This application note presents data recently generated in reproducibility and repeatability tests performed on five LaserNet Fines (LNF) Wear Particle Classifier and Particle Counter instruments. The test was conducted with Medium Test Dust dispersed in mineral oil. Procedure –Sample Preparation A master oil sample was prepared by mixing Medium Test Dust with approximately 700 ml of purified mineral oil with a viscosity of 63 cSt at 40°C. The mineral oil was initially analyzed on the LaserNet Fines to check for particulate contamination and it was found to be very clean with a 13/11/6 ISO code. It contained only 48 particles per ml greater than 4 µm.
Experimental Conditions for Reproducibility and Repeatability Tests of the LaserNet Fines (LNF)
A one U.S. quart (950 ml) container was used to prepare the master sample. Approximately ¼ of the mineral oil was poured out of the container before the Medium Test Dust was added. The oil was removed to allow the contents to be easily shaken so that the dust would be well dispersed throughout the approximately 700 ml of mineral oil remaining in the container. The master sample with the Medium Test Dust was then used to prepare five identical samples. Five 125 ml plastic wide mouth sample bottles were opened and lined up on a lab bench. Immediately after vigorously shaking the master sample, about 10 to 12 ml of master sample was poured into the first plastic sample bottle, then the second, then the third, etc. As soon as the first 10 or 12 ml was poured into each of the 5 sample bottles, the process was repeated 7 or 8 times until each bottle was approximately ¾ full. The purpose of this procedure was to make the particle population in each of the 5 bottles as similar as possible.
LaserNet Fines (LNF) Analysis of Custom Samples
The five LNF’s chosen for this test were a combination of old and new instruments. The oldest unit, S/N 0021, was manufactured approximately 6 years ago and S/N 0226 and S/N 0239 were brand new instruments removed from stock for the purpose of the test. The other two instruments, S/N 0079 and S/N 0186 have been in use for multiple years. To verify proper operation, a check fluid was run through each of the five LNF’s. The LNF check fluid used is a PartiStan, Medium Test Dust in Hydraulic Fluid; the commercial version of NIST Standard Reference Material 28062. The reproducibility testing was conducted in such a way as to randomize the effects of the sample bottles as well as the order from which the samples were withdrawn from each bottle. Each of the 5 sample bottles was run once on each instrument.
Each instrument sampled one of the bottles first from one of the five bottles, each instrument sampled one of the bottles second from one of the five bottles, each instrument sampled one of the bottles third from one of the five bottles, etc., until five samples were run from each of the five bottles.
Each time, before inserting the LNF sipper tube into the sample bottle to pull sample into the LNF for measurement, the sample bottle was vigorously hand-shaken to thoroughly disperse the particles it contained. Then air bubbles were removed by placing the bottle in an ultra-sonic bath with a power density > 4000 W/m2 for two minutes. The sample bottle was then brought promptly to one of the five LNF’s to begin the measurement procedure.
Analytical Results from Reproducibility and Repeatability Tests
The analytical results from the five samples as analyzed on the five LNF’s are shown in Table 1 Reproducibility and Table 2, Repeatability. Reproducibility for this test is defined as test results obtained with the same test method on identical samples and on different instruments. Table 1 shows the order in which the samples were run, from which bottle each sample was withdrawn, on which instrument the sample was run and the corresponding analytical result. The reproducibility for the 25 runs is summarized as an average, standard deviation and relative standard deviation for each particle size range.
Calculating Percentage Relative Standard Deviation (RSD)
The percent RSD (Relative Standard Deviation) is 3.2% for particles greater than 4µm. The counts are for particles with an equivalent circular diameter greater than the stated size. The LNF software converts the pixel image of each particle into an area by filling in any transparent interior pixels. An equivalent circular diameter is then calculated from the resulting area. Repeatability is defined as test results obtained with the same test method on identical samplesm and the same instrument. Table 2 shows the repeatability for individual LNF instruments. As one would expect, repeatability on the same instrument is better than reproducibility on multiple instruments, and under these conditions, the percent RSD (Relative Standard Deviation) is an average of 2.7% for particles greater than 4 µm.
The reproducibility and repeatability of the LaserNet Fines is very good on samples with particles of varying size ranges. The relative standard deviation is excellent when there are large numbers of particles in a size range, but, as expected, becomes statistically meaningless for large particles since there are so few present. As with most analytical instruments, repeatability on the same instrument is better than reproducibility on multiple instruments
This information has been sourced, reviewed and adapted from materials provided by AMETEK Spectro Scientific.
For more information on this source, please visit AMETEK Spectro Scientific.