Insane Hydraulics
Bold and audacious blog about fluid power
It seems to me that I am somehow able to predict the future because when I make a post about a rare hydraulic phenomenon, it often happens to me "just like that" a couple of days later! So, today, I want to put the last nail in the coffin of the matter of hydraulic pumps that should prime after an oil change because they are under a large positive head of oil - and yet they don't unless you give them a "bleeding nudge" and I want to show you how (just yesterday) a large fixed displacement pump failed to prime even after the air pocket was bled from the suction line! Impossible? I would probably be saying the same thing, too, had I not seen it with my own eyes! And to those who haven't read my previous posts on this subject - do check them out - here and here - this is a critical topic!
Now, as you can imagine, when you are working on a hydraulic installation and suddenly come about a fixed displacement pump that is rotating, but instead of displacing the oil is sloshing about an air/oil foam (or at least that's what I imagine a pump would be doing in a situation like the one I am about to describe) - you are supposed to be lightning fast to stop the prime mover and correct the issue because your pump will be toast pretty soon! So, as soon as I saw what was happening - I did my best to document this with the FLIR camera of my phone, but you can definitely tell the picture was taken quite hastily. My apologies for that in advance - but I did take the thermal image, and despite the "less than ideal" framing, it still clearly shows what is going on.
Let me set the stage real quick - a massive 5000 liter (1300 gallons) HPU, an oil change, and a hydraulic tech (yours truly) overseeing the operation. Nothing too complicated - especially given my extensive negative experience with HPUs with treacherous pseudo-flooded inlets. And this, here, is the "star of today's show:"
This is a Rexroth PVV5 193 cc (11.8 cubic inch) fixed-displacement vane pump. The square manifold underneath it is just a custom-made suction connector, and as you can see in the second picture - there's even a 1/4'' BSPP bleeding plug on top of it. The pump is "sitting under" at least a meter (3 ft) of oil, by the way. To remove the air pocket from the suction line, I closed off the suction valve, removed the plug, and then cracked the valve open and waited until oil came out of the bleeding port (and quite a mess it did, by the way, because "cracking " a large butterfly valve is harder than it may seem) - and then was pretty confident that this pump was "good to go". To be honest - at certain point a thought crossed my mind - maybe I should get a short test hose from my gauge box and bleed off air from the Minimess socket mounted directly at the pump's outlet (you can see it in the first picture), but then I thought: "Na-a-a-h.. it's good...", which actually meant: "The gauge box is in the car, and the car is downstairs, and I will be damned if I'm going to walk up and down those stairs one more time in this heat" - it's been consistently hitting 40C plus (104F) around here, so... yeah...
I bled the air from all other suction lines, double-checked everything, and left the commission for the next morning. So far so good. I came back the next day and started commissioning the circuits one at a time. There's no need to go into much detail about the convoluted hydraulic circuitry of the HPU, all you need to know is that it had several circuits, big and small, running single and tandem pumps (total of 5 electric motors) and that the service mode in the PLC allowed me to run and test them one at a time. Oh yes - I forgot to mention one very important thing - the vane pump circuit was mirrored - meaning there was a second equal pump sitting right next to this one, and they both were feeding the same circuit via the respective check valves. The second pump was bled in exactly the same manner - the plug on top of the suction manifold was removed until the oil was "gushing out" of the hole.
So, I started with the pilot pump, then the charge pressure pumps, and everything worked just fine, then I started the first of the two main motors and checked for the presence of flow at the outlet of the vane pump, which was easy to do because the pilot and the charge pump circuits had raised the oil temperature and I could feel the line warming up with my hand and, of course, literally see the "flow appearing" through my FLIR camera. Then I did some adjustments on the closed-loop pump - you know - the usual stuff that needs to be done. And then, finally, I turned the motor of the second vane pump on. At that point, I was pretty relaxed - it was the last motor/pump group to check, and so far, everything was going problem-free. So I started the motor, casually walked up to the HPU - put my hand on the pressure hose of the vane pump, and... wait!.... That can't be right at all! (On a side note - putting my hands on stuff is an old habit of mine I'll never be able to get rid of). The body of the pump was warm, but the pressure hose was cold - which could mean only one thing - that freaking 193 cc pump, the suction line of which I had so diligently bled the air from a day before, did not prime! I held up my phone and took a thermal just to prove I wasn't daydreaming - and here are the shots - the IR and the respective visual:
The oil was pretty warm at this point, so there's no way I could have faked this picture - the pressure hose is appearing dark blue because it's still cold - no oil had passed through it yet (with pump turning at 1500 rpm!). Look at the size of the suction line, look at the placement of the plug that I used for bleeding off the air pocket. Look at how close the oil came to the pump - and yet it failed to self-prime because the pump had to build at least 20 bar to open the check valve and start feeding the circuit it was connected to. Note that the check valve was not directly at the pumps' outlet, but rather downstream a meter-long -20 hose, which, supposedly, could "absorb" some of the air cushion and let those vanes come in contact with enough oil to "begin the pumping action" - but despite all this - it did not prime! And I (finally) have a picture that shows that something like that is possible!
Stopping the motor, and connecting a short test hose to the Minimess testing point at the pump outlet let the air out - and then, when I restarted the motor, the pressure hose immediately heated up!
This case shows that in situations where you have pumps with compromised priming ability (either by the construction of the suction line or the peculiarity of the hydraulic circuit it is connected to, or both) - bleeding the air from the inlet alone is not enough. It is equally important to double (and triple) check if "the flow happened" after the prime mover was turned on. Also - in the case of vane pumps, which have a pretty much straight connection between the inlet and the outlet when they are not running (there's no centrifugal and hydrostatic force on the vanes) - bleeding the air from the pressure end is definitely a better bet (and something I should have done).
I got really lucky with this one. I caught it early, and vane pumps are pretty forgiving to such abuse. But I can totally see how this recommission could have been way more "entertaining" had it not been for my arguably unsafe habit of putting my hands on hoses.
So yeah - here are three more "oil-change recommissioning" bullet points: