Reliable assessment of particle concentration is important for the condition monitoring of fluid power systems. Good filter efficiency and low contamination levels usually imply well working systems. But what should be regarded as low or acceptable levels? Historically bottle sampling has been the reference method for assessment of particle concentration. Recently more and more commercial equipment for on-line connection is available. Many times there is a difference between the two methods, so the question rises: Can we rely on these new on-line methods or is it the bottle sampling that is not good enough for modern, well filtrated fluid power systems?
Sending bottles to different laboratories very often gives different results. This is of course not satisfying. In Scandinavia there are several laboratories that perform particle counting of bottle samples from hydraulic and lubricating systems. Eight of these laboratories formed a group with the aim to produce results with better reproducibility. Good results were achieved on a round robin with MIL-H-5606 hydraulic fluid and standard Air Cleaner Fine Test Dust (ACFTD). Unfortunately similar round robin results from some systems in service with other fluids were completely different. The concentration was far higher compared with on-line measurements and the spread was very large.
The conclusion is that bottle sampling can be very accurate and reliable if the contaminants are standard test dust at high concentrations, with suitable fluids. Good applications can then be filter testing or automatic particle counter calibration. In ordinary fluid power systems, with good filtration, it seems to be very difficult to achieve relevant results.