Used oil analysis has been on the scene for decades. Many users probe their in-service fluid for evidence of deterioration to decide if the oil is still fit for service, or regularly check for the debris and wear particles that suggest machine parts are due for early overhaul.
Analysis of in-service greases, however, is another story.
Were all comfortable with used oil testing, said Bryan Johnson, an engineer at Arizona Public Service Co.s Palo Verde Nuclear Generating Station in Tonopah, Ariz. We look at oil for its condition, and also for evidence of machinery failure.
Grease, however, is far more difficult to assess. First, the lube points often are in places where not a lot of grease is available for sampling, Johnson observed in Philadelphia last month at the annual meeting of the Society of Tribologists & Lubrication Engineers. Samples taken from a port somewhere on the machinery may only be representative of the grease near the port – not the grease that is actually in contact with the moving parts. Inside a bearing, in reality theres very little grease providing lubrication, he said.
Then you have to ask, what do I see in the sample? So the sample is hard to get, or the sample size is too small, and finally I have to wonder, is it representative?
This doesnt mean grease sampling cant be done, but it needs to be organized, just like oil condition monitoring, if you want to know the condition of the grease and the condition of the part, urged Johnson.
Another hurdle is that there are virtually no standards for how to test in-service grease, and few experts in the field. The most common ASTM tests are intended for grease producers to use as quality controls during manufacturing, and require far more grease than an end user may be able to coax from a particular bearing or motor drive.
The cone penetration test for grease consistency, for example, requires 290 cubic centimeters of grease; even the modified 1/4-cup version of the test takes 38 cc to perform, Johnson noted. Only about 20 percent of the time has he been able to draw enough grease to do even the 1/4-cup cone penetration test. So Johnson for years has been using a stress rheometer, which only requires a 0.5 cc sample, to check grease consistency. As a starting point, its a good way to get data out of a very small volume of grease. ASTM is working on a rheometer test method that would be valid for checking new and used grease consistency, but its not approved yet.
Still, shrewd users are checking grease consistency using the rheometer, and examining the greases base oil quality and basic chemistry, using FTIR (Fourier-transform infrared spectrography) and RULER instruments to check grease oxidation and remaining useful life. They also count wear particles and identify debris found in used grease, through microscopic examination of the particles.
This is all valuable information. Johnson offered one case where used grease testing averted a catastrophic failure of critical motor bearings. The rheometer test showed significant stiffening of the grease. We also saw wear particles, a lot of iron particles. Palo Verdes laboratory determined that the grease had been exposed to massive amounts of radiation, which badly damaged its thickener system, causing it to lose strength and elasticity. Knowing this helped the nuclear station avert a costly breakdown.
Palo Verde also has been using the rheometer to provide insight into what may happen when greases mix, to understand compatibility issues that may arise when considering changing from one grease type to another. We mix various samples, and then track whether they become harder or softer versus the originals. In some cases the whole of the mixture was not as stiff as either part, so we could see that the structure of the grease was being affected by mixing with another type.
Proper grease sampling is tricky, agreed Rich Wurzbach of Maintenance Reliability Group in York, Pa., who also spoke at the STLE meeting. There is a well-established history of liquid oil analysis, and its easy to get samples. Grease analysis is performed in far fewer cases, though. First, theres the problem of obtaining consistent, representative samples, which is a major barrier to acceptance. Theres also less of a slate of tests you can do, he said. However, consistency, drop point, wear particle analysis, FTIR, RULER for oxidation, and rheometer tests are now becoming more commonly used for greases.
Wurzbach, whose company focuses on maintenance practices at power generation facilities, quipped that this field might be known as CSI: Grease. Like detectives on popular television shows, we come in after the fact to evaluate unusual conditions. The failed component is disassembled and the grease extracted for analysis, and maybe you can do root-cause analysis too.
Bearings, couplings, motors and cranes are all good candidates for used grease analysis – if you can get sufficient grease in meaningful quantities for testing, he added. In the mid-1990s Wurzbachs company analyzed more than 400 samples of grease from motors, gearboxes and limit switches. Overall, we saw poor correlation to what the mechanics found, Wurzbach recalled. But why? Was the sampling methodology inconsistent? The mechanics did their best, they grabbed samples from mating parts, he said, but these yielded little useful information.
This may be because those taking the grease samples have used flexible tubing, he added; the grease then is difficult to get back out of the tubing. It sticks, and that reduces the amount available for testing.
To help get larger amounts of grease out of equipment, Maintenance Reliability Group developed a grease collection device. This plastic, piston-and-cylinder device allows grease to be captured from places such as motor drains. It can also be used like a syringe, suctioning out slightly more than 5 grams in an average pull.
The company hopes to expand the sampling devices use to gear boxes and motor operated valves. It also has made cutaways of motors, adding a glass pane to view what precisely happens when grease is added. This may answer the longstanding puzzle of where exactly incoming grease goes – does it fill the space between bearings and moving parts or just push past the components?
Especially, Wurzbach warned, proper greasing procedures should be followed, but often are not. Properly, a two-part preventive maintenance drill should be followed: First, the drain plug should be removed, the grease added, and the plug left open for a while – a half-hour or longer – to allow the old grease to purge. Then the site should be revisited, so the grease can be topped off and the plug closed up.
Observation has shown that this rarely happens; usually the maintenance person comes by, shoots in fresh grease, closes the plug and considers the job done. This may mean that residual grease has not been properly cleaned out.
Despite the pitfalls, in-service grease holds valuable data for failure prediction, reliability assessment, and even trending and root-cause analysis, Wurzbach asserted. Consistency and streamlining the sample collection process can make grease analysis an economical method for reliability monitoring.
Casting for Answers Another view into grease forensics came from Patrice Mortreuil, senior scientist with Alcan Research Center in Voreppe, France. Alcan uses grease to coat its casting molds when it makes aluminum ingots.
The casting mold looks like a coffin, he pointed out, and the grease is spread all over the inside of this box before the molten aluminum is poured in. Water circulates around the outside of the mold, to quickly drop the metals temperature, and finally the bottom of the mold is lowered away from the sides – similar to a cake pan – revealing an 1,800-pound slab of raw aluminum.
Grease is used to coat the mold, because its cling properties mean it wont flow away before the hot melt is poured. The casting house also expects the grease to have some corrosion inhibiting ability, to be resistant to high temperatures and humidity, and to remain in storage without much oil bleeding out.
We need the grease to reduce friction between the ingot and the mold, and to prevent sticking. Its also good that it not burn too much, because the temperatures will be about 700 degrees C, Mortreuil told STLE. And we want the grease to leave no stains.
So why analyze the grease? Well, its fun to see whats in it, he grinned. But mainly its because the casting house engineers want to feel that everything during a cast is under control. Casting is hot and dangerous, and no one enjoys a bad result. The engineers want to know exactly the pour speed, temperature, cooling speed and so on. And since they dont know whats in it, if theres a problem the first thing they suspect is the grease. So were mainly called in for defect identification.
Mortreuil said Alcans grease lab work falls into the category of examination rather than in-depth analysis. We compare a given sample to a reference grease, that is, a sample that is known to be of acceptable quality. This doesnt require any heavy techniques.
The easiest thing in this detective work, he said, is to look for ash content. Infrared spectrography is easy, cheap and gives much information about the greases organic compounds. However, mineral-based additives are not detected, he said.
For more detailed examination, we have to analyze the soap and extract the mineral oil from the additives. The base oil must be carefully rinsed away with solvents, and the additives painstakingly dried for further analysis; this usually takes overnight. You must ensure the sample is fully evaporated, or you may see products not in the original profile of the grease, Mortreuil cautioned. Skill is needed to avoid such traps.
Along with FTIR, the Alcan researchers arsenal includes a solid-phase extractor, HPLC and light diffusion detector. This last takes about 20 minutes to run, and separates the additives according to their polar component. It can be another fingerprint of the additives. Using these devices, Mortreuil can pinpoint a component that is 1 percent of the grease – and also can spot an unknown component.
In the end, the goal is to examine the grease to uncover any drift in composition that may have led to a casting failure, he said. This means one more issue has to be tackled: The grease specifications have to be carefully written and rigorously followed, to avoid disagreement with the grease supplier. My experience with the cast house is that specifications are never read until there is a problem.
Alcans researchers do not screen new greases before they are put into service. If theres no problem, we dont exist, said Mortreuil. Were only called in when the engineers in the casting house are trying to eliminate grease as the possible source of a problem theyre seeing.