This article was first published in 2001.
All serious users of lubricating oils - from race teams to industry - use oil analysis to indicate the health of both the oil and the system that's being lubricated. Having an oil analysis performed is also quite cheap - but how are the results of that testing to be interpreted? Here a leading expert in oil analysis takes us through the laboratory techniques that are used to examine used oil, and guides us in how to interpret the results that result from the testing.
An oil analysis results in complex but vital information about both the health of the oil and of the engine that it is being used in. But many tests are so interrelated and mutually dependent that interpretation of the results is not always straightforward. Knowing this, it is easy to appreciate the impossibility of putting the whole process of diagnosis into a few lines - but we'll try anyway! Another problem is that some oil analysis results are apparently contradictory. For example, in an engine a sodium (Na) reading of 300 ppm (parts per million) may be ignored in one case, while 50 ppm may generate a warning in another.
Another point which must be emphasised is that trends are a better indicator than limits. Trends are established through regular sampling and analysis. By comparing the results of the latest sample against those of previous samples, it can be seen which readings have changed substantially from previous samples, and an appropriate diagnosis made accordingly. For example, in a differential, 300 ppm iron might be the norm for one type of operation, but on the same vehicle doing a different type of operation, and with a different period oil in use, 1500 ppm iron may well be acceptable. A pre-imposed limit of say 1000 ppm here would have excluded a perfectly good oil from further service.
When samples arrive for analysis they are grouped according to the following broad classes:
- Drivetrains (gear systems, such as manual gearboxes, differentials and industrial gearboxes)
- Transmissions (automatic transmissions)
- Compressors and turbines
There are also other smaller special classes, such as aircraft engines and refrigeration compressors.
Every sample gets four basic tests:
1. ICP spectroscopy
2. Particle quantification
3. Viscosity at 40°C
4. Water screening
This week we'll start off by looking at ICP spectroscopy.
Spectroscopy is the study of light (or more generally, electromagnetic radiation) and its interaction with matter. There are approximately 30 different types of spectroscopy. One type, inductively-coupled plasma (ICP) spectroscopy, measures light in the visible and ultraviolet regions of the spectrum. It is an atomic emission (AE) procedure whereby the diluted oil is passed through an argon gas plasma. The plasma is produced by induction and is maintained at a temperature of approximately 8000&°;C, hotter than the visible surface of the sun! In the upper region of the plasma, acquired energy is released as a result of the electronic transitions, and characteristic 'light' emissions occur. Different elements produce different frequencies or colours. The intensity of the light emitted is directly proportional to the concentration of the element.
ICP spectroscopy is used to measure the concentration of different elements in the oil.
The elements are divided into three broad categories:
- wear metals, such as iron from gears
- contaminants, such as lithium, which would indicate the presence of grease
- oil additives, like phosphorus which is found in extreme pressure and anti-wear additives.
Some elements can belong to more than one category. For example, silicon can be a component of wear debris (piston crown material), of the additive package (anti-foaming agents), and of contaminants (dirt). Only by looking at a complete set of results is it possible to predict the source of the particular element.
In certain cases oil labs use limits for contaminants. In the case of dirt, the lab at which I work generally observes the limits in the table below:
Silicon Contamination Limits
|Hydraulic / compressor / turbine
ICP spectroscopy is perhaps the most important and useful test in used oil analysis, but is does have its limitations. Perhaps the biggest drawback is the size limit of the particles it can see. Particles above approximately 5 to 8 microns in size do not get detected. Looking at an extreme example: processing a sample of oil with a solid ball-bearing ball sitting at the bottom would return an iron reading of zero! Clearly there is a lot of iron in the sample though...
But if that same ball were ground to a fine powder and the sample re-analysed, the iron reading would be very high. So perhaps it would be better to refine the definition of an ICP's function from measuring the concentration of different elements to that of measuring the concentration of elements found in particles of less than 5 to 8 microns in size.
While this limit does not affect the detection of most wear situations, there are times when it could be a problem. For instance, when a component begins to fail due to fatigue, the wear particles generated tend to be larger than normal (this process is called 'spalling'). These larger particles will not get detected by the ICP, so upon examining the trend, the iron level might be seen to be dropping, even though the component is actually in trouble. Given that this limitation exists, other tests need to be employed to provide an effective monitoring solution.
Common Elements in ICP
||Gears, roller bearings, cylinder/liners, shafts
||Roller bearings, piston rings
||Roller bearings, camshafts and followers, thrust washers, valvestems, valve guides
||Piston rings, additive, solid additive (Mo-di)
||Pistons, journal bearings, dirt
||Brass/bronze bushes, gears, thrust washers, oil cooler cores, internal coolant leaks
||Bronze bushes, washers and gears
||Journal bearings, grease, petrol contamination
||Silver solder, journal bearings (seldom)
||Dirt, grease, additive
||Internal coolant leaks, additive, sea water contamination
||Additive, sea water contamination
||Additive (antiwear, extreme pressure)
||Additive, internal coolant leak, brake fluid contamination
||Lubricant base stock, additive
It is also generally not possible to use ICP analysis to measure the additive depletion of an oil. Take, for example, the detergent additive found in an engine oil. This would reflect in the calcium (Ca) value. If the calcium level of both a new and a used oil was measured, the readings would be very similar, even though in the used oil the detergent has been depleted. The reason for this is that the amount of actual calcium in the oil has not changed. What has changed is the form, or compound, in which the calcium exists. Before being 'used', the calcium was present in a compound with detergent properties. After being used, the calcium is still present, but now in an inactive form. Therefore the ICP should not be used to measure additive depletion.
There are sometimes exceptions to this, albeit ones which cannot be relied upon. Most notable is the case of borate gear oils contaminated with water. In this case, the extreme pressure additive containing the boron precipitates out of solution and forms a sludge at the bottom of the gearcase. If this precipitate is not captured in the sample, the boron level will read much lower than normal, indicating the oil is not fit for further use due to extreme pressure additive depletion. The opposite, however, is still not necessarily true: if the boron level is correct, the oil may not necessarily still be fit for use.
Common wear situation as indicated by the ICP
||Si and Al present, usually in the ratio range of Si:Al between 2:1 and 10:1. Watch the increase in the trend. Often accompanied by associated wear when present over acceptable limits
||Al and Si in ratio Al:Si = 2:1. The Si originates from silicon carbide in the piston crown used to reduce the co-efficient of expansion. Seldom seen, as failure is usually rapid, and statistically there is little chance of getting a sample whilst occurring.
|High Fe (alone)
||As iron is the most used construction material, sources are often varied. Consider valve gear and oil pump wear. Rust formation also produces high Fe.
|High Si (alone)
||Silicon by itself comes from a few main sources - antifoaming agent additive, grease and silicone sealant. Usually seen in new/recently overhauled components. Usually can be ignored.
||Characterised by increased Fe (cylinder liner), Al (pistons) and Cr (rings) levels. The presence of Ni usually indicates camshaft/cam follower wear.
||Characterised by increased Fe (crankshaft) and Pb, Cu, Sn (white metal bearings and bronze bushes) levels. Often this wear is precipitated by reduced TBN or overcooling as bearings become subject to corrosion from combustion by-products (acids).
|Overheating (some cases)
||Increased additive levels (Mg, Ca, Zn, P & S) and viscosity. When light ends in the oil vaporise off, the oil level decreases. Topping up increases the additive concentrations, as the additives themselves do not evaporate. Oxidation often not evident, as topping up replenishes anti-oxidants and boosts the TBN. Often accompanied by Pb, Sn and Cu as bearing wear can result from this situation.
|Bronze Bush Wear
||Increased Cu and Sn levels. Cu:Sn ratio usually approximately 20:1.
|Bronze gear/thrust washer wear
||Increased Cu and Sn levels. Cu:Sn ratio usually approximately 20:1.
|Internal coolant leaks
||Increased Na, B, Cu, Si, Al and Fe. Not all elements may be present. Often accompanied by increased Pb, Cu and Sn as white-metal bearing wear often accompanies this. Water usually not evident, as it tends to boil off at normal operating temperatures.
||Increased levels of Fe, Cr and Ni levels, all components of race and roller materials. Increased Cu might result if brass/bronze cages are employed.
* Ashley Mayer is technical consultant for the Wearcheck Division of Set Point Technology. WearCheck Africa is the leading oil analysis company in Africa serving the earthmoving, mining, freight, passenger transport, rail, aircraft and marine industries, as well as the industrial sector. http://www.wearcheck.co.za
Next week: PQ and Viscosity testing