Insane Hydraulics
Bold and audacious blog about fluid power
I've been using our Hydac FCU 2210-1 particle counter a lot, and I want to share one of my recent experiences because it provides enough data to conclude that sporadic oil cleanliness analysis is utterly useless.
So, I got a request to investigate a 4000-liter 500kW HPU, that was tripping the clogging alarm of one of the return filters, shortly after the element was replaced.
The HPU housed two Rexroth A4VSG-based closed loops (total displacement of 1710 cc!) and a single relatively small open loop in the pressure-source configuration. No need to go into more detail here.
So, how clean should the oil be in such a system? If you look up the Rexroth Oil Cleanliness Booklet, you'll find the following chart:
This means that, theoretically, anything better than the good old NAS7 (18/16/13-ish ISO) would be OK for that system.
I checked the oil with the Hydac FCU, and got the results oscillating between 16/13/08 and 18/16/12, so.... we're good?
Well, even though the numbers seem to be OK, we're not good here. Absolutely not. Why? Because a similar hydraulic unit, with exactly the same filtering system, was routinely logging values of 13/10/08 and better. So something was definitely not OK.
So, after a conference with the client, the contamination spike was attributed to a top-up with unfiltered oil, the system was stopped, and I used an external bank of filters to bring the counts down to 12/10/07. Then, when the system was recommissioned, I came back a week later to check the numbers and read values between 14/12/08 and 13/10/07. It was definitely better than the 18/16 I saw last time, but... it was still slightly worse than the consistent 13/10 (or less) that I was used to seeing in an almost identical system.
Well... I found out why the numbers were a bit high a couple of weeks later, when I was, once again, called to investigate the same HPU because it was "acting out". The diagnostic was quick and simple, and I discovered that three of the four high-speed cylinders that were fed by a pressure-source accumulator-based circuit had so much of an internal leakage that the pump couldn't "keep up", which was causing all sorts of troubles.
And when the cylinders were disassembled, we saw that the piston seals and the guide rings were gone, and there was serious damage to the cylinder walls and the pistons. Deep scores and all!
I flushed the circuit, cleaned the lines, flow-tested the pump and the rest of the components, the cylinders were repaired, and when the system was re-assembled and put back into service, can you guess what values I got from the FCU after a couple of days of full-speed operation? See for yourself:
Consistent 12/10/07! With the pumps operating at full speed, and auxiliary cooling circuits regularly kicking in. Now this is a result I like seeing. Rexroth engineers did a very good job designing this HPU, with no expense spared on filters - and the particle count is a solid testament to that!
Now, consider this - all of the numbers that I got are OK if you take them out of context. The 14/12 I saw the other day is perfectly fine. Even the 18/16. However, if I was 100% sure that the habitual baseline cleanliness level for that particular HPU is 12/10, and if the particle readings were being taken at least weekly, most likely the contamination spike would have been noticed before the cylinder's wear go to the point it got.
My point is - sporadic particle analysis of hydraulic oil is useless. It is useless, because the range of values that can be considered as "OEM-recommended or better" is huge, and therefore, without an established baseline for a particular hydraulic circuit, you can never tell if the particle counts you are reading are OK or not! The same goes for chemical analysis too, by the way
So, if the cost of a hydraulic system warrants oil analysis for the sake of preventive maintenance, it must be done regularly and frequently. Anything less than that is a waste of time and money.
Baselines rule preventive maintenance, you know...