Today I continue the "investigation" of the A10VO100 pump that dumps about 20 liters per minute into the casing even after the complete replacement of the rotary group.
Two weeks ago we checked if the excessive leakage could be caused by the wear of the bias servo piston, and saw that this was definitely not the case, which left us with only one suspect - the DFR control valve:
From what I've been seeing and hearing - evaluation of the wear of such controls often gives rise to (heated) debates between representatives of the two industrial overhauling parties - the "replacers" and the "keepers".
Indeed, evaluating a tiny spool/sleeve assembly "by eye" is a difficult task. Let us look at this particular controller, and see if the overhauler made the right choice of keeping it:
Ok, so the spools do look pretty scored up. But the sleeves, on the other hand, don't look that bad, and when you try to "wiggle" the spoole in their place you don't detect anything suggestive of excessive play. So, the question then is - since a new control valve is several hundred bucks a pop - can you still use a controller in this shape at least for another production season (whatever amount of hours this may be) and, if you decide to "keep" it, how much would such wear be adding to the internal leakage of the pump?
It is obvious that scores mean wear, and wear means leakage, but would it be an extra liter per minute? Two liters per minute? Maybe three? Three doesn't seem to be too much of an add-on to our case flow, does it? I can live with three.
Well, my friends, I have the perfect answer to this question:
If you can test it - then test it and make up your mind after the test, if you can't test it - replace it and be done with it! Yes, I am a "replacer". But I am also a "tester". So, let me show you the test results for this controller and then compare it to a brand-new one, to see how much of an internal leakage difference we will detect. Here's the test setup:
I am using a custom-made base plate to connect the controller to the test bench (takes a piece of scrapped rod, basic shop tools, a pair of hands, and a couple of hours to make one). The P port is connected to the regulated pressure source, the T drains to the tank through a flow meter, and I put one pressure sensor to read the servo pressure and another to read the P.
The LS spool is locked by screwing the adjusting screw all the way in, and I decided to first see how the leakage flow behaves with the PC adjustment set to about 200 bar, and then see how it behaves when the spool is locked by the adjusting screw.
By the way, you can see that I prepared a sheet of paper and a pen before the test - a minor detail but it is very important in my opinion, and I always advise techs to do this. Having a sheet of paper in your hand with the parameters laid out in a table makes sure that you have a plan and intend to follow it. (A pen is a tech's best friend, remember?)
And now, here are the test results (pressures in bar, flows in l/min, Vg46 oil at 40-45 Cº) :
|Pressure||Servo P (200 bar set.)||T flow (200 bar set.)||Servo P (spool locked)||T flow (spool locked)|
How about them liters per minute?! As you can see, with the spools locked, the controller leaks a whopping 17 l/min at 280 bar! This is absolutely unacceptable. This and the "parasitic" servo-pressure increase. And check out these thermals - I wish you could see it light up in infrared, with the temperature constantly climbing and a bright white spot, showing the path of the fluid, revealing itself in front of your eyes in real-time:
No wonder - 17 l/min at 280 bar is an 8 kW heater!
Now, here's how a brand new controller, set at 280 bar, fairs:
|Pressure||Servo P||T flow|
Nothing is notorious here, virtually zero leakage (and also virtually zero increase of the servo pressure up until the PC setting), with some flow appearing when the PC spool shifts, which is normal, because there's a damping orifice that connects the servo to tank line.