For situations where spectrochemical oil analysis alone may not provide the most complete machinery wear profile, the results from specialized microscopic and instrument-based tests that focus on particle and wear debris may prove to be a critical resource.
Like with all oil analysis tests, knowing the relative strengths and weaknesses of each wear debris analysis test will help oil analysis practitioners choose when and how to use these tests in conjunction with their existing oil analysis program to obtain maximum benefits from the insights they provide.
A posting on Noria’s Message Boards perfectly summarizes the oil analysis practitioners’ point of view for the topic at hand. Jeff St. Amand - then with American Electric Power - wrote, “In selecting tests for wear metals, is there any known rule of thumb for when to switch between atomic emission spectrography (AES) and rotrode filter spectroscopy (RFS) (or another method) based upon direct reading (DR) ferrography DL/DS proportions, ISO cleanliness levels or other criteria to obtain a more accurate idea of the extent of wear debris? Are there other tests that should be considered based upon increasing levels of particles greater than the 3 microns to 8 microns limit of AES?”
What Jeff may not have known was that if he already had a robust oil analysis program, he had the tool he needed to make these decisions – the particle count.
Oil analysis has been called, with varying degrees of accuracy, a blood test for fluid-lubricated mechanical systems. It’s interesting to think about how many blood tests depend upon an analysis of solid particulates. For instance, any blood test involving cells - red or white blood cells, T-cells (immune system) or platelets (clotting) - depends on solid particle analysis.
As in oil analysis particle counting, these tests are predominantly microscopic identifications by type and count, but this is where the analogy breaks down. Unlike the oil-borne solids that interest oil analysis practitioners, blood-borne solids tend to be found in a narrow particle size range - only about 8 microns to 15 microns - due to the minimum blood vessel size in the human body.
Oil analysis laboratories and practitioners must consider a much wider range of solid particle sizes - submicron (<1 micron) to more than 1,000 microns - when diagnosing wear and its modes.
Complicating this effort, both rotating disc electrode (RDE) and inductively coupled plasma (ICP) atomic emission spectrochemical analysis methods, long the backbone of conventional oil analysis, have well- documented upper limitations on the particle size(s) they can effectively measure. So is particle counting a way to solve spectrochemical oil analysis weakness?
One of the “givens” in mechanical system operation is that system parts will wear abnormally, and the wear particle quantity and size progress from smaller to larger. This size change is related to stress increases as loaded surfaces depart from their original shapes and clearances, the effect of higher temperatures on lubricant films and alloy structures, and the cascade effect of wear as already worn parts release successively larger particles which act to dent and abrade these same damaged surfaces. This accelerating deterioration in mechanical condition is a primary driver for the early diagnosis of abnormal wear.
The goal of oil analysis should be to specify and direct additional testing to supplement spectrochemical analysis with a tool easy for any oil analysis practitioner to obtain and use. The particle count is such a tool, and it’s a powerful tool indeed. The particle count is a familiar, well-standardized and almost universally available method, already part of many oil analysis packages. Particle counting possesses an ability to look straight across the multiple particle size ranges symptomatic of abnormal wear.
Also, it is not limited to a particular element, and shifts in the counts by range can be correlated to increasing wear. Simply put, using the particle count to compare the particle distribution to the sensitivity ranges of the various particle analysis technologies makes it easier to choose the most effective way to supplement conventional oil analysis.
Whenever particle counts change, a proactive approach will consider all possible causes for the phenomenon. Figure 1 illustrates how far particles in a plant environment can travel from their source and how long they can linger to enter seals, vents, breathers and other vulnerable openings in machinery.
Figure 1. Particle Travel and Suspension
Time from Point of Origin
A crucial step in expanding any wear-detecting vision with particle counting involves understanding the general size ranges into which typical wear and contaminant particles fall. The forces and conditions that generate particles form them into typical particle configurations and sizes. These size ranges are primarily influenced by material type, wear mode and severity progression. (Author’s note: The size ranges are secondarily influenced by a series of engineering and lubrication considerations beyond the scope of this article.)
The important point is that as wear worsens, the average particle size increases, so any proactive approach to wear monitoring must detect this size shift immediately. Figure 2 illustrates some basic information on this topic, and provides perspective by relating the wear particles to the real-world objects of similar size.
This illustration is not an exhaustive listing of wear forms or types, but it shows that most wear particles progress from smaller to larger as wear continues. The size ranges illustrated are those associated with wear or contamination that has reached a demonstrably serious phase, one where proactive intervention will be well justified.
There is more than one route available to the oil analysis practitioner who is attempting to investigate wear modes in the larger, more critical size ranges beyond conventional AE/ICP spectrochemical methods. He or she can opt for quantitative instrument-based tests that provide numerical results, or choose visual/microscopic analyses that are qualitative by nature.
Instrumented methods have historically focused on iron when attempting to fill in this so-called “blind spot.” This is because iron is typically the most important single element in mechanical system wear analysis and researchers and instrument designers alike know that ferrous alloys are subject to manipulation with magnetism. Ferromagnetic approaches underlie some, but not all, of the most widely accessible lab and field analytical techniques less influenced by - or deliberately designed to avoid - particle size limitations.
However, any generally available laboratory or field procedure that points to the presence of or measures large wear particles is a tool with which the oil practitioner should become familiar. Figure 3 is a graphic survey (though not an exhaustive list) of off-line, generally encountered technologies. These tests should be accessible to any on-site or off-site laboratory when needed.
Figure 3. Particle Analysis Technologies and Size Sensitivities
In Figure 3, the five-range ISO 11171 particle count runs through the heart of the ranges of most of the large-particle detection technologies. (The asterisk after the 70 µm range references the fact that this is a 70 µm and greater range; some counter sensors go up to 400 µm).
Table 1 provides tips and notes that will help an oil analysis practitioner define which monitoring or investigative technology will work best for his or her situation.
The following is an example of how the information presented in Table 1 might be used:
A lab has been trending particle counts on a particular screw compressor and the third ISO contamination class code (which references total particles greater than 14 µm) has increased from ISO range code 14 to ISO range code 16 since the last sample. By reviewing the count, it is discovered that the increase is based primarily in count range five (>38 µm). This particle size range may be associated with serious wear or contamination.
In addition, Figure 2 indicates that sliding wear is a possible wear product that falls into this size range. After making arrangements to reduce load on the compressor pending further investigation, the information presented in Figure 3 and Table 1 can be used to determine the most appropriate investigative technologies for that particle size range, and to choose the best analytical ferrography tool.
Once the report returns, indicating the presence of iron alloy wear fragments potentially associated with misalignment of the compressor screws, the maintenance team can take action and the lab can then adjust the testing slate to include PQ (ferrous wear), which will provide added insight on whether or not larger iron-based particles are present.
A simple test such as particle counting can be used to help plan a strategy for overcoming the analytical limitations of rotating disc and ICP emission spectrometers. By using the information presented here and any specifications available on recommended cleanliness levels, an oil analysis practitioner should be able to make informed decisions on the need for additional testing and which tests to select.
When using the particle count as an indicator of abnormal wear, it is helpful to consider the typical time and size profiles within which wear particles are most commonly generated during a system’s normal operating lifespan. Smaller particles are less likely to be filtered and are more susceptible to residual concentrations from service to service than larger particles. These profiles will help the oil analysis practitioner project equilibrium particle concentration baselines against which he or she can evaluate sudden increases in particle count.
As a rule of thumb (and it’s just that, not a rigid guideline), if a particle count’s ISO contamination code was between 10 and 20 and changes by two ISO classes (for example from ISO 13 to ISO 15 or ISO 19 to ISO 21), or if it was above an ISO code of 20 and changes by one ISO code number (for example ISO 21 to ISO 22), it’s time to investigate the reason for the change.
Keep two things in mind: overall cleanliness guidelines and targets, and that low ISO codes such as 10 or 11, are proportional only to 10 or 20 actual particles per milliliter of sample. The cause of the change may be due to maintenance, a known performance problem, a charge of new unfiltered oil, or an unseen wear problem reaching for your bottom-line profitability.
Anderson, D. (1982). Wear Particle Atlas (revised edition). Report NAEC-92-163.
Bensch, Dr. L.E. (1991). A Modern Review of Field Contamination Levels Based On Analyses of 25,000 Samples.
Ding, Dr. J.(2003, September-October). Determining Fatigue Wear Using Wear Particle Analysis Tools. Practicing Oil Analysis.
Dory, S.H. and Hansen, T. (2003, September-October). Magnetic Plug Inspection Enhances Condition-Based Maintenance. Practicing Oil Analysis.
Hunt, T. and Roylance, B. (1999). The Wear Debris Monitoring Handbook.
Omole, S. (2001, November-December). New Dimension to Failure Analysis, Using XRF Technology. Practicing Oil Analysis.
Rhine, W.E., Saba, C.S. and Kaufman, R.E. (1986, Vol. 42, #12). Metal Particle Detection Capabilities of Rotating-Disc Emission Spectrometers. Lubrication Engineering.