out. The plunger runs at pressures where a rubber lip seal would be extruded, so the seal is a stack of packing rings compressed around the plunger, and the contact is maintained by a combination of preload and the pressure itself.
The packing is a wear item with a characteristic failure. It does not fail suddenly; it weeps. A small amount of water escapes past the rings and runs out of a drain port on the pump body, which is why that port exists and why a pump is usually mounted with a visible drip tray.
The consequence for cleaning is subtle. A weeping packing reduces the delivered volume, not the peak pressure, so the gauge still reads what it always read while the flow through the nozzle falls. On a gallery that needs volume to displace what is inside it, a pump with worn packing cleans progressively worse without a single alarm.
What The Drip Port Is Telling You
The drain between the packing stack and the crankcase exists so that leakage has somewhere to go, and it is the most under-used diagnostic on a high-pressure machine.
A pump that drips a few drops a minute is normal. A steady trickle means the packing is due. A stream means it is overdue and the plunger may already be scored, because water that escapes past the rings carries no lubricity and the plunger runs dry against the next ring in the stack. That progression takes weeks, and a glance at the drip tray once a shift catches it at the stage where it is only a packing job rather than a plunger replacement.
2. The Two Check Valves
Each cylinder has an inlet valve and an outlet valve, both one-way. The inlet opens on the suction stroke and closes on the pressure stroke; the outlet does the reverse. They are the parts that convert reciprocation into flow, and they are the parts that fail on dirty water.
A check valve that does not seat fully produces a leak back through the pump. The symptom is a pump that reaches pressure slowly, or fails to reach it at all, and it usually appears after a period of running with inadequate inlet filtration rather than as a discrete event.
The practical point is that a high-pressure pump is sensitive to particles that a low-pressure circuit would tolerate. A 100 micrometre particle that passes through a 1.0 MPa washer without consequence can hold a check valve off its seat at 50 MPa. That is why the filtration on a high-pressure circuit is tighter than the nozzle orifice, not equal to it.
3. The Unloader
A positive displacement pump must never be dead-headed: if the flow stops and the pump keeps stroking, something breaks. The unloader is the part that prevents this. When the spray valve closes, the unloader senses the rise in pressure and diverts the flow back to the tank at low pressure, holding the system in a standby state.
Its failure modes matter because they are intermittent. An unloader that sticks closed produces a pressure spike each time a valve closes, which stresses every seal in the circuit. An unloader that sticks open produces a system that will not build pressure at the start of a cycle, then builds it partway through as the valve reseats. The second case is the one that produces mystery defects, because the first part of every cycle runs at low pressure and nobody sees it.
Why The Filtration Grade Tightens With Pressure
The clearance a check valve can tolerate does not change with pressure, but the consequence of a particle holding it open does. At 1.0 MPa a leaking check valve costs a little flow. At 50 MPa the same leak is a path for high-pressure water to escape backwards through the pump, and it erodes the seat as it goes.
That is why the inlet filter on a high-pressure circuit is specified finer than the pump maker's minimum and finer than the nozzle orifice. The filter is protecting a clearance measured in microns, and on a machine fed from a shared tank it is also the only thing separating the high-pressure circuit from everything the low-pressure stages loosened.
4. The Accumulator
A plunger pump delivers flow in pulses, one per stroke. At 50 MPa those pulses are violent, and the accumulator is a gas-charged vessel that absorbs them, smoothing the delivery into something closer to continuous.
This part is the one most often neglected, because it has no obvious wear surface. Its gas charge slowly permeates through the bladder or the diaphragm, and over months the precharge falls. The pump still runs, the gauge still reads a plausible average, and the pulses become sharper.
The consequence is a pressure profile that oscillates around the setpoint rather than holding it. On a gallery that has to hold pressure for a defined time to displace its volume, an oscillating supply means the effective hold varies from cycle to cycle. Two parts from the same shift can see different treatment, and the resulting spread in cleanliness looks like a process variation rather than a pump fault.
5. The Gauge And The Relief Valve
Two instruments, and both are easy to misread.
A pressure gauge shows the pressure at the point it is connected, which is rarely the point of interest. Mounted at the manifold, it reports the header; mounted before a valve, it reports the supply. Neither tells you what the nozzle sees, and on a pulsed system either one may be showing a peak rather than a mean, depending on its damping.
The relief valve is a safety device, not a control. It should never open in normal operation, and if it does, the pressure setting is wrong or something downstream is blocked. A relief valve that has lifted repeatedly has probably been damaged by its own discharge, and a damaged relief valve is a risk rather than a protection.
6. What Each Failure Looks Like On The Part
Because the parts fail differently, the cleanliness result carries a signature.
Worn packing shows as a slow, uniform decline in cleaning effectiveness across all features, with the gauge unchanged. Check valve leakage shows as a failure to reach pressure, most visible on the first part after a stop. A sticking unloader shows as good parts and occasional bad ones, clustered at the start of a run. A discharged accumulator shows as a wide spread on a part that used to be consistent.
None of those four appears in a maintenance schedule based on running hours, which is the reason they are usually found by a customer rather than by a technician.
The Spread Test
The simplest way to separate a pump problem from a process problem is to measure the same feature on five consecutive parts and look at the spread rather than the average.
A process problem gives five values clustered around a wrong number. A pump problem gives five values scattered around a correct one, because the pump is not delivering the same treatment each cycle. A discharged accumulator and a sticking unloader both produce that scatter, and both are invisible in an average that sits comfortably inside the limit.
7. Why The Number On The Drawing Is Not Enough
A working pressure is a target, and every pump in the circuit spends some of its life below that target. What determines the result is how far below, for how long, and whether anyone would know.
The useful specification is not a single figure but three: the pressure at the nozzle, the tolerance on that pressure during a defined hold, and the time allowed to reach it at the start of a cycle. Those three describe the pump's contribution to the result, and none of them can be read from a gauge mounted upstream.