Insane Hydraulics

Site theme image

The Most Embarrassing Mistake of My Career as a Hydraulic Technician

This is the story of how I came within a hair's breadth of experiencing an avalanche of consequences of the most embarrassing mistake of my career as a hydraulic technician, and it has everything to do with the hydraulic system that I was adjusting with lasers in my last post.

I've been blogging for a long time so it's quite possible that there is an article somewhere on this website that already claims to be a record of my most embarrassing mistake, but I never stop surprising myself with unexpected achievements on the blunder front, which, I guess, can be considered as an alternative way of a subscribing to the philosophy of "constant growth".

I don't mind sharing my errors, though, no matter how humiliating they may be - if at least one single person in this world is saved from making a mistake while working on a hydraulic system because he/she read about it in the IH blog - I am super happy!

So, let me set up the stage real quick for those who haven't read my earlier blogs:

I am staining in front of a monster HPU, powered by two 500 kW electric motors, "brandishing" a tandem closed-loop 500cc pump each, and at least one of them appears to have a bad null.

The closed loop system drives two differential double-acting cylinders of a large back-fill pump, and it is quite complex, with more pumps and electric motors powering the rest of the hydraulic circuitry necessary for its operation, but for the context of this blog, all these details don't matter. All you need to know is this:

Simplified hydraulic schematic

The control system is equipped with an advanced service mode, which allows me to power the big motors one at a time (along with the auxiliary circuits) - which is very convenient because I can perform the checks "one pump per loop" at a time.

And here is my "game plan":

Important points:

Lowering the settings of the relief valves was necessary because the system was tripping the high-pressure alarm, set to 300 bar, as soon as the pumps would start, and I wanted to do the adjustments with as few starts/stops as possible (preferably in a single "go") because of the cycling limit of the star-triangle relay at those starting currents.

Making sure the cylinders were extended was necessary because closed loop systems that drive differential area cylinders have a peculiarity: neither the cylinder creep nor the fact that the pressures in the loop legs are unequal are reliable indicators of an off-center null. This is so, of course, because of the pressure intensification of a differential area cylinder (but it is still quite counterintuitive when you see it). Adjusting the null on such systems is only possible when you either lock the cylinders mechanically, cap them off, or, which is often the easiest thing to do - make sure that they are extended all the way. So the idea was to extend the cylinders (which is very easy to do manually with the EO2 control of the A4VSGs by simply moving the position feedback coil up or down) - then set the electro-hydraulic null, mark the correct neutral with the help of the laser pointer, power the system down and check if the "weakly" mechanical zero of the A4VSG "hits the right spot" (see the laser post for detail).

Cool! I am at the top of my game. I have the perfect game plan, I have the tools, I have the expertise, I am young and handsome... Wait... Cross out the last two. But still - pretty confident in what I am doing I proceed with the "expert intervention". I see that the cylinders are already extended (how lucky!), I start the M1 motor from the PLC touch screen, go up the stairs to the HPU room, and "initiate diagnostics".

And very soon I come to a very disturbing conclusion that something really wrong is happening to the pump. Something that should not be happening at all, and can only be happening if the pump has a major mechanical failure inside. I read the unbalanced pressures in the loop legs, I see the swash-plate move, and I see no changes to the pressures. No effect at all! Even at a pretty substantial swash plate angle!

So, I immediately draw images in my head of how it can be physically possible for a swash plate of an axial piston pump to be tilted a good 10 degrees without affecting the piston stroke of the rotary group... I look through the parts list of the pump. I can see that something like this is impossible and yet - I am staring at it! The system is running and there are uneven pressures in the loop legs, I move the swash plate back and forth - but I see no effect on the freaking pressures!

What do you do when you see that something bad is happening to a client's pump? You document the occurrence as best as you can and sound an alarm, and of course that's exactly what I did. I took pictures, I made a video, and I began making arrangements for an emergency meeting with the maintenance department because I just detected a serious failure that would condition the production of a whole mine and lead to a major expense! I was also imaging in my head how I would be confronting the OEM and Rexroth with the facts that I had just so thoroughly documented, and how the engineers from Rexroth would be surprised to see such a failure, and, possibly, praise my ingenuity and hard work!

But, luckily, the chief maintenance engineer was out of reach, which gave me a couple of extra minutes to walk around the HPU wondering what could be happening to the damned pump. My imagination was storming through all possible scenarios, and it occurred to me that since the pumps were connected in a criss-cross pattern, the off-centered pumps on the idle motor side could be rotating like motors and maybe, affect the system in a way I wasn't seeing yet, and I decided to check it out. I asked a colleague to peek into the hole in the bell housing (on the idle motor side) and see if the flexible coupling was rotating.

So, I start the M1, as usual, go upstairs, and the guy says: "Oh, it's spinning all right, and fast! It began moving as soon as you started the motor!" I am sorry.. what?! Once again my brain rushed through a thousand theories explaining what could be happening and then.... the simplest conclusion hit me like a freaking Niagara Fall:

Say... Which one is the M1 motor?

That's right! I started the M1, but I was adjusting the pump connected to the M2! Can you imagine the (very expensive) techs from Rexroth flying in to see me point to the pumps mounted on M2, start the M1, and then wonder why the adjustments don't work? That's a career-ending blunder right there! It's like amputating the wrong leg or giving a demolition crew the wrong house number. Boy, am I glad I caught this one in time!

- Have I seen similar systems before?

- Yes I have.

- Do I have a habit of always checking if a prime mover is working?

- I always do that (except for the times I don't).

- Was it hard to see if the electric motor shaft was actually turning?

- No, it was not. There was an opening in the bell housing with a clear view of the coupling, right in front of me.

- Why haven't I looked then?

- No excuse and no explanation. The term "brain fart" does come to mind, though.

So, there you have it, folks. I am hoping that this story brought a smile to your face. And I hope that you understand now why I always question opinions of even very experienced professionals - in the end, we are all human and therefore bound to blunder.