Insane Hydraulics
Bold and audacious blog about fluid power
The best way to gain "real" knowledge about hydraulic pumps (and I mean usable knowledge - like control function, adjustment procedures, troubleshooting techniques, and such) is to run them in test benches - nothing can beat this! Test benches are workshop machines, yes, but they also are, by far, the best learning tool you will ever find. Toying around functional parameters, like the input speed, outlet pressure, control signals, and fiddling with the adjustments with the objective of seeing how it affects the operation of a pump gives invaluable practical knowledge - something you will never find in manuals and catalogs. A manual will tell you to turn an adjusting screw clockwise to increase a pressure setting, but it will not tell you how the pump will sound when you do it.
And let us not forget that benches offer the benefit of being able to do all of the above in the comfort of your workshop, as opposed to field conditions, when you adjust a hydraulic pump that's hidden under the belly of an industrial vehicle while lying on your back on a piece of cardboard and your "learnings" are constantly being disturbed by the hydraulic oil streaming down your elbows on the way to your armpits as you unscrew the lock-nut of an adjustment screw... Still educational? - Maybe yes! But far "less captivating", if you ask me.
Anyhow, I like using our test bench for "didactic purposes", and sometimes Lady Luck throws in a unit "with a twist", a unit that bears a particularly uncommon malfunction, and it is these units that are the best both for teaching and learning, as well as for putting one's troubleshooting skills to a test - and it is one of such cases that I would like to describe in this post.
A couple of days ago I was introducing a new colleague to the world of hydraulic pump troubleshooting. The man had some (but not much) experience in this field, and a malfunctioning Sauer Danfoss series 90 closed-loop pump seemed like a perfect opportunity to show the recruit how things are done. The pump had been working flawlessly till the day a scheduled hydraulic oil and filter replacement was performed, after which it stopped responding to the control signal (the unit was equipped with a standard electric displacement control). The routine maintenance had been done by the same crew countless times before, but that time something went wrong, and nobody could figure out exactly what. Several mechanics troubleshot the transmission, but to no avail - all attempts to apply the standard "let's replace something and see what happens" technique failed, and in the end, the unit was dismounted from the vehicle and sent to our shop.
The whole troubleshooting process took about an hour, and the machine owner was quite surprised when I told him why his pump had stopped pumping so abruptly! But since I already started to talk about the academic value of testing hydraulic pumps on proper test stands, I would like to depict this troubleshooting episode as a process that consists of two parts - the Information Gathering and the Troubleshooting, each subdivided into a series of steps, which will give you an idea about the logical choices that lead to a quick discovery of the root cause of the failure:
Listen to the client tell the pump's story (patiently, and with a polite smile on your face, but don't overdo it - I mean the smile - this smiling business can get pretty awkward if you overdo it) and ask questions about the circumstances of the failure. Squeeze out as much information as you can. Talk to the operator or the person directly involved with the machine in question. Remember - you are the Sherlock Holmes of the hydraulic crime world!
The complaint is - the pump is not responsive to the input signal. The background story is - regular maintenance was performed (oil and filter change), after which the pump "never pumped again".
Even if the client reassures you that the pump is "all good inside", inspect it thoroughly before mounting it on your expensive test bench, (e.g. perform the three simple checks) - to make sure that you don't contaminate the bench with all sorts of "goodies" as soon as you turn it on.
All looks good!
Mount the pump on the bench and run it to confirm the existence of the problem.
The existence of the malfunction is confirmed - the pump is "blind and deaf" - which means that now it's time for
Troubleshooting a closed-loop pump is a simple process of measuring pressures in key points and drawing logical conclusions from the readings. Tools required - test fittings, pressure gauges, a sound head, and background knowledge.
The first test point - the charge pressure. Why? If the charge pressure isn't stable or is too low and you can't adjust it - then you'll have to troubleshoot the charging system (the charge pump, the charge pressure relief valve, the filter, the check-valves, etc) or the rotary group (excessive internal leakage), and most likely disassemble the unit.
The charge pressure is OK.
Test point number two - servo-pressures. Measure pressures in the servo-cylinder (both sides at the same time). Why? Reading servo-pressures gives you an answer to the question of why the pump isn't stroking - whether it's a hydraulic problem (bad "P", i.e. low servo-pressure, or bad "T", i.e. oil can't come out of the servo-cylinder) or a mechanical issue (the servo-pressure differential is present, but the swashplate isn't moving, which can be caused by a damaged swashplate bearing, a broken servo-link, etc...).
When the input signal is "on" the servo pressures don't change. Now we are getting somewhere - the problem is definitely hydraulic (control-related) and not mechanical.
Now let us check the electric displacement control by measuring pressure in the control module ports X1 and X2 (both at the same time) - to check if the PCP (pressure control pilot valve) is working properly. Why? The series ninety displacement control is a two-stage device - the PCP pilots the spool of the displacement control, and the spool directs oil from the charge pressure gallery to the servo-cylinder chambers to position it according to the input signal. If the PCP hydraulic signal is OK - the displacement control spool is probably stuck, if the PCP hydraulic signal is bad - the problem is most likely the PCP itself:
Very low pressure is read, less tan one bar on both sides (normal value would be around 8 bar in both ports at null), no changes with the input signal change. Great! We're on the right track! Now we've narrowed down our search - the PCP is not supplying any oil to move the spool of the four-port proportional valve that controls the servo-piston. At this point, by the way, I was pretty sure that I would find the PCP inlet screen blocked with contamination - something I'd seen many times before. In any case, these pressure readings meant - PCP out! So we took it out - and the filter looked pretty darn clean, the small orifices also turned out to be unobstructed - bummer! Passing on to
Well, since we just found out the PCP was all good and unobstructed, the only explanation for the oil not coming out of it had to be no oil coming in, in other words, something was blocking the connection between the charge pressure gallery and the PCP. Let's have a look, shall we? - displacement control valve out! Maybe the feed orifice is clogged?
So, we took this one out too - and, what do you know, no blocked orifices were found - bummer!
Ok, so we have no blocked orifices in the control module, is it really possible that the charge oil isn't reaching the control? I wish there was a way to test this, wait a minute - there is! This is a test stand - you can do whatever you please here, so if you want to check if the oil is really reaching the control, simply have a look! Although such a procedure may seem somewhat redneck-ish, it is definitely the "take away all the doubts" test, and it boils down to running the pump (ve-e-ery slowly) without the displacement control valve in place - just to see if oil is coming out of the charge pressure hole (servo pressure supply) under the control. Don't laugh now, visual confirmation of flow (if you know what to expect) is a valid troubleshooting technique, as long as you know that it's exclusively for low flows and controlled environments, of course. Under any other circumstances an open port test can do serious damage to your precious self, so, please, never do this!
Done - and no oil is coming out of the charge pressure hole! Hurray! It's not the control, not the PCP, not mechanical, not electrical - the problem is the obstructed oil passage between the charge pressure line and the displacement control. Now, to find the obstruction (before throwing in the towel and passing on to the disassembly stage) let's take out the small screen in the hole to inspect the channel more thoroughly.
Found it! - a small piece of plastic film got inside the oil channel and blocked it completely when it was pushed against the screen:
After removing it the pump was as good as new!
Where did the piece of plastic come from? - might you ask - Quite simple, lads, the piece of plastic is a small piece of the wrapping of the new charge filter, which fell inside the filter when the wrapping was removed, and eventually found its way to the small screen behind the control.
Morals of the story: