Lubricant Sampling Techniques

Introduction

The particles in an oil sample tend to settle out; if they are large and dense, they settle out rapidly. The particles of wear, corrosion, oil degradation, and contamination provide valuable diagnostic information about the condition of the oil as well as the condition of the wearing surfaces of the machine. Therefore, it is important that these particles, if present in the machine, are also present in the oil sample sent to the oil analysis lab. This tech note considers some of the factors important in taking a proper oil sample.

Valid oil samples must contain a representative selection of wear and contaminant particles. Likewise, valid oil analysis must be capable of analyzing particles of all sizes. Unfortunately, most oil analysis being done today, is blind to all but the very smallest particles.

Some of the factors considered herein are:

• Non-uniform distribution of particles
• Life of a particle
• How long to reach equilibrium after start-up
• When and where to sample
• Sampling from pipes
• Sampling from tanks
• Grease samples
• Sampling frequency
• Sample bottle kits
• Sampling accessories
• Training

Non-Uniform Distribution of Particles

Since particles exist as a separate phase in the oil, they are not evenly distributed in the system. All the freshly made wear particles will be present immediately after a wearing mechanism such as a roller bearing, gear, sliding surface, etc. The largest metal particles, which are of such critical interest to the analyst, are soon removed by settling in the slow moving sections, the sump, or they are filtered or otherwise separated. The very smallest particles tend to remain suspended and pass through all but the finest filters so they are generally distributed evenly throughout the oil piping system. Consequently, in order to capture a representative sample, the sampling location must be chosen carefully.

Life of a Particle

At equilibrium operating conditions, the rate of wear particle generation is constant. Under the same conditions, the particle separation and removal rates, although size dependent, are also constant, surprisingly, with or without a filter. Not surprisingly, the population of all but the smallest particles reaches an equilibrium level. Departure from that “normal” level is an indication of an abnormality in the system; perhaps the filter is bypassing or has failed, or a severe wear mode has begun. The smallest particles remain in the system, suspended and unaffected by the filter, so their population tends to increase during the life of the oil. Therefore, very fine particles are composed of the oldest particles, some of which may have been generated as larger particles but which are now reduced in size by various chemical and physical processes. The larger particles are relatively new and are better evidence of the current condition of the wearing surfaces.

It is the larger particles in an oil sample that represent the current wear condition; the small particles are primarily the residue of history.

How Long to Reach Particle Equilibrium After Oil Change

When a machine is started after an oil change, coarse particle concentration will increase, and the removal rate by settling and filtration also increases, until such time when the generation rate and the removal rate are in equilibrium. The time to reach equilibrium will vary machine-to-machine and even within a machine because of the many variables:

1. Very fine filters will cause equilibrium to be reached quickly; systems without filters reach equilibrium very slowly.
2. The gas turbine of a military aircraft will pump the oil system volume in seconds and reach equilibrium in minutes; the oil volume of a steam turbine may be pumped in many minutes and time to coarse particle equilibrium measured in hours or even days.
3. If the flow is turbulent throughout a system, the settling rate is reduced so it will take longer to reach a higher equilibrium population.

Where and when to sample

In addition to advice and precautions given elsewhere in this note, a representative oil sample should be taken at the same location using the same method each time; and should be taken during steady-state operation or soon after shutdown.

Sampling from pipes

1. The pipe ideally will contain oil soon after draining from the wearing part.
2. The flow rate should ideally be turbulent to keep particles entrained in the oil.
3. The oil should be hot; do not sample after cooling.
4. If pipe is large and flow laminar, take sample from near center of pipe to avoid lower wall where old debris may accumulate.
5. If the sample valve is installed with a “dead leg”, be sure to flush thoroughly before collecting sample.

Sampling from tanks

1. Take sample from near center and well clear of bottom or sides.
2. Sometimes a special spring steel sample stick can be useful in placing the suction tube just where you want it each time, thereby gaining repeatability. The narrow flat steel bar can be fitted with a stand off to prevent bottom sampling, as well as clasps to hold the tygon tubing.

Grease samples

Normally, new grease is introduced through the normal fitting to purge the old grease. A power greaser may be used. The effluent grease will be dark where most of the wear debris is concentrated. A sample of the dark grease, free of ambient dirt, should be taken and placed in a sample bottle. Care should be taken not to damage the bearing seals by over greasing. In the laboratory, a suitable dilutent is used to make what amounts to an oil sample, which can be processed like a normal sample.

However, the testing provides primarily qualitative results. Quantitative results are not as reliable as they are for oil samples because wear debris is not always evenly distributed in grease. Nevertheless, information about the wear mode underway in the bearing or surface can often be established.

Sampling Frequency

Machinery manufacturers will often suggest a sampling interval, but that should only be a rough guideline. The operator is the best judge of sampling intervals. He should ask himself some pertinent questions in arriving at a sampling interval, such as:

• What is the failure history?
• How costly is a failure? In repair cost? Lost production? Life and safety?
• Have operating conditions changed to put more stress on the machine?

In general, a quarterly or monthly sample interval is appropriate for most important industrial machinery. The answers to the above questions will help decide which regimen is more appropriate. In a new program, it usually makes sense to start with a monthly interval and then extend it as experience dictates.

Sample bottle kits

The use of clean bottles and caps, together with proper labels and convenient shipping containers, help to make an oil analysis program efficient and effective.

Sampling Accessories

Proper sampling valves, sampling pumps, tubing and other accessories will help make oil sampling easier and more repeatable; an important objective.

Training

A person experienced in oil sampling should walk through a “routine” with the trainee, advising him on each machine to be sampled. Follow up with documentation on the method of sampling and the necessary tools.