Correlation of portable wear metal analyzer (PWMA-2) and Baird A/E35U-3 oil analysis data

Abstract

Correlation of the Perkin-Elmer Corporation portable wear metal analyzer, version two (PWMA-2) and the Baird Corporation A/E35U-3 spectrometric oil analysis data is accomplished by curve fitting of data from aircraft, bearing test rigs, and organometallic oil samples analyzed on both instruments. Each of nine wear metal elements (iron, silver, aluminum, chromium, copper, magnesium, nickel, silicon and titanium) was separately fitted to linear, logarithmic, power law and low order polynomial curves. Power law curve fits were best for most of the elements, especially in the important low concentration (0–5 ppm) range, where U.S. Air Force gas turbine aircraft engines typically operate. The power law curve was applied to all the elements to construct correlation algorithms, with a general uncertainty in correlation values of about 0.5–1 ppm or 10%, whichever was the greater. The coefficients of determination for the power law curve fits ranged from 0.935 for aluminum to 0.993 for iron. The fit for silicon was poor for any type of curve (the best coefficient of determination was 0.564 for the power law curve fit) and therefore correlation of silicon values is not recommended.