62 M A Y 2 0 0 6 T R I B O L O G Y & L U B R I C A T I O N T E C H N O L O G Y

By Jack Poley

On Condition Monitoring

e have begun to explore thesignificance of oil analysis test

data, as well as the evaluation of thatdata toward practical maintenanceand equipment management deci-sions and action. Distilling this down,there are two specific processes: Rat-ing Data and Interpreting Data.

Rating dataæassessing ‘severity’This is primarily a mathematicaltreatment based on statistical limitsand trending. However, the contextfor such analysis is extremely impor-tant and includes such things as:

Application in which the compo-nent* is used• Application is arguably the most

important attribute in terms ofassessing a component’s data lev-els and trends. A simple examplewould be to contemplate a certaindiesel engine, “A,” operating in apiece of mobile equipment in aquarry operation versus that sameengine make/model operating as aconstant-speed generator in a rel-atively clean, enclosed environ-ment. Clearly the risk of ‘dirt’(abrasives) entry alone would sug-gest that the mobile applicationwould result in greater wear; butthere are additional aspects suchas intermittent engine revving orvarying load conditions that alsoaffects contamination and wearmetals levels/trends. (It is usuallycorrect to have separate boundarytables and trend limits when appli-cations are significantly different).

*For this discussion’s purpose, “component”is a single-lubricated compartment such as abearing or turbine or hydraulic system, not

the entire equipment, such as a bulldozer,which will have an engine, hydraulic system,transmission, gear drives, etc. At times mul-tiple mechanisms within the component arelubricated by the same oil, (e.g., transmis-sion-brakes). This would still be considered asingle component. Such components are fre-quently found in off-highway equipment.

Time or mileage on the lubricant(Aging the oil drain)• It is reasonable to expect that the

longer an oil is left in service, thegreater the wear metals levelsshould be. However, such a devel-opment will not usually be linear,nor should such an assumption bemade. There tends to be an equili-brating point dependent on oilconsumption, sump size relativeto the surfaces lubricated andpresent wear trauma experiencedby the component. Further thereare side conditions, such as filtra-tion effects or amounts of residuallube left in the compartment atdrain that may preclude a ‘perfect’progression. What is important isto be aware of these possibilitiesand compensate appropriately.

• Differences of less than 20% areprobably not significant in mostcases, but each situation should beassessed in its own context. Adefault time or mileage may haveto be imposed if the sampler failsto provide these data. Most timesthis forced assumption does notresult in a blown diagnosis, but itobviously poses a modicum of risk.

Time or mileage on the component(component age)• There is a general theory or expec-

tation that an initial ‘wear-in’ peri-

od is incurred,followed by a leveling-off period,finally culminating in an acceler-ated wear-out period. The wear-inperiod is the shortest; the levelperiod is the longest and thewear-out period is somewhere inbetween the other two. As withmileage data, some interpretationmust be considered in as much asthere is frequent deviation orexception to the theoretical pat-tern.

• Similar to the caveat for time/mileage on the lubricant, a num-ber of samples may be analyzedwithout mileage informationeither from broken hour meters orlack of diligence on the sampler’spart. In such instances one shouldagain consider making the forcedassumption that the component isin its ‘sweet’ life or level periodand proceed accordingly.

Careful: a conscientious evaluatormay notice a previously providedcomponent time/mileage figure anddiligently calculate or estimate theapparent time/mileage of the currentsample. What if the component hadbeen, in fact, overhauled prior to thecurrent sample? Then such anassumption could result in a poordiagnosis. Best policy is to demandthe information, then disclaim theinterpretation/evaluation to theextent of the lack of such information.Customers have minimum responsi-bilities in the oil analysis programchain of information and events. Adecent sampling procedure, includ-ing appropriate accompanying infor-mation, is part of that responsibility.

W

Breaking down the process of interpreting data

T R I B O L O G Y & L U B R I C A T I O N T E C H N O L O G Y M A Y 2 0 0 6 63

In some applications, such as off-highway, it could be reasonable torefer to the component by its equip-ment type (e.g., dozer) because thesame component, regardless of hav-ing been manufactured by other thanthe original equipment manufacturer(OEM), is used throughout that prod-uct line. Thus, one might successful-ly evaluate such components withoutevery knowing the actual componentmanufacturer. This is a shortcut,clearly, and could backfire if the OEMchanges vendors for its componentand fails to make this informationreadily available. The OEM is notparticularly concerned with individ-ual components so much as the mar-riage between all the components toproduce the best performing equip-ment it can under competitive mar-ket conditions.

There are also technical considera-tions within the analytical processthat sometimes get overlooked:

1. The quality and limitations ofthe sample taking and handling.• At the time sampling occurs. Even

with near-perfect sampling tech-nique, the final sample is not a100% accurate representation ofthe entire (effective) sump. As anobvious example, recall we notedthat water is almost never evenlydistributed in an oil system.

• When an aliquot is removed fromthe sample container for testing atthe laboratory, the technicianmight carelessly fail to shake thesample sufficiently to attempt toget the best homogeneity possi-ble, resulting in slightly skewedanswers.Testing instrumentation itself has

inherent limits, also discussed earlierin this series. The point to keep inmind is that a given number has atesting tolerance. It is important forthe evaluator to know what that toler-ance is for each type of instrumentinvolved.

Bottom line of the above discussion:

1. Test result numbers must beproperly and completely qualifiedrather than accepted at face value.

2. Once this process is accomplishedwith high confidence, one is ready toactually conduct an evaluation of com-ponent and lubricant condition. <<

Jack Poley is managing partner of Condition

Monitoring International, LLC, Miami, con-

sultants in fluid analysis. You can reach him at

poleyj@bellsouth.net. For more infor-

mation about CMI, visit www.cmi

global.biz.

www.nceed.com

“I better get a wagon...”

Castor oil and its derivatives have been a staple in the NCeed productmix from the beginning. We're constantly growing our business so youcan grow yours. Here’s a sample of our offerings:

• sebacic acid • esters of sebacic acid • tolytriazole • benzotriazole• maleate esters • denatonium benzoate

Visit www.nceed.com or e-mail Noreen Cherry atchem616@nceed.com with any question you mayhave. And when you order, don’t forget the wagon!