Monitoring used oil samples from Plant and Equipment by WDA (Wear Debris Analysis) and Spectrometric Analysis ICP (Inductively Coupled Plasma) or RDE (Rotating Disc Electrode) used oil Spectrometric Analysis.

ICP (Inductively Coupled Plasma) Spectrometric Analysis oil testing works by injecting a solvent reduced spray of the oil sample into the chamber with usually inert Argon gas that has been ionised by passing the gas through a high energy High Frequency Radio Coil.
This excites the Argon gas into a high temperature 8,500°C Plasma (Ar+ plus electrons).

As the sample passes through the high energy of the Ionised Argon plasma this excites the electrons from the differing elements in the sample and as the electrons drop back to a lower energy position they emit light that is characteristic for each element. The emitted light is measured by the detector providing an accurate rendition of the elements in the sample as parts per million i.e.: Fe 35 ppm, Cu 60 ppm.

RDE (Rotating Disc Electrode) Spectrometric Analysis oil testing works by loading a porous carbon disc with a solvent reduced portion of the used oil sample and when being analysed the disc is rotated under a high voltage discharge arc that induces a plasma like state vaporising the oil sample emitting each element contained characteristic spectrum giving an accurate count of the elements vaporised away in a plasma state.

Both ICP & RDE Spectrographic Analyses are size limited by the energy level of the analysis system with 5-10 µm being the largest sized particle that Spectrographic Analysis can vaporise and thus sensitive to so both ICP and RDE Spectrographic Elemental Analysis are size limited to particles less than 10 µm with larger particles appearing basically “invisible” to both these oil analysis systems.

If internal combustion engines are being monitored where the internal combustion engine applications are well filtered the wear mode will trend producing smaller sized particles first enabling Spectrographic Analysis Oil Testing to detect and trend abnormal wear.
With Gear Drive and Transmission oil samples the wear mode is different to engines as when gear drives and transmissions commence to wear they often produce large wear debris >150µm possessing far too much mass for the ICP or RDE Spectrographic Analysis to vaporise causing this larger size steel wear debris also to be invisible (undetectable) using spectrographic analysis.

The following examples indicate the limitations of ICP & RDE Spectrographic Analysis when monitoring Transmissions, Gear Drive and Hydraulic Application using spectrographic analysis Testing only.

These two images @ 500X indicate the particle size detection limits of ICP & RDE Spectrographic Analysis Oil Testing when monitoring Drives and Transmissions.

Small Fatigue/Normal Rubbing Wear Particles @ 500X Sized 2-15µm

The Left Hand image’s wear particles sized at 6-8 µm would just be detectable by ICP as it would be able to vapourise and register these particles. The Right Hand image’s larger particles would only have the edges burnt off with the bulk of the particle remain intact and undetected.

And the seriousness of these three abnormal wear modes below could remain undetected for up to 6 months before the generated wear debris becomes “Milled” & broken own small enough by the action of the gears and bearings for the ICP Analysis testing to vapourise and detect.

Steel Cutting Particles @ 500X sized 5-65µm Steel Sliding Wear Particle @ 500X sized 115µm Steel Fatigue Chunk Wear Particles @ 500x sized 5-45µm

*See Wear Debris Analysis on this site menu for information on the analysis process.

The following tests have two oil samples taken from an Industrial site.
Spectrographic Analysis and Wear Debris Analysis oil test results have been performed on both to allow comparison of the information provided by each test.

This sample is from in the field from an Oxide Injection Pump, the sample has been analysed spectrographically and by Wear Debris Analysis sample with results set out below.

ICP Analysis Oil Test Results in PPMNote how the Above ICP Analysis are all reasonably normal but when the sample is checked by microscope with WDA we can show there has been a bearing spall and the steel fatigue wear debris should be removed
Because the broken bearing particles are all larger than 8µm ICP Spectrographic analysis testing cannot detect the larger particles. 

200x Sized 15-30µm

Here is the application again with a Cycle Water Turbine

ICP Analysis Oil Test Results in PPMNote again how the above ICP Spectrographic analysis are all reasonable OK, but when the sample is checked by a microscope as with WDA, we can show there is heavy three body abrasive wear occurring and this bronze wear debris should be removed.
These large bronze wear particles cannot be detected by spectrographic analysis, and do not show up in the above results (which would be a spike in Cu). 

200x Sized 15-200µm

 

Accuracy and Immediacy difference of Wear Debris Analysis to ICP Spectroscopy

The difficulty for Industries being monitored with only SE or ICP Spectroscopy Analysis is the size limitation of this type of analysis is not completely understood and can give false alarms and/or allow failures to occur without any warnings being evident from the results of these types of analysis.

However when a Wear Debris Analysis is completed on an oil, the results indicate immediately whether the drive is wearing in a normal or abnormal wear mode.
*See Wear Debris Analysis on this site menu for information on the analysis process.

 

Large Excavator Right Hand V12 Engine Failure

This failure was unfortunately in a large V12, the ICP is designed to monitor this type of application but in this instance there was not enough evidence from the results for the lubrication technician to get the machine shut down and checked, we have included our conclusion for the sample.

The ICP results are below, note how the problem disappeared in the 190705 analysis.

The damage occurred in the 050705 & 190605 samples, next sample 190705 was more towards a normal wear mode so they relaxed a little as no one “saw” the wear debris, however the Pb elevated in the 280705 sample and the engine failed with a broken con rod 2 days after the sample was taken.

Note the low level of Si in the lab results, there was a crack in the induction system allowing fine hard silica crystals into the engine, the crystals were too big for the Spectrographic Analysis to see giving no warning. We retrieved the following contamination from the engine oil filter. By completing a WDA on the engine oil filter media when the first increase occurred on the 19th June this contamination would have been seen and the induction system damage hopefully discovered before the failure could occur.

Magnetic Separation by Ferrogram

Image @ 1000x 

Silica Crystals Fluorescing under the polarised light. Sized 0.5-15µm

3µm Filter

Image @ 500x 

Silica Crystals Fluorescing under the polarised light. Sized 0.5-15µm

3µm Filter

 Image @ 500x 

Silica Crystals Fluorescing under the polarised light. Sized 0.5-15µm

11µm Filter

 Image @ 500x 

Close Up of Heat Coloured Bearing Debris deposited on the filter

11µm Filter

 Image @ 500x 

Close up of Heat Coloured Bearing Debris deposited on the filter

Note the hard crystalline particles cut the bearing surface up.

This mini report is what Condition Monitoring is about, find and remove any abnormal wear that shortens the life of the machinery.
With Gear Drives, Transmissions and Hydraulics both the ICP and RDE Spectrographic Elemental Analysis can be of use once the plant is running normally across the board, but while we are realigning and improving plant reliability the WDA method is far more accurate and will pull target cleanness back into line 4-5 times faster with far less unnecessary work than can be completed only using ICP or RDE Spectrograph results.
Spectrographic Elemental Analysis without WDA does not work efficiently as there is no positive feedback for staff leading to recommendations being ignored wasting the time and money spent oil testing as maintenance staff will often be paralysed watching the results without enough information being present in the reports for staff to justify changing procedures to improve reliability.

Rob Simmonds
Rob Simmonds
Reliability Manager R&T