Computer-assisted analytical techniques in quality assurance of fuels and lubricants

Abstract

Fuels and lubricants derived from crude oils are very complex heterogeneous systems containing a mixture of numerous and analogous molecules. Their separation and structural characterization are tedious and quite difficult. Generally, average molecular structures are derived in a trial-and-error fashion from a large quantity of analytical data and used widely to represent these fractions. This paper describes a few of computer-assisted novel analytical techniques useful for quality assurance of fuels and lubricants: artificial neural networks (ANN); step-wise discriminant function analysis (SDFA); computer-assisted molecular structure construction (CAMSC); and, computer-assisted NMR spectroscopic analysis of organic mixtures.