For more than 40 years, spectrometric oil analysis has been applied as a routine and cost-effective condition monitoring technique. It is used to determine the elemental concentration in parts per million of wear metals, contaminants and additives in a used oil sample.
Figure 1. RFS Disc Electrode Filtering Oil Through Its
Circumference
With the knowledge of the wear metal and contaminant concentrations of the machine or engine being monitored, this technique may be used to determine if a sampled machine is operating properly.
It is a fact that spectrometric oil analysis detection efficiency decreases as the wear particle size increases. This partial limitation is not a significant issue with reciprocating engines that produce mostly small particles, but can be a problem in fatigue failures of rolling element bearings, such as those in military gas turbines that generate large particles at the outset of failure without generating small particles.
In addition, some wear modes such as spalling, severe sliding wear and cutting wear produce large particles that may go undetected by standard spectroscopy methods. The extent of this problem varies by machine and spectrometer type.
Atomic absorption spectroscopy (AAS) and inductively coupled plasma (ICP) spectrometers suffer from particle detection inadequacies. Rotating disc electrode (RDE) spectrometers are responsive to larger particles, but the upper limit has traditionally been approximately 10 µm. Ferrography and X-ray analyzers address this limitation and are capable of identifying large wear particles as a complementary technique.
Carbon disc electrodes used in RDE spectrometers are porous and can be used as a filter to trap particles in oil samples, which is an advantage for the RFS method. Early versions of the RFS fixture used standard disc electrodes with some success and acceptable sensitivity. However, standard disc electrode geometry is a slow process, and lack of manufacturing control of disc porosity resulted in poor repeatability.
Due to its inherent advantages, the RFS concept was recently revisited with the intent of improving performance. The updated RFS technique uses custom ring-shaped graphite electrodes and a semiautomatic process to improve sensitivity and repeatability and to reduce sample preparation time.
This article describes the RFS technique’s theory of operation, the evolution of the concept based on experience and how it is applied in commercial used oil analysis laboratories.
Theory of RFS
A microscopic inspection of the carbon disc electrodes used in rotating disc electrode (RDE) spectrometers will reveal they are porous. Rotrode filter spectroscopy (RFS) makes use of this fact, and a funnel-based fixture is used to clamp the discs; therefore used oil samples can be drawn through the outer circumference of the disc when a vacuum is applied (Figure 1). The filtering process captures the particles in the oil.
The oil is then washed away with solvent and the disc is allowed to dry. The particles are left on the outer circumference of the disc electrode so they are vaporized and can be detected on the RDE spectrometer.
The RFS technique is used as a comparative method due to the unavailability of oil standards with known gravimetric concentration of particles for each element measured by the spectrometer. In practice, a used oil sample is first analyzed using the standard RDE technique, which provides an analysis of dissolved and small wear particles. A second analysis of the same sample using the RFS technique detects large particles.
The two analyses indicate the wear particle size distribution in the sample. A sudden presence of large wear particles will not be seen by conventional analysis alone. However, large wear particles will be evident through the RFS analysis.
The RFS Fixture
The original RFS fixture to prepare electrodes for the analysis of large particles was developed in the early 1990s (Figure 2). The fixture used standard disc electrodes that were clamped in the funnels to filter and capture large particles. The fixture’s five stations could process five samples simultaneously. A vacuum pump was used to pull the sample and then filter it through the electrode.
A manual rinse with heptane was used to wash away any remaining lubricant. Sample preparation time of an electrode for RFS analysis varied with the viscosity and contamination of the oil sample. It could be as short as four or five minutes for relatively clean used oil samples. Such samples could come from turbines, electric motor bearings and hydraulic systems. Engine oil samples with high soot levels require the longest filtration times, sometimes half an hour or more.
Figure 2. Early RFS Fixture
Unfortunately, standard disc electrodes are not manufactured with consistent porosities, which affects the capture efficiency of the particles. This variability hurts repeatability, making data trends more difficult to establish. Furthermore, long sample preparation times for highly sooted samples from diesel engines were not practical for some laboratories.
These factors led to a new system that offered additional improvements. The updated system is known as automated rotrode filter spectroscopy (A-RFS) (Figure 3). Repeatability was improved through the use of controlled electrodes, automation and the use of a vacuum/pressure pump to reduce sample preparation times.
Figure 3. A-RFS System
The A-RFS system is a semiautomated sample preparation instrument with a five-station fixture of funnels. It works on the same principle as the original RFS system. However, it uses a special electrode designed for more consistent porosity specifications, has an automated cleaning cycle, and filtration times are reduced by the application of a vacuum/pressure pump to both pull and push the used oil sample through the electrode.
The new A-RFS electrode is custom-manufactured to consistent porosity specifications. Thinner walls improve repeatability and reduce filtration times. The dimensions of the standard disc electrode used for conventional RDE analysis and the new RFS electrode are shown in Figure 4.
Figure 4. Conventional Electrode and RFS Electrode
During routine operation, the A-RFS disc electrodes are mounted on an electrode clamp assembly for installation in one of the five preparation stations. The clamps are numbered so they can easily be identified and associated with their matching oil sample.
The sample and a fresh disc are introduced and with the start of the automated process, a vacuum pump creates a pressure drop of approximately five atmospheres across the disc electrode, which causes the sample to flow through the disc electrode. The filtration process captures large wear particles on the surface. A sensor determines when the sample has been filtered through the electrode and automatically starts the cleaning cycle followed by a drying process.
The RFS preparation process is complete when all of the oil has passed through the disc electrode, residual oil has been washed away and the electrode is dried. The operator then removes the electrode clamp assembly and installs the shaft, complete with the electrode in the spectrometer for analysis (Figure 5c).
Figure 5. Preparation Process of the A-RFS Electrode
RFS Data Interpretation
In order to provide the basis for benchmarking RFS calibration and to assure data consistency, spectrometers are calibrated with certified commercial or military metallo-organic standards. This ensures the spectrometer has the same response over time and permits trending of wear metals and contaminant particles.
RFS is a powerful analytical tool because it provides additional information on actual wear and contaminant particles in an oil sample. In routine operation, laboratories perform two analyses on a used oil sample. The first is conventional oil analysis which provides analysis of dissolved and small particles in the sample. RFS, the second analysis, reveals large particles and their elemental composition. Together, trends of the two analyses can be used to provide a clearer idea of machine condition.
A-RFS Field Experience
RFS has proven to be a valuable tool to military oil analysis programs; however, this was not always the case. RFS initiated in industrial applications and continues to provide important information for in-house and commercial laboratories.
Case History
The following is a typical case history from a Northeastern power company where a condensate vacuum pump was sampled. The commercial laboratory provided oil analysis services for lubricant physical properties and metals including RFS for larger particles. Although the laboratory supplied data for 20 wear metals, contaminants and additives, a trend was observed only for iron and silicon (Figure 6).
Figure 6. Spectrometric Analysis Trends
for Normal (Fine) and RFS (Coarse) Analyses
(click here to enlarge)
Normal spectrometric analysis (fine) did not show significant wear. In most industrial systems, such as this pump, concentrations of wear metals will fluctuate in accordance with oil added, and can vary by as much as 10 ppm. Therefore, the wear, as indicated by normal spectrometric analysis, did not exceed expected limits and no maintenance recommendations were made. On the other hand, the RFS analyses, as indicated by “coarse”, showed a clear trend resulting in maintenance recommendations (Figure 8).
Individually, the silicon readings were not cause for alarm. However, when combined with RFS data for iron, a red flag appeared to the laboratory. The RFS trend for iron jumped enough that a severe alarm was issued after the January 9, 2003 sample. The maintenance response was to “bleed and feed”, that is, drain some lubricant and top it off.
The next two RFS analyses showed a corresponding reduction in the iron. However, the May 22, 2003 sample indicated there was still a problem and a severe alarm was again generated. This time the maintenance personnel opened up the pump and found a loose-fitting bearing that was fretting on the bearing housing, bearing and shaft (Figure 7). The RFS data also prompted the laboratory to prepare a ferrogram of this sample. Ferrographic analysis verified the presence of large particles and the occurrence of cutting wear (Figure 8).
Figure 7. Bearing Housing, Outer and Inner Race Fretting
Due to Loose Fit
Figure 8. Ferrogram Showing Cutting Wear
In this example, normal spectrometric analysis did not detect the large wear particles emanating from the bearing-fretting problem. The RFS analysis, however, did provide early warning to repair, clean-up and solve the problem before it could lead to a major failure with possible secondary damage.
RFS is an important enhancement to standard RDE analysis. It provides new information on the condition of the machine being monitored by expanding the particle detection capability of the RDE technique.
In actual field applications, rotrode filter spectroscopy (RFS) has been shown to provide important additional information about large wear particles, information that may be missed with conventional techniques. It is applied with existing instruments to provide two analyses on the same used oil sample. The first, using the standard RDE atomic emission technique, detects and quantifies dissolved and small wear metals and contaminants. The second analysis, through RFS, qualifies and semiquantifies larger particles.
RFS analysis is not used by itself, but as a second test that expands the particle detection capability of spectroscopy to provide a comprehensive tool that directs the laboratory to identify and solve a maintenance problem. RFS analysis can also be used as a screening tool to determine if additional tests, which are typically too time consuming or expensive to perform on all samples, should be conducted.
Editor’s Note:
A similar technical paper on this subject was published in The Journal of ASTM International (JAI). (www.astm.org)
References
- Sawyer, J. W. Sawyer’s Turbomachinery Maintenance Handbook. Turbomachinery International Publications, Conn. 1980, p. 8-5.
- Lukas, M. and Giering, L. P. “The Effects of Metal Particle Size in the Analysis for Wear Metals Using the Rotating Disc Atomic Emission Technique.” International Symposium on Oil Analysis, Erding, Germany. July 1978.
- Anderson, D.P. “Wear Particle Atlas” (Revised). Prepared for Advanced Technology Office, Naval Air Engineering Center, New Jersey. June 1982.
- Rhine, W. E., Saba, C. S. and Kaufman, R. E. “Metal Particle Detection Capabilities of Rotating-Disc Emission Spectrometers.” Lubrication Engineering, Vol. 42, #12, p 755. 1986.
- Anderson, D.P. and Lukas, M. “Rotrode Filter Spectroscopy - A Method for Multi-Elemental Analysis of Particles in Used Oil Samples.” Proceedings of the International Condition Monitoring Conference. Pensacola, Fla. November 1992.
- Humphrey, G.R. “Filter Debris Analysis by Energy Dispersive X-ray Fluorescence Applied to J52 Engines.” Joint Oil Analysis Program International Condition Monitoring Conference, Pensacola, Fla. April 2004.
- Yurko, R.J. “Applying Rotrode Filter Spectroscopy (RFS) and LaserNet Fines (LNF) Technology to Detect Large Wear Particles of M50 Alloy from J52 Engine Bearing Failures.” Joint Oil Analysis Program International Condition Monitoring Conference. Pensacola, Fla. April 2004.
- Case History, Compliments of National Tribology Services, Peabody, Mass.
