Specification-Writing for Petroleum Lubricants 270061
ALL large users of petroleum lubricants are endeavoring to reduce to printed form their individual ideas of what the lubricants they want should contain and what their physical and mechanical properties should be. The lubricants manufacturer finds, however, that anarchy prevails among the requirements and that the technique of writing the specifications is distinctly amateurish.
One method followed is to analyze a satisfactory lubricant and embody the results in the specifications, but the specifier does not know that the product is the best for his purpose and does not possess the facilities for accurate analysis and the ability to determine the pertinent from the irrelevant factors. Another method is to select from a number of analyses and specifications items that seem important and incorporate them in the writer's specification. The result calls for a non-existent hybrid that may be impossible to produce.
A third method, and the only proper one, is to make a comprehensive study of each lubrication problem and summarize the results in the specifications. This is the method followed by the larger manufacturers of lubricants and, to some extent, by the larger consumers.
Such studies show that narrow limits of viscosity, flash-point and other properties are not necessary in oils and that exact soap-percentage, oil-viscosities and consistencies are not essential in greases.
Two underlying principles in specification-writing are that they should (a) describe fully the product desired so that the manufacturer can supply it and (b) enable the purchaser to determine whether the desired product has been supplied.
Numerous examples are given by the author to show that many specification writers are lamentably ignorant of the characteristics that are and are not relevant, of the nature of materials and of the limitations of analysis and test methods. Incorporation in specifications of a statement that, whenever possible, tests are to be made by the latest methods approved by the American Society for Testing Materials will avoid much uncertainty. If limits are fixed for quantities that are not determined by these methods nor covered by standard analytical practice, the specification should give in workable detail the method to be used or should make reference to a publication containing a description of it.