Insane Hydraulics

Site theme image

Evaluating Leakage of a Rexroth A10VO100 Bias Servo-Cylinder

I want to share with you an interesting experiment.

I've been going through tons of Rexroth A10VOs lately (I imagine those of you who repair hydraulic pumps professionally will say something like "who doesn't?!"), and a doubt emerged the other day - I received an A10VO100 with a complaint about an excessive case drain flow (20 l/min), but when I opened it I saw that almost all of the parts looked new. When I inquired about the history of the unit, I was told that it had been overhauled very recently, and when the mechanic installed it on the machine, he ran some tests, and since one of the tests included the measurement of the case drain flow, he discovered the excessive value, and decided to sent the pump for a "check out".

The rotary group parts looked new, but the bias servo-piston had some pretty heavy marks from the bias spring, and the shoe surface was also visibly marked, which meant this part was reused::

Bias servo-piston shoe with wear marks
Bias servo-piston shoe with wear marks
Wear marks left by the bias spring

Now, when I find this wear in a recently overhauled unit, I know for a fact the repair was done "in a hurry" because even if you opt for reusing a worn-out servo-piston, you should at least lap the shoe surface (which takes very little time since the surface is tiny), but hey - I've done "hasty" repairs myself, so no blaming here. But I began to wonder:

How much does the bias servo-piston, which is always connected to the pump outlet, actually add to the case drain flow? I understand that this knowledge has no practical value, but I got curious - is it something like half a liter per minute? 100 cc per minute? Does the seemingly deep wear mark on the shoe surface add much to it? Damn it! Now I have to find out!

So, I got a brand new bias servo piston, then a slightly worn-out one from a different A10VO100, and compared the radial plays of the three (measured at the shoe head with the piston extended in full displacement position, which is about 50 mm). I got the following numbers:

Measuring radial play

The servo cylinder extends about 50 mm, and the liner part is about 48 mm long, so since I was measuring the play at the shoe-head, I would "guesstimate" the radial clearance of the assembly to be about 40% of the measured value, which puts the numbers at 0.04, 0.05 and 0.06 mm respectively.

Then I quickly devised this "futuristic-looking contraption" out of scrap metal:

The apparatus I devised to measure leakage of a servo-cylinder
The apparatus I devised to measure leakage of a servo-cylinder

And then I used it to see (literally - see) how much of the internal leakage a bias piston of a Rexroth A10VO100 pump produces.

Well, I can tell you one thing. I knew that the leakage would be small, but I never expected it to be that small! I was expecting the slightly worn-out servo cylinder to behave OK, but I also thought the one with the pretty noticeable wear pattern on the flat shoe face would be spraying oil all over the place. But this was not the case. The servo piston shoe and the matching swash-plate (bronze) shoe are not perfectly centered in my "test rig", and I did try rotating the piston to see if I could find a position in which it would leak more because of the wear pattern not matching the bronze shoe - but no! The damned thing would not leak more! Some measly drops of oil were all I got! There was a clear difference between the "slightly worn" and the "heavily worn" assemblies, but the leakage was always in the order of "drops per second".

Note that when I say "heavily worn" here I mean "the one that has the most wear". There were no deep scratches or anything that would suggest that the part needs immediate replacement. I actually tried my best to take pictures of the surface wear of the servo-piston rod but you can't see any details in them because the surface is so shiny.

A YouTube video is worth a thousand words, so I made a small video of the radial play measurements and the leakage test (please, don't mind the auto-focus of my phone - it has a mind of its own):