In their paper, "The Timken Lubricant Test - 1932 to 1972", presented at the 40th annual meeting of the National
Lubricating Grease Institute, D.V. Culp and J.E. Leiser of the Timken Company reviewed the application and
significance of the OK Load Test with EP lubricants.

Two items discussed in the paper are of particular interest. The hypothesis that gear oils which have higher Timken OK values will provide better scoring protection for gears and bearings was developed in the early 1930's.
Recent developments in the chemical aspects of EP additives indicate that this is no longer true. In fact, it now appears that many of the more modern type chemical additives, particularly those successfully being used in the MIL-L-2105B type oils, have shown improved performance in preventing scoring in high temperature gear and bearing applications although they show a very decreased Timken load as compared to the leaded type or SCL type gear oils.

The authors also point out the lack of a direct across-the-board correlation between high Timken EP values and increased performance characteristics. It also appears that there is little correlation between Timken test results
and other EP tests results, such as the Falex, the four-Ball tester and the more recently used FZG gear type tester.

The significance of the observations made by the Timken Company concerning the limitations of the OK Load test in predicting actual performance of EP oils should be recognized when making a product recommendation. A proven record of satisfactory performance in similar applications is of greater importance in recommending a product than merely selecting one whose Timken OK Load Test rating meets the specification. Documentation of actual successful performance of EP oils is still the most valid way to support the claim that an oil will provide the
required protection of gears and bearings.

According to the Society of Tribologists and Lubrication Engineers, the Timken test is "questionable for use in evaluating levels of EP".

you are right this test has very little to do with oil performance and should never have been published, we also entered the metric system 40 years ago! The primary role of lubrication is to keep parts apart and maintain oil integrity, not see what happens when metal to metal takes place, a symptom of some other problem...

Testing to reveal actual oil performance would have been beneficial, this may involve: testing oil chemical and physical properties , inspection (microscopic particle examination) of oil debris for wear modes/indicators and testing oil cleanliness to established ISO standards. These are the true tests of oil/machine performance.

For our application the best thing you can do is to use the correct viscosity (quality stable oil, may or may not be synth) and change out the oil regularly. We do not change oil out because it has 'broken down', we change it out to remove normal wear debris, the oil chemical and physical properties at this point should still be good.

so long life oils may hold there spec (synth for eg.) but you are still pumping all that wear debris through your bearings!!

BTW oils that 'polish' or chemically etch components when put through studies of MPE and ISO cleanliness can actually initially create a lot of wear debris before the ISO codes settle down.. go figure.